TT TT       ^     ^T A TT
Hazardous Wastes
and Hazardous
Material
     Site Remediation • On-Site Treatment • Risk Assessment
  Contaminated Groundwater Control • Permitting • Monitoring • Incineration
    • Underground Leak Detection • Fixation • Cost/Economics

-------
             PROCEEDINGS OF THE
          NATIONAL CONFERENCE ON
            HAZARDOUS
            WASTES  AND
            HAZARDOUS
             MATERIALS
     Site Remediation • On-Site Treatment • Risk Assessment
  • Contaminated Groundwater Control • Permitting • Monitoring •
Incineration • Underground Leak Detection • Fixation • Cost/Economics
         March 4-6, 1986 • Atlanta, Georgia
                    AFFILIATES
              U.S. Environmental Protection Agency
           Hazardous Materials Control Research Institute
                 Department of Defense
           Agency for Toxic Substances and Disease Registry
                Portland Cement Association
            National Environmental Health Association
                 National Lime Association

-------
Printed in the United States of America
Library of Congress Catalog No. 86-80269
Copyright © 1986
Hazardous Materials Control Research Institute
9300 Columbia Boulevard
Silver Spring, Maryland 20910

-------
                       Acknowledgement
  The National Conference and Exhibition on Hazardous Wastes and Hazardous
Materials:  Site  Remediation  •  On-Site  Treatment   • Risk  Assessment  •
Contaminated Groundwater Control • Permitting • Monitoring • Incineration •
Underground Leak Detection • Fixation • Cost/Economics required dedication and
talent from many individuals and commitment from a number of organizations. We
gratefully express our thanks and appreciation to the following conference affiliates
for all their contributions to a successful conference.


U.S. Environmental Protection Agency
Hazardous Materials Control Research Institute
Department of Defense
Agency for Toxic Substances and  Disease Registry
Portland Cement Association
National Environmental Health Association
National Lime Association


  We also wish to express our gratitude to all of these knowledgeable individuals for
their advice and guidance in planning and producing a highly effective and infor-
mative program:


Gary F. Bennett, Ph.D., The University of Toledo
Hal Bernard, Hazardous Materials Control Research Institute
Ken Gutschick, National Lime Association
Robert Knox, U.S. Environmental Protection Agency
Charles Mashni, U.S. Environmental Protection Agency
Thomas Potter, National Lime Association
Jerry Steinberg, Ph.D., Water & Air Research,  Inc.
Andres Talts, Department of Defense
Ralph Touch, Agency for Toxic Substances and Disease Registry


  Producing a document of the magnitude of  this proceedings requires a highly
skilled team, much cooperation and communication, and a tremendous amount of
effort by all involved. We are fortunate to have such a team and would like to
convey our special thanks to Dr. Gary Bennett, Professor of Biochemical Engineer-
ing, The University of Toledo, and Hal Bernard, Executive Director,  HMCRI, for
the excellent  editing; to the  typesetters, proofreaders, and graphic artists who
completed a tremendous amount of work in an incredibly short period of time; and
to the staff of HMCRI for coordinating the myriad details and activities of this
conference.

-------
                                                      Preface
       The RCRA Amendments  of  1984 (the new RCRA) have
       radically changed hazardous waste management through-
       out the country and considerably improved the control of
 hazardous wastes. Whereas it was estimated that the old RCRA
 provisions would cost the regulated industry between $1  billion
 and $3 billion per year, the new  RCRA amendments, when com-
 pletely implemented and  fully in effect, may cost as high  as $20
 billion per year.
   The new RCRA includes 58 congressionally mandated  statu-
 tory deadlines that go into effect by 1987 or early 1988. Congress
 wrote, into the statute, provisions that are detailed and directive.
 The bill  does not wait for EPA to write regulations or issue
 guidance; provisions go into effect automatically on prescribed
 dates.

 Land Application  Management
   The law establishes a hierarchy of management practices. There
 is a very strong presumption against the disposal of hazardous
 wastes on the land and a very strong preference for treatment and
 destruction of hazardous wastes.

 Site Remediation
   The bill provides EPA with "corrective  action authority" at
 existing hazardous waste  facilities which are very similar  to the
 enforcement authorities under Superfund. The bill also closes the
 gap in the existing  program by  expanding the regulated com-
 munity.

 Leaking Underground Tanks
   A major new subtitle of the  legislation  dealing with leaking
 underground storage tanks went into effect on May 8, 1985. This
 section probably affects two to  five million underground tanks
 across the U.S. including tanks storing petroleum products and
 gasoline as well as  very  complex tank  storage systems storing
 hazardous chemicals.
   As of May 8, 1985, there is a ban on new tanks that are not de-
 signed to prevent  releases due to corrosion  or structural failure.
 An estimated 100,000  new tanks must meet these requirements.
 By May  1986, a  nationwide registration program will require
 state notification of the age, size, type and location of the tank, as
 well as its uses.

Burning Hazardous Wastes
  The provision concerning burning hazardous wastes and the
blending of hazardous wastes into fuel requires  notification of
the fuel user that they are receiving hazardous waste. Some can-
not be burned in residential and commercial boilers.

RCRA Permit Program
  The  bill makes permitting of a hazardous waste facility far
more difficult than it has been. Permit applicants must now sub-
mit exposure information on the potential for public exposure to
hazardous substances from landfills and surface impoundments.
This exposure information will be used to write new permit con-
ditions. EPA has the authority to write any permit condition
necessary to protect human health in the environment, inde-
pendent of whether  or not  regulations  are in place  for that
purpose.
  Congress has also set some very strict deadlines for the issuance
of permits.  By November  1985,  all applications for land disposal
facility permits must have been  submitted to EPA or an author-
ized state.  Final determination on these  land  disposal  permit
applications must be made by November 1988.
  By November 1986, all incinerator applications  must be sub-
mitted  and EPA must make final decision on these by November
1989.  All remaining applications have to be in by November
1988,  and EPA must make a decision on these by November
1992.

Small Quantity Generators
  The first small quantity generator (SQG) requirement went into
effect in August 1985. Any SQG that ships wastes offsite must do
so by using the Uniform  National Manifest Form. Effective in
April 1986, SQGs must send their wastes to fully regulated haz-
ardous waste facilities.  These requirements will have a  tremen-
dous impact on small generators; costs for their waste manage-
ment will be considerably  greater.

Other RCRA Amendments
  There are many other provisions of the new RCRA. There is a
set of provisions that deal with non-hazardous  solid waste and
small quantity generator wastes disposed of in landfills and sur-
face impoundments. The  bill also provides for  federal enforce-
ment of the requirements for lagoons. There are  a series of waste
minimization requirements for generators to certify  that they are
doing everything economically feasible to reduce the amount of
wastes  generated.  There are requirements for listing additional
hazardous wastes. Delisting of wastes is much more difficult and
complex under the new bill. New tests for ascertaining toxicity are
included in the new RCRA.

-------
                                                 CONTENTS
                    MONITORING
Equipment for Data Collection at Hazardous Waste
Sites—An Overview for Environmental Management
Professionals	1
  James P. Mack & Thomas J. Morahan
A Statistician's View of Groundwater Monitoring	8
  Douglas E.  Splitstone
Data Evaluation in a Groundwater Study of Waste
Management Practices in the Phosphate Processing
Industry	13
  Edward W. Mullin, Jr., Jack S. Greber & William
  E. Thompson
Plant Cuticular and Dendrochronological Features
as Indicators of Pollution	17
  O.K. Sharma, Ph.D.
Detailed Stratigraphic and Structural Control: The Keys
to Complete and Successful Geophysical Surveys of
Hazardous Waste Sites	19
  H. Dan Harmon, Jr., P.O.
Detection and Measurement of Groundwater
Contamination by Soil-Gas Analysis	22  /
  H.B. Kerfoot, J.A.  Kohout &  E.N. Amick
Environmental Appraisals and Audits: A Case Study
of Their Application	27
  Anthony R. Morrell

             DETECTION OF RELEASES
Chlorinated Organics and Hydrochloride Emissions
Sampling from a Municipal Solid Waste Incinerator	31
  Thomas A.  Driscoll, James P.  Barta, Henry J. Krauss,
  David H. Carmichael & J. Maxine Jenks
CARE—Modeling Hazardous Airborne Releases	34
  M. Gary Verholek
The Use of PC Spreadsheet-Based Graphics to Interpret
Contamination at CERCLA/RCRA Sites	39
  George A. Furst, Ph.D.


   CONTAMINATED GROUNDWATER CONTROL
Emergency Response to Toxic Fumes and Contaminated
Groundwater in Karst Topography: A Case Study	44
  P. Clyde Johnston, Mark J. Rigatti & Fred B. Stroud
Use of Low Flow Interdiction Wells to Control
Hydrocarbon Plumes in Groundwater	49  >
  John H. Sammons, Ph.D. & John M. Armstrong, Ph.D.
Computer Groundwater Restoration Simulation at a
Contaminated Well Field	58
  Shih-Huang Chieh, Ph.D. & Jeffrey E. Brandow, P.E.
Enhancement of Site Assessments by Groundwater
Modelling  	64
  Joseph R. Kolmer, P.E.  & John B. Robertson, P.G.
Alternative Treatment Techniques for Removal of Trace
Concentrations of Volatile Organics in Groundwater	69
  Mark E. Wagner & Brian V. Moran
        CONTAMINATED SOIL TREATMENT
Cost-Effective Soil Sampling Strategies to Determine
Amount of Soils Requiring Remediation  	76
  Gregory J. Gensheimer, Ph.D., William A. Tucker,
  Ph.D. & Steven A. Denahan
Land Treatment of Wood Preserving Wastes	80
  John R. Ryan & John Smith
Method for Determining Acceptable Levels of
Residual Soil Contamination 	87
  William A. Tucker & Carolyn Poppell
Objective Quantification of Sampling Adequacy and
Soil Contaminant Levels Around Point Sources
Using Geostatistics	92
  Jeffrey C. Myers
                ON-SITE TREATMENT
Innovative Application of Chemical Engineering
Technologies for Hazardous Waste Treatment	98
  Robert D. Allan & Michael L. Foster
Field Studies of In Situ Extraction and Soil-Based
Microbial Treatment of an Industrial Sludge Lagoon	102
  David S. Kosson, Erik A. Dienemann & Robert
  C. Ahlert, Ph.D., P.E.
Cleanup of Contaminated Soils and Groundwater
Using Biological Techniques	110
  Paul E. Flathman & Jason A. Caplan, Ph.D.
Physical/Chemical Removal of Organic Micropollutants
from RO Concentrated Contaminated Groundwater	120
  L. Simovic, J.P. Jones & I.C. McClymont
State-of-the-Art Technologies of Removal, Isolation
and Alteration of Organic Contaminants Underground	124
  Walter W. Loo & George N. Butter

-------
 Assessment of Volatile Organic Emissions from a
 Petroleum Refinery Land Treatment Site	127
   Robert G. Wet herald, Ph.D., Bart M. Eklund,
   Benjamin L. Blaney, Ph.D. & Susan A. Thorneloe
 Technology for Remediation of Groundwater
 Contamination	133
   David  V. Nakles, Ph.D. & James  E. Bratina

       TREATMENT OF HAZARDOUS WASTES
 Anaerobic Biological Treatment of Sanitary
 Landfill Leachate	136
   A.K. Mureebe, P.E., D.A.  Busch & P.T.  Chen, Ph.D.
 On-Site Versus Off-Site Treatment of Contaminated
 Groundwater—An Evaluation of Technical Feasibility
 and Costs	143
   Kent L. Bainbridge & Daniel  W. Rothman, P.E.
 Optimization of Free Liquid Removal Alternatives in
 the Closure of Hazardous Waste Surface Impoundments	148
   David K. Stevens
 Assessment of Chemical Treatment Technologies and
 Their Restrictive Waste Characteristics	154
   Hamid Rasiegar, Ph.D.. James Lit, Ph.D. &
   Chris Conroy
 Treatment Technologies for Hazardous Materials	157
   Jeffrey M. Thomas & Phillip T. Jarboe
 Update on the PACT Process	163
   Harry W. Heath, Jr.
 Treatment of PCB-Contaminated Soil in a
 Circulating Bed Combustor 	171
   D.D. Jensen, Ph.D. & D.T. Young
 Treatment of Hazardous Waste Leachate	175
   Judy L. McArdle, Michael M. Arozarena, William
   E. Gallagher, P.E. & Edward J. Opatken

        BARRIERS & WASTE SOLIDIFICATION
 Field Testing of a New Hazardous Waste
 Stabilization Process	180
   Richard H. Reifsnyder
 Innovative Techniques for the Evaluation of
 Solidified Hazardous Waste Systems	186
   Harvill C. Eaton, Marty E. Tittlebaum & Frank K.
    Cartledge
 Site Characteristics and the Structural Integrity
 of Dikes for Surface Impoundments	
   Jey K. Jeyapalan,  Ph.D., P.E. & Ernest R. Hanna
 Use of X-Ray Radiographic Methods in the
' Study of Clay Liners 	
   Philip  G. Malone, Ph.D., James H. May,  Kirk W.
   Brown, Ph.D. & James C. Thomas
 Closure Design and Construction of Hazardous
 Wastes Landfills Using Clay Sealants	
   John F. O'Brien, P.E.,  Lonnie E. Reese & Ian
   Kinnear, P.E.
                                                     190
                                                     198
                                                     202
Soil Liners for Hazardous Waste Disposal Facilities	206
  D.C. Anderson
Slurry Wall  Economical  in Dewatering of
Sydney Mine Disposal Site 	210
  Bruce J. Haas, Mark R. Nielsen & Norman N. Hatch
Utility of Soil Barrier Permeability Data	216
  Walter E. Grube, Jr.
                                                             Quality Assurance and Quality Control Procedures for
                                                             Installation of Flexible Membrane Liners	221
                                                               James R. Woods,  P.E. & Salvatore  V. Arlotta, Jr., P.E.
                                                             Mechanisms for the Fixation of Heavy Metals in
                                                             Solidified Wastes Using Soluble Silicates	224
                                                               Ella L. Davis, James S. Falcone, Scott D. Boyce &
                                                               Paul H. Krumrine
                                                                                                                  .229
                    INCINERATION

Illinois Plan for On-Site Incineration of
Hazardous Waste	
  James F.  Frank & Robert Kuykendall
An Introduction to EPA's New Trial Burn Data Book	233
  M.P. Esposito & N.J. Kulujian
Low Temperature Thermal Stripping of Volatile
Compounds	2-**
  John W.  Noland, P.E..  Nancy P. McDevitt & Donna
  L. Kolluniak
Comparisons Between Fluidized Bed and Rotary Kiln
Incinerators for Decontamination of PCB Soils/
Sediments at CERCLA Sites	242
  Henry Munoz, Frank L.  Cross, Jr., P.E. & Joseph
  L. Tessitore, P.E.

        UNDERGROUND LEAKING TANKS
New Requirements for Underground Storage Tanks	246
  A nna O.  Buonocore, P. E.,  Gerald F  Kotas &
  Kevin G.  Garrahan, P.E.
Cost-Effectiveness Evaluation of Leak Detection and
Monitoring  Technologies for Leaking Underground
Storage Tanks	251
  James Lu, Ph.D.. P.E. <$ Wayne Barcikowski
Underground Storage System  Assessment, Testing
and Remediation 	269
  Scott J. Adamowski, Angela J. Caracciolo, III
  & G. David Knowles,  P.E.
Case Study of Product Detection in Groundwater	273
  Pratap N. Singh, Ph.D..  P.E.
Underground Storage Tanks—Leak Prevention,
Leak Detection, and Design 	278
  Jey A. Jeyapalan. Ph.D., P.E. & James B.
  Hutchison. P.E.

                 SITE MANAGEMENT
Performance Evaluation of Commercial Hazardous
Waste Treatment Facility Operations	292
  Ronald J. Turner & Joan V. Boegel
An Examination of Siting Problems for
Off-Site TSDFs	298
   Douglas  B. Taylor, P.E.
Spatial Data Research for Hazardous Waste Sites	301
   Timothy  W. Foresman & Lynn K. Fensiermaker
Sensitivity Analysis of Remedial Action Alternatives
for Hazardous Waste Sites	305
  Elio F. Arniella, P.E. & E. Lawrence Adams, Jr., P.E.

     RCRA SITE REMEDIATION & EXPANSION
Design of a Lateral and Vertical Expansion at an
Existing Interim Status Landfill	310
  Rodney T. Bloese & Thomas G.  Ryan, P.E.

-------
                                                    .314
                                                    .318
                                                    .327
                                                    .331


                                                    .336



                                                    .342

                                                    .346
Cost Estimating for RCRA/CERCLA Remedial
Actions 	
  Michael R. Morrison & Gregory P. Peterson, P.E.
A Modular Computerized Cost Model for Remedial
Technologies at Superfund Sites	
  William Kemner, John Abraham, P.E., Jack
  Creber & Jay Palmisano

                 RISK ASSESSMENT
Using Cost/Risk Analysis in Waste Planning:
The New England Pilot Project	
  Debora C. Martin
Assessment of Potential Public Health Impacts
Associated with Predicted Emissions of
Polychlorinated Dibenzo-Dioxins and
Polychlorinated Dibenzo-Furans from a
Resource Recovery Facility	
  David Lipsky, Ph.D.
Improving the Risk Relevance of Systems for
Assessing the Relative Hazard of Contamianted Sites	
  Ellen D. Smith, Lawrence W. Barnthouse, Ph.D.,
  Glenn W. Suter II, Ph.D., James E.  Breck, Ph.D.,
  Troyce D. Jones & Dee Ann Sanders, Ph.D.
Soil Cleanup Criteria for Selected Petroleum Products
  Sofia K. Stokman  & Richard Dime, Ph.D.
Risk Management: Personnel, Equipment and Indemnity
  Denny M. Dobbs & Arlene B. Selber

         RCRA AMENDMENT EXPERIENCE
Innovative Approach to Site Remediation Involving
an RCRA Part B. Permit	
  Michael J. Conzett, P.E. & Michael E. Harris, P.E.
Overview of the Proposed Natural Resource
Damage Assessment Regulations	
  Richard J. Aiken,  Willie R. Taylor, Ph.D.
Regulatory Impact Analysis for the Toxicity
Characteristic	
  John L. Warren
Preliminary Assessments and Site Investigations
Under the Corrective Action Authorities of RCRA:
Analysis of Early Experiences	
  John W. Butler &  Robert D. Volkmar
Assessment of the Application of RCRA Part 264
Standards  to CERCLA Site Remediation	
  Rebecca N. Fricke, P.E.
     STATE, REGIONAL & LOCAL PROGRAMS
Small Quantity Generators: The Maryland
Approach to Regulation and Assistance	375
  Alvin L. Bowles, P.E.
Siting Efforts in Southern California	378
  Kieran  D. Bergin
The California Site Mitigation Decision Tree	380
  Paul W. Hadley & William Quan
                                                    .348
                                                    .353
                                                    .360
                                                    .364
                                                    .369
Hazardous Waste Management: The Role of the
Local Health Department	384
  Michael J. Pompili & Philip G. Brown

            PUBLIC COMMUNICATIONS
U.S. EPA's Initiatives for Expanded Public
Involvement in the RCRA Permitting Program	386
  Vanessa Musgrave & Edwin Berk
Cleanup in the Sunshine: Florida DOT's Public
Information Program at the Fairbanks Site	390
  Robert C. Classen & Charles C. Aller
Public Involvement in the RCRA Permitting Process—
A Facility Perspective 	394
  Gordon Kenna
Public Participation in Siting Hazardous Waste
Management Facilities in Alaska	397
  Sharon O. Hillman
Developing  a Comprehensive Public Affairs Policy
for New and Existing Industrial Operations	401
  Ann E. Burke
Behind the "F"  in Public Education for Hazardous
Waste Management: A Case for Special "Tutoring"
—Plus Some Tips for Better Grades 	404
  Howard A. Coffin, Patricia Hunt & L.T. Schaper, P.E.
                REUSE & RECOVERY
Recycling of Dust from Electric Arc Furnaces—
An Experimental Evaluation 	409
  E. Radha Krishnan, P.E., William F. Kemner &
  Copal Annamraju, P.E.
Trends in Used Oil Composition and Management	419
  Jacob E. Beachey & William L. Bider
Used Solvent Elimination Program	424
  Renato G. Decal, P.E.
        WASTE MINIMIZATION PROGRAMS
Waste Reduction Audit Procedure—A Methodology
for Identification, Assessment and Screening of
Waste Minimization Options	
  Carl H. Fromm, P.E. & Michael S. Callahan, P.E.
Minnesota Technical Assistance Program: Waste
Reduction Assistance for Small Quantity Generators
  Cindy A. McComas & Donna Peterson
Waste Minimization at Air Force GOCO Facilities	
  Douglas L. Hazelwood, Brian J. Burgher, P.E. &
  Charles Alford
Defense Environmental  Leadership Project  Study
of Industrial Processes to Reduce Hazardous Waste	
  Thomas E. Higgins, Ph.D., P.E. & Drew P. Desher
                                                             Exhibitors' List	450
.427


.436

.440



.445

-------
                             Equipment  for  Data  Collection  at
                      Hazardous Waste  Sites—An Overview for
                      Environmental Management Professionals
                                                  James P. Mack
                                               Thomas J.  Morahan
                                          Fred C.  Hart Associates, Inc.
                                              New York, New York
INTRODUCTION
  Industry and government are in the process of cleaning up
America. Industry is becoming increasingly concerned with ques-
tions of environmental liability. Regulatory agencies are actively
protecting public  health and the environment. Future liability
can be minimized by the use  of environmental audits. Present
liabilities, however, must be thoroughly understood in order to
develop cost-effective liability  management plans that meet the
goals and objectives set for environmental cleanup.
  This understanding generally is gained through environmental
investigations at individual sites  or facilities. CERCLA,  RCRA
and other statutes require regulatory compliance. In some cases,
the buyer or  seller of a property may run into unforeseen en-
vironmental liabilities. Some states have recently passed legisla-
tion  that requires  environmental conditions be defined and
problems investigated prior to real estate transactions. In cases of
corporate mergers or other processes where ownership of a facil-
ity changes, new owners or sellers may be held accountable for
the environmental liability at  a facility. While the  government
agencies  ensure that cleanups  are effective,  industries which  fi-
nance the cleanups can encounter huge expenditures. Business-
men and legal professionals handling environmental affairs stay
close to technical events, carefully monitoring costs and cleanup
strategies. These environmental  management professionals are
sometimes unfamiliar with techniques for data collection and the
creative use of equipment used for data collection activities. This
paper is directed toward business, legal and environmental man-
agement  professionals seeking knowledge of various types  of
equipment available for environmental data collection. An under-
standing  of the use of this equipment will help in the develop-
ment of cleanup strategies for a proper, complete and cost-effec-
tive cleanup.
  While  each site investigation is unique and requires a certain
degree of site-specific modification, there is a logical sequence of
events that is followed  in order to collect the desired informa-
tion. While the purpose of this  paper is to discuss the field equip-
ment available to perform the investigation, it is important  to
understand the sequence of events in order to define the situa-
tions in which the equipment should be used.
  The  first phase of the  investigation should be  devoted  to
collecting enough background data on site conditions as possible.
While there may not be much  data available  on actual site con-
ditions, there is usually a certain amount of regional information
available. Data can be obtained from local health departments,
state and federal environmental and geological agencies and uni-
versities.  This information  is important in defining the general
environmental setting of the site.
  The second phase of the investigation should be devoted to
understanding the source of the potential contamination. At this
point, certain indirect techniques such as surface  geophysical
instruments or portable analytical instruments  are very useful.
They can be used to define the possible horizontal extent of the
source, detect buried objects and initially determine the general
types of contaminants. Soil borings and soil sampling techniques
also are used to collect actual waste or contaminated  soils for
chemical analysis to further define the characteristics of the con-
taminant.
  Once it has been determined that there is a contaminant source,
the question arises as to whether the contaminants are migrat-
ing. This usually requires more extensive sampling of surface
water, sediment, soils and groundwater. Normally, groundwater
is an important issue and becomes the focus of further investiga-
tions. Wells usually are installed adjacent to the source to deter-
mine  the subsurface  conditions as well as to detect any immed-
iate migration. Proper soil sampling  and well drilling techniques
are essential to the adequate completion of this task. Downhole
investigative  equipment becomes useful in defining subsurface
geology. Water levels are measured in wells to define flow direc-
tions. As the scope of the potential problem develops, additional
wells may be required at further distances to define long-term mi-
gration pathways.
  The final component is the data evaluation and risk analysis.
This phase is as important as the field effort. The data must be
properly organized and the actual risks correctly established in
order to arrive at a cost-effective remedial solution.

SURFACE GEOPHYSICAL TECHNIQUES
  Surface Geophysical Techniques have been used for  years to
give indications  of subsurface rock  and soil conditions. These
techniques vary greatly in theory and technology, resulting in very
specific conditions where their use is warranted. In recent years,
these techniques have been applied successfully to investigations
at hazardous waste sites.
  Surface geophysical  surveys are performed as indirect tech-
niques to provide information to investigators in advance of
direct sursurface investigatory techniques such as test pitting, test
borings or well drilling activities. At hazardous waste sites, these
techniques generally are used in three different ways. First, some
methods can be  employed to provide information about subsur-
face conditions  such as depth to bedrock, depth to the water
table or thickness of certain soil units. Next, some methods can
be  employed  to examine the migration  of contamination in
groundwater. Finally, some techniques can be utilized to search
for buried metal, specifically containerized wastes such as drums
or tanks which could act as ongoing sources of contamination.
                                                                                                     MONITORING    1

-------
Magnetometry
   Magnetrometry has been used successfully in the past to iden-
tify trends in regional geology. More recently, it has been ap-
plied to search for buried metals at hazardous waste sites. The
proton precession magnetometer measures the total intensity of
the earth's magnetic field at one point. The earth's magnetic field,
which is fairly constant over a small site area, is affected by fer-
rous metals.  Ferrous metal objects create their  own magnetic
fields which change the intensity of the  magnetic field over a
site. Using the magnetometer, these ferrous metal objects can be
located.
   Several key steps in a magnetometry survey must be observed
to obtain quality data. First, all measurements must be referenced
to a grid system. Second, a point free of magnetic inferences
must be located, known as a reference point. The magnetic  field
of the  earth  changes  throughout the day. These changes arc
known as diurnal changes. A full record of diurnal changes must
be kept for corrections of the  intensity measurements taken at
grid points during the day. Often it is necessary to have a  base
station magnetometer dedicated to this task. These measurements
also can guard against the collection of poor quality data such as
that collected during magnetic storms.
   Once correct data are plotted, it is sometimes possible to model
the depth, mass, shape and angle of orientation of an object.
Theoretically, that  degree of data interpretation is possible  only
under ideal conditions.
   Variations on  this technique are possible. Gradiometers can be
used to measure the point changes in a field, instruments can be
hand held, sensors  can be placed on staffs and  instruments may
be towed along traverse lines for continuous readings.
   Limitations are critical to this type of survey. Magnetic induc-
tion caused by alternating current in power lines will affect read-
ings. In this case, or in  the case  of extensive surface metal inter-
ference, surveys should not be performed. Additionally, it is pref-
erable to have a map of shallow buried surface metal from a metal
detection survey to  aid in data interpretation. With these precau-
tions in mind, magnetometry can be a useful tool  to indicate the
presence of buried ferromagnetic objects.

Metal Detention
   Metal detention  has been used for years to locate pipes and
other buried metal  objects.  It relies on the effects of a conduc-
tive metal object or a radio signal which it broadcasts and re-
ceives.  The effective depth of penetration is only a few feet, but
it can be used very  accurately to locate shallow buried metal tar-
gets.

Electromagnetic Conductivity
   Electromagnetic conductance, generally  referred to as "EM,"
also responds to  the conductive properties of materials.  EM, also
called terrain conductivity, usually is used to indicate the presence
of a conductive contaminant plume.
   EM does not require ground contact. The EM transmitter coil
radiates an electromagnetic field which induces eddy currents in
the earth. Each of these eddy current loops, in turn, generates a
secondary electromagnetic field  which is intercepted by the re-
ceiver coil. An output voltage fis produced which is linearly re-
lated to subsurface conductivity.
   Continuous survey  instruments are available with  transmitter
and receiver fixed in the same instrument, but their use is limited
to shallow depths. Instruments are available with separate trans-
mitter and receiver coils for deeper surveys. Initially, several EM
soundings are made by varying the intercoil spacing at each loca-
tion.  After  a determination of  background  conductivity  and
appropriate  intercoil  spacing, EM profiles can  be completed
along previously surveyed transverse lines.
Electrical Resistivity
  Electrical Resistivity (ER) allows a measurement of  subsur-
face resistivities in soil, rock and groundwater. Application of the
method requires that an electrical current be  introduced into the
ground through a pair of surface electrodes. The resulting poten-
tial field  is measured at the surface between a  second pair of elec-
trodes. The electrodes can be arranged into several different con-
figurations.
  Initially, several ER soundings can be made to establish sub-
surface background conditions and determine  the appropriate
electrode ("A") spacing  for  the ER profile lines. The sounding
data will be collected  by establishing a central reference  point
and varying the spacing between the current  and potential elec-
trodes.
  ER profiles  then can be assembled to indicate areas of low re-
sistance,  indicating the presence of potential contaminant plumes.
Areas of low resistivity also could indicate perched water zones
or low lying topographic areas.
  ER is best used, as with most geophysical  techniques,  in con-
junction  with other indirect testing. Soundings always should be
done to determine the depth to the water table and depth of any
clay layers  that might be present. ER is a quick  and inexpensive
technique generally available for site investigations.

Ground Penetrating Radar
  Ground penetrating  radar  (GPR) is a  reflection technique in
which high frequency radio waves  are radiated downward into
the subsurface and then  reflected back to a receiving antenna.
Variations in the return signals occur when subsurface materials
have different electrical properties. An interface between two lay-
ers having sufficiently different electrical properties will show up
in the radar profile. The effectiveness of GPR can be limited by
the penetration depth  of the radio waves. The effective pene-
tration is highly  site  specific and  dependent upon subsurface
boundaries, water content and clay content.
  The radar system electronics usually are mounted in a van, with
the antenna towed behind. An impulse radar transmits electro-
magnetic pulses of short duration into the ground from a broad-
band antenna. Pulses from the antenna are reflected from vari-
ous interfaces in the subsurface and then are picked up by the re-
ceiver section of the antenna.
  Unfortunately,  GPR systems are very expensive and  usually
are  used to locate contaminant plumes which are lighter than
water.

PORTABLE ANALYTICAL INSTRUMENTS
  Hazardous materials emergency response has spawned the de-
sign of many field analytical instruments due  to the added safety
requirements  inherent  in  the performance of such dangerous
work. Many of these instruments have been readily adapted for
use in site investigations to collect data which can be used in the
assessment  of a  site.  These  instruments can be used either to
gather data directly or as a screening technique to indicate where
further investigation is needed. This section of the paper dis-
cusses the more common types of instruments which can be used
to gather field data and their uses and limitations.

Organic Vapor Analyzers
  Two general types of instruments are currently used to analyze
for  the presence of volatile organic compounds.  Both detectors
rely on the ionization potential of the organic  compounds. Flame
ionization detectors (FID) rely  on a flame  which burns  com-
pounds to break chemical bonds to release  energy. Photoioniza-
tion detectors (PIDs) rely on ultraviolet light which is capable of
breaking the chemical bonds to release energy. The major differ-
ence in their application is that methane and  some very light  or-
     MONITORING

-------
ganic compounds cannot be ionized by ultraviolet light. As a re-
sult, PIDs can be used only if methane concentration is not a con-
cern. If qualification of methane is a  concern, a flame ioniza-
tion detector should be used to measure its presence.
  PIDs, such as the HNU portable photoionizer, and  FIDs, such
as  the Century Systems (Foxboro) Organic Vapor Analyzer
(OVA), have been used  in hazardous waste site investigations to
survey for the presence of organic compounds. Since  compound
separation and identification are not possible, total  concentra-
tions are measured as the benzene response equivalent. When
used in the  survey mode, the OVA is an efficient and accurate
indicator of total organic compound concentrations on a con-
tinuous sampling  basis  with a  response time of one  to  two
seconds. If instantaneous information is required on the presence
of methane  and non-methane hydrocarbons,  both instruments
can be used  in tandem. An example  of this application is a meth-
ane-vinylchloride  gas mixture which can  be common at landfill
vents.
  In cases where compounds identification or quantification is
required, portable gas chromatographs can be used.  The Model
128  OVA is  a portable FID gas chromatograph. The OVA has an
injection port which enables a sample to be injected from the air
intake line or by a syringe. As an initial approach to sample eval-
uation, this  technique can quantify the amount of volative organ-
ics  in a particular sample. This technique is useful when screen-
ing soil or water samples to select screen placement or selecting
worst  case samples for laboratory  analyses. This type of field
screening can lead to significant savings. If a gas chromatograph
(GC) column is present,  separation of each of the individual com-
ponents in a mixture is possible for identification and quantif-
ication. The OVA provides data in the 1 ppm concentration
range.
  Low ppb concentrations  can be  measured with a Photovac
portable  PID gas chromatograph. Several Photovac models cur-
rently are available. The most advanced is the model 10S70 which
has  two columns, is fully programmable, has a 100-compound
memory with computer peak search  and integration and an inter-
face to allow communications with  a home-base computer. This
type of  instrument can  separate  non-methane  organic  com-
pounds, identify them and perform  integration to provide accu-
rate concentrations. When used with prepared standards and lab-
oratory analytical checks, the Photovac can provide confident
analytical analysis on-site in a matter of minutes without the ex-
pense of an on-site laboratory.


Applications of Portable Analytical
Instruments
  Portable analytical instruments typically are employed in sev-
eral different approaches during site investigations. One approach
is an investigation  of the lateral of extent of shallow  volatile
organic soil  contamination. This is commonly done by opening a
hole in the ground with a slam-bar or other tool and inserting the
probe  of a  survey instrument to measure the concentration of
total organics. If measurements are referenced to a grid, they can
provide an indication of areas to be sampled for analytical analy-
ses or remediation.
  Another common use is to screen samples taken during an in-
vestigation.  To prevent costly  analyses  of every sample, only
certain samples are sent  to the  laboratory based on  the sample
screening. A small amount of sample is placed in a VOA vial and
heated to volatilize the organics into the air, or "headspace," in
the  vial. An aliquot of air is removed by a syringe and injected
into a GC column for analysis. This technique can also be used to
assist site geologists in deciding the depths at which to  finish wells
designed to monitor for volatile organic contamination.
  The photovac is gaining widespread use as a quantitative ana-
lytical tool to monitor drilling discharge water to prevent releases
of contaminants into the environment.
  Portable units to scan for concentrations of trace metals are ex-
pensive and relatively insensitive. They can be useful for sites with
widespread metals contamination in the 100 ppm and up range.
The  Colombia X-Met 700 portable x-ray  fluorescence (XRF)
spectrophotometer recently has been used successfully for this
application. Although it is expensive, the portable XRF currently
is gaining acceptance as a field analytical tool.
  Other direct-reading instruments which could be adapted from
safety-oriented tasks to portable analytical data gathering tools
include scintillometers, mercury vapor  analyzers and Hydrogen
Sulfide indicators.
  In addition to safety related instrumentation, some field ana-
lytical tools have been adapted from the laboratory and have been
around for some  years. They include pH meters, conductivity
meters, specific ion probes  for the measurement  of  dissolved
metals and colorimetric test kits for indicator parameters such
as chlorides and COD.
  The advantages of  using these techniques  are time and cost
savings. Field analyses provide data rapidly and allow data to be
factored into an  investigation while work is still taking  place.
Plans can be developed concurrently  for further investigation.
The number of field analyses need not be limited in these situa-
tions. A limited number of samples can be selected and analyzed
by the laboratory  for confirmation  of contamination, saving ex-
pensive laboratory time and reducing lab costs. Scheduling and
costs  are crucial on any project but are  paramount at sites  re-
quiring hydrogeologic investigations or soil investigation and ex-
cavation projects.

SUBSURFACE SAMPLING EQUIPMENT
  A large  portion of the effort devoted to any site investigation
should be toward understanding the geologic and hydrologic
properties of the soils or rock, the  extent of soil contamination
and the potential  for contaminant migration through the unsat-
urated zones. Since most hazardous waste site investigations  in-
volve surface or near surface disposal of liquid and solid ma-
terials, one of the  primary questions that needs to be answered is
how much of the material remains in the soil and how much con-
tinues to leach. Source investigations involve the combined appli-
cation of surface geophysics and direct soil sampling.
  Additionally, contaminant migration potentials within the un-
saturated zone and upper water table are important issues to be
addressed. This determination requires the collection and analysis
of subsurface  soil and rock samples. Many methods are avail-
able  for advancing boreholes to obtain samples or details  of
strata; the selection of a method is  dependent on the  extent of
potential contamination, the complexity of the site geologic con-
ditions and  the risk of contaminant  migration. The  principal
methods in use are: hand augers, power augers and rotary core
drilling methods.

Augers
  The auger technique is useful as a tool to collect preliminary
samples, to use with an OVA, HNU or  Photovac in a soil gas
survey or to collect near surface waste samples. At least six types
of light portable augers are  available  for sampling  soft to stiff
rolls.
  Hand augers may be used by one or two people. The hole is
advanced by pressing down on the cross bar as the tool is rotated.
Once the auger is full or has  collected sufficient material, it is
brought back to the surface and the soil is removed. The most
commonly used augers are the post hole, the helical and the spiral
augers.
                                                                                                          MONITORING

-------
  Hand augering usage is commonly limited to 3 to 7 ft. In tills
or clays that contain gravels or cobbles, hand augering may be
impossible. In uncemented  sands or gravels, it will not be pos-
sible to advance the hole below the water table since the hole will
continually collapse. Only samples of very limited size can be ob-
tained from the hole.
  In addition to collecting shallow subsoil samples, hand augers
also are used to create small diameter holes for a soil gas survey.
The holes are installed based on a grid system and the probe of an
OVA, HNU or Photovac is inserted into the hole. A sample of
the gas is collected and analyzed. This technique is a useful initial
method to define  the extent  of soil contamination by volatile
organic compounds.
  Because  hand augers are limited by soil conditions, depth of
penetration, quantity of sample, size of borehole and thickness of
the unsaturated  zone, they usually are confined to preliminary
surveys intended to generally define the limits of the contamina-
tion.

Power Augers
  The most common power  augering procedures used at  haz-
ardous waste site investigations are the continuous flight  solid
stem and hollow stem augers. Another type of augering  tech-
nique  is the bucket  auger. This procedure  uses open-topped
cylinders with base plates which have one or two slots reinforced
with cutting teeth which break up the soil and allow it to enter
the bucket as it is rotated. However, because of the limited depth
of penetration and the expense involved with operating a bucket
auger rig, it is rarely used.
  Solid stem augers drill much deeper holes with fewer problems.
However, this technique presents a serious problem when sub-
surface soil samples are required. To obtain the sample, the auger
string  must be pulled from the borehole.  Unless the soils are
stiff clays  or silts above the water table, the borehole is likely
to collapse and any sample  obtained would not be representative
of the stratum.
  Hollow stem augers consist of 3- to 6-in. diameter pipe with
continuous flights attached to the outside. Auger tailings are re-
moved from the hole by traveling up the flights and emerging at
the surface. When samples of the soil in advance of the augers
are required, the drilling is  stopped  and  a sampling device  is in-
serted into the pipe. The device is gently lowered to the bottom
of the hole and then either driven or pressed into the soil ahead of
the augers. Hollow stem augers allow drilling and sampling below
the water table, the acquisition of samples without pulling the
drill stem and the collection  of "undisturbed" samples.
  Auger rigs are mounted on four wheel drive trucks and all ter-
rain vehicles, allowing access  to most locations. Auger rigs will
not drill through rock and certain  cobbly  or till soils. Augers
usually are restricted to 200 or 250 ft in depth and still require cas-
ing if the hole is to remain after auger removal.


Rotary Core Drilling
  The  most common  use of rotary drilling  in site investigations
is to obtain intact samples of rock.  To do this, a "core-barrel"
fitted with a "core-bit" at its lower end is rotated and grinds
away an  annulus of rock. The stick of rock in the center of the
annulus passes up into  the core-barrel and is subsequently re-
moved from the  borehole when the core-barrel is full. The length
of core drilled before it becomes necessary to remove and empty
the core-barrel is termed a "run."
  In its simplest form, the core-barrel consists of a single tube
with an abrasive lower edge  which is rotated against the  rock
while fluid is passed under pressure.  However, most rock drilling
in  the United  States utilized the double  tube  swivel  type  core-
barrel. The importance of this tool is  that it contains an inner
barrel connected to the outer tube at the top via a swivel which
allows the inner barrel to remain stationary while the outer barrel
is rotated. The rock core "rides up" into the inner barrel.
  Double tube rock core barrels normally obtain a core between
5 and 10 ft long. Rock cores should be at  least 2 1/8 in. diameter
or larger. In soft to moderately hard rock, cores 12 in. in diameter
or larger can  be taken. A core barrel that collects a  2 1/8-in.
diameter rock core will produce a 3-in. borehole.
  Deep rock coring can be expedited with the use of the "wire-
line" technique. A wireline can produce  rock cores without the
removal of drill rods on the  outside core barrel. The inside core
barrel is retrievable as a separate  unit with the  use of a method
cable, or "wireline." An  empty  barrel can be sent down on  a
messenger to continue coring.  This type of technique saves time
and money for deep continuous coring operations.
  Core barrels should be held horizontally while the cores are ex-
tracted onto a  rigid surface.  The core should be properly tagged
and then placed in a core-box. Wooden spacer blocks should in-
dicate the top and bottom of each run.

Soil Sampling Devices
  The primary purpose of soil augering is  to collect subsurface
samples for visual examination and chemical and  physical tests.
In unconsolidated material capable of being drilled with an auger,
an important  component  of  any investigation should be  the
collection of soil samples ahead  of the  drill bit.  Hollow stem
augers are the most  common  drilling  technique,  but soil sam-
ples also can be collected ahead of drilling while using hydraulic
rotary methods.
  A  large number of samplers are available, most adapted from
geotechnical or soil  survey  investigations.  However, the  most
common device used in conjunction with hollow stem augering is
the split spoon sampler. In this device,  the sampler barrel is split
longitudinally  into two halves. The device  is  lowered into the
inner space of the hollow stem auger and driven into the soil by
repeated blows of a 140 Ib hammer falling through 30 in. Dur-
ing driving, the longitudinal halves are held together by the shoe
and  head which are screwed into each other.  The split barrel
allows easy examination and extraction of soil samples.
  Soil obtained with the split barrel sampler can be subject to  a
wide variety of field and laboratory tests. Samples can be  field
screened with an OVA  to determine presence of volatile organic
compounds. Soil should  be  logged  by  describing its texture,
color, grain size distribution, moisture  content and odor. If the
split  spoon has been properly decontaminated before sampling,
soil can be obtained for chemical analysis.
  Occasionally, large diameter "undisturbed" soil samples are re-
quired for complex physical  tests  such as hydraulic conductivity
determinations. Shelby or  push tube samplers often are used to
obtain undisturbed samples  of medium to  stiff consistency co-
hesive soils. The sampler is pressed into the soil. Care should be
taken because the sampler can be damaged,  either by buckling or
blunting or tearing the  cutting edge, when driven into very stiff,
hard or stony  soils. Undisturbed  samples of soft or loose  soils
must be taken  with a piston  sampler. Undisturbed samples  con-
taining much gravel or soft rock must  be obtained with a Deni-
son sampler.

SURFACE WATER/SEDIMENT SAMPLING
  One of the  primary  waste migration pathways is via surface
runoff to rivers, streams and lakes. The contaminant may move
as a dissolved constituent in  the runoff or attached to entrained
soil particles. One of the most cost-effective means of determin-
ing the extent of waste migration  is to sample surface water and
sediment.
     MONITORING

-------
Surface Water Sampling Techniques
  Surface water sampling locations are selected on the basis of
their probability for showing contaminants migrating from the
site. Prior to sampling, the surface water drainage in and around
the site must  be characterized using all available background
maps, topographic maps, serial photographs, river basin surveys
and other sources.
  Either grab or composite samples may be collected.  Grab sam-
ples are collected at one particular point and at one time. Flow
or time-weighted composited  samples are composed  of more
than one specific aliquot collected at various sampling sites and/
or at different points in  time. Because of the unknown safety
risks, as  well as the changes in chemical nature of the sample
that  may occur through  compositing, samples  containing haz-
ardous materials at significant concentrations shall not be com-
posited.
  If it is necessary to enter the water to  obtain the sample, it
should be done carefully  to leave the bottom sediments undis-
turbed. If the water is moving, a grab sample should be obtained
by pointing the open end of the container into the direction of the
flow. An attempt should  be made  to obtain the sample in the
middle of the stream and at mid-depth.
  The choice of a particular sampling device is dependent upon
the size of the water body,  the purpose of the sample and the
types and concentrations of wastes anticipated to be in the water
body. The types of samplers available include open tube, pond
sampler, manual hand pump, weighted bottle sampler, kemmerer
sampler and extended bottle  sampler. Of these, the pond sampler
and weighted bottle are used most often.
  The pond or dip sampler consists of a container attached to the
end of a long pole by an adjustable clamp. The pole can be of any
non-reactive materials such as  wood, plastic or metal. The sam-
ple is collected in a jar or beaker made of stainless  steel, glass
or non-reactive plastic. Preferably, a disposable beaker which can
be replaced at each station should be used.
  The weighted bottle sampler consists of a glass bottle, a weight
sinker, a bottle stopper and a line that is used to open the bottle
and to lower and raise the sampler during sampling. The sam-
pler can be either fabricated or purchased. This sampler is used to
take discrete samples  at predetermined intervals. The  pond sam-
pler can be used to develop composite samples from a single loca-
tion by collecting individual samples at regular intervals.

Sediment Sampling Techniques
  Sediment  samples are valuable  for locating pollutants  of low
water solubility and high soil binding affinity. Heavy  metals and
high molecular-weight halogenated hydrocarbons are examples
of contaminant groups which might be  found in greater con-
centrations in  sediment. A background sediment sample  should
be obtained  from sediments  upstream from the suspected source
for comparison. This is  especially  important if contamination
with heavy metals is suspected, because they occur naturally.
  Very simple techniques usually  are employed  for  sediment
sampling. Most samples will be grab samples from one particu-
lar locations although, for preliminary studies, several locations
may  be composited to reduce  the analytical requirements. Sug-
gested techniques include:
• In small, low-flowing streams or  near the shore of a pond or
  lake, the sample  container (typically an 8-ounce wide-mouth
  glass jar) may be used to scoop up the sediments.
• TO obtain sediments from larger  streams or farther from the
  shore of a pond or lake,  a Teflon beaker attached to a tele-
  scoping aluminum pole  by means of a clamp may  be used to
  dredge sediments.
• To obtain sediments from rivers or in deeper lakes and ponds,
  a spring-loaded sediment dredge or benthic sampler may be
  used by lowering the sampler to the appropriate depth with a
  rope. The  sediments thus obtained  are then placed into  the
  sample container.

DOWNHOLE TECHNIQUES
  Geophysical borehole logging has been used for oil, coal and
mineral exploration for years. Recently,  these techniques have
gained widespread acceptance in the performance of hydrogeo-
logic investigations. There are many different techniques avail-
able. Lithologic logging can be used  to measure properties of the
rock or soil in a borehole or well. Hydrologic logging can be used
to measure the properties of the fluids in  a borehole.  Downhole
television also can be used  to examine downhole  conditions. A
review of these  techniques and their  applications is presented
below.

Lithologic Logging
  Lithologic logging is used to measure the properties of down-
hole soils and rock. As with all logging techniques, it  is first ad-
visable to log at least one hole by physical examination of either
split spoon samples from soils or rock core samples for bedrock.
Rock core samples are more useful  than  chips because fracture
occurrence and widths (which are important conduits of ground-
water flow in bedrock) can be determined. Once lithology of at
least one hole is  known, the information can be correlated to
other holes.
  The simplest of the lithologic  techniques is caliper logging.
The caliper tool measures the diameter of a rock borehole using
three spring loaded prongs.  As the tool is  pulled up the hole, the
prongs spread out indicating the width of the borehole. Results
ared recorded electrically at the surface. The caliper method in-
dicates the relative hardness of  rock units. Caliper logging also
can indicate fracture location and occurrence. Caliper logging can
be the best tool to examine the  presence  of bedding plane frac-
tures.

Spontaneous Potential Logging
  Spontaneous potential logging is one form of electric logging.
It serves to measure the natural  electrical  potential that develops
between the  formation and the borehole fluids. This logging
technique must be performed  in  an open borehole  filled with
fluid. The  logging device consists of a surface electrode  and a
borehole electrode with a voltmeter to measure potential.
  Generally,  SP logs are read in terms of positive and negative
deflections from  an arbitrary  base line  which might correlate
either with permeable or impermeable zones. Information  re-
garding zones of higher permeability will  indicate likely contam-
inant pathways that may require further  investigation. The rate
at which the particular logging tool  is  lowered into the borehole
is another important criterion in  the data evaluation. The rate
should be sufficient  to detail specifics of the geology and pos-
sibly detect isolated  zones  of contamination or  likely contam-
inant pathways.

Resistivity Logging
  Resistivity  logging is another electric logging technique. Two
types of resistivity logging  techniques  are commonly employed
in downhole methods.  The  single point resistance  log is the sim-
plest form of electric log. In this  method, a single electrode is low-
ered into the hole and the return path for the current flow is furn-
ished by the  ground electrode. The single point resistance  log
measures the total resistance of earth materials.
                                                                                                          MONITORING

-------
  The second type of resistivity logging involves measurements
taken in the borehole in a similar manner as surface resistivities
investigations. In this system, four electrodes are commonly em-
ployed, two for emitting current (I) and two for potential meas-
urement (P). A number of different electrode configurations can
be used in resistivity logging to provide specific information. The
short normal spacing indicates the resistivity of the zone close to
the borehole where the drilling fluid might be an influence. The
long normal spacing  has more distance between the electrodes
and thus measures the resistivity further away from the borehole,
presumably beyond the influence of the drilling fluid. Both con-
figurations, short  and long-normal, measure a greater radius  of
influence than the single point resistance.
  A third configuration involving lateral devices utilizes widely
spaced electrodes for measuring zones  that are far from the bore-
hole. Because of the wide  spacings, lateral devices will not de-
tect thin beds of different  resistivity.  Boundaries of formations
having different resistivities are located  most  readily with short
electrode spacing, whereas  information on fluids (e.g., contam-
inated groundwater)  in permeable formations can  be  obtained
best with long spacings.

Density Logging Techniques
  Many downhole techniques are available to  measure the den-
sity of rock units. Natural gamma ray logging is perhaps the most
common. Natural gamma  logging measures  the  quantity  of
gamma rays naturally emitted by radioisotopes contained in clay
minerals. The higher the clay content of a rock, the more gamma
rays will be  emitted. This information is useful in indicating the
permeability of rock  units, since lower  permeability units gen-
erally contain more clay size particles.
  Other techniques which can be used to indicate density are high
resistivity density,  gamma-gamma density  and  neutron logging.
As with natural gamma ray logging, these techniques can be run
in cased boreholes. In other words, if the geology at a well needs
to be investigated and no records exist,  a cased hole can be logged
to gather qualitative  information. Natural gamma ray  logging
can be especially useful to  locate the  depth and thicknesses  of
bentonite seals on grout packs in monitor wells.

Vertical Seismic Profiling
  Recently,  some success has been met utilizing vertical seismic
profiling, or "cross-hole seismic." This  technique utilizes  an
array of detectors set in a  borehole to produce a three-dimen-
sional image of subsurface structures such as fractures. A seismic
source is used on the surface at some distance from the borehole
to create the seismic movement measured by  the detectors. Al-
though this technique is still under investigation, it shows promise
for  hydrogeologic investigation in the area of the development  of
three-dimensional data.

Hydrologic Logging Techniques
  The two most common hydrologic borehole logging techniques
are fluid conductivity and temperature logging.  These open bore-
hole techniques examine the properties of fluids within a bore-
hole and can provide  important information on  the natural cir-
culation of fluids in a borehole or the presence of contamina-
tion.
  These two probes are commonly available together and can  be
run as a hydrologic suite. Fluid conductivity logs provide a con-
tinuous record of  the conductivity or resistivity of fluid in the
borehole, which may be related to the conductivity of fluids in the
adjacent formation. A temperature log made simultaneously with
a fluid conductivity log allows  the most accurate conversion  to
specific conductance and also may identify contaminated or more
permeable zones in the formation.
   Fluid conditions in the boreholes should stabilize prior to im-
plementing this technique. The longer the period of time between
the drilling of the borehole and the logging of the hole, the more
accurate the results will be. Hydrologic logging can detect geo-
thermal gradients in groundwater in holes on the  order of 100 ft
thick.  Fractures contributing to groundwater movement some-
times  can  be noted  by changes  in  the geothermal  gradient.
Changes in temperature or conductivity also may be an indication
of contamination.

Downhole Television
   Downhole television has been  utilized  successfully to gather
in-situ information on boreholes and  wells in several  ground-
water  monitoring  programs. Borehole  television surveys are a
viable  alternative to  other  downhole instruments in the subsur-
face investigation stages of a groundwater monitoring program.
   Miniature borehole television cameras, developed for use in the
examination  of  nuclear reactor cores,  have been modified for
use in  borehole investigations. The lens attachments are capable
of looking sideward  or downward and include built-in lighting
assemblies.
  The in situ characterization of fractures that can provide path-
ways  for contaminant migration is critical in some investiga-
tions. Borehole television inspection can provide information on
the frequency, size and orientation of these fractures.  Vertical
correlations of rock cores in areas where voids are present (i.e.,
deep mining  or karst  topography) also can be simplified by this
technique.  Borehole  television can be used to check monitoring
well integrity. Casing inspections are especially useful  for con-
struction inspections when construction details are not known.
Well screens may be inspected in place to determine if rusting has
enlarged the screen openings or if screens have been damaged dur-
ing emplacement or well development operations. This informa-
tion may be  invaluable in the decision to decommission a well.
This technique is quick, inexpensive and creates a permanent
record for potential court cases.

Packer Assemblies
  Downhole  packers were developed for in situ measurements of
permeability  in geotechnical investigations such as dam  building
where data are critical. The packer assembly consists of two sets
of packers, upper and lower,  separated  by a perforated  pipe.
Each  packer is surrounded by a rubber seal  which is  fastened
to the assembly. Packers are inflated with compressed  air from
the surface to seal against  the  sides  of the borehole. Particular
zones then can be isolated for testing.
   For in situ  permeability testing, water is pumped in at a low and
constant pressure,  creating a constant  artificial head. Operating
pressures are stabilized with a gate valve until they remain oon-
stant.  Pressures then  are checked with a pressure gauge and  re-
corded for later use in permeability calculations. Each zone is
tested  for a defined period of time, and the amount of water re-
ceived by the zone is recorded with a standard water meter.
  Another  recent development is the insertion of electric or blad-
der pumps inside the assembly to remove water from the  bore-
hole. A particular zone can be isolated and tested with a  portable
GC for identification of zones carrying volatile organic  contam-
ination. Screens then can be set at optional levels for the study of
groundwater contaminant migration.

SPECIAL EQUIPMENT
   Equipment which was not discussed in this manuscript includes
air sampling equipment. Generally,  air sampling programs are
highly specific. Particulates such as asbestos can be collected on
filters  by high volume air  samplers, while organics  can be col-
lected  in packed tubes for  thermal desorption  by the use of low
     MONITORING

-------
volume air samples. Air investigations generally include the use of
continuously reading portable meteorologic stations.
  Air sampling technology recently has been extended to plume
tracking through the analyses of soil gas. Tubes are placed in the
ground, usually for several days. Organic compounds which vol-
atilize  from groundwater or soil travel upward over time in the
unsaturated zone of the soil and collect a trace levels in the tubes.
The tubes then are removed,  thermally desorbed and  analyzed
for volatile organics. The technique is not yet accepted as a stan-
dard methodology, but has been  useful in many cases and cur-
rently is gaining wide acceptance.
  Groundwater  pumps continue  to be refined, especially dedi-
cated systems. These systems include small diameter electric sub-
mersible pumps and bladder pumps. Materials used in the pumps
are being researched. Many bladder  pumps currently  available
for purchase are built almost completely of Teflon.
  Technological developments continue in  the area of contin-
uous  measurement devices for groundwater levels in monitor
wells. These devices range from single well units with solid-state
memories to full-scale 16-channel devices. Some multi-channel
devices can be operated up to one mile away from a well, making
them useful for aquifer tests.  Some actually have  built-in com-
puters, disk drives, monitors and printers and can analyze aquifer
test data, keep records on each well and even generate reports.
CONCLUSIONS
  The general structure of an investigation requires an analysis of
data gaps, particularly with regard to source definition, pathway
examination and assessment of potential for contaminant migra-
tion. Cost and time constraints involved in collecting certain types
of data can be balanced against the usefulness of the results. For
instance, it sometimes may be less expensive to install additional
wells than to use a costly indirect geophysical technique.
  Investigations generally rely on  indirect investigatory tech-
niques to start, followed by necessary confirmation steps. Spe-
cifically,  a technique such as magnetometry can be used to search
for buried containerized wastes, but this generally is confirmed
through test pitting and subsurface waste analyses. It is impor-
tant to note, however, that the limitations inherent in some tech-
niques may preclude their use in certain situations. A knowledge
of these limitations  is critical to cost-effective investigations and
proper cleanup strategies.
  Knowledge of indirect measurement techniques and field ana-
lytical equipment is critical to the collection of proper data and
the control of costs. Innovative technologies constantly are being
considered for their usefulness in these investigations. Use of this
type of equipment to paint a complete picture of site conditions
can  help the environmental management  professional  reduce
long-term liability and provide  useful data  for risk assessments
and feasibility studies. We hope that we have provided a basic
overview of the equipment,  its  limitations and its usefulness  to
environmental site investigations.
                                                                                                           MONITORING

-------
               A Statistician's  View  of Groundwater  Monitoring
                                              Douglas E. Splitstone
                                                  IT Corporation
                                            Pittsburgh, Pennsylvania
 ABSTRACT
   Owners and operators of hazardous waste facilities are required
 to monitor the groundwater around the hazardous waste facility
 and perform prescribed statistical tests on the resulting data to
 detect the existence, if any, of groundwater contamination. The
 statistical test prescribed by the U.S. EPA, which is referred to as
 Cochran's Approximation to the Behran's-Fisher Student's t-test,
 has  received increasing  criticism  as  being  inappropriate  and
 resulting in the allegation of groundwater contamination when, in
 fact, contamination did not occur. In addition, its specified ap-
 plication requires the existence of a well that  clearly monitors
 groundwater upgradient  of the site. The  U.S.  EPA recognized
 these problems. In publication of its regulations for owners and
 operators of permitted waste facilities, the U.S. EPA indicates the
 admissibility of statistical procedures alternate  to the Student's
 t-test.
   The inadequacy of the  Student's t-test is in part due to the lack
 of recognition that background groundwater quality varies both
 spatially and temporally. This paper describes a statistical model
 appropriate for describing spatially and temporally varying back-
 ground  water quality. An alternate statistical test procedure for
 assessing groundwater contamination as  a  deviation  from this
 background model is proposed. Use of this test procedure permits
 the owner/operator to control  the risk o' "false positive" test
 results to a small specified probability. Thus, the risk of unfairly
 being required  to conduct  an expensive  groundwater quality
 assessment program is  controlled while  assuring  protection
 against groundwater degradation.

 CURRENT GROUNDWATER
 MONITORING PROCEDURE
  A sampling program to determine whether a particular hazar-
 dous waste facility is contaminating groundwater  should be
 designed so that any contribution from the particular facility can
 be distinguished from other sources. These other sources, which
 may be both anthropogenic and natural, provide what is loosely
 referred to as the background. The design of a sampling program
 to distinguish the contribution  of the site of interest from the
 background is complicated by temporal and spatial variations in
 the background.
  Temporal variations in  groundwater may, in part, be explained
by recharge rates which  are weather related. Spatial  variations
among wells may be due to a variety of  reasons.  Some of this
variability may be reduced by ensuring that the wells monitoring
the aquifer beneath a particular waste disposal  site are located
relatively close together and drilled to the same depth. Neverthe-
less, spatial variations may occur even if there is no contribution
from the site being monitored. The variations might be expected
to be horizontal gradients across the site.
  The Behrens-Fisher Student's t-test recommended by the U.S.
EPA for monitoring  the groundwater near possible sources of
contamination1-2  has  increasingly become the topic of critical
reviews.3-4-5 The U.S. EPA procedure is based on establishment of
a background level before the disposal facility is put into opera-
tion and is, therefore, insensitive to spatial and temporal varia-
tions in the background. It can be used, however, if variations in
the background levels are taken into account.
  Considering the hypothetical situation in which measurements
of groundwater  quality  are  obtained without  sampling or
measurement  error,  the  sample variance  of upgradient  well
measurements, SB2,  reflects  only  real background  variations.
Measurements  on subsamples from  the sample  taken from a
downgradient well all have the same value Xm. The test statistic
given by U.S. EPA, which is referred to as Cochran's approxima-
tion to the Behrens-Fisher Student's t-test,' reduced to:

                   (xm  ' XB)/(SB2/MB)1/2               (1)
where XB  is the mean of the background values and NB is the
number of background values. This test is inappropriate because
in the absence  of contamination, Xm can vary as widely as the
background values. The test is appropriate for the inappropn tte
null hypothesis that Xm  equals the mean of the background
values. The appropriate null hypothesis for testing is that Xm is a
member of the same population as the background values. The
appropriate statistic for this test is:
  Compared to the first test statistic, the estimate of the standard
deviation used in the second is considerably larger. Therefore, for
many sets of measurements, while the first test statistic may er-
roneously indicate  that there is contamination, the second test
statistic may indicate the opposite.
  The U.S. EPA has recognized this problem. In publication of
its  regulations for  owners  and  operators of permitted waste
facilities,  the  U.S.  EPA acknowledges  the admissibility  of
statistical procedures alternate to the Students t-test.1 An alter-
nate statistical  procedure is described below.

PROPOSED ALTERNATE MONITORING
DATA EVALUATION PROCEDURE
  Although the background quality of groundwater varies both
spatially  and temporally,  temporal variations in background at
     MONITORING

-------
nearby points are expected to be related. Extending this assump-
tion, temporal variations (apart from sampling and measurement
error) are the same at nearby points. This procedure leads to the
expression of a mathematical model for the background quality
of groundwater around a waste disposal site. In this model, the
background quality is given by the sum of a general mean (/*), a
spatial effect (Wi) and a temporal effect (Tj).
  If contamination of the groundwater occurs  from the  waste
disposal facility, and if the monitoring wells surround the site so
that they are located both upgradient and downgradient from the
site, then contamination is unlikely to change the levels of a con-
stituent in all  wells by the same amount. Thus, contamination
from a disposal site represents a deviation from  the background
model, and such a deviation  can be detected.
  Monitoring groundwater quality around a disposal site is an ex-
periment with two factors: a spatial factor (or well factor)  and a
temporal factor. In conducting such an experiment, samples are
collected and measurements  made at each well at each sampling
time.  If the  background model holds,  these  measurements,
denoted by Yij, can be represented by:
                         M  +  W.
                                         .
                                                        (3)
                           = 1
                           = 1
 where N is the number of wells, M is the number of sampling
 times and Ejj represents the random sampling and measurement
 error. The measurements can be arranged in a two-way table and
 can be analyzed by a two-way analysis of variance.6
   Estimates of n, Wi and Tj may  be obtained from  the data.
 Denoting these estimated by m, w; and tj, respectively, they are
 given by:
                                                         (4)

                                                         (5)

                                                         (6)
 The differences between the observed values, yij, and the values
 predicted by the model are the residuals
                 rij = Xij - wi  -  tj - m               (7)

  As mentioned above, the contamination of groundwater by a
 waste disposal site will result in a deviation from the background
 model. Therefore, it is the residuals which contain the answer to
 the question of whether the site is contaminating the  ground-
 water. However, the residuals also are  affected by the sampling
 and measurement error. Because the assumed background model
 does not indicate whether the residuals  occur as a result of sam-
 pling and measurement error or are indicative of contamination,
 the model is incomplete.
  The analysis of  groundwater data  can  proceed no  further
 because there is no way to decide whether the residuals from the
 background model can  be  attributed  solely to  sampling and
 measurement error. The solution to this problem is the collection
 of replicate samples.  Replicate samples differ from each other
 only because they each contain an independent replicate of the
 sampling and measurement error. For example, a sampling design
 for groundwater monitoring might  specify collection   of two
 replicate samples from each well at each sampling time. Replicates
 are important in environmental sampling because the adequacy of
 environmental models is always in question.7 Replicates allow the
 variations unexplained by the model to be compared with the
 variations caused by sampling and measurement so that model in-
adequacies can be detected.
  The exact specification of the replicates that are needed is a
complex process since it depends on the environmental model.
For this reason,  replicate laboratory measurements obtained  by
splitting samples are rarely adequate because they do not reflect
sampling error. Specification of replicates often involves choosing
the time period between successive replicate samples. The model
should account for variations over long time spans with the term
Tj. Sampling  error usually accounts  for local and short-term
variations.
  Assume that true replicate samples can be obtained by resam-
pling the wells after a specific short time period. Consider a sam-
pling  program design which requires L replicate samples from
each well at each occasion that the wells are sampled. In most
cases, two (L = 2) replicates seem reasonable.  The measurements
on the samples may then be described by the  following model:

                                                        (8)
Tj
                                           Eijk
where, as before,
    i = 1 . . .  N (number of wells)
    j = 1 . . .  M (number of sampling periods)
    k =  1  . . . L (number of replicates)
  This model is the same as Model (3)  except that another
subscript has been added to index the replicates and a term (WT)jj
has been added to account for the various changes that occur
when the site is contaminating the groundwater. This term is
referred to as the well by period interaction and describes changes
in the relative relationship of the wells between time periods.
  If the sampling and measurement errors Eijk are statistically in-
dependent and normally distributed with the same variance, then
measurements obtained under this model can be analyzed by the
two-way  analysis  of variance appropriate  for models  with
replicate sampling.6 Estimates of /t, Wi, Tj and (WT)jj, which are
denoted by m, wj,  tj and ry, can be obtained as before with yij
replaced by:
The appropriate analysis  of variance (ANOV) table is given in
Table 1.
  The crucial comparison of the interaction with sampling error
which is indicative of groundwater contamination is made using
the F-statistic.6 The value of this statistic is given by:

                  NT(L-l)    45rij2           ,        (7)
            F =
                (N-D(T-l)
  If the interaction is found to be significantly greater than the
sampling error, then one concludes that the wells do not vary in
the same way over time, and thus there is local influence on the
measured contamination.
  True replicate samples can be obtained by sampling the same
well after a fixed period. The duration of this period may be as
short as hours or as long as days. The determi.iation of this period
can be established for a given well system with an initial sampling
experiment. Such  an experiment has been conducted on the
monitoring wells of a major industrial facility.

DETERMINATION OF REPLICATE SAMPLES
  The need for determining what constitutes a replicate sample
has been discussed above. It also was indicated that replicate
analyses of the same sample do not provide for the determination
of sampling variability. To determine true replicate sampling for
the monitoring wells, a sampling experiment was conducted over
a four-day period. This experiment required the sampling of each
well once a day on each of the four days. The samples were col-
                                                                                                         MONITORING

-------
                         Table 1
           Appropriate Analysis of Variance (ANOV)
                 for Groundwater Monitoring
               Table 4
Groundwater Monitoring Study TOC (mg/l)
SOURCE OF
VARIANCE
Between Wells
Between Sampling
Interaction
Sampling Error
Total
DECREES OF
FREEDOM
N-l
Periods T-l
(N-lHT-1)
NT(L-l)
NTL-1
Table 2
Groundwater Monitoring Study pH
DAY
ANALYSES
B i
SUM OF
SQUARES
LT'w2
t 1
LN1 t2
1 1
, II 2
L. . r. .
'J ']
"J ("ijk" " ij.)
"i t"ijk m'2
(Standard Units)
WELL
B2 B3 B4
1 1
2
3
4
2 1
2
3
4
3 1
2
3
4
4 1
2
3
4
7.00
7.00
7.00
7.00
6.50
6.60
6.60
6.70
6.60
6.60
6.70
6.80
6.60
6.60
6.60
6.60
7.00
7.00
7.00
7.10
6.90
6.90
6.90
7.00
6.90
6.90
6.90
7.00
6.90
6.90
6.90
7.00
7.30
7.30
7.40
7.40
7.30
7.30
7.40
7.40
7.30
7.40
7.40
7.40
7.30
7.40
7.40
7.40
6.50
6.50
6.60
6.60
6.50
6.50
6.60
6.60
6.50
6.50
6.60
6.60
6.50
6.50
6.50
6.60
                         Table 3
          Groundwater Monitoring Study Conductivity
                    OtMHOS @ 25 °C)

DAY ANALYSES
1 1
2
3
4
2 1
2
3
4
3 1
2
3
4
4 1
2
3
4
Bl
1039.00
1037.00
1012.00
1021.00
1062.00
1074.00
1068.00
1041.00
1080.00
1067.00
1071.00
1064.00
1045.00
1050.00
1033.00
1039.00
B2 B3
913.00
904.00
911.00
917.00
1097.00
1090.00
1087.00
1087.00
1062.00
1060.00
1050.00
1055.00
1011.00
1027.00
1005.00
966.00
599.00
587.00
592.00
599.00
688.00
683.00
686.00
672.00
698.00
696.00
695.00
696.00
712.00
712.00
712.00
717.00
B4
1355.00
1353.00
1265.00
1352.00
1339.00
1332.00
1337.00
1341.00
1339.00
1341.00
1339.00
1340.00
1327.00
1351.00
1345.00
1356.00
DAY
1



2



3



4




DAY
1



2



3



4



ANALYSES
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
Groundwater
ANALYSES
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
Bl
28.40
27.80
27.10
27.90
37.60
33.80
36.80
32.20
35.50
33.40
35.50
33.90
32.20
31.80
32.40
31.70
TableS
Monitoring
Bl
21.00
21.00
22.00
18.00
21.00
21.00
23.00
21.00
11.00
10.00
18.00
16.00
12.00
10.00
11.00
12.00
WELL
B2 B3
102.00
98.80
102.00
98.50
72.60
75.30
75.80
73.90
64.80
64.80
62.40
64.00
64.60
61.70
61.30
62.80
Study TOX
28.80
29.30
28.10
31.30
44.70
42.90
42.70
44.70
40.80
40.30
38.50
40.50
38.90
39.40
38.20
41.70
0*/1>
WELL
B2 B3
29.00
21.00
32.00
29.00
35.00
36.00
33.00
34.00
28.00
31.00
32.00
31.00
26.00
27.00
31.00
28.00
<10.00
<10.00
<10.00
<10.00
11.00
<10.00
<10.00
<10.00
16.00
16.00
19.00
17.00
<10.00
<10.00
<10.00
<10.00
B4
101.00
97.90
98.70
98.20
103.00
100.00
106.00
98.20
102.00
100.00
102.00
103.00
93.40
93.10
97.30
96.70

B4
19.00
16.00
17.00
18.00
19.00
22.00
21.00
25.00
25.00
22.00
25.00
27.00
17.00
21.00
19.00
23.00
                                                              lected after three well volumes of water were removed from the
                                                              well. Each sample was analyzed four times for pH, conductivity,
                                                              total organic carbon (TOC) and total organic halogens (TOX).
                                                              The data collected during this experiment are found in Tables 2
                                                              through 5.
                                                                 The sampling  plan was  designed to compare the variability
                                                              among samples from a given well taken within a relatively short
                                                              time span (approximately 24 hr apart) to the variability among
                                                              repeated analyses on the same sample. The  sampling plan is,
                                                              therefore, hierarchical, and the data were analyzed accordingly
                                                              using the appropriate  ANOV technique.' The  results of these
                                                              analyses are found in ANOV Tables 6 through 9.
                                                                 The values of the F statistics  shown in these  tables are all
                                                              greater than the tabulated critical values of F for the appropriate
                                                              degrees of freedom.* They clearly indicate that highly significant
                                                              variability exists  in all the measured constituents among the wells
                                                              and among samples from the same well during a comparatively
                                                              short period  of  time. The  significance of variability along the
                                                              wells is as expected for reasons given above. The significance of
10
     MONITORING

-------
the variation among analyses of a given sample clearly illustrates
that sampling variability is important.
  To  further illustrate this point, assume that the four samples
taken from Well Bl describe the background groundwater quality
for TOX. Thus, the background  mean TOX is 16.19 /ig/1, and
standard deviation  of the background TOX  is 6.11  /ig/1.  Now
assume that downgradient Well B4 was to be samples on one of
the four days of this experiment chosen at random. The mean and
standard deviation of the four  repeated analyses for TOX of the
daily samples is given in Table 10.  Also presented in this table are
the calculated and critical values of the Student's "t" statistic as
prescribed by Appendix IV of 40 CFR 264.
  Comparing the calculated and critical values of the Student's
"t" statistic for these four consecutive days serves to illustrate the
importance of the  sampling variability. On two of the four days
(Day 2 and Day 3), the calculated value of Student's "t" exceeds
the critical value. Thus, a conclusion of groundwater degradation
would be made. However, this is not true of the samples taken on
Day 1 and Day 4. If the quarterly sampling day were picked at
random  from these four, there would be a 50% chance of con-
cluding that groundwater degradation occurred solely due to ran-
dom variation introduced by the act of sample collection.


VARIATION
SOURCE
Wells
Samples
Within
Well
Analyses
Within
Sample
TOTAL


VARIATION
SOURCE
Wells
Samples
Within
Well
Analyses
Within
Sample
TOTAL


VARIATION
SOURCE
Well
Samples
Within
Well
Analyses
Within
Sample
TOTAL
Table 6
pH ANOV
DEGREES SUM QF MEA[) p
FREEDOM SO-UARES SQUARES STATISTIC
3 6.0162 2.0054 51.03
12 0.4713 0.0393 12.61


48 0.1500 0.0031


63 6.6375
Table 7
Conductivity ANOV
DEGREES SUM QF MEAN F
™.?«L,., SQUARES SQUARES STATISTIC
FREEDOM
3 3,578,628 1,192,876 127.81
12 111,990 9,333 43.21


48 10,348 216


63 3,700,968
Table 8
TOC ANOV
DEGREES SUM op MEAN p
FREEDOM SQUARES SQUARES STATISTIC
3 48,326 16108.70 44.22
12 4,372 364.32 132.96


48 132 2.74


63 52,829


SIGNIFICANCE
LEVEL
.999
.999








SIGNIFICANCE
LEVEL
.999
.999








SIGNIFICANCE
LEVEL
.999
.999






Table 9
TOX ANOV
VARIATION DEGj>EES SUM OF MEAN F
SOURCE FREEDOM SQUARES SQUARES STATISTIC
Well 3 2,918 972.63 17.26
Sample 12 676 56.36 12.44
Within
Well
Analyses 48 218 4.53
Within
Sample
TOTAL 63 3,811
SIGNIFICANCE
LEVEL
.999
.999




                Table 10
TOX Concentration for Downgradient Well B4
DAY
1
2
3
4
MEAN
Ug/l
17.50
21.75
24.75
20.00
STANDARD
DEVIATION
Pg/t
1.29
2.50
2.06
2.58
CALCULATED
STUDENT'S "t"
0.79
2.82
4.65
1.91
CRITICAL
STUDENT'S "t"
1.84
1.99
1.94
1.96
                                                                                          Table 11
                                                                        Expected Mean Squares for Groundwater Monitoring
                                                                             VARIATION
                                                                              SOURCE
                          EXPECTED MEAN
                             SQUARES
                                                                               Well

                                                                           Samples Within
                                                                               Well

                                                                          Analyses Within
                                                                              Sample
                                                                                          Table 12
                                                                    Variance Components for Groundwater Monitoring Parameters
VARIATION
SOURCE
Analyses, o.
Sampling , o
PH
0.003
0.009
CONDUCTIVITY
215
2,279
TOC
2.74
90.39
TOX
4.53
12.95
                                                                   The selection of a sampling plan for continued groundwater
                                                                 monitoring can be made with the information generated by the
                                                                 analysis of variance. By equating the mean squares estimated by
                                                                 the analysis of variance to their theoretical expectations, one can
                                                                 estimate the contribution of each component (e.g., sample collec-
                                                                 tion or analysis) to the  total variability of the measurement. If a%
                                                                 and as are used to symbolize the variability due to analysis and
                                                                 sample collection respectively, then the expected mean squares for
                                                                 the analysis of variance are given in Table 11.
                                                                   In Table 11, W2 symbolizes the natural background variation
                                                                 among wells.  The resulting estimates of a2, and a^are given in
                                                                 Table  12 for the indicator parameters observed.
                                                                   The characterization of the contaminant concentration for a
                                                                 well undergoing quarterly monitoring is by the mean of the obser-
                                                                 vations taken during the quarterly monitoring period. A sampling
                                                                 plan should be designed to minimize the variation in this mean. If
                                                                 S samples are taken during the quarterly monitoring period and A
                                                                 analyses are run on each sample,  the variance of the quarter mean
                                                                 is given by:               2     2
                                                                                         SA
                                                                                                        MONITORING    11

-------
  The effect on this variance of varying the number of samples,
S, and analysis per sample, A, then can be investigated. If the
analytical cost is constrained by requiring the S + A £.4, then the
minimum  variance, and  hence maximum  precision,  of the
quarterly mean value for the indicator parameters for a given well
is obtained for the same analytical cost if the well is sampled four
times each  quarter approximately 24 hr apart and one analysis is
performed  on each sample. These results are given in Table 13.
The information  on the last line of the table indicates  that the
precision of the quarterly mean is approximately doubled (the
variance halved)  from  the U.S. EPA  recommended  sampling
scheme of one sample per quarter by sampling each well twice and
analyzing each sample once. This reduction in analytical cost may
be sufficient to compensate for increased sampling costs.
                           Table 13
                 Variance of Well Quarterly Mean
NUMBER OF
SAMPLES
1
2
4
2
NUMBER OF
ANALYSES
PER SAMPLE
4
2
1
1

0
0
0
0
pH
.00975
.00525
.00300
.00600
CONDUCTIVITY
2332
1193
.75
.25
623.50
1247
.00
TOC
91.
45.
23.
46.
08
88
28
56
TOX
14
7.
4.
8.
.08
61
37
74
 CONCLUSIONS
   The salient conclusions are summarized as follows:
 • The portion of the total measurement variability due to the
   act of sampling is much greater than that due to repeated chem-
   ical analyses of the same sample. Thus,  repeated chemical
   analyses of the same sample do not necessarily improve quar-
   terly characterization of groundwater.
 • Adequate replicate samples may be obtained by repeated sam-
   pling of each well approximately 24 hr apart.
 • The precision  of the quarterly measurement for groundwater
   contamination will be greatly improved from that specified in
   Federal Regulations if each well is sampled twice within a four-
   day time span and only one  analysis is performed on each
   sample. This sampling  scheme also  will permit the use of the
   analysis of the variance statistical procedure to adequately test
   the hypothesis of groundwater contamination.
•  Continued use of the statistical procedure specified in the Fed-
   eral Regulations  will, more  than likely, yield results alleging
   that groundwater contamination has occurred when the appli-
   cation of adequate statistical  procedures  will conclude that
   there is not contamination.
REFERENCES
1.  U.S. EPA,  "Interim Status Standards for Owners and Operators of
   Hazardous  Waste Facilities." 40 CFR 265.92.
2.  U.S. EPA, "Regulations  for Owners and Operators of  Permitted
   Hazardous  Waste Facilities," 40 CFR 264.97.4.ii.
3.  Liggett, W., "Statistical  Aspects of Designs for  Study Sources of
   Contamination," Quality  Assurance for Environmental Measure-
   ments,  ASTM STP867, June 1985.
4.  Miller, M.D. and Kohoul.  F.C., "RCRA Ground Water Monitoring
   Statistical  Comparisons:  A  Belter  Version of Student's T-Test,"
   Proc. of the NWWA/AP1 Conference on Petroleum Hydrocarbons
   and Organic Chemicals in Ground Water Prevention, Detection, and
   Restoration, National Water Well Association, 1984.
5.  Ross, L. and Elton, R., "Maximizing the Statistical Performance of
   Ground Water  Monitoring Systems,"  Proc.  of  the NWWA/APl
   Conference on Petroleum  Hydrocarbons and Organic Chemicals in
   Ground Water  Prevention.  Detection, and  Restoration, National
   Water Well Association, 1984.
6.  Snedecor, G.W. and Cochran, W.G.. Statistical Methods, 5lh ed.,
   The Iowa State University Press, Ames, I A, 1950.
7.  Krumbein,  W.C., "Experimental Design in the  Earth  Sciences,"
   Transactions of the American Geophysical Union,  36, 1955, 1-11.
8.  Kempthorne, O., The  Design and Analysis of Experiments,  John
   Wiley and Sons, New York. NY. 1952. 104-109.
12    MONITORING

-------
                       Data  Evaluation  in a  Groundwater Study
                               of Waste  Management  Practices
                          in  the  Phosphate Processing Industry
                                             Edward W. Mullin, Jr.
                                                  Jack  S. Greber
                                              William E. Thompson
                                               PEI Associates, Inc.
                                                 Cincinnati, Ohio
ABSTRACT
  The phosphate processing industry  has  been the focus  of
several groundwater monitoring programs within the past several
years. As a result of these studies, a relatively large data base has
been   assembled  which,   in   addition  to   evaluating  the
characteristics of the various wastes generated by this industry,
also includes groundwater monitoring data  which were used to
evaluate the impact of the industry on groundwater quality. The
interpretation of the monitoring data involved the use of several
different evaluation techniques which will be discussed.
  Student's t-test is one traditional method of evaluation that was
used.  This method compares the average and standard deviation
of a set of paired values. This method has been used extensively in
hazardous  waste-related groundwater  studies  and has been
criticized for its lack of sensitivity.
  Another traditional method of evaluation is the comparison of
Stiff diagrams.  These diagrams are graphical representations of
the basic quality of a water sample. This method is very useful
when dealing with the wastes from the phosphate industry,  but
the method is somewhat subjective.
  Several newer methods also were used in the  course of these
projects. These  methods included the calculation of ion  indices
and the use of cluster analysis.
  The ion indices are calculated using basic water quality com-
ponents such as  sodium and chloride. Through the comparison of
indices from impacted wells with indices from background wells,
an evaluation of contamination and attenuation can be made.
  Cluster analysis was used to identify any trends or relationships
between monitoring wells.  Cluster analysis  is a collective term
covering a variety of statistical techniques for delineating natural
groups or clusters. Evaluations were made based upon the degree
and the order in which the individual wells were clustered. This
method proved  to be  quite useful in confirming trends which
other methods also identified.
INTRODUCTION
  Over the past several years, PEI Associates has been involved
with evaluating the waste management practices in the phosphate
industry. These evaluations have involved the  installation  of
groundwater  monitoring  programs  at  phosphate processing
facilities. In the course of these projects, eight different facilities
with a total of 85  wells were monitored. Samples were collected
from these wells at varying times; each well was sampled at least
twice and analyzed for a wide variety of water quality parameters.
The interpretation of the monitoring  data involved the use  of
several evaluation techniques which will be discussed.
STUDENT'S T-TEST
  Groundwater monitoring data collected for the purpose of ful-
filling RCRA groundwater monitoring permit requirements have
been evaluated using Student's t-test. The Student's t-test is the
principal criterion for testing hypotheses concerning two popula-
tion means. The t statistic is the deviation of a normal variable (X)
from its mean (/*) measured in standard error units (Sx).
    t = (X - M)/SX                                     (1)
  This test compares the average and standard deviation of the
measurements for each parameter with the same parameter as
measured in another well, usually the background well. This test
was used on some of the data. However, it was not used on all the
data since  there  were  not enough replicates  to make the test
statistically valid. At least three or more replicates are required to
make the test valid. Due to various constraints on the projects,
only two sets of samples were collected at some of the sites. In ad-
dition to the lack of sufficient replicates, the Student's t-test has
been  shown  to  generate false  positive  results  as  has been
documented  in  the Federal  Register.  Therefore,  additional
methods for evaluating the data were used.

STIFF DIAGRAMS
  Stiff diagrams  are a traditional method of depicting the basic
water types. Only the major cations and anions typically are used
in the  diagrams (Ca + 2, Mg+2, Na+, K + , C1-, alkalinity and
SO4-2). The  concentrations  of the ions are converted to milli-
equivalents/liter and then plotted on a Stiff diagram.  The shape
of the resulting polygon is indicative of the type of water being
evaluated.  For example, the Stiff diagram of a sodium carbonate
water  is distinctly different from  a calcium sulfate  water.
Therefore the comparison of a Stiff diagram of a sample of liquid
waste from a processing facility to the diagrams from monitor
wells will enable  one to graphically determine if there has been
any impact on the groundwater from the liquid waste.
  Figure 1  shows  such a comparison for one of the test sites. The
waste sample  (GL-1) and the background  well are compared to
two different wells from the site (PEI-2 and PEI-3). Neither of the
diagrams for PEI-2 or PEI-3  resemble the diagram for the back-
ground well, indicating that the water quality in these wells differs
from background conditions. The Stiff diagram for PEI-3 bears a
close resemblance to the diagram of the waste (GL-1), indicating
that some  contamination  has occurred in this  well. The Stiff
diagram for PEI-2 does not  resemble the diagram for GL-1 as
closely as  PEI-3, but  is  more similar to GL-1  than  to the
background well,  and therefore, it can be concluded that PEI-2 is
contaminated but not as severely as PEI-3.
                                                                                                    MONITORING    13

-------
       MO+K
        Mg
          50       25       0       25       50
           Cations (mg eq'l)  Amons (mg eq/l)
             BACKGROUND vs. PEI-2
       Na+K
        Mg
          100      50       0       50      1UU
           Cations (mg eq/l)  Anions (mg eq/l)
               GYP LIQUID vs. PEI-2
                                                            Na+K
                                                                       Mg
                                                                                                    S04
                                                                50       25      0       25       50
                                                                 Cations  (mg eq/l)  Anions (mg eq/l)
                                                                   BACKGROUND vs. PEI-3
                                                                       Mg
                                                                100      50      0       50      100
                                                                 Cations (mg eq/l)  Anions (mg eq/l)
                                                                    GYP LIQUID  vs. PEI-3
                                                        Figure 1
                                              Stiff Diagrams from the Test Site
 ION INDICES
   Another method which has been used with some degree of suc-
 cess is the calculation of ion indices. These indices were developed
 by the U.S. Geological Survey during an evaluation of phosphate
 mining and processing in  Florida.1 The indices are a numerical
 evaluation  of the concentration of the major anions and cations
 within a sample and in some cases the sample concentrations were
 compared to the concentrations in the background samples.
   The two  most useful indices are the index of Nonmineral Input
 (INMI)  and the Percentage  of Ion Increase (PII). The  INMI
 measures the percentage increase in concentrations of sodium
 above chloride and calcium plus magnesium above sulfate plus
 alkalinity compared to the same parameters in the background
 samples.
                            '-"«'MCl-CDl
                                C1 + A1
|(Ca+Ca'
Mg-Mg ') - (S04-S04 '+Alk-Alk ') I - 1C-A ] - | C' -A'
              c, + Ai
                                                      (2)
                                              x 100
where:
  INMI

    Alk
      C
     A
  Index of Nonmineral Input, in percent above
  background
  Alkalinity, in milliequivalents/1
  Cation sums, in milliequivalents/1
  Anion sums, in milliequivalents/1
  Background conditions
                                                      The  PII is used to evaluate the increase in ions above back-
                                                    ground concentrations and is based solely on the sum of selected
                                                    cations and anions.
                                                                                       100
                                                      (3)
                                                     where:
     PII =  Percentage of Ion Increase, in percent above
           background
      C =  Cation sums, in milliequivalents/1
      A =  Anion sums, in milliequivalents/1
       ' =  Background conditions

  The senior author of the U.S. Geological Survey report1 consid-
ered an INMI of approximately 5% to be above background
levels for the groundwater systems in Florida and to reflect an im-
pact from the waste management practices. The relationships be-
tween the INMI and PII also can be used to evaluate the extent of
seepage and whether  or not the soil is attenuating the seepage
through ion exchange. For example, if the INMI and PII are both
significantly elevated above background conditions, this probably
indicates that the seepage has not been altered by ion exchange. If
the INMI is low and the PII is high, then this result may indicate
that the seepage is being  altered through ion  exchange with the
soil or that another source of contamination is influencing the
system.
 14    MONITORING

-------
  At  the  test site  previously  mentioned,  ion indices  were
calculated for each sampling period. Table 1 shows the indices
(calculated with PEI-1 as background) for the second sampling
round. As can be seen from the data, the conclusions concerning
PEI-2 and PEI-3 are supported by the indices.  These two wells
have the highest indices for both INMI and PII. The lower INMI
for PEI-2  indicates that more attenuation has  taken place  at
PEI-2 than PEI-3,  while the other wells show some degree of im-
pact. The high INMI for PEI-5, the second background well, is
most likely a result of it being completed in a different lithology
than PEI-1, rather than it showing the effects of seepage.

                          Table 1
                   Ion Indices at the Test Site
  INMI
   PII
                    32%
                   695%
 411%
1332%
 32%
123%
32%
 5%
  0%
100%
  These methods have proven to be quite useful for evaluations
of the phosphate industry where the waste being studied consists
mainly of calcium and sulfate, both of which are included in the
Stiff diagrams and  the indices. However, these methods would
not be very useful in evaluating seepage  from wastes which did
not impact any of the major cations or anions, such as  organic
contamination, and these evaluations often may be subjective
when there is not a great deal of dissimilarity in the diagrams or
the indices.

CLUSTER ANALYSIS
  The ground water data collected during these projects also were
analyzed by cluster analysis to identify any water quality trends or
relationships between the monitoring wells and the waste samples.
Cluster analysis  is a collective  term covering a wide variety of
statistical techniques for delineating natural groups or clusters in
data sets.2 The objective of this type of analysis is progressive
grouping of the variables into clusters according to their degree of
statistical similarity.
  The  cluster procedure in  the  BMDP  Statistical  Software
package was used for this analysis.3 The exact procedure used was
Cluster Analysis of Cases (Method 2M). This procedure uses a
hierarchial approach to building the clusters. Initially, each  case
(an individual sample from a monitoring well or waste sample) is
treated as a single member cluster; then, the two most closely
related cases are combined into a new cluster and the "distances"
between the new cluster and the remaining ones are computed.
  The  distance  measure calculated is the  Euclidean  distance
radius between the  clusters. The  general formula used for this
calculation is as follows:
                                                         (4)
where:
  djk   = distance between two cases of clusters j and k
  Xjj   = value of the ith variable in the jth case
  This process continues until all the cases are combined into one
large cluster. Figure 2 is a graphical representation of this process;
the initial cases are placed on a line in an order that is determined
by the program. Vertical lines then are drawn from each case until
it joins with another. By using the computed distance between
clusters as the vertical scale, the vertical location of each joint or
cluster shows the dissimilarity of the joined groups. Therefore,
clusters formed  near the  top represent homogeneous groups,
whereas clusters  formed further down represent clusters  formed
only because the process  goes to its logical end  of one large
cluster.
                                                                                  Cluster Diagram
                                                                                           Cases
                                                                      E  --
o
o>
e
O)
0>

                                                                                            Figure 2
                                                                                 Example of Hierarchial Clustering

                                                                    For the purposes of these studies, each sample from a monitor-
                                                                  ing well or waste sample was treated as one case, and all samples
                                                                  from a given sampling period were  analyzed together; i.e., no
                                                                  clustering was conducted over the different sampling periods. The
                                                                  variables  used in the analyses included the basic  water quality
                                                                  parameters (Ca, Mg, Na, etc.)  and several parameters chosen as a
                                                                  result of their  concentrations in the waste (sulfate,  fluoride,
                                                                  Ra-226, etc.).
                                                                    The results of the cluster analysis were evaluated based on the
                                                                  criterion that any increase in distance between clusters that was
                                                                  twice the  previous distance (a  100% increase) was considered to
                                                                  be a significant difference between clusters.
                                                                                                Cases
                                                                                     PEI  PEI  PEI  PEI   PEI  PEI   0L
                                                                                      1564231
                                                                      ~   100-1-
                                                                       39
                                                                      =   150-f-
                                                                      £   200- -
                                                                    a E   250- -
                                                                    O (ft
                                                                    S »   300- -
                                                                           1000- -
                                                                           2000- -
                                                                           3000
                                                                      a
                                                                         10000	
                                                                         20000- -
                                                                         30000- -
                                                                                           Figure 3
                                                                                        Test Site Cluster
                                                                                                         MONITORING    15

-------
  The cluster analysis  of the data represented previously in the
Stiff diagrams supported the conclusions of the Stiff diagrams.
As  shown  in Figure 3, the six  wells analyzed showed definite
clustering.  The background wells (PEI-1 and PEI-5) were deter-
mined to be similar on the basis of the clusters, as were PEI-4 and
PEI-6 (these wells were furthest downgradient  from the waste
source). Wells PEI-2 and PEI-3  also clustered together and were
determined to be the least similar to the other well clusters and the
most  similar to the waste sample GL-1.
  At  this site, all the data evaluation methods provided the same
conclusions as  to which wells were being affected by the waste
management practices. Cluster analysis is a worthwhile method to
assist in the evaluation of data when traditional methods may not
have  the necessary statistical power to sort out all the variables.
  The power of the cluster analysis may be increased if a factor
analysis is performed on the data set prior to the cluster analysis.
The factor analysis will identify which analytical parameters are
responsible for the majority of the experimental variance. Use of
factor analysis also will prevent the researcher from inadvertently
weighting the cluster  analysis by the inclusion of two or more
parameters which represent the same effect. The primary variable
should be identified  by the  factor analysis,  and  only those
variables should be included in the cluster analysis.
CONCLUSIONS
  As can be seen, there are a wide variety of methods available
for the evaluation of groundwater monitoring data. The use of
cluster analysis with or without factor analysis is a valuable tech-
nique which can help identify trends in the  data that otherwise
may not be recognized.

ACKNOWLEDGEMENTS
  The authors would like to thank the U.S. EPA and the Florida
Institute of Phosphate Research for their assistance in funding the
projects  which were discussed in this paper.

REFERENCES
I.  Miller, R.L. and Suicliffe, H.. Jr., "Effects of Three Phosphate In-
   dustrial Sites on Ground Water Quality  in Central Florida, 1979 to
   1980," U.S. Geological Survey Water-Resources  Investigations Re-
   port 83-4256.
2.  Andberg, M.R., Cluster Analysis for Applications, Academic  Press,
   New York. NY, 1973.
3.  Dixon, W.J.,  Ed., BMDP Statistical Software, University of Cali-
   fornia Press, Berkeley, CA, 1981.
16    MONITORING

-------
               Plant Cuticular and  Dendrochronological  Features
                                      as  Indicators  of Pollution

                                                G.K. Sharma, Ph.D.
                                                Biology Department
                                              University of Tennessee
                                                  Martin, Tennessee
                                                          and
                                                 Harvard University
                                             Cambridge, Massachusetts
ABSTRACT
   In a highly industrialized world, it is imperative that scientists
explore different methods of monitoring environmental contam-
ination. One new field of environmental research, begun in re-
cent years, may do this. This field is the study of leaf cuticle and
annual growth rings in woody plants.
   Several plant taxa growing in habitats characterized by varied
levels of environmental contamination were studied by the author
of this  paper for  dendrochronological and cuticular patterns.
Growth rings, ring and porous wood, and related features were
analyzed with a dendrochronograph. Cuticular patterns such as
stomatal frequency, stomatal size and  trichome frequency and
type were studied in relation to environmental contamination in
various habitats. Most plant populations under investigation were
in the industrial areas of Nashville and Memphis, Tennessee—
known for their high levels of environmental pollution.
   Statistical analysis of the data revealed the significance of plant
features as  indicators of environmental pollution in  the  area.
Growth  rings were,  for  example, extremely narrow  in highly
polluted areas, while the plant populations exposed to relatively
low levels of pollution exhibited fully  developed annual incre-
ments. Growth rings tended to be unusually irregular in plants
growing in highly polluted areas. Cuticular patterns such as tri-
chome frequency  and type were quite diagnostic for determin-
ing pollution levels in the environment.
   With an increase in pollution level, the stomatal frequency de-
creased  and the trichome frequency increased—possibly a re-
sponse to offset the detrimental effects  of pollutants on various
metabolic reactions and reproductive strategies in plants. In addi-
tion,  floral productivity  seemed  to  be adversely  affected in
polluted areas. The results demonstrate that tree ring character-
istics and leaf cuticular patterns can be used as monitors for eval-
uating the growth  responses of a plant species to varied levels
of environmental pollution.
  Analysis of annual increment variation  can be used to deter-
mine the degree of stress exerted by environmental pollution on
plant growth. Results of several cuticular and experimental stud-
ies conducted by the author substantiate  the findings reported
herein. These results suggest that plants can be used as extremely
reliable indicators  of environmental pollution, including hazard-
ous waste sites, provided  long-term monitoring is done for  a
variety of plant species.

INTRODUCTION
  Studies3'5 have demonstrated the relationships between plants
and the environment. Damaging effects of fluorides and sulfur
 dioxide on plants have been documented.'
   In 1934, Chamberlin1 determined that industrial pollutants of a
 large mid-western city in the United States were detrimental to
 coniferous flora of the area. He especially referred to the fatal
 effects  of industrial pollution  on  Pinus banksiana.  Hill and
 Thomas3 observed that there was a decrease in alfalfa yield after
 exposure to sulfur dioxide.  Pyatf suggested lichens as possible
 indicators of air pollution in a steel-producing town in Wales. His
 studies  revealed that thallus size decreased and the lichen flora
 decreased in the number of  species present with increasing prox-
 imity of the pollution source.
   While working in the forests of the northwestern United States,
 Scheffer and Hedgcock5 found that sulfur dioxide from smelters
 produced  a characteristic mottling of leaves.  Coniferous  plants
 were more affected than deciduous species.
   Feder2 reported that geranium and  carnation plants had re-
 duced branching and retarded floral productivity after exposure
 to  low  levels  of  oxidant-type pollutants. Although  numerous
 plant taxa have been studies to determine the significance of
 various plant  morphological features as indicators of  environ-
 mental pollution,  relatively little work has been done to establish
 relationships between  cuticular dynamics and environmental
 pollution, especially hazardous waste sites. Recent studies7'8 on
 the subject suggest the potential of leaf cuticular dynamics in en-
 vironmental pollution research.
   The purpose of this study was to collect data for three plant
 species in  the relatively clean environment of Reelfoot Lake and
 to be compared to samples collected in the industrialized areas
 of Memphis and  Nashville. The present investigation is,  there-
 fore, a  continuation of a comprehensive study involving plant
 cuticular features as  indicators of environmental pollution with
 special reference to hazardous waste sites.

MATERIALS AND METHODS
  Plant  populations of Polygonum pensylvanicum (smartweed),
Platanus occidentalis (sycamore) and Catalpa bignonoides (catal-
pa) were studied; comparisons of samples taken from the relative-
ly unpolluted sites of Reelfoot  Lake and Martin in  northwest
Tennessee  were made with ones collected from the-contaminated
sites in  middle  and northwest Tennessee—the latter  sites were
characterized t>y exposure to  additional pollutants emitted by
vehicular traffic and the industrial complex of the  surrounding
metropolitan areas.
  Twenty-five randomly selected leaves  were gathered from each
plant species.  Their  lengths,  widths,  internodal  lengths and
length-width ratios were determined. Gross morphological meas-
                                                                                                     MONITORING    17

-------
urements were supplemented with cuticular data.
  For cuticular studies, representative leaves were washed with a
mild detergent and distilled water. After air-drying, the upper and
lower surfaces of the leaves were coated with Duco-cement™.10
Upon drying, fine layers of Duco-cement showing the cuticular or
epidermal complex of leaves were removed; slides were prepared
from the central portion of the layers for microscopic analysis.
Cuticular dynamics representing in stomatal frequency stomatal
size, epidermal  wall  undulation, trichome  frequency, trichome
length  and type  and  subsidiary cell  complex were recorded
(N = 25) for the  upper and  lower surfaces representing upper and
lower surfaces of leaves using 43x objective and lOx oculars.
  Dendrochronological studies were conducted on  the  woody
taxa by getting core samples in the two sets of habitats.  Annual
increments were recorded from the core samples. The entire data
were analyzed by a computer.

RESULTS
  Statistical analysis  of the data  revealed several significant dif-
ferences in the gross morphological, cuticular and dendrochron-
ological features of the two sets of samples (polluted versus rela-
tively unpolluted). Plants growing in the contaminated areas had
smaller leaves with both their length and width significantly re-
duced. In addition,  internodal length  also  was reduced. Plants
growing in the rural,  Reelfoot Lake area were healthier and had
larger leaves and longer internodes. These leaves were dark green,
while the leaves of plants from the contaminated sites had yellow-
ish,  light green  pigmentation.  It is  obvious that the photosyn-
thetic productivity of the plants at the contaminated sites was ad-
versely affected  by the  environmental pollution prevalent in  the
area.
  A reduction in the total  biomoass of plant populations in  the
contaminated  areas  was an additional evidence to  suggest  the
detrimental effects  of environmental  degradation on  plant
growth. A reduction  in metabolic activities, especially in  photo-
synthesis, must mean reduced bimoass; this  result was evident in
all the populations  of contaminated sites. Cuticular measure-
ments revealed additional features affected by the environmental
contamination of  the  sites under  investigation.  Stomatal fre-
quency values  of the lower leaf surfaces in all plants under study
were high in the rural, less polluted habitats, while the plants
from the contaminated sites exhibited extremely  low stomatal
frequency values.
  In addition, trichomes on the leaf surfaces of plants from  the
latter sites had more numerous and much longer trichomes with
large bases. Similar results have been found in earlier studies6'7
in which affected plant populations were growing in sites char-
acterized by heavy vehicular traffic.  Stomatal size  differences in
the two sets of samples were not significant. Subsidiary cell com-
plex remained  unaffected by environmental  contamination in all
three plant taxa, and hence must be regarded as a reliable feature
for identification purposes of the species.
  The fact that some cuticular  and morphological features  of
some plant species were different in the two habitats is indica-
tive  of the evolutionary  trend exhibited  by plants growing  in
areas exposed to various kinds of environmental contamination.
Monitoring of these plant features over a long  period of time is
needed. This  study is underway in order to determine if these
morphological and cuticular trends are permanent.  If so, these
and  similar plant species might be used as indicators of environ-
mental complex, especially at contaminated sites.
  Preliminary investigations of the core samples of woody taxa
revealed a general decrease in the size of annual  increment. Addi-
tional  detailed studies are  underway to determine the usefulness
of plant features as monitors of environmental  pollution,  espec-
ially al hazardous  waste sites. Such  investigations might  reveal
plant species that  are able to  withstand  specific contamination
and  hence  be  of significance for  environmental and commercial
considerations.

ACKNOWLEDGMENTS
  Financial support for this work under the Research Grant Fund
of The University  of Tennessee at Martin is gratefully acknowl-
edged.

REFERENCES
 I. Chamberlain, C.J., Gymnosperrru: Structure and Evolution, Dover,
   New York. NY. 1934.
 2. Feder, W.A., "Plan! Response to Chronic Exposure to Low Levels
   of Oxidant type Air Pollution," Environ. Pollut. I. 1970, 73-79.
 3. Hill, G.R.  and  Thomas,  M.D., "Influence of Leaf Destruction
   by Sulfur dioxide and Clipping on Yield of Alfalfa," Plant Physiol.
   8. 1933.334-345.
 4. Pyatt, B.F., "Lichens as Indicators of Air Pollution in a Steel pro-
   ducing Town in South Wales," Environ. Pollul. I, 1970, 45-55.
 5. Scheffer, T.C. and Hedgcock, G.C.,  "Injury to  Northwestern For-
   est  Trees by Sulfur dioxide from Smelters," U.S. Dept. Agr. Tech.
   Bull. 1117. 1955.
 6. Sharma, O.K. and Butler, J., "Leaf Cuticular Variation in Trifolium
   repens  L. as Indicators  of  Environmental  Pollution," Environ.
   Pollul. 5. 1972.287-293.
 7. Sharma, O.K.. Chandler,  C. and Salemi, L., "Environmental Pollu-
   tion and Leaf  Cuticular Variations  in  Kud/.u (Purrario  lobalo
   Willd.)," Ann. Botany. J5. 1980. 77-80.
 8. Sharma, O.K. and Tyree,  J., "Geographic Leaf Cuticular and Gross
   Morphological Variations in Liquidambar slyraciflua L. and their
   Possible Relationship to Environmental Pollution," The Botanical
   Gazette, 1.1-1, 1973, 179-184.
 9. Solbcrg. R.A. and Adams, D.F., "Histological Responses of some
   Plant  Leaves to  Hydrogen fluoride and Sulfur  dioxide," Am. J.
   Botany.  43. 1956, 755-60.
10. Williams, J.A.,  "A Considerably Improved Method  for Preparing
   Plastic Epidermal Imprints," The Botanical Gazette.  134,  1973,
   87-91.
18
      MONITORING

-------
                  Detailed Stratigraphic and  Structural  Control:
              The  Keys to Complete and  Successful Geophysical
                             Surveys  of  Hazardous Waste Sites
                                           H. Dan Harman, Jr., P.G.
                                            Engineering-Science, Inc.
                                                 Atlanta, Georgia
ABSTRACT
  Since the passage of RCRA and CERCLA legislation, count-
less geophysical surveys have been conducted at hazardous waste
sites as part of site investigations in an attempt to delineate the
horizontal and vertical extent of groundwater contamination as
well as to delineate the limits of the sites. Of utmost importance
during a geophysical survey is the identification of Stratigraphic
and structural control features  which influence the  potential
migration of leachate away from a site.
  Stratigraphic features such as  sand zones, clay lenses, outwash
valleys and top of rock zones are critical in  the overall hydro-
geological assessment of a site. Structural features such as faults
and fractures are equally critical in the  assessment of a site.
Proper use and interpretations of geophysical data can result in a
better understanding  of a site's subsurface characteristics even
before a drilling program begins.
  The geophysical data interpretations explained in this paper,
using  the "Modified  Wenner" method, are empirical  and have
resulted in the site-specific identification of subsurface details
which have been reasonably accurate as compared to subsequent
drilling results. The acquisition of detailed knowledge of a site's
subsurface is  the key to  a complete and successful geophysical
site survey.

INTRODUCTION
  Geophysical surveys consist of surveying a site and its immed-
iate vicinity utilizing one or more remote sensing techniques. The
techniques  may  include  earth  resistivity,  electromagnetics,
magnetometry, seismic, metal detection and ground penetrating
radar.  Each technique has its  own advantages and  disadvan-
tages  at a particular  site. When groundwater contamination  is
a suspected problem, either earth resistivity or electromagnetics
normally is chosen as the technique in a geophysical survey.
  Among the many methods utilized within the  earth  resistivity
technique, the "Modified Wenner" method1 has been reasonably
accurate in terms of  depth investigated as compared  with sub-
sequent drilling results.  By utilizing the "Modified  Wenner"
method and  empirical interpretations,  site-specific identifica-
tions of Stratigraphic and structural control features are possible.
The identification of these features  is the key to performing a
complete and successful geophysical survey of a hazardous waste
site.

STRATIGRAPHIC AND STRUCTURAL
CONTROL
  Stratigraphic and structural control is the understanding of a
site's  subsurface characteristics. When Stratigraphic and struc-
tural control is understood, a site's geological characteristic can
be understood. Furthermore, a site's hydrological characteristics
can be understood better in terms of aquifers, confining layers
and groundwater migration routes. Also, this  control enables
the investigator to more effectively perform  resistivity profiles
in an attempt to define the horizontal and vertical extent of iden-
tifiable contaminate plumes. By understanding the Stratigraphic
and structural control, the resistivity profile measurements can
be effectively placed in the appropriate subsurface zones in which
contaminant plumes are suspected.
  Stratigraphic  and  structural control is normally  established
only after an exploratory drilling program, but the control can
be established by the proper interpretation of resistivity sound-
ings. Soundings are  measurements of the earth's resistivity  at
various depths at a single land surface point.

RESISTIVITY SOUNDINGS
  Resistivity  soundings,  utilizing  the  "Modified  Wenner"
method and empirical interpretations, have resulted in  site-spe-
cific subsurface identifications which have correlated well with
existing and/or subsequent drilling data. The key factor in the
interpretations of "Modified Wenner" method soundings is that
the potential or inner electrode spacing across the land surface
very closely approximates the depth below the land surface  at
which a measurement is taken; or, in general terms, "electrode
spacing equals depth below ground."
    Formula lor Apparent Resistivity
                   Current Meter       Battery
                  	(T)	1 ill	
                         Volt Meter
                         	(v)	
                                      P'
    SOURCE' Carrtaglon A wmon. lost
                          Figure 1
                   "Modified Wenner" Array
                  Diagram of Electrode Spacing
                                                                                                    MONITORING    19

-------
  The electrode arrangement in the "Modified Wenner" method
is shown in Figure  1. In this arrangement, the current or outer
electrodes (C and C) are stationary while the potential or inner
electrodes (P and P') are moved  at equal distances from the
center of the array. The arrangement of the electrode array can
be varied depending  on the objective and depth to subsurface
targets.
  Numerous investigators have discussed the advantages and dis-
advantages of  empirical versus theoretical interpretations,  and
a discussion here will not be attempted. A partial list of refer-
ences  is included at the end of this paper.2"4 The author has ap-
plied the "Modified  Wenner" method and  empirical interpre-
tations  in  numerous subsurface investigations  because  of the
method's ease of operation and the relative simplicity of the inter-
pretations. The correlation between the sounding interpretations
and actual subsurface drilling data has been reasonably accurate.


EXAMPLE SOUNDINGS
  Five example soundings  are  presented to illustrate the appli-
cation of the "Modified Wenner" method and empirical inter-
pretations. The method has proven to be an asset in solving strati-
graphic and structural control problems, the  solutions of which
have  improved the effectiveness of subsequent resistivity profil-
ing, monitoring well  placements and groundwater contaminant
plume tracking.

SHALLOW AND DEEP SOUNDING
COMPARISONS
  Resistivity soundings at a landfill in the  mid-western  states
were  conducted to aid in the  understanding  of the stratigraphy
underlying the  landfill as well as determining the thickness of the
landfill cap. Figure 2 illustrates a shallow sounding to 5 ft below
ground. An  apparent resistivity change is evident at approxi-
mately a 2 ft depth which was interpreted as the thickness of the
landfill cap. Similar soundings over the landfill were conducted,
and the apparent resistivity responses varied only slightly (mainly
due to the variation of cap saturation and lithology). A limited
number of shallow borings on the landfill had confirmed the cap
thickness to be  approximately 2 ft deep. Resistivity profiling iden-
tified variations across the landfill  which aided in the determina-
tion of possible recharge occurring through the cap.
                          P-Pl SPACING  (FEETI
  A deep sounding to 100 ft over the landfill was conducted to
aid in understanding the  deep stratigraphy underlying the site.
Figure 3 illustrates this sounding showing interpreted features
such as the water table, sand/clay lenses, limestone rock and pos-
sible fractures in the rock.  The actual drilling logs from both land-
fill borings and a nearby water supply well are presented to show
the correlation  with  the interpretations. Note the depth correla-
tion of the apparent resistivity graph to the boring logs, especially
the deflection of the  graph at 2 ft (cap-fill interface) and between
90 and 96 ft which correlates with the 90-ft water-bearing zone
identified by the water well log. The deep sounding at this land-
fill site aided greatly in  the  understanding of the  site-specific
stratigraphic and structural control.
             \
 I  40
   24



   16
        -—3	j-sf—-I
I
        SI
                           «     SO    60

                           P-Pl V1CI-* !<•££']
                           Figure 3
                   Deeper Sounding at Land Till
SOUNDINGS IN HIGHLY CONDUCTIVE
GROUNDWATER
  Soundings were conducted downgradient of a leaking lagoon at
a site in the southeastern states. This site is underlain by crystal-
line  rocks. The groundwater downgradient of the site contains
highly conductive salts which created special problems during the
soundings. Figure 4 illustrates -a 60-ft  sounding conducted with
all four electrodes in the center of the highly conductive ground-
water plume.  The graph  is essentially  flat,  and interpretations
are impossible to make.
                                                                                             !4    JO    36

                                                                                             P-Pl SPACING tfEETI
                          Figure 2
           Results of Resistivity Soundings at a Landfill
                           Figure 4
        Results of Sounding Downgradient of a Leaking Lagoon
20
      MONITORING

-------
  Figure 5 illustrates a 200-ft sounding conducted with the cur-
rent electrodes placed outside the plume. This graph is more rep-
resentative of the subsurface than is the graph in Figure 4. Note
that the low apparent resistivity between 50 and 69 ft on Figure 5
is not seen on Figure 4. The low resistivity values at this zone
were confirmed  to represent contaminated groundwater within
weathered and fractured crystalline rock. The sounding in Figure
5 aided in the understanding of the  stratigraphy and structure
below a highly conductive groundwater plume.
  200  ,

  ISO

  ISO

  140

ui 120

I 100

a
i  eo
&
   60
   ao

    o
      0    20    10    60    30    100   120    140   1EO    160   200
                           P-P1 SPACING (FEET)

                            Figure 5
     Results of Sounding Near and Leaking L.agoon with Electrodes
                        Outside the Plume
CONCLUSIONS
  Understanding the stratigraphic and structural control under-
lying and in the vicinity of a waste site is critical in the overall
hydrogeological assessment of a site. Surface resistivity sound-
ings aid this understanding by yielding data which can be corre-
lated with existing and/or subsequent drilling results. Subsequent
geophysical surveys, resistivity  profiles for  example, can be
planned by the proper interpretation of the soundings in terms of
site-specific stratigraphy and structural features.
  1500


  1350


  1200


  1050


5  900
u.

§  750
in
£  500

|
   450


   300


   150


     0
                           P-P1 SPACING (FEET!
                            Figure 6
  Results of Soundings in Sedentary Rocks to Identify Solution Cavities
 SOUNDINGS IN SEDIMENTARY ROCK
 CONTAINING SOLUTION CAVITIES
   Soundings were conducted to various depths at a waste site in
 the southeastern states to aid in identifying solution cavities with-
 in sedimentary rock. The rock at this site is limestone with solu-
 tion cavities commonly occurring. The solution cavities are pos-
 sible avenues  of contaminant migration; therefore  the  under-
 standing of the subsurface structure at the site is critical. Figure 6
 illustrates one of the soundings conducted at the site.  The sound-
 ing graph correlates well with the actual drilling log. Note the
 low resistivity between 11 and 29 ft and the zone of staining and
 vugs present at 14.5,19.8 and 24.9 ft.
   As the degree of weathering in the rock decreases  with depth
 to 30.3  ft, the resistivity graph shows an  increase (more resistive
 rock). Yet, between 40.1 and 48.0 ft where a solution pitted zone
 and small vugs are present, the resistivity graph shows a corres-
 ponding decrease (more solution and weathering). The monitor-
 ing well subsequently installed  27  ft deep  at the sounding sta-
 tion  yielded groundwater  with higher conductivity  levels than
 background wells.  The resistivity sounding in Figure 6 aided in
 the understanding  of the solution  cavity distribution below the
 site.
   Application of the "Modified Wenner" method and empirical
 interpretations has resulted in reasonably accurate correlations
 between the interpretations and actual drilling data as well as
 groundwater quality data. The ability to predict the presence of
 subsurface features with reasonable accuracy is a real advantage
 and time-saving element in the decision-making processes con-
 cerning a hazardous waste site.

 REFERENCES
 1. Carrington, T.J. and Watson, D.A.,  "Preliminary Evaluation of an
   Alternate Electrode Array for Use in Shallow-Subsurface Electrical
   Resistivity Studies," Ground Water, Jan./Feb., 1, 1981.
 2. Moore, R.W., "An Empirical Method of Interpretation of Earth-
   Resistivity Measurements," American Institute of Mining and Metal-
   lurgical Engineering, Technical Publication No. 1743, July, 1944.
 3. Muskat,  M.,  "The Interpretation  of  Earth-Resistivity Measure-
   ments," American Institute of Mining and Metallurgical Engineering,
   164,224-231.
 4. Moore, R.W., "Geophysical Methods of Subsurface Exploration in
   Highway Construction," Public Roads, 26, Aug., 1950, 49-64.
                                                                                                               MONITORING     21

-------
                     Detection and Measurement  of Groundwater
                            Contamination by Soil-Gas Analysis
                                                   H.B. Kerfoot
                                                   J.A. Kohout
                                                    E.N. Amick
                      Lockheed Engineering and Management Services Company, Inc.
                                                Las Vegas, Nevada
 ABSTRACT
   A soil-gas sampling probe which can penetrate calcareously
 cemented alluvium was developed and evaluated above a chloro-
 form-contaminated groundwater plume. A linear correlation of
 greater than 95%  significance between groundwater and soil-gas
 chloroform  concentrations was observed. The precision of the
 method is controlled by sampling, with relative standard devia-
 tions between closely spaced samples ranging from 12 to 43%. A
 linear chloroform  depth profile was observed; the profile, as de-
 lineated by  the probe, agreed  with calculations made assuming
 diffusion-controlled  vertical flex of soil gases through the vadose
 zone.

 INTRODUCTION
   Contamination of groundwater is a problem of increasing con-
 cern. In efforts to detect and measure groundwater contamina-
 tion, groundwater sampling and analysis is the traditional method
 of choice. Recently, preliminary surveys using remote-detection
 methods have been used as a tool for planning more cost-effective
 groundwater monitoring networks.' These surveys typically have
 been made  using  geophysical  instruments such as electromag-
 netic techniques that can be effective for locating inorganic spe-
 cies  but are  not useful in detecting organic compounds. For the
 detection and measurement of subsurface contamination by vola-
 tile organic compounds (VOCs), a new technology, soil-gas meas-
 urement, has been  developed.
   Soil-gas surveying originally was developed for oil exploration.1
 Applications of the technique to delineate subsurface contamina-
 tion  by VOCs have been developed by several workers.3'4
   Measurement of soil gases for detection of groundwater con-
 tamination takes advantage of Henry's Law, which states that
 the concentration  of a volatile compound in vapors that are at
 equilibrium with a VOC solution is directly proportional to the
 VOC concentration in solution. The relationship between these
 two concentrations is quantitatively described by the Henry's Law
 constant for that compound. The magnitude of this constant is
 directly proportional to the vapor pressure  of the compound and
 inversely proportional to the compound's solubility in water.
  Because of their  relatively low water solubilities and high vapor
pressures, VOCs in contaminated groundwater tend to be present
 in soil gases above the source. Table 1 lists the 17 substances most
 frequently encountered at Superfund  sites and their Henry's Law
constants; of these chemicals, 10 are amenable to soil-gas survey-
ing. In addition, major components of petroleum products which
can leak from underground storage tanks can be detected using
this technique.'
  This paper describes a study undertaken to validate equipment
and procedures for soil-gas surveying. The validation of the sys-
tem involved assessing  the bias of the technique for  indicating
chloroform contamination of groundwater, as well as assessing
the precision and short-range variability of the method.
                         Table 1
    Most Frequently Identified Substances at 546 Superfund Sites*
               Sub«C4nc«
                                Henry*e L*" Constant  Percent
                                    (ppbv L/y9)      of aitea
   1   Trichloro«thyl«n*


   2   Le«d and compound*

   3   Tolu«n«


   4   B«nz«ne

   5   Polychlorinated blphcnyl* (PCB*)

   6   Chloroform


   7   Tctr*chloro«thylene
 40

133
              30

              28b

              26»>

              22

              20 b
8
9
10
11
12
1]
14
IS
16
17
Phenol
Araenlc and coapounde
Cadpdun and coe^xMinda
Chroadua and coapounda
1,1. 1-Trlchloro«chane
Zinc and coag>ounda
Bthy Ibeniene
«ylena
Hethylene chloride
trana- 1,2-dlchloroethylene
« 1
NA
HA
HA
30
NA
59
4]
21
580
15
IS
IS
IS
14b
14
Ub
Ub
12b
lib
"Source: Kcrfoot, H.B. and Barro*\, L.J.. "Soil Gas Measurement for ihc Detection of Sub-
surface Organic Contamination," U.S EPA, l.*u Vegas, NV, 1986,
 Compound amenable 10 soil-gas surveying.
EXPERIMENTAL DESIGN
  The objectives of the study were to assess the bias and the pre-
cision of this newly developed technique for detection and meas-
urement of groundwater contamination by VOCs and to study the
vertical distribution of VOCs in the vadose zone. The study loca-
tion has a known chloroform-contaminated groundwater plume
22   MONITORING

-------
               Pittman
                           Figure 1
             Survey Site Location, Henderson, Nevada
and an existing system of groundwater monitoring wells.  These
monitoring wells are separated by 200 ft and lie along a line per-
pendicular to the direction of groundwater flow.
  For evaluation of the method bias, soil-gas samples were taken
at a 4-ft depth at four locations 20 ft to the north, east, south and
west of  four monitoring wells  within  the boundaries of the
groundwater plume. To delineate the edges of the plume, addi-
tional samples were taken at a 4-ft depth 20 ft to the east of a well
at the western edge of the groundwater plume and 20 ft to the
                                                                            END CAP
                                                                         1/8" TUBING
                                                                       3/4" THREADS
                                                                       TUBING NUT
                                                                      1/2" THREADS
                                                   Figure 3
                                             Sampling Probe Design
                                      Benzene/Chlorobenzene
LEGEND
   Well Location
   Water Table
   Sand and Gravel
   Clay
                                                                             Chloroform
Elevation (feet)
1660—,
1640 —
1620 —
1600
1580 —
o
o
~^^~ T— — • -r *
^ [ -] -j- •" 	 1 — r--T — '
14
655 650 645 640
West

•Sand 8. !
Gravel j
9' deep ! \
,,-S_ ~
^^^^ " .i
o
i ! I 1 ! i i i**^~
-_J 	 H 	 1 	 IWate Tabe 	 l__}-'l,^""
j(^^~TT^
—
635 630 625 620 615 610
Stations Teit Well Eagt
0       600
 Scala in Feat
0          200
 Scale in Meter*
                                                             Figure 2
                                            Subsurface Hydrogeology at the Pittman Lateral
                                                                                                            MONITORING    23

-------
west of a well on the eastern edge. Samples were taken in triplicate
and were analyzed by gas chromatography on site. The triplicate
analyses were used to assess the method precision.
  At one sampling location, four points separated by 3 ft were
sampled as a check on the short-range variability of the method.
Halfway between two of the wells, three points separated by 3 ft
were sampled in duplicate at depths of 1, 2, 3,4 and 5 ft.

STUDY LOCATION
  The geohydrology  of the study area  is relatively simple. Un-
confined groundwater occurs at a depth of 7 to 15 ft in calcified,
unconsolidated alluvium overlying a clay acquiclude. The ground
surface, water table and acquiclude all slope downward about 1
degree to the north, and the groundwater moves northward about
1.5 ft/day. Figure 1 shows the site location, and Figure 2 shows
the geohydrologic profile along the line of our survey.

EXPERIMENTAL
  A shallow probe was used  to obtain soil-gas samples from a
depth of 4 ft. The probe consists of an outer pipe constructed of
3/4-'m. o.d. high-strength steel that ends in a tapered head with six
horizontal sampling ports, each 1/8  in. in diameter. A  1/8-in.
o.d. stainless-steel tube is connected  to the sampling ports and
runs through the pipe to  the sampling  manifold. The sampling
manifold is assembled from commercially available stainless-steel
fittings. Figures 3 and 4 show the probe design. Figure 5 shows
the sampling manifold.
  After insertion into the soil, the sampling manifold was purged
with soil gas.  Samples were then  withdrawn from the manifold
using gaslight syringes and were transported to a mobile labora-
tory.
  Soil-gas  samples were analyzed on-site  using an  Analytical
Instruments Development Model 511  gas chromatograph with a
3H electron-capture detector. The gas chromatograph was oper-
ated at 43 °C, and the output was processed using a Shimadzu C-
R3A integrator. Calibration standards were prepared by serial
dilutions of chloroform headspace vapors.
  Figure 6 shows the  soil-gas sampling locations relative to the
monitoring wells. Groundwater samples were taken and analyzed
in April and August of 1985. The samples were analyzed by the
purge-and-trap  gas chromatography/mass  spectrometry  (GC/
MS) method specified in the U.S. EPA Contractor Laboratory
Program.'

RESULTS
  The bias of the method was evaluated by comparing the mean
soil-gas chloroform concentration measured to the results of the
groundwater analyses.  The soil-gas chloroform concentrations
correlate strongly with the groundwater concentrations. Table 2
lists the soil-gas and groundwater chloroform concentrations at
each well. Figure 7 is a plot of these data. A linear regression of
the soil-gas concentrations on the groundwater concentrations in-
dicates a correlation of better than 95% significance (r  =  0.85,
n = 6).' In addition, the spatial delineation of the chloroform
groundwater plume, as shown in Figure 8, is very good.
  The mean chloroform concentration from the four locations
around each well was used in the above calculations; however, the
relative standard deviation (RSD) of these mean values was often
quite high (above 100%). We interpret this result to be an effect
of spatial variability in  the soil-gas chloroform concentrations
over the  20- to  40-ft lateral distances separating these sampling
points.
                                                                                 'x/s */",?./"•:*!	i.
                                                                                 D      vl	-...
                                      SHAFT *4130 COLO DRAWN, CHROM-MOLY. CONDITION N STEEL
                          END CAP *400 CHROM-MOLY
                           "'i r
                                                                             -I -«IO	
                                                                       INNER TUBING
                                                          Figure 4
                                                Probe Construction Specifications
24    MONITORING

-------
                                                                                                 Table 2
                                                                             Groundwater and Soil-Gas Chloroform Concentrations
                            Figure 5
                       Sampling Manifold
                                                                                              Soil-gaa concentration (ppbv)a
                                                                              Ground-water
                                                                              concentration  20  ft.    20 ft.    20 ft.     20 ft.
                                                                         Well     (U9/L)       West     North     East      South
                                                                  631

                                                                  629

                                                                  627


                                                                  625
                                                                                   11

                                                                                  175
                                                                                                   25(2)
                                                                                                             5(0)

                                                                                                            27(5)
                                                                                                        5

                                                                                                       23
                                                                                           28(5)   72.9(0.1)  124(53)c   45.6(0.2)d   67
                                                                                           (WSW)     (WNW)      (ENE)   (25 ft.  SSE)
                                                                                           266(6)   326(10)
                                                                                                            376(6)
100 ft.
E of
625
623
621


150
555 115(6) 12(5) 6(3) 27(2) 40
28 10.5(0.3) 10.5
                                                                      triplicate analyses; standard deviation in parentheses.
                                                                      Not detected; 5 ng/1 used in regression.
                                                                      ^Mean of four closely spaced points (see text).
                                                                      Duplicate determinations.
                                                                 The effect of short-range geologic variability on analytical re-
                                                              sults was assessed in  two locations. At one location,  four sam-
                                                              ples were taken at a depth of 4 ft, at separations of between 3 and
                                                              7 ft. Figure 6 shows the pattern of the sampling points. The rsd
                                                              value of the chloroform concentrations among these points was
                                                              42%. At  another location, three points 3 ft apart along a north-
                                                              south line were sampled at five depths in 1-ft increments between
                                                              1 ft and 5 ft. The mean rsd value of the mean chloroform con-
                                                              centration for all  five depths  at the  three locations was 12%.
                                                              These results indicate the short-range  geologic variability can be
                                                              a major factor in soil-gas  surveys and that the magnitude of this
                                                              factor can vary.
  The precision of the method was evaluated by analysis of mul-
tiple samples.  The analytical  precision, based on multiple daily
analyses of calibration standards, showed an rsd value below 4%.
The combined sampling/analysis precision,  based on triplicate
samples from each location sampled at a 4-ft depth, was charac-
terized by a mean rsd of 12 %.
 631

  O
     629
627
625
623
    N
       |—200ft.—|

     O Well Location
      • LGAS Probe Location
                   Figure 6
Soil-Gas Sampling Points (Sampling Depth  = 4 feet)
                                                                      S   100-
                                                              a 1
                                                              "o "H.
                                                              o -^
                                                                                                            100
                                                            Log~Ground Water Chloroform Concentration
                                                                         lug/L)
                                                                                                Figure 7
                                                                               Plot of-Chloroform-Soil-Gas Concentration and
                                                                                        Groundwater Concentration
                                                                                                                           1000
                                                                                                                MONITORING     25

-------
                                                   Ground-Watp r
                                                   Concent rat ion
            631     629      627      625      623     621      t, I0
                                                                                                Table 3
                                                                                    Chloroform Concentration Vertical
                    629      627     62i      62]     621      619
                           Figure 8
          Spatial Distribution of Chloroform in Soil-Gas and
                     Groundwater Samples


   The results of an evaluation of the chloroform concentration
 vertical profile (Table 3) showed a linear dependence of concen-
 tration upon sampling depth. A correlation coefficient indicating
 greater than 99% significance was obtained (r =  0.999, n  =  5).
 This result is in agreement with a model proposed  by Swallow
 and  Gschwend'  that  attributes  vertical  transport of  VOCs
 through the vadose zone to gaseous diffusion. The water table
 below the sampling locations  was at a depth of approximately
 13ft.

 CONCLUSIONS
   Soil-gas surveying accurately indicated groundwater chloro-
 form contamination at a site in Pittman, Nevada. The precision
 of the technique is controlled by sampling and short-range geo-
 logic variabilities. At  the  site  studies,  results of a  depth study
 agreed along with a model for vertical transport of volatile organ-
 ic compounds through the vadose zone by gaseous diffusion.

 ACKNOWLEDGEMENTS
   The authors would like to acknowledge contributions made by
 J.W. Curtis, K.L. Ekstrom, and L.J. Barrows to this study.
   Although this work was supported in part  by the U.S. EPA, it
 has not undergone review by  that agency and does not reflect
 agency policy.

 REFERENCES
 1.  Walther, E.G., LaBrecque, D.J., Weber, D.D., Evans, R.B. and van
   Ee, J.J.,  "Study  of  Subsurface Contamination  with  Geophysical
   Monitoring Methods  at  Henderson,  Nevada," Proc.  Fourth  Na-
   tional Conference on Management of Uncontrolled Hazardous Waste
   Sites, 1983, Washington, DC, 28-36.
Location
base
3 ft. N
6 ft. N
base
3 ft. N
r, ft. N
base
3 ft. N
6 ft. N
base
3 ft. N
6 ft. N
6 ft. N
base
3 ft. N
base
Depth
1
1
1
2
2
2
3
3
3
4
4
4
5
5
5
6
Mean chloroform
(ppbv) (SOa)
23.0 (0.1)
22.9 (0.8)
19 (1)b
76 (1)
70 (2)b
58 (3)
109 (1)
111 (1)
99 (14)*>
153 (9)
149 (1)
132 (14)
205 (2)b
183 (8)
167 (23)
236 (29)
 Duplicate measurements unless noted
 triplicate measurements
2. Horowitz, L., "Geochemical Exploration for Petroleum " Science
   229,(1985), 821-827.
3. Marrin, D.L. and Thompson, G.M.,  "Investigation in the Unsatur-
   ated Zone Above TCE Polluted Groundwater."  U.S.  EPA, Ada,
   OK, U.S. EPA Project Number CR811018-01-0.
4. Voorhees, K.J., Hickey,  J.C. and Klusman,  R.W., "Analysis  of
   Groundwater Contamination by a New Surface Static Trapping/Mass
   Spectrometry Technique," AnalyticalChem.. 56. 1984, 2604-2607.
5. LaBrecque, D.J., Pieretl, S.L., Baker, A.T., Scholl, J.F. and Hess,
   J.W.,  "Hydrocarbon  Plume Detection  at Stovepipe Wells, Cali-
   fornia," U.S. EPA, Las Vegas, NV, 1985.
6. McGhee. J.W., Introductory Statistics. West Publishing New York
   NY, 1985.

7. U.S. EPA, "Chemical Analytical Services for  Multi-Media Multi-
   Concentration Organics  GC/MS  Techniques," WA-85J680 US
   EPA, Washington, DC, 1985.

8. Swallow, J.A.  and  Gschwend, P.M.,  "Volatilization  of  Organic
   Compounds from Unconfined Aquifers," Proc.  National Symposium
   on Aquifer  Restoration  and  Ground water Monitoring^ National
   Water Well Association, 1983, 327-333.
26    MONITORING

-------
                         Environmental  Appraisals and Audits:
                            A Case  Study of Their  Application

                                               Anthony R. Morrell
                                          U.S. Department of Energy
                                       Bonneville Power Administration
                                                Portland, Oregon
ABSTRACT
  The Bonneville Power Administration (BPA) has had an envir-
onmental audit and appraisal program for 2 years. BPA's unusual
approach in implementing its appraisal program resulted in the
program's quick acceptance and organizational effectiveness.
  The first audits conducted under the appraisal program iden-
tified problems  in the use of polychlorinated biphenyl (PCB)
equipment at BPA substations. Subsequent U.S. EPA inspections
confirmed these same findings. In response to this situation, the
U.S. EPA and BPA entered into a Memorandum of Agreement
(MOA) specifying corrective actions to be taken.
  The U.S. EPA, recognizing the strength of BPA's appraisal
program, agreed to  forego their normal ad hoc  inspections of
BPA facilities. Instead, the U.S. EPA deferred to BPA's own
audits with U.S. EPA inspections limited to followup or verifica-
tion inspections.
  In addition to ensuring compliance with environmental regula-
tions, the Agreement's reliance upon BPA's appraisal program
maximizes the effective use of limited staff resources in both
agencies. The Agreement interjects a  large measure of predic-
tability and  manageability into  BPA  efforts to achieve  com-
pliance.

INTRODUCTION
  As a power marketing agency within the U.S. Department of
Energy (DOE), the Bonneville Power Administration (BPA) sells
and transmits the electrical output from 30 federal hydroelectric
dams in the Columbia River Basin. The BPA service area includes
300,000  square miles primarily in the  states of Oregon, Wash-
ington, Idaho and Montana, with small service areas in Califor-
nia, Nevada, Utah and Wyoming. The  BPA transmission system
serves as the backbone  for the  interconnected utilities in the
Pacific Northwest and is connected with 18 other transmission
systems at over 100 locations. The BPA transmission system con-
sists of approximately 14,200 circuit miles of  high  voltage
transmission lines and about 400 substations. Approximately 100
of the BPA substations contain polychlorinated biphenyl (PCB)
equipment, primarily PCB capacitors. A total of about 140,000
PCB capacitors are in service at these substations.
  In 1983 BPA instituted an environmental appraisal  and  audit
program. Upon adoption of the program, BPA decided as a mat-
ter  of  policy to  use  its appraisal and  audit program  as  a
mechanism to ensure that BPA facilities comply with applicable
environmental standards.

BPA APPRAISAL AND AUDIT PROGRAM
  Equal to the importance of its adoption is the special approach
BPA used in implementing its new program. BPA is convinced
that, given its own organizational culture, any other implementa-
tion  approach  would not have been as  readily accepted and
therefore would have been much less successful in achieving the
objectives of the program.

   BPA appraisal program objectives include the following:

 • Assure that  DOE environmental policy and requirements are
   appropriately interpreted and implemented by BPA and BPA
   contractors.
 • Help line managers achieve BPA's compliance commitments.
 • Increase employees' awareness  of environmental regulations
   and BPA's commitment to compliance.
 • Provide management with objective, timely and reliable in-
   formation on  BPA and BPA-contractor  environmental per-
   formance, including significant achievements and efficiencies.
 • Evaluate the effectiveness and efficiency of BPA's implemen-
   tation of measures to avoid, minimize,  rectify or otherwise re-
   duce adverse impacts to the environment and measures to com-
   pensate for impacts.
 • Provide management with recommendations for improvement
   of BPA's  environmental program performance.
 • Develop and recommend long-term solutions to current  en-
   vironmental  problems in anticipation of future standards or
   conditions.
 • Evaluate the accuracy  of  environmental  analysis or  impact
   predictions and identify methods for improvement.

   These objectives are relatively standard. What distinguishes the
 BPA program  is the unusual approach taken in its implementa-
 tion. Without  compromising the objectives of the program, the
 approach taken was a positive one designed to find the facilities
 audited in compliance. As a practical matter, this simply meant
 that every effort was made  to communicate, beforehand, what
 was required and what would be audited. This was accomplished
 by providing advance copies of checklists to be used during the
 audit and by meeting with the affected organizations prior to their
 actual audits.  This approach was very well received by the af-
 fected field organizations. Given BPA's functional organization
 and culture, any other approach would have met with resistance.
 Other factors which ensured successful implementation were pro-
 visions to: (1) resolve problems at the lowest organizational level;
 (2) include representatives of affected offices on each appraisal
 team;  and (3) resolve smaller problems in the field (i.e., labeling
 equipment), so that they need not be mentioned in the written
 audit reports. Taken together, these provisions were viewed as a
 sincere attempt by central headquarters  staff to assist the field
 facilities to come into compliance without unnecessarily embar-
 rassing them organizationally.
                                                                                                    MONITORING    27

-------
         PHASE 1
     PRE-PLANNING ACTIVITIES
  PHASE 2
FIELD APPRAISAL
     PHASE 3
POST-APPRAISAL ACTIVITIES
                              Figure 1
                 BPA Environmental Appraisal Process
 DEVELOPMENT OF THE APPRAISAL PROGRAM
   A useful strategy in the initial development  of the Environ-
 mental  Appraisal Program was to seek management consensus
 regarding both the scope of the program and the priorities of sites
 to be appraised.
   To begin the program, meetings were  held with  all affected
 managers. The objectives and  purposes of the appraisal program
 were discussed.  As a result of these discussions, it was agreed that
 BPA's  appraisals  initially  would  focus  on  operations  and
 maintenance activities.
   It was also agreed that the first year's program should examine
 the requirements of those environmental  laws and  regulations
 dealing  with  oil  spill prevention,  hazardous and toxic waste
 management, noise  pollution control, safe drinking water protec-
 tion and  commitments made  to the  public through documents
 prepared to comply with the National Environmental Policy Act.
   After a schedule was developed for the first year and appraisal
 teams were identified, training  was conducted for those  who
 would  participate  in  the program.   BPA  sought  contractor
 assistance in conducting  the initial training and obtained the ser-
 vices of a nationally recognized firm that had published informa-
 tion in the area of environmental auditing.
   Through the  training,  the teams  obtained  skills  regarding the
 use of checklists and how to document audit findings in a report.
 Also, environmental  staff  members  contacted  other  electric
 utilities who had completed environmental  audits or appraisals in-
 order to  utilize  their  experience in  performing environmental
 audits on facilities associated with the electric utility  industry.
   In the program's  first  2 years, 23 field appraisals were  con-
 ducted.  Each of these appraisals required 1 -2 days of field work at
 the  site  followed by the  preparation of  a written  report;  the
 equivalent of approximately four full-time positions have been re-
 quired to  implement the  appraisal program.
   Figure 1 illustrates the basic steps used  in conducting  the ap-
 praisal process.


STRUCTURE OF THE APPRAISAL  PROGRAM
  The BPA appraisal program contains three components. Each
component has  a different focus (site versus program)  or em-
                             Figure 2
           BPA Audit/Appraisal Program TSCA Checklist

                Polychlorlniled Blphenyli (PCBi) In Service

Definitions:  PCBs in service al BPA facilities may  fall into any of the following
categories: (I) non-PCB equipment, where the level  of PCBs is below 50 ppm; (2)
PCB-contaminaled equipment, levels of JO to 499 ppm PCBs; and (3) PCB equip-
ment with levels above 500 ppm. This checklist addresses all PCB articles in service,
whatever the distinction in PCB levels. If appropriate, remarks regarding the levels
of PCBs should be made in the Remarks Column or on the back of the page.

Nole 1: Definition of Transformer: All transformers (e.g., power transformers,
potential transformers, current transformers) are defined as any piece of equipment
which contains the description "transformer" on the manufacturer's nameplate.
  a)  The transformers that have no PCB ppm  tag, and which have the word "oil"
     marked on the manufacturer's nameplate, arc assumed to be 50-499 ppm PCB
     until these are lagged with the actual PCB ppm level.
  b)  The transformers that have no PCB ppm tag.  which have markings such as
     "liquid-filled" or "contains dielectric" (other than the word "oil") or do not
     give any indication of the content of the transformer, are assumed to contain
     more than 500 ppm PCB until these are tagged with the actual ppm level.
                                                                         Yea  No
                                                                         Yes  No
                                                                         Yes  No
                                            Type
                                                       0.  Does the facility have any equipment which contains any levels
                                                       of PCBs (cither in service or in storage)?
                                                       I   PCB-Conlaining Equipment (Transformers, Circuit Breakers.
                                                          Reactors. Reclosers. etc.)
                                                          1.1  Has this facility ever conducted an inventory of its equip-
                                                             ment \thich contains PCBi (or an inventory of all equipment
                                                             including pieces containing PCBs)?
                                                             I.I.I If yes, when was the inventory conducted? (dale of
                                                                  pnntoul)	
                                                             1.1.2 If yes, is the inventory complete? (Check each piece of
                                                                 equipment to see if it corresponds to the inventory.)
                                                             1.1.3 Describe each type of RGB-containing equipment,
                                                                 quantity, status. PCB concentrations and volumes (or
                                                                 include inventory list as working paper):
                                         Quinllly
                                                                          Stilus
                                                                                     PCB Cone.    PCB Volume
                                            Yes  No
                                            Yes  No
                                                         1.2  Are all pieces of equipment marked with an appropriate
                                                             PCB ppm level tag and or the appropriate PCB label (ML)?
                                                             1.2.1  If no. list the ones that are not marked.
                                                         1.3  Are any pieces of equipment leaking? (Identify)
                                                             1.3 I  What action is being taken to control these leaks?
                                                                  Describe:	
                                            Note 1: Action to contain and cleanup leaks in 500+ ppm PCB Transformers must
                                            be initiated within 48 hours of having knowledge of the leak. All leaks from any
                                            (ransformer or other piece of equipment should be contained and cleaned as soon as
                                            possible.
                                           Yes  No
                                           Yes  No
                                           Yes  No
                                           Yrs  No
                                      1.4  Are all pieces of equipment inspected?
                                          1.4.1  At » hat frequency?	
                                      1.5  Are inspection records kept?
                                               Where?	
                                           Yes
                                           Yes
                             No

                             No
                                           Yes  No
                 I.S.I Arc operators' inspection records maintained for three
                     (3) years after equipment disposal?
                     Where? . _		
             1.6  Arc pieces of equipment stored for use (spares) listed on the
                 inventory?
                 I .ft. 1 Arc these spares inspected regularly as described above?
             1.7  Have any pieces of equipment been serviced or repaired
                 (rebuilt) at the facility?
                 1.7.1 Arc maintenance records for these pieces of equipment
                     kept? Where	
28
       MONITORING

-------
Figure 2 (continued)

Yes  No       1.8  Does the facility ever top-off or refill any existing pieces of
                  equipment?
                  1.8.1 What materials are used? (include PCB ppm levels)
Note 3: No new oil will be procured with over 2 ppm PCB level.
Yes  No

Yes  No




Yes  No



Yes  No




Group
  1.9 Have any pieces of equipment been removed from service
      at this facility?
      1.9.1 Are records kept on the disposition of these pieces of
           equipment? Where?	
2. Capacitors
  2.1 Has this facility ever conducted an inventory of capacitors?
      2.1.1 If yes,  when was the inventory conducted? (date  of
           printout)	
      2.1.2 If yes, is the inventory complete?
      2.1.3 Describe each type (or group) of capacitor, quantity,
           status, PCB concentrations and volumes (or include
           inventory list as working paper):
     Quantity
                               Status
                                          PCB Cone.    PCB Volume
 Yes  No       2.2  Are all capacitors marked with an appropriate PCB label
                       (ML) or in an appropriately marked area?
                  2.2.1  If no, list the ones that are not marked.
Yes  No
Yes  No
Yes  No
Yes  No


Yes  No


Yes  No

Yes  No
Yes  No

Yes  No
Yes  No
Yes  No
                                                                         Yes  No
                                                                           2.4 Is there a designated area for capacitors stored for use?
                                                                               2.4.1 Is this area appropriately marked?
                                                                           2.5 Are all capacitors inspected?
                                                                               2.5.1 At what frequency?	
                                                                                       2.6  Where are inspection records kept?_
      2.6.1 Are operators' inspection records maintained for three
           (3) years after equipment disposal?
           Where?	
  2.7 Are capacitors stored for use (spares) listed on the
      inventory?
      2.7.1 Are these spares inspected regularly as described above?
  2.8 Have any capacitors been removed from service at this
      facility?
      2.8.1 Are records kept on the disposition of these capacitors?
           Where?	

3. General Procedures
  3.1 Has or is a PCB analysis done before any oil handling?
  3.2 How are solvents, filters or rags disposed of if used on
                 equipment of unknown PCB level?
                                                                                       3.3
                                                                                          Does the facility have the SPIFs on handling, inspecting,
                                                                                          storing PCBs and PCB equipment?
                                                                                          3.3.1 Where are they kept?	
                                                                                           3.3.2 List the SPIFS found at the facility:
                                                              Yes  No
           4. Other Equipment Containing PCBs (Light Ballasts, Relays, etc.)
             4.1  Does the facility have any other equipment containing
                 PCBs? List:	
 Yes  No
              2.3  Are any capacitors leaking?
                  2.3.1  How many? 	
                  2.3.2 What action is being taken to control these leaks?
                       Describe 	
 Note 4: Action to contain and cleanup leaks in PCB capacitors must be initiated
 within 48 hours of having knowledge of the leak.
                                                              Yes  No      4.2 Has any of this other equipment leaked?
                                                              Yes  No      4.3 Was cleanup initiated?
                                                                           4.4 Describe how the material was contained, cleaned up and
                                                                               disposed of:  	
                                                              Yes  No      4.5  Has any of this other equipment been removed from service,
                                                                               but retained for backup at this facility?
phasis and each  is the responsibility of a different organization
(field personnel versus central staff). These three components are
listed below.

Functional Appraisal
   The functional appraisal is an evaluation of the agency's overall
performance based upon a compilation of findings obtained from
the individual  field appraisals. This appraisal is  conducted an-
nually by BPA's Environmental Manager's office. The functional
appraisal focuses its attention on major or  repetitive problems
identified  in the field appraisals. In addition to apprising  top
management of its findings,  the functional appraisal is used as a
means for securing management support for  implementing those
recommendations that require an organizational commitment of
resources.

Field Appraisal
   The field appraisal is a documented, on-site appraisal of pro-
gram effectiveness for specific disciplines, namely  hazardous and
toxic waste management,  oil spill  containment,  noise and safe
                                                             drinking water regulations.  Following the field appraisal, a writ-
                                                             ten action plan is prepared. The action plan outlines an agreed
                                                             upon course of action to implement recommendations designed to
                                                             correct problems identified in the field appraisal. The action plan
                                                             is prepared within 60 days after completion of the field appraisal.

                                                             Internal Audit
                                                                A self-audit is a less formal evaluation, conducted by the  line
                                                             organization on a particular part of its environmental program.
                                                             These audits are scheduled  on an as needed basis.
                                                                The three types of appraisals described above permit appraisals
                                                             to be specific to a particular audience or level within the organiza-
                                                             tion and to focus upon issues of greatest concern to that level.

                                                             MEMORANDUM OF AGREEMENT
                                                             WITH THE U.S. EPA
                                                                In February 1985, the BPA and the  U.S. EPA entered into a
                                                             Memorandum of Agreement (MOA)  to clarify each  agency's
                                                             responsibilities and commitments for conducting actions required
                                                             and authorized by TSCA and CERCLA.
                                                                                                                      MONITORING
                                                                                                                                         29

-------
   Among other things, this Agreement specifically provides that
 BPA will audit its field installations for compliance with TSCA
 PCB regulations. The Agreement requires that  these audits be
 done on a scheduled basis and that they be administered by cen-
 tral headquarters staff. Any deviations from the PCB regulations
 that are found in the conduct  of these  audits are  to be
 documented, and a compliance plan is to be included in the audit
 report. All reports are forwarded to the U.S. EPA. A copy of the
 PCB audit checklist, which  reflects U.S. EPA Region  10's inter-
 pretation of the TSCA regulations, is included in Figure 2.
   With the Agreement in place, the U.S. EPA agreed to forego
 ad hoc inspections of BPA  facilities. Instead, U.S. EPA inspec-
 tions are limited to verification inspections to be  conducted only
 after the facilities have been audited by BPA.
CONCLUSIONS
  The utility of BPA's environmental appraisal and audit pro-
gram was evident even before the Agreement with the U.S. EPA,
but its usefulness to the agency was certainly increased when the
U.S. EPA agreed to use the appraisal program as a substitute for
its ad hoc inspections of BPA facilities.
  From the U.S. EPA standpoint, this ensured not  only that
potential violations would be corrected, but that the underlying
causes would be corrected as well. From BPA's standpoint, rather
than reacting to the U.S. EPA's inspections, this  arrangement
resulted in an opportunity to prioritize corrective actions at BPA
facilities based on environmental and operational considerations.
Consequently, a large measure of predictability and manageabil-
ity was introduced into what could have been a counterproductive
and possibly adversarial relationship with the U.S. EPA.
30
      MONITORING

-------
              Chlorinated  Organics and Hydrochloride  Emissions
              Sampling  from  a Municipal  Solid Waste  Incinerator
                                               Thomas A- Driscoll
                                                  James P. Barta
                                                 Henry J. Krauss
                                              David H. Carmichael
                                                 J. Maxine Jenks
                                            Texas Air Control Board
                                               Waxahachie, Texas
INTRODUCTION
  As the use of solid waste incinerators in Texas increases, so do
the questions regarding the environmental impact of their emis-
sions. Are the incinerators creating or releasing potentially toxic
materials from the waste?
  Currently, there are eight municipal incinerators operating in
Texas, and permit applications are pending for several more. The
purpose of this paper is to report the levels of chlorinated organic,
chlorine and chloride measured during the stack  sampling con-
ducted at the City of Waxahachie municipal waste incinerator.
The sampling was performed in August 1985, by  staff from the
Texas Air Control Board (TACB) Sampling and  Analysis  Divi-
sion. Three sets of 3-hour samples were collected.
  The results of this sampling will be reviewed to determine
whether or not  additional preventive measures are needed to
reduce the emission of toxic chlorinated organic compounds or
hydrochloric acid. The hydrochloric acid emissions generally are
thought to result from the burning of plastics in the refuse.1 The
chlorinated organic compounds (i.e., polychlorinated biphenyls
(PCBs), polychlorinated dibenzodioxins (PCDDs) and polychlor-
inated dibenzofurans (PCDFs) may originate as contaminants in
the refuse or as byproducts of the combustion of precursors in the
incinerator.2 In addition, these data will be used by the Permits
Division of the TACB to develop emission factors to be used in
permitting new incinerators.
  The stack sampling procedure was a modified version of U.S.
EPA Reference Method Five. The sample is collected  by XAD-2,
florisil, alkaline arsenite solution and glass fiber filter media.

RESULTS
  At this time, analysis of all the samples collected is not  com-
plete. Results of the hydrochloric acid testing show the emissions
to be on the order of 2 Ib/hr. A preliminary list of compounds
detected (but not quantified) is presented in this paper. The com-
pounds that were detected are listed below with some information
about their  use, toxicity and exposure limits in air:
• Dichlorobenzene—used as a process solvent and  as an in-
  termediate in the synthesis of dyestuffs and herbicides;3 causes
  hemolytic anemia and liver necrosis in humans;  can also cause
  irritation  of eyes and nose; the OSHA occupational exposure
  limit (OEL) has been set at 50 ppm upper limit.4 The TACB
  Health Effects staff uses  1 % of the OEL as a guideline for per-
  mit and health effects review.
• Hepta and hexachlorodibenzofurans—similar properties  and
  health effects to dioxins of similar molecular weight. The On-
  tario Ministry of Health has set 1 x  10~3 /*g/m3 as a standard.
• Hepta and hexachlorobiphenyls—used in electricity transmis-
  sion transformers; can cause cancer and adverse skin, liver and
  reproductive health effects. The NIOSH standard is 1 jtg/m3.
• Hexachlorobenzene—used as an herbicide, wood preservative
  and also is a byproduct from chlorinated hydrocarbon produc-
  tion;5 can cause death in breast-fed infants, skin  sores, skin
  discoloration and enzyme disruptions.' There are no standards
  set for this pollutant, however, the TACB Health Effects staff
  uses  1 % of the estimated OEL of < 1 ppm as a guideline for
  permit and health effects review.
• Pentachlorobenzene—used as a precursor to fungicide produc-
  tion and as  a flame retardant; also occurs as a product of the
  degradation of Lindane;7 may cause mutagenic and carcino-
  genic effects in mice.' There are no air standards set for this
  compound,  but 1 % of the estimated OEL is used as a guideline
  by the TACB.
• Tetrachlorodlbenzo-p-dioxin—occurs as a byproduct of herbi-
  cide production;' causes liver cirrhosis, spontaneous abortions,
  kidney disease and chloracne.10 The New York Air Pollution
  Control has set a standard of 9.2 x 10-' jtg/m3.
• Trichloroethane—used as a degreaser, "cold" cleaner and dry-
  cleaning agent;" acts as  a narcotic to depress the central ner-
  vous system; can cause dizziness, uncoordination,  drowsiness
  and death systematically. Locally,  it is irritating to eyes and
  causes dermatitis.12 Air  limits include TLV of 10 ppm. The
  short-term exposure limit (STEL) value is 20 ppm (tentative).13
  The OSHA  standard  is 10 ppm.'4
    The compounds detected in the analyses of these samples are
  consistent with compounds  measured  in stack sampling con-
  ducted in other areas.15'16  Stack sampling conducted by the
  state of California and  Scott Environmental (contracted by
  EPA) indicates the small amounts of PCDDs and PCDFs were
  detected in municipal solid waste incinerator emissions.

INCINERATOR DESCRIPTION
  The municipal solid waste incinerator stack sampling was con-
ducted from Aug. 13 to 15, 1985. Three separate sets of samples
were collected over a 3-hr period each time. A representative por-
tion of the stack emissions was collected isokinetically using stan-
dard operating procedures described in the TACB Sampling Pro-
cedures Manual.
  The incinerator sampled is located on Singleton Drive in North
Waxahachie, Texas. It  has two dual-chambered incinerators that
operate at approximately 1600°F. The incinerators are permitted
to destroy 2088 Ib/hr of solid waste. The residence time is approx-
imately 45 min to 1 hr in the primary  burn chamber and 1.22 sec
                                                                                        DETECTION OF RELEASES    31

-------
                                                                                      TC  THERMOCOUPLE
                                                          Figure I
                                              Chlorinated Organic*, Sampling Train
for the secondary burn chamber which is often used to burn efflu-
ent gases from the first chamber.
  The California Air Resources Board  recommends  a flue gas
temperature of 1800°F + / -  190° with a residence time of 1 sec
under  well-mixed conditions.  These conditions should destroy
99.9% of the toxic organics compounds."
  The facility is operated 24 hr/day, 6 days/week. There is down
time for maintenance and cleaning each day. The  facility is per-
mitted to emit 12 Ib/hr or 53 tons/yr of particulate matter.

SAMPLING TRAIN
  The sampling train used to collect chlorinated organic  com-
pounds is shown in Figure 1. The sampling apparatus is made en-
tirely of glass except  for the stainless steel nozzle and teflon tape
seals used between the ground glass joints. Following the stan-
dard front-half, the back-half is modified to consist of a water
jacketed condenser followed by two XAD-2 and one Florisil resin
traps in series. The remainder of the train consists of two im-
pingers. The first impinger contains 200 ml of sodium arsenite to
collect chlorides/chlorine and to protect equipment downstream.
The  second impinger contains silica gel to insure the complete
dryness of the sampled gas. The Organic Laboratory Section per-
formed the analyses of  the samples using the Finnigan triple
quadrupole  mass spectrometer (TSQ).
  The hydrochloric acid emissions were captured using the collec-
tion system as described in Chapter 5 of the TACB Sampling Pro-
cedures Manual. The stack  exhaust gas  entered  the  system
isokinetically through a stainless steel probe and traversed a non-
heated probe across a glass fiber filter which was  maintained al
250 to 275 °F. The condenser  train  consisted of a series of im-
pingers containing deionized  water, alkaline arsenite and  silica
gel. The analyses followed the procedures described in the TACB
Laboratory Procedures Manual.
  Leak checks were performed several times during the process to
insure  sample integrity. If acceptable, the  sampling process was
continued.  The sample air was circulated through a condenser
filled with cold water. Due to the large volume (approximately
100 ft3) and  the high  moisture  content of the air drawn from the
stack,  a large  volume of condensate was  collected in the flask
following  the  condenser. As necessary,  the  condensate was
removed from the flask, the volume recorded, transferred to an
amber glass container and returned to the TACB laboratory for
analysis. A check for air leakage in the sampling system preceded
and followed the condensate transfer to insure sample integrity.
  Following the completion of sampling, the three  adsorption
tubes and impingers  were weighed again  and their final  weights
were recorded to determine stack air moisture content. The probe
and glassware in front of the filter were rinsed  with  tetrahydro-
furan (THF). All  rinsate was collected  in  amber bottles.  The
amber bottles were used to prevent rinse solutions from reacting
with ultraviolet  light. The glass fiber filter was returned to its pro-
tective alumi.ium  foil  and  stored   in  plastic  envelopes  until
analysis. The connecting glassware and two impingers  were rinsed
with deionized water and the rinsate was collected for analysis at
the TACB laboratory.

ANALYTICAL METHODS
  The XAD-2 and Florisil adsorbers were extracted using  hexane
in a Soxhlet extractor for 8 hr.  The distillate then was  evaporated
down to 2 or 3 ml using a rotary evaporator.  The product  was in-
jected into a gas chromatogram/mass spectrometer (GC/MS).
The pollutant results were interpreted by a  TACB chemist  who
used graphic responses compared to a chemical library to identify
the pollutants.
  The front-half and  back-half rinsates were  concentrated by
evaporation. Separately, the products of the evaporation were in-
jected into the GC/MS. The results were interpreted as discussed
in the previous  paragraph.
  The impinger containing deionized water was used to collect
hydrochloric acid. The samples were analyzed turbidimetrically.
Chlorine was collected in the arsenite solution. The solution also
was analyzed turbidimetrically.
  The glass fiber filter also was analyzed for chlorides and chlor-
inated organics. A 37 mm diameter circle was cut from each of the
three filters. The exposed filters  were analyzed versus the field
blank for chlorine on an x-ray fluorescence spectrometer system.
Duplicates also were run to test the method's precision. In  addi-
tion, the filters were  analyzed for chlorinated organics. They were
extracted separately  using hexane and a Soxhlet extractor. The
resulting solution then was reduced  to 2 or 3 ml with nitrogen.
The product was injected into  the TSQ and analyzed  as described
previously.
32
      DETECTION OF RELEASES

-------
CONCLUSIONS
  More stack sampling of municipal solid waste incinerators in
Texas is needed. During this initial stack sampling exercise, there
were some problems with the plant operation and with the stack
monitoring process due to  malfunctions  in  the  incinerator.
Preliminary determinations of the hydrochloric acid test  have
been made. The emissions of hydrochloric acid are approximately
1 Ib/hr. This value is lower than expected.
  Particulate matter collected during the test was analyzed and
found to contain significant amounts of chloride salts. It is be-
lieved  that the chlorides  exiting this  process  were primarily
chloride salts with relatively low hydrochloric acid emissions. Fur-
ther testing is required to determine if this is the case.
  Also, some of the pollutants detected in preliminary  analysis
are very toxic  and warrant additional attention. PCDDs and
PCDFs have been described by  some  sources as unsafe at any
level. This project has not produced  enough  information  to
prescribe  incinerating  at  higher temperatures, having  longer
residence times in the burn chamber, requiring additional pollu-
tion control devices or sorting out plastics before incineration.
  Whether or  not the Waxahachie municipal  solid waste in-
cinerator  is representative of incinerators in Texas also is not
known. Stack sampling at other sites may be needed to adequately
characterize  incinerator emissions.  Other types  of air  quality
monitoring also may  be warranted.  Perhaps  monitors placed
downwind may help correlate stack emissions levels to levels near-
by residents may eventually breathe. Therefore, developing con-
clusive emission factors is premature.

ACKNOWLEDGEMENT
  The authors would like to acknowledge the Texas Air Control
Board Laboratory, Dallas Regional personnel, and  S. Thomas
Dydek of The Health Effects Section for their assistance with this
project.


 REFERENCES
  1. Krakower, T.,  Resource Recovery Facilities Overview, Texas Air
    Control Board (Health Effects and Research Division), Sept. 1984,
    Section III.
  2. Ibid., Summary.
  3. Sittig, Handbook of Toxic and Hazardous Chemicals,  Noyes Publi-
    cations, Park Ridge, NJ, 1981, 227-228.
  4. Ibid., 228.
  5. Ibid., 358.
  6. Ibid., 359.
  7. Ibid., 522.
  8. Ibid., 522.
  9. Ibid., 632-633.
 10. Ibid., 632.
 11. Ibid., 669, 671.
 12. Ibid., 670-671.
 13. Ibid., 669, 671.
 14. Ibid., 669.
 15. Sheffield, A., "Sources and Releases of PCDDs and PCDFs to the
    Canadian Environment," Chemosphere, 14,  1985, 811-812.
 16. Nunn,  A.B., III, "Gaseous HC1 and Chlorinated Organic Com-
    pound  Emissions from  Refuse Fired  Waste-to-Energy Systems,"
    Scott Environmental Services (for U.S. EPA Environmental Sciences
    Research Laboratory), Plumsteadville, PA, 1984, 30, 37.
                                                                                                 DETECTION OF RELEASES    33

-------
                CARE—Modeling Hazardous Airborne  Releases

                                                M. Gary  Verholek
                                     Environmental Systems Corporation
                                              Knoxville,  Tennessee
ABSTRACT
  Emergency preparedness for airborne hazards means being able
to answer such questions as where the material will go, who it will
effect, what areas are safe as evacuation centers, how  long the
chemical will take to disperse and which preplanned response op-
tion would be most effective in the event of an incident. To help
answer such questions,  Environmental  Systems Corporation
(ESC)  has  developed  an assemblage  of  sophisticated
mathematical models into a user-friendly and truly capable com-
puter system called the CARE (Computerized Airborne Release
Evaluation) System.
  CARE  provides the capability to perform a variety  of func-
tions,  such as risk  assessment, hazardous materials  training,
monitoring and emergency response.  In short, CARE  provides
the  tools needed  to prevent an emergency from becoming  a
disaster.

INTRODUCTION
  Are you ready for a hazardous materials accident? What would
you do if a toxic cloud from a spill  or a ruptured tank threatened
your facility? What if burning chemicals were threatening your
town today? Where would the cloud go? Who would be effected
and who  would be out of danger?  What areas would need to be
evacuated? Where could people  go to be safe? What is the best
way to respond to the emergency? How do you  get answers to
these questions in time to prevent the emergency from becoming a
disaster?
  To answer such questions. Environmental Systems Corpora-
tion (ESC) has developed a system called CARE, which stands for
Computerized Airborne  Release Evaluation System. CARE gives
you the capability to perform a variety of emergency preparedness
functions:
• Risk Assessment
• HazMat Training
• Real-Time Monitoring
• Emergency Response
  CARE   is a  system   that puts  truly  capable  emergency
preparedness on  your desk. In the event of an accident  that
releases a hazardous material into the air, CARE will tell you
where it will go, who it will effect, which areas are safe as evacua-
tion centers, how long  it  will take to disperse and which  pre-
planned response options would be most effective. CARE gives
you the ability to neutralize the  danger, to assess  alternative
defense strategies and to keep abreast of changing situations. In
short,  CARE gives you  the tools you need to prevent  an emer-
gency from becoming a disaster.
                                                                                      CARE
Computerized Airborne Release Evaluation System
       SOURCE Mode!
                                      METEOROLOGY Model
      Calculates source
      cna'actensiics
            r
Calculates winds
stability, lemp
over modeling region
                       DISPERSION Model
                       Calcinates pollulani
                       conce^ra'.o^ 'te'ds
                        EFFECTS Model
                       Describes lie Hazards
                       due to ambient
                       concentrations
                       RESPONSE Model
                      Provides protective
                      action guidance
                          Figure 1
                   CARE System Diagram
 SYSTEM DESIGN
   CARE uses sophisticated mathematical models, assembled by
 ESC's meteorologists and engineers, to provide a realistic depic-
 tion of the hazard due to a  cloud of airborne material. These
 models are grouped into five  modules (Fig. 1):
 •  Source Module
 •  Meteorological Module
 •  Dispersion Module
 •  Effects Module
 •  Response Module
34    DETECTION OF RELEASE-IS

-------
  The Source Module provides one of the two basic mathematical
models required to estimate atmospheric concentrations of air-
borne materials. For chemical spills, it calculates the evaporation
rate of the chemical into the atmosphere. This rate of evapora-
tion, determined through heat and/or mass transfer mechanisms,
it used as the emission rate in the second mathematical model, the
air  dispersion model. The final product is then the calculation of
the  geographical  distribution  of  pollutant  concentration at
selected times  since the spill,  via the atmospheric dispersion
model. Specific source modules are available for various applica-
tions such as chemical spills, fires and transportation accidents.
nuclear plant accidents, fossil plant emissions,  airborne pesticide
releases, military smoke generators, etc.
  The Meteorological Module is a group of several models that,
in combination, characterize the  weather conditions that effect
the dispersion of hazardous materials. This module uses observed
meteorological  data, such  as wind speed,  wind direction, tur-
bulence and temperature and provides the necessary inputs to the
dispersion module.
  The Dispersion Module,  using the source and weather inputs,
provides a realistic depiction of the movement of the  cloud of
hazardous material and calculates its concentration with respect
to time and distance from the release. The model, which is the key
to systems realism, is a variable-trajectory, puff-advection model.
This kind of model treats the airborne release as a series of puffs
which are moved by the air flow that has been calculated by the
meteorological  module. The puffs are diffused along the way ac-
cording to the character of the atmosphere. The resultant disper-
sion of the material is displayed on the computer as a meandering
plume which flows over and around obstacles, much as one would
expect a plume to move once released to the air.
  The Effects Module assesses the critical  concentration values
and provides information on the hazard to health and welfare
posed by the materials in the plume. This module examines the
concentration fields for critical levels based on the application.
For example, the  effects module for chemical accidents deter-
mines exceedances of Threshold  Limiting Values or Lower Ex-
plosive Levels;  for  nuclear releases,  it  calculates dose and dose
rate;  for smoke obscurant  releases, it calculates transmission or
visibility; for ambient air quality, it determines exceedances of the
National Ambient Air  Quality Standards; and  so forth.
  The Response Module provides the user with recommended re-
sponse actions  based on the effects assessment. The  response
module is tailored by ESC to meet user requirements, special con-
ditions and capabilities.
  The CARE system assembles the aforementioned capabilities
into a user-friendly system that is fully documented and tested.
The near real-time display  of the location and concentration of
hazardous airborne materials provides the information needed to
make informed decisions that can save  lives and property.

SOURCE MODULE
  The Source Module for toxic releases is based on the SPILLS
model developed by Shell Development Company.' SPILLS is an
unsteady-state  model  which calculates the  evaporation  of a
chemical spill to determine the source strength of the vapor cloud.
Used in conjunction with the atmospheric dispersion model, it is
used to estimate concentrations of the vapors as a function of
time and distance downwind of the spill.
  Three options depending on the nature of the spill have been in-
corporated in the model: (1) continuous spills, such as leaks from
tank cars, tanks or pipelines; (2) instantaneously-formed pools of
liquids or liquified gases; and (3) stacks, where the emission rate is
assumed to be known.  For options 1  and 2, thermophysical pro-
perties (available as a subroutine in the computer program) of 36
potentially hazardous chemicals are  used  to calculate, through
heat and mass transfer mechanisms, the evaporation rate, which
becomes the emission rate for the atmospheric dispersion calcula-
tions.
  The computer program was  adapted for use with complex at-
mospheric dispersion models to provide a  realistic prediction of
hazard caused by an airborne toxic cloud. The program now con-
tains  the necessary  properties of  36 potentially  hazardous
materials commonly handled,  used or transported by industry,
but the data easily can be expanded to include the thermophysical
properties of other substances.  The air dispersion model takes in-
to account the atmospheric conditions, type of source and emis-
sion rate  to calculate  downwind  concentrations. The  source
strength can be specified by the user (stacks) or is determined by
the evaporation rate models  if the  chemical is  known. At-
mospheric and soil conditions and chemical properties are utilized
to predict the effects of heat and mass transfer on the vapor cloud
formation.

Evaporation Rate Models
  The objective of an evaporation rate model is to predict the
amount of material emitted in the case of a continuous or an in-
stantaneously formed spill from a transport line or a storage tank.
Although the basic theory is the same for both cases, the calcula-
tions are different, and thus they are  presented separately.
  Three evaporation rate models are discussed in this section.
First, the continuous  leak case which corresponds to emissions
generated  from small ruptures  in transport equipment  (e.g.,
pipelines, storage tanks or tank cars).  In this case, the chemical is
assumed to flow at atmospheric conditions through the rupture to
form a quiescent pool on the ground. Since the chemical will be at
ambient temperature, no heat transfer will occur between the
material spilled and the surroundings, and the amount of mass
taken by the wind blowing over the pool will be limited by mass
transfer.
  Two key  variables  involved in the mass transfer model are
simultaneously unknown: (1) the area of the spill and (2) the emis-
sion rate into the atmosphere. It was assumed that the area of the
pool would be the more difficult variable to predict during the ac-
cidental spill.  Therefore, the  computer program  requests the
value of the emission rate through the rupture,  which the  model
takes  as the emission rate into the atmosphere. The model then
determines the spill area by assuming that this emission occurs by
convective mass transfer. The evaporation then is assumed to occur
indefinitely at  the same rate unless the user specified the elapsed
time from the beginning to the end of the flow through the rupture.
At this point, the program would set the emission rate to zero.
  Evaporation from instantaneously  formed pools, the second
option in SPILLS, corresponds to the spill of a chemical from a
storage tank or transport line. In this case, however, the spill and
the resultant pool at  the ground are  assumed  to occur instan-
taneously. The area of the  pool is assumed to remain constant
during the evaporation process so that its estimation at the scene
of the accident would be easier than the continuous spill case. The
computer program then requests the value of this spill area in ad-
dition to the total amount spilled.
  Chemicals  with  a  normal  boiling  point   below  ambient
temperature will first flash off due to the pressure drop between
the storage pressure (the storage temperature has to be specified
by the user) and atmospheric pressure. The evaporation rate due
to this adiabatic flash calculation is assumed to occur during the
first minute after the spill. The chemical will then form a pool of
liquified gas at its normal boiling point. The difference between
this temperature and the ambient temperature will cause heat to
be transferred from the ground (the soil is  assumed to be  at am-
                                                                                             DETECTION OF RELEASES    35

-------
bient temperature) to the pool by conduction  and from the at-
mosphere to the pool by convection.
   Mass transfer, due to the wind blowing over the pool,  takes
over when the heat transfer evaporation rate becomes equal to the
mass transfer evaporation rate. Note that this occurs because the
heat transfer rate decreases as time increases, whereas the mass
transfer rate is  independent of time.
   Finally, the  emission rate  is equated to  zero when  all  the
chemical  is evaporated. Chemicals with a normal boiling  point
higher than ambient temperature are  assumed to form a liquid
pool at ambient conditions.  Here, the chemical is  assumed to
evaporate only by convective mass transfer.

Continuous Leaks
   Continuous leaks are regarded in the present work as emissions
from a small rupture of a transport line (such as a pipeline) or a
storage tank (such as a tank car). It is assumed that the mass flow
rate through the orifice is known and  that the chemical flows at
atmospheric conditions, i.e., as a gas or as a liquid depending on
the  normal boiling   point  being  below  or  above  ambient
temperature, respectively.
   The chemicals of interest in the present program are all  more
dense than air  so that a gaseous plume or a liquid pool will be
formed at ground level after the chemical exits the original car-
rier. Convective mass transfer is assumed to be the limiting pro-
cess for the chemical to be transported  by the wind. The spill area
then is predicted by equating the mass transfer rate to the flow
rate through the rupture.
   The pool size is determined for both laminar and turbulent flow
cases and is used to predict the corresponding Reynolds numbers.
The value that gives the consistent flow regime is chosen as  the
final pool length.
   The area of the spill is used as a ground level area source in the
air dispersion model.  In addition, the flow rate through the rup-
ture is utilized  as the emission rate in  the air dispersion calcula-
tions. The user has the option of specifying the length of time
from the beginning  to the end  of the flow.  This time  can be
estimated by knowing the amount of material to be emitted. This
case is approached by the air  dispersion model as an unsteady-
state case where the constant emission rate is dropped to zero at
the  corresponding time. The program assumes a  steady-state
operation if no time is specified by the user.

Instantaneously Formed Pools
   The time-dependent evaporation rates of stationary, instantan-
eously formed  pools are calculated by two different procedures
depending upon the chemical spilled: (1) liquified gases or (2) liq-
uids. However, both methods assume that the chemical pool does
not spread on the land as a function of time, such that the spill
area specified by the user remains constant.

Liquified Gases
  An initial adiabatic flash calculation  is performed based on the
pressure drop of the  liquified gas (i.e., from  the pressure of the
chemical  while stored, to atmospheric pressure).
  The original  enthalpy of the chemical is evaluated as the en-
thalpy of the saturated liquid at the user's specified temperature.
For conservative purposes, the amount of vapor which is flashed
off initially is assumed to be emitted in a period of 1  min.
  The rest of the liquified gas remains in the liquid pool specified
by the user at its normal boiling point. From this point on, the
chemical  can undergo two different  transport  processes:  heat
transfer or mass transfer.
  Due to heat transfer, which occurs mainly by conduction  from
the soil assumed at ambient temperature and by convection  from
the atmosphere, the  chemical  evaporates off the  boiling liquid
pool. It is assumed that solar radiation is negligible and that no
thermal resistance exists between the soil and the boiling liquid.
The  mass range at which the chemical evaporates is then deter-
mined.
  The air dispersion model uses the time-dependent mass rate
calculation to estimate the source strength during a heat transfer
evaporation mechanism  based on  the total amount  of  mass
evaporated between a time t(0) and any later time t.
  Since all the chemicals  treated in the present work are  more
dense than air, the emission rate into the atmosphere is  mass
transfer limited. This means that only a portion of the evaporated
mass will be transported by the wind. Mass transfer governs this
take-up flux into the atmosphere. The rest of the evaporated
chemical remains as a gas cloud above the liquid pool. For conser-
vative purposes, and due to the fact that the person at the scene of
the accident will report an estimate of the size of the pool, it is
assumed in the present work that all the mass evaporated by heat
transfer will be emitted into the atmosphere.
  Mass transfer occurs  predominantly by forced convection over
the liquid pool due to the wind. The amount of chemical emitted,
used in the air dispersion model, is obtained from integration with
respect to time of the amount of material emitted  by mass transfer
during the specified interval of time.
  Heat  transfer dominates at the beginning of  the process.  As
time increases, the  heat  transfer emission  rate decreases and
reaches a point  in  time where it  becomes equal to the  mass
transfer rate.

Liquids
  Liquids are  defined as those chemicals with a normal boiling
point,  T(B), higher  than the ambient  temperature, T(0). These
chemicals are transported  at conditions very close to ambient so
that  emission into the  atmosphere after a  spill will occur only
through mass transfer by convection. The risk to health and life
from an atmospheric release of hazardous effluent is a function of
the ambient downwind concentrations and the length of the  ex-
posure period.

THE DISPERSION MODEL
  A  vital element in such a consequence assessment is the ability
to estimate the location  and concentration of an airborne effluent
plume. To provide this estimate, the dispersion models employed
must include both the transport and diffusion properties of the at-
mosphere.  Consequently,  the  models must account  for  the
physical, temporal  and  spatial  changes  that  the atmosphere
undergoes during  the course of the release and during transport.
  To estimate the downwind location  and concentration of air-
borne effluents, CARE uses a highly sophisticated atmospheric
dispersion model. The model includes  both transport and diffu-
sion  components  which account for  the  temporal  and spatial
changes that the atmosphere undergoes during the course of the
release and during transport.
  Atmospheric dispersion calculations are  performed usin$  the1
ESC  variable  trajectory, puff-advection  model—TRAJ.' The
model simulates a release as a series of superposed puffs which are
transported by the local atmosphere. Each specification of new
meteorology is considered a new transport interval. During such a
transport interval, the  puffs are diffused according to  standard
Gaussian theory. The dimensions of an individual puff are pro-
portional to its travel distance (or travel time). The model accom-
modates multiple point  sources and includes algorithms for plume
rise and shoreline effects. The model  provides a. rigorous  treat-
ment of the spatial and temporal variations in the atmosphere and
accounts for influences due to local  meteorological effects.
  The advection of the puffs is accomplished through inputs of
local meteorological variables, wind speed and direction. The
36
      DETECTION OF RELEASES

-------
model can accommodate several meteorological stations by apply-
ing a weighting to each station variable. For these applications in
areas of complex flow fields (e.g., mountains, valleys, shorelines,
etc.), a three-dimensional wind field model can be added to pro-
vide a rigorous, dynamic treatment of the spatial and temporal
variations in the atmosphere.
  The dispersion model calculates the concentration over a grid
covering the modeling region and also at specified special receptor
locations within the grid. Special receptor locations are stored as
data to facilitate modification should additional stations be needed.
Site-specific run parameters, which are input as data, define  the
various site-specific characteristics unique  to this model applica-
tion. The calculation is performed in a time-wise, step-by-step
analysis of the plume  path from the release point through  the
limits of the calculation grid.

THE METEOROLOGICAL MODULE
  Dispersion  modeling  requires  the  input  of   critical
meteorological data. Wind speed and direction are needed to
define the transport of airborne material;  stability data are used
to determine the spread of the material as it travels downwind.
The depth of the layer of the atmosphere where mixing occurs is
called Mixing Depth; this parameter defines the maximum height
of  the  plume.  Temperature  data  are  used to determine  the
bouyant rise of the plume and, where applicable, the existence of
a sea breeze.
  At each modeling time  increment  (generally  15 min),  input
meteorological data are used to determine:
• Transport wind (u,v,w) components
• Plume dispersion coefficients (sigma-y; sigma-z)
• Atmospheric temperature
  The meteorological data  can be acquired in real-time from an
on-site meteorological  tower  or,  if data are acquired manually
from alternate sources such as National Weather Service stations
or United States Air Force bases, they can be input from the con-
sole via menu responses. Wind data, input as speed and direction,
are converted to transport wind (u, v and  w) components at  the
grid points. Horizontal and vertical plume  dispersion coefficients
(sigma-y and sigma-z) can be derived from direct measurements
of turbulence  (sigma-theta  and sigma phi) or from routine me-
teorological observations using well-known stability classification
schemes, depending on the available input data.

Stability Classifications
  The most commonly used method to determine dispersion coef-
ficients  is  the indirect method, requiring the determination of a
stability class which  then determines the  dispersion curve. The
Pasquill-Turner stability classification  scheme  or the  NWS-
Turner  method are typically used  with  the Pasquill-Gifford
dispersion curves. Stability categories cannot be measured directly
but must be determined from other meteorological parameters.
Any of the following approaches  can be used:
• Sigma phi (standard  deviation of the vertical wind direction)
• Sigma theta (standard deviation of wind direction)
• Delta temperature (the difference in temperature with height)
• Cloud conditions (i.e., clear,  scattered, broken or overcast,
  and ceiling height)
• Default conditions (e.g.,  F-nighttime; D-daytime)
• Manual override inputs (operator discretion)
  When using sigma theta or sigma phi to determine stability clas-
sifications, and changes in dispersion are expected because of sur-
face roughness, a factor of (Zo/15 cm) A0.2, where Zo is  the
average surface roughness in centimeters, may be applied to  the
tabular values. Suggested Zo values that may be used as a guide to
estimating  surface  roughness  are  given  in  Hogstrom  and
Hogstrom.2
  The  sigma stability classification  schemes  are  adequate for
daytime use,  but during nighttime (1 hour to sunset to 1 hour
after sunrise), adjustments  adapted from Mitchell and Timbre3
are used.
  If cloud cover inputs from NWS or USAF sources are used, the
values  are compared to criteria  published in  the  U.S.  EPA
CRSTR model. Default dispersion classes  (e.g.,  D-daytime;
F-nighttime) can be provided for the operator  in the event no
other data are available.

Dispersion Coefficients
  The  purpose of establishing a stability class is to be able to
calculate  the  horizontal and vertical dispersion  parameters,
sigma-y (Sy) and sigma-Z (Sz), which are used in the calculation
of downwind concentration. Sy and Sz are approximated by a
curve fitting process which describes the PGT  curves with  power
laws  as a function of the total distance  traveled, x:
  Sy(x) = a*vAb
  Sz(x) = c*vAd + e
(1)
(2)
where the coefficients (a,b,c,d,e) are a function of stability class.
The coefficients are taken from the POLYN routine found in
NUREG/CR-2919, the manual for the XOQDOQ model, written
by Sagendorf, et al.' but have been modified slightly to improve
the continuity of sigma-z from the region of x < 1000m to x < =
1000m.

Mixing Height
  The  boundary layer height  (H)  is used to limit the  vertical
dispersion parameter (Sz)  to values  < 0.8*H. The mixing depth
will be determined by applying the climatological mixing heights
according to the method  described  in the U.S. EPA CRSTER
manual,' which is the standard Holzworth method.

Sea Breeze
  In regions near  large bodies of water where a  sea breeze
phenomenon may be a significant  meteorological feature, the
Thermal Internal Boundary Layer (TIBL) will be used as the  max-
imum mixing height. The software will determine if a sea breeze
condition is likely to exist, using the following rules:
• The sun is above the horizon (based on the data and time)
• The lower-level air temperature is greater than the sea water
  temperature, or monthly climatological temperature  if  tem-
  perature data are not  available
• The  lower  wind direction  measurement  indicates on-shore
  winds within a 180 degree sector
  If a sea breeze is determined to exist, the mixing height  is set
equal to the height of the Thermal Internal  Boundary Layer
(TIBL):
  H = C * sqrt (d)
(3)
where H is the TIBL height (in meters) at distance d (in meters)
inland6'7 and C is a constant determined for the local area.
Wind Field Model
  For applications in areas of complex flow (e.g., valleys, moun-
tains, seashores, etc.) a three-dimensional,  mass-consistent flow
field model is used to provide a rigorous, dynamic treatment of
the spatial and temporal variations in the movement of the cloud.
  The ESC windfield model (ESCWIND) was adapted from code
available  in  the  public domain  from several  sources. It  was
developed  to provide  the user with  an economical windfield
                                                                                              DETECTION OF RELEASES    37

-------
predictor.  The  model  produces three-dimensional,  terrain-
dependent,  divergence-free windfields given  observed surface
and/or upper-air data as input. Given the wind components for
each meteorological station and at each level, a 1/r weighted in-
terpolation is used to "fill  out" the  rest  of the windfield along
each horizontal plane.
   The terrain surface boundary is approximated  using obstacle
cells to represent the terrain in a stairstep fashion. For computa-
tional purposes, the irregular  surface boundary is removed by a
coordinate transformation  in  which the  terrain surface also
becomes a coordinate surface.  This "sigma-space"  computa-
tional domain has two main advantages. First,  the bottom boun-
dary condition is defined more accurately; second, the option of
variable vertical  zoning improves  the model's  accuracy and
economy.
   ESCWIND follows a procedure  of extrapolating and  inter-
polating  the input  data  to  determine  an  estimated  three-
dimensional  windfield on a specified finite difference grid. This
prospective  windfield then is adjusted to account for  terrain
effects and atmospheric stability considerations constrained  by
the condition that the resulting windfield be nondivergent.
   The basic  set of model  equations was  derived from the ap-
proach that minimizes the squared variation of  the windfield sub-
ject to the constraint that the adjusted field be  nondivergent.8'°

APPLICATIONS
   The CARE Dispersion Modeling System has many applications
such as toxic spills, nuclear dose assessment, pesticide drift, am-
bient air quality assessments  in complex terrain,  chemical fires
and visibility obscurations.  The dispersion model is designed to
provide physically  realistic calculations of the position and con-
centration  of downwind airborne effluents. These calculations
then are available  for application in any scenario that relies upon
atmospheric  dispersion calculations. The type of application for
which the system is to be used  will dictate the type of source term
calculations and the effects  and response modules that would  be
useful in the  CARE system.

Nuclear Accidents
   The system is applicable  for risk assessment and operational
dose  assessments   from  radioactive  releases.  It  has  been im-
plemented  for real-time, operational  dose assessments using ac-
tual meteorological data signals,  for manual and back-up dose
calculations and for training purposes. The model can be used for
dose  projection   estimates based  on  forecase  or  historical
meteorological data. For training exercises,  the operator would
run the model using the  interactive menus where input data and
run parameters have been selected to simulate the desired exercise
conditions.

Toxic Spills
   For  toxic spills emergency  response, a  chemical source  term
model has  been added which calculates the appropriate  release
rate based on chemical properties, release mode and weather con-
ditions. The  effects model  provides recommended response ac-
tions based on the concentration Threshold Limiting Value or the
Lower Explosive Level.

Visibility Obscurations
  By using a  visibility model to calculate the effects of an aerosol
released to the air, the  system has been  used to calculate the
obscuration effects of various smoke releases for military applica-
tions.  This capability also is useful in determining  visual impacts
on U.S. EPA Class I areas.
Ambient Air Quality
  The system is extremely useful for calculating ambient air qual-
ity  in complex terrain  where traditional straight-line Gaussian
models typically perform unrealistically. The use of the variable-
trajectory calculation capability of the system helps  to account
for  terrain-induced dispersion.  This  capability is  useful  in
transport studies such as odor and  toxics impacts.

CONCLUSIONS
  CARE gives one the capability to perform a variety of emergen-
cy preparedness functions:
• Risk Assessment
• Hazmat Training
• Real-Time Monitoring
• Emergency Response
  CARE is a system that puts truly capable emergency prepared-
ness on one's desk. In the event of an accident that releases a
hazardous material into the air, CARE will tell one where it will
go, who it will effect, which areas are safe as evacuation centers,
how long it will take to disperse and which preplanned response
options would be most effective. CARE gives one the ability to
neutralize the danger, to assess alternative defense strategies and
to keep abreast of changing situations. In short, CARE gives one
the tools  needed  to prevent an emergency from becoming a
disaster.
REFERENCES
 I.  Fleischer, M.T., "SPILLS, An Evaporation/Air Dispersion Model
    for Chemical Spills on Land," Shell Development Company, Hous-
    ton, TX, Dec. 1980.
 2.  Hogstrom, A.S. and Hogstrom, U.. "A Practical Method for De-
    termining Wind Frequency Distributions for the Lowest 200m  from
    Routine Meteorological Observations." J. Appl. Meteor., No. 17,
    1978, 942-954.
 3.  Mitchell, A.E. Jr. and Timbre, K.O., "Atmospheric Stability  Class
    from Horizontal Wind Fluctuations," paper presented at the  72nd
    Annual  Meeting of the Air Pollution Control Association, Cincin-
    nati, OH, June 1979.
 4.  Safendorf, J.F., et  a/.,  "XOQDOQ:  Computer Program for the
    Meteorological Evaluation of Routine Effluent Releases at Nuclear
    Power Stations," NUREG/CR-29I9, Sept.  1982.
 5.  U.S. EPA, "User's  Manual for Single-Source (CRSTER) Model,"
    EPA-450/2-77-013, July 1977.
 6.  Raynor,  G.S.,  el  a/.,   "Recommendations  for  Meteorological
    Measurement  Programs  and  Atmospheric Diffusion Prediction
    Methods for  Use at  Coastal Nuclear Reactor Sites," NUREG/CR-
    0936, Oct. 1979.
 7.  Beebe, R.C. and Sorge,  J.M., "Coastal and Overwater Dispersion
    Characteristics  of a  Warm  Water Environment,"  unpublished
    paper, Apr. 1975.
 8.  Patnaik, P.C. and Freeman, B.C., "Improved Simulation of Meso-
    scale  Meteorology Phase I," Science  Applications, Inc.,  Report
    SAI-77-915-LJ, Mar. 1977.
 9.  Sasaki,  Y., "Some  Basic Formalisms in Numerical Variational
    Analysis," Won., H'ea.,  Rev., No. 98, 1978, 312-319.
10.  Sherman, C.A.,  "A  Mass-Consistent Model for Wind Fields  Over
    Complex Terrain," J. Appl. Meteor., 1978, 312-319.
 38
      DETECTION OF RELEASES

-------
                    The Use  of PC  Spreadsheet-Based  Graphics
            to  Interpret Contamination at CERCLA/RCRA  Sites

                                            George  A. Furst, Ph.D.
                             U.S.  Environmental  Protection Agency, Region 1
                                             Boston, Massachusetts
ABSTRACT
  During site investigations, large amounts of environmental data
are generated in the form of lists of chemical analyses from differ-
ent sampling locations over various periods of time. These tables
of data are difficult to analyze and are seldom fully interpreted
because of the magnitude of the task and the difficulty of visualiz-
ing trends from long lists of numbers. Electronic spreadsheet soft-
ware for personal computers, originally developed for use by the
business community for financial analysis, may be employed to
store site data and to graphically display contamination patterns
at CERCLA/RCRA sites. This software allows the input  of
numbers, formulas and words into a matrix of rows and columns
and provides an ideal format for listing contaminant concentra-
tions in the air, soil, surface water and groundwater.
  When the above data are entered into a spreadsheet  and in-
tegrated with  graphics software, it is a simple task to rapidly
visualize contamination distribution at a site over space and time.
The following spreadsheet-based graph types are illustrated: the
line, bar, stacked bar, x-y  and pie chart. Each type is well suited
to a particular application.  For example, the pie chart is useful for
showing a connection between a source area (such as  a  lagoon)
and a receptor (such as a downgradient well). The expanded bar
chart can be used to indicate if new sources of contamination are
affecting the downgradient groundwater. Specific applications of
the spreadsheet-based graphics  are illustrated  using data from
Region 1 CERCLA/RCRA sites.

INTRODUCTION
  The  advent  of the personal computer and  its increasing
availability, as well as the increase in user sophistication, has led
to new applications of this  powerful instrument. The spreadsheet
has become the basis for many financial as well as scientific ap-
plications. One recent example of the use of the spreadsheet is
that of Olsthoorm.'  In this case, the spreadsheet was used to
model groundwater flow in two and three dimensions.
  The spreadsheet is used  because it is a general purpose, inex-
pensive  software product  and  one need not  be a  computer
specialist to use it. In the above application,  the spreadsheet
replaced three or more specific programs. This paper applies the
spreadsheet analysis to aid in  understanding the flow of con-
tamination from an uncontrolled hazardous waste site  and from
an active manufacturing site.
  Often there  is  a great deal  of information  known about a
specific site. This information is available as tables of chemical
data collected while sampling the groundwater, surface water,
soils, sediments and air at the site. Without being able to organize
and visualize this data, one often does not utilize the data to the
full potential. The computer-based electronic spreadsheet, com-
bined with graphics, provides the means to more fully interpret
data.
  The spreadsheet,  more than any other software program, has
revolutionized the use of the personal computer. The spreadsheet
consists of a matrix of rows and columns in which numbers, for-
mulas and words are input. It is also a very user-friendly software
tool that is limited only by the imagination of the user. Examples
of popular spreadsheet-based software are: Visicalc, Multiplan,
Lotus 1,2,3 and Symphony.
  The spreadsheet used in this paper is Symphony, which along
with Lotus 1,2,3 represents a new generation of integrated soft-
ware; word processing and graphics are  directly available while
working with the spreadsheet. The Symphony spreadsheet con-
tains 256 columns and 8,192 rows. This capacity allows all known
chemical  information from a specific  site  to be stored on one
spreadsheet. The size of the spreadsheet  is demonstrated by the
fact that  the above matrix may hold 63,000 individual analyses
consisting of 30 chemical elements or compounds. In total, the
spreadsheet contains over 2 million available data points,  limited
only by the computer's Random-Access Memory. The advantage
of storing all the data from one site on one large spreadsheet is the
ability to rapidly compare different sections of data and to con-
duct direct statistical analysis on specific ranges of data. The ap-
plication described in this paper is the ability to focus on sections
of data in the spreadsheet and to make a picture or graph of the
numbers. An important assumption made is that the data are ac-
curate and have passed a quality screening  test.

APPLICATION OF SPREADSHEET-BASED
GRAPHICS AT A SUPERFUND SITE
  Table 1 is a portion of one spreadsheet of data from a Super-
fund site in southern New Hampshire. It  is displayed in the stan-
dard spreadsheet  format: i.e., the  information is listed in rows
and columns. Note that both numbers and words are used. At
this site, there are a sufficient number of monitoring well sam-
pling locations to define the cross-section of a contaminant plume
adjacent  to a water supply. The  horizontal  axis of the table
represents the distance in feet from the assumed center of the con-
taminant  plume to the center of the  sampled monitoring well
screen. All samples were taken in the same period of time.
  An examination of the data shows that  total contamination de-
creases with distance from the center of the plume. Also, the con-
centration of 2 -butanone decreases rapidly, but it is not as ap-
parent that t-l,2-dichloroethene is  decreasing much less  rapidly
across the plume cross-section.
                                                                                       DETECTION OF RELEASES    39

-------
                                                           Table 1
                                    Distribution of Contamination, Plume Center to Water Supply


              Distance  from Plums Center        0         70       130        175      250                 310

              Sample  number             NUS 2-3   NILS  2-2    NIJS 2-1    N'JS  -2   C5X  11     NUS 2BR    WP-4
              Benzene
              1,2-Dichlorde thane
              1,1,1-Trichloroethene
              1,1-Dichloroethane
              Chloroform
              1,1-Dichloroethene
              t-1,2-Dichloroethene
              1.2-D ichloroethene
              Ethylbenzene
              Tetrachloroethene
              Toluene
              Trichloroethene
              Vinyl Chloride
              2-Butanone
              4-Methyl-2-Pentanone
              Total Xylene
              Chlorobenzene
              Total:
                         (PPB)
10
51
24
48


760

17
14
410
27
13
1100
69
48

2591

7
41
38
43
7
5
810
5
15
12
350
23
10
520
58
43

1987

7
32
24
37
5
5
590
5
9
10
240
20
10
310
37
24

1365



34
25


520


15
11
36





641


19
35
12


300


5

22




11
404




32


18


5

11

27



93



0.5



5




0.5





6

               Note: All concentrations in ppb (or jig'l)
   Figure 1 is an x-y plot of the distribution of specific organic
contaminants versus distance from the center of the plume. Using
this plot, one can see the partition of different organic chemicals
in the plume diffusion gradient perpendicular to a groundwater
flow  direction. This distribution may be due  to differences in
sorption as discussed by Mackay, el al.1 This hypothesis can be
modeled by taking additional data from the spreadsheet (such as
analyses from downgradient wells) and comparing these analyses
to the above theory. Because the data are readily available on the
spreadsheet,  examination  of contaminant  distribution  in the
water supply well and the bedrock  well is facilitated. These two
analyses are listed in the columns headed by  WP-4 and NUS 2BR
on Table 1. The pie chart (Figs. 2 and 3) is jsed because it graph-
                     CONTAMINANT PUT .E CROSS SECTION
                         VOLATILE COKT/umMON NEAB WP «U
                          DISTANCE fDOU PUIME CENTER (IT)
                      TOtUENE        6  H.2-OPCHUWOETHENE
                          Figure I
  X-Y Diagram of Volatile Contamination in a Groundwater Plume
 ically depicts the relative percentages of organics impacting these
 different  receptors and specifically  shows the  persistence  of
 t-1,2-dichloroethene.

 APPLICATION OF SPREADSHEET-BASED
 GRAPHICS TO AN RCRA SITE
  Groundwater monitoring at manufacturing facilities is a critical
 aspect of the U.S. EPA RCRA permitting process. Implementa-
 tion of the 40 CFR Part 265 and 270 groundwater monitoring se-
 quence generates much data related to the permitting process,
 either closure or post-closure. If the detection monitoring wells at
 a land disposal  site detect contamination,  the  site immediately
 goes into assessment monitoring.
  The goal of this monitoring is to clearly identify the rate am ex-
 tent of hazardous waste constituent migration and to establish the
 concentration of individual constituents in the plume. For all of
 the above  assessment  data, the spreadsheet provides an ideal
 storage file for later retrieval and interpretation.
  The following discussion is an example of the  ways in which
 one can use the above data once they have been entered into the
 electronic spreadsheet.  The  site utilized for  illustration  is  a
 northern New England metal machining and plating company
 that has been active since the early 1950s. From  1970 to 1985, an
 unlined waste lagoon was used to manage hazardous waste at the
 site. Detection monitoring wells were installed in 1981 to assess
 the site groundwater flow and  potential contaminant migration
 irom the waste lagoon. During the early 1980s,  about 100,000
 gal/day of waste from the manufacturing process entered the
 lagoon. The waste stream was  composed principally of solvent
 wastes from metal degreasing, waste oils from metal  machining
 and sludge from electroplating operations. In order to assess the
changes in contaminant migration at the site, all data collected
 from  1981  to 1985 were entered onto  the spreadsheet. Table 2
contains the data collected during the fall of 1982. The following
 figures are based upon this table.
40
      DETECTION OF RELEASES

-------
                 MONITORING WELL NUS 2 BEDROCK
                       TOTAL VOLATILES - 93 PPS
    2-Bt/TANOtC (».OX)
 TTflCHLOROETHENE (tl.«)
                                             1.1-OCHLOROETHANE (J4.4X)
        TETHACHLOROETHENE (S.4*>
                                   1-1.2-OCHLOROETHENE (19.4%)
                          Figure 2
Pie Diagram of Volatile Distribution in Monitoring Well NUS-2BR
                      WP WATER SUPPLY
                      TOTAL VOLATILCS - « ffB
            TRICHLOROETHE)* (».3»)
                                  1.1.1-TRCHLOROETMANE (8.3*)
                            1-I.Z-DICHlOroCTHeNE (83.W)
                         Figure 3
     Volatile Organic Contamination in WP Water Supply
                  Figure 4 is a stacked bar chart that illustrates the composition
               of the oily waste lagoon. This graph  is used because it gives a
               visual summary of the contaminant profile at different depths in
               the waste lagoon in 1982. Figure 5 shows the distribution of con-
               taminants at mid-level in the lagoon. Figures 6, 7 and 8 illustrate
               the volatile distribution in wells 3, 4 and 7, located 50 ft adjacent
               to and 50 and 1,450 ft downgradient from the lagoon. The main
               difference  in contaminant profiles is the absence of methylene
               chloride in the farthest downgradient well MW-7.  The  other
               organic chemicals are  present at lower concentrations than in the
               lagoon.
                               MAJOR COMPOSITION OF OILY WASTE LAGOON
                                            SAMPLES Of 10/5/82
                   35000 -


                   3OOOO -


                   25000 -


                   20000


                   15000
                                                                    E7"3 METHYI^NE CHLORIDE

                                                                    [V~l 1.1.1-TOCHLOROETHANE
                                                 IT^ TMCHLOROETHYIENE

                                                 rvTl 1.1-DICHLOROETHANE
                                                                                             Figure 4
                                                                           Volatile Organic Concentrations in Waste Lagoon
                 The pie diagram is an ideal graph to show the hydrologic con-
               nection between source area, lagoon and receptor which in this
               case is the downgradient well. At sites where there are multiple
               source areas,  such  as  separate lagoons used for waste oil and
               plating and degreasing wastes, the pie chart represents a signature
               that is useful  for indicating which downgradient wells are con-
               nected hydrologically with a specific lagoon.
                                                           Table 2
                               Lagoon Composition at Surface, Mid-Depth, Lower and Sludge (10/82)
                                        Monitoring Well Groundwater Composition (10/81)
DATt:                      10/5/82
SAMPLE NUMBER           Surface
DISTANCE FROM LAGOOM         0

1,1-Dichloroethane           630
1,1-Dichloroethylene
Ethylbenzene
Methylene Chloride
Tetrachloroethylone
Toluene
t-1,2-Dichloroethylene
1,1,1-Tr ichloroethane
Trichlorethylene

Total:  (PPB)
                                     10/5/82   1-1/5/8 ^
                                     M id-Dap th   lower
                                            0       0
                            12960
                                          7550
1390
10/5/5'2
 sludge
   0

  13700
                             10/16/81
                              MW-1UPG
                             150

                                 0
22775   43731
           164320
               34
                                 1.0/1.6/81
                                 MW-2ADJ
                                 150
                                                                                        11
                                                                                        45
1.0/16/1)1
  MW-3
  50

   830
   170
10/16/131 10/81
   MW-4 MW-7
  50    1450
                                                                                                              1500
           570
           110
7500

174

2586
2070
9900

0

3000
2325
30500

121

6350
5370
96900
1240
1180

23100
28200


0

34



12
22


820
340
21
140
1000
270
470
14

1300
33
1300

70

1100
430
950
                                               3591
                4617    3230
Note: All concentrations in ppb 0
-------
                     OILY WASTE LAGOON. MID-DEPTH
                          IOIAL voutus - 2277s m
             HWCHUXlOCTmVCW (I0.2X)
M.I-TWCHlOfWCTHKNC (13.2X)
                                                  I.I-OCW-OHOCTHWC (31 !»)
             urnmrxc cmonct <«3 M)
                               Figure 5
     Volatile Organic Concentrations at  Mid-Depth in Waste Lagoon
             MONITORING WELL 4, OOWNGRAOlENT
                    TOT*. KXATUI - «M4 m
                                                                                                                            I.I-OCMU*OCTNW*OCT»*M*t (07X)
                                                                                                                    TTTHMXJCMOCTMtUM (OJX?
                        Figure 7
Volatile Organic Contamination of Groundwater at MW-4
                    MONITORING WELL 3. DOWNGRADIENT
                                  i - 3M t ^0
            MONITORING WELL 7. OOWNGRA0IENT
                  I0t«t MX*rU3 - 3241 »»•
                mO«.OHOCTHTU>C (
                                               1.1-DICHt.O
-------
                           SO MONITORING WELL 3
                            PERIOD SAMPLED 10/81 - 10/85
                         SO MONITORING WELL 3
                       CONCENTRATION OF METHYLENE CHLORIDE
              1.1 -OICHLOROETHANE
                              t. 1.1 -TR1CHLOROETHANE
                           BATE SAMPLES ANALYSED

                    U7~* 2/U/82       r^J 10/5/82
                                                TRICHLOROETHYUNE
                                                   r\7\ 8/11/&J
                    E53 5/25/84
                                      | 10/85
                           Figure 9
         Volatile Organics Present in MW-3 (10/81 - 10/85)
  The spreadsheet also provides the basis for comparing changes
in organic contaminants at a specific well over a number of years.
This comparison is facilitated since data can be moved rapidly
from one section of the spreadsheet to another. Table 3 contains
data from monitoring well 3 which were collected from October
1981 to October 1985. Figure 9 is a bar graph of the major organic
contaminants, which,  with the exception of methylene chloride,
are contained in the groundwater. The graph illustrates that the
highest organic  contamination at well 3 occurred in late 1982.
Since then, there has been a decrease in the concentration of the
three compounds, as shown in this figure. Figure 10 shows the
concentration of methylene chloride over the same time period.
The highest concentrations of this  species were observed in Oc-
tober 1982 and  October  1985. The later date coincides with the
final closure of the lagoon and may be related to it. This line chart
is particularly useful for showing changes in contamination over a
period of time.
  In summary, spreadsheet-based graphics allow one to visualize
the  changes  in  concentration of  organics in the source  area,
lagoon and downgradient monitoring  wells and changes over a
4-year  period in  well number 3.
                                                                                     O DATES OF SAMPLING
                                                                                            Figure 10
                                                                      Concentration of Methylene Chloride at MW-3 (10/81   10/85)
CONCLUSIONS
  The spreadsheet is the fundamental way to store data in the
computer. Once data from a hazardous waste site or manufactur-
ing site are entered into the spreadsheet, one can readily combine
data, move data and, when the spreadsheet is  combined with
graphics software, one can generate graphs from any portion of
the spreadsheet. In this paper, the spreadsheet has formed the
basis for quickly assessing the contamination at two sites. The
software also has been used by the author to  prepare graphs for
overhead projection at a meeting, to prepare tables and graphs
for reports and to forecast when increasing contamination in a
well will cause the water to exceed Safe No Action  Response
Levels (SNARLS). With the increasingly available portable com-
puters, the above site-specific information can be taken  to the
manufacturing  company and used as a resource during discus-
sions regarding contaminant distribution at an RCRA site. As
stated earlier, the application of this new generation of software is
limited only by the imagination of the user.

REFERENCES
1.  Olsthoorm, T.N., "The Power of the Electronic Worksheet:  Model-
   ing Without Special Programming," Groundwater, 23, 1985, 381-390.
2.  Mackay,  D.M., Roberts, P.V. and Cherry, J.A., "Transport of Or-
   ganic Contaminants in Groundwater," Environ. Sci.  Technol., 19,
   1985, 384-391.
                                                                                              DETECTION OF RELEASES    43

-------
         Emergency Response to  Toxic Fumes  and  Contaminated
                          Groundwater  in Karst Topography:
                                             A  Case Study

                                              P. Clyde  Johnston
                                               Mark J. Rigatti
                                             Roy F. Weston, Inc.
                                      Jacobs Engineering Group,  Inc.
                                               Atlanta,  Georgia
                                               Fred B.  Stroud
                            U.S. Environmental Protection Agency   Region 4
                                               Atlanta,  Georgia
ABSTRACT
  This paper presents a preliminary report on the on-going con-
trol of contaminated aquifers in karst topography. Over the past
several years, there has been a growing concern over the presence
of volatile hazardous chemicals in the karst formation cave sys-
tems in the Bowling Green, Kentucky area. By April 1985, con-
tamination and related chemical fumes warranted an "emergency
response action" by the U.S. EPA, Region 4. The management
of migration procedures under these conditions are discussed and
summarized. Dye tracing, geophysical studies, speleological car-
tography, exploratory drilling, down hole cameras and sampling
techniques have  been utib'zed to locate the primary  flow char-
acteristics of the  "Lost River" and its tributaries and to identify
the source(s) of contamination and mitigate its effects.
  Karst formations  present peculiar  and distinct management
difficulties. The  impact of leaking underground storage tanks
and illegal dumping  may be magnified by the motility of spilled
substances. Pollution source determination is especially proble-
matic.
  Many methods used in this emergency response action were
relatively  experimental with few  strong  precedents.   These
methods are comparatively analyzed for the benefit of  future
cleanup activities.

INTRODUCTION
  Karst topography, evident to some degree almost anywhere
carbonates are found at or near the surface of the earth and sub-
jected to  humidity and water movement, presents  unique and
perplexing difficulties to the professional involved in hazardous
waste management. By definition, karstic terrains are character-
ized by closed  depressions, i.e., sinkholes or dolines and uvalas,
caves and the prevalence of underground drainage. Although
these conditions can  be beneficial, as in the case of high capacity
groundwater accessibility (it is the closest thing in  nature to a
fresh water pipe), extreme liabilities also may be present. For ex-
ample,  the impact of  leaking underground storage tanks and
illegal dumping may be greatly magnified by the motility of lost
substances. While permeability in non-karst areas may be meas-
ured in ft/day, permeability in karst can be measured in miles/hr.
Discrete sources of groundwater recharge are numerous and var-
ied and the velocity of recharge so high that pollution source de-
termination can be especially complex.
           APPROXIMATE SCALE IN MILES

           1 LOST RIVER UVALA I
             CAVE ENTRANCE

           3 LOST RIVER RISE
                       Figure 1
                 Bowling Green Location
  Located on the textbook karstic terrain of south central Ken-
tucky, Bowling Green is the only city of such size (pop. 41,000)
in the United States to rest almost entirely over a cave system.
The Lost River, which is exposed in a uvala in southwest Bowling
Green, flows into the Lost River Cave  entrance and  has been
mapped for 4 miles. At this point, the river sumps and takes an
undetermined flow path through the St. Genevieve and  St. Louis
limestones underlying Bowling Green until it exits north of the
city at the Lost River Rise.
44   CONTAMINATED GROUNDWATER CONTROL

-------
  Fixed  facility  and  transportation related  spills,  deliberate
dumpings and undetected leaks, all associated with oil and haz-
ardous materials, have become daily occurrences in every major
city across the nation. Bowling Green is no exception. Its loca-
tion in a sinkhole plain and over an extensive cave system, makes
the city highly susceptible to groundwater pollution and to the re-
lated problems associated with karst areas.
  Many  contaminants can immediately  affect the  watershed
either through unobstructed flow into the cave system or through
rapid percolation through thin cover soils. In Nicholas Craw-
ford's earlier research, he showed that heavy metals and road
salts from urban detritus quickly entered the  groundwater sys-
tem,1 but volatile materials pose special problems.  As the sub-
stances flow through the subterranean labyrinth, they can volatil-
ize in the cave air space and ultimately find their way to the sur-
face.  The control of gases and vapors in the atmosphere usually
is  not attempted because natural dispersion dilutes  vapors and
gases rapidly. However, fumes rising from  cave systems may con-
centrate in areas of concern. If these fumes collect in a basement,
crawl space or simply in or near public areas, a potentially explo-
sive situation exists.

 Bowling Green's Problem
   The toxic and explosive fume problem in Bowling Green dates
 back to 1969, when explosive fumes were detected and believed
 to have caused  a residential  explosion.  In 1981, the residents of
 five homes in the Riverwood area  were evacuated  when fumes,
 rising into their basements and crawl  spaces, reached explosive
 levels.
   In 1982, benzene and  methylene chloride were detected in the
 Lost River Cave. Subsequently, the U.S. EPA, Region 4, initiated
 an emergency response action in March of 1983. Preliminary in-
 vestigations showed alarmingly high levels of volatile chlorinated
 hydrocarbons2  flowing  into the spring-fed Keith  Farm pond,
 which discharges into the Lost River drainage system. Dye traces
 pinpointed the source of the  contamination to be a series of leak-
 ing industrial underground storage  tanks.  The U.S. EPA treated
 1.9 million gal of contaminated pond water, and the responsible
 party removed the leaking tanks and excavated the contaminated
 soils.3
   Throughout 1984, toxic and explosive fumes plagued the homes
 in the Forest Park and Parker-Bennett School areas. In January
 1985, toxic and explosive  fumes were  detected in  the Dishman-
 McGinnis Elementary School. To date, fumes of this type have
 been investigated in over  100 homes and commercial buildings,
 two schools, one church, 18 drainage wells and 24 sinkholes.
   After an extensive public  health  evaluation of  the fumes by
 the Center for Disease Control, a health advisory was issued on
 the grounds that explosive levels existed and that the detectable
 fume levels of benzene,  toluene and xylene exceeded standards
 for  "non-occupational settings,"  At that time, the U.S.  EPA
 Region 4 Emergency Response Branch initiated a  "Superfund"
 emergency response to address the problems designated in the
 public health advisory.  Mitigation efforts were initiated to  re-
 duce the threat, identify the  source or sources of contamination
 and  define the  underground  drainage system controlling the
 transport of chemicals in the Bowling Green area.
upon the assumption that  continuous releases or intermittent
dumpings into the Lost River drainage system, i.e.,  sinkholes,
solution joints and fractures, have resulted in a chronic flow of
chemicals within the caves.
  The first  hypothesis is that, once in the  cave, the chemicals
volatilize; a volatilization that is accelerated by natural struc-
tures. Waterfalls and riffle areas provide an indigenous turbu-
lence which increases the rate of volatilization, increasing, in
turn, the  fume concentration within the cave atmosphere.  Dur-
ing high water flow,  the air space within the cave is decreased,
forcing the gases to the surface. This process  not only occurs dur-
ing high water flow, but also during normal cave breathing due
to changes in barometric pressure.
  The second hypothesis is that water-filled  passages,  forming a
natural underflow dam, trap floating chemicals  and restrict them
from flowing downstream. This flow restriction, in turn, causes
low density  and volatilize chemicals  to become concentrated in
these areas. This  frequently occurs  during  increased river  dis-
charge when heavy rainfall causes the river within the cave to rise.
Consequently, the floating chemicals and the fumes are pushed
upstream.
  As the water recedes, the floating chemicals adhere to the ceil-
ing and walls forming a "bath tub ring" within the system. This
bath tub ring phenomenon coats the surface of upstream perched
pools with floating chemicals.  The rise and fall of the  river level
results in a  repetitive deposition of  chemicals  within the cave,
leaving a stratified layer of aliphatic hydrocarbons and mud.
  Finally, it is proposed that contaminants have been dumped or
spilled in soils and areas underlain by perched water tables or
water pockets and are flushed into the Lost River Cave  system
when meteoric water  percolates through the soil or overflows a
perched water table. Perched water tables and water pockets are
associated with the pinnacle  nature of the limestone at the lime-
stone/soil interface (Fig. 2).
WATER POCKET
                                                 SOIL
                                                 (Terra Rossa)
                                                PERCHED WATER
                                                   TABLE
                                                (VARIABLE SIZE)
             SOLUTION CONDUITS
            FLOWING TO LOST RIVER
                           Figure 2
    Generalized Illustration of Pinnacle Nature of Limestone Showing
          Possible Perched Water Table and Water Pockets
 FUME OCCURRENCE HYPOTHESES
   Although not fully documented, there is an apparent direct re-
 lationship between increased reports of fumes and wet weather;4
 similarly, there have  been  fewer fume incidents  during  dry
 weather. Thus, the fumes appear increasingly manifest in periods
 of heavy precipitation. The  hypotheses which follow are based
  These hypotheses are not mutually exclusive. Investigative tech-
niques in verifying them and solving the resulting problems asso-
ciated with them have been varied. In this verification, geological
techniques were used in conjunction with sound environmental
studies and common hazardous waste site cleanup practices.
                                                                            CONTAMINATED GROUNDWATER CONTROL   45

-------
SEARCH FOR THE LOST RIVER
  The logical place to start this investigation  was to locate the
contaminated medium, i.e., the Lost River. In  this search, an in-
tegrated team approach based on multidisciplinary scientific sup-
port was utilized to address the diverse problems associated with
the fumes rising from the cave system.
  Excavation of sinkholes and crevasses initiated the physical ex-
ploration of the project. In the search for possible entrances into
cave passages,  sinkhole breakdown was removed with a track-
hoe or a backhoe and by hand. Excavations were made adjacent
to homes and buildings at identified channels from which fumes
were  reaching the surface. Excavations also were made within
sinkholes along the hypothesized Lost River flow path and where
historical accounts indicated the closing of cave entrances.
  Concurrent  with these  excavations, exploratory  drilling was
initiated.  However, with  few means of locating drilling sites for
cave passages other than topographic features,  the production of
exploratory wells at 10-ft intervals appeared to  be a  costly and
time-consuming method of cave discovery. Clearly, a method was
needed to locate drilling sites more precisely and economically.
This need led to the implementation of gravity survey techniques.

Gravity Survey
  Gravity surveys are utilized by, and in fact originated in, the oil
industry,  where there is great interest in  detecting  large under-
ground structures.  In microgravity surveys, which  include cor-
rections  for tidal effects, more  sensitive instrumentation and
more precise techniques,  it is possible to  detect very small and
finely localized variations  in  geologic structure.  In Bowling
Green, it was  anticipated that microgravity surveys  would de-
tect the presence of void space (or absence of solid rock) along a
given transect.
  Using a Lacoste and  Romberg  Model D  Microgal  Gravity
Meter, 10-ft stations  were measured  over thousands of  feet of
transects in the study area. The readings were plotted and large
anomalies were investigated by exploratory drilling.  It often was
found that anomalous  readings represented  unusually deep
troughs of soil between pinnacles of limestone  or  break down
zones  of possible cave collapse  (Fig. 3). Nevertheless,  micro-
gravity surveys proved to  be the most reliable  means of locating
void space or cave passages when no historic  or surface indica-
tions were available.
                                                         9
                                                       .„;
     2000   IBOOcoo  1*001200   ioootoo   «oo   400   zoo

                           Figure 3
   Representative Comparison Between Topographic (Solid Line) and
   Microgravity (Dotted Line) Surveys Showing Probable Breakdown
               Zone (A) and Small Solution Void (B)
Drilling
  Exploratory and core drilling, utilizing rotary water and forced
air hammer drill rigs, continued throughout the investigation to
verify  the  microgravity data  and to correlate  localized strati-
graphic columns. Many of the exploratory and core wells later
became monitoring wells used to record a wide variety of rele-
vant variables. In certain instances, monitoring wells were placed
by survey and drilled into the stream thalweg of cave passages,
eliminating  the  need for  lengthy  and  labor  intensive  cave
sampling and/or monitoring trips.
  Standard hydrologic methods, including dye tracing and piezo-
metric surface contouring, were used to locate primary flow char-
acteristics, while air monitoring was used to locate and character-
ize areas  of  fume contamination.  A 1.5-in.  downhole camera
with both downhole  viewing  (Lighted  Actual Viewing  Attach-
ment,  LAVA)  and sideviewing (Right  Angle Viewing  Attach-
ment, RAVA) capabilities also were used  to investigate the sub-
terranean environment.
  The investigation team occasionally experienced difficulty  in
distinguishing between solution voids and breakdown zones and
between true void space  and mud-filled cavities.  The downhole
camera allowed visual inspection of these features, ensuring cost-
effective and time-saving verification  of the presence of  cave
passages. Also, the downhole  camera confirmed the indications
of microgravity data and historic  accounts  that a large  cave
existed in  the vicinity of Creason Street and  Robinson Avenue
area.
  This discovery prompted drilling  of  a  new caliber. A 30-in.
core barrel drill rig  was employed to drill entrance wells  into
Robinson's Cave and Creason Cave. These caves were actually
one prior  to  a  1960s  road construction blasting accident which
collapsed  100 ft of road, effectively severing the original  cave
passage.
  Utilizing a team of experienced  spelunkers, detailed speleo-
logical investigations and complete surveys of the caves were in-
itiated. From these explorations it was determined that the Robin-
son's and  Creason Cave  system represented  a paleo-channel of
the Lost River,  allowing the  realization that the river was not
confined by the resistant Lost River chart,' but that it had broken
through into what Crawford calls a "Karst Sandwich" (Fig. 4).
 ST. CENEVIEVE
    LIMESTONE
                                                                             GRANULAR LIMESTONE
                                                                             (FOSSIL FRAGMENTS t COLITESI
                                        DOLOMITE
                          Figure 4
  Partial Section Showing Generalized Flow of Lost River into "Karst
               Sandwich" of Chert and Limestone
46    CONTAMINATED CROUNDWATER CONTROL

-------
  Unfortunately, the Lost River proper has yet to be found by
way of Robinson's Cave. The cave does, however, represent a
tributary conduit into the Lost River system contributing con-
taminants from perched water tables and/or soil percolation.
  Another cave  investigation was prompted by fumes rising to
the surface. In one residential crawl space and in an adjacent
storm sewer and drainage well, fumes reached explosive levels
frequently. The investigation team began hand excavation accom-
panied by continuous air monitoring in the drainage well. During
this apparently minor excavation, fumes peaked at levels  above
300 ppm and 100% LEL, prompting a more comprehensive ex-
cavation. Using spark-proof tools, an entrance to a new, hitherto
unknown,  cave was discovered.  This cave, Napier's Cave, be-
came the focal point of further study and investigation. Physical
exploration of this cave required positive pressure flushing of con-
taminant fumes, utilizing two high volume smoke  exhaust fans
blowing into the cave.  Exploration revealed  floating contam-
inants, wall staining and contaminated perched pools, all indi-
cating contamination transport at high and low water flows.
SAMPLING
  The ultimate success of determining locations where investiga-
tions could  be efficiently performed, of determining mitigation
procedures  and of initiating possible enforcement actions de-
pended on accurate and valid analysis of the chemicals involved
and of the vulnerability of the environment to their effects. This
information was obtained by extensive sampling and monitoring
of soil, air and water matrices characterizing the contaminants
and affected medias. Potential  and real effects on the environ-
ment, immediate and long-term risk to public health and to the
investigating team were established  and monitored.  Several
methods were employed.
  Air monitoring was performed with the use of real-time, direct-
reading, portable  instrumentation, -photoionization  detectors,
combustible gas indicators and oxygen meters. This highly mobile
instrumentation proved very convenient in screening cave atmos-
pheres and monitoring stations  over the large study area. How-
ever, instrumentation limitations were revealed. The HNU photo-
ionizer, Model PI  101, gave unreliable readings within the cave
environment and in the monitoring wells. This problem was due
to the cave  humidity and subsequent condensation on the ultra
violet light source.
  The GasTech, Hydrocarbon Super Surveyor, Model 1314, was
the most practical  portable instrumentation for use in cave en-
vironments.  It  enabled monitoring  for the parameters of total
organics, combustible gas and  oxygen with one instrument.  Its
light weight  and size, ease in calibration and relative insensitivity
to moisture  made  it the  preferred instrument for  speleological
monitoring.
  Once initial contaminant parameters were identified,  species-
specific instrumentation provided accurate quantification. Specif-
ically, monitoring  was  aided by the use of a Photovac, a pro-
grammed photoionization gas chromatograph. The Photovac was
placed within a crawl space in the Parker-Bennett Elementary
School. Scheduled samples were  automatically collected and
immediately  analyzed before and  during building occupation.
Generated chromatographs provided the specific qualitative and
quantitative  data utilized for public health considerations.
  The U.S.  EPA  Environmental  Response Team's mobile en-
vironmental unit (SCIEX) also was utilized in the study. The com-
puter-coupled tandem mass spectrometer, mounted  in a specially
adapted bus, provided real-time measurements of contaminants
in ambient air from the atmosphere, monitoring wells and within
buildings and homes. This versatile field laboratory was not lim-
ited  geographically and provided  quality assured  quantitative
and qualitative analysis in minutes. It proved to be the most effic-
ient  and accurate means of screening air samples and identify-
ing compounds.
  Tedlar  air sample bags  were used  for capillary soil  gas
sampling and occasionally for atmospheric sampling when sub-
sequent qualitative  laboratory analysis was indicated. The soil
gas technique was employed at industrial sites to establish a corre-
lation between volatile chemicals associated with those sites and
those identified  rising from the Lost River Cave system,  and
where  fumes were detected  in the absence of any surface  evi-
dence of escape paths. In addition to real time air monitoring and
sampling, Tenax collector  tubes were employed, while subse-
quent desorption and analysis provided time weighted average
concentrations for public health considerations.
  Soil,  sediments and water samplings completed  the charac-
terization of chemicals involved and the assessment of this com-
plex problem. Surface and cave sediment samples were collected
from drainage ditches, cave walls and cave  stream beds to iden-
tify contaminants and to establish correlations between industrial
releases of chemicals and cave contamination. Soil samples also
were collected, specifically as a means of documentation during
underground storage tank excavation.
  Due to the scarcity of surface streams in the Bowling Green
area, surface water samples were limited to the Lost River Cave
Entrance and Rise. However, groundwater samples were collected
directly from the caves, monitoring wells and potable wells. Cave
stream and  monitoring  well samples  helped to verify contam-
inant flow paths, while potable well samples were useful in pub-
lic health determination and helped to  locate  contaminated
perched water tables.


MITIGATION
  A  priority throughout the  investigation was mitigation  of
immediate and potential threats to health and safety. Prompted
by the  1983 U.S.  EPA identification of leaking underground
storage tanks as a source of Lost River contamination, an exten-
sive  search was performed  to  locate  all storage  tanks in study
areas. Data were compiled to indicate the integrity of the tanks.
Several underground storage tanks were tested and removed or re-
paired if leaks were detected.
  Similarly,  the discovery of one residence that relied entirely on
wellwater and the subsequent determination that this water was
contaminated, led to an extensive data search and public survey
in an effort to identify all water wells in the study area. Residents
on contaminated potable water supplies were connected into the
city water system.
  The desired culmination of mitigation with regard to hazardous
fumes is the  ultimate elimination and removal of those fumes and
their sources. However, prior to the identification of the chem-
ical  nature  of the fumes and their  precise sources, the most
immediate and obvious mitigation procedure was the venting of
homes and public buildings. Using exhaust fans in crawl spaces,
basements and  cave entrances, fumes were dispersed into the
atmosphere instead of being allowed to concentrate in closed and
inhabited areas.

CONCLUSIONS
  As a part of an on-going study, the following  conclusions on
particular aspects of the investigation and mitigation of contam-
inated aquifers in karstic terrains are tentative. Based on informa-
tion  derived from this study, the body of this paper might pro-
vide  guidelines for future investigations of  hazardous waste gen-
eration and handling in karst areas.
                                                                           CONTAMINATED GROUNDWATER CONTROL    47

-------
 Ventilation of fumes
   Venting provides a convenient short-term means of lowering
 or eliminating fume  concentrations  in  living  areas  and public
 buildings.  Venting is, of course,  a short-term solution to a long-
 term problem.

 Monitoring
   Clear-cut advantages were noted between instruments applied
 to various monitoring tasks. The GasTech proved  superior in
 speleological monitoring due to its relative insensitivity to mois-
 ture and multiple function capacity. Single function instruments
 used in conjunction  with multi-function instruments provided
 data checks.  The  HNU with its increased sensitivity was ideal
 for surficial monitoring of organic  vapors. Similarily, the ex-
 plosimeter's specialized function provided a comparison  and
 verification of data accuracy.

 Microgravity Survey
   Not overlooking topographic features and historic information
 of an area,  microgravity surveys were  the most effective tech-
 nique utilized in this  study for the remote sensing of karstic  fea-
 tures.

 Smoke Testing
   Although smoke testing met with little success in this investiga-
 tion, we believe that a very dense paniculate and aromatic smoke
 with an efficient injection system could be an effective tool in
 locating small or hidden cave openings.

 Drilling
   Although  rotary water drilling initially was used  as a safety
 precaution to prevent spark ignition of cave  fumes, it is now
 thought that forced air rotary and hammer drilling are acceptable
 methods in all but the most dangerous areas; the forced air would
 flush fumes away from any spark generation.

 Tank Testing
   As national attention to the  Leaking Underground  Storage
 Tank Program and hazardous material handling grows, special
 attention should  be  paid  to karst  areas.  Precision  testing of
 underground  storage  tanks identified  leaking tanks. Their re-
 moval or repair possibly may have eliminated the source of con-
 tamination and the subsequent  fume problem in a  Lost  River
 tributary.

 Chemical Assessment
   The only substance releases whose consequences might be min-
 imized in karstic terrains could be those of acids which rapidly
 neutralize  in contact  with limestone.  In all other instances, oil
 and hazardous substances  releases  in karst areas  can rapidly
 migrate and generally pose more complicated and variable prob-
 lems than in non-karst regions.
  These studies further concluded that all of the water wells in
the study area, with one exception on its boundry, were contam-
inated with a variety  of  hazardous substances. Some of these
contaminants were not present in the Lost River; hence it was in-
ferred that  these contaminants were trapped  in perched water
tables, migrating down well into the phreatic zone.
  As general methods and specific techniques of hazardous waste
management continue to develop and improve across the country,
special attention should be paid to environmental influences on
hazardous  waste dispersion. Karst areas,  in particular,  require
specific guidelines  and possibly  regulative  measures governing
hazardous  material generators, transporters and emergency re-
sponses.

FOOTNOTES
I. Crawford, N.C., "Sinkhole Flooding Associates with Urban Develop-
  ment upon Karst Terrain: Bowling Green, Kentucky," in B.F. Beck,
  ed., Sinkholes: Their  Geology, Engineering, and Environmental Im-
  pact, A.A. Backema,  Rotterdam,  1984, 283-292.
2. U.S. EPA, "Superfund Cleanup,  Keith Farm Pond, Bowling Green,
  Kentucky," Vols.  I &  II,  1983.
3. Ibid.
4. This section owes much to the formal work  of, and  informal dis-
  cussions with, Nicholas C. Crawford; see footnote I, above.
5. The relative position of the Lost River chart in the stratigraphic col-
  umn is subject to  some disagreement. The Geological Survey places
  the Lost River chart at the base of the St. Genevieve formation; how-
  ever, authorities in the field place the Lost River chart in the Horse
  cave member of the St. Louis formation. Cf. Palmer, N., A Geologi-
  cal Guide lo Mammoth Cave National  Park,  Zephyrus Press, Tea-
  neck, NJ,  1981, 196; Pohl, E.R.,  "Upper Mississippian Deposits in
  South Central Kentucky," Kentucky Acad. of Sciences, Trans., 31,
  1970, 1-15. This placement has  been further confirmed through per-
  sonal communication with Nicholas Crawford.

REFERENCES
1. Beck, B.F., ed.. Sinkholes: Their Geology, Engineering and Environ-
  mental Impact, A.A. Backema, Rotterdam, 1984.
2.  "Bowling Green Toxics," U.S. EPA Report, Vol. I., 1985.
3.  Crawford, N.C., "Toxic  and Explosive Fumes Resulting from Con-
  taminated Groundwater Flow through Caves Under Bowling Green,
   Kentucky," unpublished  research proposal submitted  tothe City of
   Bowling Green, KY, April 1985.
4.  Palmer, A.N., A Geological Guide to Mammoth Cave National Park,
  Zephyrus Press, Teaneck, NJ, 1981.
5. Pohl, E.R., "Upper  Mississippian Deposits  in South Central  Ken-
  tucky," Kentucky Acad. of Sciences, Trans.,  31, 1970, 1-15.
6. U.S. EPA, "Superfund Cleanup,  Keith Farm Pond, Bowling Green,
  Kentucky," Vols.  I &  II,  1983.
7. Tejada, S., "EPA Goes Underground at Kentucky Superfund Site,"
  EPA J. 2. July/Aug.  1985, 26-27.
48    CONTAMINATED GROUNDWATER CONTROL

-------
                        Use  of  Low Flow Interdiction Wells  to
                  Control Hydrocarbon Plumes  in Groundwater

                                          John H.  Sammons, Ph.D.
                                       Ninth U.S. Coast Guard District
                                                Cleveland,  Ohio
                                         John M. Armstrong, Ph.D.
                                           The Traverse Group, Inc.
                                             Ann Arbor, Michigan
ABSTRACT
  This paper presents a case history and discussion of a situation
where the initial hydrocarbon contamination occurred in 1969, re-
mained undiscovered until 1980 and since has resisted massive
efforts to  even adequately define the situation. The  paper de-
tails the installation of low-flow wells across the plume to block
flow of contaminants. This interdiction resulted in a serendipi-
tous activation of the naturally occurring soil microbes in the
sediments down-gradient from the well field. The net result of the
interdiction and microbiological activation was to  effect a rapid
and dramatic reduction in the hydrocarbon contaminants in the
plume down-gradient from the interdiction well field.

INTRODUCTION
  Hydrocarbon contamination of groundwater is rapidly becom-
ing one of the most troublesome and costly problems to be con-
fronted  by both the regulators and the regulated community.
Virtually every day, an article is  written announcing another
underground water supply that has become  contaminated  and
that it appears virtually impossible and prohibitively expensive
to contain and completely decontaminate the aquifer.

SITE HISTORY
  In July 1942, the United States Navy established an Air Sta-
tion at Traverse City, Michigan, a small and isolated community
located in the northwestern section of  the Lower  Peninsula
(Fig. 1). The purpose of the Air Station was to conduct highly
classified research and  development of pilotless  drone aircraft
that could be remotely guided  by television to the target from
chase aircraft. This research effort was continued until 1944 when
it was suspended. When the war  ended, the Air Station  was
turned over to the United States Coast Guard (USCG) to serve as
a major Search and Rescue base for Lakes Superior, Huron and
the upper portion of Lake Michigan.

Coast Guard Operation
  Coast Guard operations at the site commenced in  1944 and
have continued until the present with a mix of rotary-wing and
fixed-wing aircraft. A  review of operations revealed nothing
extraordinary or unusual had been  reported although historical
records of the station are sketchy.

History of Problem
  In 1979, during the removal of two fuel farms preparatory to
the  installation of a new system, there was some indication that
leaks had occurred. The only soil contamination was in the Jet
Fuel (JP-4) storage area. These soils were removed and disposed
of under the direction of the State of Michigan Department of
                         Figure 1
  Location of East Bay Township and U.S. Coast Air Station in Michigan
 Natural Resources (MDNR). This area was located some 1,500 ft
 upgradient and to the north of the area that was ultimately impli-
 cated as the "geographical origin" of the plume. The Aviation
 Fuel (115/145) fuel farm, located  immediately  adjacent to the
 "geographical origin" of the plume,  was excavated at the same
 time with little indication that any unusual occurrences had taken
 place. There was some odor noted in the soil,  but laboratory
 analyses did not confirm gross contamination, so no soil was re-
 moved.
  In 1979 and  1980, residents in the Avenue E area of the Pine
 Grove Subdivision of East Bay Township complained to the local
 health department that  their  wells were producing  discolored
-water, were foaming and had bad tastes and foul odors. The first
                                                                     CONTAMINATED GROUNDWATER CONTROL   49

-------
 residence reporting the problem is located 1,200 ft to the north-
 east of the Coast Guard Air Station (CGAS). At that time there
 was no explanation for the contaminated wells and, unfortunate-
 ly, the health department did not test for  any of the possible
 organic contaminants. Later in 1980, the MDNR  did a limited
 hydrogeologic study in the area and concluded that the source of
 the contamination was from some unspecified site on the CGAS.'
 In May 1982, the government was notified of the findings and be-
 gan an attempt to elucidate the problem.

 COAST GUARD ACTIONS
   The U.S. Geological Survey (USGS) was retained in June of
 1982 to do a thorough hydrogeological study of the area to define
 the plume of contaminants and to find the source. By April 1983,
 the USGS had determined the direction and velocity of ground-
water flow through the area and  had tentatively identified  the
boundaries of the plume (Figs. 2 to 5). The USGS was not able to
determine the source of the contaminants. They did, however,
conclude that the majority of the contaminants identified were re-
lated to components in fuels with  some chlorinated compounds
also present.1

Detailed Studies
  In November 1983, the Coast Guard contracted with the Uni-
versity of Michigan (UM)  for  a scientific study and  feasibility
study of the site. Objectives for these studies  included a tem-
poral analysis of the plume  and a positive identification and loca-
tion of the contaminant origin(s). The final  report' included a
description of the time variation of contaminants and adsorptive
characteristics  of soils,  including  side-by-side  comparisons of
                                                           n  n w  n to w
                             WEST ARM GRAND
                              TRAVERSE BAY
                                                                             EAST ARM GRAND
                                                                               TRAVERSE BAY
                                                                       Principol
                                                                       study area
                                                                         EXPLANATION
                                                                -roo-LINE OF EQUAL ALTITUDE OF
                                                                       LAND SURFACE--Interval
                                                                       20 feet NGVO of 1929
                                                                  10-WATER-TABLE  CONTOUR--
                                                                       Shows altitude of water table.
                                                                       Contour interval 10 feel.
                                                               	»- GROUND -WATER FLOW--Arrow
                                                                       indicates direction of flow
                                                               	GROUND-WATER DIVIDE
                      Base from U S Geological
                      Survey I 62.SOO quadrangles
                                                         Figure 2
                                             General Direction of Groundwatcr Flow
50   CONTAMINATED GROUNDWATER CONTROL

-------
water and soil concentrations found by MDNR, USGS and UM;
a description of the quality assurance program for collection and
analysis of groundwater samples; a preliminary risk assessment
for chemicals found in the groundwater and a description of the
effort to numerically model the behavior of  the groundwater
plume, a description of the groundwater sampling program and a
preliminary  remediation analysis that identified several cleanup
alternatives.
  During this time, the USCG was pursuing an internal investiga-
tion which resulted in the discovery of an Aviation fuel spill inci-
dent. In November or December of 1969, a flange in an under-
ground pipe line under a 115/145 high octane Aviation Gasoline
fueling station failed, resulting in the loss of approximately 2,000
gal of product over a 12-hr period.
  The Traverse Group Inc.  (TGI), an Ann Arbor-based multi-
disciplinary consulting firm, was contracted by the USCG to do a
detailed feasibility study that included  the design,  construction
and operation of groundwater treatment plants (involving carbon
adsorption and air stripping techniques) and to continue a sam-
pling program for selected wells to monitor the plume. This part
of the investigation was completed in February  1985." TGI was
then asked to continue the project to include the following tasks:
• Selection of a specific cleanup technology
• Design, construction  and operation  of an interdiction well
  system
• Design and installation of monitoring wells
• Design, construction,  operation and evaluation of a full-scale
  carbon adsorption system
• Assembly,  installation and evaluation of an advanced rotary
  air stripping and vapor incineration system
                                                          Figure 3
                                  Water Table and Direction of Groundwater Flow on Apr. 5-7, 1983
                                                                          CONTAMINATED GROUNDWATER CONTROL    51

-------
  In February 1985,  new hydrocarbon  contamination  was dis-
covered at the JP-4 fuel farm south of the Hangar/Administra-
tion building (HAB). The four fiberglass underground storage
tanks at the station were tested; three were found to be leaking
and were removed. In March 1985,  routine sampling indicated
the presence of high levels of benzene, toluene and lighter hydro-
carbons in  monitoring wells  along the south boundary fence of
the USCG  property upgradient  from the  JP-4 Fuel Farm. The
origin of this contamination  was not then, nor is it now, under-
stood since these wells are now  clean. One possible explanation
for the contamination in these wells was a spill of approximately
2,000 gal of Jet-A at the Republic Air Lines fuel farm some  1,800
ft directly  upgradient from  the south boundary of the USCG
property.  What  cannot be explained  is the high contaminant
levels when the wells were sampled in March 1985, and the current
absence of contaminants in the wells.

Plume Characterization
  Much of the investigative effort  focused on the location and
characterization of the groundwater plume. All agencies involved
installed and monitored observation wells for this purpose. Due
to temporal variations of the  contaminants and differences in
analytical  techniques, not all results were in complete agreement;
however, there is sufficient similarity to delineate locations and
concentrations of the plume's major constituents.
                                                                            \    EXPLANATION
                                                                            *l\   • «VU iOUtiOi
                                                          Figure 4
                                  Water Table and Direction of Groundwater Flow on July 19-20, 1983
52    CONTAMINATED GROUNDWATER CONTROL

-------
                                                                              LOCATION  OF
                                                                              INTERDICTION WELLS
                                                                                                       EXPLANATION
                                                                                                — 590— WATER-TABLE CONTOUR--Showi
                                                                                                     olliludc ot limuloled -aw table
                                                                                                     Aoril 1383 Contour interval I loot.
                                                                                                     NCVO ol 1929
                                                                                               	»-GROUND-WATER FLDW--Arrow
                                                                                                     indtcotci direction of fto*
                                                                                               	  AIR STATION BOUNDARY

                                                                                                :•:•:•:•  PLUME OF CONTAMINATION

                                                                                                     O      500    ICOOFEET
              Bole odapfrrd Irom U.S Cootl Guard mop and
              mopi bv Gojrdic-Frotcr and Atiocialtf. Inc
                                                                                                        IOO   200 METERS
                                                            Figure 5
                                    Water Table for Apr. 1983, Simulated by Groundwater Flow Model
  The UM study3 identified benzene and toluene as the compon-
ents in the plume presenting the greatest health risk.  The largest
concentrations occurred in the vicinity of the Hangar/Adminis-
tration Building (HAB) at the geographical  head of the plume
(Figs. 6 and 7) although significant amounts of some  compounds
such  as  benzene were found at some distance  downgradient.
Other chemicals also were found in the plume but at smaller con-
centrations and reduced distributions.3
  The USGS provided a  detailed map of the plume's size  and
location (Fig. 1). From the HAB, the plume follows groundwater
flow to the northeast and off the base, passes under an industrial
park and turns slightly north,  narrowing as it passes  underneath
Parsons Road and widening out again  under Avenue E (Fig. 5).
The plume is approximately 4,300 ft long and varies from 180 to
400 ft in width. Its vertical dimension ranges from 25 to 80 ft.
  Small  concentrations of benzene and toluene  have been de-
tected in the water of East Bay. The USGS2  reported maximum
values of 20 /ig/1 benzene  and  3.1 /tg/1 toluene aproximately
330 ft from shore. The vast majority of subsequent measurements
by the TGI have been less than those found by the USGS with the
majority being below the detectable limit.

Soil Contamination
  Both the UM study and the USGS study reported numerous
measurements of organics in the soils  at the Air Station, again
with discrepancies due to analytical, temporal and  areal varia-
tions. The UM study found maximum concentrations of 25.4 jtg/1
benzene, 27.6 /jg/1  toluene and 229  /itg/1 xylene.  Analyses were
done for seven other hydrocarbons with negative results. Analysis
of soil borings indicates that much of the organic material is re-
tained in the soil in  a 6 to 12 in. thick layer in the capillary zone
immediately above  the water table.  It  has been suggested that
this zone is slowly  leaking organics  into the groundwater over
time and is thus serving as a contaminant source for the plume.3

Proposed Cleanup
  After consideration  of the various  long-term treatment  or
cleanup options available, it became clear that the most logical
first step in any remediation action program would be to decrease
or stop  the further  movement of contaminants off U.S. Coast
Guard property.
  This option was judged to have several advantages:
• Reduce any possible increase in risk to human populations that
  may have been related to fuel based contaminants present in
  the groundwater.
• Promote reductions  in  contaminant  concentrations  in  the
  groundwater either by dilution or possible biodegradation of
  fuels by indigenous microbial consortia present in  the subsur-
  face soil-water  system.
• Provide a better  opportunity (e.g., more time) to efficiently
  select and design appropriate method(s) for dealing with the
  contaminants present in the geographical origin of the plume.
                                                                             CONTAMINATED GROUNDWATER CONTROL    53

-------
                      K9S.O
   I
     ^
   /r-   »S20   »SI7
  H&«ix
   
-------
system to withdraw the minimum amount of water yet still cap-
ture the contaminants passing through the  aquifer so that no
contaminants,  or at least a significantly reduced amount, were
leaving government property.
  After reviewing several of the USOS groundwater model runs,
the team selected a well field configuration shown in Figure 8.
Seven wells were located laterally across the plume in the east-
northeast area  of the Air Station. These wells were 6 in. auger-
drilled wells with full 10-slot stainless  steel screens running from
the top of the  aquifer to the clay confining layer at the bottom
(Figs. 5 and 8).
  The system initially was constructed  with  all elements  above
ground. In 1985, the interdiction system was modified by install-
ing pitless adapters and providing a heated building to protect the
manifold and control systems from the extreme cold experienced
at this site. The water produced from  the interdiction wells is
piped to a carbon treatment system consisting of four 20,000-
Ib carbon reactors. The carbon reactors  were specified to reduce
the levels of benzene and toluene in the water to less than  1 /ig/1
as  measured using  headspace technique on  a HP5710A  Gas
Chromatograph.
  The USGS model gave clear indications that these flow rates
would produce closure of the equipotential lines at the interdic-
tion well line. One difficulty was that the two-dimensional nature
of the model did not yield any prediction on the vertical move-
ment of water at or  near the individual well locations. However,
because of the uniform nature of the saturated zone and the rapid
mean field velocity of the aquifer, which was measured at 5 ft/
day, it was believed that vertical movement in a fully screened
well at 15 gal/min would be sufficient to capture contaminants
flowing in the lower areas of the aquifer.
                           Table 1
         Specifications for Groundwater Interdiction Systems
North Interdiction Field
Date Wells Installed:
  Wells ID 1-ID6      Feb. 26-Feb .28,1985
  WellID-7          Mar. 22, 1985
Date of Water collection
  system construction: Mar. 19-Apr. 18,1985
Date Pumping
  commenced:        Apr.  19, 1985
Well and Screen Data:

                       North Interdiction
                        (6 in. diameter)
                              Table 2
                Monitoring Wells for North Interdiction
We) 1 *
Flow cpm
Top of
screen
(below
ground)
dottom
of
screen
uel 1
Dump

[ bottom
of pump
Below ;
ground ',
level) 1
.' ID-1
; is


13


53




43




10-2 i 10-3
; is : is


13 ' 13


53 : 43




43 ; 33



;
10-4
15


13


»3




33




10-5
15


13


43




33




ID-6
15


13


43




33




[0-7 \
25 ;


13 :


58




48 :




  Five full screen wells installed—Mar. 20-27, 1985
  Downgradient of North Interdiction Field on Mar. 20-Mar. 27, 1985
  Four-inch diameter wells extending to clay liner beneath the aquifer
Uell «
Depth below
well screen to
bottom of well
screen { top of
water at 13'
below ground
level)
Vertical loca-
tion of sample
Dumps in we 1 I s .
feet below
ground
Sample pump HI
HZ
«3
*4
»5
rci-i


n'-53'









15'8"
25 '8"
35 '8"
«5'8"
54 '5"
TGI-2


13'-69'









14'10.5"
24'
40'
54'
66'9" ;
roi-3


12'-68'









14 '6"
24'
41'
57'
68'
rci-4


n'-6S'









5'5.5"
8'
I1
4 '
6'
rci-s


13'-54'









14'10"
i?' :
33' :
43' :
53'5.P;
 TGI-2 sample well located 50' downgradient from ID-3 on a line parallel to the flow lines of the
 plume drawn NE from ID-3. TGI-3 located 50' downgradient from TGI-2 on a line parallel to
 the flow lines through ID-3 and TGI-2. TGI-4 located 90' northwest of TGI-3 on a line through
 TGI-3 perpendicular to the flow lines. TGI-1 located 180' northwest of TGI-3 on a line through
 TGI-3 and TGI-4. TGI-5 located at the end of a line drawn perpendicular to a line through
 TGI-3 and TG1-4 starting at a point 90' SE of TGI-3 running 50' NE.
   In most situations, contaminants such as benzene and toluene
 are found more predominantly in the upper areas of the aquifer.
 However, in this  situation the original spill had been  in  the
 ground so long that visible product was no longer present and
 the product not contained in the interstitial area of the capillary
 zone had solubilized and had mixed downgradient so that it was
 present throughout the vertical cross-section of the aquifer. Thus,
 placement of the pump in the well had to be in conjunction with
 a full screen configuration to capture contaminants from all ver-
 tical locations.
  A second consideration in the design was the monitoring of the
 interdiction system to  obtain data on  the effectiveness of  the
 blocking  of contaminant flow. Since the area of the  aquifer
 downgradient from the interdiction system had been contam-
 inated for several years, the monitoring wells must yield  data in
 sufficient detail to prove any decrease in contaminant levels after
 the "valve" had been closed.
  A sampling network of five  wells was located downgradient
 from the interdiction well system (Fig. 9). The wells were  located
 outside the zone of influence of the interdiction wells.  Sampling
 was done using a dedicated "Well Wizard" sample pump system.
 Groundwater samples were collected following standard field pro-
 cedures for the collection of volatiles and analysis was done with-
 in a few  hours of collection, always on the same day that  the
 sample  was collected. Five sample pumps were located in each
 monitoring  well at the top  and bottom of the saturated zone
 and at three equidistant points (Table 1).

 RESULTS
  The results downgradient from the interdiction  system have
been satisfying to date. Toluene and benzene levels in the down-
gradient monitoring wells were monitored in one of the  five wells
(M2/TGI2)  on a daily basis for the first 8 weeks and on a bi-
weekly basis thereafter. The  other four wells are monitored once
a week.
                                                                             CONTAMINATED GROUNDWATER CONTROL     55

-------
  Since benzene is the more mobile of the two constituents of in-
terest and has  the higher solubility, it should be slower than
toluene to respond to the interdiction. In addition, there was
speculation that some biodegradation by selective demethyoxyla-
tion might be seen and be reflected in  the benzene and toluene
concentrations  found in the downgradient wells. A considerable
amount of data have been collected which allow trend  observa-
tions to be made regarding the fate of benzene and toluene in the
downgradient area (Figs. 8 to 10).
  Figure 8 shows the variations and trend of benzene concentra-
tions as a water column average by month since the interdiction
field was activated on Apr. 15,  1985. Figure 9 provides  the same
information for toluene. These  data are from Well M-2 (TGI-2)
located immediately downgradient and outside the zone  of influ-
ence of the  interdiction well  field. This well logically would be
the first monitoring well to provide data on  the effectiveness of
interdiction. Although microbiological activity would  not be sep-
arated from hydraulic blocking, toluene levels decreased from a
baseline level of 10329 jig/I to less than  10 /ig/1 in  approximately
 100 days.
  Figure 10  shows benzene  and  toluene levels plotted against
time from April to December 1985. Since this figure shows the
variation of the two compounds, it could substantiate the micro-
biological activity. The decrease in toluene levels, together  with
the concomittant rise in benzene  between May and July, could
be attributed to the demethoxylation of toluene to  benzene. Both
compounds then began a steep downward trend with  the toluene
being the first  to approach non-detectable levels  in August. As
can be seen, the benzene reached its lowest level during October.
  The increase  in benzene levels since October is attributed to the
appearance of a red slime that plugged  pump screens  and coated
the inside of the piping systems,  resulting in a reduction of the
hydraulic capacity of the system. The groundwater in  this aquifer
is high in iron  content, and  the  substance has been tentatively
identified as oxidized iron compounds  and  complexes, mineral
deposits and biomass. Microbial action is believed to play a signif-
icant role in the dissolution  of these solids in the aquifer and
subsequent precipitation within the system. The slime is complete-
ly soluble in  an acidic solution, and gamma  logging of the wells
does not indicate the presence  of any clay layers that  could be
contributory. Obviously this situation must be better  understood
and controlled  or the field will lose its effectiveness as may be in-
dicated by the gradually increasing benzene levels.
  The USCG has an active project with  the U.S. EPA Robert S.
Kerr Environmental Research  Laboratory (RSKERL)  in Ada,
Oklahoma to evaluate the soil and water chemistry as well as the
microbial consortia present at the site. The results of these studies
should provide a mechanism to control the slime and  elucidate
the biodegradation phenomenon.
  While monitoring of the interdiction system continues, there al-
ready is substantial evidence that  the desired effect was achieved
in terms of contaminant containment. Toluene has decreased to
virtually non-detectable levels and, while it continues to be vari-
able, benzene concentrations clearly demonstrate a downward
trend.  The benzene variability may be due in part  to some, all or
a portion of the slime problem  discussed earlier. Biodegrada-
tion variability, the effect  of recharge and the higher  mobility of
benzene  in the groundwater  system also may contribute to the
variability of the benzene.


CONCLUSIONS
  Based  on the data collected over the past 9 months, it appears
that, for well defined plumes,  low flow interdiction is a viable
alternative to intercept and block contaminant flow.
                                                                                         HELL M-2
      Hit   a*   ui  JJ<   4J.I  U  OT'   Kl  •>•   OK
      (.:«

                           Figure 8
           Benzene Concentration as a Function of Time
      iiv   *•   wt  JLM  juf   »4   ari  XT   «*   ore

                          Figure 9
           Toluene Concentration as a Function of Time
     ST   **»    ui   JU<   all  14*   apt   XT   4*   OCC

                          Figure 10
      Benzene and Toluene Concentrations as a Function of Time
  Careful attention must be paid to the  mechanical as well as
contaminant systems. The appearance of  the biomass is a good
example of this since no homeowners in the area using wells have
reported a problem with slime.
  The possibility of triggering increased microbial activity is a
potential factor to consider.
  Particular attention must be given to the vertical  distribution
of contaminants since withdrawal volumes may be significantly
different for new plumes as compared to old plumes.
  Minimum volume requirements as dictated by available process
water disposal  options  are a major  determinant  in  interdiction
system design.
 56
       CONTAMINATED GROUNDWATER CONTROL

-------
DISCLAIMER
  The opinions or assertions contained herein are the private
ones of the writers and are not to be construed as official or re-
flecting the  views of the Commandant or the Coast  Guard at
large.

REFERENCES
1.  Sibo, K., "Groundwater Investigation of East Bay Township," Un-
   published, Mar. 3, 1982.
Twenter, F.R., Cummings, T.R. and Grannemann, N.G., "Ground-
water Contamination in East Bay Township, Michigan," U.S. Geo-
logical Survey Water—Resources Investigations Report 85-4064, 1985.
Rossman, R., Rice, C.P., Hartung, R., Simmons, M.S., Armstrong,
J.M. and  Wright, S.J.,  "Scientific  Study and Feasibility Study
Groundwater Contamination at Traverse City, Michigan,"  Unpub-
lished, The University of Michigan, Ann Arbor, MI, Mar. 1985.
"TGI Final Report, Phase I," Unpublished, Apr. 1985.
                                                                            CONTAMINATED GROUNDWATER CONTROL    57

-------
                 Computer  Groundwater Restoration  Simulation
                                 at  a  Contaminated  Well Field

                                            Shih-Huang Chieh, Ph.D.
                                        Ecology and Environment, Inc.
                                                Buffalo, New York
                                            Jeffrey E. Brandow,  P.E.
                               State Department of Environmental Conservation
                                                Albany, New York
ABSTRACT
  The most obvious and direct  means of contaminated aquifer
rehabilitation  is to remove the contaminated water by pumping
from recovery wells. Factors affecting the technical and economic
feasibility of this procedure include the number of recovery wells,
the rate and duration of pumping and the method used to dispose
of  the contaminated  water.  Since the aquifer rehabilitation  is
assumed  to be the reverse of the contaminant spreading process,
the theory and models that have been developed for predicting the
spread of a contaminant should provide a basis for predicting
contaminant withdrawal.
  The drinking water of western  Vestal, New York, is supplied by
Water District 1, which consists of wells  1-1,  1-2 and 1-3. Well
1-1, located on the bank of the Susquehanna River about 400 ft
from the Endicott-Vestal Bridge, is contaminated with volatile
organic compounds. Currently, a remedial investigation/feasibil-
ity  study is being conducted at the  site. The purpose of the
remedial  investigation/feasibility study is  to locate the contami-
nant source, to define  the extent of the contaminant plume and to
evaluate the remedial alternatives. As a practical tool to assist in
the feasibility study, a  two-dimensional finite element model is be-
ing applied to this site. First, the model is calibrated against field
data. Then the model  is utilized to simulate the movement of the
contaminant plume under the operation of a recovery well. The
model results  will be  used to evaluate and refine the  remedial
alternatives.

INTRODUCTION
  A groundwater contamination problem due to toxic organic
chemicals exists in Water District 1 of Vestal, New York. In early
1980, organic chemicals were detected in a municipal well in
Vestal. As a result of  this finding, all wells in Vestal were tested
for synthetic organic compounds. Well 1-1 of Water District 1
was found to contain large quantities of organic compounds. This
triggered  an assessment study of the potential of the pollutant to
contaminate other water supply wells in the district. Currently, a
remedial investigation/feasibility study is being conducted at this
contaminated  well field site. As a result of the remedial investiga-
tion, the contaminant  source areas were identified and the extent
of the contaminant plume was defined.1'2
  A feasible alternative to cleanup of the contaminated well field
is to remove the contaminated groundwater by pumping from
recovery wells. Since the aquifer rehabilitation is assumed  to be
the  reverse of the contaminant spreading process, the theory and
models that have been developed for predicting the spread of a
contaminant  provide a  basis  for  predicting contaminant
withdrawal.
  A two-dimensional finite element model is applied to this con-
taminated well field site to simulate the transport of the pollutant
under a recovery well is being operated.  The model utilized a
quadrilateral element and the Crank-Nicholson  method  for
calculation of the time marching scheme. The model result can be
used to evaluate the effect of the recovery well and to refine the
remedial alternatives.

DESCRIPTION OF THE SITE
  The drinking water for most of western Vestal is supplied by
Water District 1, which consists of well I-I and water supply welk
1-2 and 1-3. Well 1-1, located on the south bank of the Susque-
hanna River about 400 ft from the Endicott-Vestal Bridge, is con-
taminated with volatile organic compounds.  Contamination was
detected first in early 1980; since spring 1980. the well has been
pumped directly into the Susquehanna River. At present, well 1-2
is supplying the Vestal drinking  water;  well 1-3  is serving as a
backup supply. Because the water in well  1-3 is highly corrosive
and well 1-2 has a limited capacity, it is important that well 1-1 be
restored, especially  to meet anticipated future peak demands for
Water  District 1 and other interconnected districts.  Figure 1
shows the site of the contaminated well field and the groundwater
table contours  under wells 1-1 and 1-2 while in operation. The
groundwater table contours were developed from field measure-
ments.'
  The area immediately surrounding  well 1-1 is within the Sus-
quehanna River floodplam and consists of marshland, wooded
areas and commercial and residential land use areas. The river oc-
cupies a bedrock  valley filled with glacial deposits ranging in size
from clay to gravel. The bedrock surface, situated between 130
and 160 ft below ground level, is overlain by a productive aquifer
consisting  of  about 40  ft of permeable sand and gravel. The
aquifer is overlain by 120 ft of poorly permeable clays and silts.
Within these deposits are locally permeable pockets of material
which apparently have acted as contaminant pathways between
the ground surface and the producing aquifer.
  The high rate of pumping from this aquifer (approximately 1.5
million gallons daily, including the pumping of well 1-1 to waste)
has induced organic solvent movement toward well 1-1 from the
source area.  The source  area is an industrial park containing a
variety of industries. Volatile organic levels exceed 10 mg/1 in one
monitoring well located  within the area, but no information is
available regarding spills or disposal or organic solvents.

DESCRIPTION OF THE MODEL
  The governing equation of pollutant transport in porous media
is the following:
 58    CONTAMINATE!} GROUNDWATER CONTROL

-------
                                                           Figure 1
                       Map of Vestal Well Field and Groundwater Table Contours with Wells 1-1 and 1-2 in Operation
   3c    f 3  /    3c      3c \   3 /    oc       3c\l
   — = n\ — \D,,, — + £>xv— +— £v,- — + DVV —
   3t    lax \    3x    y3y/  3y\  yx dx    yy 3y/J
         r 3         3
       - —  (Vxc) + —
         Idx         3y
                                                         (1)
in which:

                  c
    R = 1  + esA://z
                 Ss
                  k
                 n
      Dxv, D
            •yx>   yy
             V   V
             'x>  'y
                  \
                      = concentration of the pollutant
                      = retardation factor
                      = bulk density of the porous medium
                      = distribution coefficient of the
                        pollutant
                      = porosity of the  porous medium
                      = component of dispersion tensor
                      = Darcian velocity component
                      = 1st order decay constant
                      = concentration of the source fluid
                      = flow rate of the source fluid
Let L denote the operator on c in equation (1), thus:
Lc = 0.
For the finite element method:
                                                         (2)
                 {N}T{c}                                 (3)

where £/V} is the shape function. The residue becomes Lc. The
principle of Galerkin method requires that:
  Lc,Wt
           = 0 .
(4)
  Equation  (4)  states  that  the  inner  product of L? and the
weighting function Wj over the solution domain B vanishes. The
integrations  of Equation (4) are now carried out, and the results
are expressed in matrix and vector notation as follows:4
                                                                 {c}T[Af{c} +(K-A+E) {c} -{p}] =0
                                                                 in which:

                                                                       dc
                                 dW
               —  — •}  +nDxy  — •
               dx   3         x
K   -//  [n£)x
                                                    nD
                                                                                                                   ay
                                                                                                                       3
                 + nDyy  {	\l	
                                                                                             dxdy
        E   =//   n\R{W}{N}'* dxdy
               B

        M   -ff   nR{W}{N}'T dxdy
                                                                 {p} =      Qc*{A'}dxd;y+
                                                                                                   3c        3c
                                                                                                 xs — + "-D^y	VjcC Inx
                                                                                                   ax        ay       '
                                                                                   (3c       3c      \   "1
                                                                              nDyx — + nDyy -- VyC rty   { W } dS .
                                                                                   dx       3y      /   J
 (5)





 (6)

 T


 (7)



 (8)

 (9)


(10)




(11)
The advantage of the finite element method over the finite dif-
ference method is its ability to handle complex boundaries and
normal derivatives. In the time dimension, these advantages are
not present, so the finite difference method will be used for the
time derivative term of the solute transport Equation.5
  With finite difference  in time, Equation (5) becomes
                                                                            CONTAMINATED GROUNDWATER CONTROL    59

-------
                                                                                              Table 1
                                                                     1,1-Dichloroelhane Concentration Measurements at Vestal Well Site
 in which ^ =  1  for backward-differences and /i  = 0.5 for the
 Crank-Nicholson differencing scheme. Along the boundary of the
 study area, Dirchlet boundary conditions were specified.
 SIMULATION OF CONTAMINANT PLUME
   For the preparation of input files for the computer simulation
 study, it was necessary to set up a grid system for the discretized
 site and determine several model parameters. According to the
 site history, production wells 1-1 and  1-2 are constantly in opera-
 tion. Therefore, a non-uniform but steady-flow field  was as-
 sumed from the measured ground water table under wells 1-1 and
 1-2 while in operation'  as presented in  Figure  I. As shown in
 Figure 2, the Vestal well site is discretized into 103 quadrilateral
 elements with 125 nodals.
   Among  the dissolved organics  in groundwater,  1,1-dichloro-
 ethane was chosen to model  the groundwater solute  transport in
 the study  area.  The  measurements  from the  samples  at  the
 monitoring wells are summarized in Table I. The locations of the
 monitoring wells are shown in Figure 1.  The 1- series wells were
 installed during the assessment study. The s-  series wells were in-
 stalled during the remedial investigation  by E & E.
   Since  no field measurements are available for the dispersion
 coefficient, this value was selected based on the literature. The
 molecular  diffusion is assumed to be  small compared  to  the
 hydrodynamic  dispersion. Values of 205  ft  for  longitudinal
 dispersivity and  12.9  ft for  transverse  dispersivity were used.
 1,1-dichloroethane could be  adsorbed on the solid phase of the
 aquifer.  However, the retardation effect is believed to be small
 and will not seriously affect  the qualitative prediction of solute
 transport.  Therefore, a retardation factor of 1 was used for this
 study. The values of model parameters are listed in Table 2.
                                                                             Well
                                                                                                       1.1-dlchloroethane

NO.
S-l
S-2
S-6
S-7
S-ll
1-33
1-34
Node
No.
65
77
103-112
94
68
84
78

Date
4/26/85
4/26/85
4/26/85
4/26/85
4/26/85
4/26/85
4/26/85

Cone.
0
945
58
1.280
158
860
l.JOO
                            Table 2
           Model Parameters Used in Computer Program
  Number of elements
  Number of nodes
  Number of we)It
103       Longitudinal dliperslvlty
125       Transversal dlsperstvlty
7        Retardation factor
  Hydraulic conductivity  47 ft/day  First order decay constant
205 ft
12.9 ft
1
0
  In order to verify the dispersion coefficient chosen for  this
study,  a simulation of the dispersion of 1,1-dichloroethane  was
conducted. Two constant-strength sources located at nodals 84
and  94 were assumed in this computer run. The contaminant
sources were found during the remedial investigation. The simula-
tion  time was 6 years. The assumptions were based on our best
knowledge of the history of the site. The computed contaminant
plume  and comparison with  field  measurement  are presented in
Figure  3. The results show that  the simulated  concentration is
very  close to the field measurements. Other values of longitudinal
and lateral dispersivity also were selected in simulating the disper-
sion  of 1,1-dichloroethane. No satisfactory' results were obtained
from these tests. Therefore,  the longitudinal and lateral disper-
sivities selected for the current study represent a rational choice
for this site.
                                                            Figure 2
                                         Finite-Element Grid System for the Vestal Well Field
60    CONTAMINATED GROUNDWATER CONTROL

-------

1597 Slmul.i.d Cone.
11300) M.I,
ltd Concin
imHon
».„.„
                                                          Figure 3
                                         Computed Contaminant Plume and Comparison
                                                   with Field Measurement
  Since spring 1980, well 1-1  has been pumped and discharged
directly into the Susquehanna River. A possible remedial action
would be the removal of the contaminant source and would con-
tinue use of well 1-1 as a recovery well to withdraw the pollutant
from the aquifer.  Computer simulation runs were conducted to
simulate  the movement of the  contaminant  plume after the
removal of the contaminant source and while well 1-1 was in
operation. Figure  4 presents the simulated concentration at the
end of 5 years. A comparison of Figure 4 with Figure 3 shows that
the 1000-/tg/l iso-concentration contour is moving toward well
1-1, and the iso-concentration area is shrinking. Figures 5, 6 and 7
present the simulated concentration at  the end of 10,  15 and 20
years, respectively. It can be seen  clearly in these figures that the
contaminant plume is moving toward recovery well 1-1 and that
the high concentration region is shrinking.
                                                          Figure 4
                                           Computed Contaminant Plume, 5 Years After
                                                Removal of Contaminant Source
                                                                          CONTAMINATED GROUNDWATER CONTROL    61

-------
                                                          Figure 5
                                          Computed Contaminant Plume, 10 Years After
                                                Removal of Contaminant Source
 CONCLUSIONS
   The model predicted the movement of the contaminant plume
 under the operation of well 1-1 and the removal of the contami-
 nant sources. The prediction was made for the time periods of 5,
 15 and 20 years. The study provided a basis to evaluate the effec-
 tiveness  and the cost of using well 1-1  as a  recovery well. This
 summary paper demonstrates  the usefulness of  the computer
 model in assisting the feasibility  study at uncontrolled hazardous
waste sites. Further  applications  of computer  groundwater
restoration  simulation can be conducted by varying the pumping
rates, locations and numbers of recovery' wells.

ACKNOWLEDGEMENT
  The authors would like to express their appreciation to U.S.
EPA and  the  New  York Slate  Department  of  Environmental
Conservation for their support in this study.
                                                         figure 6
                                          Computed Contaminant Plume, 15 Years After
                                                Removal ol Contaminant Source
62   CONTAMINATED GROUNDWATKR CONTROL

-------
                                                              Figure 7
                                             Computed Contaminant Plume, 20 Years After
                                                   Removal of Contaminant Source
REFERENCES
1.  Ecology  and Environment,  Inc., "Town of Vestal  Water District
   No. 1 Focused Feasibility Study," May 1985.
2.  Chieh, S.H., Cook, D., Hwang, J.C. and Brandow,  J., "Computer
   Model Study of a Contaminated Well Field—Groundwater Pollution
   Source Identification and Contaminant Plume Simulation," Proc. of
   the 1985 ASCE Hydraulics Division Specialty Conference, Orlando,
   FL, 1985, 330-335.
3.  Martin, R.J., Coates, D.R. and Timofeefe,  N.P., "Well Field Con-
   tamination Investigation for Town of Vestal Water District No. 1,"
   Report for New York State Department of Environmental Conserva-
   tion, 1983.
4.  Yeh, G.T.  and  Ward, D.A., "FEMWASTE:  A  Finite-element
   Model of Waste Transport  Through Saturated-unsaturated Porous
   Media," Oak Ridge National Laboratory Report No.  5601, 1981.
5.  Hwang, J.C. and Koerner,  R.M.,  "Groundwater Pollution Source
   Identification from Limited  Monitoring Well Data—Part 1, Theory
   and Feasibility," J. of Haz.  Mat,, 8, 1983, 105-119.
6.  Bear, J., Hydraulics of Groundwater, McGraw-Hill, Inc., New York,
   NY, 1979.
7.  Gray,  W.G.  and Hoffman, J.L.,  "A  Numerical Model Study of
   Ground-Water Contamination  from Price's Landfill, New Jersey-II.
   Sensitivity Analysis and  Contaminant  Plume Simulation," Ground
   Water, 21,  1983, 15-21.
8.  Finder, G.F., "A Galerkin-Finite  Element Simulation of Ground-
   water Contamination on  Long  Island, New York," Water Resources
   Res., 9, 1973, 1657-1669.
                                                                               CONTAMINATED GROUNDWATER CONTROL    63

-------
                          Enhancement of  Site Assessments by
                                      Groundwater Modelling

                                            Joseph R. Kolmer,  P.E.
                                            John B. Robertson, P.G.
                                               Roy F.  Weston, Inc.
                                           West Chester, Pennsylvania
ABSTRACT
  It is a commonly  held misconception that extensive ground-
water monitoring and  other geologic and groundwater related
data must  be  available before a groundwater model  can  be
selected, calibrated and implemented. The primary use of models
in this mode of operation has been to predict long-term ground-
water flow  and groundwater quality due to implementation of
engineered groundwater supply systems and/or remedial cleanup
programs.
  Advances in computer technology have dramatically revised the
utilization of groundwater models and expanded their utility with
respect  to  groundwater investigation  programs. Groundwater
flow models are now  readily adaptable to personal computers. To
use these models in a site  investigation, only a very  limited
amount of information is needed initially, and assumed data can
be used to supplement this initial  information.
  Given the limited data, the model is responding primarily to the
general equations of groundwater flow and not to unique site con-
ditions.  This uncalibrated flow model, however,  provides the
technical skeleton to which additional  groundwater monitoring
information can be input and assessed. As additional site data are
added to the model,  it is refined and calibrated. During the data
gathering process, the model is used not only to identify data
gaps,  but  also to determine when  sufficient  data  have  been
gathered to  quantify  site conditions.
  The groundwater model is not limited to utilization as a predic-
tive model based upon exhaustive site information. Rather, the
groundwater model is now most properly  and  cost effectively
used in the groundwater monitoring program to assess data gaps,
quantify  site  conditions  and evaluate engineered remedial
programs.

INTRODUCTION
  The purpose of this paper is to discuss and present a method of
site  investigation using groundwater  computer  models to guide
the  investigation as well as to indicate when sufficient data have
been collected. The methods used to implement this approach are
neither unique nor  technically complex, but  they do  require
utilization of an integrated geoscience and engineering staff. This
technical integration,  plus the general availability of groundwater
models on microcomputers, makes this approach to site investiga-
tion technically feasible and advantageous.
  The need for quantitative groundwater  resource  assessment
tools is growing at an  accelerating rate. Water resources investiga-
tions for domestic, agricultural and industrial uses are increasing,
and the investigation work associated  with definition of con-
taminated groundwater due to hazardous materials is expanding
daily. These latter investigations are particularly difficult because
they require quantitative analysis methods to define the contami-
nant migration conditions as well as the remedial solutions. The
use of groundwater flow models to quantify the site investigation
work and standardize the basis upon which remedial systems are
comparatively evaluated makes use of the inherent attributes of
these methods.  A brief discussion of the current practices in
groundwater modelling is presented below followed by a detailed
discussion of how models can be used in site assessments.

EXISTING METHODS OF
SITE INVESTIGATION
  The predominant practice for projects involving groundwater
quality and water resource investigations is illustrated in Figure 1.
The objective or the field investigation work is to obtain sufficient
information to describe site conditions, specifically the geologic
conditions. After these data are gathered, they are assembled and
assessed to determine their adequacy.
           Report
                                              Sit.
                                         	Groundwater
                         Figure 1
       Silc Investigation Followed by Groundwater Modeling
64    CONTAMINATED GROUNDWATER CONTROL

-------
  This data evaluation exercise might reveal the need for addi-
tional information to describe an anomalous or vague site condi-
tion, and a second phase of field investigation would commence.
This interactive process would continue until the  objective of
understanding  the geology, ground water hydrology  and  water
quality conditions is accomplished. A report describing the in-
vestigative procedures  and the defined site conditions is finally
prepared.
  After  the  field investigation is complete,  the  engineer or
engineering  group receiving the final report develops a set of
remedial action alternatives  designed to mitigate  any adverse
chemical quality conditions identified during the field investiga-
tion. To conduct a comparative analysis of the identified remedial
alternatives, a groundwater flow model is usually employed.  The
flow model is  calibrated based  upon the findings of the field
report and run (exercised) to satisfy the identified  needs of the
engineer. If, during the conduct of the engineering  analyses, the
need for additional data is identified, it usually is assumed as  part
of the model input. The mathematical operations inherent to the
flow model check these assumed inputs against measured site con-
ditions. Thus, assumptions made at this stage of a site investiga-
tion are not totally unchecked, but they usually do  not have the
advantage of being field verified.
  After the site model has been developed and calibrated, it is ex-
ercised to simulate  the  various remedial solutions being  con-
sidered for the site. For example, a groundwater barrier wall, such
as a soil bentonite slurry cut-off wall, could be simulated by the
insertion of no-flow boundaries around the perimeter of the  site.
Guswa1 noted that implementation of a remedial action plan can
be very expensive, and it is desirable to have the preimplementa-
tion assurances that the proposed plan can be effective. While no
type of modelling provides guarantees that any particular plan
will be effective, they do provide a comparative basis against
which all remedial alternatives can be evaluated.

THE MODEL AS INTEGRATOR
  The use of the groundwater model is a good tool to evaluate
various remedial solutions. This use of the model is obviously ad-
vantageous, but additional information can be obtained. That is,
the groundwater model can be used not only to evaluate the facts
gathered after conclusion of the field investigation, but also can
be used to better define what information needs to be gathered
and where it can be found.
  Figure 2 illustrates how the groundwater  flow model can be
used during the site investigation phase of work and in the subse-
quent analysis of remedial alternatives. Comparison of this figure
to Figure 1 shows that the steps leading to the final report have
been reduced. This reduction primarily centers around integration
of  the groundwater modelling work into  the analysis of data
obtained during the field investigation exercise.
  To initiate use of  groundwater modelling during the field in-
vestigation phase of any project, a basic amount of site informa-
tion must be known.  General site geology,  basic  groundwater
flow conditions and surface water conditions should be known to
establish  the initial flow parameters and boundary conditions for
the model. Published literature is replete with geologic definition
and groundwater ami surface water factors for a large portion of
the United States.  If an assessment of existing information on a
site area is conducted in a fairly exhaustive manner, the general
information needed initially to establish the groundwater model
usually can be accumulated.
  One of the primary questions raised during conduct of a field
investigation is how does one know when  sufficient data are
available to satisfy the objectives of the project? The  groundwater
flow model  can provide significant assistance in answering  this
question. As indicated in Figure  3, in order  to  assemble or
                          Figure 2
                  Integrated Site Investigation

establish the groundwater model, existing data as well as input
from numerous technical disciplines must be obtained. Questions
as to the area to be covered by the model, the geologic conditions
to be  considered and the  overall engineering requirements
associated with the model must all be discussed and incorporated.
Thus, at this early stage of the project, the flow model has in ef-
fect served as a catalyst for integration of the technical disciplines
that eventually will be needed to resolve the identified problem.
This integration,  properly carried to fruition,  will result in the
establishment of a set of objective criteria for the site investiga-
tion and the overall remedial action study. Specific data needed to
satisfy these criteria would be identified for acquisition during the
field investigation.
  As field investigation work is conducted, new site information
is developed and  can be used  with the established groundwater
model. As these data are added, the aquifer characteristics are ad-
justed so that the model representations are consistent with iden-
tified field conditions. As more information is gathered in the site
area, the adjustments that need to be made to the model will
        Model
       Calibration •
                          Figure 3
                Groundwater Model Development
                                                                            CONTAMINATED GROUNDWATER CONTROL   65

-------
become relatively minimal. As this point is approached, acquisi-
tion of additional field information is not needed to refine the site
model, and field investigation work is complete. Site conditions
are now defined consistent with the set of objective criteria iden-
tified for the project.  Assessment of remedial alternatives can
commence.
   The model assembly, identification of additional  data  needs
and model revision work tasks identified in Figure 3 are in  effect
the data analysis  and  additional data input loop identified  in
Figure 2. The exercise  of  identification of additional data  needs
and acquisition and  input of these data to the  flow model con-
stitute calibration of the flow model. An important distinction be-
tween this approach to use of groundwater modelling and the
approach discussed  in the previous subsection  is that  model
calibration is integral to the field investigation exercise.
   Finally, no (or at  least very little) assumed data are needed  to
calibrate the groundwater model.  Thus,  the flow model that
ultimately will be  used to  compare various remedial alternatives
already  will  have undergone the  rigors  of development and
calibration placed upon it not only by the simulation of known
site conditions, but also by the technical requirements of the disci-
plines involved in  overall problem resolution.

THE COMPUTER MODEL
AND SITE INVESTIGATIONS
   The principal concerns surrounding the use of computer model-
ling  have  been  its  availability  to  the  project's principal
investigator and the cost associated with development and opera-
tion of these models. Over the past few years, both of these con-
cerns have been significantly diminished by  the development  of
the desk top personal computer. Thus, the principal investigator
or project manager has  a significant  library  of groundwater
models readily available to him that can be developed and exer-
cised at a  relatively  inexpensive cost.  In  fact,  the graphics
capability as well as the data base management  capabilities  of
these machines can reduce the overall project costs compared  to
hand manipulation and graphics development for a  medium  to
large scale site investigation.
   In  the past, groundwater flow models have been misused by
misrepresenting their capabilities  and  application. One of the
common problems associated with groundwater modelling  is the
tendency to use the most sophisticated model available for  prob-
lem resolution. This  temptation is difficult for the technical com-
munity to resist, but can be managed if integration of technical
skills is accomplished in the early stages of the project. This in-
tegration  of  the  groundwater  modelling  expertise with the
pragmatism of the engineer usually results in  model development
consistent with the identified objectives for  the site study.  This
management of technical  expertise must be  coordinated by the
project's principal investigator.
   Another technique that can be used to help prevent  over exten-
sion and misrepresentation of groundwater  model usage is the
phased implementation of  a particular model's capability. For ex-
ample, throughout this paper, only groundwater flow  models
have been discussed  and solute transport models have not been
strongly  considered. This analysis  does  not  say  that  solute
transport models are not appropriate for the problems discussed
herein; it does, however,  say that in the early  stages of model
development  and utilization in site investigations, the first step
should be a  characterization  of  groundwater flow  hydrology.
Once   groundwater  flow  characteristics  (especially velocity
parameters) have been identified, solute transport models can be
included. One must remember that the model only need to detailed
enough to satisfy the  stated objectives of the project.
  The following brief example shows how various data elements
can be obtained and input for site investigation modelling. The
advantage of a phased field investigation program is highlighted
in the example.

Existing Data  Review and Model Development
   During the review of existing information for a site area, data
concerning the history of operation of the site, the geologic and
geohydrologic conditions of the area, precipitation, surface water
hydrology and so forth, are compiled. The completeness of this
information depends upon the location of the site, the notoriety it
has received, site management and numerous  other factors. For
the purposes of this paper, only that  available data pertinent to
development of the groundwater model will  be discussed. The
lack of consideration of any other data does not mean it is not
needed for a site investigation nor does it mean it may not be per-
tinent to model development at some  sites.
   Figure 4 presents  a map of the example site area and contains a
significant amount of information. The example site is located in
a river floodplain below an upland area. Thus, the aquifer of con-
cern is  more than likely alluvial in nature, probably containing a
wide distribution of grain-size materials. In addition, channeliza-
tion of sediments due to alluvial action is very possible. Finally,
the upland area probably  forms a recharge  boundary for the
alluvial  aquifer,  and  the  river probably  forms a  discharge
boundary.
                          Figure 4
                      Example Site Area

  There are four existing monitoring wells identified at the cor-
ners of the site. These wells could provide information on ground-
water  levels as well as the  type  of  sediments  contained  in the
saturated zone. Figure 5 presents preliminary geotechnical find-
ings showing a gravel seam through the center of the site flanked
by  finer grained  sediments. If no aquifer characteristic data is
available, the existing monitoring wells could be used to conduct
abbreviated aquifer tests, such  as  slug  tests. These tests are
66    CONTAMINATED GROUNDWATER CONTROL

-------
described by Cooper, Bredehoeft and Papadopulos2 and Bouwer
and  Rice.3  Thus,  existing site information as well as readily
obtainable information can provide the needed data.
                           Upland Area
                           Figure 5
                Preliminary Geotechnical Findings

   In some site areas, no existing monitoring wells are available. In
such  cases,  geologic and hydrogeologic reports concerning the
aquifers of  the area as well as the  estimated permeabilities and
transmissabilities would  have to be utilized. In these cases, it
might be advantageous to postpone initial model development un-
til specific site information could be obtained from installation of
groundwater monitoring  wells. Generally speaking, sites being in-
vestigated today have a limited number of monitoring wells or soil
borings available for use in the initial stages of  the projects.

FIELD INVESTIGATION
   Based upon the  findings of the data review, it is often advan-
tageous to conduct the field investigation work in a phased man-
ner. That is, early acquisition of general site area  information
(such  as aquifer characteristics determined by slug  tests  and
geophysical  data) would  be advantageous to more detailed plan-
ning of the later stages of field investigation.
   Figure  6  shows a  probable set of conditions related  to
geophysical  findings for  the example site area.  The geophysical
methods employed might include determination of terrain  con-
ductivity by electromagnetic techniques.  The results of this type
of investigation  usually are mapped so that  anomolous conduc-
tivity  patterns (inferred to be indicative of contaminant transport
conditions) can be identified. In our example problem, it appears
that the gravel seam through the site area is a preferred avenue of
groundwater movement  and contaminant transport.  In a pre-
ferred groundwater movement area, it is likely that water levels in
the gravel will be slightly lower than in the immediately adjacent
less permeable zones.
  Given the information obtained from the data review as well as
the preliminary work conducted in the initial stages of the site in-
vestigation, a plan of overall investigation effort can  be for-
                         Figure 6
                 Geophysical Survey Findings

mulated. Figure 7 displays how additional data with respect to
groundwater flow conditions and contaminant levels might be ob-
tained by monitoring well installation.  At the same  time, the
available data can be used to establish the initial conditions of the
groundwater flow model.  The groundwater elevations obtained
from the existing monitoring  wells coupled with  the assumed
groundwater levels associated with the gravel seam can be used to
predict the findings of the investigative work. The actual data
from each well would be put into the model as each well was in-
stalled and the model would be rerun to refine predictions for the
remaining wells scheduled for installation. Based upon these find-
ings,  monitoring well  locations would be modified to  maximize
data acquisition efforts. The use of the model to guide and refine
the field investigation exercise can result in the elimination of
monitoring wells and the early cessation of the field investigation
exercise.

Simulation of Remedial Work
  After the  field  investigation  exercise has been completed and
groundwater flow calibration has been accomplished, the impacts
on groundwater movement due to various remedial scenarios can
be simulated. In the case of the example problem,  groundwater
flow  barrier walls intersecting the gravel zone,  coupled with
groundwater pumping wells, could be simulated.  Groundwater
pumping without barrier walls also could be modelled and the
results could be compared to estimate the savings in  treatment
costs that could be realized due to the difference in groundwater
pumpage. These simulations, as well  as numerous other simula-
tions, can be made on a consistent comparative basis through the
                                                                           CONTAMINATED GROUNDWATER CONTROL    67

-------
 developed groundwater  flow  model. This consistency of com-
 parison is essential to the quantitative evaluation of alternative
 remedial measures.
     jLLLOU-LLLLi
     I  ILJLl_U-i_LiJ.
                          Figure 7
                Additional Data Needs Definition

 THE OVERALL PICTURE
   Not all projects may be appropriate for application of ground-
 water flow models as described in this presentation. Certain site
 problems may be so small that a conceptual model may be suffi-
 cient and a computer model may not be warranted. The ground-
 water model is simply a tool that can be used to help organize and
 streamline the site investigation work and the process of remedial
 system selection.
   Whenever a model is being developed during a site investigation
 and is intended to be used to compare and predict remedial action
 impacts, care must be taken to properly apply historical factors
 specific  to  the  site  area.  For  example, long-term  historical
 precipitation  patterns should  be  considered  when  designing
 groundwater   pumpage,  treatment  and  recharge  networks.
 Anomolously high rainfall conditions could cause an increase in
 the volume of water requiring treatment or in other ways impair
 operation of the remedial system. Anomolously low rainfall could
 result in  a general lowering of the groundwater table and cause
 flow conditions to be significantly different than those considered
 by the predictive model. In either of these cases, if the predictive
 model did not consider the long-term fluctuations in precipitation
and the resultant impact on infiltration quantities, the predictive
analyses  compiled by computer modelling might not  be valid.
Mercer and Faust* recommend that groundwater flow models be
 calibrated with known  historical site information and that the
 period of prediction not exceed twice the period of historical data
 used in the calibration.
    When evaluating the overall factors of a site investigation and
 development of remedial alternatives, a general review  of the
 work should be made to insure that it is consistent with sound
 technical "intuition." One of the worst mistakes that can  be
 made in groundwater modelling is unchecked acceptance of the
 model results. Data cross-checking and integration are advisable
 with  or without  a  groundwater  computer  model,  but must
 definitely be utilized when  employing models to protect against
 blind acceptance of computer output data.
    A good conceptual understanding of the site conditions and
 general influences in  the site area is a prerequisite to initiation of
 the site investigation  work and utilization of groundwater model-
 ling. Without such understanding and cross-checking of technical
 information, little or  no integration of technical disciplines will be
 accomplished,  the  overall   program  effectiveness  will  be
 significantly reduced  and the investigation may, in fact, become a
 waste of time and money.

 CONCLUSIONS
   Until now, groundwater models have not been extensively used
 as tools in the conduct of site investigations. Generally, models
 have been applied after completion of  field  investigations to
 predict the impact of  various remedial alternatives. The prolifera-
 tion of sophisticated  computer equipment and the adaptation of
 groundwater model codes to this equipment have literally brought
 the groundwater flow modelling capability to the desk top of each
 project  engineer and scientist. Thus, groundwater modelling is
 now a readily available tool  that can be used to identify data col-
 lection requirements,  evaluate these data, conduct data sensitivity
 analyses with respect  to groundwater flow impacts and assess the
 long-term impacts of various remedial actions on the groundwater
 flow conditions.
   When employing groundwater flow models, a good conceptual
 understanding of the  aquifer and other pertinent subsurface con-
 ditions  must be available  as  the  underlying  basis for model
 development. The extent to which these conceptual understand-
 ings backed by actual observed field conditions can be simulated
 by the model is  the  quantitative measure of  the ability of the
 model not only to adequately represent existing field conditions,
 but also to predict future conditions. The ultimate responsibility
 to ensure both proper integration  of technical talent in model
 development and calibration, plus proper technical usage of the
 model in simulating remedial activities, rests with the principal in-
 vestigator or overall  project manager.  Proper  integration  of
 technical talent in model development and usage is key to suc-
 cessful site investigation and remedial action alternative analysis.

 REFERENCES
 I. Guswa, J.H., "Numerical Models, Practical Tools in Groundwater
   Investigations," The  Weston Way. Fall 1984, 10-13.
 2. Cooper, H.H., Bredehoeft,  J.D. and Papadopulos,  I.S.. "Response
   of A Finite-Diameter Well To An Instantaneous Change of Water."
   Water Resources Res.. 3, 1967,  263-269.
 3. Bouwer, H. and Rice,  R.C., "A Slug Test for Determining  Hy-
   draulic Conductivity of Unconfined Aquifers with Completely or
   Partially Penetrating Wells," Water Resources Res.. 12.  1976.423-428.
4.  Mercer, J.W. and Faust,  C.R., "Groundwater Modelling," National
   Water Well Association, 1981.
68    CONTAMINATED GROUNDWA TLR CONTROL

-------
         Alternative Treatment  Techniques for  Removal  of  Trace
             Concentrations  of Volatile  Organics  in Groundwater
                                                Mark E. Wagner
                                                Brian V. Moran
                                            Geraghty & Miller, Inc.
                                             Annapolis, Maryland
ABSTRACT
  In response to the detection of volatile organics in wells serving
the process and drinking water needs of an industrial manufactur-
ing facility, a groundwater quality investigation was initiated for
an industrial manufacturer in Maryland. The objectives of the
investigation were to define the extent of groundwater contamina-
tion, to identify probable  sources and  to  develop remedial
measures to rehabilitate the  affected aquifer while ensuring a
potable water supply.
  Field studies confirmed the sources of volatile organics to be
burial trenches located 350 ft southeast of one of the affected
wells.  A hydraulic-control recovery  system was recommended,
using the affected well as a pumping center.
  Several treatment techniques were identified which could treat
the low-levels of tetrachloroethylene and trichloroethylene. The
alternatives included aerated  lagoon volatilization, spray irriga-
tion/land application and packed-tower air stripping. Cost com-
parisons and  treatment  efficiencies for  each  process were
evaluated. Air stripping was selected as the preferred alternative
based on efficiency, cost and reuse of the treated water to satisfy
the facility's water needs.
INTRODUCTION
  An industrial manufacturer in Maryland depended on wells
on the plant property for process cooling and drinking water.
When low levels of tetrachloroethylene (PCE) and trichloroethy-
lene (TCE) were detected in several of the supply wells, the facility
installed individual granular activated carbon (GAC) cartridges at
various points in  the plant. In response to increasing levels of
volatile organics found in the water supply, a groundwater quality
investigation was initiated to determine the probable sources of
the contamination and the degree of aquifer degradation, and to
recommend  remedial measures to rehabilitate  the aquifer  while
ensuring a continuous potable water supply for the plant.

GEOLOGY/HYDROGEOLOGY
  The plant site is located within the eastern division of the Pied-
mont physiographic province. The region is  characterized by
moderate relief, gentle  slopes and  rounded hills. Locally, the
bedrock geology consists  of folded biotite and chloritic-albite
schist of the Wissahickon formation. The Wissahickon is overlain
by saprolitic material maintaining parental bedding-plane struc-
tures. The saprolite consists of silty clay varying in thickness from
                                              PROPERIY_L|NE.	
                                                       X STUDY AREA  _. :^°'
                                                    WOODED AREA
                                    TREATMENT
                                      PLANT
                      EXPLANATION
               300FEET
                      H
                            I WATER SUPPLY WELL
                                                             MANUFACTURING
                                                                 BUILDING
                                                     Figure 1
                                        Site Plan of the Groundwater Investigation
                                                                    CONTAMINATED GROUNDWATER CONTROL   69

-------
             SURFRCE  MRGNETIC SURVEY  (Gammas+50K)
                            Figure 2
       Geophysical Survey Results Using Magnetics Instrumentation
                                                  "'     SURFRCE CONDUCTIVITY SURVEY  (mmhos'm)

                                                                       Figure 3
                                                 Geophysical Survey Results Using Conductivity Instrumentation
                                                     \

                                                     "/ .N
                                                     /•*.
                                                    ?»«
                                                   -
                                                                           \
            f~.   Scare* *!•••
    .„    .5"Ts~r
\          S^b\NS-^f $
                                           100      200
                                           F..I
                                                                                                  •SS
                                                                                                    \
                                                                                                        \
                                                                                                        3 ^
                                                           Figure 4
                                         Well Locations and Source Areas of Contamination
70    CONTAMINATED GROUNDWATtR CONTROL

-------
                                                   as*
                           Figure 5
        Geologic Cross-Section and Groundwater Flow Patterns

 25 to 125 ft. Groundwater exists in both bedrock and the overly-
 ing  unconsolidated  deposits  predominantly  in  fractures  of
 bedding-plane   microfractures.1  The secondary  permeability
 results in a wide areal variation in well yields dependent on prox-
 imity of the well to fracture zones. The facility and surrounding
local populace depend on the Wissahickon formation for water
supply.  Wells in the area typically penetrate the bedrock from 50
to 450 ft, with well yields varying from 5 to 250 gal/min.

FACILITY DESCRIPTION
  The manufacturing facility is located on top of a hill with a
small stream located along the northwestern and western property
boundaries.   This  stream locally  represents  the  ground water
divide. On-site water-supply wells for the plant are located along
the northwest property line. At the southwest corner of the facil-
ity, a lagoon collects and treats stormwater runoff from the boun-
dary swaleway  and treated wastewater effluent from the plant
(Fig.  1). Water-supply well 7  (WS-7), located at the northwest
property boundary, was found  to contain the highest concentra-
tions of volatile organic compounds (VOCs) with 1300 /ig/1 tetra-
chloroethylene.
SOURCE INVESTIGATION
  Facility personnel tentatively identified possible VOC  source
areas  as waste pits  previously excavated in a clearing  350 ft
southeast of Well WS-7. Wastes, which were suspected of being
present in these pits, included process waste drums, construction
debris and defective products. A surface geophysical investigation
was  conducted  in  this area using a proton precession mag-
netometer and terrain electromagnetic conductance instrumenta-
                                                                         FRACTURED ROCK
                                                                                                    MOVEABLE PACKER /PUMP STRIN6
                                                                                                    INFLATABLE PACKER
                                                                                                   INFLATABLE PACKER
                          Figure 6
        Straddle Packer Apparatus Used for Stressing Wells

 tion to locate buried metallic objects. Both surveys were run on
 common transects with 20-ft grid spacings. The proton precession
 method delineated three anomalous areas of low and high signals.
 Terrain conductance delineated three areas of anomalous signals
 located between the proton precession  highs and lows  (Figs.  2
 and 3).
WATER QUALITY INVESTIGATION
  Subsequent to the surface geophysical work, numerous nested
well clusters were installed around the major anomalies to: (1) de-
fine horizontal and vertical flow patterns and (2) determine local
groundwater quality. Seventeen monitor wells were installed at
discrete water-bearing intervals in  the saprolite material. Wells
were designated by series P, W or S, representing the first water-
bearing zone (20 ft), intermediate water table (40 ft) and bedrock/
overburden interface (60 ft), respectively. The well locations are
shown in Figure 4.
  Based on water-level measurements from the installed monitor
wells, groundwater flow is directed to the west toward the small
stream. In  addition, three significant  seeps  were  discovered
downgradient of the area under investigation.  These seeps are
believed to represent surface discharge of groundwater from the
first water-bearing  zone. Vertically, a strong downward com-
ponent  of groundwater flow also  exists. Groundwater samples
from  the wells and  seeps were  collected and  analyzed for
suspected VOCs. Based on the sampling, Anomaly B was iden-
tified as the major contributor of VOCs  to  the groundwater
regime (Fig. 5). Wells located between Anomaly B and Well WS-7
exhibited  tetrachloroethylene  concentrations  equal  to levels
discovered in the water-supply well.

AQUIFER TESTS
  In an effort  to better define the groundwater regime in the
vicinity of Well WS-7,  several aquifer  test programs were in-
itiated.  A downhole pneumatic straddle packer/pump assembly
was used in Well WS-7 to stress discrete intervals of the openhole
well (Fig. 6). Short-term pumping was performed at various inter-
                                                                            CONTAMINATED GROUNDWATER CONTROL    71

-------
                                                                Table 1
                                                     Lagoon Disposal Cosl Estimates
                                         EquIpment
                                                                 Llni-
-------
                                                          Table 2
                                         Slow Infiltration (Spray Irrigation) Cost Estimates



•a
-. I
' U)
a





TOTAL
CAPITAL
COSTS

4J
10
0

-------
                                                           Table 3
                                                 Stripping Tower Cost Estimates
S t r i p|i l nq Towe i (with b 1 owo r )
P t p i nq

French Drat n
s ?s,ooo
$ 10,000
S 3,000
$ 15,000
$ 19,000
C4M, T«Bpa
Mean* C CD
Means CCD
G*« I
Means CCD
                                            CoetB Un Iquc to
                                            Stripping  IUC)
                                           Total Capital
                                           Costa (TCC)
Subcontracting Fees
Eng 1 nee r ing and De» ign


Start-up/Shake Down

C
TOTAL
CWITU,
COSTS
Operating Labor
Electricity for Pumps
$ 7,600 10% o( TCC
S $, 100 9% of TCC


S 3,300 M of TCC

$ 32,700
5108,200
$ 11,000 2 nan-hri/day 1-30
S 3,000 6.25 h.p. * 1-30
Be/kilowatt
                               Electricity for Blower
                                                                            3 h.p. •
                                                                            ec/kllo«att
                               Monitoring
                                                                            Not  Included
                               Administration
                                                               S  1,000
                               Maintenance Contingency Reserve
                                                               S  1,000
                      TOTAL
                      OPERATING
                      COSTS
 evaluated with and without a liner for the swaleway. Mass balance
 calculations were made, treating the lagoon as a continuous reac-
 tor with first-order kinetics for volatilization of the PCE and
 TCE. Evaluation of flow  conditions in the lagoon at both max-
 imum and average flow rates showed that there  would be suf-
 ficient retention time for volatilization of PCE to concentrations
 below 10 /ig/1 throughout  the year.
   Total capital costs were  estimated to be $41,000 for the unlined
 swaleway option and $79,000 for the lined swaleway. Operational
 costs would be  about $16,000 yearly. Cost estimates  for both op-
 tions are presented in Table 1.

 Spray Irrigation
   Treatment of the volatile organics in the groundwater also was
 evaluated  using  a spray  irrigation method.  With  this option,
 groundwater would be pumped through a network of sprayers
 located throughout part of the wooded  land located above the
 swaleway. It had been  estimated that approximately two acres of
 land would be required to  allow infiltration of the  100 gal/min of
 water. Approximately 60% of the PCE and TCE would volatilize
 during spraying. Water containing residual PCE and TCE would
 enter the swaleway and flow to the lagoon; however,  we deter-
 mined that some could infiltrate to deeper sediments.
   Since spray irrigation for this site  represented a combination of
 slow rate and overland flow methods (for water to the swale), the
method was evaluated using both design procedures. Both design
procedures confirmed that sufficient land was available at the site
for this treatment option. Volatilization rates of the contaminants
were evaluated using a mass transfer approach as in the lagoon
option; volatilization  rates compared  favorably to values in the
literature for similar studies by others.'
   Estimated capital costs for the spray irrigation/recovery system
were $108,200; ytarly operational costs  would be approximately
$15,000 (Table 2). One problem with this option was that the
spray irrigation system would  not be  operational during the
months of January and February due to cold weather conditions.
An additional cost consideration for spray irrigation would be the
need for  a pilot study to determine actual PCE removal rates and
the ability of the  soils to accept the sprayed fluids.

Air Stripping and Reuse of Groundwater
   The last  option  considered was pumping and treating the
groundwater using  a  packed  column tower air stripper.  Using
mass transfer  calculations, we calculated a 99<% removal of the
volatile organics  (PCE and  TCE). This option  would involve
pumping the water to be treated to the top of the aeration tower
and allowing it to flow downward  against a countercurrent flow
of blown air.  The system would be designed so that the treated
water  could achieve volatile  organic concentrations  as low as
1  /ig/1 which,  in turn, could be used as  the water supply for the
74    CONTAMINATED GROUNDWATER CONTROL.

-------
                                                             Table 4
                                            Summary of Alternatives, Costs and Evaluation

1 .

Option
Lagoon Disposal
wo/Liner
Total
Capital
Costs

57,730
Annual
Operating
Costs




could be
capital
Comments

desirable on
costs basis
                               2.  Spray Irrigation
                                  Air Stripping/Reuse
                                                           S  88,600   S 15,000
                                                           $108,200   S 18,000
                                                                    (S  5,060)
                                                                                    Cold weather operation
                                                                                    problems makes this option
                                                                                    less attractive
                 Option would allow for
                 S13,000/yr. savings by
                 eliminating in-plant GAC
                 system.
plant. Locating the tower adjacent to the plant buildings would
make it convenient to use excess steam, if available, to ensure
operation of the treatment system throughout the winter.
  Total capital costs for the completed stripping-tower/recovery
system were estimated to be  about $108,200; yearly operational
costs would  be $18,000. The cost  estimiates  are presented in
Table 3. These costs assumed that the volatile organic air emis-
sions would be permissible and that no carbon recovery system
would be required.

CONCLUSIONS
  The three treatment options were evaluated for effectiveness
and cost (Table 4). Assuming that the only contaminants of con-
cern are trichlorethylene and tetrachloroethylene, both the strip-
ping tower and the lagoon systems can effectively reduce VOC
concentrations in the groundwater to acceptable levels; the spray
irrigation  option  would remove approximately  60%  of  the
volatiles. The capital costs of the lagoon system would either be as
expensive as the stripping tower or 50% less expensive,  depending
upon  whether or  not lining  the  swaleway would be required.
However, considering the operating cost of the in-plant activated-
carbon filters currently used for drinking water (which  would still
be necessary if the lagoon system or the spray irrigation system
were used), the $18,000 per year operating cost of the stripping
tower would be offset by a $13,000 savings by eliminating the ex-
isting GAC system currently  installed. This option would also
allow reuse of the treated groundwater in the plant, since the con-
taminants could be reduced to a "safe" level.
  The air stripping option was recommended to the client and will
be  constructed for  treatment and  reuse  of the contaminated
groundwater. The lagoon treatment system, although competitive
from a capital cost perspective, was eliminated from further con-
sideration  because it did  not eliminate the  high  costs of the
presently installed in-plant activated carbon system nor did it pro-
vide the water-supply advantages that resulted from the packed
tower air stripping treatment system.
REFERENCES
1.  Meyer, G. and Beall, R.M., "The Water Resources of Carroll and
   Frederick Counties, State of Maryland," Department of Geology,
   Mines and Water Resources, Bull. 22, 1958, 355.
2.  Jenkins, T.F., Leggett, D.C., Martel, C.J., and Hare, H.E., "Over-
   land Flow: Removal of Toxic Volatile Organics," Cold Regions Re-
   search and Engineering Laboratory, Special Report 81-1, Feb.  1981.
                                                                             CONTAMINATED GROUNDWATER CONTROL    75

-------
             Cost-Effective  Soil  Sampling  Strategies  to  Determine
                        Amount  of Soils  Requiring  Remediation

                                          Gregory J. Gensheimer, Ph.D.
                                            William A. Tucker, Ph.D.
                                                Steven A. Denahan
                                 Environmental Science and Engineering, Inc.
                                                Gainesville, Florida
INTRODUCTION
   Designing a sampling strategy to determine the extent of con-
tamination in uncontrolled hazardous waste sites is not as simple
as putting out some transects or an evenly spaced grid. Similarly,
contouring the data should not be considered a simple exercise in
running software. The spatial distribution of  the data plus the
statistical distribution will determine the type of sampling scheme
and the type of contouring (interpolation scheme) that will repre-
sent field conditions with greatest certainty.
   Sampling schemes  and contouring programs  are most easily
designed and applied when the spatial distribution  of con-
taminants is continuous across the field and when the statistical
distribution tends toward normality or lognormality. If these two
conditions are met, the certainty about the boundaries separating
clean from contaminated  soils will be highest.  This high level of
certainty also can be achieved  by using  a  minimal sampling
strategy.
   Such ideal conditions of spatial and statistical distributions are
rarely met on  hazardous waste sites.  Hazardous  waste sites
typically are very heterogenous. The statistical distributions most
often are dominated by outliers or exist as multiple populations,
and the spatial distributions usually are discontinuous across the
landscape.  Such non-ideal conditions reduce the efficacy of sim-
ple sampling  schemes or  simple contouring programs. Non-ideal
conditions  also increase the  chances of non-optimal remediation
costs; costs that are much higher  than necessary considering the
"true" extent of contamination.
   A sampling strategy designed to overcome the non-ideal condi-
tions will be  expensive to accomplish, but the extra expense of
sampling can be compensated by reducing the amount of material
removed during remediation. An example of sampling scheme/
remediation optimization is described in detail later in this paper.
   Designing the appropriate soil  sampling strategy and choosing
the methods of interpolation depends on a variety of factors in-
volving the spatial and statistical distributions of the data set as
well as the expected remedial measures. These factors include: (1)
the method of disposal, spatial scale of disposal areas, site history
and previous sampling  results;  (2)  chemical and  physical
characteristics of the contaminants and site including mobility of
contaminants, soil organic carbon, water table and groundwater
flow direction; (3) present spatial extent of the contamination tak-
ing into account the homogenizing effects of the environment; (4)
appropriate sampling, analysis, methods and their costs consider-
ing site conditions, depth of contaminant penetration into the
subsurface  and anticipated  compounds; and (5)  most  likely
remedial alternatives and their costs.
Extrinsic vs Intrinsic Variability
  The firsi three factors above can be broken into  two major
groups relative to the variability contributed by each  factor. Ex-
trinsic variability can be thought of as the variability applied to
the system by non-natural forces (humans).
  Methods of disposal,  spatial  scale and distribution of the
disposal areas are all factors directly imposed onto the system by
humans. There is no way to predict the locations  or size of
disposal areas from viewing the natural system. A series of aerial
photos may  indicate  the presence  of roads, process buildings,
burning grounds or landfills, the location of which can be used in
designing the appropriate sampling scheme. Human  actions set
up the heterogeneity for natural factors to homogenize.
  On the other hand, intrinsic variability can be thought of as the
variability  applied  to the system by the natural environment. In-
trinsic factors  such as temperature,  precipitation and  physico-
chemical characteristics of the media act  in conjunction with the
characteristics  of the contaminant  to homogenize the  extrinsic
variability.
  When extrinsic factors are more evident than intrinsic factors,
the spatial and statistical distributions tend to be non-ideal (i.e.,
very  heterogenous).  As the effects of intrinsic factors become
more apparent, the distributions tend to be closer to ideal. This
certainly is not always the case as evidenced by the extreme
variability  of contaminant distribution  in naturally occurring
fractured  bedrock.
  For an example of the intrinsic/extrinsic effects of a hazardous
waste site, consider a drum storage area  with some leaky drums
interspersed through the area. At time zero, numerous small areas
of heavily  contaminated  soil  would  exist   in  an  area of
predominantly clean  soils. A gridded sampling scheme would
probably  miss most  of the contamination, but the statistical
distribution would be dominated by a few extreme outliers. The
extrinsic factors dominate in this situation.
  If the same site  is studied 10-20  years  later, the contaminants
will have begun migrating vertically and horizontally through the
soil. The homogenizing effects of the intrinsic factors are evident.
The same gridded sampling scheme will identify more of the con-
taminated  areas, especially at depth. The statistical distribution
will not be dominated by the extreme  outliers as before. The
statistical distribution will tend toward lognormality (more ideal),
exhibiting  some outliers but not as extreme as at time zero. The
spatial distribution will be more continuous than at time zero.
  Factors 4 and 5 consider the economics of the total remedial ef-
fort including sampling and analysis costs as well as cleanup costs.
The importance of designing an optimum sampling strategy is
76   CONTAMINATED SOIL TREATMENT

-------
minimized if the costs of the expected remedial alternatives are
minimal. If  one of these feasible remedial alternatives is to ex-
cavate, the number of samples to be taken can be optimized by
considering the estimated volumes of soils needed to be removed
considering sampling density.

OPTIMIZATION OF SAMPLING DENSITY
  When the  pattern of contamination is relatively simple (a con-
tiguous contaminated area with little or no contamination outside
that area) an optimal sampling density can be defined to minimize
total remediation costs,  including both  the  investigation and
cleanup phases. This  concept  is illustrated best by considering
some extreme examples.  Assume that  100 m2 of soil are con-
taminated. Next assume that samples  were taken to define this
contamination: one sample at its center and then samples on a
uniform rectangular grid with a sampling interval of 100 m2 for a
total of 9 samples as in Figure 1. As has been discussed, soil con-
tamination is extremely heterogeneous, and interpolation between
data points  is risky at best.  It is prudent, based  on the  data
available from this hypothetical site, to remediate all soils within
the region defined by the clean samples, approximately 40,000 m2.
 X SAMPLING LOCATIONS
                                      CONSERVATIVE ESTIMATE OF
                                      CONTAMINATED AREA
                             ACTUAL
                             CONTAMINATED AREA
                                 I—-
                          Figure 1
            Example of Inappropriate Sampling Density

   On the other hand, one might have started by taking a sample
every meter along a similar rectangular grid. Since phased sam-
pling frequently is not feasible, this approach would have  re-
quired 40,000 samples to define the contamination distribution
over the same area. Even if we had been clairvoyant, or had some
independent  information on  the soil contaminant distribution,
sampling at a density of 1 m would have required 121 samples to
define a contaminated area of only 10 m2.
   Obviously these are ridiculous extremes, but they clarify an im-
portant problem. How does one define an appropriate sampling
density?  If we sample intensively,  sampling and analysis  costs
could exceed the cleanup costs. If the sampling grid is too coarse,
we may not accurately define the contaminant distribution. Based
on a reasonable requirement to ensure the public health and safe-
ty, remediation of large quantities of clean soils may be indicated,
wasting valuable resources.
  This problem is resolved as a simple linear optimization prob-
lem where the objective is to minimize total remediation costs, in-
cluding the wasteful remediation of clean soils and the cost of
sampling and  analysis.  The cost  of remediating contaminated
soils must occur in any case and can be ignored in the analysis.
  The cost of wasteful remediation is calculated as follows. It can
be assumed that two rings of sample locations will define the con-
taminated perimeter. The perimeter of most simple shapes is 3 to
4 times the width (a circle is 3.14 times; a square is 4 times). It can
be assumed approximately that the perimeter, P, is given by P =
3.6 W,  where  W is a length scale, such  as the width. The area
where wasteful remediation  can occur is the distance between
sampling points, L, times the perimeter of the contaminated area.
  The sample density, D, is the number of samples per unit area.
Thus, L = D - '/2. Combining these definitions, the area in which
wasteful remediation is likely to occur is AR  = 3.6 WD- '/*. The
cost of wasted remediation is  given by a unit  cost, R$,  the
remedial cost per unit area times the area of wasteful remediation.
  The sampling and analysis costs will be simply the S&A costs
per sample, SA$, times the number of samples, N. The number of
samples is given approximately by N =  DW2 (sample density
times area). The total costs thus are given by:
  $ = SA$ (W2) D + R$ (3.6 W)D -
(1)
                                                                   The solution of the optimization problem is derived by differ-
                                                                 entiating with respect to sample density and setting the derivative
                                                                 to zero. This will define the sample density at which total costs are
                                                                 minimized. The result is given by:
                                                                   D=
               ^_T
               SA$J
(2)
Consider the following example:
    SA$  = $1,000 per sample
     R$  = $900/m2
      W  = 75 m
Then D = 0.069 sample/m2.
  The sampling interval, L = D -l/z = 4 m, and one sample every
4 m would optimize overall cost. Approximately 400 samples
would be indicated with a sampling program  cost of $400,000.
Contrast  this  with the cost of wasted remediation, estimated in
this example to be $810,000, even though a relatively dense sam-
pling grid is employed. If the sampling interval were 20 m,  the
sampling  costs would be reduced to  $15,000, but the cost of
wasted remediation would increase to $4,000,000. Total costs
would  be approximately $3,000,000 higher than the optimal
response.
  By inspection, Equation 2 indicates that a lower sampling den-
sity (greater spacing between samples) is optimal if the con-
taminated area is large (large W) or if remedial costs are less than
in the example. If remediation is inexpensive, then overdoing it is
less costly than sampling so there  is less need to define the con-
taminated area precisely.

GEOSTATISTICS
  Geostatistics are best  applied to normal populations that are
spatially  continuous. As normality  and  continuity  criteria
deteriorate, geostatistics  become less optimum in ability to inter-
polate. Somewhere between ideal  and non-ideal data sets, inter-
polation schemes such as method of triangles, inverse distance
weighting and kriging become useless.
  Kriging is a valuable interpolator because, unlike other geosta-
tistical methods,  kriging produces an error value for each inter-
polated value. By paying close attention to the errors, the user can
                                                                                    CONTAMINATED SOIL TREATMENT   77

-------
determine when kriging becomes a liability in determining extent
of contamination. There  is no easy way to tell when other inter-
polators fail to produce accurate estimates.
  The Semi-Variogram (SV) is the tool within  kriging that iden-
tifies the expected ability of this interpolator to work. A combina-
tion of SV characteristics and goodness-of-fit values helps iden-
tify the usefulness of  kriging before interpolation even begins.
Goodness-of-fit values are typically called jackknifing errors' and
have been the standard method of determining the usefulness of
the kriging system and its SV.
  Characteristics of the  modeled SV  are as valuable or more
valuable than jackknifing errors as indicators of the ability of
kriging to  produce  optimum  estimates.  The  important  SV
characteristics include the nugget value (y-intercept), sill value
and shape (value of gamma where the  SV levels  out) and noise.
The ideal SV has a nugget value of zero, a well defined flat sill and
minimal noise. The noise can be defined as the variability of the
SV data points about the fitted model.

Pitfalls of Using Geostatistics
  User knowledge of SV  characteristics and the Cumulative Den-
sity Function (CDF)  is  vital  to  be  confident that the kriging
algorithm actually is producing optimum estimates.
  Using jackknifing  errors alone may be misleading.  Certain
hazardous waste site characteristics tend to result in problems if
kriging is  to  be used as the interpolator. The problems  are
evidenced by SV or CDF characteristics.
  Waste sites consisting of numerous  small waste spills  inter-
spersed through generally uncontaminated  soils  most likely will
result in a statistical population dominated by outliers. The CDF
would be heavily skewed. The effects of the outliers also will be
noticeable as significant  noise and/or a significant nugget value
on  the SV. Sometimes the SV is so noisy that a model cannot be
fit to it.
  Log transformation is the simplest remedy  for this situation
but, depending on the extremity of the outliers, it  does not always
work. The transformation will usually smooth the SV, enough to
be modeled, but rarely affects the nugget value.1 Transformation
removes the extreme variance caused by the outliers but does  not
remove the baseline variance between the majority of points. In
one case,  log transformation reduced the noise but resulted in a
pure nugget (horizontal SV) indicating that  the sampling scheme
was not of sufficient density to identify spatial variability.
  Kriging this system  will tend to greatly overestimate the extent
of contamination. The peak true contamination values will be re-
duced  while the true extent  of contamination will be overesti-
mated. The interpolation scheme will not generate any new peaks
in the field but will incorrectly increase the extent of low level con-
tamination especially around the measured outliers. This becomes
significant when the cleanup criterion is near background.
  Another waste site  characteristic that causes problems is one
that contains several lagoons or larger areas of contamination in-
terspersed through relatively uncontaminated soil. A grid system
of points  covering this site most likely would show a bimodal
distribution, one peak of near background values and  another
peak of high values. The SV may show a poorly defined sill or
possibly even two sills representing the effects of the two popula-
tions of values. If the true SV model cannot be well estimated,
one of the populations will be overemphasized causing a severe
over or under estimate of the true extent of contamination. One
solution to this problem  is to work with the two populations in-
dividually to calculate individual SVs and then use both SVs as a
comparison for interpolation.
  Waste sites characterized by basin environments are conducive
to kriging. In a basin, the extrinsic variability is decreased by the
intrinsic factors involved with seasonal flooding. Each time the
system floods,  the  contaminants disperse somewhat creating a
more continuous spatial distribution and decreasing  the extreme
outliers to create a more lognormal population. These spatial and
statistical distributions are more ideal than those found on other
types of hazardous  waste sites.

Case Histories
  The  following example shows how information about extrinsic
and intrinsic factors can be used to optimize the utility of kriging.
The site is a basin environment, so the values are dominated by in-
trinsic  variability.  Boundaries of  expected contamination  were
identified from   previous investigations (extrinsic variability) and
used to help locate  the sampling grid (Fig. 2). The results from
this first phase are not available yet, so we've described one of the
probable outcomes. We assume that the eastern third of the site
was uncontaminated while the samples up to the western boun-
dary were contaminated. Our confidence along the western half
boundary is bolstered by the previous information on  the location
of the perimeter (extrinsic factors). There is less need  of sampling
along this boundary since it is more certain. The dashed line, P2,
represents the line of proposed phase  2 sample  locations based
only on seeking a precision of 35%  in estimation of contaminated
soil areas. This line would encompass 1.35 times the area cir-
cumscribed by P|. This precision of estimation of contaminated
soil  lies  in  the requirement under recently  issued CERCLA
Guidance' that feasibility study cost  estimates  be  accurate to
within  + 50% or - 30% of the actual costs. Where the perimeter
extends beyond line Pi, an additional line of samples is proposed
(P3) which again would encompass an area 1.35 times larger than
the previous area.
                                                   L-PNA
                                                      SE I
       EXPLANATION

     =, BOOING LOCATIONS FOUND 1
     u TO QE UNCONTAMINATEO
     . BORING LOCATIONS FCLNO I
       TO DC CONTAMINATED   j
     • PHASE II BORING LOCATIONS
P,  	IDENTIFIED PERIMETER
P,	MINIUM EXTENT OF CONTAMINATION
^  	I JS • AREA CIRCUMSCRIBED 8» P
Pj  	I 33 • AREA CIRCUMSCRIBED 8Y P
   	 TS\ CONFIDENCE INTERVAL
   —	 90% CONFIDENCE INTERVAL
                           Figure 2
   Sampling Scheme for Hazardous Waste in a Basin Environment
78    CONTAMINATED SOU. TREATMENT

-------
  We have less confidence in the boundary along the eastern half
of the site since it is completely estimated. A parameter,  C,  has
been used to help locate subsequent sample locations.3
  Parameter C is defined as follows:
  C =  [(Criterion) - (Est + (Conf)(KE)]/Standard Deviation       (3)
Where:
  Criterion =  Cleanup criterion
        Est =  Interpolated value
      Conf =  Confidence interval (1.28 for 90% confidence)
        KE =  Kriging error associated with interpolated point.
  The 75 and 90% confidence intervals about the boundary of
minimum extent of contamination were drawn  using parameter
C. It is obvious that many more samples are required to identify
this boundary than the western boundary due to increased uncer-
tainty in this area.
  Parameter C is unique since it combines the cleanup criterion
with the kriging error to arrive at the most likely Phase 2 sampling
locations. Areas that are significantly above or below the criterion
are eliminated  from further  consideration even if uncertainty is
large  in those areas. The use of kriging error alone to indicate
future sampling  areas discriminates based  only on uncertainty
rather than boundary locations which can be misleading.
  Kriging  was  used in  conjunction  with a phased  sampling
scheme on another site and reported by Tucker et al.3 Parameter
C was used here to identify confidence intervals for the whole site
since the expected extent of contamination was not well known on
this site. This site was not a basin area as described previously, so
it had less than ideal spatial and statistical  distributions. Log
transformation  of the  data eliminated  SV noise resulting in a
smoother curve.  The uncertainty of boundary locations was  still
significant however, and caused a large portion of the site (1/3 of
the area) to  be considered  for Phase  2 sampling. This level of
uncertainty was caused mostly by a few outliers in the data set  and
resulted in the  need  of a significant Phase 2 effort.

CONCLUSIONS
  With the advent of user-friendly computer contouring pack-
ages,  there will be an  increased tendency to misuse kriging or
overemphasize the positive benefits of kriging when used on inap-
propriate data.  Kriging along with other geostatistic methods
should not be used as standard procedure at all hazardous waste
sites.  The data must be  examined  for  statistical  and  spatial
distribution patterns before initiating geostatistics. Geostatistics
become more valuable as the spatial distribution tends to be con-
tinuous and the statistical  distribution tends to  be normal  or
lognormal. Both patterns are relatively rare on hazardous waste
sites since extrinsic factors usually dominate intrinsic  factors in
controlling  site  variability. Extrinsic factors  tend to result  in
discontinuous spatial data exhibiting statistical outliers.
  Kriging is more useful than other interpolators since it produces
an error term with each estimated value. This error can be used to
plan a second phase of sampling using Parameter C when the data
from Phase 1 sampling are less than ideal. This error also may in-
dicate that kriging or other interpolators are not useful on that
particular site.
  Finally, sampling schemes can be designed to optimize the total
costs of remediation (sampling costs + remediation costs) by op-
timizing the relationship between the number of samples taken
and analyzed versus the costs of soil remediation. By examining
this relationship mathematically, it appears that costs of extra
sampling can be easily overcome by reducing the final costs of soil
remediation.

REFERENCES
1.  Gambolati, G.  and Volpi, G., "Groundwater Contour Mapping in
   Venice  by Stochastic Interpolators. 1.  Theory,"  Water Resources
   Res., 155,  1979,281-290.
2.  Gensheimer, G.J., Flaig, E.G.  and Jessup,  R.E., "Effects of Log
   Transformation on Semi-Variogram Modeling Using Kriging Error,"
   Agronomy Abstracts, 1985, Chicago, IL.
3.  Tucker, W.A., Gensheimer, G.J.  and Dickinson,  R.F., "Coping
   with Uncertainty in Evaluating Alternative Remedial Actions," Proc.
   Fifth  National Conference on Management of Uncontrolled Haz-
   ardous  Waste Sites, Nov. 1984, Washington, DC, 306-312.
4.  U.S. EPA, Memorandum from William Hedeman, Director of Office
   of Emergency and Remedial Response, and Gene Lucero, Director of
   Office of Waste Programs Enforcement, to Waste Management Di-
   vision Directors,  Regions 1-10;  Subject: Preparation of Records  of
   Decision for Fund Financed and Responsible  Party Remedial Ac-
   tions, Mar. 1984.
                                                                                     CONTAMINATED SOIL TREATMENT    79

-------
                   Land  Treatment  of Wood Preserving  Wastes
                                                   John R. Ryan
                                            Remediation Technologies
                                               Ft. Collins, Colorado
                                                     John  Smith
                                                    Koppers  Inc.
                                             Pittsburgh, Pennsylvania
ABSTRACT
  Numerous wood preserving plants currently are evaluating dis-
posal alternatives for sludges and contaminated soils generated
at these plants. Land treatment is a potentially environmentally
sound and cost-effective means of treating these wastes on-site.
Results of four pilot-scale studies completed at wood preserving
plants in California, Montana,  Minnesota and the northeastern
United States are summarized. The land treatment studies in-
volved a detailed evaluation of  the degradation, adsorption and
toxicity  reduction of the hazardous constituents present in these
wastes.

INTRODUCTION
  This  paper presents a description of on-going industry spon-
sored studies concerning the land treatability of sludges and con-
taminated soils resulting from wood preserving operations. The
objectives of this paper are to:
• Identify the types of field  and  laboratory studies being con-
  ducted in this area
• Identify the methods being used in the treatability studies
• Present preliminary results of the on-going studies

Background
  Of the estimated 400 wood treating plants in the United States,
over 80% use creosote and/or pentachlorophenol as wood pre-
servatives. These plants may generate sludges and contaminated
soils containing the above preservatives as a result of spills, leaks
and  settled material from wastewater  impoundments. Many of
the wood preserving plants are regulated  under RCRA or  are on
the  National Priority List  and  must identify  alternative  means
for disposal of sludges and contaminated soils.
  This  paper briefly discusses four current laboratory and  field
studies   concerning the land  treatability of  wood  preserving
wastes. Table 1  summarizes the location at each  study and the
types of wastes involved. All the studies have been designed to
meet the objectives of  a  land  treatment demonstration  under
RCRA as defined in 40 CFR 270.20. These objectives are to:

• Accurately simulate proposed full scale conditions including
  soils, climate and design and operating conditions
• Demonstrate that the wastes can be degraded, transformed and
  immobilized within the treatment zone of the land  treatment
  facility

  The four studies  described  in this paper will provide valuable
data in determining whether land  treatment of wood preserving
wastes can successfully meet RCRA standards. It should be em-
phasized that additional studies concerning  the land treatability
of wood preserving wastes have been completed. These include
but are not limited to studies by Sims', Sims and Overcash2, Um-
fleet el al.\ Umfleet,4  Edgehill and Finn,' Baker and Mayfield'
and  McGiniss.' Laboratory  studies also  are  being conducted
under the sponsorship  of the U.S.  EPA Robert S. Ken Labora-
tory in Ada, Oklahoma.

CASESTUDY A
  The objective of this study was to develop design and operat-
ing criteria for full scale on-site treatment of  creosote-contam-
inated soils at a wood  treatment plant in Minnesota. The study
included both bench scale and pilot scale evaluation of several
performance, operating, and design parameters. These param-
eters include:

  Soil characteristics
  Climate
  Treatment supplements
  Reduction of organics, phenolics and PAH compounds
  Toxicity reduction
  Effect of initial loading rate
  Effect of reapplication

Soil Characteristics and Climate
  The study soil is a fine sand which comprises the  upper 2 ft of
the bottom of a RCRA impoundment  previously  used for the
storage of wastewater from a tie treatment plant. The soil is con-
taminated with creosote constituents which primarily consist of
PAH and phenolic compounds. Total benzene extractable hydro-
carbons in the contaminated soil range from approximately 8 to
12% by weight.
  The soil has a natural pH of approximately 6.5 and, therefore,
no pH adjustments were necessary prior to initiating the studies.
Specific conductance values and  metal  contents of the soil are
generally within the range of values reported  for  natural soils.
The carbon to nitrogen (C/N) ratio of the contaminated soil is
high, and nutrient additions were necessary to reduce the C/N
ratio to a range which promotes microbial growth.
  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 leachate break through.
To replicate the proposed  full scale conditions, the pilot studies
consisted of five lined, 50 ft1 test  plots  with leachate collection.
Table 2 shows the experimental conditions for each of the test
plots.
  The field studies were conducted from July 25  through Oct.
30, 1984. Average ambient temperatures generally exceeded 50° F
during this period. July and August were exceedingly dry months,
80   CONTAMINATED SOIL TREATMENT

-------
                                                            Table 1
                                                Summary of Land Treatment Studies


Fj





3C 11 i ty
A

B





Loca tion
Minnesota

C<

s 1 i f or nia
Type
of
Study
L,F

L,F






Type(s) of waste
Creos
soil
:ote
and
Creoso te
con tamina ted
sludge
and pentachlorophenol

Principle
Con taminants
PAH, T. Phenol

PAH, clorinated phenol





ics
contaminated soi 1




C

D

Montana


North East


F

L

Creosote
soil
Cr eoj
con tf
and
>ote
imi n;
contamina ted
sludge
and pentachlorophenol
ated soil
PAH, T. Phenol



PAH, chlorinated phenolic s


L— Laboratory Study
F— Field Study
and the field  plots required weekly watering to maintain them
near field capacity. October was a very wet month, and the plots
were near saturation during this  period. The active degradation
period appears to extend through the month of October at this
site.

Treatment Supplements
  The studies were designed to maintain soil conditions which
promote the  degradation of hydrocarbons. These conditions
include:
• Maintain a pH of 6.0 to 7.0 in the soil treatment zone
• Maintain soil carbon to nitrogen ratios of 25 to 1
• Maintain soil moisture near field capacity
  In addition, the studies evaluated the effect of seeding the soil
with commercially available microbes adapted to hydrocarbon.
  The soil pH was within the desired range at startup and no lime
additions were necessary. Soil pH decreased below 6.0 in most  of
the plots after 2 months of operation. Subsequent additions  of
1.0 to 1.5 tons/acre of agricultural grade limestone raised the pH
of the test plots to the desired range and no subsequent drops in
pH were observed during the test period.
  The equivalent of 10-20 tons/acre of 10:10:10 fertilizer (10 Ib
nitrogen, 10 Ib phosphorous and 10 Ib potassium per 100 Ib fertil-
izer) were added to all the plots at the  beginning of the study
based on the estimated carbon content of each  plot. The fertil-
izer additions  successfully reduced the carbon to nitrogen ratios
of each plot below 25:1. Residual ammonia  levels following the
fertilizer application were fairly high, however,  creating a con-
cern about possible ammonia toxicity problems.
  All the test plots were watered based on the results of daily rain-
fall and pan evaporation data as  well as daily soil  tensiometer
reading. Soil irrigation was necessary on a weekly basis during
most  of the study. The monitoring and irrigation program  suc-
cessfully maintained the soil near field capacity. No significant re-
ductions in organics were observed in the plot which was not
watered. These data illustrate that maintaining the soil  near field
capacities is a critical  operational  parameter for obtaining  suc-
cessful degradation. Field capacity of the fine sands appears to be
approximately 10% moisture by weight.
  No significant advantages in organic degradation can be attrib-
uted to seeding the plots  with the adapted microbial culture used
in the study. It is hypothesized that an active natural soil micro-
bial population exists in the contaminated soil. Due to the length
of time the soil contamination existed, the existing soil microbes
are well adapted to the contaminants present in the soil.

Reduction of Organics, Phenolics and PAH
  Reductions of benzene extractable hydrocarbons were fairly
similar between all the field plots.  Similar reduction trends were
observed in the laboratory reactors, however, at lower rates.  Per-
cent removal efficiencies and first order kinetic rate  constants
were fairly similar for all the field plots. Average removals for all
field plots over the 4 months were approximately 40% with a cor-
responding first order kinetic constant (K) of 0.004.
  The greatest mass of benzene extractable removals  was asso-
ciated with the highest initial loading rates. Total  removals  over
the  4-month test period  on the basis of  pounds of  organics re-
moved/ft3 of soil/degradation month are presented in Table 3.
                                                            Table 2
                                              Experimental Specifications for Test Plots
Test
Plot
1

2

3

t,

5

Soil Initial Oil1
Specifications Content (%)
Con t aminated
Hydrobac CL
Contaminated
Hydrobac CL
Contaminated
no microbial
Contaminated
Hydrobac CL
Visibly non-
Hydrobac CL
soil and clean sand; 10

soil and clean sand: 5

soil and clean sand, 5
seed
soil and peat moss, 5

contaminated soil, 3

Fertilizer Cultivation
(Ib) Frequency
50 Weekly

24 Weekly

25 Weekly

25 Weekly

15 Weekly

                 'Benzene Extractable Hydrocarbons.
                 •10-10-10 Fertilizer.
                                                                                     CONTAMINATED SOIL TREATMENT    81

-------
                           Table 3
         Total Benzene Extractable Removals In Field Plots

Extractable Content
Plot Number (percent by weight)
1 8.83
2 4.00
3 4.06
4 10.59
5 1.73
Total B
ExtracLnbl
( Ibs/f t of
0
0
0
0
0
e n z e ne
e Removals
soil/month)
.94
. 56
. 34
.91
.20

   These removal rates can be compared to data from petroleum
 refinery land treatment facilities.' Typical  removal rates at well
 operated refinery facilities range  from 0.1  to  1.0  Ib of oil  re-
 moved/ft5 of soil degradation month with  the lower end corres-
 ponding to a "low" initial concentration (such  as in Plot 5) and
 the high end corresponding to "high" initial concentrations such
 as Plots 1 and 4. The data, therefore, are in good agreement with
 removals documented at refinery land treatment  facilities.
   Phenol removals did not exhibit the same linear relationship
 over time as the  benzene  extractable contents.  Phenol removals
 were observed over the first three  months followed by a market
 increase in the phenol content of  the soil in the fourth month.
 The trends observed may  be  partially a result of analytical vari-
 ability but also may be affected by the breakdown of complex
 organic compounds releasing phenolic constituents.
   First order rate kinetic  constants for PAH were fairly similar
 between the field and laboratory studies with the exception of 4 +
 ring compounds in test runs 1 and 4.  The laboratory tests had
 higher first order rate constants than the field plots for these com-
 pounds.
   Table 4 compares the  first order  rate constants for benzene
 extractables and  PAH compounds between the laboratory and
 field studies which had an initial benzene extractable content of
 4 to 5% and the laboratory and field studies which had an initial
 benzene extractable content of 8 to  10%.
   The median kinetic values (of the 4 to 5% studies and the 8 to
 10% studies) are approximately equal for all the parameters. The
 most notable variance is for the 4 + ring PAH compounds. The 4
 to 5% initial loading rates resulted in slightly higher kinetic rates
 for these compounds as compared  to the 8 to 10% initial loading
 rate. The kinetic  rates appear independent  of the initial loading
 rate within the range of loading rates tested.
                          Table 4
             Comparison of the Range and Median
           Kinetic Values at Two Initial Loading Rales
                     4 to 5 percent
                 Range
                      (K. day-')
                                Median
                                                to 10 percent
Range       Median
     (K,  day-')
Ben zene

Extractables
2 Ring
3 Ring
4 + Ring
PAH
PAH
PAH
Total PAH

0,
0,
0.
0.
0,

.001-0.
,021-0.
.014-0.
.002-0.
.008-0.

007
024
017
007
Oil

0.
0.
0.
0.
0

,003
,023
.016
.004
.009

0,
0.
0,
0.
0.

.002-0.
.01 3-0.
,013-0.
,000-0.
006-0.

004
033
023
005
013

0.
0.
0.
0.
0.

003
023
016
001
008
                        Toxicity Reduction
                          A battery of toxicity assays was completed on 3 test plots (1,2
                        and 5) at the completion of the 4-month study as well as on creo-
                        sote sludge and contaminated soils which approximately repre-
                        sented the start up conditions in plots 1 and 2 at the beginning of
                        the study. The battery of assays included:
                                                                  • Microtox
                                                                  • Ames mammalian mulagenicity
                                                                  • 96-hour static acute fish bioassay
   The battery of assays resulted in the following observations:
 • Test plots 2 and 5 were non-toxic after 4 months
 • Test plot 1 showed intermediate toxicity after 4 months
 • Both contaminated soils and the creosote sludge were toxic to
   all the bioassays

   The microbial assays (microtox and Ames test) on the contam-
 inated soils would  have suggested that  the  initial concentrations
 in test plots 1 and 2 would have resulted in a toxic response to the
 soil microorganisms  and  therefore no significant  degradation
 would occur. The results from these plots, however, have shown
 significant degradation.  This indicates that  an acclimated micro-
 bial population existed in the contaminated soil prior to the initia-
 tion of the studies.  The creosote constituents (within the range of
 initial concentrations tested) do not  result in toxic effects  to the
 acclimated soil microorganisms used in the toxicity tests.
   From an environmental toxicity standpoint, however, the toxic-
 ity tests show  that the toxicity of the contaminated soil has been
 decreased through  the treatment process.  Complete toxicity re-
 duction appears to fall between 2.5 and 5% benzene extractable
 content. Complete toxicity reduction was achieved after 4 months
 in plot 2.

 Effect of Initial Loading Rate
   First order rate constants were fairly similar between all  the
 loading rates.  The  intermediate loading rate (4 to 5% benzene
 extractionables)  may demonstrate a slightly higher removal of
 high molecular weight PAH compounds. Toxicity reduction also
 appears to occur at a faster rate for a 4 to  5% initial loading rate
 than at the higher initial loading rates.
  The higher loading rates, however, showed the greatest mass
 removals  and  a  clear detoxification trend. The studies suggest
 that all the loading rates tested are feasible. The  selection  of an
 initial loading rate  should balance additional land area require-
 ments against  time requirements for completing the treatment
 process. Moderate loading  rates (5%) will result  in  a faster  de-
 toxification  whereas higher loading rates will decrease  land area
 requirements.

 Effect of Reapplication
  Greater kinetic rates  were observed  after waste reapplication
 to a treated soil. The results suggest that average loading of 1 Ib of
 benzene extractable/ft' of soil/degration month to a treated soil
can be treated effectively when the application does not exceed
3.0 Ib/ft.1 The  first application should occur after the initial
application has been shown  to be successfully treated.

Operating and Design Criteria
  The studies  have successfully developed operating and design
criteria for a full-scale system at this site. These criteria include:
• Treatment period can be extended through October
• Soil moisture should be maintained near field capacity
• Soil pH should be maintained between 6.0 to 7.0
• Soil carbon:nitrogen ratios should be maintained at 25:1
82    CONTAMINATED §OIL TREATMENT

-------
• Fertilizer applications should be completed in small frequent
  doses
• Initial benzene extractable hydrocarbon contents of 5 to 10%
  are feasible
• Bioassays can be used to determine treatment effectiveness
• Waste reapplication should occur after initial soil concentra-
  tions have been effectively degraded
• Waste reapplication rates of 2 to 3 Ib of benzene extractables/
  ft3 of soil/2 degradation months can be effectively degraded
  The  studies completed  have shown that the visibly contam-
inated  soil at the site can be effectively degraded and detoxified.
The data indicate that the organic content of  the visibly contam-
inated  soil present in the upper two feet of the RCRA impound-
ment can  be degraded to levels similar to the organic content of
the soils beneath the visibly contaminated soils. The data also
indicate that the treatment is effective in detoxifying the waste.

CASE STUDY B
  This study involves a field pilot program to evaluate the feas-
ibility  of land  treatment of contaminated soils from a wood pre-
serving plant in California.  Principal hazardous constituents  in-
clude polynuclear aromatic hydrocarbons (PAH), pentachloro-
phenol, dioxins and furans. The studies were initiated in the sum-
mer of 1985 and may be continued through the fall of 1986.

Treatment Medium Soils
  The test plot soils  are located  east of a spray irrigation field
which is proposed as the full-scale treatment area. The general
characteristics of the surface soils at the  spray field and the test
plots nearby are characterized as clayey silts which grade to silty
clays with depth.  Depth to groundwater in the area is generally
30  to  40  ft beneath the ground surface.  Specific soil character-
istics of the test plots based on core samples  to greater than 7 ft
show  the soils to be relatively homogeneous with depth. The
soils through  the test plots are characteristic of a brown silty
clay, tight to very tight, with small gravel clasts to 1 in.

Field Plot Experiments
  Three experimental field plots were constructed at the facility
 near the  spray irrigation field. The plots were graded to a  1%
 slope and earth berms were constructed between the plots. Run-
 off is completely collected at the low slope position of each plot.
The collection areas for the plots are sized to collect the run-off
 from the 25-year, 24-hour storm (i.e.: approximately 4.5 in).
  Also included in the plot design was  the  installation of soil-
 pore liquid samplers. The samplers are used to collect water from
 the unsaturated zone for the purpose of determining if hazardous
 constituents are migrating out of the treatment zone.  The plots
 were fitted  with three types of lysimeters per plot: a commer-
 cially  available vacuum lysimeter, a fabricated  glass block lysi-
 meter and a gravity flow trench lysimeter. During the course of
 the study, the vacuum and glass block lysimeters have proven to
 be  severely limited in extracting  water for analysis. The trench
 lysimeter has worked effectively  in  collecting macropore flow
 through the treatment zone which is the most critical factor in
 determining if hazardous constituents are moving through  the
 treatment zone.
  The plot design also included the installation  of tensiometers.
 Tensiometers  are important in monitoring  soil suction  or  the
 force that determines which way moisture will move in soil. For
 the study, tensiometers are used to measure moisture in the zone
 of  incorporation and at 18 in. to determine irrigation scheduling
 and at 5 ft to  monitor soil moisture movement through the treat-
 ment zone of the three plots.
Startup
  After the plots were constructed and the treatment area was
readied, the contaminated soil was spread and  mixed in place
with native soil to achieve an approximate benzene extractable
content of 2% in plot 1  and 3000 ppm of pentachlorphenol in
plot 2. Sewage sludge was applied at a rate of 20 tons/acre to
Plots 1 and 2 to increase the organic matter of the contaminated
soil as well as reduce the  carbon:nitrogen:phosphorous ratio to
a range to promote microbial growth. Also, agricultural lime was
spread as a step to raise soil pH to 6.0 to 7.0. The final step of
startup activities was the mixing of contaminated soil, fertilizer
and lime with a tractor mounted rototiller to a depth of 6 to 8 in.
  Table 5 summarizes the operational activities conducted dur-
ing the pilot scale studies. These activities included test plot culti-
vation, irrigation, sampling and on-site climatological and sub-
surface monitoring.
                           Table 5
                  List of Operational Activities
Activity
Test Plot Cultivation
Lime Additions
Fertilizer Additions
Moisture Additions
On site monitoring
-soil temperature
-pan e va poration
-ra in fa 11
-soi 1 suction
-soil moisture
Freq uenc y
Weekly
As necessary to maintain pH 6-7 '
As necessary
As necessary

Daily
Daily
Daily
Weekly
Weekly
 'Determined from monitoring data.
Soil and Water Sampling
  Soil samples were collected from Plots 1 and 2 at 2-week inter-
vals for the first 6 weeks of the study and then monthly until the
end of the study. Each plot was subdivided into three subplots for
sampling purposes. The subplots were each 12 ft by 16 ft  with
1-ft grid spacings for a total of  192 grid squares per plot. Four
subsamples were collected in each subplot during each  sampling
event. These four subsamples were then composited into one sam-
ple for each subplot. Each  subsample was collected from one of
the grid squares. The four grid squares sampled in any one event
were selected from a random number table. Each grid square was
sampled only once during the course of the study. The random-
ized sampling was recommended as part of quality assurance pro-
cedure to account for field variability.
  Also, during the study, nine grab samples were taken from one
subplot during one sampling event as well as the normal com-
posite sampling. This sampling was done to compare composite
sampling versus grab sampling.
  Soil core sampling was conducted prior to waste application,
60 days and 90 days after waste application. Cores were taken in
each  subplot including the  control plot. The cores were collected
every 1 ft to a depth of 7.5 ft.
  Two water samples were taken during the study. Runoff sam-
ples were  collected  from  each  plot  in  September,  and trench
lysimeter samples were collected from each plot in December. All
samples were shipped in coolers packed  with ice under standard
chain-of-custody procedures. Sampling equipment  was decon-
taminated with deionized  water and methanol  after each sam-
pling round.
                                                                                     CONTAMINATED SOIL TREATMENT    83

-------
                           Table 6
                 Range of PAH Kinetic Constants
 Subplot

   1A

   IB

   1C

   2A

   2B

   2C
2 Ring

0.020

0.014

0.035
0.019

 (1)

 (1)
3 Ring

0.200
0.014

0.029
0.027

0.020

0.018
4 Ring

0.006

0.007

0.027

0.017

0.012

0.001
Tola 1

0.010

0.009

0.028

0.019

0.013
0.004
(1) not delected at startup.
Organic Reduction Rates
   Complete analytical results for  the study were not available
during the preparation of this paper. Preliminary results, how-
ever, indicate that benzene extractable removals in  each of the
subplots ranged from 34-66% for the 4-month time period with
a  corresponding  first order kinetic constant (K) ranging from
0.003 to 0.009.
   Table 6 summarizes the first order kinetic constants computed
from the preliminary analytical results for various ring classes of
PAH compounds. The following compounds are listed under the
noted ring classes:
2 Rings           3 Rings         4+ Rings
Naphthalene      Fluorene        Fluoranthene
Acenaphthylene    Phenanthrene    Pyrene
Acenaphthene     Anthracene      Benzo (a) anthracene
                                Benzo (a) pyrene
                                Benzo (b&k) flouranlhene
                                Chrysene
                                Benzo (ghi) perylene
                                Indeno (1,2,3-cd) pyrene
                                Dibenzo (a,h)anthracene
   Considerable variability was observed in the pentachloraphenol
(PCP) data. Table 7  summarizes the measured values over the
course of the study  and the corresponding  calculated removal
rates. The  PCP data are non-conclusive and continuations of
these studies into 1986 are necessary  to draw any  treatability
conclusions.
  As a byproduct of PCP production, furans and dioxins may be
present at variable concentrations. Tetra and 2378 furans and
dioxins were not detected at startup in Plot 1  and  were below
5 ppb and 1 ppb for the Tetra and 2378 furans and dioxins re-
spectively in Plot 2. The remaining furans and  dioxins  (Penla,
Hexa,  Hepta) show removals after 4 months  between  75 and
95%.  The octa furans and dioxins have the least percentage re-
movals with values ranging from 0-81%.
  In summary, significant removals of furans and dioxins were
observed over the 4 month period, and initial startup concentra-
tions were relatively low in comparison to PCP and PAH startup
concentrations.
  The studies may be extended  through 1986 to evaluate  long-
term performance of the plots and evaluate the immobilization
of the contaminants.

CASE STUDY C
  This study is an  on-going  RCRA land  treatment  demonstra-
tion for a proposed full scale land treatment facility in Montana.
The wastes being evaluated are creosote contaminated soils and
sludges resulting from the closure of a RCRA surface impound-
ment. The proposed full scale land treatment facility will be used
to treat and detoxify the wastes removed from the impoundment
at closure. Design parameters for the field studies were developed
based on results of Case Study A.

Soil and Climate
  The treatment area soils consist of loamy silts  and silty sands.
The caution exchange capacity of the surface soil is approximate-
ly 20 milliequivalents/100 g of soil and the organic matter content
of the surface soil is approximately 1.5%.
  The climate of the area is  semi-arid. Annual  evapotranspira-
tion exceeds annual precipitation and there is limited recharge to
the  groundwater  resulting  from  precipitation and run-off.
Monthly ambient  temperatures  exceed 50° F approximately 6
months of the year.

Description of Study
  The soil and climatic conditions at the site are considered high-
ly favorable for attenuation of the creosote contaminants. There-
fore, unlike  Case Study A, this system is  not lined.  The results
                                                           Table 7
                                       Summary of Pentachlorophenol Removals in Plots 1 and 2
                                   Concentration  (ppm)

                                   at Time (t)  Daya
                                                               Percent of  Removal
                                                               at Ti«e(l)  Days(l)
                                      Da to
                                                                                    Date
                          7/12/85 8/16/85 8/27/85 10/10/85  11/11/85  8/16/85 8/27/85  10/10/85  11/11/85
                     Plot   t-0     t-35    t-46     t.90      1-122    t-35    t-46     t-90      1-122
PIA
PIB
PIC
P2A
P2B
P2C
120
130
160
6800
1400
1700
1 10
95
69
5300
5200
4200
260
1 20
150
8200
5700
5600
54
44
48
1500
2300
1900
210
1 10
1 1
310
1000
480
8
27
57
22
0
0
0
8
6
0
0
0
55
66
70
78
0
0
0
1 5
93
95
29
72
                    (I) Percent Removal at Time   (C(O)-C'(T)/C(O))IOO.
84    CONTAMINATED SOIL TREATMENT

-------
of Case Study A, however, were used to select the design param-
eters  for the study including waste application rate and treat-
ment supplements.

Waste Application Rate
  Waste application rates are based on gross organic loading as
defined by benzene extractable hydrocarbons. The loading were
selected on the basis of results from Case Study A and involved
initial loadings of 2% and 5% benzene extractables. The 2% plot
includes subsequent bi-monthly applications of 2% benzene ex-
tractables to evaluate the effect of waste re-applications. The 5%
plot does not include waste re-application in order to evaluate the
effect of one batch loading.

Treatment Supplements
  To minimize  the need for replicate plots and minimize the
analytical costs of the study, certain management practices were
selected as "good management practices" for the study and were
applied consistently to all the plots.  These "good management
practices" are anticipated to optimize treatment (based on liter-
ature data) and include:
• Maintain surface soil pH between 6.0 and 7.0
• Maintain a soil carbon:nitrogen ratio of 25:1
• Maintain soil moisture near field capacity through irrigation
• Apply manure or organic  amendments  to provide  nutrients
  and increase the organic carbon content of the surface soil
• Till the soil once every 2 weeks
• Prevent storm water run-on from entering the plots and capture
  run-off

Construction of the Plots
  The field plots were constructed in June  1985. One plot is a
control where no  wastes have been  applied but  all the  "good
management practices" are implemented. The other two plots are
for the two waste loading rates being evaluated.
  Each plot is 12 ft x 50 ft with a defined treatment depth of 5 ft.
The plots are separated by earthen berms which prevent run-on
from entering the site. Each plot is graded to a  1 % slope,  and a
lined run-off collection sump capable of handling two 25-year,
24-hr storm events is present  at the toe of the slope. The  entire
study area is fenced:
  Three types of soil-pore water collection devices were installed
in each plot. These include:
• A vacuum lysimeter capable of placing a suction on the sur-
  rounding soil
• A pan lysimeter constructed from a hollow glass block approx-
  imately 1 ft2
• A trench  lysimeter consisting of a 10-ft long 6-in.  diameter
  PVC drain, placed at the bottom of the treatment zone
  In addition soil tensiometers were placed at 3 ft and 5 ft. These
tensiometers in conjunction with a hand tensiometer used to take
soil suction readings at the  surface  are used  to evaluate  the
hydraulic flux through the soil.

Monitoring
  Monitoring at the site includes monthly sampling of the zone of
incorporation to evaluate the degradation of specific organic con-
stituents and bi-annual sampling of soil cores, soil pore water and
groundwater to evaluate contaminant immobilization.
  Table 8 presents the analytical  parameters evaluated in each
medium. The principal hazardous constituents  being evaluated
are polynuclear aramatic hydrocarbons PAH. These constituents
are monitored on a monthly basis along with indicator  para-
meters such as benzene extractable hydrocarbons and total phen-
                          Table 8
                 Analytical Parameters, Site C

               Parameter/Method  	
                                                    Medium
                                                      Z

                                                      C

                                                      Z

                                                      Z

                                                    2,0,1

                                                    Z,C,L

                                                      L

                                                     Z,C

                                                     C.L

                                                     Z,L

                                                      Z
-Benzene Extractables/Soxhlet extraction

-Metals/ICAP

-Total kjeldahl nitrogen/automated titrimetric

-Ammonia nitrogen/automated titrimetric

-pH/combination electrode

-Conductivity/conductivity bridge

-Total organic carbon/combustion or oxidation
-PAH/Method 8270

-Base neutrals/Acid extractables

-Microtox/Toxicity Analyzer System

-Fish bioassay
Z   zone of incorporation soil samples.
C = core soil samples
L   lysimeter water samples.
ols. In addition, nutrients,  soil electrical conductivity and soil
pH are monitored on a monthly basis to evaluate the need for
treatment supplements. Soil moisture is checked on a weekly basis
to evaluate  the  need for supplemental irrigation.  The relative
toxicity of the surface soil is evaluated through monthly micro-
tox analyses.
  Bi-annual soil core sampling includes collection of soil cores in
1 ft increments down to the bottom of the treatment zone (5 ft).
The soil cores, groundwater and lysimeter samples are analyzed
for indicator parameters and for a complete scan of base neutral
and acid extractable organics using GC/MS techniques. Inorgan-
ic metals also are included in the analyses using ICAP techniques.
  All sampling and analyses include triplicates from each plot to
determine statistical variability. The startup of the study was in
July 1985, and it is scheduled to be completed in 1986. Insuffic-
ient analytical data were available at the time of this writing to
evaluate degradation trends. However, the soil core, lysimeter
and groundwater analyses have not detected  any migration of
hazardous constituents from the treatment zone.

CASE STUDY D
  This project involves a detailed laboratory  study using creo-
sote and pentachlorphenol  contaminated soils from an  aban-
doned wood preserving  plant in the northeastern United States.
The study was initiated in December 1985 and is scheduled to run
for 8 months.
  The laboratory  study program evaluating  land  treatment  is
divided into the following three areas:
• Evaluation of leachate movement through  soil  columns and
  contaminant volatilization
• Examination of contaminant degradation rates in soil pans
• Development of soil/water partition coefficient for PAH and
  pentachlorophenol associated with the site soil

Soil Column Evaluation
  The studies outlined below are designed to evaluate the rate
and extent of treatment, including biodegradation  and retarda-
tion. A mass balance of selected constituents will be determined
by taking into consideration the initial loading rate, the concen-
tration of constituents in the leachate for the  duration of the  8-
month experiment and the concentration of constituents in the
soil core at the termination of the experiment.
  The soil treatment process will be evaluated using 4-in. diam-
eter, 5-ft long glass column reactors packed with clean soils  to
represent actual site  conditions. The columns contain the same
                                                                                     CONTAMINATED SOIL TREATMENT    85

-------
soil  profile materials that exist at  the site.  Native  soils  were
collected in 6-in. intervals and packed into the columns in the
same order as obtained in the field to obtain a representative
soil column profile.
   Before any contaminated material was applied to the soil col-
umns, chloride tracer studies, using sodium chloride, were  per-
formed. This work was completed to determine the proper hy-
draulic loading rate once contaminated material is applied.
   Actual  contaminated soil  from the site was mixed  into the
upper 6 in. of the clean soil in a porportion to give approximate-
ly 5% benzene extractables. Common commercial fertilizer and
manure also were added  and the  soil pH was adjusted to  and
maintained between 6-7 by lime additions. The columns are to be
operated for approximately 8 months with only one initial waste
application.  Watering of the columns is to be done with its pH
adjusted to approximately pH 4. The pH adjustment is done to
simulate the existing pH of rain water for the area.  In  addition
to routine watering, the columns are to be watered with the equiv-
alent of a 25-year, 24-hr storm event in January and May during
the study. The soil is to be tilled every other week. Volatile emis-
sions are to be addressed as part of the column work.
   For QA/QC purposes, two duplicate columns containing the
waste material are operated. A third column containing  only
clean soils serves as a control and is operated in the same man-
ner as the columns with contaminated material. All three columns
are to be sampled periodically for analyses.

Soil Pan Reactor Evaluation
   In conjunction with the operation of the soil columns, a soil
pan study phase is to run  for approximately 8  months  with
degradation rates of specific compounds addressed. Four soil pan
reactors are operated. The soil pans are constructed of aluminum,
have a surface area of approximately 1.5 ftj and contain approx-
imately 6 to 8 in. of soil. The soil is a mixture of contaminated
material and clean surface soil from the site in a proportion to
give a concentration of approximately 5% benzene extractables.
Fertilizer and manure have been added to give a C:N:P:ratio of
approximately 50:2:1, respectively. Manure was applied corres-
ponding to a loading rate approximately 10-20 dry tons per acre.
This is the same soil misture that is to be used in the column re-
actors. The soil pH is maintained  within  a range of 6-7 by lime
addition as needed, and the  soil moisture is  maintained within
50% of field capacity by laboratory tap water adjusted to a pH of
4. During the course of the experiment, hand tilling is to be done
every other week.
   For QA/QC purposes, three duplicate pans containing the soil
waste material lime, fertilizer, manure are to  be set  up. Two of
the pans are  to be maintained in a moist environment by water
applications as needed. The third pan is not to be watered and is
to serve as a contaminated  soil control.  A fourth pan, contain-
ing only clean soil, is to be operated in the same manner as the
contaminated soil pans.

Soil/Water Partition Coefficient Development
   Soil/water partition coefficients are to be developed for penta-
chlorophenol and PAHs as these compounds specifically relate to
the contaminated  site soils. For this work,  different solution
pHs are to be evaluated to examine the effect of pH on the leach-
abilities of PAHs and pentachlorophenol.
   This  information is needed to develop a better  theoretical
understanding  between pentachlorophenol and  PAH attenua-
tion by the soil matrix.
CONCLUSIONS
  Land treatment is a potentially technically effective and eco-
nomic means of treating and disposing of wood preserving wastes
from sites currently regulated under RCRA or CERCLA. Rigor-
ous 2-year studies are currently under way to determine appro-
priate engineering design features for these systems.  The data
from those studies will provide a sound scientific basis for eval-
uating the feasibility of this technology at other wood preserv-
ing sites.
REFERENCES
1.  Sims,  R.C., "Land  Treatment  of Polynuclear Aromatic Com-
   pounds." Ph.D. dissertation. Depi. Biol. Agr.  Eng.. No. Carolina
   State Univ., Raleigh, NC, 1982.
2.  Sims, R.C. and Overcash,  M.R., "Fate of  Polynuclear Aromatic
   Compounds in Soil-Plant Systems," Residue Reviews 88. 1983, 1-68.
3.  Umfleet, el al..  "Reclamation of PAH  Contaminated Soils."  Pre-
   sentation made at  Environmental Engineering 1984. Specialty Con-
   ference, ASCE, Los Angeles, CA. June 1984.
4.  Umfleet, D.A.,  "Land Treatment of PAH Compounds,"  Unpub-
   lished M.S. thesis. Department  of  Civil and Environmental Engi-
   neering, Utah State University, Logan, UT, 1985.
5.  Edgehill and Finn, "Microbial Treatment of Soil to Remove Penta-
   chlorophenol. " Appl. Environ, .\ficrobiol. -IS.  1981, 1120-1125.
6.  Baker, M.D. and  Mayfield, C.I.. Water. Air and Soil Pollul., 13,
   1980,411.
7.  McGinnis, G.D.,  "Biological and  Photochemical  degradation of
   pentachlorophenol and creosote," Mississippi Forest Products  Lab-
   oratory. Mississippi Slate, Ml, 1985.
8.  American  Petroleum  Institute. "Land  Treatment  Practices in the
   Petroleum Industry." API. Washington, DC, 1983.
86    CONTAMINATED SOIL. TREATMENT

-------
                      Method  for  Determining  Acceptable  Levels
                                of Residual  Soil  Contamination

                                                 William A. Tucker
                                                   Carolyn Poppell
                                 Environmental Science and  Engineering, Inc.
                                                 Gainesville, Florida
INTRODUCTION
  With the exception of PCBs, regulatory standards or estab-
lished numerical criteria that define acceptable levels of soil con-
tamination at  hazardous waste sites do  not  exist.  However,
numerical criteria are readily available for air and water. In con-
trast to air and water, soils are not fluid and are more readily sub-
ject to property rights. People experience little direct exposure to
contaminated soils.  For example,  adults inhale approximately
17,000 I/day and ingest  roughly 2  I/day  of  water, but they
generally are believed to accidentally ingest less than 0.00005 1 of
soil per day. For these and other reasons, Congress has not
enacted the "Clean  Soils Act," and  soil quality standards have
not been  implemented.
  The extent of human exposure to  residual contamination in
soils can be extremely  variable, depending on  such  factors as
climatic factors and  soil properties. The most important variable
is land use.  Contaminated soils in a  backyard playground and
subsistence garden (common in many residential areas) pose a
much greater health risk than in a paved shopping mall develop-
ment.
  Consequently, it is reasonable to determine acceptable levels of
residual soil contamination on a site-specific basis, considering
the probable (or  possible) future uses of the land  during  the
period that it will  remain contaminated. Many feasibility studies
for remediation of contaminated sites and remedial actions im-
plemented at sites reflect  this  concept. On the other  hand, im-
plications of this concept have been addressed only implicitly or
qualitatively in many Remedial Investigation/Feasibility Studies
(RI/FS)  conducted  to  date.  Systematic and  quantitative im-
plementation of this concept can be  accomplished with multi-
media exposure and risk assessment procedures that are readily
available  and well  documented.
  The purpose of this paper is to present these procedures in the
context of the endangerment assessment process as they currently
are implemented by the authors. The methods used by the authors
are similar  to the  PPLV method  developed  by Dr.  David
Rosenblatt and co-workers of the U.S. Army Medical Bioengi-
neering Research  and  Development Laboratory; the  interested
reader should consult Rosenblatt and Dacre7, Rosenblatt, et a!.'
and Small9 for further discussion of their procedures.  Following
the presentation of the method, a few examples will be presented
to illustrate some features of its implementation.

APPROACH
  Establishment of acceptable residual soils concentrations often
will be an essential element of the endangerment assessment pro-
cess. Endangerment assessment links the remedial investigation to
the feasibility study in two ways:
• The process results in determination that an endangerment or
  public health risk exists. If the  determination is positive,  it
  virtually rules out the no-action alternative and demonstrates
  the need for  the feasibility study.
• The endangerment assessment should clearly identify the en-
  vironmental  quality  criteria that would be required  for the
  CERCLA alternative; that is criteria which, if met, would sub-
  stantially eliminate the endangerment.
  The establishment of acceptable levels  of residual soils con-
tamination contributes to  both objectives of the endangerment
assessment. Endangerment may be suspected if the "endanger-
ment criteria" are exceeded. The role of "endangerment criteria"
in the RI/FS process is presented in Figure 1.
  The recommended approach for determining acceptable levels
of residual soils contamination is presented in Figure 2. Results of
the  remedial investigation  will identify the contaminants of con-
cern for the site. Inputs from  the endangerment assessment team
will be useful at an early  stage in the remedial investigation to
refine the list of contaminants  of concern to focus RI resources on
the  "riskiest"  contaminants.  Further refinement and  screening
undoubtedly will be required as the RI data are acquired and re-
viewed.  In some circumstances a formal hazard ranking system
may be useful to screen contaminants, and applications of a for-
mal screening  model may be a component of the uppermost
"box" of Figure 2.
   Once contaminants of  concern have been identified, investiga-
 tions proceed along two parallel and simultaneous paths which in-
 terface at intermediate points in the  process as well as in the  in-
 tegrative final steps. The two paths  are contaminant fate (down
 the right in Fig. 2) and contaminant  toxic effects (down the left).


 Effects
   First, the available toxic effects data for the contaminants of
 concern are reviewed. It is important to identify any particularly
 sensitive subpopulation (e.g.,  children, women of childbearirig
 age, aquatic species or age classes). Considering this information
 on sensitive subpopulations  and  site-specific factors (e.g.,  in-
 dustrial site where children  would not be  exposed) permits the
 identification of populations at risk.  Populations at risk are those
 potentially exposed populations who also are sensitive to the con-
 tamination. The most sensitive population at risk normally will be
 considered as the basis of the endangerment criteria. However, if
 the number of individuals in  the most sensitive population at risk
 is likely to be very small, it may in some instances be reasonable to
 base the criteria on another  sensitive population while allowing
 for special or auxiliary protection of the special, rare subpopula-
 tion.
                                                                                  CONTAMINATED SOIL TREATMENT    87

-------
REMEDIAL INVESTIGATION
COLLECT AND ASSESS SITE
SPECIFIC DATA


DEVELOP PRELIMINARY
ENOANOEFIMCNT CRITERIA
                                                                             NO FS NEEDED
                                                                          NO ACTION ALTERNATIVE
                                                                             ACCEPTABLE
                                                             Figure I
                                           Role of "Endangcrmenl Criteria" in RI/FS Process
   The next step in the effects side of the process is to determine
 acceptable exposure levels for  the sensitive population at risk.
 Determination of an acceptable exposure level is often the most
 controversial step in the process depicted in Figure 2. Much of the
 effort, techniques and procedures frequently referred to as carcin-
 ogenic  or quantitative  risk assessment  are  almost exclusively
 devoted to this element of the process.  Later in this paper, it will
 be shown that other aspects of the process also introduce substan-
 tial uncertainty and variability in estimates of acceptable residuals
 levels. For carcinogenic contaminants, this step also introduces
 the controversial issue of acceptable risk: recent  policy guidance
 from the U.S.  EPA indicates  that a 10-*  lifetime  individual
 cancer  risk  level  should  be evaluated  as a cleanup goal  at
 CERCLA sites, consistent with  its guidance for groundwater
 Alternative Concentration Limits,  a site-specific standard setting
 mechanism under RCRA. Alternative cleanup standards resulting
 in lifetime individual cancer risks  in the range of  10 ~8 to 10-4
 also should be considered.

 Land Use Scenario
   As illustrated in  Figure  2, a critical requirement of the soil
 criteria setting procedure recommended here is the need to define
 an exposure scenario. As used in this discussion,  an  exposure
 scenario is a characterization of  the behavioral  patterns  of  the
 population at risk that could result in, or reduce, exposure to  the
 contaminants at the site.  The  critical  element of the  exposure
 scenario will be  the land use scenario used in the analysis.
   Depending upon  the degree  of conservatism  that may seem
 warranted to affected parties, and the persistence of site-related
 contaminants, the exposure scenario may be based on: (1) the  ex-
 isting land use, (2) probable future land  use over a  specified
 future period, based  on existing  or predicted patterns of land
 development and (3) the use (of all possible future uses) that leads
 to the greatest  potential for exposure.  The choice among these
 alternative scenarios  is not strictly a technical decision, but the
 technical implications of selecting among these alternatives can be
 addressed by a technical specialist. Ultimately, the  selection of a
 land use scenario for the analysis and the selection of an "accep-
 table risk" are political/regulatory decisions, not technical ones.
 It will be shown later that these  decisions can affect the ultimate
soils criteria by factors of a thousand or more.

 Environmental Fate
  Parallel to and  simultaneous  with the review of toxic effects
data, the environmental physical chemical properties of the con-
                                              RCVIEW ENVIRONMENTAL
                                             BEHAVIOR OF COMPOUNDS:
                                              ENVIRONMENTAL SETTING
                                               IDENTIFY DOMINANT
                                              PATHWAYS OF MIGRATION
                                                  FROM SITE
                          DETERMINE QUANTITATIVE RELATIONSHIPS
                         BETWEEN SOIL CONCENTRATION AND HUMAN
                            EXPOSURE ALONG EACH PATHWAY
                       EXPOSURE VIA DRINKING WATER • EXPOSURE VIA
                       CONSUMPTION OF FISH • _< ACCEPTABLE DOSE
                            ACCEPTABLE SOIL CONCENTRATION
                           Figure 2
         Determination of Acceptable Soil Concentrations
                by Exposure and Risk Assessment
taminant are reviewed in the context of the environmental setting
of  the contamination/soils  to evaluate  the site-specific en-
vironmental behavior of the contaminants.
  Dominant pathways of migration from the site are identified by
screening level analyses and calculation. Screening level analysis
procedures are diverse, and flexibility must be maintained in their
application in order to derive the most information possible from
the available data,  while conserving "level of effort" resources.
Identification of the dominant migration pathways usually can be
deduced on the basis  of the judgment of senior environmental
scientists familiar with the site. Occasionally, formal screening
level calculations as represented, for example, by methods of En-
field, el at.' or Tucker, el a/.10,  can be applied at  this step.
  Integrating  the information  from the exposure scenario with
the dominant  migration pathways yields the dominant pathways
of exposure. These are the pathways, from soil to an exposed
member of the population at risk, that are capable of carrying
      CONTAMINATED SOIL TREATMENT

-------
and delivering a significant dose of contamination. An example of
an exposure pathway would be "surface run-off erodes surficial
soils and carries both dissolved and adsorbed soil contaminants to
a nearby estuary;  fish and shellfish  bioconcentrate  the  con-
taminants from the water,  and the organisms are harvested for
human consumption."
  Once the dominant exposure pathways  are identified, each is
expressed as a mathematical equation relating the dose (exposure)
via the pathway to the soil contaminant level. These equations are
representations of actual processes and are designed to express the
relationship conservatively  (inferring a higher-than-expected ex-
posure for a given soil concentraton) to favor protection of public
health. The most important pathways  should be evaluated with
the greatest  degree  of sophistication  to  make their  exposure
estimates more reliable.
  Finally, the exposures via all of the individual pathways are
added together, and the sum (total exposure) must be less than the
acceptable exposure. This step links the exposure assessment with
the toxic effects evaluation to produce the risk assessment-based
criterion. Each pathway is defined as a factor times  the soil con-
centration, and   the  acceptable  exposure  and  the  factors are
known. Consequently the equation may be inverted  to derive the
acceptable  soil concentration.

APPLICATIONS OF THE PROCEDURE
  In  most applications  of  this procedure  conducted by the
authors, the  following pathways of exposure  have required con-
sideration:
• Incidental  direct soil ingestion
• Dermal absorption
• Uptake by plants in the human food chain
• Leaching by infiltrating rainwater leading to contamination of
  shallow water table aquifer
• Groundwater or  surface  run-off discharge to surface  water
  bodies supporting a fishery
• Resuspension of contaminated  soil particles to the atmosphere
  by wind erosion or mechanical entrainment  associated with
  earth moving activities
  Consequently, the assumptions related to exposure via the pro-
cesses and  the uncertainty  associated  with procedures used to
quantify these processes, become critical issues in determining ac-
ceptable levels of residual soil contamination.

Importance of Land Use Scenario
  The land use issue can be investigated with reference to these
processes by  contrasting the assumptions involved for a residen-
tial land use scenario vs.  an alternative  land  use.  Consider  a
hypothetical example: rural land now used for lumber production
has been determined to  be contaminated. Assume it  has  been
estimated that the contaminants may persist in soils for 50 to 200
years. If the existing land use is maintained for that  period, then
the exposure  potential is very low. Workers involved in lumbering
will be exposed to soil contamination via dust inhalation, inciden-
tal  direct soil ingestion and dermal absorption. Small9 recom-
mended the use of 10 mg/m3 as a reference exposure level for this
situation. Since levels higher than 15 mg/m3 are not permitted in
the workplace, this is a useful conservative estimate. Actual levels
may  be  much lower. Depending on the importance of this
pathway relative to others, a more realistic estimate considering
level of activity, soils and climatic factors may be warranted. Ex-
posure to airborne dust also depends on the pulmonary ventila-
tion rate. Data are presented in Anderson, et al.' and Small.*
  Adults probably ingest very little soil. Soil ingestion occurs as a
result of contamination of food and food wrappers, dirt on hands
(especially  prior  to  eating) and  ingestion of inhaled particles
trapped in the upper respiratory tracts. Definitive data on typical
or extreme rates of soil ingestion among adults are not available.
Small' recommends the use of 100 mg/day, an estimate consistent
with Ford and Gurba" recommendations. Although there is very
little basis for this estimate, it is commonly used. General support
for the estimate relates to studies of juvenile soil ingestion that in-
dicate ingestion rates from 100 to 1,000 mg/day.2 It is generally
accepted that most adults ingest less soil than most children.
  A basis  for estimating dermal absorption  exposure  to  con-
taminated soils was presented recently by Layton, et al.6 Their
analysis indicates that it is conservative to assume that adults are
exposed to the contamination contained in 43 mg of soil per day
via this exposure pathway. Alternative approaches to estimating
dermal absorption  are available, but few address the details of ab-
sorption from contaminated soils, so the Layton, et al.6 method is
a useful basis for screening level analysis.
  The information presented above may be combined to estimate
the total dose experienced by the individual as follows:
          Incidental
          Ingestion
         Airborne
Inhalation   Dust
  Dermal
Absorption
 Dose = [0.1 g/day  + 17 mVday (0.01 g/m3 + 0.043 g/day]Csoil  (1)
Expressed in terms of the soil concentration:

  DoseO»g/day) = 0.31 g/day x Csoilf4g/g
                                    (2)
  To determine the acceptable soil concentration, ACS, we sim-
ply require that the actual dose be less than or equal to the accep-
table dose:
  Dose = 0.31 g/day x'ACS S Acceptable Dose
implying ACS  <  Acceptable Dose
                     0.31 g/day
                                    (3)

                                    (4)
  To  complete the hypothetical example, insert an  acceptable
dose for a 70 kg adult of 2,000 /ig/day. Then the acceptable soil
concentration is 6,450 /tg/g.
                           Figure 3
      Characteristic Elements of a Residential Land Use Scenario
                                                                                    CONTAMINATED SOIL TREATMENT    89

-------
  The recommended residual contamination level would have to
be much lower  if it were  anticipated that  the  land  would  be
developed residentially. Figure 3 illustrates some of the exposure
pathways  operating  under a  residential land use.  It  must  be
assumed that children may spend most of their time playing on
contaminated ground. Subsurface contamination can be exposed
during normal residential landscaping which also may  resuspend
dust into the air. Families may derive a significant fraction  of
their diet from a backyard garden. The residential water supply
may be derived from shallow individual wells, depending on local
hydrogeologic conditions and the availability of community water
supplies. These factors can  lead to substantially higher exposure,
especially for children.
  Assume that the population at risk is children ages 1 to 3 having
a typical body weight of 12.5 kg. Their acceptable daily dose for
the (same) contaminant would be 350 /ig/day. Based on the data
presented by Binder, el at..1 it seems  likely that the average child
ingests 100 to 200 mg of soil per day and a value of 150  mg/day
will  be  used in this  sample calculation. Kolbye, et  al.'  provide
estimates that a 2-year-old ingests 0.4 I/day of water and  125 g of
potatoes and other vegetables  per day  (20 g/day, dry weight);
while Layton, el a/." estimate that a  child  is exposed to the con-
tamination found in 41 mg  of soil per day, by the dermal absorp-
tion route. These exposure  factors will account for the major ex-
posure  pathways to children in a residential setting.  To relate
these exposure  factors back to the  soil concentration, it  is
necessary to establish the relationship between soil concentrations
and: (1) plant concentrations (plant bioconcentration factor) and
(2) groundwater/drinking water concentrations.
  The best method for estimating plant bioconcentration  factors
will depend on the combination of contaminant and plant(s) and
the data available for that combination.  Assume that Small's"
guidance has been used to indicate that plant concentration on a
dry weight basis will be approximately five times soil concentra-
tion.
  Groundwater/drinking water concentrations may or may not
be related to soil concentrations.  However, if the shallow water
table aquifer yields water of acceptable quality and quantity for
residential potable use, then it usually would be assumed to be the
source of drinking water. Further, and for the sake of simplicity
in this example, if the contaminated land is situated at  a ground-
water divide, then the quality of water in the wells will be similar
to the quality of the soil solution in equilibrium with the con-
taminated soils. Then the concentration in drinking water will be:
     mg/1
                  SOil
                                                         (5)
where: Kj is the adsorption coefficient. Assume that Kj =  125
for the contaminant in site soils. Combining this information for
the various pathways, the total dose for the 1-3 year old children
comprising the population at risk can be calculated in proportion
to the soil concentration.

             Soil          Drinking         Garden         Dermal
           ingestion         water         vegetables       absorption
Dosc=   CSO,| (0.15 g/day  + 0.4 I/day li'** + 2(1 g/day • 5  • 0.041 g day)  (6)
                                kd
Expressed in terms of the soil concentration:

  Dose O^g/day) =  103 Csoi, (/tg/g)

97% of the dose (100/103) is calculated to come from the garden
vegetable pathway.
  To calculate an acceptable soil concentration (ACS), the  dose
must be less than the acceptable dose.

  Dose  = 105 g/day Csoi, < 350 ^g/day
  ACS<3.4Mg/g
  Comparing this value with the ACS for the lumber production
land uses scenario (ACS S 6,450/ig/g), it is seen that the land use
scenario assumption can affect the recommended soil criterion by
more than three orders of magnitude. This degree of sensitivity to
assumptions is apparently at least as important as the uncertainty
regarding  toxic effects and quantitative carcinogenic risk assess-
ment procedures.

Refinement of Exposure  Factors/
Reduction of Uncertainties
  The results of any predictive method, such as the one presented
here, always  will be uncertain.  In multimedia exposure  and risk
assessment,  the number of variables  and  uncertainty  in each
results in large uncertainties. A  traditional response to uncertain-
ties in environmental  assessment  has been  a reliance on conser-
vative estimates. The true situation is unknown, but estimates and
decisions arc  based  on extremely adverse assumptions.
  The simple predictive methods presented in the examples above
can be refined by acquisition of more data or application of more
sophisticated predictive methods. As the estimates become more
reliable, the degree  of conservatism can  be reduced. This process
usually  will result  in   a  less stringent,  yet  more  firmly based,
recommendation.
  As an example, consider the basis for  estimating inhalation ex-
posure for the lumber production scenario.  It was assumed that
air  concentration of  dust that had  been stirred  up from con-
taminated areas could be as high as 10 mg/m'. The basis for that
estimate was conservative:  if  levels were higher,  the  workers
would wear respirators. This estimate could be  refined either by
sophisticated wind erosion/fugitive dust modeling or by measur-
ing air  concentrations under   the conditions of  the exposure
scenario.  Assuming that  fixed air  monitors were placed in a
similar,  though  uncontaminated  area,  and  that  personnel
monitors also were employed in a 1-month study of worker ex-
posure to nuisance dust, a more realistic dust concentration of 0.4
mg/m-1  was  determined. Applying  this  finding, the previous
analysis can be  refined. The  dose  is now estimated to be:
             Soil
           ingestton        inhalation
Dose -  CSO,| (0 I g da>  *   17 m.Y das •
     = 0.15 mg da> C'ioil I eg g)
  Airborne        Dermat
    dim         absorption
0 IXKM g m» *  0.043 mg day) (7)
and the acceptable soil concentration would be:
                                                                    ACS=
              0.15
                                                                                             =  13.400
  Such refinements almost invariably should result in a relaxation
in the recommended criterion, as long as uncertainties always are
resolved with a conservative assumption.


CONCLUSIONS
  A method for determining safe  levels  of residual soils con-
tamination as a component of the endangerment assessment pro-
cess for hazardous waste sites has been presented.  The method
presented  is based closely on the PPLV approach developed by
Dr.  David H.  Rosenblatt  and  co-workers at the U.S.  Army
Medical Bioengineering Research and Development Laboratory,
Ft.  Detrick, MD."* The method has been  restated here, using
slightly different terminology, to illustrate its relationship to the
endangerment assessment process.
  Development of a hypothetical example illustrates the sensitiv-
ity  of  the  recommended criteria to  the exposure  assumptions
adopted, particularly those relating  to the future land use of the
site. The land  use  scenario  adopted for the analysis affected the
recommended  acceptable soils concentration by more than three
90
      CONTAMINATED SOIL TREATMENT

-------
orders of magnitude in the example problem. It is possible to con-
struct alternative scenarios that affect the criteria by even greater
amounts. Thus,  uncertainty  regarding  future land  use  and the
assumptions adopted in the face of that uncertainty can influence
the endangerment assessment, the evaluation of the no-action
alternative and the costs of  remedial action just  as severely as
uncertainties regarding the "toxic effects/quantitative risk assess-
ment" aspects of the problem. The exposure  assumptions are
readily understood and thus readily debated by all concerned par-
ties, and so they can and should receive a thorough airing during
the RI/FS. Decisions regarding appropriate future land  uses for
the site are quintessentially political, rather than technical deci-
sions.  Affected parties should be consulted, and these decisions
should be made by regulatory authorities and/or affected parties.
   The second major  conclusion  of the paper is based on  an
analysis in which one of the criteria was refined by acquisition of
additional data.  This allowed a more realistic  analysis  and the
discarding  of an excessively  conservative assumption.  More
sophisticated theoretical analysis frequently can achieve a similar
result. To develop criteria in the face of great uncertainties, it fre-
quently is prudent to adopt  conservative assumptions.  This af-
fords the decision-maker the certainty that the criterion level is
safe and that concentrations  below the criterion do not pose an
unacceptable risk. The converse, however, is untrue: levels above
the derived criteria are not necessarily unsafe. The refinement of
estimates presented bears this out: with a conservative  assump-
tion, applied in lieu of site-specific data, it was determined that
soils having 6,450 jtg/g of contaminant "X" were "safe"  and
presented no endangerment. Upon acquisition of additional data,
it became apparent that soils  as high as  13,400 /ig/g can be left in
place. This does  not contradict the previous analysis.  Before ac-
quisition of the new data, the analysis indicated that 6,450 jtg/g
was a safe level. The  new  analysis confirms  this conclusion.
Generally, the public and their representatives interpret results of
this type to mean that soil contamination  at 6,500 /ig/g, or any
level above 6,450 /ug/g, is harmful. Drawing this interpretation
from the conservative analysis would be inappropriate and should
be avoided.
REFERENCES

 1.  Anderson, E., Browne,  N., Duletsky, S. and Warn, T., "Develop-
    ment of Statistical Distributions or Ranges of Standard Factors"
    used in Exposure Assessments, U.S. EPA Contract No. 68-02-3510,
    1984.
 2.  Binder, S., Sokal, D. and Maughan, D., "Estimating the Amount of
    Soil Ingested by Young Children Through  Tracer Elements,"  Un-
    published Manuscript. Div. of Env. Hazards and Health Effects,
    Center for Env. Health, Centers for Disease Control Public Health
    Service, U.S. Dept. of Health and Human Services, Atlanta, GA.
 3.  Enfield,  C.G., Carsel, R.F.,  Cohen, S.Z.,  Phan, T. and Walters,
    P.M., "Approximating Pollutant Transport  to  Ground Water,"
    Ground Water, 20, 1982, 711-22.
 4.  Ford, K.L.  and Gurba,  P., "Health Risk Assessments for Contam-
    inated Soils," Proc. of the Fifth National Conference on Manage-
    ment of Uncontrolled Hazardous  Waste Sites,  Washington,  DC,
    Nov. 1984,  230-731.
 5.  Kolbye, A.C., Mahaffey,  K.R., Fiorino, J.A., Corneluissen, P.C.
    and Jelinek,  C.F., "Food Exposures to Lead," Environ.  Health
    Perspectives,  May 1974, 65-74.
 6.  Layton,  D.W., Hall,  C.H.,  McKone,  T.E., Nelson, M.A.  and
    Ricker, Y.E., "Demilitarization of Conventional  Ordnance: Priori-
    ties for Data Base  Assessments of Environmental Contaminants,"
    UCRL-53620 Draft prepared  for USAMBRDL,  Ft.  Detrick, MD,
    1985.
 7.  Rosenblatt,  D.H. and Dacre, J.C., "Preliminary Pollutant Limit
    Values  for  Human Health Effects," Environ. Sci.  Techno/.,  14,
    1980, 778-84.
 8.  Rosenblatt,  D.H.,  Dacre, J.C. and Cogley, D.R., "An Environ-
    mental Fate Model Leading to Preliminary Pollutant Limit Values
    for Human Health Effects" in R.A. Conway  (ed.), Environmental
    Risk Analysis for Chemicals,  Van Nostrand Reinhold, New York,
    NY, 1982, 475-505.
 9.  Small, M.J., "The Preliminary Pollutant Limit  Value Approach:
    Procedures and Data Base," U.S. Army Medical Bioengineering Re-
    search and Development Laboratory, Ft. Detrick, MD, Tech.  Re-
    port 8210, 1984.
10.  Tucker, W.A., Dose, E.V., Gensheimer, G.J.,  Hall, R.E., Pollman,
    C.D. and Powell,  D.H.,  "Evaluation of Critical Parameters  Af-
    fecting Contaminant  Migration through  Soils,"  Report AMXTH-
    TE-TR-85030, prepared by Environmental Science and Engineering,
    Inc. for U.S. Army Toxic and Hazardous Materials Agency, 1985.
                                                                                         CONTAMINATED SOIL TREATMENT    91

-------
                  Objective Quantification of  Sampling Adequacy
                           and  Soil  Contaminant Levels  Around
                               Point Sources Using  Geostatistics
                                                  Jeffrey C. Myers
                                          Geostat Systems  International
                                                 Golden,  Colorado
ABSTRACT
  This paper describes the events and conclusions resulting from
a study focused on two lead smelters within the Dallas city limits.
These smelters were implicated  as the sources of elevated con-
centrations of lead in the surrounding soils. The presence of high
lead is of considerable concern to nearby inhabitants.
  This report focuses on  the application of geostatistical tech-
niques to sampling around point sources. The application of geo-
statistical methods to smelter contamination had not been prev-
iously attempted. Many  of the questions concerning applicability
of geostatistics,  geometric and spatial  structure of the contam-
inant plume, modeling of the area, sampling strategies and gen-
eralization to future sites have been resolved. Geostatistical tech-
niques have proven to be very useful in the resolution of these
questions.

INTRODUCTION
  Increased environmental awareness dictates that virtually any
type of industrial "smokestack," either  existing or planned, is
subject  to public scrutiny. Coal-burning  power plants and steel
mills are being blamed for acid rain problems. Other heavy metals
such as uranium that are emitted from coal-powered plants also
are causing concern. These emissions  pose a threat  to nearby
plant and animal life. Insights gained from modeling heavy metal
airborne contamination  hopefully can  improve models in these
other situations or provide a starting point  for understanding
these unique problems.
  The results and conclusions of  the Dallas Lead Project have
provided analytical tools and valuable insights into the nature of
contamination at other  locations.  This information then can be
applied to future modeling projects.
  This paper discusses the techniques applied to the modeling of
the Dallas project  in detail as well as the insights that have been
gained by  the authors. The U.S. EPA commissioned this study
as an experiment to determine applicability of geostatistical tech-
niques in an attempt to  develop more objective procedures. The
following conclusions are reached in the paper:

• Geostatistical  methods  are very objective  and effective  in
  modeling  contamination around  smelters.  Two separate
  smelters in the Dallas  area have been studied. Both sites have
  produced excellent  results and  nearly  identical geostatistical
  models indicating that other sites may follow a similar pattern.
• Geostatistical  modeling has many advantages over  traditional
  estimation techniques.  Its Best Linear  Unbiased  Estimator
  (BLUE) property provides more objective answers to questions
  that are difficult to provide otherwise. Results of studies such
  as this one influence  decisions  in cleanup  phases that  affect
  human health and safety and should be made based  on the
  best possible data. The results also will have significant legal
  ramifications and are subject to close scrutiny in courts of law.
  Thus the need for objectivity and reproducibility becomes in-
  creasingly important as issues enter the courtroom.
• Modeling with regionalized variable techniques permits the
  earth scientist to isolate different phenomena acting simul-
  taneously within the  same geographic area. Different patterns
  of dispersion from the smelter can be recognized and isolated
  from other  sources.  Trivial contributions from  secondary
  sources (e.g., auto emissions) can be identified. In this way
  liability can be correctly assessed.
• Objective methods to determine sampling sufficiency easily can
  be defined. Sufficient grid sizes to model and assess  the area
  with  an  acceptable level of  confidence  can be obtained with
  minimum sampling.  If too many samples already  have been
  taken, sampling can  be reduced to determine sufficient levels
  for obtaining both variograms and desired accuracy as a start-
  ing point for future projects. The technique known as the com-
  position of extension variances has been applied and discussed
  with respect to overall precision.
  The results can  be applied to the design of future smelters;
they also provide a method of limiting liability in environmental
impact  situations by isolating sources; finally, the results can be
used by the environmental and legal professions to further the
cause of public health and safety within a framework of cost con-
sciousness and objectiveness.

BACKGROUND: THE DALLAS LEAD
PROJECT
  During the last  quarter of  1980,  a  sampling  team from the
University  of Texas at  Arlington,  under contract with the U.S.
EPA's regional office in Dallas, Texas, collected samples of soil,
house dust, paint and tap water from a number of private resi-
dences  and schools.  These sample  locations were situated near
and in the general vicinity of two lead smelters and one battery
plant in Dallas,  Texas. In addition to the two smelter sites and
the battery plant, a fourth site identified  as the Reference area
(REF) also was selected as a control  area for sampling. The pur-
pose of this initial phase of the investigation was  to determine
whether additional effort was warranted, i.e.,  did  a  problem
exist?
  All four sampling sites were located within the Dallas city limits
and all except  the REF site were geographically located  in the
city's western section. Figure 1  shows the location of the two lead
smelters: RSR Corporation (RSR) and the  Dixie Metal Company
(DMC). General Battery Corporation (GBC) is located  near the
intersection of Bengal and Amelia Streets.  The GBC site was not
analyzed in the geostatistical portion of the study, however. The
92
     CONTAMINATED SOIL TREATMENT

-------
                          Figure 1
     Location of the DMC & RSR Smelters and the Reference Area

REF site was selected for its similarity to the RSR,  DMC and
GBC sites in automotive emissions, population and ethnic back-
ground.

SAMPLING
  Data collected by the University of Texas at Arlington have
shown some soils in the  immediate vicinity of the Dixie Metal
Company (DMC) and the RSR Corporation (RSR) to have lead
levels that exceed 1,000 ppm, believed to be the critical danger
level for exposure. Due to the presumed behavior of airborne
lead from smelting facilities such as these, the sampling area was
confined to  a 2-mile radius around each site. To better evaluate
the degree of contamination of these two industrially  exposed
areas, a site known as the Reference area (REF) was selected.
  The sampling scheme, except for the total number of samples
collected in  the reference area, was identical to that used at the
DMC and RSR sites.  In  addition, a 2-mile area surrounding a
third site, the General Battery Corporation (GBC), was sampled
using the same scheme.

SAMPLING AND GRID DESIGN
  The information collected in sampling phase  1 (20 data points)
was used to construct a variogram for each of the three study
areas.  The  calculated and recommended  sampling scheme  to
identify the geographical extent and level of soil lead contamina-
tion for phase 2 of the project was as follows:
• The dimensions of the square grid should be 750 ft x 750 ft.
• The  total  area covered by the grid should be contingent upon
  the minimum lead concentrations of concern.
• The grid should be oriented along the axis of the plume.
• The grid used in the REF area should be identical to that used
  in the DMC and RSR areas.
• Soils within a radius of 100 ft of each grid intersection should
  be sampled using procedures identical to  those described in
  phase 1.
• Six samples should be collected for each site to be composited
  into a single sample.
• Duplicate samples should be collected at 5% of the grid loca-
  tions.
• Samples should be separated into three  "splits" at 5% of the
  grid locations and sent to different laboratories.

  During the summer of 1980, the U.S. EPA collected more than
3,000 individual core  samples that were composited into more
than 500 grid samples throughout the three study areas. These
data are the foundation  of the subsequent geostatistical analysis
and mapping portion of the project.

STRUCTURAL ANALYSIS
  Variograms of the  Ln-transformed soil sample  values were
computed in each of the three areas. In each case, an average
variogram and four directional variograms were produced. The
four directional variograms have  been calculated along the east-
west  (0°),  north-south  (90°), northeast-southwest (45°)  and
northwest-southeast (-45°) directions. For each directional vario-
gram, an angular tolerance window of 22.5 ° (total width of win-
dow) around the given direction was allowed. A distance lag sep-
aration of 800 ft also was used. For the average variogram, which
considers all points, the angular tolerance was 180°.
  The results of the calculations were very good. Salient features
of the variograms include the following:
• All the variograms show  some kind of  continuity between
  holes. They all tend to increase with distance until they reach
  a sill value.
• Continuity in the DMC and RSR areas shows very similar pat-
  terns when compared to the reference  area. Modeled range
  and sill values bear this out.
• As expected, the REF area shows a much lower sill and a lesser
  degree of continuity.
    VARIABLE LEAD
    DM-LAS LEAD	DHC
                            LOGARITHMIC
                         Figure 2
    DMC Variograms of Ln-Transformed Lead Values for Phase 2
                                                                                   CONTAMINATED SOIL TREATMENT   93

-------
                                                    t i-coo  Q I s i
                            LOGARITHnIC
                          Figure 3
     RSR Variograms of Ln-Transformed Lead Values for Phase 2

   Isotropic spherical models have been fitted to the experimental
average variogram curves. In all cases, they are the sum of a small
nugget effect (calculated from the duplicate samples) and  one or
two spherical structures of varying ranges. The curves of the fitted
models are shown as a solid line in Figures 2 to 4.
       ^ LCAO -   K' I
  Finally, one anticipated problem in the calculation of the vario-
grams did not arise. This "non-problem" was the possible pres-
ence of a drift or a  trend. A drift is defined in geostatistical
terms to be a systematic increase or decrease in  the data along
given directions.  This drift causes  problems in the variogram.
Modeling such a structure is difficult and also causes problems in
the kriging process. Special  forms of kriging (universal,  general-
ized IRF) have been created just to handle this problem. Although
the possibility of a drift seemed reasonable, one did not arise.

MODELING THE AREAS: KRIGING
LEAD VALUES
  The  soil  sample data were kriged in  each  of the  three areas
using their respective variograms. The kriged area  has been made
large enough to include three  reference  locations of geographic
interest. These locations have been used for map orientation.
  The contour maps resulting from  the kriged grid are shown in
Figures 5 to 7. The round symbol in the center represents  the lead
smelter. The contour lines show the estimated soil lead concentra-
tion in ppm.  In the DMC area (Fig. 5),  a highly visible  concen-
tration or plume has developed around and to the  north and east
of the source. The large number of concentric contours encircling
the  smelter shows a steep gradient of rapid change in a short dis-
tance between a low of 200 ppm outside and a high of 3,000 in-
side.
  The effects  of  flooding by the  Trinity River are well  defined
(to  the north) as a distinct area of low lead concentration which
can be seen along the flood  plain of the river. On either side of
the  river, the concentrations are noticeably higher. Several pos-
sible anomalous sample values also occur to the north of the river.
In at least one case a sample whose concentration was 10,400 ppm
contamination was found later to  have been exposed to automo-
tive contamination.
  In the RSR area, the smelter is located within the major con-
taminant plume that disperses rapidly east-west and more slowly
north-south (Fig.  6). The  southern end of the plume also contains
                                                                    SCO	2700	SJOO	7900	»»OO
                                                                I23OO
                                                                                                                         12300
                                                                 T80O-
                                                                 SKX>-
                                                                 2700-
                          Figure 4
    REF Variograms of Ln-Transformed Lead Values for Phase 2
                                                                    »002700^00        TioO       9900
                          Figure 5
         Isomap of Estimated Lead Values in the DMC Area
94    CONTAMINATED SOIL TREATMENT

-------
                29OO       S3OO      7TOO
                                                                            1500   3OOO  45OO   60OO   75OO   90OO  IOSOO
                                                                                                                        H20OO
                                            IOIOO
                            Figure 6
          Isomap of Estimated Lead Values in the RSR Area

a possibly anomalous value that has the effect of extending the
plume in this direction. This finding suggests that the true dis-
persion nature of the plume is primarily north and slightly west
from the smelter.
  As expected, the REF area shows no systematic areas of lead
concentration (Fig. 7). Local highs and lows  fluctuate randomly
and at considerably reduced levels overall from those  seen in the
DMC and RSR sites. There is no evidence to suggest  that this is
anything other than a background lead level map.
  Accompanying each isopleth map is a second contour surface
that  contains isovalues for the kriging standard deviation (Figs.
8 to  10). The values shown are the logarithmic kriging deviation
multiplied by a factor of ten (to facilitate plotting). The "swiss
cheese" effect noted here is common for cases such as this where
the sample grid spacing is a relatively large percentage of the
range of the variogram. Because the quality of estimation is solely
a function of geometry and the variogram, estimates are better as
they approach a sample point.
37OO       6IOO       85OO       \O9OO
            1300        3700       6100       6500       I09OO
                           Figure 7
         Isomap of Estimated Lead Values in the REF Area
                                                                    I05OO-
                                                                   9000 -
                                                                    6000
                                                                   3OOO -
                                                                            I50O    3OOO   450O   6OOO   75OO    9OOO  K55OO
                                                                        Figure 8
                                                      Isomap of the Estimation Variance in the DMC Area
                                                                                     IOIOO     "TOO
                                                                                                                        RD.a
                                                                                                                        THE MKTRft TRACK
                                               IOIOO
                                                                                                  -6900
                                                370O
                                                                       500
                                                                                                           IOIOO     II7OO
                                                                                              Figure 9
                                                                            Isomap of the Estimation Variance in the RSR Area
                                                                           /?      °
                                                                           y * iLLiNoia AVC a
                                                                                     2IOO
                                                                                                          6500     IO9OO
                                                                                 •ucxtrr AVE
                                                                                                                      -8500
                                                                                                                      JNNTV»LE IT.
                                                                                                                      -5300
                                                                                                                      "2IOO
                                                                                                                       .-IIOO
                                                                                                           85OO      IO9OO
                                                                       Figure 10
                                                      Isomap of the Estimation Variance in the REF Area
                                                                                       CONTAMINATED SOIL TREATMENT    95

-------
SAMPLING SENSITIVITY: REDUCED
SAMPLE SIZES
  The results of this first mapping stage proved that kriging and
geostatistical  techniques apply very well to soil data  and point
source contamination. The second stage of this study addressed
the concerns of the U.S.  EPA and others as to the adequacy of
sampling: have enough samples been taken to assure reasonable
level of certainty?  Also, what changes might be made in future
studies of similar phenomena? These questions will  not be ad-
dressed.
  The techniques used to answer these questions will be the analy-
sis of extension variance and the use of reduced data sets to de-
termine the effects on the variogram. The results of the prelim-
inary geostatistical analysis provided models of the spatial con-
tinuity structure in each area through their respective variogram
models. These  models will  now be  assumed  to be the "true"
underlying spatial structure. The data sets of 206 and 208 samples
in the DMC and RSR sites respectively will be taken as the "uni-
verse" of data.
   The procedure, then, is to reduce the level of sampling  in the
areas of study and to  recompute the variograms using a smaller
data set. Multiple data sets will be produced and analyzed at each
level of reduced sampling to minimize any  possible selection bias.
The new parameters of the variograms then will be compared to
the "true" values to see how accurately the subset can predict the
"real" structure.  This information from  small data sets can  be
used to design future two-phase sampling schemes.
   The two smelter areas have been divided into equal sized grid
cells from which one sample per cell has been taken at random.
This random selection has been  done by  computer to eliminate
human  bias.  Since the "true" variograms are isotropic (equal
zone of influence in all directions), the sampling grids were de-
signed to be  isotropic (square) so that  this relationship can  be
honored.
   From the variograms,  it is  possible to  quantify the  variances
of polygonal (block) extension errors and then combine  them,
since they are the variances of independent variables. This has
been used as a check on total precision.
   Small sample subsets were selected in  both the DMC and RSR
areas on grids of 11 x  11, 7 x 7, 5 x 5 and 4x4. These represent
sample  sizes of approximately 100, 50, 25 and 16 respectively.
The 11x11  grid was selected as the first  grid  size with the idea
that subsequent sizes would be enlarged or reduced as needed de-
pending upon the outcome of this first attempt. Ten sample sub-
sets were taken using the same ten random seeds in both smelter
areas. The ten new data sets in each area then were variogram-
med, their sample locations  were plotted  and their  histograms
were plotted.
  The 20 variograms then were modeled.  The  results showed re-
markable similarity to the original data  sets both visually and in
variogram parameter comparison. The results of the 11 x 11 grid
data were so successful that the data set was again halved by us-
ing a 7  x 7 grid.  Excellent variograms were produced although
more variation was noticed in the model  parameters. The RSR
models  tended  to  have a lower  sill  than  the  true one, but the
ranges were generally the same as for the  true  model. The  DMC
models showed on the average the same  sill but generally showed
a shorter range. Because the  object of this exercise  was to de-
termine sampling grid sizes from small initial data sets, the range
was the  most  important parameter. For  this grid size, both areas
provided accurate  ranges with the DMC  models  being a little
more conservative.
  Again the data set was  halved by going  to a  5 x 5 grid. At this
level of sampling, the directional variograms had deteriorated to
such a point  that  they were valueless.  A  lack of data was the
cause.  However,  the average variogram  still could be used to
advantage  as  good structures  still were found.  Sill  variance
showed a much greater fluctuation, with a number of low-valued
variances where sampling failed to pick up high values near the
smelters. Ranges generally were shorter than the  "true" ranges.
This corresponded to the shorter range found in phase 1  of the
sampling where only 20 samples were taken. A few very  long
range values did appear, but they resulted mainly from an attempt
to fit a good model at short distances where the variance  rose
quickly for a time and then slowly crept toward the sill.
  The  final level of sampling was produced on a 4 x 4 grid. Again
directional  variograms  were of no value. The  sill values  in the
RSR area showed a large fluctuation,  but the ranges were quite
stable and not too far from the  "true" parameters. In the DMC
area, sills were generally near or above the "true" parameters
and the ranges were at or below  those expected. Overall, this grid
produced about the same quality estimates as the 5x5 grid and
showed slightly better stability of ranges.
                                 DMC AREA
             "sooo
             2
             t-
             m
             til


               000
                   0    497    H   14 19
                      NUMBER  OF GRID CELLS

                          Figure 11
       Global Precision vs. Sampling Density in the DMC Area

                               RSR AREA
             I
               2000.
                         497     II    14 19
                       NUMBER OFGRIDCELLS

                          Figure 12
        Global Precision vs. Sampling Density in the RSR Area
96    CONTAMINATED SOIL TREATMENT

-------
GLOBAL PRECISION
  An estimate of global precision provides a measure of the cer-
tainty with which the global mean is known. This value is anal-
ogous to the "standard error of the mean" in classical statistics.
In the geostatistical sense, obtaining a high value for precision of
the global mean suggests that local estimates also will remain
quite accurate. This condition is not strictly true in all local cases
but offers a starting point for decision making.
  The precision for symmetric grids of sides 4 through 15 have
been calculated and have been plotted in Figures 11  and 12. The
14 x 14 and 15 x 15  intervals represent the precision when con-
sidering a full sample set of 206 or 208. The graphs show that in
terms of global estimates the sampling has reached  the point of
diminishing returns. Both figures indicate that further sampling is
unnecessary.  What also is critical here is  that variograms have
been produced for even the smallest of data sets.
CONCLUSIONS
  Geostatistical techniques have proven to be an extremely valu-
able tool in the analysis of pollutant contamination. A strongly
identifiable plume of lead is associated with smelter source. No
such structure is found in the REF area. The plume is marked by
a central core of concentrated high values, generally within 1 mile
of the source, with a continual decrease in lead values to the
borders of the areas. It appears certain that these two smelters
were indeed the source of the contamination  and that automo-
bile emissions are of negligible contribution.
  The smelter areas have been adequately sampled with the 800-ft
grid. The  range of the variograms was  five to seven times this
sampling interval and proved quite sufficient to model the spatial
continuity of the lead values.
  Global estimates of precision for the second stage sampling
campaign  are very conservative and  indicate  that additional
sampling would have been of little value. Variogram parameters
remain surprisingly constant down to extremely small levels of
sampling.
  Two-stage sampling, though already in regular use in the en-
vironmental industry, can be greatly enhanced by the use of geo-
statistical technique. Optimal sampling geometries can be  devel-
oped for a variety of situations and constraints. As a result, cost
benefit  curves can be produced to select development drilling
grids. These analyses will provide a measure of confidence and
objectivity.
                                                                                     CONTAMINATED SOIL TREATMENT    97

-------
                 Innovative Application of  Chemical  Engineering
                   Technologies  for Hazardous  Waste Treatment
                                                   Robert  D. Allan
                                                  Michael  L. Foster
                                                   IT Corporation
                                                Knoxville, Tennessee
ABSTRACT
  Regulatory and public pressure are creating an increasing de-
mand for more innovative methods of hazardous waste manage-
ment. In response to this need, some proven technologies are
being adapted to provide alternatives to the more classical remed-
iation techniques in certain applications. This paper reviews the
site-specific parameters that need to be considered before select-
ing one of these alternatives. The operational details of the can-
didate technologies are discussed, and specific examples of appli-
cations are given.

INTRODUCTION
  Application of the RCRA and CERCLA statutes often creates
the need for corrective action, particularly with regard to ground-
water treatment and surface impoundment closure. Of particular
interest  to industries are the on-site treatment alternatives due to
the cost and liability factors of land disposal and the lack  of con-
venient incineration options.
  This paper concentrates on proven treatment  solutions avail-
able to meet the needs of corrective action situations.  Particular
emphasis is placed on those technologies offering the opportun-
ity  for cost-effective,  on-site treatment or waste volume reduc-
tion. Included is a review of  technologies  utilized in  mobile
and/or in situ applications.

PROBLEM DEFINITION
  Unfortunately, no single treatment technology is going to be
universally  applicable  for  all corrective action  situations. A
thorough evaluation of the site-specific  parameters involved is
necessary in order to choose the most technically suitable  and
cost-effective treatment technology for the job.  Because major
corrective  action  programs, especially  those involving  aquifer
restorations, are  likely to take several years to accomplish, the
investment  of time and effort to properly evaluate the situation
initially  will more than pay for itself in the long run.
  Among the major site-specific parameters that should be de-
fined are those discussed below.

Contaminant Composition and
Concentration
  An accurate analysis of the site contamination is  important.
The contamination could be a single component from an under-
ground solvent storage tank leak or a wide variety of compon-
ents from  a disposal site leachate. The  level of contamination
could range from  low >jg/l to several hundred mg/1. The treat-
ment costs of technologies such as carbon adsorption and oxida-
tion are dependent on concentration  and type  of compound,
while costs for technologies such as air stripping are more closely
related to the treatment rate and desired removal efficiency.
Desired Cleanup Criteria and
Effluent Disposition
  Stringent cleanup criteria (e.g., drinking water standards for
aqueous streams) will eliminate some technologies from consid-
eration immediately,  but  it must be recognized that different
levels of cleanup are likely to be required for many remedial pro-
grams. The ultimate disposition of the effluent for aqueous treat-
ment will have significant  bearing on the required effluent qual-
ity. If the treated  stream can be discharged to the sewer for
further treatment by a municipal waste facility (POTW), effluent
guidelines are likely to be less strict than if the stream is to be re-
charged  to an aquifer or discharged directly to  a surface water
source under an NPDES permit.

Volume to be Treated
  Treatment volumes for remedial programs can vary widely. A
small volume may be treated best by a technology with low capital
costs and high operating costs, while economic remediation of a
larger magnitude is more likely  to be accomplished utilizing  a
technology with low operating  costs. An evaluation of treatment
options  should  include consideration of leasing equipment to
treat small volume sources.

Fate of Treatment Byproducts
  Local  requirements for disposition of treatment byproducts
will have a major impact on the economics and viability  of vari-
ous treatment technologies. Air stripping  can be an extremely
cost-effective technology if the organic-laden air can be  emitted
directly to the environment; but strict air emissions standards may
require that air stripping be followed by a vapor-phase  adsorp-
tion system, significantly increasing treatment costs. Similar con-
siderations  exist  with  the generation of  potentially toxic by-
products from chemical oxidation processes and the possible need
for nutrient removal (phosphorus and nitrogen) from the  effluent
of biological  oxidation processes. The economics of activated
carbon adsorption  (especially on a throwaway carbon basis)
can be affected by  the accessibility of an approved disposal site
for the contaminated carbon.

Utilities Availability and Cost
  The selection of certain technologies will be influenced by the
availability of utilities such as  steam at the treatment site. Like-
wise, some processes such as UV-catalyzed  oxidation can require
a substantial source  of  electrical power.  The  energy-intensive
processes are likely to be more attractive in the southern part of
the country where electrical costs are approximately $0.05/kWh
than on the West Coast where costs can approach $0.15/kWh.
  A clear definition of the goals of a remedial action program
along with an understanding of the above site-specific parameters
98
      ON-SITE TREATMENT

-------
and the technical limitations of the technologies under considera-
tion  will allow  selection  of the most  cost-effective treatment
option.

TECHNICAL EVALUATION
  A number of technologies are potentially applicable to remed-
iation  programs. Some are best suited  for specific applications
and others may be eliminated from widespread consideration be-
cause of the inability to meet stringent discharge standards or
other technical limitations. In this section, the technologies with
the greatest potential in remedial actions are identified and spe-
cific applications are discussed.

Incineration
  Incineration, or thermal destruction, is an effective way to de-
stroy organic liquids and sludges and organics in a solid matrix.
It can  be used for a relatively broad waste profile depending on
the type of incineration system utilized. Incineration is defined
generally as a controlled, high-temperature, oxidation reaction.
The  end products of the incineration reaction are generally car-
bon  dioxide and water from hydrocarbon wastes; but industrial
and hazardous wastes also can generate hydrogen chloride, sulfur
dioxide and less desirable combustion byproducts depending on
the nature of the wastes, the operating features of the specific in-
cineration system and the compatibility of the waste profile  and
incineration system.
  The technology of high-temperature incineration is in  wide-
spread use, both at private and commercial installations. It often
is considered in  remedial  action alternative evaluations, but the
waste volume must be substantial to justify the expense of permit-
ting and building an on-site unit. For more limited volumes, com-
mercial incineration  facilities offer the advantage of destruction
of the wastes involved.

Thermal Desorption
  Thermal desorption removes organic  chemicals from soil  and
solids by heating the soils to temperatures sufficient to convert the
organics to vapors and then holding the soil at that elevated temp-
erature long enough for the removal process to occur. Desorbed
organic and water vapors  can be recovered by solvent scrubbing
or condensed and collected in an accumulator where the  water
and  organics are separated into their respective phases. Water
can be treated with activated carbon before discharge,  while the
organics phase can be sent to a commercial incinerator or treated
chemically or photolytically.
  Thermal desorption is effective for any organic  contaminant,
with more volatile compounds such as solvents, being more read-
ily removed. The types of compounds for which this technology
is expected to be applicable range from chlorinated  solvents  and
gasoline/fuel components  to  phenols,  aromatic hydrocarbons,
PCBs  and dioxin. The  concentration of the contaminant does
not have a  significant influence on  the applicability. Although
the soil characteristics will impact the operating rate and process
design requirements, thermal  desorption is amenable to a wide
range of soil/site conditions.
  IT  has  developed and demonstrated thermal desorption in
bench  scale and pilot scale equipment. Laboratory testing  has
been performed  on a variety of soil types containing such chem-
icals as Agent Orange (containing chlorinated dioxins), penta-
chlorophenol and creosote materials (PNAs).  Pilot scale work,
conducted  as part of the  Air Force's  Environmental Restora-
tion Program, will be reported at an upcoming American Chem-
ical Society meeting.1 Briefly, a pilot plant was assembled to treat
up to 100 Ib/hr  of soil contaminated with 2,3,7,8 TCDD from
Agent  Orange. The desorber was operated at 460 to 560 °C. The
vaporized contaminants were captured by a solvent scrubber and
destroyed in a UV photolysis reactor.  Soil analysis  showed  a
greater than 99% reduction in the TCDD concentration and resid-
ual soil concentrations of less than 1 ppb.
  The thermal desorption process offers several advantages,  as
compared to thermal  oxidation,  for decontamination of organ-
ically contaminated soil. These advantages include lower temper-
ature operation resulting in less energy consumption and simpler
design criteria, relative ease of regulatory approval due to classif-
ication as a physical/chemical treatment process and the ability
to easily skid or truck mount the equipment  for mobile treat-
ment.

UV Photolysis
  Ultraviolet light can catalyze normal chemical oxidations by
creating a supply of "free radical" species. IT has applied this
technology to the destruction of dioxin at the Syntex Agribusi-
ness plant in Springfield, Missouri.2 The dioxin destroyed was in
a distillation residue which was a byproduct  of previous pro-
duction of  hexachlorophene by the Northeast Pharmaceutical
Company. The process involved extraction of dioxin from the
residue  using a common  solvent and  then destruction  of the
dioxin in reactors equipped with  10-kW,  high  intensity  ultra-
violet lamps. Approximately 13 Ib of dioxin were destroyed using
this process.

Chemical/Biological Oxidation
  Oxidation technologies  use  chemical  or biological  means to
completely oxidize organic contaminants to carbon dioxide and
water, or partially oxidize them to non-toxic intermediates. Most
of the oxidation options are best suited to specific applications.
  Chemical  oxidation  accomplishes  detoxification  of  waste
streams through the addition of chemical oxidation agents. Near-
ly all chemical oxidation processes use either chlorine or oxygen
as the oxidizing agent.
  Chlorine oxidation  is of questionable value  in most remedial
applications because of the difficulty in assessing the impact of
potentially toxic chlorinated byproducts. Oxygen itself has very
limited effectiveness but may be  of use in the form of ozone or
hydrogen peroxide.
  As an example of an application of chemical  oxidation, IT has
utilized hydrogen peroxide to successfully treat  a sizeable spill of
formaldehyde. The spill resulted  from a leak in an underground
pipeline and had contaminated both the subsurface soil  and the
groundwater. Because the  spill contamination was below a ship-
ping and receiving area, common cleanup techniques involving
excavation would have caused an expensive disruption of plant
operation. Instead the formaldehyde was chemically oxidized  in
situ by injecting a hydrogen peroxide solution  into  the contam-
inated zones. Formaldehyde contamination levels in the soil, orig-
inally as high as 8% (80,000 ppm), were reduced to 15 ppm or less
by this method.
  Biological oxidation uses active microorganisms to biodegrade
organics to acceptable forms. The two major forms of biological
treatment are aerobic (which produces carbon dioxide and water)
and  anaerobic  (which produces  carbon dioxide  and methane).
Biological treatment is getting increased attention as a remedial
alternative because  of its  potential for in situ treatment. Bio-
reclamation basically is the use of indigenous soil bacteria to de-
grade organic contaminants. Nutrients,  such as oxygen, and spe-
cific biological cultures can be added to enhance the degradation.
  Vandalism of a rail car  parked on a railroad siding in Ukiah,
California resulted in  an opportunity for the application of bio-
logical oxidation to an emergency situation. Twenty thousand
gallons of a strong formaldehyde solution drained from the rail
                                                                                                  ON-SITE TREATMENT   99

-------
car into and through the rail-track ballast and into the associated
drainage system. Initial efforts focused on containing and isolat-
ing the spill.
   Once the spill had been controlled, the more difficult task of
decontamination  of the affected area  was addressed. The rail-
way company was vary concerned that normal excavation pro-
cedures would disrupt rail service on the adjacent mainline track.
Thus,  IT developed an alternative approach which initially in-
volved treatment of the  area with alkaline  hydrogen peroxide.
This chemical oxidation  process was successful in reducing the
formaldehyde concentrations to a  level where biological oxida-
tion could be used.
   The biological process made use of a portable aeration tank, a
spray system and the railroad ballast itself. Liquid was pumped
from a sump dug next to the ballast, passed through the aeration
tank and sprayed back onto the ballast. The system was inocu-
lated with  a  specially cultured microorganism that  degraded
formaldehyde but was free of pathogens. The ballast, being com-
posed of coarse rock, supported the growth of the biological med-
ium in a manner similar to a trickling filter. Nutrients were added
to the system as needed. Over a 25-day period, the system suc-
cessfully reduced the formaldehyde concentration from several
hundred ppm to less than 1 ppm.'

Carbon Adsorption
   The use of activated carbon adsorption for aquifer  restora-
tion programs has drawn widespread attention. Several literature
sources point to  its ability to achieve exceptionally good efflu-
ent quality, and the U.S. EPA has endorsed it as  the preferred
treatment  method for meeting drinking water standards.  This
section discusses three basic ways in which carbon can be used:
•  Throwaway carbon
•  Thermal regeneration
•  Nondestructive regeneration

Throwaway Carbon
   One way to consistently ensure good effluent quality is to use
activated carbon absorption on a once-through carbon  basis.
Virgin carbon is capable of removing a broad range of organic
contaminants to low ug/1 levels. A once-through carbon  adsorp-
tion system is easy  to operate,  requires a minimum of operator
attention and its capital cost requirements are relatively low.
   Unfortunately, carbon replacement costs associated with once-
through carbon adsorption systems are very high. The large treat-
ment volumes and/or high concentrations usually associated with
remedial programs result in a high carbon consumption  rate. In
addition,  hazardous substances, when loaded onto activated car-
bon, make the carbon a hazardous waste, requiring disposal in
an approved hazardous waste facility.

Thermal Regeneration
   The most common regeneration technique for activated car-
bon is thermal oxidation, usually accomplished  in a multiple
hearth, fluidized-bed or rotary-kiln furnace. A thermal regenera-
tion unit can be built at the treatment site, but the level of carbon
consumption associated  with most remedial programs  usually
makes it more economical to utilize an off-site thermal regenera-
tion service.
  The advantages of activated carbon adsorption with a  thermal
regeneration service are low capital requirement and ease of oper-
ation,  as with throwaway carbon,  but  the disadvantages of
thermal regeneration are numerous.  Using a thermal regenera-
tion service basically  amounts to "renting"  the carbon. How-
ever, in many cases, the  cost per pound of carbon for  thermal
regeneration is only slightly less than that for purchasing virgin
carbon. Additionally, carbon losses associated with thermal re--
generation are typically 5 to 10% per cycle due to handling losses
and  carbon that  is destroyed during the regeneration  process.
However, thermally regenerated carbon can be expected to have
different performance characteristics and reduced removal effic-
iency as compared to virgin carbon.
  A major problem associated  with thermal regeneration is that
many hazardous substances cannot be thermally regenerated for
one reason or another. Because the hazardous substances are de-
sorbed from  the carbon and destroyed in an afterburner during
thermal regeneration, air  permits for thermal regeneration sys-
tems often contain restrictions on the types of materials that may
be accepted.  Consequently, thermal regenerators often refuse to
deal with carbon that  has been used to adsorb toxic materials
such as PCBs.

Nondestructive Regeneration
  There are three basic ways  that granular activated carbon can
be nondestructively regenerated:
• Using pH shift for weak acids or bases
• Using steam for volatile organics
• Using a solvent for a wider variety of organics
  Steam regeneration, which would be applicable in many remed-
ial situations, is accomplished by passing steam through a spent
adsorber to a condenser and  then to a decanter where the con-
densate  and  immiscible organics  are separated. Carbon losses
associated with nondestructive regeneration are significantly less
than those associated  with thermal regeneration. The carbon is
not physically altered by the nondestructive regeneration process,
and if the carbon  adsorbers are made of the proper material, the
regeneration  can  be accomplished  in  situ, eliminating carbon
handling losses.
  The process is relatively easy to operate and the nondestruc-
tive  nature of the process allows  organic recovery, if  desired.
Even if recovery is not desired,  the disposal requirements are re-
duced from several thousand pounds of contaminated carbon to a
few gallons of organic material.
  IT has evaluated the applicability of carbon adsorption systems
with nondestructive regeneration in a variety of remedial  situa-
tions. A recent feasibility study compared  air stripping, air strip-
ping with vapor-phase adsorption, adsorption using virgin car-
bon, adsorption  using steam-regenerated carbon,  steam  strip-
ping and  chemical oxidation  using UV peroxide for an aquifer
restoration. The critical design parameters were: a contaminated
groundwater  concentration of  30  mg/1 or halocarbons includ-
ing EDB (1,2-dibromoethane);  a maximum flow of 200 gal/min
and  an  average flow of 100 gal/min; and a mandated  effluent
concentration of ^1 ug/1 of each halocarbon present.'
  Details on the  result  of this evaluation can be  found in the
referenced paper. Based on these design parameters, the costs of
air  stripping and steam-regenerated  carbon adsorption were
basically equal and fell in  the range of $300-400/day of system
operation. Subsequent laboratory work,  using samples of the
actual groundwater, verified that  carbon adsorption with  steam
regeneration could achieve acceptable effluent quality and  main-
tain a stable working adsorption capacity.
  Laboratory evaluation of the air  stripping option was post-
poned  until the  issue of  exhausting the  vaporized organics to
the atmosphere could be resolved. This issue is frequently a re-
striction in the application of air stripping.

Air Stripping
  Air stripping is a contaminant removal technique based on con-
centration differentials between a liquid phase and a contacting
gas phase. As air is contacted  with a wastewater stream in a strip-
ping tower, the concentration differential drives the organic con-
 100    ON-SITE TREATMENT

-------
taminant from the liquid to the gas phase. The major advantage
of air stripping is its low overall treatment cost. Both capital and
operating costs  requirements are low compared to most other
technologies.
  The key to air shipping's low overall treatment cost is the
assumption that it can stand alone as a treatment technology. In
many cases, however, air emission standards will require that air
stripping be used in conjunction with a vapor-phase adsorption
unit, significantly affecting the cost-effectiveness of the technol-
ogy.
  Air stripping  is most applicable to compounds  of  a volatile
nature with relatively low solubility in water. Chlorinated organ-
ics such as trichloroethylene, 1,1,1 trichloroethane and aromatic
compounds such as benzene and toluene are  good examples of
chemicals that can be successfully removed by air stripping.  The
overall applicability of the technology can be  expanded to com-
pounds with lower Henry's law constants by preheating the water
prior to its entering the stripping column.
  IT has evaluated air stripping a number of times in treatability
evaluations for a wide variety of clients. In a study conducted for
the American Petroleum Institute, IT engineers conducted side-
by-side, bench-scale evaluations of air stripping and carbon ad-
sorption with steam regeneration to remove  dissolved gasoline
components  from  groundwater.5 Air stripping effectively  re-
moved the aromatic gasoline components from groundwater even
at the normal groundwater  temperature  of SOT.  However, as
would be expected, air stripping was ineffective in removing more
soluble components of gasoline such as tert-butyl alcohol.

MOBILE TREATMENT
  As the expense and regulatory limitations of waste disposal be-
come increasingly important factors, the innovative option of
mobile, on-site  treatment has become an  important  remedial
alternative. Many of the technologies described above can be, and
have been, adapted to mobile applications.
  The U.S. EPA and a limited number of commercial firms  cur-
rently operate mobile incineration systems.  A number of aque-
ous treatment technologies have been mobilized  for on-site re-
mediation including activated carbon, air stripping and ion ex-
change. In addition, mobile  equipment is available to dewater
sludges and reduce the volume of material that ultimately must be
sent to disposal or destruction facilities. These options include
mobile centrifuge and mobile belt filter press equipment.

CONCLUSIONS
  A number of technologies  exist and have been proven  in re-
medial action situations. Experienced, knowledgeable personnel
in the hazardous  waste field  are  able  to efficiently evaluate
options for  corrective action programs, but, in order  to  do so
properly, they must evaluate the site-specific parameters involved
and fully  understand  the constraints, options and goals of the
program.


REFERENCES
1. Helsel, R., Alperin, E.,  Geisler, T., Groen,  A., Fox, R., Stoddart,
   T. and Williams,  H., "Technology Demonstration of a Desorption/
   UV Photolysis Process for Decontaminating Soils Containing  Herbi-
   cide Orange," To be presented before the Division of Environmental
   Chemistry, American Chemical Society, April 1986.
2. "Destroying Dioxin: A Unique Approach,"  Waste Age, Oct. 1980,
   60-63.
3. Sikes, D.J., McCulloch,  M.N. and Blackburn, J.W., "The Contain-
   ment and Mitigation of a Formaldehyde Rail Car Spill Using Novel
   Chemical and Biological In Situ Treatment  Techniques," Proc. of
   the Hazardous Materials Spill Conference, Nashville, TN, Apr. 1984,
   38-44.
4. Parmeli, C.S.,  Allan, R.D. and Mehran,  M.,  "Steam-Regenerated
   Activated  Carbon: An  Emission-Free, Cost-Effective  Groundwater
   Treatment Process," Paper presented at American Institute of  Chem-
   ical Engineers Annual Meeting, Chicago, IL, Nov. 1985.
5. Parmeli, C.S. and Allan, R.D., "Treatment Technology for Removal
   of Dissolved Gasoline Components from Groundwater," Third Na-
   tional Symp. and Exposition  on Aquifer Restoration  and Ground-
   water Monitoring, Columbus,  OH, May 1983.
                                                                                                  ON-SITE TREATMENT    101

-------
                            Field  Studies  of In Situ  Extraction
                          and Soil-Based  Microbial Treatment
                               of  an Industrial  Sludge  Lagoon

                                                 David S.  Kosson
                                               Erik A. Dienemann
                                        Robert C. Ahlert, Ph.D., P.E.
                                         Rutgers,  The State University
                          Department of Chemical and  Biochemical Engineering
                                            Piscataway, New Jersey
ABSTRACT
  Over a period in excess of 10 years, several industrial sludges
were disposed of by landfilling in a surface impoundment. The
resulting  lagoon contains  more  than  30,000  yd3  of sludge.
Leachate from the  sludges has impacted the local groundwater,
thus requiring remediation. Laboratory  studies indicated that in
situ extraction with aqueous sodium hydroxide is a viable method
for removal of organic materials from the sludges. Treatment of
recovered extract was accomplished through use of a soil-based
microbial treatment system. Sequential  aerobic and anaerobic
biodegradation in the laboratory reduced TOC by greater  than
99.5%.
  A field, pilot-scale treatment system has been designed,  con-
structed and operated to demonstrate feasibility of the proposed
renovation program. Field results indicate that exhaustive extrac-
tion of the sludges is possible. Up to a 50-fold increase in the rate of
removal of contaminants  over natural processes can be achieved.
Biodegradation treatment efficiencies in excess of 95°'o have been
demonstrated.

INTRODUCTION
  Over a decade or more,  several  sludges  were  disposed of  by
landfilling in a surface  impoundment.  During  this period  of
operation, the composition and  rate of deposition of sludges
varied. The resulting lagoon contains more than 30,000 yd1  of
sludge. The principal sludges in the lagoon are primary (lime neu-
tralized) and secondary (biological) sludges from treatment of ef-
fluent from diverse chemical  manufacturing  operations.  The
sludges range from solid  to gelatinous in physical state and are
layered in the lagoon. Leachate from the sludges has impacted the
local groundwater.
  Cleanup of the lagoon  is viewed a,s two interrelated problems.
The first problem is the removal of contaminants from the lagoon
without major excavation. The second problem  is treatment  of
the stream containing the stripped contaminants, including  both
organic and inorganic species. To achieve these goals, a process
corsisting of in situ extraction of the sludges followed by on-sile
treatment of recovered extract has been developed.
  Laboratory results have indicated that exhaustive  leaching of
the sludges with an alkaline aqueous extractant (sodium hydrox-
ide) is possible.1 Natural leachate and the sludge extracts can  be
treated effectively  for total  organic  carbon  (TOC) removal
through  use of a soil-based aerobic/anaerobic  microbial  treat-
ment process employing an acclimated mixed microbial popula-
tion in a soil matrix.2 Aerobic biodegradation occurs near the soil
surface where oxygen is available. Anaerobic biodegradation oc-
curs at greater depths. Conversion  of organic solutes to carbon
dioxide and methane is the ultimate result. Treatment efficiencies
in excess of 99% were attained.'
  Preliminary design estimates for implementation of these pro-
cesses indicated that renovation  of the lagoon could be ac-
celerated  to a period of less than 5 years with substantial cost
reductions compared to traditional treatment technology. Thus, a
field pilot plant  was designed, constructed and  operated for
demonstration of the process.

PILOT-PLANT DESIGN
  The pilot system consisted of several sequential process steps
(Fig. 1). The first process step is the extraction of sludges present
in a representative section of the lagoon. Sodium hydroxide solu-
tion is mixed batch-wise in a 200 gal process tank (Tank 1). This
solution is applied to sludges present in the Extraction Bed. The
solution either can be applied to the surface of the Extraction Bed
through perforated pipe or injected  into the sludges through six
well-points. Extract is recovered from the Extraction Bed through
Wells 1 and 2. Extract removed from each well is pumped through
a basket strainer and cartridge filter and into a  1,000 gal storage
tank (Tank 2).
  The  second  process  step is adjustment  of pH, dilution  if
necessary and addition of  nutrients  to extract stored in Tank 2.
This process occurs as a continuous  process in Tank 3 which is a
200 gal process tank, divided into  four equal sections by two
overflow  baffles and one  underflow baffle.  Extract is pumped
from Tank 2 to the first chamber of Tank 3. Carbon dioxide  is
bubbled through the extract to adjust the pH to between 7.0 and
7.5.  Extract passes under an underflow baffle  to the second
chamber  of Tank  3. In  the second chamber, floe created  by the
recarbonation process is allowed to settle. Clarified extract passes
over an overflow baffle into the third chamber of Tank 3. In the
third chamber, dilution with  recycle or potable water and addi-
tion of nutrients  occurs.  Finally, extract underflows into the
fourth chamber from which it overflows and is applied to the sur-
face of the Treatment Bed.
  The third process step is treatment of the modified extract (ef-
fluent  from Tank 3) through an  aerobic/anaerobic  soil-based
microbial treatment process. The treatment process consists of a
soil bed in which an aerobic microbial population is maintained in
the upper region and an anaerobic microbial population is main-
tained in the lower region.1 Extract  applied to the surface of the
Treatment Bed  percolates through the soil  bed where it  is
biodegraded. Effluent  from  the Treatment  Bed  is  recovered
through Well 3. Recovered effluent is pumped to a 500 gal storage
tank (Tank 4) from which it can be recycled onto the Treatment
Bed, recycled to Tank 1 or discharged.
  The  entire process was designed  to be sufficiently  flexible  to
compensate  for fluctuations in either the extraction or treatment
processes as well as weather conditions. Ranges of expected condi-
tions and flow rates for each process step are presented in Table 1.
102   ON-SITE TRLATMhNT

-------
                                                       * BLEED 
-------
                                           WEU.2
                                                                  WCU.3
- n




»*
0^




•T









1.








—
.OT771
"
. ,



\A
SUU06E
^








VWEL







OS
or^fr VCNTdWD
rxwii *
IS






ii l
L"


^
SOIL

ri







T.6AS VEKT
/O'PVC)
i J
1
r
,

vj


--HJ-Y-"

•r^se


^-— GRA
u-^w
^
^1C ^•r°00

DETAIL A
JIIIUHVilllllllllll [&&
KOTO.** LAC«» ..J[J?T^~
                                                                                              T
                                                                                                 1
                                                                                              i:  O
                                                           Figure 2
                                            Extraction and Treatment Bed Construction
placed in the center. An additional 2 ft of sand were backfilled
around the well. Two 1-in. methane vents were installed on top of
the sand, and the remainder of the bed was backfilled with 5 ft of
soil mixture.
  The soil mixture used to backfill the Treatment Bed consisted
of local top soil,  sand  and  granular activated carbon IS: 1:1  by
volume, respectively. The soil mixture was mixed  on-site  by a
bucket loader prior to backfilling.
  Modified extract is applied to the surface of the Treatment Bed
through two 0.5-in. perforated PVC pipes. Two water level sen-
sors are located on the surface of the bed to prevent flooding. Ef-
fluent from the process is collected through Well 3 which contains
a stainless steep deep well pump, a low water level sensor and a
high water level sensor.  Water collected from Well  3 is pumped
directly to Tank 4.

Process Area Design
  The Process Area consists of  a 30  x  35-ft compound  located
orr top of shale fill approximately 100 ft east of the Extraction and
Treatment Beds at an elevation of approximately 10 ft above the
surface of the lagoon (Figs. 3 and 4). Tanks 1,  2, 3 and 4, a
storage shed for CO2 cylinders and nutrient solution, electrical
power supply and controls, a wash basin and safety shower are
located within the compound.
  Interconnections between tanks and the Extraction and Treat-
ment  Beds consist of PVC  solvent-cemented pipe. All piping,
where possible, is buried in shallow trenches approximately 1  ft
deep and backfilled with top soil. The surface of the compound is
covered with roadstone.

Process Controls
  The process controls consist of four operationally independent
controllers. Three of the controls are identical units,  i.e., one
pump controller each for Wells I,  2 and  3. The well pump con-
trols are a high water level sensor and a low water level sensor
(float switches) in each well.  Liquid rising to the level of the high
water level sensor activates the pump in that well. When the liquid
level falls below the level of the low water level sensor, the pump
is deactivated. The actual operating time of each pump is ac-
cumulated on a timer located on the control panel.
104    ON-SITE TREATMENT

-------
                               TOP OF BERM
                 	   PROCESS
    8
                      oo
                      (VI
                           •II
                                        •TO UTILITIES
                                         (POWER AND WATER)
                                                           "i-WELL
                                                                         TO DISCHARGE
                                            EXISTING ROAD
                                       LAGOON SURFACE
                                            	GRADE AND LEVEL WITHIN PERIMETER

                                            	CHAIN LINK FENCE
                                                                                      I' DEEP TRENCH FOR PIPING
                                                                                      (TO BE BACKFILLED)
                         T
                          3
BED
      BED
                      TREATMENT
                      BED
      EXTRACTION
      BED
                                                          Figure 3
                                                      Pilot Plant Layout
  The fourth controller controls operation of pumps, valves and
sensors associated with Tanks 2, 3 and 4 and sensors on the sur-
face of the Treatment Bed. Tank  1 is operated manually, only.
The primary function of the fourth controller  is to regulate ap-
plication of influent to the Treatment Bed. A master 24 hr/7 day
timer regulates  operation cycles  for application. When set on
automatic, the controller will activate whichever of the Tank 2, 3
and 4 valve and pump combinations are set on  automatic during
the preset cycle. These combinations include:
                                  • The actuation valve and pump on Tank 2 (extract feed to Tank
                                    3 or Tank 1)
                                  • The actuation valve and pump on Tank 4 (recycle to the third
                                    chamber of Tank 3, Tank 1 or discharge)
                                  • The valve controlling potable  water for dilution in the third
                                    chamber of Tank 3
                                  • The pump for  addition of nutrients to the third chamber of
                                    Tank3
                                  • The valve controlling recarbonation in the first chamber of
                                    TankS
                                  • The valve controlling floe removal from Tank 3
                           LINK /=jr/vcr
r—
n
•v i
v » i . .e



*











T*

-n


A*
r~*

T
CAT*1



ti-L£ D


\. 	 Sfi^S 	 ^
r«e ivcr* fo «
                               (SHALE)
                                                         Figure 4
                                                  South View of Pilot Plant
                                                                                          SLUD6£^
                                                                TREATMENT
                                                              AMD EXTAAcr/0/J
                                                                   BEOS
                                                                                                             PILIN&~
                                                                                               ON-SITE TREATMENT    105

-------
   The system will be shut off when set on automatic when:

 • A  sensor  on  the  surface of  the Treatment  Bed  indicates
   flooding
 • A sensor in Tank 2 or 4 indicates the tank is empty, if the valve/
   pump combination for that tank is set on automatic
 • An application cycle is completed

 In addition,  each pump and  valve combination can be activated
 or deactivated  manually. The  system  includes  accumulating
 timers to monitor operating time  for each pump in the system. A
 main power  disconnect is provided for this controller, also.

 PILOT-PLANT CONSTRUCTION
   Construction of the pilot-plant began in May 1985.  Brush was
 cleared from the surface of the lagoon in the area proposed for
 placement of the Extraction and Treatment Beds. Approximately
 1-2 ft  of  cover were scraped from the lagoon  surface. The
 resulting surface was leveled.  In addition, an area on top of shale
 fill on the eastern portion of the lagoon was leveled for placement
 of the Process Area.
   Wells 1 and 2 were installed on May 24 and 25, 1985. Three bor-
 ings were conducted before an appropriate location for placement
 of the Extraction Bed was found. It was considered important to
 locate the Extraction Bed in a section of the lagoon that was filled
 continuously with sludges rather than in an area containing semi-
 continuous lenses of shale  fill.  Sheet piling was installed around
 Wells 1  and  2 between June  10 and  13, 1985.  Construction and
 water batching of the pilot-plant was completed on July 15, 1985.

 PILOT-PLANT OPERATION
   The pilot-plant has been inspected either daily or every second
 day during operation.  Field crews consist of at least two trained
 workers. Routine maintenance and inspection of the pilot-plant
 includes:

 • Recording of elapsed time on the control timers for each  pump
 • Recording of the liquid levels in Tanks 2 and 4
 • Visual inspection of the Extraction and Treatment Beds
 • Measurement of the water  level elevation in Wells 1, 2 and  3
 • Mixing NaOH solution in  Tank 1 and feeding  the Extraction
   Bed, if required
 • Adjusting  feed stream  controls and timers for Tank 3
 • Sampling Wells 1, 2 and 3, and Tanks 2, 3 and 4 at times re-
   quired
II
^-d
^ 3
So
 9 -
 8 -
 .7 -
 6 -
 5 -
 4 -
 .3 -
 .2 -
 . I
 I
09-
o.a -
0.7 -
0.6 -
o.s -
0.4 -
0.3 -
0.2 -
0 I -
 0
                          10        GO

                              TIME (DAYS)
                                                             Routine maintenance and inspection usually requires 1 to 2 hours
                                                             of site work per day.

                                                             Extraction Bed Operation
                                                               Recovery of leachate from Wells 1 and 2 began on July 17,
                                                             1985. Application of potable water to the surface of the Extrac-
                                                             tion Bed began on July 23. The purpose of the potable water addi-
                                                             tion was to test the application system from Tank I and the
                                                             hydraulic response of the Extraction Bed. It was indicated that
                                                             the apparent  hydraulic response  of the  bed was  greater  than
                                                             several days.
                                                               On  July  31, influent to  the Extraction Bed was changed to
                                                             0.005  N NaOH.  Throughout August,   low  infiltration  and
                                                             recovery rates  of extractant  were observed (Figs. 5 and 6). Pond-
                                                             ing frequently  occurred at the sludge surface. In response to this
                                                             problem, six well-points were installed in the Extraction Bed to ef-
                                                             fect subsurface injection of extractant. Installation of the well-
                                                             points was  completed  on  August  30 (Day 40). Subsequently,
                                                             sodium hydroxide solution  was injected through  the well-points
                                                             and applied to the surface of the Extraction Bed. Substantial in-
                                                             creases in extract  recovery  were observed (Fig. 6).  On Sept. 19
                                                             (Day 60), influent  to the bed  was increased in strength  to 0.1
                                                             N NaOH. This change  was  made in an attempt to increase TOC
                                                             concentration and quantity  of extract recovered. Recent data in-
                                                             dicate the permeability  of the Extraction Bed is increasing.
                                                               100
                                                                90 -


                                                                BO -


                                                                70 -


                                                                CO -


                                                                50 -


                                                                tO -


                                                                30 -


                                                                20 -


                                                                10 -
                                                                                     40        60
                                                                                         r/.ur (DATS)
                           Figure 5
                 Extraction Bed Influent Volume
                                                                                           Figure 6
                                                                                 Extraction Bed Effluent Volume
Treatment Bed Operation
  Addition of potable water to the Treatment Bed surface began
on July 15. This application continued until field status of the soil
system was established. On July 19, influent to the Treatment Bed
was  changed to a nutrient solution  containing dextrose as the
primary carbon source (Table 3). Inoculation with an activated
sludge culture occurred on July 24.


                           Table 3
       Makeup of Nutrient Solution Applied Beginning July 19

    50   g/1   Dextrose
     8   g/1   NH4C1
    45   g/1   Ca(NO3)2 • 4H2O
     0.3  g/1   K2HP04
This solution is diluted approximately 100:1 in Tank 3
106    ON-SITE TREATMENT

-------
   11
  10. S -


    to -


   9.5 -


    9 -


   8.5 -


    8 -


   7.5 -
                             TIME (DAYS)
                   D   WELL I       +  ITJiY.i 2

                          Figure 7
               Extraction Bed  Well 1 and 2 (pH)
  Operation with nutrient solution as influent continued until
adequate  quantities of  extract  for  treatment  had  been  ac-
cumulated in Tank  2. Treatment of extract began on Sept.  13.
The target TOC concentration for influent to the Treatment Bed
(effluent from  Tank 3) was  1,000 mg/1. The nutrient solution
composition during this  phase of operation included 45  g/1
Ca(NO3)2 and 0.3 g/1 K2HPO4. This solution was diluted by a
factor of approximately 100:1 in Tank 3. The target influent rate
to the Treatment Bed  during this phase  of operation was 1
in./day. The typical application cycle was 4 hr/day.
  On Oct. 17,  the  flows to Tank 3 were adjusted to the target
Treatment Bed influent TOC of 2,000 mg/1. The target influent
rate  of application  was reduced  to 0.5  in./day. These changes
were made so that operation at a similar mass loading with a dif-
ferent hydraulic flux and influent concentration could be exam-
ined. Target flow and  concentration were  difficult to  control
because of variability in water pressure.

RESULTS
Extraction Bed
  Influent and effluent volumes for  the  Extraction  Bed  are
presented in Figures 5 and 6. Initially, influent volumes exceeded
the  volume  recovered  through Wells 1 and 2 by considerable





->^
If
Si






14 -
13 -
12 -
11
10 -
a -
8 -
7 -
6 -
5 -
4 -
3 -
9 _

+

+
% + + + * ++ +
9
D
D +
+ •*• +
D + ++ ++
DmD ++ + 0°° Q+ &
DaDfiQD P a Da++
D
— . -i 	 1 	 r- 	 1 	 1 	 1 	 1 	 1 	 1 	 1 	
amounts. These differences are attributed to several factors. In-
itially, the Extraction Bed was unsaturated and a large volume of
liquid was required to saturate the sludges. Leading to the dif-
ference between the two flows were losses through the joints of
the sheet piling.  Although the area for leakage was considered
small, a significant hydraulic gradient may have existed across the
joints. Losses through the Extraction Bed bottom also accounted
for more liquid loss.
  Recently,  increases  in the volume of extract recovered from
Wells 1 and 2 have been observed. These increases correlate close-
ly to the installation of the well-points for subsurface injection of
extractant. Additional increases have been observed, most likely
the result of increased sludge permeability occurring from the ad-
dition of more concentrated sodium hydroxide solution. Current
influent application rates are approximately 0.5 in./day;  current
effluent recovery  rates are approximately 0.3 in./day. Both in-
fluent application rates and effluent recovery rates are increasing.
During the interval between Days 60 and 100, extract recovery
was approximately 60%.
  Effluent pH,  TOC and  TDS for  the Extraction Bed  are
presented in Figures 7, 8 and 9. Initially, all three quantities were
high. The TOC and TDS of extract recovered from Well 2 were
approximately 11,000 mg/1 and 28,000 mg/1, respectively; the pH
was 10. However,  all three parameters decreased between Days 40
and 60. On Day 60, the TOC and TDS of extract recovered from
Well 2 were 5,200  and  11,000, respectively; the pH was 9.2. Initial
high pH and TDS values are attributed to lime previously applied
to  the  sludges  in the Extraction  Bed.  Decreases  in these
parameters between Days 40 and 60 are considered to be the result
of the lime dissolution and removal from the sludges.
   32

   30
   28

   2B

   24

   22


-------






•"V
Cfmy/2;
tnLsandsi
g£















*
t.
1
^
s














1
a»
^•5
p E
UENT I
(T/icm»,

fe
^






15 -
14 -

13 -
12 -
II -
10-
9 -
B -
7 -

e -
5



3




B.S -
g .


7.5 -
7 -

S.S -

s -
5.5 -





z -
.9 -
.8 -
.7 -

6 -
S -
.4 -
.3 -
2 -
j

t
0.9 -
0,8 -
0.7 -
06-
0.5 -
0.4 -
0.3 -
0.2 -
O.I -
0 -
I


*


+
*** '*
+
D
«m +
D* 3 +
0 * *
_ +

D ^ + ft * *
D ri
cP* o of
" OKLD^
	 1 	 1 	 1 	 1 	 1 	 1 	 1 	 1 	 T— T —I 1 1 '
8 8.4 8.8 3.2 9.6 10 10.4 108
pH
D ITELL i » "ELL 2
Figure 10
Extraction Bed - Well 1 and 1



DO
0 D D^ D

^^ i — " — i 1 1 1 1 1 1 ' 20 40 SO tO 100 TlUf fDArsi Figure 12 Nutrient Tank Tank 3 (to Treatment Bed) Figure 15 Treatment Bed - Well 3 (Effluent TOC) attributed to development of the microbial population and modification of the target application rate to 0.5 in./day. Signifi- cant rainfall occurred during this period, also. Effluent TOC for the Treatment Bed is presented in Figure 15. The peak observed between Days 70 and 90 is the response ex- pected during development and acclimation of the microbial 108 ON-SITE TREATMENT


-------





p,
EFFLUENT







7.8 -
7.B -
7.4 -
7.2 -
7 -
C.8 -
6.B -
B.4 -
6.2 -
6 -

s.a -
S.B -
5.4 -
S.2 -
S -




Rfu

D
O DO
00° D ° °
one ° D D
D QDnD n
O
O D





                                   so        ao        100
                             TlilE (DAYS)
                          Figure 16
              Treatment Bed   Well 3 (Effluent pH)

population. This peak was accompanied by a decrease and subse-
quent increase in effluent pH (Fig. 16). This variation in pH also
was an expected response of the system.  Both phenomena were
observed in previous laboratory column experiments.
  Throughout operation, process effluent pH was between 6 and
8 and effluent TOC and TDS were  less  than 50 mg/1 and 900
mg/1.

CONCLUSIONS
  A pilot-plant  to demonstrate  a state-of-the-art process  for
cleanup of an industrial sludge lagoon has been designed, con-
structed and operated successfully for a period of over 100 days.
The cleanup process consists of two coupled treatment steps. The
first step is removal of contaminants from the sludges through in
situ extraction with aqueous sodium hydroxide. Extractant is in-
jected into the sludges through well-points or applied to the sur-
face through a perforated pipe distribution  network.  Extract is
recovered by means of two wells screened near the bottom of the
sludge deposits.  Operation of this system indicates that up to a
15-fold increase in removal efficiency, over natural processes, can
be obtained through increased  extract concentration. An addi-
tional 4-fold increase in the rate of renovation is obtained through
increased hydraulic flux through the sludges.
  Extract recovered from the sludges is  treated for removal  of
TOC through a soil-based sequential aerobic/anaerobic microbial
treatment system. Treatment occurs on-site,  immediately adja-
cent to the extraction process. A diverse, mixed microbial popula-
tion is developed in the soil system. Neutralized extract is applied
to the surface of the Treatment Bed and allowed to percolate
through the soil system. Aerobic and anaerobic microbial pro-
cesses metabolize organic contaminants to carbon dioxide and
methane.  Treatment efficiencies in excess of 95%  have been
demonstrated.

ACKNOWLEDGEMENTS
  The authors would like  to  acknowledge  several people  at
Rutgers who have performed principal roles in the design, con-
struction and daily operation of the pilot-plant. These people are
Nick Bosko, Betsey Brush, Jim  Claffey, Irene Legiec,  John
Magee,  David Robertson, Judy Svarczkopf and  Valerie Ver-
miglio. In addition, the authors would like to acknowledge Joe
Evankow and Terry Burke for design and fabrication of the con-
trol panel.

REFERENCES
1. Kosson, D.S., Ahlert, R.C., Boyer, J.D., Dienemann E.A. and Ma-
  gee II, J.F., "Development and Application of On-Site Treatment
  Technologies for Sludge Filled Lagoons,"  Proc. International Con-
  ference on New Frontiers for Hazardous Waste Management, EPA/
  600/9-85/025, Sept. 1985, 118-127.
2. Kosson, D.S. and Ahlert, R.C.,  "In-situ and On-site Biodegradation
  of Industrial Landfill Leachate," Environ. Prog. 3, 1984, 176-183.
3. Dienemann, E.A., Magee II, J.F.,  Kosson, D.S. and Ahlert, R.C.,
  "Rapid Renovation of a Sludge Lagoon," paper  presented at the
  AIChE Annual Meeting, Chicago, IL, Nov. 1985.
                                ON-SITE TREATMENT    109

-------
                Cleanup  of  Contaminated Soils and Groundwater
                                   Using Biological Techniques
                                                 Paul E. Flathman
                                             Jason A. Caplan, Ph.D.
                                                O.H.  Materials Co.
                                              Biotechnology Division
                                                    Findlay, Ohio
ABSTRACT
  On-site  biological cleanup following spills  of biodegradable
hazardous organic compounds in lagoon, soil and groundwater
environments is a cost-effective technique when proper engineer-
ing controls are applied. Biodegrdation of hazardous organic con-
taminants by microorganisms minimizes liability by converting
toxic reactants into harmless end products.
  The two case histories presented in this paper describe:
• Bench-scale evaluation of the potential for biological cleanup
  in the spill  site matrix
• Field implementation of biological cleanup systems
  O.H. Materials Co. has performed biological cleanups of spilled
substances since 1978 when a railroad incident resulted in spillage
of acrylonitrile. Subsequent biological environmental restoration
projects have included additional acrylonitrile spills and cleanup
of other materials such as methylene chloride, methylethylketone,
crude oil, petroleum hydrocarbons, butylcellosolve, ethylacry-
late, n-butylacrylate,  toluene,  styrene, acetone,  isopropanol,
tetrahydrofuran and various phenolics.
  Cost-effectiveness, minimal disturbance to existing operations,
on-site destruction of spilled contaminants and permanence of the
solution are several of the advantages identified for implementing
biodegradation as a  technique for spill cleanup and environmen-
tal restoration.

TREATMENT OF METHYLENE CHLORIDE-
CONTAMINATED  GROUNDWATER
  Many current and emerging technologies are available for the
on-site removal of groundwater contaminants. For cost-effective
cleanup following spills of hazardous organic materials, combina-
tions of treatment alternatives often are employed.
  The case history presented here describes the physical and bio-
logical  treatment of methylene chloride-contaminated ground-
water following  rupture of an underground  pipeline. After 2
months of field operation, air stripping techniques provided an
estimated  97% reduction  in  the  concentration of methylene
chloride in the groundwater. In a  downgradient monitoring well
located within 20 ft of the pipeline break, the methylene chloride
concentration was reduced from 9,300 mg/l to 300 mg/  by the
end of the second month.
  Since it became increasingly difficult  to remove the residual
methylene chloride from the groundwater by physical treatment
methods, biological  techniques were initiated.  Biological treat-
ment, using adapted microbial strains, was implemented by the
end of the third month and achieved an estimated 97% reduction of
the residual methylene chloride in groundwater. After 4 months of
field operation, greater than a 99.9% reduction  in the concentra-
tion of methylene chloride was attained.
  On Aug. 16, 1983, it was discovered that a buried  methylene
chloride line, located in proximity to a water main, had ruptured
and an undetermined amount of methylene chloride had leaked
into the soil and groundwater. Following discovery of the methy-
lene chloride leak, OHM was summoned to conduct an emergen-
cy response site investigation. Free product was collected first and
staged in vessels for eventual on-site treatment. Monitoring wells
that were installed established that the contamination was within
a confining clay  layer at the 20-ft level and did not reach the
aquifer located at 100 ft. Monitoring wells were utilized to iden-
tify the contaminant plume, and pumping techniques  were em-
ployed to contain the spilled methylene chloride.
  Emergency response to the spill involved only containment. An
investigative phase was initiated immediately to examine various
alternatives for environmentally restoring the site once the spill
was contained. Several alternatives were considered,  including:
excavation and  disposal; physical containment; and  biological
techniques. After reviewing ihe alternatives, it was determined
that  a  combination  of the alternatives  would  be  the  most
beneficial for site remediation.
  Following removal of 160 yd-' of highly contaminated soil to a
Class A secure landfill, positive placement and suction lift tech-
niques were  implemented.  Air stripping was  used to remove
methylene chloride from the recovered groundwater. When  it
became increasingly difficult to remove methylene chloride from
the groundwater,  biological techniques utilizing adapted micro-
organisms were employed.

Physical Treatment
  Air stripping was selected as the preferred treatment technology
for  removing methylene chloride from the recovered ground-
water. This decision was based on several important factors. The
first  factor was OHM's documentation of the strippability of this
compound in past emergency and remedial cleanup operations.
Second, the regulatory agencies did not require the implementa-
tion  of  vapor phase  scrubbing units to cleanse the air  emissions
from aeration devices. Finally, unlike granular activated carbon,
the   use of  air  stripping  is generally  maintenance-free, and
regeneration and/or disposal of  contaminated media is not
necessary.

Bench-Scale Evaluation
  Characterization and bench-scale testing of recovered ground-
water was performed prior to final selection of the packed column
air  stripping  treatment alternative.  Among  the wastestream
parameters characterized were:

• Influent concentration of volatile organic(s)
• Presence of phased product(s)
• Hardness and alkalinity
110    ON-SITE TREATMENT

-------
FEED
FROM
VACUUM — —
RECOVERY
UNIT


1 SAND I
1 FILTER I





EQUALIZATION
VESSEL
1
                                                                      t\  HEAT
                                                                      UEXCHANGER
                                                                    CONDENSATE
                                                                                            AIR STRIPPER
                                 TO PLANT
                                 WASTE
                               -+• WATER
                                TREATMENT
                                 FACILITY
                                                      SOLVENT
                                                     RECOVERY
                                                           Figure 1
                         Physical Treatment System Schematic for Methylene Chloride-Contaminated Groundwater
• Suspended solids concentration
• Scale-forming tendency at influent pH
Characterizing these parameters determined if pretreatment was
necessary prior to implementation of the primary treatment pro-
cess.  A  particular primary  treatment  process  may  be  cost-
effective, but the pretreatment costs may render the process cost-
prohibitive.
  Methylene chloride was obtained when the groundwater was in-
itially pumped from several monitoring wells.  The recovered
groundwater often  had a murky brown color with  visible par-
ticulate  matter that slowly settled  with the heavier-than-water
methylene chloride.
  Following  the  separation  of methylene  chloride (by phase
separation) from recovered groundwater in quart jar samples, the
supernatant was analyzed for methylene chloride concentration.
The methylene chloride concentration approached the accepted
limit of solubility at ambient temperature.
  Dilling et al.1  calculated the volatilization rate of methylene
chloride  for concentrations ranging  from  several hundred  to
several  thousand mg/1 from  the  liquid-air interface in  open
laboratory vessels. OHM assumed that similar volatilization rates
of methylene chloride would  occur in process settling chambers
following gravity separation. OHM  also assumed that the action
of system pumps and surface winds blowing over water surfaces
into open vessels would result in even greater rates of volatiliza-
tion. Evaluations of the remedial alternatives were based upon
those assumptions.
Field Implementation
  Figure 1 is a schematic of the treatment system designed to treat
groundwater at a rate of 10 to 15 gal/min. The recovered ground-
water was pumped through a mixed-media filter bed to remove
sand and other particulate matter. Three types of porous media
were used in this filter: (1) anthracite, (2) silica sand and (3)  pea
gravel, with the  sand placed above the gravel and the anthracite
above the sand.  This media configuration provided a large pore
size near the filter surface to capture particulates with a  relatively
low head loss. Smaller sized particles were trapped in the lower
sand layer.
  Groundwater leaving the mixed-media  filter bed was piped to
an equalization vessel having a residence time of 14 hr. This vessel
was basically a rectangular holding tank which allowed the denser
methylene chloride to separate from the water by gravity. The
vessel was equipped  with sludge collection equipment to remove
accumulated solids and float controls in the last chamber to main-
tain a constant flow  to the next treatment step.
  Liquid from the equalization vessel was then pumped through a
skid-mounted  shell  and  tube heat  exchanger  to  raise  the
temperature of the  water from 10°C to more  than 40 °C. The
heated water was then pumped to  the top of the air stripping  col-
umn for removal of  soluble methylene chloride.
  Data collected during the project demonstrated that methylene
chloride,  as  pure product,  settled  in the equalization  vessel
together with  fines  and other  particulate matter that  were  not
trapped by the mixed-media filter. While the total volume of the
                                                           REMOVAL EFFIENCY

                                                           —O CALCULATED
                                                             •  OBSERVED
                                                         40        45        50

                                                        COLUMN OPERATING TEMPERATURE

                                                                 CC)
                                                           Figure 2
        Removal Efficiencies of Methylene Chloride from Groundwater by Air Stripping at a Function of Column Operating Temperature
                                                                                                   ON-SITE TREATMENT    111

-------
pure product collected in this manner did not exceed more than a
few gallons,  the vessel operated efficiently. Methylene chloride
concentrations in the water entering the air stripping column rare-
ly exceeded 150 mg/1. The mixed-media filter removed virtually
all paniculate matter  in the water, and backwashing of the filter
was infrequently required. The tower operating temperature was
maintained between 27 and 60 °C.
   Figure 2 is a summary of actual removal efficiencies obtained
during the 3  month operation of the air stripper. The same plot
contains predicted  data. The predicted  removal efficiency was
calculated  from the mass transfer model of Onda et a/.2-M At op-
erating temperatures less than 38 °C, removal efficiencies agreed
with those predicted.
   However,  significant deviations from expected values  were
observed at the higher operating temperatures. One explanation
for those deviations is the non-adiabatic nature of the air strip-
ping system. The column was constructed of mild steel and was
not insulated since  the system was meeting the discharge criteria
of 20 mg/1 methylene chloride. As ambient temperatures became
progressively  colder,  column heat losses  were greater. Those
observations  suggested  that  substantial deviations  from the
predicted removal efficiency would occur as the average operating
temperature of the air stripper was increased.
   The method used for predicting mass transfer rates of methy-
lene chloride during the design phase of the project was adequate
for estimating removal efficiencies  at elevated temperatures. This
same procedure has been implemented  by OHM in numerous air
stripping operations where mobilization of available equipment
was used for emergency situations.  Unlike other volatile organics,
equilibrium and mass transfer data for methylene chloride have
not  been  documented for  aeration  applications at elevated
temperatures.
   Results for the removal of methylene chloride from monitoring
well B-5 are  presented in Figure 3. This well was located  20 ft
downgradient from the ruptured methylene chloride line. An ex-
ponential decay curve was  used to quantify the removal rate
(decline in concentration) of methylene chloride from the ground-
water surrounding the well. Holding other variables constant, the
rate of decrease was assumed to be  a  function  of methylene
chloride concentration, i.e.
     dC
    —   =  -kC                                       (!)

where,

   C  = the concentration of methylene chloride remaining
        (mg/1)
   t = time (days)
   k = the rate constant (day-')

   The curve generated  was  fitted to  the  following first-order
equation:

    C = C0e-i«                                         (2)
where,

   C0 = methylene  chloride concentration at time zero (mg/1)
   The  first-order rate constant, k,  was determined  by  linear
regression using least  squares, and the first order equation was
converted to:

   InC = InC0 - kt                                      (3)

   A  50%  reduction  in methylene chloride  concentration was
observed within the first 11  days of physical  treatment. After  2
months of  field operation, air stripping techniques provided an
estimated  97%  reduction  in the  concentration of methylene
chloride in the groundwater being extracted. Since it became in-
creasingly  more  difficult  to remove  the  residual  methylene
chloride from the groundwater,  biological techniques were con-
sidered for implementation.

Biological Treatment
Bench-Scale Evaluation
  Prior to initiation  of a  methylene chloride biodegradation
feasibility study, the following screening analyses were performed
on 17 selected water samples collected from the site: methylene
chloride,  ammonia-nitrogen, nitrate-nitrogen,  orthophosphate-
phosphorus, pH and aerobic heterotrophic bacterial plate counts.
  Results  of the screening analyses indicated that the site matrix
samples contained viable microbial populations; i.e.,  the  site
matrix did not appear toxic to microbial growth (geometric mean
= 2.0 x  104 CFU/ml, n =  17).  The analyses also indicated that
inorganic  nitrogen and phosphorus nutrient additions would be
necessary  for effective biological treatment of the contaminated
groundwater. Ammonia-nitrogen, nitrate-nitrogen  and ortho-
phosphate were not detected  in any of the groundwater samples
analyzed.  The average pH for all 17 wells was 7.7, an acceptable
value for microbial growth.'
  The biodegradation study was performed using the electrolytic
respirometer  (Exidyne Instrumentation Technologies, Inc., Col-
orado Springs,  CO). For this study, both the  fate of methylene
chloride and  chloride release, resulting from the biodegradation
of methylene chloride, were used as the test parameters. Chloride
concentration was determined using an Orion model 96-17B com-
bination chloride electrode.
                       PHYSICAL TREATMENT
                          DAYS 0-Si
                         UETHYLEMC CHLORIDE
                         r* m 0.7t
                         M C • » 11000 - 00«It
                                      INJECTION OP
                                     BIOLOGICAL MEDIUM
                                     INTO GROUND WATCH
                                       INITIATED
                20      40
                                              too      ito
                           Figure 3
             Physical Removal of Methylene Chloride
                   from Monitoring Well B-5
112    ON-SITE TREATMENT

-------
  An injection pool sample collected in November 1983, was used
for the electrolytic respirometer study. Sludge collected from the
plant's industrial wastewater treatment basin was used to inocu-
late the respirometer vessels. According to reports, the sludge in
the wastewater treatment basin had been continuously exposed to
low levels of  methylene chloride  for years  and was, therefore,
thought to contain populations of microorganisms acclimated to
methylene chloride. The electrolytic respirometer vessels  were
prepared as  follows:
Treatment
Vessel
Number
U
3,4
Inoculum
Source
S
S
Test
Matrix
IP
IP

Poisons
	
X
Poisons added were KCN and NaN3
 S  =  wastewater treatment basin sludge
IP  =  injection pool

  Methylene chloride was added to all respirometer vessels at a
concentration of 100  mg/1.  This  concentration approximated
average methylene chloride concentrations found in the ground-
water at the site. Vessels 1 and 2 were replicates that demonstrated
the feasibility of using the natural microbial flora. Any nonbio-
logical loss of methylene chloride from the test environment
would be quantified with the abiotic control (i.e., vessels 3 and 4).
Respirometer vessels were prepared using procedures described by
Young and Baumann34 with diammonium phosphate and sodium
dihydrogen phosphate  added to each culture vessel. Nitrification
was not suppressed.
                           Figure 4
            Biodegradation of Methylene Chloride with
             Chloride Release in a Microbial Reactor
  On a  periodic  basis, 40  ml aliquots were  removed from
respirometer  vessels  and analyzed for methylene chloride and
chloride  concentrations. Throughout the study, aliquots were
periodically removed from culture vessels and analyzed for pH
and for ammonia-nitrogen, nitrate-nitrogen and orthophosphate
concentrations. When necessary, pH and inorganic nitrogen and
orthophosphate concentrations were adjusted to ensure that the
chemical environment remained optimal for microbial growth.
  Methylene  chloride biodegradation and chloride release data
for respirometer vessel  1 are presented in Figure 4.  Results for
vessel 2  were similar.  Neither loss  of methylene chloride nor
chloride release was observed in the abiotic control vessels.
  The theoretical chloride release for mineralization of methylene
chloride was calculated from the following relationship:
                                                                    CH2C12
                           + 2HC1
The theoretical chloride release is 0.835 mg chloride per mg of
methylene chloride following complete oxidation.
  Methylene chloride was  biodegraded  stoichiometrically  to
chloride in respirometer vessels 1 and 2 (Fig. 4). Using the follow-
ing relationship, it  was determined that 130 and  120%  of the
theoretical chloride release for methylene chloride biodegradation
was met in respirometer vessels 1 and 2.
  The results indicated that methylene chloride  was completely
oxidized to carbon dioxide, water and the chloride  ion. The lack
of  a  significant lag  period  indicated  the  presence of adapted,
naturally occurring microbes that could biodegrade methylene
chloride.

Field Implementation
  Use of  the underground  recovery  and  treatment system for
aquifer restoration has been described previously.n'15' 25>27> 28 A
schematic of the biological treatment system is shown in Figure 5.
  A modified activated sludge system was the preferred method
for aboveground treatment. In addition to providing efficient
biological treatment, the activated sludge system provided a sup-
ply of adapted microorganisms for inoculation  of the ground-
water through the injection system. The injection system was used
to inoculate the groundwater with naturally occurring  microbes
capable of biodegrading methylene chloride and  to  provide  a
suitable chemical environment necessary  to support microbial
growth.
                                                                                NUTRIENTS
                                                                                                   »H
                                                                                               ADJUSTMENT
                                                                                                                       TO
                                                                                                                   UNDERGROUND
                                                                                                                     INJECTION
                                                                                                                     SYSTEM
                          Figure 5
           Biological Treatment System Schematic for
         Oxidation of Methylene Chloride in Groundwater
                                                                                                  ON-SITE TREATMENT    113

-------
   I!
               BIOREACTOR

               BIOREACTOR INFLUENT
                      * BIOREACTOR DRAINED INTO INJECTION
                         POOL WHILE BEING REFILLED
                     INJECTION OF    CONTINUOUS
                    BIOLOGICAL MEDIUM BIOLOGICAL
                    MTO GROUND WATER TREATMENT
                       HT1ATED      INITIATED
                           Figure 6
        Chloride Release in the Bioreactor Resulting from the
        Biodegradation of Methylene Chloride in Groundwater
  The recovery  system  was  used to withdraw contaminated
groundwater for  aboveground treatment. The supernatant from
the treatment system was reinjected into the subsurface environ-
ment,  creating  a  closed-loop   system.   Biodegradation   of
methylene chloride took place in the ground as well as above the
ground in the biological treatment system.
  On day 82 of field operation, the bioreactor was prepared for
batch  treatment  of methylene chloride-contaminated ground-
water to  provide a large quantity of adapted microorganisms for
injection. To maintain a methylene chloride concentration in the
bioreactor capable of supporting microbial growth, air containing
methylene chloride vapor was bubbled into the bioreactor.  Be-
tween days 82 and 87, an 80.4% increase in the chloride concen-
tration was observed as a result  of  chloride release from  the
biodegradation of methylene chloride. Between days 87 and 90,
the bioreactor  was drained  into the injection pool while being
refilled. Injection of methylene chloride adapted microorganisms
into the groundwater environment began on day 89 (Fig.  6).
  Batch biological treatment of methylene chloride-contaminated
groundwater again was performed in the bioreactor between days
90 and 92. A 110% increase in chloride concentration was  ob-
served as a result of  methylene  chloride  biodegradation. The
bioreactor again was drained into the injection pool while being
refilled. On day 95, continuous biological treatment was initiated.
Over the 10-day operational phase, bioreactor effluent chloride
concentration averaged 131 ±65% greater than influent concen-
tration. Therefore, the bioreactor contained  a population  of
microorganisms adapted to methylene chloride.
  Removal of  methylene chloride from  the groundwater  was
rapid. A 50% reduction in methylene chloride concentration was
observed in monitoring well B-5 (Fig. 7) within 8 days after the
commencement of biological treatment.  The rate of decrease was
assumed to be first-order;" i.e., holding other variables constant,
the rate of decrease was assumed to be  a  function of methylene
chloride concentration.
  With  the onset of winter, biological treatment was temporarily
suspended on day  123 (Jan. 2, 1984). At that time, a 91% reduc-
tion in methylene chloride concentration had been achieved since
the injection of adapted microorganisms into the groundwater.
  The theoretical chloride release for biodegradation of 186 mg/1
methylene chloride (i.e., from 192  to 6 mg/1 methylene chloride)
is  155 mg/1.  The  observed chloride release was 156 mg/1; i.e.,
chloride concentration in monitoring well B-5 increased from 175
mg/1 on day 80 to  331 mg/1 on day 123. Therefore, we concluded
that biological  treatment was  responsible  for the removal of
methylene chloride from the groundwater.

TREATMENT OF ETHYLENE
GLYCOL-CONTAMINATED GROUNDWATER
  At the Naval Air Engineering Center in Lakehurst, New Jersey,
biodegradation techniques were used successfully to treat ethylene
glycol-contaminated   groundwater  following the  loss  of  an
estimated 4,000 gal of cooling water from  a lined surface storage
lagoon.   The cooling  water  was  estimated  to contain 25%
(vol/vol) ethylene  glycol.
  The problem developed on Jan. 5, 1982, following a storage
lagoon  liner  break.  A  subsequent investigative program con-
firmed soil contamination around the  lagoon and identified a
180-ft long by  45-ft wide contaminant  plume extending to the
east. At the start of the project,  the average ethylene glycol con-
centration in the groundwater was  1,440 mg/1. Approximately 85
to 93%  of the ethylene glycol was removed from the groundwater
within the first 26 days of biological treatment. By the completion
of the project, ethylene glycol was  reduced to below the limits of
detection (<,50 mg/1) in all production wells at the site.
       I ,0
                 INJECTION OF >IO(.OaiCAl
                 UEOAJUINTO OROUHD
                 WATER INITIATED
                                 METHYLCNE CHLORIDE


                                 •* • 0.71
                                 In C B In 100000 - 0.0851
           •o    »o    100    no   no   iJo    MO   i»o
                             DAYS

                           Figure 7
     Biodegradation of Methylene Chloride with Chloride Release
         Based on Samples Taken from Monitoring Well B-5
114    ON-SITE TREATMENT

-------
Biofeasibility Evaluation
  The biofeasibility study assessed the biodegradation potential
of ethylene glycol in the ground as well as the presence of a toxic
or inhibitory environment to microbial growth. In addition, tech-
niques to increase the biodegradation rate in the ground also were
evaluated. Such techniques commonly include the use of surfac-
tants, cosubstrates, primary substrates, vitamins, trace elements
and selected microbial strains.  The results of the  biofeasibility
evaluation form the basis for subsequent field implementation, as
discussed in this case history.
  Ethylene glycol can be used as a carbon and energy source for
aerobic migroorganisms.6'10.17^1 The aerobic metabolism  of ethy-
lene  glycol is  relatively  common,  and  the  pathways of its
metabolism are known.4,5,18,19,26,31-33
  Anaerobic metabolism of ethylene glycol has been reported by
Dwyer  and Tiedje.8  Using  a  sewage  sludge inoculum under
methanogenic  conditions,  ethylene glycol was  converted to
ethanol, acetate and methane. The ethanol produced was further
oxidized to acetate with methane as the final end product. It has
been shown that clostridium glycolicum fermentation of ethylene
glycol yields equimolar amounts of acetate and ethanol.16
  Prior to initiation of the feasibility study, screening analyses, as
previously described for the first case history, were performed on
representative soil and groundwater samples collected from the
site. Ethylene  glycol  concentration was determined by direct
aqueous injection into a gas chromatograph.
  Screening analyses indicated that the groundwater contained a
viable microbial population; i.e., the environment did not appear
toxic  to microbial growth. Aerobic  heterotrophic bacterial
population densities ranged from 102 to 106 colony forming units
(CFU)/ml.  Bacterial population densities of soil samples collected
within the spill area varied from 102 to 106 CFU/g oven-dry soil.
Screening also indicated that both pH adjustment and inorganic
nitrogen and phosphorus nutrient additions would be necessary
for biological treatment of ethylene glycol. The average pH of the
13 samples  was  4.5.
                                   •  TOTAL BOD
                                       n = 40
                                       r2 = 0.85B
                                       y = 1740 - 4480 •"'
                                     ETHYLENE OLVCOL
                                       n = 6
                                       r2= 0.813
                                       In y = In 44100 - 0.7»»t
  The biodegradation study was performed using the electrolytic
respirometer with the composite groundwater sample. All samples
were prepared in a liquid nutrient medium containing basal salts
and yeast extract.
                       Treatment
Vessel
Number
1
2
3
4
Test
Matrix
OW**
GW
LGWf
GW
Glucose
(1000 mg/1)
—
X
X
Poisons*
—
....
X
*Poisons added were HgCl2, KCN, NaN3
**GW = Composite groundwater sample
tLGW = Laboratory grade water

  Results from vessel 1 would measure the rate of biodegradation
of ethylene glycol by the natural microbial flora. Comparison of
oxygen uptake rates in vessels 2 and 3  would demonstrate the
presence of significant quantities of substances toxic or inhibitory
to microbial growth in the composite groundwater sample. Any
nonbiological loss of ethylene glycol would be quantified with the
abiotic control (i.e., vessel 4).
  Ethylene glycol biodegradation data for respirometer vessel 1 is
presented in Figure 8. A loss of ethylene  glycol was not observed
in the abiotic control (vessel 4). The  theoretical oxygen demand
(ThOD)  for  aerobic mineralization of ethylene  glycol was
calculated from the  following relationship:

  2 (CH2OH)2 + 5  O2 	> 4 CO2  + 6 H2O
The ThOD is 1.29 mg O2 per mg of ethylene glycol. With an in-
itial concentration of 1440 mg/1 ethylene glycol,  1860 mg O2
would be required for complete oxidation. The ultimate BOD
(BODu)  was determined by fitting  oxygen uptake data to  a
modified crescent curve:30
  BOD = a + be-kt                                     (4)

where,
BOD =  the amount of BOD expressed or exerted at time t
         (mg/1)
    a =  BODu, the ultimate amount of oxygen uptake to be
         expressed (mg/1)
    b =  a lag period parameter (mg/1).  (Note: b = a if the fitted
         crescent curve intersects the origin)
    k =  the rate constant (hr-i)
    t =  time (hr)
The ultimate BOD is defined as  the total amount of oxygen re-
quired to biodegrade the  immediately available organic matter
present in a sample.24
  Using the following relationship, it was determined that 94% of
the theoretical oxygen demand for ethylene glycol biodegradation
was met:
                                                                   BODu

                                                                   ThOD
          x 100
(5)
                          Figure 8
         Biodegradation of Ethylene Glycol in Composite
                Groundwater Sample (Vessel 1)
Thus, the results indicate that ethylene glycol was completely ox-
idized to carbon dioxide and water without the accumulation of
incomplete oxidation products. With a BODu of 1700 mg/1, an
estimated minimum of 170 mg/1 NH3-N and 17 mg/1 PO,}-?
would be required in the groundwater to prevent nitrogen and
phosphorus  from  limiting  microbial  growth during biological
treatment."
                                                                                                 ON-SITE TREATMENT    115

-------
        1000

        too

        BOO

        700
                                      -»-• VESSEL,
                               COMPOSITE OROUNO WATER SAMPLE
                                4- BASAL SALT! 4- OLUCOSE
                                    O.B8S
                                  Ts In SB.4 + O.OSB7I
                             • BASAL SALTS » OLUCOSE
                                t MICROaiAL INOCULUM
                                 n   7
                                 rf = O.BBS
                                 In y = hi BS.1  4- 0.04781
                             30
                            HOURS
                           Figure 9
    Oxygen Uptake Data for the Composite Groundwater Sample
           Containing Glucose and for a Glucose Control
   The lack of a significant lag period in both oxygen uptake and
 ethylene glycol biodegradation by the natural microbial flora in
 vessel 1  indicated the presence of adapted microbes that could
 biodegrade ethylene glycol. That Finding was significant because
 it indicated that in situ biodegradation of ethylene glycol already
 was occurring and that the management approach should be to in-
 crease the natural biodegradation rate.
   The  groundwater appeared  to  be slightly stimulatory  to
 microbial growth, as evidenced by the slightly greater rate of oxy-
 gen uptake in the composite groundwater sample compared to the
 glucose-basal salts control  (Fig.  9). At  the termination of the
 study, initial oxygen  uptake data in the toxicity/inhibition control
 (respirometer vessels 2 and 3) were compared on a semi-log plot.
 A comparison was  made  of the  slopes of the lines-of-best-fit
 through the linear portion of the natural log transformed oxygen
 uptake data. In the linear portion of the semi-log plot, the rate of
 oxygen uptake in the glucose-containing composite groundwater
 sample was 1.2 times greater than in the  glucose-basal salts con-
 trol. Thus, evidence for an environment toxic or inhibitory to
 microbial growth was not found.
   The feasibility study results indicated that biodegradation tech-
 niques were a viable  option for reducing  ethylene glycol concen-
 trations in the groundwater.

 Field Implementation
   The biological treatment  program at the site was divided into a
 14-day operational phase,  a 3-month monitoring phase and a
 9-month  maintenance program.  The operational  phase was de-
 signed primarily to provide maximum recovery, treatment and en-
 hanced bacterial growth in  the groundwater both within the spill
 area and within the contaminant plume (Fig. 10). The monitoring
 program  was designed to assess  the ethylene glycol degradation
 rate  in  the groundwater following  nitrogen and phosphorous
 nutrient addition, pH adjustment and enhanced  bacterial growth.
The maintenance program  was designed  to  provide an environ-
ment suitable for the continued  biodegradation of any residual
ethylene glycol remaining in the groundwater environment.
STORAGE LAQOOr^

0
\
jO.
V" ,
\

o
• 0
                                                                                                                   CONTAMINANT
                                                                                                                     PLUUE.
                                                                                                                     OnOUNO WATCH
                                                                                                                        'LOW
                                                                       O RECOVERY WELL

                                                                       O (OIL SORINO

                                                                       • BOIL IORINO / MONITOR WELL
                           Figure 10
            Schematic of Ethylene Glycol Storage Lagoon
               and Contaminant Groundwater Plume
   Ethylene glycol contamination at the site was divided into two
 zones. The first zone was the unsaturated zone between the sur-
 face and the water table where ethylene glycol had been retained
 through  capillary  action.  The  highest  contamination  level
 detected in that zone was 4,900 mg/1 ethylene glycol. Surface con-
 tamination also was indicated  following analysis  of shallow
 samples (0 to 2 ft) collected adjacent to the lagoon.
   The second  zone of  contamination  was  the  groundwater.
 Groundwater samples within the spill area had ethylene glycol
 concentrations as high as  2,100 mg/1. Monitoring wells were in-
 stalled, and  subsequent  water analyses  indicated significant
 groundwater contamination. A downgradient contaminant plume
 was estimated to be 180 ft long by 45 ft wide.
   A two-phased approach was implemented to deal with both soil
 and groundwater contamination. Using  injection  and recovery
 wells,  initial  efforts addressed  the highly contaminated  soils
 underlying the storage lagoon. The second phase concentrated on
 groundwater cleanup. Its goals were (1) to prevent further migra-
 tion  of contaminated groundwater  and (2)  to  treat any  con-
 tamination released during treatment of the unsaturated zone.
   The injection system was used to adjust groundwater pH as
 well as provide the inorganic nitrogen and phosphorus necessary
 to support microbial growth. The recovery system withdrew con-
 taminated groundwater  for  above ground treatment in an ac-
 tivated sludge treatment system (Fig.  11). Supernatant  from the
 treatment system then was reinjected into  the subsurface environ-
 ment, creating a closed-loop system.
                            pH ADJUSTMENT
                     PREMIX I NUTRIENT ADDITION
                    1 TANK '
RECOVERY—.
 WELLS  «
                                                   WASTE SLUDGE
                                                    FOR SURFACE
                                                     APPLICATION
                           Figure 11
      Flow Diagram for the Activated Sludge Treatment System
           of Ethylene Glycol-Contaminatcd Groundwater
116    ON-SITE TREATMENT

-------
                                                                   ETHYLENE GLYCOL
                                                                    O BELOW DETECTION LIMIT

                                                                   MICROBIAL POPULATION

                                                                   DH
                                                           Figure 12
     Ethylene Glycol Concentration, Bacterial Population Density, and pH as a Function of Time for a Contaminant Plume Production Well
  Based  on the information obtained during the site investiga-
tion, five recovery wells were installed to recover contaminated
groundwater (Fig. 10). Three wells located near the lagoon pro-
vided zones of attraction for treated water injected in that area.
The  remaining two recovery wells  were positioned  east of the
lagoon to  recover contaminated water  from  the contaminant
plume and aid in the distribution of treated water from the plume
injection system.
  Following pH adjustment and nutrient addition, supernatant
from the biological treatment system was recharged through a
three-phase  injection  system.  The lagoon  injection  system  was
employed to flush contaminated soil and thereby transmit con-
taminated water to the three recovery wells located in the vicinity
of the lagoon. The plume  injection system  was similar to the
lagoon system in  both construction and operation. The primary
functions of the plume injection system were to enhance bacterial
growth through pH adjustment and nutrient addition  and to
create a  gradient from the fringe of the plume toward the  two
recovery  wells located at the center  of the contaminant plume.
The third injection phase was implemented through surface ap-
plication. Surface application was used primarily in the lagoon
area to flush the unsaturated  zone and  enhance bacterial growth
in the contaminated soil.
  The activated sludge biological treatment system was designed
to reduce ethylene glycol concentration through microbial bioxi-
dation using the indigenous microbial flora. A second role of the
bioxidation reactor was to  provide adapted microorganisms for
the three-phase injection system (Fig. 11). The biofeasibility study
had demonstrated that the  natural microbial flora were adapted
to biodegrading  ethylene glycol and that the management ap-
proach  for  the project  should  be  to  increase the natural
biodegradation rate using  microbes indigenous to the ground-
water.
  Throughout the  operational  phase, treatment system and
recovery well samples  were analyzed for  aerobic heterotrophic
bacteria, pH, dissolved oxygen, inorganic nitrogen, phosphorus
and ethylene glycol. Based on the results of those analyses, en-
vironmental parameters were modified to maintain an effective
                                                                        ETHYLENE QLYCOL
                                                                          O BELOW DETECTION LIMIT
                                                                        MICROBIAL POPULATION
                                                                        pH
                                                           Figure 13
    Ethylene Glycol Concentration, Bacterial Population Density, and pH as a Function of Time for a Downgradient Spill Area Production Well
                                                                                                   ON-SITE TREATMENT    117

-------
 rate of biological treatment. During the monitoring phase, those
 same  parameters were quantified  on  a monthly  basis from
 recovery well samples.
   In the initial phases of treatment, ethylene glycol concentra-
 tions in two plume recovery wells were reduced from 690 and 420
 mg/1 to below limits of detection (LOD  = 50 mg/1) within 26 days
 of treatment. With  the exception of one monitoring phase data
 point (day 40),  ethylene  glycol concentration remained below
 detection limits. As shown  in Figure 12, groundwater pH  was
 maintained within an optimal range for bacterial growth (i.e., pH
 6 to 7). Bacterial population density was returning to background
 levels (i.e., 105 CFU/ml) following a high of 10« CFU/ml.
   In the two downgradient recovery wells adjacent  to the storage
 lagoon, the groundwater concentration  of ethylene glycol  was
 reduced by more than 85% and 92%, respectively, within the first
 26 days of treatment (Fig. 13). Initial concentrations were 3,400
 and 1,100 mg/1, respectively. In the upgradient recovery well ad-
 jacent to  the lagoon, the ethylene  glycol concentration in the
 groundwater  increased from 100 mg/1  on day 0 to 880 mg/1 by
 day 13. However, monitoring data showed a continued decrease
 in  ethylene glycol concentration after day 13. The increase in
 ethylene  glycol  concentration  prior  to day  13  reflected  the
 project's  aggressive operational phase, whereby,  the  injection
 system flushed pockets of ethylene glycol from the unsaturated
 zone into the groundwater.
   The maintenance  program focused on the removal of those re-
 maining pockets of contamination,  particularly in  the  lagoon
 area. As part  of that program, both lime and diammonium phos-
 phate were applied to the soil surface in the lagoon  and contami-
 nant plume areas. Lime increased the pH of the soil and ground-
 water to a  favorable range for microbial growth. Diammonium
 phosphate,  a  readily  available  source  of  nitrogen  and
 phosphorus,  supported  the growth  of an increased microbial
 population.
   By the completion of the project, ethylene glycol was reduced
 to below the limits of detection in all production wells at the site.

 CONCLUSIONS
   The two case histories presented demonstrate the application of
 biological techniques for environmental restoration of areas con-
 taminated  with spilled organic materials.
  The combination  of physical  and biological techniques effec-
 tively removed methylene chloride from the groundwater  at a
 relatively rapid rate. If it  had been necessary to remove the con-
 taminated  soil from  the site, transportation and disposal (T&D)
 costs for 14,000 yd3 of contaminated soil would have been at least
 $1,050,000. The estimate for traditional treatment was based on a
 transportation cost of $3/mile in 20-yd3 sized trucks to a hazar-
 dous waste disposal facility 100 miles from the treatment site. The
 disposal cost for the contaminated soil was $60/yd3.
  By successfully decontaminating the soil and allowing ii to  re-
 main on-site, a two-fold cost savings was achieved and all future
 liability was reduced substantially. This cost  estimate  for T&D
 was conservative since removal of soil from an area increases void
 volume for transportation and disposal often by a factor of 1.2 to
 1.3. Moreover, hazardous materials cannot be shipped off-site in
 bulk,  but  most be  packaged into  drums. Additional  chemical
 analyses generally are required to characterize the waste prior to
 off-site transportation and disposal. Inclusion of these additional
 factors that must be considered for off-site disposal  resulted in an
 even greater cost benefit for in situ cleanup.
  Biological techniques, as described in  the second case history,
were  also effectively used to remove ethylene glycol  from  the
groundwater at a relatively rapid rate. The flexibility of the injec-
tion/recovery system in maintaining an environment conducive to
 biodegradation while flushing ethylene glycol from the ground-
water was a key factor in the removal of scattered pockets of con-
tamination. If it had been necessary to remove the soil from both
the spill area and the contaminant plume to a hazardous waste
disposal facility 500 miles from the treatment site, the T&D cost
would have been more than 16 times the cost for on-site treat-
ment.
  Biodegradation as a method  for spill cleanup and environmen-
tal restoration is considered to be a  promising technology. Land
treatment techniques have been engineered and are accepted as an
economical and environmentally sound means of destruction for
many types of industrial wastes.  With regard to the cleanup of
contaminated soil and groundwater, excavation has been a com-
mon method for remediation. However, combinations of physical
and biological techniques now are gaining increasing acceptance as
a practical, cost-effective alternative for environmental restoration.
REFERENCES
 I.  Adas, R.M. and Bariha, R.. Microbial Ecology: Fundamentals and
    Applications, Addison-Weslcy Publishing Company, Reading, MA,
    1981.
 2.  Bolles, W.L. and Fair. J.R., "Improved Mass-Transfer Model En-
    hances Packed-Column Design," Chem. Eng.. 84(14), 1982, 109-116.
 3.  Bordner, R., Winter, J. and Scarpino, P.,  Eds. "Microbiological
    Methods for  Monitoring  the  Environment-Water  and Wastes,"
    EPA 600/8-78-017. Environmental Monitoring and Support Labora-
    tory. U.S. EPA, Cincinnati, OH, 1978.
 4.  Caskey, W.H. and Taber, W.A., "Oxidation of Ethylene Glycol by
    a Salt-Requiring Bacterium," Appl.  Environ. Microbiol.. 42, 1981,
    180-183.
 5.  Child,  J. and  Willetts, A.,  "Microbial Metabolism of Aliphatic
    Glycols.  Bacterial  Metabolism  of  Ethylene  Glycol."  Biochem.
    Biophs.  Ada. 538, 1978, 316-327.
 6.  Claus, D. and  Hempel, W.. "Specific Substrates for Isolation and
    Differentiation  of Azotobacler vinelandii," Arch. Mikrobiol, 73,
    1970, 90-96.
 7.  Dilling,  W.L.,  Tefertiller, N.B. and  Kallos.  G.J., "Evaporation
    Rates of Methylene Chloride. Chloroform,  1,1,1-Trichloroethane,
    Trichlorocthylene,  Tetrachloroethylene,  and  Other  Chlorinated
    Compounds in  Dilute Aqueous Solutions," Environ. Sci. Technol.,
    9, 1975, 833-837.
 8.  Dwyer,  D.F. and Tiedje. J.M., "Degradation of Ethylene Glycol
    and Polyethylene Glycols by  Methanogenic Consortia," Appl. En-
    viron. Microbiol., 46, 1983, 185-190.
 9.  Anon.,  "Current Developments,  Hazardous Waste," Environ.
    Reporter. May  6. 1983. 11-12.
10.  Fincher, E.L. and Payne, W.J., "Bacterial Utilization of Ether Gly-
    cols," Appl. Microbiol.. 10, 1962, 542-547.

11.  Flathman. P.E.,  McCloskey.  M.J., Vondrick. J.J. and Pimlett,
    D.W., "In Situ Physical Biological Treatment  of Methylene Chlor-
    ide (Dichloromethane) Contaminated Ground Water,"  Proc.  of the
    Fifth  National Symposium on Aquifer Restoration and Ground
    Water Monitoring, Columbus. OH,  May 1985, 571-597.
12.  Flathman, P.E. and Caplan, J.A., "Biological Cleanup of Chemical
    Spills,"  Proc.  of HAZMACOM 85.  Oakland, CA, Apr. 1985, 323-
    345.
13.  Flalhman, P.E. and Githens, G.D., "In Situ Biological Treatment
    of Isopropanol, Acetone, and Tetrahydrofuran in the Soil 'Ground-
    water Environment," E.K. Nyer,  Ed., Ground  Water Treatment
    Technology. Van Nostrand Rcinhold  Company,  New York. NY,
    1985, 173-185.
14.  Flalhman, P.E..  Quince,  J.R. and Bottomley,  L.S.. "Biological
    Treatment of Ethylene Glycol-Contaminated Groundwater at Naval
    Air Engineering Center, Lakehursl, New Jersey," Proc. of the Fourth
    National Symposium on Aquifer Restoration and Ground  Water
    Monitoring. Columbus, OH, May 1984. 111-119.
 118    ON-SITE TREATMENT

-------
15-  Flathman, P.E.,  Studabaker,  W.C., Oithens,  G.D. and  Muller,
    B.W., "Biological Spill Cleanup," Proc. of the Technical Seminar
    on Chemical Spills, Toronto, Ontario, Oct. 1983, 117-130.
16.  Gaston, L.W. and Stadtman, E.R., "Fermentation of Ethylene Gly-
    col by Clostridium glycolicum sp. /. Bacterial., 85,  1963, 356-362.
17.  Gonzalez,  C.F., Taber, W.A. and Zeitoun, M.A., "Biodegrada-
    tion of Ethylene  Glycol by a Salt-Requiring Bacterium," Appl.
    Microbiol., 24, 1972, 911-919.
18.  Harada, T. and Nagashima, Y., "Utilization of Aklylether Com-
    pounds by Soil Bacteria," J. Ferment. Techno!., 53, 1975, 218-222.
19.  Jones, N.  and Watson, G.K.,  "Ethylene  Glycol and Polyethylene
    Glycol Catabolism by a Sewage Bacterium," Biochem. Soc. Trans.,
    4, 1976, 891-892.
20.  Kemmer,  F.N., Ed., Nalco Water Handbook, McGraw-Hill  Book
    Company, New York, NY,  1979.
21.  Klecka, G.M., "Fate and Effects of Methylene Chloride in  Acti-
    vated Sludge," Appl. Environ.  Microbiol., 44, 1982, 701-707.
22.  Kobayashi, H  and Rittman,  B.E., "Microbial Removal of Haz-
    ardous Organic Compounds," Environ.  Sci.  Tech.,  16, 1982, 170A-
    183A.
23.  Larson, R.J., "Role of Biodegradation Kinetics in Predicting En-
    vironmental Fate," A.W. Maki, K.L. Dickson, J. Cairns, Jr.,  Eds.,
    Biotransformation and Fate of Chemicals in the Aquatic Environ-
    ment, American Society for Microbiology, Washington, DC,  1980
    67-86.
24.  Mitchell,  R.,  Introduction  to Environmental Microbiology,  Pren-
    tice-Hall,  Inc., Englewood Cliffs, NJ, 1974.
25.  Ohneck, R.J. and Gardner, G.L., "Restoration of an Aquifer Con-
    taminated  by  an Accidental Spill of Organic Chemicals," Ground
    Water Monitoring Review, 2(4), 1982, 50-53.
26.  Pearce, B.A.  and  Heydeman,  M.T., "Metabolism of Di(Ethylene
    Glycol) [2-(2'-Hydroxyethoxy) Ethanol] and Other Short Poly (Ethy-
    lene Glycol)s by Gram Negative Bacteria," J. Gen.  Microbiol, 118,
    1980, 21-27.
27.  Quince, J.R. and Gardner, G.L., "Recovery and Treatment of Con-
    taminated Ground Water: Part I," Ground  Water Monitoring Re-
    view, 2(3), 1982, 18-22.
28.  Quince, J.R. and Gardner, G.L., "Recovery and Treatment of Con-
    taminated Ground Water: Part II," Ground  Water Monitoring Re-
    view, 2(4), 1982, 18-25.
29.  Roberts, P.V., Hopkins, G.D., Munz, C. and Riojas, A.H., "Evalu-
    ating Two-Resistance Models for Air Stripping of Volatile Organic
    Contaminants  in  a Countercurrent, Packed Column," Environ.
    Sci. Tech., 19, 1985, 164-173.
30.  Shammas, N.C., "Modified Crescent Curve Fitting, Program Num-
    ber 341C," Users' Library, Hewlett-Packard Company, Corvallis,
    OR, 1983.
31.  Thelu, J.,  Medina, L. and Pelmont, J., "Oxidation of Polyoxy-
    ethylene Oligomers by an Inducible Enzyme from  Pseudomonas
    P400," FEMSLett., 8, 1980, 187-190.
32.  Wiegant, W.M. and DeBont, J.A.M., "A New Route for Ethylene
    Glycol Metabolism in Mycobacterium  E44," /. Gen. Microbiol.,
    120, 1980, 325-331.
33.  Willetts, A., "Bacterial Metabolism of Ethylene Glycol.  Biochem,"
    Biophys. Acta, 677, 1981, 194-199.
34.  Young, J.C. and Baumann, E.R., "The Electrolytic Respirometer—
    I Factors Affecting Oxygen Uptake Measurements," Water Res.,
    10, 1976, 1031-1040.
ACKNOWLEDGEMENTS
   The authors wish to acknowledge the contribution of their col-
league,  Lucy  S.  Bottomley,  P.E.,  Environmental  Engineer,
Public  Works Department,  at  Naval  Air Engineering  Center,
Lakehurst, New Jersey. At O.H. Materials Co., we specifically
thank Robert H. Panning and Robert J. Ohneck for their support
throughout these projects.
                                                                                                       ON-SITE TREATMENT    119

-------
        Physical/Chemical  Removal  of Organic  Micropollutants
            from  RO Concentrated Contaminated  Groundwater
                                                  L. Simovic
                                            Environment Canada
                                     Environmental Protection Service
                                       Wastewater Technology Centre
                                             Burlington, Ontario
                                                  J.P. Jones
                                          University of Sherbrooke
                                             Sherbrooke, Quebec
                                               I.C. McClymont
                                              ICAN Consultants
                                             Burlington, Ontario
ABSTRACT
  The reverse osmosis (RO) concentrate of contaminated ground-
water from the Gloucester Landfill site near Ottawa, Ontario, was
treated by air stripping (AS), ozonation (O3) and ultraviolet (UV)
irradiation for the removal of organic micropollutants.
  Fourteen compounds, predominantly volatiles, were monitored
during this 6-week study. The concentration ranges of some of the
compounds in the reverse osmosis concentrate were:  trichloro-
ethylene  2-280 /ig/1;  benzene  28-1,350  pg/1; chloroform
800-40,100  pg/l;  diethyl ether  136-2,300  /ig/1  and acetone
750-36,000 jig/1.
  Based on the results of this study,  the air stripping process
alone is considered to be adequate for treating the RO-concen-
trated contaminated groundwater at the Gloucester Landfill site.
Residual concentrations of four of the six compounds with pro-
posed groundwater quality  objectives met  or exceeded those
levels.
  The removal efficiency achieved by all three  unit processes
combined, however, was better than that obtained by AS alone.
For the degree of additional improvement that might be achieved,
it is doubtful that the addition of Os/UV to the AS process would
be economically justified.

INTRODUCTION
  From 1969 to 1980, federal government laboratories in Ottawa,
Ontario, disposed of their  hazardous  wastes (mainly organic
solvents) in a Special Waste Compound (SWC) at the nearby
Gloucester Municipal Landfill.1 In  1981 contaminants including
diethyl ether,  acetone  and some  chlorinated  solvents  were
detected in the groundwater migrating from the landfill site. This
contamination  was  a potential danger to the drinking water
source of nearby residents. Subsequently, Environment Canada
and Transport Canada undertook a number of studies to define
the extent  of the problem  and identify possible  remedial
measures.'
  In 1984 Environment Canada's Wastewater Technology Centre
(WTC),  in cooperation with the  Environmental Emergencies
Technology Division, carried out a pilot scale study on the
removal of toxic organics from the contaminated groundwater
within the SWC. To define one remedial action, the objectives of
this study were: (1) to evaluate the effectiveness of selected reverse
osmosis (RO) membranes for removing toxic organics from the
contaminated groundwater and (2) to evaluate the effectiveness of
air  stripping,  ozonation  and ultraviolet  (UV) irradiation for
treating the concentrate from the RO system.
  The results related to the RO  tests  have been  presented
elsewhere.' This paper presents the results for the removal of
organics from the RO concentrate by air stripping, ozonation and
ultraviolet irradiation. For each treatment process, the results are
discussed with  reference to specific toxic organic  compounds
which were present in the groundwater at concentrations higher
than the proposed groundwater quality objectives.4

METHODS AND MATERIALS
  A trailer facility outfitted for this project was set up at the
Gloucester site beside test well 36 W. This well, located within the
SWC, was found to have the  highest concentrations of organic
micropollutants.' A schematic  of the pilot plant used during this
study is shown in Figure 1.
  The air stripping unit in the pilot plant consisted of two 1.22 m
by 10 cm (I.D.) plexiglass columns connected in series. The col-
umns were packed with 1.3  cm  Intalox*  saddles.  A  Whispair
1707Jt blower was used to force air through the stripping columns.
  The ozonation experiments were conducted using a bubble col-
umn 91 cm by 10 cm (I.D.). Ozone was generated from oxygen us-
ing a Welsbach T-816J ozonator. The Oj concentrations in the inlet
gas  and  the off-gas were measured using a standard iodometric
method.' Ozone in the liquid was measured using the indigo
method.'
  The UV irradiation chamber  used was a Trojan System 2000.**
It consisted of four 9 W low pressure mercury lamps. The total UV
output of all the lamps in the reactor was determined by ferriox-
alate actinometry'and was found to be 4.1  x 10 ~4 einsteins/min.
The flow of liquid through the reactor was maintained during the
tests to give a theoretical retention time (RT) of 22 min.

Air Stripping
  During air  stripping experiments, the following three variables
were studied: air flow rate, the packing height and the water flow
rate. A full 23 factorial experimental design was used to allow for
the determination of the relative significance of the effect of these
variables (Table 1). The air and water flow rate levels were chosen
to bracket a  wide range below the conditions at which the col-
umns would become flooded.  From the calculated stripping fac-
tor  for the volatile compounds present in the concentrate, the
values for the  packing height were chosen to give a  sufficient
number of transfer units for ?  99% removal of these compounds.

• Norton Chemical Company, Akron, OH
t Roots Blower Operation, Dresser Industries, Inc., Cornerville, IN
t Polymetrics Inc., Sunnyvale, CA
"Trojan Technologies Inc., London, Ont.
120    ON-SITE TREATMENT

-------








AIR
STRIPPING
COLUMN


+
RESERVOIR



EXHAUST



^




' t

PUMP

           00
       CONCENTRATED
       GROUNCVATER
                                                           Figure 1
                                        Process Schematic for Groundwater Treatment System
                           Table 1
                 Air Stripping Process Variable
                                         Levels
Variables
Air flow rate (1/min)
Packing height (m)
Water flow rate (1/min)
15
1.22
0.5
300
2.44
2.0
Ozonation/Ultraviolet Irradiation
  The effects of ozonation, UV irradiation,  air stripping and
water flow rates on the removal of organic micropollutants were
observed using a full 24 factorial experimental design (Table 2).

                           Table 2
                AS/Os/UV Experiment Variables
                                         Levels
Variables
Air stripping
Os dose (mgO3/l of gas)
UV irradiation
(einsteins/min)
Water flow rate (1/min)
OFF
0

0
0.5
ON*
65

4.1 x 10-4
2.0
•Air now rate = 300 1/min; packing height 1.22 m
Sampling and Methods of Analysis
  The locations where samples of pilot plant influent and treated
effluent were collected during the tests are shown in Figure 1. For
each experiment, two sets of samples were collected: one set was
collected at the beginning of the test and another set after 2 hr of
continuous operation of the unit process(es) being tested.  After
collection,  the samples  were stored at  4°C in glass vials with
teflon-lined septums. Sample analyses were completed within 10
days of collection of a given set of samples.
  The organic compounds, predominantly volatiles, were analyzed
by the purge and trap  method using a  Hewlett Packard model
5830A Gas Chromatograph coupled to a Flame lonization Detec-
tor (FID). The following pairs of compounds coeluted during the
analysis  and could not  be  quantified  individually:  (1) 1,1-di-
chloroethane (RT  = 13.6) and tetrahydrofuran (THF) (RT =
13.7;  (2) 1,1,1-trichloroethane (RT = 17.4) and 1,4-dioxane (RT
= 17.4).

Water Characteristics
  During this 6-week study the concentrations of various com-
pounds  in  the  RO concentrate varied  widely. These  data  are
presented in  Table 3, together with the proposed groundwater
quality objectives for these compounds.

STUDY RESULTS
  Forty-seven tests were conducted during this study.  Of these
tests,  27 used AS, 7 used O3 and 3 used UV alone to remove the
organic micropollutants  from the concentrate. In 11 tests, O3  was
                                                                                                  ON-S1TE TREATMENT     121

-------
                           Table 3
 The Influent Concentration for Some of the Organic Mlcropollulanls
Organic Contaminant

Dichloromethane
Acetone
1,1 -Dichloroethylene
Chloroform
Diethyl Ether
1,2-Dichloroethane
Trichloroethylene
Benzene

NA — Nol Available
           Groundwaler
RO Concen- Quality Objectives
Irate 0*g/l)  (/'g/l)
  9-  1,670
750-36,000
  2-  1,531
800-40,100
136-  2,300
 32-   653
  2-   280
 28-  1,350
150
NA
  0.3
 30
NA
 10
 30
 10
 followed by UV irradiation. Air stripping was followed by UV in
 2 tests, by Oj without UV in 5 tests and by O3 in the presence of
 UV in 8 tests. Not all of the samples collected during these tests
 could  be analyzed for all the parameters, so some of the data
 points required to complete the factorial analysis of the data are
 missing.  A  stepwise  regression  using backwards  elimination
 routine was, therefore, applied for the analysis of the data.
   Although the concentration of organic compounds in the RO
 concentrate varied greatly (see Table 3), the removal efficiency of
 the various processes tested was not affected by this variability in
 the concentration. Variations in the removal efficiencies were due
 to the experimental conditions applied.

 Air Stripping
   Table 4 summarizes the results  for air stripping experiments.

                           Table 4
       Percentage Removal and Residual Concentration Ranges
          Achieved During AS Treatment of  Groundwaler



Contaminant
Dichloromethane
Acetone
1,1 -Dichloroethylene
1 , 1 -Dichloroethane/THF
Chloroform
Diethyl Ether
1,2-Dichloromethane
1,1,1 -Trichloroethane/
1 ,4-Dioxane
Trichloroethylene
Benzene
Range of
Removal
Efficiency
CVo)
57- > 99
32- 99
43->99
25- 99
31- >99
25- >99
24- 99

2->99
41->99
25->99

Residual
Concentration
Oig/D
2.3- 38.2
67.4-9,340.4
0.5- 130.2
6.4- 667.9
33.6-2,775.7
0.7-1,294.0
1.6- 228.6

12.5- 813.7
0.5- 63.4
1.2- 78.2
  In addition to the chemicals listed in Table 4, the removal effi-
ciencies of  four other  compounds,  listed in  Table 5, were
estimated. The estimates were based on the percent reduction in
the peak  areas for  the compounds  observed during the  gas
chromatographic analysis of the RO concentrate and the air strip-
ping samples.
  Analysis of  the data from the  AS experiments showed that
water flow rate had  the greatest effect on the removal of most
compounds. The best removals  were obtained in  the experiments
where the water flow rate was low (0.5 1/min). Packing height and
air flow rate did not have a substantial effect on the efficiency of
removal within the range of values used in these experiments.
                        Under the best conditions (water flow rate of 0.5 1/min), four
                      of the compounds  were removed to below  their groundwater
                      quality objective concentrations (see Table 6).

                                                Table 5
                            Percent Reduction Range Estimated from the Peak Areas
Chemical

Dichloropropane
Bromoform
Toluene
Chlorobenzene
Estimated
Removal 0
21->99
15-799
 7-  98
 5- >99
                                                               Table 6
                                            Effluent Concentrations Achieved with Air Stripping



Compounds
Dichloromethane
1 ,2-Dichloroethane
Trichloroethylene
Benzene
Ground water
Quality
Objectives
Oig/l)
150
10
30
10

Residual
Concentration
99%,  having residual concentrations of 67.4  /tg/1
                                     and 0.7 /»g/l, respectively. Since there are no proposed ground-
                                     water quality  objectives  for  these compounds, it is not clear
                                     whether the residuals obtained through AS  alone would be con-
                                     sidered acceptable. If they  were not acceptable,  then AS would
                                     have to be augmented by some other process(es).

                                     Ozonation
                                       Results from direct ozonation of the RO concentrate are shown
                                     in Table 7.
                                       The  removal efficiencies for  ozonation alone were generally
                                     either similar to or lower than those achieved by AS alone.  The
                                     reason for lower  removal during ozonation  was probably due to
                                     the lower gas flow rate through the column. In AS, the gas flow
                                     rate was in the ranges of 15 to  300 1/min; for ozonation, the max-
                                     imum gas flow rate was 2 1/min.
                                       In one  of the tests using  Oj alone, the ozonation column was
                                     operated  at a gas  flow rate of 2 1/min but the ozone generator was
                                     turned off. From the results of this test, it appears that approx-
                                     imately 30%  of the removal observed was due to the stripping ef-
                                     fect of the gas and 70% was due to the chemical oxidation of the
                                     organics.
                                       With ozonation of air-stripped  RO concentrate (Table 7), the
                                     removal of the organic micropollutants generally improved,  par-
                                     ticularly for acetone and diethyl ether. It is not clear whether this
                                     additional removal could be sufficient to justify the use of ozona-
                                     tion in conjunction with air stripping.

                                     Ultraviolet Irradiation
                                       Ultraviolet irradiation was  used alone, as well as in combina-
                                     tion with  O3  or AS. For most compounds, the removal was not as
 122    ON-SITE TREATMENT

-------
good as that achieved by AS alone. For this reason, the data from
these tests are not reported. The  reader is referred to an earlier
data report.8 However, since the best removal efficiency for most
of the organic compounds was achieved when all three unit pro-
cesses were applied, the removals and residuals from these tests
are presented in Table 8.

                           Table 7
   Average Removals of Organic Chemicals from Groundwater after
              Ozonation and Air Stripping/Ozonation
Contaminant
Dichloromethane
Acetone
1 , 1-Dichloroethylene
1 , 1-Dichloroethane/THF
Chloroform
Diethyl Ether
1 ,2-Dichloroethane
1 , 1 , 1-Trichloroethane/
1 ,4-Dioxane
Dichloropropane (1,2- or 1,3-)
Trichloroethylene
Benzene
Bromoform
Toluene
Chlorobenzene
Removal
After AS
(%)
93
57
73
61
90
84
86

69
57
92
91
83
77
60
Removal
After Oa
W
37
28
56
67
33
74
39

53
99
93
73
82
53
64
Removal
After AS/
03 (%)
84
64
76
83
70
95
76

81
100
99
99
95
78
83
                           Table 8
      Percentage Removal and Residual Concentration Ranges
      Achieved During AS/Os/UV Treatment of Groundwater
Contaminant
Dichloromethane
Acetone
1 , 1-Dichloroethylene
Chloroform
Diethyl Ether
1,2-Dichloroethane
Trichloroethylene
Benzene
Range of
Removal
Efficiency (%)
89- 99
38- 95
29->99
87- > 99
84- > 99
63- > 99
99- > 99
85- > 99
Range of Residual
Concentrations
Oig/i)
7.1-45.5
36.5-1,933.8
0.5-45.4
6.2-2,020.0
0.7-95.0
1.6-56.5
0.5-3.5
0.3-13.5
  Generally, the range of removal efficiencies obtained by the
combination, AS/O3/UV, was higher than AS alone (Table 4).
The residuals were either similar or lower than the ones obtained
by AS alone.
  Based on these results, AS alone appears to be adequate to
remove volatile  organic compounds from the RO concentrate.
However, all three unit processes combined provided marginally
better removal efficiencies. It is doubtful, though, that the degree
of improvement would justify the increase in cost that would oc-
cur by including O3 and AS units in the treatment train.

CONCLUSIONS
  The conclusions drawn from this study follow below.
  Two pairs of compounds (1,1-dichloroethane/tetrahydrofuran
and  l,l,l-trichloroethane/l,4-dioxane)  coeluted during  the gas
chromatography analysis and could not be quantified individually.
  The results of data analysis by stepwise regression showed that
during air stripping experiments  the water flow rate had the
greatest effect on the removal of most organic compounds. The
best  removals were obtained at a water flow rate of 0.5 1/min.
  During  air  stripping  experiments,  four  compounds
(dichloromethane,  1,2-dichloroethane,  trichloroethylene and
benzene) were removed to below the proposed groundwater quali-
ty objectives.
  In the absence of groundwater quality objectives for acetone
and diethyl ether, it is  not clear whether the effluent concentra-
tions obtained by air stripping (67.4 and 0.7 /xg/1) would be con-
sidered as acceptable residual values. If they were not acceptable,
O3/UV in conjunction  with air stripping could be considered.
  Removal efficiencies observed after O3 or UV irradiation alone
were generally either similar to or lower than the results obtained
by AS alone. Combinations of AS/O3, AS/UV and O3/UV unit
processes compared to AS alone did not  offer significant im-
provements in the removal efficiencies in most of the tests.
  The removal  efficiency achieved by  all three  unit  processes
combined was better than that obtained by AS alone. For the
degree of improvement that can be achieved, it is not  clear that
the addition of O3/UV to the AS  process would be economically
justified.
  The data obtained in this study can be used to predict process
performance for those situations where the contaminated ground-
water has large  variations in the micropollutants concentration
range.

REFERENCES
1. Jackson, R.E., Patterson, R.J., Graham, B.W., Bahr, J., Belanger,
   D., Lockwood, J. and Priddle, M.,  "Contaminant Hydrogeology of
   Toxic Organic  Chemicals at a Disposal Site, Gloucester,  Ontario,"
   NHRI Paper No. 23, IWD Scientific Series No. 141, Ottawa,  Ont.,
   5985.
2. GEC, WESA, Canviro, "Gloucester Landfill Waste Site Engineering
   Study—Problem Definition and Remedial Alternatives,"  submitted
   to Transport Canada as project #  L10137-1174 by A.J. Graham En-
   gineering Consultants Ltd., Water and Earth Science Associated Ltd.,
   and Canviro Consultants Ltd., Sept. 1984.
3. Whittaker, H., Adams, C.I., Salo, S.A. and Morgan, A., "Reverse
   Osmosis at the Gloucester Landfill," Proc. of the Technical Seminar
   on Chemical Spills, Environment Canada, Toronto, Ont.,  Feb. 1985.
4. Canviro Consultants Ltd., "Treatment of Organic Contaminants in
   Landfill Leachate:  Final Report,"  submitted  to Wastewater Technol-
   ogy Centre, EPS, Burlington, Ont.,  May  1984.
5. APHA, AWWA, WPCF, Standard Methods for the Examination of
   Water and Waste-waters, New York, NY, 15th ed., 1980.
6. Bader, H.  and Hoigne, J., "Determination of Ozone in Water by the
   Indigo Method; A Submitted Standard  Method," Ozone: Science
   andEng.,  4,  1982, 169-176.
7. Hatchard, C.G. and Parker, C.A., "A New Sensitive Chemical Acti-
   nometer II.  Potassium Ferrioxalate as a Standard Chemical Acti-
   nometer," Proc. of the Royal Soc. of London, Series A.,  235,  1956,
   518.
8. Simovic, L. and Jones, J.P., "Removal of Volatile Organic from RO
   Concentrated Contaminated  Groundwater—Gloucester Site," Un-
   published manuscript, Environment Canada,  Wastewater Technology
   Centre, Burlington, Ont., 1985.
                                                                                                   ON-SITE TREATMENT    123

-------
             State-of-the-Art Technologies of Removal,  Isolation
          and  Alteration  of  Organic  Contaminants Underground
                                                 Walter W. Loo
                                                George N. Butter
                                      McKesson Environmental Services
                                             Pleasanton, California
ABSTRACT
  The rising cost and  increasing long-term liability of off-site
treatment and disposal  have caused a demand for on-site in situ
treatment of organic contaminants located underground. There
are generally three types of in situ treatment: removal, isolation
and alteration.
  The purpose of this paper is to provide the reader with a de-
scription of available state-of-the-art, in situ treatment alterna-
tives for the treatment  of various organic contaminants in  both
the vadose and saturated zones in the subsurface. Guided format
tables are developed for the reader to search for the application
that may work in an individual situation while considering regu-
latory, economic, risk and liability consequences.

INTRODUCTION
  The treatment and disposal of hazardous waste and material
off-site has become increasingly difficult due to closure of an al-
ready limited number of qualified waste handling and disposal
facilities, tightening of regulatory standards, escalating transpor-
tation and disposal costs and risks of long-term liability. While
the technology of in situ treatment is advancing at a rapid pace,
there is a need to evaluate periodically the applicability of various
technologies to specific  on-site  contamination problems and
underground conditions.
  The purpose  of this  paper is to provide the reader with a de-
scription of available in situ treatment alternatives for applica-
tion on  various organic contaminants in both the vadose and
saturated zones  in the subsurface. However, the best technology
will work only if it is in agreement with regulatory remediation
standards and permits, economic feasibility and  long-term  risks
and liability as shown in Figure 1.
  The following sections discuss the three possible types of in situ
treatment alternatives:  removal, isolation and alteration. These
alternatives will encompass a broad combined spectrum of science
and engineering areas:
  Chemistry
  Physics
  Geology
  Geohydrology
  Biology
  Environmental Engineering
IN SITU REMOVAL
  An  initial requirement for on-site remediation always  is the
removal of the source and control of the contaminated plume,
whether it is in the soil  (vadose zone)  or in the groundwater
(saturated  zone) or both. The in situ removal process may in-
volve primary, secondary and tertiary methods.
                         In-iltu
                        Treatment
                        Technical
                      Alternatives
     Regula tory
     Remedia tion
     Standards &
      Permits
                      Environaenlal
                    Risks & Liability
                     Considerations
 Economic
Feasibility
                        Figure I
  The primary methods include the following:
• Vadose zone—air vacuum venting
• Saturated zone—pumping of recovery well

  The air vacuum venting works well with volatile organic com-
pounds located underground,  in particular, permeable strata in
the vadose zone where the soil  can serve as a migration pathway.
In general, the venting process  will require an air emission permit
which may be prohibitive in some areas. If organic emissions to
the atmosphere are not allowed, an  oil stripper (OS) or an acti-
vated carbon adsorption unit  is required. If the vent stream is
steady and  loaded with high concentration volatile organics, it
may be advisable to recover the product by either an oil strip/de-
sorption or a regenerative activated carbon process (Fig. 2).
  The secondary method of removal involves both injection and
recovery wells. Usually, injection wells for the vadose zone should
be discouraged  because  the  injected  air/water  stream  may
"push" the volatile organics  toward  undesirable off-site  loca-
tions.
  The tertiary method of removal involves temperature enhance-
ment and addition of surfactants to  increase the release rate and
mobility of the organic compound movement through the porous
media toward ultimate recovery. The addition of surfactants may
create new  problems in recovery and surface treatment if not
properly applied. Temperature enhancement is generally a desir-
124   ON-SITE TREATMfiHT

-------
   Recovery
    Phase
Emission
 Phase
                   Atmosphere
Capture/Disposal
    Phase
Recycling
  Phase
                          Figure 2
         Generic Flow Diagram of In Situ Removal Process
able process (if duration of treatment can be prolonged) by a
passive solar heating system. The system works as underground
space heating under steady circulation. The system cost and main-
tenance are usually very attractive.
  For the saturated zone, the application is similar to the vadose
zone with the exception of depth of the contaminated zone and
liquid handling.         \
  The following design criteria are important for  recovery well
field applications:

• The recovery rate must exceed the recharge/safe yield of the
  geohydrologic system
• Drilling and well installation design must now allow vertical
  migration of contamination from shallow depth
• Pumping of recovery wells must begin with shallow wells and
  move to deeper wells to avoid pressure induced vertical migra-
  tion of contamination
• It is not advisable to establish pumping wells in low perme-
  ability material (generally less than 1 X 10~4cm/sec)
• Well screen design must be designed specifically for "sinkers"
  or "floaters" organic compounds
  Unfortunately there  will never be 100% removal of the con-
taminants. Some portion of the contamination will remain affixed
to the fine grain matrix in the subsurface material.

IN SITU ISOLATION
  There  are  specific geohydrologic conditions which  will not
allow in situ removal treatment. Also, less volatile  organic com-
pounds may not be easily removed by even secondary or tertiary
treatment processes. The utilization of hydraulic isolation may be
the solution in these situations. Hydraulic isolation can be used as
a follow-up to in situ treatment after removal of a major portion
of the contamination.
  There are three different  in situ hydraulic isolation methods:
surface cover, subsurface barriers and sorption processes.
  The surface cover method generally involves placement of a
clay or concrete cover  over a defined contaminated area to pre-
vent percolation of surface water into the contaminated zone to
contact the contaminants below the surface.
  Subsurface barriers include a number of isolation techniques:

  Slurry trench walls
  French drain
  Sheet pile
  Interceptor ditch
  Curtain grouting
  The first four types of hydraulic barriers are best for hydraulic
isolation of low permeability media and shallow depths (20 to 40
ft maximum). Curtain grouting involves injection of impervious
material into permeable strata to block off potential  migration
pathways in the subsurface environment.
  Sorption in situ treatment is a relatively  new technology in-
volving  replacement of the permeable strata in a known contam-
inated area with a specially formulated fluid with both absorption
and adsorption properties. In the absorption process, the material
will expand up to 30 times its original volume while absorbing the
contaminant and blocking off or filling in the pore spaces. Due to
the microscopic size of the material, the large surface area of con-
tact with contaminants will make an effective sealant to the per-
meable  pathways when  contaminants are  released.  The formu-
lated fluid  will  work  for both polar and non-polar organic com-
pounds. Since this treatment  process is still  in the experimental
stage, it may not be very cost-effective at the present time.
  A shallow interceptor  ditch may work in relatively low perme-
able saturated material. The ditch may act and perform like a line
of hundreds  of wells. Sometimes a combination of the in situ
hydraulic isolation techniques may work for sites with multiple
organic  compounds of different physical and chemical properties
when other treatment processes do not provide a total solution.

IN SITU ALTERATION
  In situ alteration treatment involves the breakdown of haz-
ardous  organic compounds into  harmless compounds. It has
been  demonstrated  that  biodegradation  of  non-halogenated
organic  compounds to harmless compounds can be effected in
surface  aerobic treatment of sludge and waste water. The same
result can be  attained in  subsurface conditions which control  the
growth  of aerobic bacteria. The parameters controlling bacterial
growth and metabolism are:
  Pressure
  Temperature
  pH
  Salinity
  Dissolved oxygen
  Nutrients
  Biotoxins
  Radiation
  Mixing

  The transport mechanism in the vadose and saturated zones is
the major problem that hinders subsurface treatment. If the con-
taminated fluid can be pumped to the surface, it is better to per-
form surface  treatment.  The application of the in situ alteration
process  should not be a primary means of treatment.
  The in situ alteration treatment of halogenated hydrocarbons is
still in the experimental stage.  The  recent advances are well docu-
mented  in the references  found at the end of this paper. The pro-
cess of breakdown of halogenated hydrocarbons depends on the
successful control of methanogenic conditions (addition of meth-
ane mixed  with air to the contaminated soil and  water).  When
aerobic  bacteria are exposed to methanogenic conditions, selec-
tive species  will degrade the halogenated organic compounds into
carbon  dioxide and  chlorides which are harmless compounds.
While the researchers are developing a better understanding of the
mechanism of degradation, forseeable problems of field applica-
tion are  as follows:
• Aerobic bacteria only occur in shallow depths
• A specially formulated carrier fluid must be coupled with meth-
  ane and other nutrient when introduced underground
• Introduction of methane gas  underground may create haz-
  ardous conditions (i.e., explosion) if not properly contained
                                                                                                ON-SITE TREATMENT    125

-------
                            Table 1
     Practical Application of In Situ Treatment of Volatile Organic*
                            Table 2
    Practical Application of In Situ Treatment of Acid/Base Neutral
               Extraclables, Pesticides and Herbicides
In- si tu Trea traenl
Techniques
Remova 1


• Flushing
tempe ra ture

Isolation

• Subsur face
barriers
• SorplLon
Alt a tto
. Aerobic Process

NA Not applicable
(a) Air stripper
(b) Oil stripper
Vadosc Zone
Less


difficult
NA
ef f ec tlve
Yes
Yes
NA



(b) or (c)
Yes, need (a)
(b) or (c)
(b) or (c)
Yei
Yes
Vol. n««d Id)
Saturated Zone
Leu

Yes, but
difficult
NA

Yet
NA

Tachn Ique a
Permeable
Renova 1
Yti , need * Pu» ptng/e vacua 1 1 on
{a) (b) or (c) to create sink
Yet. need • Flushing
(a) (b) or (c)
(a) (b) or (c) temperature
enh«nc*aen t
I tola I i on
Yt i • Surface cover
bo r r I e r t
Yet, need (d) • Sorptlon

A 1 tera t Ion




( • ) Oil s tripper
Vadote Zone
Lea*

NA
NA
NA
Yei
Yet
VA


NA
N««d (c)
Long dura t Ion
Yes
Yet
Yet, r.«ed (c)
Saturated Zone
Ul«

NA
NA
NA
Yet
NA
Pcraeablc
Need (a) or
(b)
Need («) or
(b) and (c)
Lont duration
Yea
Yea, can be
expcnilva

£«(,er .o»jntal for halo-enaled organic co« pound*. Need



(c) Detail geohydrologlc definition
   The understanding of the anaerobic degradation of halogen-
 ated organic compounds is in its infancy, but it has great poten-
 tial. Because of the low  cost of bacteriological processes, it is
 worthwhile to have engineers perform opportunistic  trial and
 error experiments on both laboratory bench and field pilot tests
 using this technique. Once demonstrated,  the cost saving over
 other in situ treatment alternatives may be substantial.
   At the present state of knowledge of biodegradation, there is
 immense room for improvement  and development of original re-
 search and application.  Hopefully, the economic  advantage  of
 innovative in situ technology will push the  alteration techniques
 to a major breakthrough in the near future.

 PRACTICAL APPLICATIONS
   The above mentioned in situ treatment techniques may work
 individually or in combination for  specific  contamination cases.
 However, there  is no general formula to  produce an optimal
 remediation solution for a case-specific problem.
   An attempt is made to provide  a general guidance  for prac-
 tical application of the specific in  situ treatment techniques on
 organic compounds in the vadose and saturated  zones  in Tables
 1 and 2. These tables will not provide a total solution for a spe-
 cific site. However, they will provide an individual with an oppor-
tunity to evaluate the "cans," "cannots" and effectiveness of the
full spectrum of the available alternatives.
  These in situ treatment techniques are only one aspect of the
total remediation plan. The engineer will  still have to consider
regulatory standards and permits, economic feasibility and long-
term risks and liability for a total remediation program.
REFERENCES

1.  Bouwer, E.J. and McCarty, P.L.. "Transformation of 1- and 2- Car-
   bon Halogenated Aliphatic Organic Compounds Under  Methano-
   gcnic Conditions," Appl. and Environ. Microbiol., 45, 1983, 1286-
   1294.
2.  Haber, C.L. el al.,  "Methylotrophic Bacteria: Biochemical Diver-
   sity and Genetics," Science 221, 1983, 1147-1153.
3.  Parsons, F., el al., "Transformations of Tetrachloroethene and Tri-
   chloroelhene  in  Microcosms  and  Groundwater," JAWWA,  76,
   1984,56-59.
4.  Wilson, J.T. and McNabb, J.F.. "Biological Transformation of Or-
   ganic Pollutants in Groundwater," American Geophysical Union EOS
   64, 1983,505-506.
5.  Wilson, J.T. and Wilson, B.H., "Biotransformation of Trichloro-
   ethylene in Soil," Appl. and Environ. Microbiol.. 49, 1983, 242-243.
126    ON-SITE TREATMENT

-------
               Assessment  of  Volatile Organic Emissions from a
                       Petroleum Refinery Land Treatment  Site

                                         Robert G.  Wetherold, Ph.D.
                                                 Bart M. Eklund
                                              Radian Corporation
                                                  Austin, Texas
                                          Benjamin  L. Blaney, Ph.D.
                                   U.S. Environmental Protection Agency
                          Hazardous Waste Environmental Research Laboratory
                                                Cincinnati,  Ohio
                                              Susan  A. Thorneloe
                                   U.S. Environmental Protection Agency
                               Office of Air Quality Planning and Standards
                                           Durham, North Carolina
ABSTRACT
  A field assessment was performed to measure the emissions of
volatile organics from a petroleum refinery land  treatment site.
As part of this study, the emissions of total volatile organics from
surface-applied and subsurface-injected oily sludge were meas-
ured over a 5-week period. The effect of soil tilling on the emis-
sions also was monitored.
  Volatile organics emission rates were measured using the emis-
sion isolation flex chamber method. Soil samples were collected
during the test periods to determine soil properties, oil levels and
microbe count. Soil surface and  ambient temperatures, both in-
side and outside the flux  chambers, were measured throughout
the test periods.

INTRODUCTION
  The U.S. EPA currently is developing background information
on air pollutants at hazardous waste treatment, storage and dis-
posal facilities (TSDFs) to provide supporting data for  setting
standards as necessary under Section 3004 of RCRA as amended
in 1984. This regulatory development  is the responsibility of the
Office of Air  Quality Planning  and  Standards (OAQPS). The
U.S.  EPA Hazardous Waste Environmental Research Labora-
tory-Cincinnati (HWERL) has the responsibility of providing
technical support to OAQPS in the area of atmospheric emis-
sions determination from hazardous waste management.
  In order to assess emissions from hazardous waste TSDFs, a
field study was performed  at a land treatment facility located at a
large West Coast petroleum refinery. This study was intended to:
(1) measure the total volatile organics emission rates from sur-
face- and subsurface-applied oily refinery sludge and (2) to de-
termine how emissions of volatile organics  varied  with time.
Based on these measurements the total emissions over a 5-week
period were estimated.
  Little work has been performed to characterize volatile organic
emissions from large full-scale land treatment sites. The study
approach and selected  results  are summarized in this paper. A
more detailed  description of the work and its limitations in in-
cluded  in a previous U.S. EPA  report.1 The emissions of indi-
vidual  measured  compounds  are provided  in that  U.S. EPA
report.
DESCRIPTION OF THE TEST SITE
  The test site was located at a large West Coast refinery. The
land treatment area covers approximately 42,000 m2 (10 acres).
Three test plots, each approximately 420 m2  (0.1  acre)  in area,
were located side-by-side at one corner of the facility. The area
containing the test plot has been in land treatment service for
several years.
  During normal operation,  sludge from  refinery operations is
collected and  stored in a fixed-roof tank  at the land treatment
area. The sludge is injected subsurface using an 8,0001 (50-barrel)
capacity injector  vehicle equipped with  four separate  injector
tines. The injector vehicle  consists of a tractor and an 8,000 1
(50-barrel) capacity tank trailer. The tines are attached to the tank
trailer. The waste is injected  15 to 28 cm below the surface and
immediately disked by a disk unit pulled behind the injector ve-
hicle. The area is cultivated with a disk and/or rototiller several
times per week until the next sludge application.
  According to the operators of this facility, the annual  applica-
tion rate at this site is about 5.4 - 9.1 x 106 kg/year, and the in-
dividual application rates are typically about 16 1/m2 (400 bar-
rels/acre). The sludge typically consists of 50  to 75% DAF/API
separator float, 20 to 30% API separator sludge and about 5%
miscellaneous oily waste. The soil is  irrigated periodically to
maintain proper soil moisture content.

EXPERIMENTAL APPROACH
  The  experimental  design,  sampling  techniques,  analytical
methods and test procedures are summarized below.

Experimental Design
  The experimental design of this program was developed to ob-
tain data of sufficient accuracy to fulfill the major objectives of
the study. The primary emphasis of the program was on obtaining
volatile organics emission rate measurements. The test design was
a synthesis of two sampling strategies  for obtaining emission
measurements: totally randomized sampling over  the test plots
and semi-continuous sampling at a single location in each plot.
  To provide an overall estimate of the volatile organics emission
rate for a total test plot area,  a number of measurements at ran-
domly selected points  must be performed. One method of ac-
                                                                                          ON-SITE TREATMENT    127

-------
complishing this is to divide the area into discrete segments and
then randomly select segments to be tested. On the other hand,
the best strategy for examining the change in emission rate with
time is the repeated sampling of a limited number of points over a
period of time.
Three test plots were used in this study. Sludge was surface-
applied to one plot (Plot A) and subsurface-injected at another
plot (Plot C). In between these two plots was a third plot (Plot B)
with no sludge applied. Plot B served as a baseline or control plot.
Each plot was approximately 420 m2 (0. 1 acre) in size, and each
was divided into 21 segments, as shown in Figure 1 , to provide for
randomized sampling.
Subturlac* Background Surlact
Application Plot Plol Application Plot
Cl
C«
C7
CIO
C1J
cie
C1I
a
CS
c<
C11
Cl«
C17
C20
Cl
C*
O
C12
CIS
C1I
C2I
SI
84
B7
BIO
B13
Bit
B1I
Bl
BS
se
Bit
B14
Bir
BJO
U
M
B9
B12


811
B21
A1
A4
A7
A10


A16
Alt
A]
AJ
At
An


AI7
A»
A)
At
At
A12


Alt
A2I
1
(
t
»
Table 1
Typical Dally Sampling Schedule
Flux Charter
frMUno
ApprmlMt* Grid Cm F/C £MU?IM Cnll«rfpd
"TIB* Point* 10 ID" Canlitw Soil Syrlng* Couwnt*
TEXT PLOT A DATE, 11/09/84 APPL 1 CAT ION METHOD i SURFACE
OBOO 	 002 003 - - X OC
0843 A07e 002 003 - X X
0930 603 002 003 X XX
1013 A03 002 003 - - X
1100 AI9C 002 003 X X X
IMJ AOI 002 003 X
1230 	 002 003 - - - TIM
r 1400 A07e 002 003 X - X
1449 Mr 002 003 X - X
1330 B14 002 003 X X
TEST PLOT A DATEi 11/09/84 APPLICATION METHOD! SUBSURFACE
0600 	 001 002 - - X OC
0845 C1IC 001 002 X X X
0930 CI3 001 002 X
1015 CI4 001 002 - - X
5 1100 CIS* 001 002 - - XX
E 1145 BIS 001 002 - - . X
: 1230 	 001 002 - - - Till
1315 C13e 001 002 X X
1400 Cll' 001 001 X
1449 C04 001 002 - X X
1530 C06 001 002 X X X
1615 BOS 001 002 X
i Prefix A. B, C refers lo plot A (surface applied). pk>l B (background or control plot), and
Eacn Plol . O3 lq m.l.fl (4.450 K). II.)
Eacn Grid > 20.1 M m.l«r« (217 M IIJ
b. F/C « flux chamber
C. Control points
                           Figure 1
       Test Plot Grid Assignment for Assess; ;ent of Emissions
             from Land Treatment of Petroleum Wastes

   Sampling schedules for each  plot  were developed for each
 separate day of sampling. Each sampling day was divided into
 two parts. During  each half day,  one  sampling crew  made
 measurements at the surface-applied plot, and the other crew per-
 formed similar measurements at  the subsurface injected plot.
 Each crew performed emission measurements  at  two  separate
 control points, at two randomly selected points  in their assigned
 plot and at one randomly selected point in the control plot. Ran-
 dom number tables were used to select the order of sampling, the
 crew used  on each plot and the equipment set used by each crew.
 The sampling order for composite soil samples, soil cores, gas
 samples and their associated duplicates within a given plot were
 also randomly  selected.  A  typical  daily  sampling schedule  is
 shown in Table 1.

 Sampling Procedures
   The main sampling technique employed at this site involved the
 direct measurement of emissions using the emission isolation flux
 chamber.  The enclosure emission measurement  approach has
 been used by others to measure emission fluxes of sulfur and
 volatile organic species.2-iA
   A diagram of the flux chamber is shown in Figure 2.  The flux
 chamber was placed on the emitting surface. Clean, dry air was
 passed through the chamber at a controlled  and measured rate.
 The concentration of the specie(s) of interest was measured at the
 outlet of the chamber. The emission rate of the measured specie(s)
 from the enclosed surface was then calculated from the flow rate
 and concentration in the gas.
                          Figure 2
      Cutaway Side View of Emission Isolation Flux Chamber
                  and Sampling Apparatus
   The flux chamber encloses a surface area of 0.13 m1 and has a
 volume of 0.031 m'. The sweep air  flow rate was 0.003 - 0.005
 mVmin. Air samples were collected from the outlet gas line using
 both  gas-tight syringes (for on-site  analyses) and evacuated
 stainless steel canisters (for off-site analyses).
   Liquid grab samples of the sludge and of the service water used
 to irrigate the soil were collected. Sludge from two tank trucks
 was transferred to the application vehicle (tractor) and applied to
 the soil. Each tank truck held two application tractor loads. After
 the first load was transferred from the tank  truck to the applica-
 tion tractor, 4-6 sample containers were filled with  sludge from
 the tractor. The samples were collected consecutively, while the
 tractor tank was full, from a gravity-fed line. While two tractor
 128    ON-SITE TREATMENT

-------
loads were required to empty each tank truck, samples were col-
lected after only one of the two transfers from each tank truck
load. One service water sample was collected from the  plant ser-
vice water line to determine the concentration of organics in this
water which the facility applies  as part of its  site  management
technique.
  Two types of solid samples were collected: soil composites and
undisturbed soil cores. Both types of samples were  collected im-
mediately adjacent to points where flux chamber measurements
were made. Soil composites were used in the determination of oil
and moisture contents. These samples were collected using a hand
auger and boring to a depth of 30 to 40 cm. The undisturbed soil
samples were used for determining the physical properties of the
soils. They were obtained by driving a brass sleeve approximately
20 cm into the ground, removing the sleeve and capping both
ends.
  Type K thermocouples were used to monitor the air and soil
temperatures both inside and outside each flux chamber. Two
separate thermocouples  were  used  to  measure  soil  surface
temperatures inside and outside the flux chamber. Each was plac-
ed in the ground to a depth of 1.3 to 1.9 cm.

Analytical Procedures
  The on-site analyses were limited to gas-phase analyses of the
air  samples collected in gas-tight syringes from the  outlet of the
flux chambers. A Byron Instruments Model 401 Total Hydrocar-
bon (THC) Analyzer was used to determine the concentrations of
total hydrocarbons (THC), CO2  and methane in the effluent air
samples from the flux chamber.
  The off-site analyses included the  chemical speciation of the
flux chamber air samples collected in stainless steel  canisters and
of the liquid sludge and water samples collected at the land treat-
ment site. The oil, moisture and microbe levels in the soil  also
were determined,  as  were the physical properties of the  soil
samples.
  Chemical speciation of the air  and liquid samples for C2 and
C10 hydrocarbons  was performed with  a Varian  3700  Gas
Chromatograph. The chemical species were cryogenically concen-
trated to increase the sensitivity of the analyses. Liquid samples
were analyzed  using a purge-and-trap technique  modified to
utilize the cryogenic trap. For both air and liquid samples, total
volatile organics concentrations were obtained as the sum of the
species concentrations.
  The oil,  water and solids contents of the sludge samples were
determined by solubilizing the oil and water in the waste with
tetrahydrofuran (THF). This liquid  phase was then  separated
from the insoluble solids. Residual solids were weighed, and the
amount of water in the THF was determined by Karl Fischer titra-
tion. The oil content was  calculated by difference. The oil and
grease content of the soil samples was determined using U.S. EPA
Method 413.1. Standard methods were used to determine the bulk
density, particle density, total porosity, moisture content and par-
ticle size distribution.

Sludge Application
  The sludge was applied as evenly as possible to the surface-
applied Plot A  (SA plot) and to the subsurface-injected Plot C
(SSA plot). The application tractor started injecting sludge at grid
point (GP) C19 (Fig. 1) and moved across Plot C from GP  C19
through GP C21. The injector then was lifted above the soil, and
the tractor proceeded across Plot B (control plot) from GP  B19
across GP B21. No sludge was applied to Plot B. Sludge then was
allowed to flow from the raised injector tines onto the surface of
Plot A as the tractor passed across it from GP  A19 through GP
A21. The tractor then made a U-turn and returned across Plots A,
B and C while applying sludge.  Several application passes were
made to complete the application to the entire area of Plots A and
C. Three  complete tractor loads  and part  of another were re-
quired. The average waste loading was 1.4  x  104 kg/plot. All
three plots were tilled with a disk  tiller pulled behind the tractor
during the sludge application.
  Emission  sampling with the flux  chambers was started im-
mediately  after application and continued for 4 days.  Emission
sampling was performed during two other 4-day periods in the 3rd
and  5th weeks following  application. Each plot was  tilled 2-3
times per week during the 5-week test period. The plots also were
irrigated with water once between the first and second sampling
periods and three times between the  second and third sampling
periods. The complete tilling and irrigation  schedule for the test
period is summarized in Table 2.
                          Table 2
   Tilling Schedule for Land Treatment Facility During Test Study
Date
10-10-84
10-12-84
10-15-84
10-16-84
10-18-84
10-19-84
10-24-84
10-26-84
10-29-84
10-30-84
10-31-84
11-01-84
11-02-84
11-05-84
11-09-84
11-12-84
Till Ing
Time
0950-1030
1224-1258
0930-1100
1300-1400
0800-0930
1300-1430
1222-1235
1218-1233
0815-0930
0900-1100
1000-1100
0900-1100
0900-1030
1300-1430
1225-1245
1047-1110
Hours from Start a
0
50.5
120.5
147.5
191
220
338.5
386.5
455
480
504.5
528
552
628
722.5
793
Activity
Apply waste and till
Till
Till and Irrigate
Till
Till
Till
Till
TIM
TIM
TIM and Irrigate
TIM
Till and Irrigate
TIM
Till and Irrigate
Till
TIM
a. Rounded off to nearest 1/2 hour.
SAMPLING AND ANALYSES RESULTS
  The results pf the sample analyses for the gas, liquid and soil
samples are summarized below.

Gas Samples
  The results of the on-site Byron THC analyses agreed quite well
with the off-site Varian 3700 analyses. The results were correlated
with a correlation coefficient of 0.977. The Varian GC generally
gave slightly higher concentrations, with an average difference of
13.8% between the Byron and Varian instruments.
  No methane was detected in any of the on-site samples analyzed
with the Bryon 401, which is capable of detecting methane at
levels of 1 ppm.
  The CC>2 concentrations in the on-site gas samples were found
to range from  about 50 ppm to almost 500 ppm. Most values fell
between 100 and 300 ppm. Background CO2 levels,  as determined
by sealing  the bottom  of the flux chambers and  passing  air
through them, were in the range of 10 to 200 ppm.
                                                                                                 ON-SITE TREATMENT     129

-------
Liquid Samples
  The sludge samples  were analyzed  several times to define the
sludge nonhomogeneity and obtain an average value for the level
of volatile organics. The analyses were performed over a period of
several months (during which the samples  were  kept  under re-
frigeration).  The  measured volatile organics content  of  the
sludge, as determined with the purge-and-trap technique, ranged
from 4,250 to 16,600 /ig/g. Most volatile organics concentrations
were between 4,000 and 6,000 /ig/g,  with the mean  value  being
7,800 jig/g. The variability was believed  to be at least partially at-
tributable to the difficulty of obtaining a representative sample in
small quantities for analysis.  The sludge volatile organic  com-
ponent was made up  principally of  paraffins (average concen-
tration = 3,600  pg/g),  olefins  (1,200 ug/g)  and  aromalics
(2,900 jag/g). The  remainder of the  volatiles were halogenated
and oxygenated hydrocarbons and unidentified organics.  From
the  characterization tests, the  sludge was  found to consist  of
3.9 to 4.5% solids, 57.5 to 59.1 % water and 37.0 to 38.0% oil.
   The sample of service water used to  irrigate the soil was also
analyzed  for volatile organics to  determine whether this  water
could be a significant source of volatile  organics (this sample was
kept under  refrigeration and analyzed  approximately  3  months
after collection).  The  volatile organics  content was found  to be
0.089 mg/1. The total amount of volatile organics applied with the
water over the duration of the test was estimated to be about 0.03
kg/plot. This amount is  equivalent  to less than 0.5% of the
volatile organics applied in the sludge.

Soil Samples
   The mean values for the physical properties determined for the
soil samples  are summarized in Table  3. Also included are the
mean moisture content  and oil/grease levels  for the collected
samples. The results are summarized for two of the three  sam-
pling periods.
                           Table 3
                     Mean Soil Properties

Test
Keek
1
3
Particle
Density
^g/cm )
2.50
2.54
Moisture
Content
lot J>
7.40
7.89
Bulk
Density Porosity,
(g/ciT5) (J)
1.21 52
1.42 44
Oil and
Grease
(mg/g)
64
56
STATISTICAL ANALYSES OF THE DATA
  A statistical analysis was performed primarily to investigate the
effects of time, temperature inside  the chamber  (soil and air),
plot samples, location within each plot and chamber shading on
emission rates. The frequency distribution of the emission rate
data was highly skewed and a log normal distribution was used
to model the frequency distribution of emission rates. To de-
termine whether measurable differences existed among the emis-
sion rates from the test plots, an analysis  of covariance was per-
formed. The analysis of covariance technique eliminates the vari-
ation in emission  rates due  to other  variables  such  as  time,
temperature  (inside the chamber) and chamber surface shading.
In this way,  the average emission rates for the test plots can be
directly compared. A discussion of the  analysis  of covariance
can be found in Brownlee.'
  The statistical analysis indicated that the emission rates  were
significantly affected by test plot and elapsed time. The results of
the statistical  analysis are summarized in Table  4. Soil or air
temperatures within the chamber were marginally significant.
None of the other variables were found to have a significant af-
fect on the emission rates.

                           Table 4
                   Components of Variability
Source of  Variability
                      Variance
                                 Percent of
                      Component   Total  Variance
Percent of Total
 Minus Temporal
   Variance
Temporal (day-to-day)      58.3          63.0

Air Temperature
 In Chamber               3.3          3.6

Plot                   28.1          30.4
      9.6

     82.2
Sanpl Ing Location
(•Ithln the plot)
Surface Chamber
Shading
Sampl Ing/ Analytical
Total
1.4
0.42
1.0
92.5
1.5
0.4
1.1
100
4.1
1.2
7.9
100
a Variance components arc equal to standard dotations squared and ihus have units which are
  the squares of that used for the log-emivsion rate% f>ig m--sccp
b Relative contribution of parameters to the total tanance


  The oil and moisture contents of the soil samples were exam-
ined to determine whether there was any significant change with
time in these parameters. There appeared to be little change in the
oil content of the soil over the duration of the test. The moisture
content of the soil appeared to decrease by about 14 to 40% over
the test period. Both changes were so small that no  correlation
with emission rates could be determined.

MEASURED VOLATILE ORGANICS
EMISSION RATES
  The volatile organics emission rates were examined with respect
to elapsed time since application, ambient conditions  at the time
of sampling and tilling activities.
                                            A  =  Surface
                                             3  =  Background
                                            C  =  Subsurface
          01234    5678     9  10 11  12
                             Sampling Days
            f = tilling episode

                           Figure 3
       Measured Emission Rates as a Function of Time Elapsed
                    Since Sludge Application
 130    ON-SITE TREATMENT

-------
                           Table 5
       Average Measured Emission Rates by Plot by Half-Day
                                               Table 6
                                     Cumulative Measured Emissions
Fmlsslon Rate
Sampl Ing
Day
0
1

2

3

4

5

6

7

8

9

10

11

12

Date
10-09-84
10-10-84
10-10-84
10-11-84
10-11-84
10-12-84
10-12-84
10-13-84
10-13-84
10-23-84
10-23-84
10-24-84
10-24-84
10-25-84
10-25-84
10-26-84
10-26-84
11-08-84
11-08-84
11-09-84
11-09-84
11-10-84
11-10-84
11-12-84
11-12-84
A Plot
(Surface)
1.88
249
142
22.5
32.6
48.2
146
33.4
46.3
16.6
15.0
8.04
49.1
6.41
13.5
21.8
71.1
2.56
7.25
1.24
12.34
2.49
6.4
2.97
10.5
(uo/nT-s)
B Plot C Plot
(Background) (Subsurface) Comments
1.48
24.7
10.0
4.09
5.11
4.09
9.93
8.18
8.86
2.71
6.13
2.73
10.0
2.04
5.80
1.32

2.59


1.22
1.39
1.43
1.19
4.42
1.95
176
52.8
13.8 early start
26.5
120
295 till
71.2
67.4
12.4
15.1
10.1
91.3 till
12.1 early start
22.0
13.0
59.3 till
2.93
5.76
3.94
16.2 till
11.5
7.13
4.06
16.2 till
 Each Plot Area = 423 m2
 Effect of Elapsed Time from Application
 on Emission Rates
   The measured emission rates for the three plots are shown as
 functions of elapsed time in Figure 3. The average measured emis-
 sion rates are summarized for  each half-day of sampling. These
 rates are tabulated in Table  5.  The  first point (Day 0)  was
 measured before sludge application. The tabulated values and the
 figure show the approximately exponential decline  in  emission
 rates with time elapsed from application (and, to some extent,
 elapsed time from tilling  episodes). The emission rates from the
 control Plot B remained substantially below those of the test plots
 throughout the duration of the testing.
   The cumulative measured  volatile organics emissions  are sum-
 marized in Table 6 and shown graphically in Figure 4. An approx-
 imately exponential decline in the emission rates is evident.
             A = Surface Applied
             B = Background Plot
             C = Subsurface Injected
                    3456789
                         Sampling Days
10  11  12
Fml<;>:lon Rate (kn)
Sampl Ing
Day
1
2
3
4
5
6
7
8
9
10
11
12
A Plot
(Surface)
1.327
2.084
2.204
2.378
2.634
3.413
3.591
3.837
3.926
4.006
4.049
4.310
4.345
4.417
4.533
4.912
4.925
4.964
4.971
5.036
5.050
5.084
5.100
5.156
B Plot
(Background)
0.1317
0.1850
0.2068
0.2340
0.2558
0.3088
0.3524
0.3996
0.4140
0.4467
0.4613
0.5146
0.5254
0.5563
0.5634
0.5772
0.5837
0.5911
0.5987
0.6051
0.6286
C Plot
(Subsurface) Comments
0.938
1.219
1.293 early start
1.434
2.074
3.646 till
4.026
4.385
4.451
4.531
4.585
5.072 till
5.136 early start
5.254
5.323
5.639 till
5.655
5.685'
5.706
5.793 till
5.854
5.892
5.914
6.000 till
                    Each Plot Area = 423 m2
                   Effect of Tilling on Emission Rates
                     Tilling of the soil had a significant effect on the emission rate.
                   The effect can be seen clearly in Figure 3 as spikes in the emission
                   rates immediately after tilling. The percentage increase in average
                   emission rates by half-day immediately following tilling episodes
                   is shown for each plot in Table 7.  The five tilling episodes for
                   which emission measurements were made were distributed among
                   the three sampling periods.
                                                                                            Table 7
                                                                               Increase in Emission Rate Due to Tilling
Observed
Tilling
Episode
1
2
3
4
5
Percent
A Plot
(Surface)
303
611
326
995
354
Increase In Emission Rate
B Plot
(Background)
243
366
0*
0°
371

C Plot
(Subsurface)
246
904
456
411
399
                          Figure 4
   Cumulative Measured Emissions as a Function of Sampling Day
(includes only measured rates for days on which sampling was performed)
a. Incomplete data set

ESTIMATED EMISSION RATE FOR
THE 5-WEEK TEST PERIOD
  The  emission  rates and  cumulative emissions  for  the  entire
5-week test period were estimated from the measured  rates. The
following assumptions were made to estimate the total  emissions:
                                                                                                  ON-SITE TREATMENT    131

-------
• Tilling had no effect on emission rates beyond 6 hours after
  tilling occurred
• Each day had 12 hours of cooler temperatures (night)
The emission rates were estimated for three distinct times during
each sampling period. These were: (1) nighttime, (2) non-tilling
daytime and (3) tilling daytime. The nighttime emissions for each
sampling period were estimated from the emission rate data ob-
tained during the early morning hours of that period. The tilling
and non-tilling day emissions for each sampling period were de-
rived  from  the appropriate emission rate measurements. Emis-
sions  for the interval between sampling periods were obtained by
interpolation among the three  emission rates for the three sam-
pling periods. Summing all the night-time, non-tilling daytime and
tilling daytime emissions, the total emissions for each of the plots
for the 5-week period were estimated to  be:
   Plot A (surface applied):       33.3 kg
   Plot B (control):               5.2 kg
   Plot C (subsurface inj.):        39.0 kg
   The estimated total emissions per plot per week are shown  in
Figure 5.
£.\J
» 18
= o IK
eo* ID
1 Jf 14
sl 12
Q) 10 _
Estimated We
Organic Emi
to *. a> CD a
o
A = Surface Applied
„ B = Background Plot
A
77

I
I
//



B.
_/

^
^
§
|
^
C = Subsurface Injected




\
*y

s
^
N
\


ll
I/A/' \~ fVyfTPxJ f77t7~T^
12345
Week
                          Figure 5
        Total Estimated Weekly Emissions from Each Plot
CONCLUSIONS
   The results  of  the  detailed sampling and  analysis of total
volatile organic emissions from this land treatment  facility in-
dicate that emissions from both surface-applied and subsurface-
injected  waste varied in a similar, approximately exponentially
decaying manner over the 5-week period of this test. This result is
in agreement  with the predictions  of the Thibodeaux-Hwang
emissions model for surface-applied plots.'
  Also,  the  effect of tilling  is similar for  both application
methods. Tilling resulted in a 2- to 10-fold increase in the emission
rates. A  large portion  of the 5-week cumulative emissions from
both plots is due to emissions which occurred within 4 hours after
tilling was performed.  The magnitude of emissions which occur-
red immediately after tilling decreases from tilling event to tilling
event approximately in proportion to  the decrease in emissions
during non-tilling periods.
  Based  on a statistical analysis of  the emissions,  it was found
that the emission rates were significantly affected by the elapsed
time. Soil and air temperatures within the chamber were mar-
ginally significant. None of the other  variables that were mon-
itored  were found to  have a significant effect on the emission
rates.
NOTICE
  The information  in this document has been funded wholly or
in part by the U.S.  EPA under contract to Radian Corporation.
It has been  subject to the Agency's peer and administrative re-
view, and it has been approved for publication. Mention of trade-
names or commercial products does not  constitute an endorse-
ment or recommendation for use.
REFERENCES
1. Radian Corporation,  "Field Assessment of Volatile Organic Emis-
   sions and Their Control at a Land Treatment  Facility," Draft Re-
   port, DCN 85-222-078-15-05, U.S. EPA  Contract No. 68-02-3850,
   Oct. 1985.
2. Hill, F.B., Aneja, V.P. and Felder, R.M., "A Technique for Meas-
   urements of  Biogenic Sulfur Emission Fluxes," J.  Environ.  Sci.
   Health, AIB(3), 1978, 199-225.
3. Adams, D.F., Pack, M.R., Bamesberger,  W.L. and Sherrard, A.E.,
   "Measurement of Biogenic Sulfur-Containing Gas Emissions from
   Soils and Vegetation," Proc. of 71sl Annual APCA Meeting, Hous-
   ton, TX 1978.
4. Schmidt, C.E., Balfour,  W.D.  and Cox, R.D., "Sampling Tech-
   niques  for Emissions  Measurements  at Hazardous Waste Sites,"
   Proc. of 3rd  National Conference on Management of Uncontrolled
   Waste Sites, Washington, DC, 1982, 334.
5. Brownlee, K.A.,  Statistical Theory and Methodology in Science and
   Engineering, John Wiley & Sons, Inc., New York, NY, 1965.
6. Thibodeaux, L.J. and Hwang, S.T., "A Model for Volatile Chemical
   Emissions to Air from Landfarming of Oily Waste," paper presented
   at the Annual Meeting of the American  Institute of Chemical En-
   gineers, New Orleans, LA, Nov. 1981.
132   ON-S1TE TREATMENT

-------
                               Technology  for Remediation  of
                                 Groundwater  Contamination

                                            David  V. Nakles, Ph.D.
                                                James E. Bratina
                                                     ERT, Inc.
                                            Pittsburgh, Pennsylvania
ABSTRACT
  The AquaDetox* technology is a high-efficiency stripping tech-
nology for the removal of organics from water. The technology
was developed by the Dow Chemical Company for internal use
through considerable process research, and only recently has the
technology become available for general use outside of Dow via a
licensing program.
  Two types  of AquaDetox*  units are available,  air or steam
AquaDetox*. Air AquaDetox* is capable of efficient removal of
chlorinated solvents and other pollutants considered easily strip-
pable from water. Steam AquaDetox* is used for the removal of
low-volatility organic materials from water including many com-
pounds not   considered  strippable   by  any  prior   process
technology.
  The technology has a wide range of applications  including
removal of 92 of the 111 organic priority pollutants as well as
other organic materials. It is applicable to the removal of trace
quantities of organics in water intended for drinking water pur-
poses with the capability of producing waters with organic con-
centrations from /ig/1 down to non-detectable levels.
  This paper describes the AquaDetox* technology, describes its
potential applications and presents economic and operating data
from  several operating commercial units.

INTRODUCTION
  Extensive groundwater contamination from both active and in-
active sites  has been identified across the United States. It has
been stated that more than 700 synthetic organic chemicals have
been discovered in groundwater resulting from over 30 different
types  of sources of contamination. The House Public Works Sub-
committee of Investigations and Oversight has estimated from an
inquiry last year that 4,000 city drinking wells were affected by
hazardous waste seepage into groundwater. Hence, the subcom-
mittee has assessed groundwater contamination as potentially one
of the most expensive problems confronting Congress and the
country.
  ERT has been involved extensively in the assessment of ground-
water contamination at a variety of active and inactive sites.
These efforts have focused largely on determining the nature  and
extent of contamination, modeling the migration of contaminants
beyond the site boundaries, assessing risks associated with  this
off-site  migration and preparing conceptual approaches  to
remediating the contamination as a means to reduce risk to accep-
table  levels.  To address the implementation of cost-effective

•Trademark and Service Mark of the Dow Chemical Company
remediation  strategies,  ERT  has licensed a  high efficiency
air/steam stripper known as the AquaDetox™ technology from
Dow Chemical.
  This paper discusses the current regulatory climate governing
groundwater remediation, describes the AquaDetox technology
(including selected performance/economic case studies) and pro-
vides  an overview  of ERT's  approach to offering it  to  the
industry.

Regulatory Overview
  In response to the growing concern for groundwater quality,
the U.S. EPA has developed a national groundwater protection
plan which is derived from eight different laws passed by Con-
gress. This plan divides  groundwater into three classifications
based on its use or potential use. There are specific groundwaters
that are irreplaceable drinking water sources and are particularly
vulnerable to contamination; groundwaters that are  currently
used or available for drinking water; and groundwaters that are
not potential sources of drinking  water because they already are
contaminated.
  Under the various statutes, drinking water aquifers will be pro-
tected using  RCRA (and its 1984 amendments) to discourage
pollution from landfills,  the Superfund law to select areas for
urgent cleanup and TSCA to develop  additional restrictions on
the use and storage of potentially threatening chemicals over these
areas. Other laws to be used by the U.S. EPA and state agencies
to control groundwater pollution are: the Clean Water Act under
its construction grants provisions; the Uranium  Mining and Mill
Tailings Reclamation and Control Act (radioactive contamina-
tion); and the Federal Insecticide, Fungicide and Rodenticide Act
(pesticide contamination).  Last,  the  National Environmental
Policy Act's sole source aquifer program and the Safe Drinking
Water Act's provisions for controlling underground  injection will
also be used.
  With regard to the Safe Drinking Water Act, current versions
of the Senate and House both refer to Phase I and Phase II lists of
contaminants which the U.S. EPA published in 1982 and 1983.
The Senate version  of the Act requires the U.S.  EPA to set stan-
dards for 85 contaminants with nine regulations  due within a
year, another 40 within 2 years and the remainder within 3 years.
  In January 1988, the U.S. EPA will be directed to publish a
new list of 25 priority pollutants with a year to propose recom-
mended and maximum contaminant levels and an additional year
to promulgate  final standards. This  process is to repeat  on a
3-year cycle. The House version of the Act addresses 64 con-
taminants and dictates either the promulgation of regulations or a
                                                                                            ON-SITE TREATMENT    133

-------
determination that no rational basis exists to believe the contami-
nant may have adverse health effects. This action must be com-
pleted on 14 volatile organic compounds within a year,  and deci-
sions on  the  remaining  contaminants are due within 3 years.
Beginning in 1988, the bill demands annual priority lists  for possi-
ble regulation, calling for standards within 3 years of publication.
   Under either version of the Act, the U.S. EPA must propose
Filtration  and  disinfection as treatment for drinking water. The
target maximum contaminant levels will be set as close to health-
based recommended maximum levels as possible, considering cost
and available  technology. In both cases,  granular activated car-
bon (GAC) is  suggested, though not mandated, as the "feasible"
technology. Generally speaking, however, GAC is a very expen-
sive treatment technology, and trade organizations such as the
Synthetic Organic Chemical Manufacturer's Association, as part
of a Safe Drinking Water Coalition, have stated that GAC is too
expensive and impractical. It is this same thinking which  led to the
development of the AquaDetox stripping technology by the Dow
Chemical Company and ERT's participation as a Licensee of that
technology.

The AquaDetox™ Technology
   The technology was developed by the Dow Chemical Company
for internal use through considerable process research  and  only
recently has it become available for general use outside of Dow
via a licensing program. AquaDetox achieves high stripping effi-
ciency at lower operating and capital costs when compared to
other technologies or unit operations. In many cases, removal of
organic contaminants, including many which have boiling points
in excess of 200°C and are typically considered "not strippable,"
can be achieved by a single, continuous stripping operation.
   The process has a wide range of applications including removal
of 92 of the  111 organic priority pollutants as well  as other
organic materials. It is applicable to the  removal of trace quan-
tities of organics in water  intended for drinking purposes with the
capability of producing waters with organic  concentrations from
jig/1  down to non-detectable levels. The  ability to achieve these
low levels of organics can eliminate the need for a polishing step
using GAC.
   Two types  of AquaDetox units are  available  as shown in
Figures 1 and 2. The air unit is capable of extremely efficient
removal of chlorinated solvents  such as perchloroethylene, tri-
chloroethylene, 1,1,1-trichloroethane, methylene  chloride and
other pollutants  considered  easily strippable from water.  The
steam AquaDetox is used for the removal of low volatility organic
mateials from water including many compounds not considered
strippable by any prior process technology. For example, it has ef-
fectively removed chlorinated phenols (i.e., pentachlorophenol)
and chlorinated  benzenes. The steam unit  also removes easily
strippable pollutants from water when vent gas emissions from the
air process are unacceptable.
                                 VENT TO ATMOSPHERE
        RAW WATER -
                         AQUADETOX
                           UNIT
                                        «. TREATED WATER
              AIR •
 Primary Advantages
   The benefits afforded by the application of AquaDetox units
 are many. First and  foremost, it is a  fully  developed,  proven
 technology. The Dow Chemical Company has 12 units operating
 at its Michigan Division treating a range of process wastewaters at
 flow rates of 10 to 3,000 gal/min. In addition, four field applica-
 tions of AquaDetox for outside firms have been installed and are
 operating on contaminated groundwater.

                       AQUEOUS CONDENSATE
  RAW WATER —


1




CONCENTRATE
ORGANIC
D
                     UNIT
                                       OVERHEAD
                                       —	» TREATED WATER
      STEAM
                           Figure 2
                    Steam AquaDetox Unit

  The continuous nature of the technology offers simplicity of
operation  as  compared  to  batch  processes  and  has  low
maintenance requirements and low operating costs (as will be
demonstrated in the following section). The technology provides
maximum flexibility for the sizing and location of units for both
permanent and mobile installations. Finally, its ability to achieve
high-efficiency removal (99 to  9.99%) of organics to achieve at
least /ig/1 levels makes it a viable alternative to GAC, which is ex-
pensive, introduces solids handling equipment and poses disposal
and/or regeneration problems.

PERFORMANCE/ECONOMIC CASE STUDIES
  Current  applications of  the  AquaDetox  technology  have
demonstrated its ability to cost-effectively remove organics from
water. Three case studies emphasizing groundwater remediation
are briefly presented to demonstrate this point.

Case Study No. 1
  An air AquaDetox was applied to groundwater contaminated
with 1,1,1-trichloroethane, methylene chloride, trichloroethylene
and  perchloroethylene. The system  was designed to treat a 10
gal/min stream and achieved extremely high removal efficiencies
for all pollutants as shown in Table  1.
  The estimated capital cost for this installation was $80,000 with
an operating cost (including utilities, labor, maintenance,  taxes
and depreciation) of $2.82/1000 gal (or $0.0028/gal). This com-
pared to an estimated GAC operating cost of $20.42/1000 gal for
the same level of pollutant removal.

                          Table 1
               Performance Data: Air AquaDetox
                      Case Study No. 1
                                                                      Pollutant
                             	Pollutant  levels,  ppm
                             Contaminated
                             Croundwater	Treated Effluent
                          Figure 1
                     Air AquaDetox Unit
1,1,l-Trichloroethane


Methylene Chloride


Trichloroethvlene


Perchloroethvlenc
                                                                                                 220
                                                                                                 180
                                                                                                                  0.08
                                                                                                              non-detectable
                                                                                                              non-detectable
                                             non-detectable
134   ON-SITE TREATMENT

-------
Case Study No. 2
  A steam AquaDetox system was installed to treat groundwater
contaminated with 6,000 mg/1 of a wide variety of organics shown
in Table 2. Typical pollutant  removals from  this 10 gal/min
stream exceeded 99% with several in excess of 99.99%.
  Estimated capital cost for the unit  was  $225,000 with an
operating cost  of $10.00/1000 gal (or $0.010/gal). Estimated
GAC operating costs to achieve these levels of pollutant removal
were $113 to $135/1000 gal of water.

                           Table 2
              Performance Data: Steam AquaDetox
                      Case Study No. 2
Pollutant levels , ppm
Contaminated Treated Percent
Pollutant Groundwater Effluent Removal
Methvlene choride 18
1,1,1-trichloroethane 434
laopropanol 5674
Acetone 772
HER 252
2-butanol 399
MIBK 130
2-hexanol 21
0.0007 99.
0.0019 99.
0.191 99.
2.7 99.
0.01 99.
0.014 99.
0.002 99.
0.059 99.
996
999
996
6
996
996
998
7
Case Study No. 3
  An air AquaDetox system was utilized to treat groundwater
contaminated with trace quantities of toluene (22,000 /ig/1) and
benzene (150 /tg/1). The technology achieved an effluent quality
with respect to both pollutants of less than 2 pg/1, the analytical
detection limits.
  This removal was achieved at an estimated operating cost of
$1.35/1000 gal versus an estimated cost for carbon adsorption of
$16.00/1000 gal.

OTHER CONSIDERATIONS
  One of the primary considerations when evaluating the use of
stripper technologies for groundwater contamination is the accep-
tability of the vent gas discharge. In general, this is evaluated on a
case-by-case basis at  the state level, but it is estimated that less
than half-a-dozen states have any specific legislation at this time.
The crux of the matter centers on a cost-benefit analysis where the
trade-offs are between the additional cost to control vent gas
emissions (or avoid them entirely) using GAC and the health risk
to these uncontrolled emissions.
  Since a number of industries already emit volatile organic car-
bon at legal levels much higher than those under consideration for
the groundwater cleanup  situation, the key  question  becomes
whether a reduction of the already low health risk, especially with
the rapid dispersion of pollutants, is worth the price of control-
ling air emissions from groundwater cleanup. To put this issue in-
to perspective, it has been determined in many instances that the
concentrations of pollutants in the vent gas prior to their disper-
sion in the atmosphere do not exceed OSHA workplace standards
for the individual pollutants. Clearly, each application will have
to be reviewed individually with consideration given to the relative
health risk  associated with the lack of control  of such emissions.
  In instances where vent gas control is dictated, the use of GAC
to remove the pollutants from the gas may be considered. Such a
combination of stripping and GAC has been applied on a Super-
fund site in Michigan and has proven cost-competitive with direct
adsorption of the organics from the water.  This competitiveness
has been enhanced by the incorporation of humidity control prior
to the carbon beds since the adsorptive capacity of GAC is re-
duced dramatically when it gets wet.

CONCLUSION
  As a process wastewater treatment technology, AquaDetox can
be used alone or in combination with GAC or biological treat-
ment systems  to  produce  a  more  cost-effective or  reliable
operating system.
                                                                                                ON-SITE TREATMENT    135

-------
                              Anaerobic  Biological Treatment
                                of  Sanitary  Landfill  Leachate

                                              A.K. Mureebe, P.E.
                                                   D.A. Busch
                                               P.T. Chen, Ph.D.
                                              Wehran Engineering
                                            Middletown, New  York
ABSTRACT
  The anaerobic treatment of sanitary landfill leachate was in-
vestigated in bench-scale studies at ambient room temperature.
The test runs were conducted in continuous operation for more
than 6 months using the leachate collected directly from the
leachate collection sump installed at a lined landfill site. No  pre-
adjustments of pH or other chemical or physical features were
made. The leachate used in the study showed highly variable
characteristics  with a COD of 2,000 to 35,000 mg/1, BOD5 of
1,800 to 24,000 mg/1 and ammonia nitrogen of 50 to 500 mg/1.
  An anaerobic fixed-flim reactor, with a volume of 22.24 1 and
with a controlled rate of sludge  recycling, was  operated with
upflow feeding at  organic loading between 0.23 and 2.48 kg
COD/mVday and hydraulic detention times of 6.9 to 10.3 days.
  The test resulted in filtered COD removals from 92 to 94%, un-
filtered BOD5 removals  of 90-94% and ammonia removal from
40 to 50%.  Despite the  wide range of loading conditions, the
system showed a high level of stability in terms of COD removal
efficiency which hovered about 93%. With influent COD of 1,700
to 17,000 mg/1, the effluent leaving the system was in the range of
100 to 1,300 mg/1, depending on the organic loading rate. The ef-
fluent can be further purified by an aerobic biotreatment unit to
reduce the COD and BOD5 levels below 250 and 35 mg/1, respec-
tively.

INTRODUCTION
  The landfill is located  off Route 17M east of the Wallkill River
in the Town of Goshen,  Orange County, New York (Fig. 1).  The
site was  designed  to protect  surface and groundwater from
leachate contamination by utilizing the following technologies:

• Low permeability on-site glacial till material used as soil liner
  to minimize  the impact of leachate  on adjacent surface  and
  groundwater
• A sloped landfill  bottom to direct leachate toward a collection
  system
• A leachate collection piping system to intercept and collect the
  leachate
• A collection  well  with level-controlled pumps that transfer the
  leachate collected in the landfill to a geomembrane-lined pond,
  thus maintaining the hydraulic head in the landfill at a low level
  The leachate management issues currently facing  the landfill
are the  establishment  of a feasible and economically  sound
disposal  scheme that will meet the needs of the current operation
and future closure of the landfill, and the acquisition of permits
from state  and/or  local agencies for  disposal of the leachate
generated.
                         Figure 1
                Location Map — Turi Landfill
  Wehran Engineering has been retained to conduct a study that
will provide a viable system for leachate treatment and to furnish
the necessary framework for implementation.
  This paper includes a detailed investigation of all relevant
features and data pertaining to leachate treatment. Specifically,
the scope included the following:

• Analysis of leachate characteristics to determine the quality and
  quantity of leachate generated by the site
• Performance of bench-scale treatability tests to confirm the
  technical  feasibility of on-site treatment options
• Development of engineering design criteria

Quantity and Characteristics of Leachate
  An extensive monitoring, sampling and analysis program was
conducted for a period of 1 year at Turi Landfill to determine the
quantity and quality of leachate generated from the operational
cell of the landfill. Weekly and monthly samples were collected
during this period and analyzed in the laboratory.
  The weekly samples were analyzed for the conventional pollu-
tant parameters of chemical oxygen demand (COD), 5-day bio-
136    TREATMENT OF HAZARDOUS WASTES

-------
chemical oxygen demand (BOD5),  total solids (TS),  ash, total
Kjeldahl nitrogen (TKN), ammonia nitrogen  (NH3-N),  hard-
ness, conductivity and pH.
  The monthly samples were analyzed for iron (Fe),  zinc (Zn),
copper (Cu), cadmium (Cd), lead (Pb), manganese (Mn) and total
phosphate (T-PO4).

Leachate Flow
  The flow of leachate generated in the active landfill operation is
shown in Table 1. The actual monthly average daily leachate flow,
measured in the  operating landfill  area for the  period between
April  1984 and March 1985, ranged from 100 to  18,000 gal/day,
with an  annual average of 6,300 gal/day. The variation in flow
probably is due to  the  absorptive characteristics of the solid
waste.

Characteristics of Leachate
   Starting in November 1984, sampling and analysis of the con-
ventional pollutant parameters described above were  performed
as part of the study program. In conjunction with this sampling
and analysis, leachate flows were recorded weekly.
   With the exception of pH, the pattern of changes in the concen-
trations  of  these pollutant parameters appears to  be  in general
correlation with the changes in leachate flows.  The concentration
changes are inversely proportional to the leachate  flow rates
                           Table 2
             Leachate Characteristics - Turi Landfill
Parameter
BOD5
COD
Am monia Nitrogen
Organic Nitrogen
Total Solids
Ash
Hardness
Conductivity
pH
Range
180- 36,900 mg/1
200 -45,800 mg/1
5 - 640 m g/1
1 - 140 mg/1
1,005-25,330 mg/1
805- 13,580 mg/1
345 - 9,460 mg/1
900 - 15,500 u mhos/cm
5.30 - 6.00
Arithmetic
Average
12,042 mg/1
19,527 mg/1
340 mg/1
46 mg/1
14,367 mg/1
7,037 mg/1
5,647 mg/1
9,360 u mhos/cm
5.55
  The monthly results for the metals and total phosphate analyses
are summarized in Table 3. These concentrations are comparable
to reported ranges for sanitary  landfills.4  All  concentrations
observed in this study are  below reported  threshold limits for
biological assimilation in a wastewater treatment facility.''4'8

                          Table 3
              Monthly Analysis of the Concentration
              of Metallic Elements and Phosphates
wmcn may DC aunouiea 10 me percoiaiion or ramiau. i ne exier-
nal water percolation in the landfill will thus act as a dilutant in
carrying leachables out of the landfill bed.
Table 1
Leachate Generation Rates
Measured Values Estimated Values (5)
Ppt(l) Perc (2) Ratio Flow (4) Ppt Perc Ratio Flow
Month (inch) (inch) (3) (gpd) (inch) (inch) (3) (qpd)
April 84 4.65 1.080 0.23 12,840 3.86 2.12 0.55 25,135
May 10.03 — -- — 4.22 1.48 0.35 17,548
June 3.95 0.270 0.07 3,212 3.86 1.35 0.35 16,006
July 6.27 0.430 0.07 4,920 4.01 1.41 0.35 16,717
Aug. 2.61 0.290 0.11 3,367 3.39 1.19 0.35 14,109
Sept. 1.18 -- - - 3.01 1.06 0.35 12,567
Oct. 2.88 — — — 3.03 1.67 0.55 19,800
Nov. 2.58 0.008 0.03 92 2.88 1.59 0.55 18,852
Dec. 2.99 0.540 0.18 6,139 2.46 2.09 0.85 24,780
Jan. 85 1.06 0.127 0.12 1,463 2.46 1.23 0.50 14,584
Feb. 1.99 1.430 0.72 18,192 2.68 1.34 0.50 15,888
March 2.21 0.516 0.23 5,916 3.88 3.30 0.85 39,127
Annual 42.4 4.7+ — -- 39.74 19.83
Average .19 6,239 .50 19,584
Monthly

Notes:
(1) Precipitation (ppt) from climatological data, Middletown Station, NY, NOAA.
(2) Calculated percolation (perc) through cover soil and underlying solid wastes based on actually
measured leachate flow within 13.1 acres of operating landfill.
(3) Ratio of perc to ppt.
(4) Actual leachate flow measured.
(5) Calculated leachate flow, based on the estimated values of percolation for 13.1 -acre area.
Metallic
Elements 10/25/84 12/20/84 1/24/85 2/27/85 4/4/85
Fe 770.00 420.00 1,100.00 1,140.00 840.00
Cu 0.05 0.08 0.05 0.05 .35
Pb 0.05 0.05 0.16 0.05 .14
Mn 190.00 41.00 310.00 200.00 200.00
In 11.00 14.00 33.00 6.30 11.00
Cd 0.01 0.01 0.01 0.01 .01
T-P04 8.40 1.90 3.40 2.50 6.50
Note: All units are in mg/1

The concentrations of priority pollutants which were present in
the leachate at levels above analytical detection limits are given in
Table 4. All organic chemicals on the list are reported to be effec-
tively removed by conventional biological treatment processes at
both POTWs and industrial wastewater treatment facilities.'

Leachate Treatability Evaluation
A review of the literature reveals that leachate may be treated
by aerobic1 '2'3'4'6'9 or anaerobic biological processes7'8, physical
methods and chemical means.4-9 Existing reports discuss mainly
laboratory-scale studies.
Bench-Scale Treatability Tests
In the selection of systems for a treatability study, considera-
tion was given to the processes available at local POTWs, in-
dustrial wastewater treatment facilities and for the purchase of
packaged treatment systems for possible on-site use. The follow-
ing process techniques were chosen for the bench-scale treatability
studies:

• Aerobic bio-oxidation system utilizing local POTW activated
sludge for the evaluation of maximum ratio of leachate to
  The range and average concentrations  for  the conventional
pollutant parameters during the study's test period are given in
Table 2. The  ranges are comparable with reported  ranges for
sanitary landfills published in technical literature.4'7
   sewage flow
   Anaerobic biological system for treating raw leachate
   Aerobic treatment for polishing the effluent of the anaerobic
   system
   Physical/chemical treatment for removing solids or COD as a
   pre-treatment or post-treatment process
                                                                                 TREATMENT OF HAZARDOUS WASTES     137

-------
                             Table 4
            Concentrations of Detectable Priority Pollutants
                      in Turi Landfill Leachatc
            Parameter
       Benzene
       Chloroethane
       Dichlorodifluoro methane
       1,1-Oichloroethane
       1,2-Dichloroethane
       1,2- Dichloro propane
       Ethylbenzene
       Methylene Chloride
       Tetrachloroethylene
       Toluene
       1,2-Trans-Dichloroethylene
       1,1,1-Trichloroethane
       Trichloroethylene
       Vinyl Chloride
       Diethyl Phthalate
       Isophorone
       Naphthalene
       Arsenic
       Cadmium
       Chromium
       Copper
       Lead
       Nickel
       Selenium
       Zinc
       Total Cyanide
       Total Phenolics
  Reported
Concentration
   (mg/1)

    0.025
    0.062
    0.089
    0.301
    0.081
    0.015
    0.051
    2.600
    0.079
    1.000
    0.228
    0.111
    0.130
    0.010
    0.076
    0.067
    0.010
    0.018
    0.015
    0.240
    0.027
    0.090
    0.390
    0.007
   12.000
    0.100
    4.000
   The  bench-scale  treatability  tests were performed using the
 following equipment:
 • For anaerobic biological system: A fixed film upflow reactor,
   6 in. diameter  x  5  ft high with hydraulic  detention time of
   7 days, was used  in a plug flow mode with recycle (Fig. 2).
 • For aerobic bio-oxidation  system: A 4-1 reactor with air dif-
   fusers was used  in  a completely mixed mode.  Sludge  was
   wasted directly from the unit or recycled from a settling  unit
   to maintain the desired MLVSS under aeration (Fig. 3).
 • For physical/chemical test: A multiple stirrer,  Phipps-Bird
   stirring apparatus was used to test the effect  of coagulation/
   precipitation with lime, caustic and/or alum.

 POTW ACTIVATED SLUDGE
 TOLERANCE STUDY
   In this test, activated sludge  seed  from the Harriman Waste-
 water Treatment Facility was aerated in a bio-oxidation unit. The
 system  was  run  to  a steady state MLVSS with a sewage  and
 nutrient feed prior  to the addition  of raw leachate. Once the
 bench-scale activated sludge system achieved this steady state, raw
 leachate was added in ratios of raw leachate to sewage of 1 to  5%.
   The results of this study are given in Table 5. At dosage ratios
 of raw leachate to raw sewage greater than 2%, the aerobic treat-
 ment system is adversely impacted. The system failed to maintain
 a steady state MLVSS at a 5%  dosage, and higher order organ-
 isms were observed. At the high dosage levels,  increased phos-
 phate addition on the order of 1,000 mg/1 was necessary for ade-
 quate cell development.

r 	 "
^-U


x
\ /
A




H
^
i — >>
^
i
"^x j
e ai 3 H <|
,4..
H ,

'
r-
k.
	
rOAS/LIOa/SOL
SEPARATOR
I •'» i a' H
r

L
Oil
METER
1
i
f
-| 1 	
10 1
^
Ln

i











T


AAS
ANALYZER

PUMPIN8 RATE
CONTROL
PANEL


                   'Q~\
                                             I
RECYCLE PUMP
                   FEED    FEED     31UDOE   EFFLUENT
                   PUMP    TANK     TANK     TANK
                             Figure 2
     Bench-Scale Anaerobic Fixed Film Biological Treatment System
                                                    EFFLUENT
                                                    TRAP BOTTLE
                                                 Figure 3
                                    Aerobic Bio-Oxidation Treatment Unit
                                                  Table S
                                      POTW Tolerance S(ud> of Aerobic
                                  Bio-Oxidation System Treatment of Leachate
                                   Raw Leichatc,
                                    Raw Sewage
                       •Ai 5°'o ihc system coult) not aliam u steady stale Ml VSS
                                              Average COD
                                                Removal
                                                                       81
                                                                       70
                                                                       54
                                                                       37
                                                                       12
                       ANAEROBIC TREATMENT
                         The  parameters monitored in studying the efficiency of the
                       anaerobic biological treatment  system were  COD,  BOD5 and
                       TVA. The total volatile acid removal (TVA) is an important step
                       in methane fermentation and is therefore included. The results for
                       the removal of these parameters are  summarized in Table 6.
                         In evaluating the anaerobic treatment of landfill leachate, the
                       removal of COD and BOD are the most important parameters to
                       study. As can be seen from the data in Table 6. the efficiency of
138    TREATMENT OF HAZARDOUS WASTES

-------
                                                          Table 6
                                          Anaerobic Test Results of Leachate Treatment
Run
1
2
3
4
5
6

Influent
--
4,500
8,775
11,250
11,625
8,700
BOD<;(mq/l)
Effluent
—
390
855
960
1,200
540

X Removal
—
91.3
90.3
91.5
89.7
93.8

Influent
1,670
6,480
13,430
16,990
16,990
9,875
COO (rag/1)
Effluent
100
355
935
950
1,275
545

X Removal
94.0
94.5
93.0
94.4
92.5
94.5
TVA
Influent
1,730
6,520
12,300
14,485
14,160
8,060
(rag acetate/1)
Effluent
405
660
1,545
925
1,710
205
% Removal
76.6
89.9
87.4
93.6
87.9
97.5
              •TVA is total volatile acids.
the removal of these two parameters is quite constant. For Run
No. 1, which had an organic loading rate of 0.23  kg COD/m3
• day, the COD removal was 94%, while with an organic loading
rate of more than 10 times that in Run No. 1, the COD removal
rate (Run No. 5) was 92.5%. The system showed an ability to
operate efficiently under a wide range of COD loading condi-
tions, which is an important characteristic in the treatment of
landfill leachate. In conjunction with the COD removal, the study
of two other system paramters is important: methane and sludge
production.

Methane Production
  One of the major advantages of the anaerobic process is that
organic material in  the wastewater is converted to methane gas.
The methane  gas  evolution rates  and composition for  the
anaerobic treatability study are summarized in Table 7.

                          Table 7
     Gas Production Data for Anaerobic Treatment of Leachate
Run
1
2
3
4
5
6
Gas Evolution
Rate
(ml/hr)
125
500
745
965
1,350
745
Methane
(%)
67
71
64
70
67
—
Methane Rate
(I/day)
(at STP)
1.84
7.81
10.48
14.85
19.89
10.97
   As can be seen, methane production rates determined from ex-
 perimental removal of organics are higher than theoretical rates.
 This may be a result of average daily influent conditions actually
 being higher than those detected in the samples analyzed;  this
 would mean that the organics removal rate is higher than the  rate
 used in the calculations. An important finding of this portion of
 the study is that the methane production rate was directly propor-
 tional to the organics  loading rate.

 Solids Production
   Another advantage  of the anaerobic system is the low solids
 production rate. Sludge production from anaerobic biological
 treatment systems  typically ranges from 0.04 to 0.10 g VSS/g
 COD removed. The sludge  production rates as a function of
 organics removal are  summarized in Table 9.  The amounts of
 solids produced were estimated from the increase in volatile solids
 concentrations in the circulation stream.

                          Table 9
   Anaerobic Sludge Production in Treatment of Landfill  Leachate
                                                                    Run     (g/day)
                                                                            2.82
                                                                            2.81
                                                                            1.67
                                                                            6.67
                              COD
                            Removed
                             (g/day)

                             18.52
                             26.99
                             34.64
                             50.92
TVS/  COD
 (g solid/g
   COD
 removed)

   .15
   .10
   .05
   .13
  To determine the efficiency of the methane fermentation, the
experimental  gas production  rate  can be  compared to the
theoretical value which is 0.35 1 CH4/g of organics removed. The
methane production rates are shown as a function of the removal
of BOD, COD and TVA in Table 8.
                          Table 8
        Methane Production Rates for Anaerobic Treatment
                     of Landfill Leachate
                 Gas Rate as a Function of Organic Removal
                                                                    Average
                                                                                                                 .11
Run
1
2
3
4
5
6
Liters CH4/
g COD removed
0.39
0.42
0.39
0.43
0.39
0.36
Liters CH4/
g BOD removed
	
0.53
0.61
0.67
0.59
0.41
Liters CH4/
g TVA removed
0.46
0.44
0.45
0.51
0.49
0.43
  The sludge production of  the  anaerobic  digestion of  the
leachate was slightly higher than normal anaerobic sludge produc-
tion, yet the amount of sludge was still significantly lower than
the amount of sludge produced by a conventional activated sludge
process.

AEROBIC TREATMENT OF
ANAEROBIC EFFLUENT
  The aerobic  treatment  of the effluent from  the anaerobic
system was conducted in a fill and draw manner. Anaerobic ef-
fluent was added to activated sludge from the Harriman Sewage
Treatment  Plant, and the system  was aerated.  Samples were
drawn from the system at 12,  24 and 48 hrs  for analysis. The
characteristics of these samples were compared to the anaerobic
effluent that  was added initially. The test was conducted with
COD and BOD values of 1,300 and 200 mg/1. The MLSS, MLVS
and SVI values were 8,100, 3,950 and 135, respectively.
                                                                               TREATMENT OF HAZARDOUS WASTES    139

-------
                         Table 10
    Results of Aerobic Treatment of Anaerobic System Effluent

Hour
0
12
24
48

COD
1,300
400
260
235
COD X
Removed

69.2
80.0
81.9

BOD
1,200
100
70
35
BOD X
Removed
__
91.7
94.2
97.1

NH3-N
135
110
45
5
NH3-NX
Removed
__
18.5
66.7
96.3
COO/
BOO^
1.1
4.0
3.7
6.7
  The removal of the conventional pollutants (COD, BOD and
NH3-N) was monitored to determine the efficiency of the pro-
cess. The results of these tests are summarized in Table 10.


Removal of BOD and COD
  In discussing the removal of organics by the aerobic system, an
important parameter to consider is the ratio COD/BOD5.
  The general trend is an increase of the COD/BOD ratio with in-
creased  aeration. This  increase  is due  to a rise  in  the non-
biodegradable portion of the leachate over time. The test  for
chemical oxygen demand will detect some organics which are non-
biodegradable and thus will not be measured by the BOD test. As
the aeration time increases,  the biodegradable organics will  be
removed by the system microorganisms, and the non-biodegrad-
able organics will represent an increasing proportion of the re-
maining organics.
  Based on the results of this test, we concluded that a removal
rate of 80% of the COD  and 95% of the BOD is attainable by the
conventional aerobic treatment systems.


Removal of Ammonia
  The  aerobic  system  removes  ammonia by  the  process  of
nitrification; the ammonia is oxidized to  nitrate. The bacteria
responsible for nitrification have a growth rate slower than that of
the bacteria responsible for the removal of organics. Thus, it is ex-
pected that a longer aeration time will be required for the removal
of NH3-N than that for BOD. As  can be seen from the  test
results, the removal of BOD was as high as 90% in only 12 hr,
while an aeration time of 48 hr was needed to remove 96% of the
ammonia.
average sludge production rate of 0.68  g  VSS per g  COD re-
moved. This increased sludge production is attributed to the ex-
tended aeration time and the lack of sludge wasting in the test to
maintain  a  steady state MLVSS. The results do indicate that
under typical conditions  for aerobic biological  treatment, the
sludge production from treatment of the leachate would  be within
the expected range for such systems.


CHEMICAL COAGULATION/PRECIPITATION
  The effect of chemical coagulation and precipitation on COD
and suspended solids removal also was evaluated in this study.
The test conditions for the study are summarized  in Table 12. In
this test, the samples were dosed with the coagulating chemicals,
rapidly mixed  using a  Phipps-Bird  stirring apparatus  at  50
revs/min  for 30 sec, flocculated at 15 revs/min for 20 min and
then allowed to settle  for 30 min. The supernatants then were
sampled and analyzed for COD and TSS.
                          Table 12
               Chemical Coagulation/Precipitation
                Treatment Operating Conditions
 Waste Stream

Raw Leachate
Anaerobic Effluent
Aerobic Effluent
  Ca(OH)2
   (•g/1)

1.  800-4200
2.  2800
3.  --

1.  1800
2.  --
3.  --
4.  --
                               Cnemical Coagulant Dosages
                                       NaOH
                                       (•9/1)
                                       1200
                                                      Alum
 0-333
 0-333

 0-333
 0-333
 0-333
0-2000

0-1455
0-1500
0-2000
   The  results  for  the  chemical  coagulation/precipitation
treatability studies on the raw leachate, anaerobic filter effluent
and aerobic polishing unit effluent are presented in Tables 13, 14
and 15, respectively.
Solids Production
  The production of sludge is an important parameter in the design
of the aerobic process. Table 11 contains a summary of the solids
production as a function of the amount of organics removed.
  The reported range of sludge production for activated sludge is
0.25 to 04 g of VSS formed per gram  of COD removed. The
results of  the  bench-scale aerobic polishing  unit indicate  an
                         Table 11
                Aerobic Sludge Production for
            Treatment of Anaerobic System Effluent
Time
(hrs)
0
12
24
48
Average
COO
(in 9/1)
1,300
400
260
235

MLVSS
(rcg/1)
1,610
1,860
2,030
3,060

COO
Removed
(mg/1)
__
900
1,040
1,065

MLVSS
Increase
(mg/1)
._
250
420
1,450

MLVSS/
COD
--
.28
.40
1.36
.68
                          Table 13
           Results of Chemical Coagulation/Precipitation
                  Treatment of Raw Leachate
Chemical Coagulant
Dosages (ng/1)

Run
1





2





3






Ca(OHj>
0
830
1,670
2,500
3.330
4,170
0
2,800
2,800
2,800
2,800
2,800
__
.-
—
—
—
--

NaOH
„
_.
..
.-
,_
--
__
-.
..
._
.. .
--
0
1.200
1,200
1,200
1,200
1,200

Alum
..
._
„
..
__
--
0
0
33
83
167
333
0
0
33
83
167
333
COD
(•q/1)
31.740
32.105
32,230
32.350
31,130
31,660
31,335
30.805
31,700
30,600
31,660
30,805
26,110
22,010
22,560
22,440
23.665
--
Parameters
Percent
Removed
..
	
	
..
1.9
0.3
..
1.7
_.
2.4
.,
1.7
..
15.7
13.6
14.1
9.4
--
TSS
(•g/il
3.700
1,400
900
—
5
600
3,780
2,045
2,585
3,050
1.540
1,495
4,915
3.340
4,535
3,560
1,765
955
Percent
Removed
„
62.2
75.7
—
99.9
83.8
_.
45.9
31.6
19.3
59.3
60.5
„
32.0
7.7
27.6
64.1
80.6
140    TREATMENT OF HAZARDOUS WASTES

-------
                          Table 14
      Results of Chemical Coagulation/Precipitation Treatment
                  of Anaerobic Filter Effluent
                            Table 16
              Operating Conditions and Performance of
                   Bench-Scale Anaerobic Filter

Chemical
Coagulant

Dosages (mg/1)
Run


1




2




3


4



Ca(OHt>

0
1,800
1,800
1,800
1,800
1,800
__
—
—
—
—
	
—
~
—
—
—
—
Alum


0
0
33
83
167
333
0
33
83
167
250
333
0
83
167
333
0
1,000
1,500
2,000
COD
(mg/1)

485
600
600
620
550
625
610
620
575
560
555
530
980
955
955
880
685
635
565
585



P ara m eters
Percent
Removed

—
—
—
--
—
__
—
5.7
8.2
9.0
13.1
--
2.6
2.6
10.2
—
7.3
17.5
14.6
TSS

--
45
5
35
0
35
120
80
40
70
45
50
39
12
19
37
45
Percent
Removed

—
—
__
—
--
	
33.3
66.7
41.7
62.5
58.3
..
69.2
51.3
5.1
—


Operating Conditions Removals
Hydraulic
Oet. Time
(days)

Run 1 7.7
Run 2 7.3
Run 3 10.3
Run 4 10.3
Run 5 6.8




9 V5/
kg COO/ g COD rem.
m 3 day "ercent Solids
Org. Loads COD rem.

0.216 94.0
0.419 94.6
1.302 93.7
2.100 94.0
2.380 92.0


Prod.

-
0.150
0.100
0.040
0.112



Products
LCH4/gCOD
rem.

0.478
0.458
0.418
0.497
0.434




LCH4/g
VFArem.

0.760
0.672
0.504
0.596
0.656


* liters of methane produced/g COD removed



Summary




Table 17
of Performance of the










Four Treatment Systems

Performance

8
—
82.2

Process

COD

(X)


in Removal of

NH3-N

(%)
                          Table 15
      Results of Chemical Coagulation/Precipitation Treatment
               of Aerobic Polishing Unit Effluent
Run
Chemical
Coagulant

Alum
0
545
910
1,455
0
500
1,000
1,500
0
1,500
2,000
Parameters

COD
235
220
205
180
320
310
270
245
485
360
330
Percent
Removed
„
6.4
12.8
23.4
__
3.1
15.6
23.4
__
25.8
32.0

TSS
..
--
—
--
70
50
20
20
240
2
--
Percent
Removed
..
—
—
--
	
28.6
71.4
71.4
__
99.2
--
Removal of COD
Raw Leachate
  Of the three tests run, only the addition of caustic and alum
showed any significant reduction in COD, and this was less than
20%. The addition of lime only, and lime and alum showed no
appreciable  decrease in COD from pH elevation or coagulation.
The results of the third run indicate that the  best COD  removal
was obtained with caustic alone, and that alum was ineffective as
a coagulating agent for COD removal.
  The results  obtained  in this study indicate  that  chemical
coagulation and precipitation processes do not appear to be ade-
quate as a pretreatment process for disposal of the leachate either
off-site at a POTW or through an on-site treatment system.

Anaerobic Filter Effluent
  The results of the chemical coagulation and  precipitation of the
anaerobic filter effluent indicate that large dosages of alum are re-
quired to obtain moderate removal efficiencies. The results of the
second and third runs indicate that a dosage of 333 mg/1 of alum
results in a 10-13% removal of COD. By increasing the dosage to
 Anaerobic  system

 Aerobic system

 Alum coagulation

 Integrated efficiency
  94                  49

  80                  95

	25                  --_

99.1                97.5
 1,500 mg/1 in Run No. 4, a removal of only 17.5% is attained.
 These results indicate that an increase in the alum dosage of 500%
 results in only a 5% increase in COD removal. This process is,
 therefore,  not considered effective  for reducing residual COD
 from the anaerobic filter effluent.

 Aerobic Polishing Unit Effluent
   The results presented in Table 15 indicate that very high levels
 of alum (1,500-2,000 mg/1) will remove only 25-30% of the COD
 in the aerobic polishing unit's effluent. This process, therefore, is
 not considered effective  for removing any residual COD in the
 aerobic polishing unit, should such removal be essential to meet
 discharge standards.

 Removal of Suspended Solids
 Raw Leachate
   The test results presented in Table 13  indicate that effective
 suspended solids removal in the raw leachate is achieved only with
 very high dosages of lime (3,330-4,170 mg/1). Such high dosages
 generate proportionally high sludge volumes which would require
 further processing for proper disposal.
   Based on these results, we concluded that this process is not ef-
 fective as a  pretreatment process for disposal of the leachate
 either off-site at a POTW or through an on-site treatment system.

Anaerobic Filter Effluent
   The results presented in Table 14 indicate that effective removal
 of suspended solids in the anaerobic filter effluent is achieved
 only with high levels of alum (2,000 mg/1). At lower dosages
(33-83  mg/1), moderate removals were attained,  indicating that
this process may be viable for solids removal in conjunction with
a  clarification system,  should  subsequent  biological  or
physical/chemical treatment for residual organics and ammonia-
nitrogen require pretreatment for solids removal.
                                                                                 TREATMENT OF HAZARDOUS WASTES    141

-------
 Aerobic Polishing Unit Effluent
    The results presented in Table 15 indicate lhat relatively high
 levels of alum (1,000 mg/1) effectively reduce the suspended solids
 level in  the  aerobic  polishing  unit's  effluent  to  acceptable
 discharge standards (20 mg/1). Full-scale clarification units, with
 tube or plate settlers, may reduce the chemical dosage required to
 achieve this effluent quality. This process is considered viable for
 consideration  as  a polishing treatment  for  suspended  solids
 removal.

 CONCLUSIONS
    The laboratory bench scale  study using an anaerobic fixed-film
 reactor   for   the  treatment of   sanitary  landfill  leachatc
 demonstrated the amenability of the process and its  application
  for the stabilization of leachate.
    The landfill raw leachate was rapidly adopted  by the sludge
 used in  the anaerobic fermentation system. Hydraulic detention
 times as short as 7 days and organic loads  up to 2.2 kg COD/m'
 day resulted in a 94% removal of COD.  The gas produced in the
 system was composed of 60-70%  combustible material. Based on
 the rate of gas production, we estimated that more than 99%  of
 the COD removed was converted to CH4 gas.
    The process performance with respect to  COD removal and gas
 production at  various  loading conditions is shown in Table 16.
 The removal of COD and the  rate of solids production were very
 steady.  The  response of process  performance  to changes  in
 organic  or hydraulic  loads was reflected in the rates of methane
 production;  however, the  efficiency of COD  removal remained
 unaffected for the conditions  tested.
    The performance of the activated sludge system for treating the
 effluent  from the anaerobic filter was much better than anaerobic
 treatment alone. With a loading  of 0.90 g  COD/g MLVSS, the
 system attained removals of 80 and 95%  for COD and NH, - N,
 respectively, within 48 hr of aeration. The estimated rate of sludge
 formation in the aerobic system was 0.41  g/VS/g COD removed.
  The series of tests conducted  in the bench-scale treatability
study have confirmed that the landfill leachate can be treated with
the use of conventional biological and physical/chemical pro-
cesses. The effluent qualities achievable are comparable to the ef-
fluent  of wastewater treatment plants with secondary treatment
facilities.  Since  the  processes  tested   are  all  conventional
technology,  the equipment  required  for use is  commercially
available.
  Based on  the results  of each test included  in this treatability
study,  a combination of anaerobic and aerobic treatment in series
alum coagulation could  achieve more than 99% removal of COD
from raw leachate.

 REFERENCES
 I. Boyle. W.C. and Ham, R.K., "Biological Trealabilily of  Landfill
   Lcachatc." JWPCI. 46. 1974, 860.
 2. Cook, O.M. and force, G.G., "Aerobic Biostabilization of Sanitary
   Landfill Leachate," JWPCI.  46,  1974, 380.
 3. Uloih, V.C. and Max ink, "Aerobic Biotreatment of a High Strength
   Lcachalc," J. Environ.  EnK. Dt\  . ASCE. EE4, 1977, 841.
 4. Chain, E.S.K. and DcWalle,  F.B.. "Evaluation of  Leachate Treat-
   ment." U.S EPA Report, Washington, D.C.
 5. Zapf-Gilzc,  R.. "Temperature Effects on Biostabilization of Leach-
   ate," J. Environ. Eng. Di\'.. ASCE. EE4, 1981. 653.
 6. Ehrig, H.J. and  Slegcman.  R..  "Treatment of Sanitary  Landfill
   Leachate:  Biological Treatment." Waste.
 7. Force, E.G. and Reid, V.H.,  "Anaerobic Biological Stabilization of
   Sanitary Landfill Leachatc," Technical Report URY TR 65-73-CG17,
   1973.
 8. Wu, V.C.  and Kennedy. J.C.,  "Anaerobic  Treatment of Landfill
   Leachate by an Upfloxx Two-Stage Biological Filter," Proc. Fixed-
   Film Bioiechnol. Seminar, 1984.
 9. Ho, S.,  Boyle. W.C.  and Ham. R.K..  "Chemical Treatment  of
   Leachate from Sanitary Landfill." JWPCF. 46. 1974, 1776.
142   TREATMENT OF HAZARDOUS WASTES

-------
                            On-Site  Versus  Off-Site Treatment
               of Contaminated  Groundwater —  An Evaluation
                            of Technical  Feasibility and Costs
                                              Kent L. Bainbridge
                                          Daniel  W. Rothman, P.E.
                                              URS Company,  Inc.
                                              Buffalo, New York
ABSTRACT
  The Pollution Abatement Services, Inc. (PAS) site is an aban-
doned hazardous waste incinerator site  located in the City of
Oswego,  New York. Groundwater contamination has  resulted
due to leakage from a  surface impoundment, buried tanks and
drums—both above and below ground.  The principal  ground-
water contaminants include volatile organic compounds, organic
acids and other low molecular weight organics, and nickel, which
also appears to be organically bound.
  Groundwater recovery and leachate collection  were included in
the site remediation plan. Alternatives considered  for treatment
of the groundwater/leachate were: (1) hauling to an off-site com-
mercial treatment facility, (2) conveyance to the local POTW and
(3) on-site treatment. The relative merits of these alternatives were
evaluated on the basis of the  degree of treatment technically
achievable and the relative  cost.
  Critical to the evaluation was the identification and demonstra-
tion of technology capability for treatment of the groundwater/
leachate. Initial laboratory  treatability studies demonstrated that
biological treatment followed by carbon adsorption was required
for cost-effective treatment, while carbon adsorption alone was
not effective. Further studies demonstrated the feasibility of the
treatment system on an intermittent basis—a necessary require-
ment based on the expected low rate of groundwater/leachate
recovery.
  In this analysis, the costs and ultimate selection of a  ground-
water/leachate treatment alternative were sensitive to several fac-
tors including: (1) the wastewater character, (2) the wastewater
volume and rate of recovery and (3) the degree  of treatment re-
quired. On-site treatment was  offered a  higher  degree of treat-
ment than would be achievable at the local POTW and a lower
cost than would be expended for hauling to an off-site commer-
cial facility.

INTRODUCTION
  The Pollution Abatement Services, Inc. (PAS) site, located
near the eastern limit of the  City of Oswego, New York (Fig. 1), is
an abandoned hazardous waste facility which is listed among the
top 10 priority sites on  the  U.S. EPA's National Priorities List.
During the period from 1982-84, URS Company, Inc. was con-
tracted by the New York  State Department of Environmental
Conservation (NYSDEC to perform a Remedial Investigation/
Feasibility Study (RI/FS) of the site. One result of the field in-
vestigation was the determination that the groundwatei underly-
ing the site and leachate taken from two on-site drainage ditches
was highly contaminated by organic priority pollutants.
  Hydrogeological data collected during the RI/FS supported the
conclusion that all subsurface contaminant  migration was being
                        Figure 1
            Location of the Pollution Abatement Site
intercepted by two creeks which bounded the site on three sides.
Based on these findings, the RI/FS report recommended specific
remedial measures for subsurface cleanup and closure of the PAS
site. These measures included construction of a leachate collec-
tion system, installation of a groundwater recovery system and
treatment of the groundwater/leachate.1
  As part of the subsequent Remedial Design Study, laboratory
treatability studies  were  conducted to  identify  process  re-
quirements to  determine the level of treatment attainable and to
develop design and cost data. Results of the treatability studies
are reviewed in this paper. Estimates of the cost of on-site treat-
ment also are presented. Finally, the on-site and off-site treatment
methods are compared to identify the most cost-effective treat-
ment alternative.
                                                                           TREATMENT OF HAZARDOUS WASTES    143

-------
SITE HISTORY
  The facility known as Pollution Abatement Services,  Inc.
(PAS) was constructed and put into operation in 1969-70. A high
temperature, liquid chemical waste incinerator was the principal
unit installed by the owners. Throughout its active life, PAS ex-
perienced  continuous operating  problems, numerous air and
water quality violations and mounting public opposition.
  During its operating period (from 1970 through 1977), large
numbers of drums containing various chemical wastes were col-
lected and stored on-site. Liquid wastes also were  stored  in
lagoons.
  Beginning in 1973, a series of incidents including liquid waste
spills and overflowing of lagoon wastes into the adjacent White
Creek led to the involvement of the U.S. Coast Guard, the U.S.
EPA and  the  New York State  Department of Environmental
Conservation  (NYSDEC). This involvement included a number
of limited and temporary remedial actions during the period from
1973-1976. In 1977, PAS was abandoned.
  During the several years immediately following abandonment,
a number of emergency remedial  actions and preliminary in-
vestigations were undertaken at the PAS site including draining
and filling the waste lagoons and removing approximately 3,000
leaking barrels and several waste storage tanks from the site. Late
in 1981, contract documents for the surficial cleanup of PAS were
prepared,  including  the  demolition  and disposal  of on-site
facilities,  the  removal of the  remaining (approximately  8,000)
drums from the site and the drainage and disposal of approx-
imately 80,000 gal of liquid chemical waste from 10 bulk storage
tanks. This surficial cleanup was completed in October 1982.
  In November 1982,  URS Company, Inc. (URS) entered into a
contract with NYSDEC to evaluate alternative remedial measures
for final cleanup of buried wastes, contaminated soil and ground-
water and site closure.

TREATMENT ALTERNATIVES
  Alternatives considered for  treatment of the  contaminated
groundwater/leachate were: (1) hauling to an off-site commercial
treatment facility, (2) treatment at a local POTW and (3) on-site
treatment. Two commercial hazardous waste disposal facilities
are  located approximately 180  miles from the PAS site.  The
Oswego East Sewage  Treatment Plant,  which provides primary
and secondary treatment, is located just 1 mile from the site.
  A preliminary estimate indicated that on-site treatment would
be less costly than off-site disposal at a commercial facility. For
this reason, laboratory treatability studies were conducted: (1) to
determine  treatment process  requirements, (2) to measure the
degree of treatment achievable and (3) to develop design and cost
data.

TREATABILITY STUDIES
  Available  hydrogeological   data indicated  that  the  strata
underlying the PAS site would be low yielding and, on this basis,
it was estimated that groundwater pumping would  yield only ap-
proximately 24,000 gal per each  1-3 week period. Since a 24,000
gal basin  remains on the site and could be used for pumped
groundwater storage, it was anticipated that the treatment system
would be operated on an intermittent basis with treatment being
performed whenever  the  storage basin was  full. A chemical
analysis of this contaminated groundwater is given in Table 1.
  As  part  of the  PAS  remedial design project,  laboratory
wastewater treatability  studies  were conducted  to  determine
specific treatment process requirements and  capabilities. These
studies have been described previously.'  The  key findings are
summarized below:

• Only a minor fraction ( <1%) of the wastewater total organic
  carbon (TOC) was removed by air stripping. This result indi-
                          Table 1
                    Wutewaler Analysis'
    HetK,W CMoridf
    Acetone
    trjns-I ,?-Olchloroeth«nr
    ?-Butinonr
    flenjene
    4-Methyl -?-Penunone
    Toluene
    fot«l I/lenet
    Phenol
    Aniline
    2-Hetnylphrnol
    4-Melhylphfnol
    2,4-OI«thylphenol        ,
    Tentatively Identified luo/l)
    [thy] Bentene
    ?-3-Olethyl Oilrine
    4.Methy1-?-Pent«nol
    Hrtn/l Sen/me
    N.N-Olethyl Forvumldr
    Bulinolc Add
    l-Hethoiy  Cthinol
    f thylbenjene
    I,4-Olmethyl Benrene
    Pennnolc  Acid
    2-tthyl Butinolc Acid
    1-Methyl-?-PyrrolIdlnon*
    N.N-Dlnelhyl Benien««lne
    ?-fthyl Heitnolc Acid
    2.3-OlMlhyl Phenol
    Beniene Acetic Acid
    3-Mrlhyl Beniolc Acid
    Convention!! PjrMgte
    Sul'ite
    Phosphate
    H-NH3
    TICK
    «U«llnlty  -
    H«rdnesi K« )
    Iron
    Klckel
    COO
    TOC
    BOO.
    IDS5
Concentration

   11.000
   42.000
    3.500
   14,000
    I .WO
   17.000
    4.300
    1.600
    5.700
    5.600
    2.300
   15.000
    3,400

     620
   28.000
    6,400
    8.400
    5,400
   14,000
    2.900
     940
    2.400
    6.300
    2.300
    7.MO
    4.700
    4.200
   20.000
    1,700
    5. POO

      12
      5.5
      80.7
      80.7
    1.500
    2.150
      96
      2.6
    1900
    1020
    1100
    6400
 Samples analyzed were 50:30 mixtures of kachate.
 A complete analysis of the US EPA Contract Lab Protocol Hazardous Substances was per-
 formed. Only the hazardous substances found at concentrations above analytical detection
 limits are reported in this table In addition, the ten nonhazardous lubstance list compounds
 present in each fraction (volatile, acid and base.'neutraf) at greatest concentration were tenta-
 tively identified and quantified. These compounds are also reported in this table.
 All concentrations in jig/ unless otherwise indicated.
  cated that very little of the TOC consisted of volatile organic
  contaminants.  Therefore, air stripping was determined to be
  unnecessary.
• Nickel was not removed by hydroxide precipitation but was re-
  moved by activated carbon adsorption. From these data, it was
  concluded that nickel was organically bound or complexed.
• Adsorption of the organic contaminants by activated carbon
  was enhanced at acidic pH values.
• Wastewater COD was reduced  from a concentration of 4,000
  mg/1  to approximately 450 mg/1 (89°7o removal) by carbon
  adsorption at pH 3.
• Carbon requirements per 24,000 gal batch of wastewater were
  computed to be  6,030 Ib based on an effluent quality  of 800
  mg/1 COD (i.e., 20% breakthrough of the influent COD con-
  centration). This usage is equivalent  to a carbon loading of
  0.106 Ib COD/lb activated carbon.
• With a period of 1 week acclimation,  biotreatment achieved a
  COD removal of 85%. Rate studies showed that  this level of
  treatment could be achieved in less than 8 hr retention time. A
  final effluent quality of less than 50 mg/1 COD was achieved by
  treating the bioreactor effluent with activated carbon.
  On the  basis of the batch bioreactor studies, we concluded that
a treatment process consisting of biological treatment followed by
activated  carbon  polishing  was feasible, would be more cost-
effective and would produce a much higher quality effluent than
carbon adsorption alone. For these reasons,  further laboratory
studies were subsequently performed to  develop the  necessary
design and cost data for a batch biological process to be operated
intermittently. The results  of  these studies, which have been
reported previously,' are summarized below.
144    TREATMENT OF HAZARDOUS WASTES

-------
Pretreatment Requirements
• The PAS waste has no toxic or inhibitory effect on biological
  treatment.  Consequently,  pretreatment of the wastewater to
  remove substances (such as iron) potentially inhibitory to the
  biological treatment process is unnecessary.
• Approximately 8 hr of aeration were required to achieve opti-
  mum COD removal at 25 °C.
• No difference in bioreactor performance was observed when
  wastewater feed volumes in the range of 10-70% of the total
  reactor volume were used.
Nutrient Requirements
• The addition of nutrients  (i.e., nitrogen and phosphorous) to
  the wastewater is not required to achieve optimum biological
  removal of COD.
• The addition of nitrogen and  phosphorous actually resulted in
  a higher total suspended solids concentration.
Effect of Extended Lags Between
Bioreactor Operation
• Little difference was observed in the COD removal efficiency
  of  the  batch bioreactors when operated  with idle periods of
  1-3 weeks.
• The rate of COD removal is inversely proportional to the length
  of the idle period.
• The cycle time required to achieve optimum COD removal in-
  creased from 8-11 hr when  the bioreactor was operated at
  1 week idle periods, to 25 hr when the idle period was 3 weeks.
Effects of Temperature on Bioreactor Performance
• During the initial  weeks of bioreactor operation, the effluent
  COD was inversely proportional to the bioreactor temperature.
• After several weeks of intermittent operation (1 week idle per-
  iod), the effluent COD level achieved at the different tempera-
  tures (i.e., 25 °C,  15 °C and 5°C) began to converge.
• Bioreactor performance as measured in terms of COD removal
  rate and efficiency is  adversely impacted by cooler tempera-
  tures. The exact temperature at which performance begins to
  degrade is not known, but lies somewhere between 5°C and
  15 °C.
Effects of Powdered Activated  Carbon
Addition to the Bioreactors
• Powdered  activated carbon (PAC) addition substantially en-
  hances COD removal rates under conditions of extended bio-
  reactor idle periods and low temperature.
• COD removal efficiency increases as the PAC dosage is in-
  creased.
• The time of  PAC addition during the reactor cycle does not
  affect the effluent  quality  achieved.
• When PAC is added to the bioreactor, the COD removal me-
  chanism involves simultaneous adsorption and biodegradation.
Oxygen Supply Requirements
• Oxygen uptake rates are greatest in the initial phases of the
  reactor cycle  and decrease rapidly with time.
• As  the length of the idle period between bioreactor react per-
  iods is increased, the rate  of  oxygen uptake during the initial
  phase of the react  cycle decreases.
Sludge Production
• The net biological  sludge production in sequenced batch bio-
  reactors operated at 25-50% fill ratios was minimal or negative.
• Total bioreactor sludge production was found primarily to be
  a result of inorganic  precipitates.  This production  averaged
  800 mg/1 solids of wastewater treated in bioreactors operated
  at 50% fill, nonpretreated  and 1 week idle periods.
• The use of PAC  in the bioreactor increased the amount of
  sludge produced in direct proportion to the PAS dosage.
Activated Carbon Column Studies
• Results of batch adsorption  isotherm  studies indicated very
  little difference in adsorption characteristics at pH 3 and 7.5.
  Based  on this result, no pH adjustment of bioreactor effluent
  is necessary to achieve optimum COD removal by carbon col-
  umn polishing.
• The carbon inventory  requirement for a granular  activated
  carbon column is a function of the allowable breakthrough
  level. Effluent from a reactor operated at a 50% fill and 1 week
  idle period was polished to determine carbon requirements.
  For breakthrough levels of 20, 30 and 40% of the carbon col-
  umn influent, carbon requirements  of 2,670, 1,460 and 700 lb/
  24,000 gal of wastewater treated were calculated.
• Approximately  95% COD removal was achieved using a se-
  quenced batch  bioreactor followed by carbon column pol-
  ishing.
• Essentially complete removal of priority  pollutants was
  achieved by biological treatment followed by granular activated
  carbon adsorption.

DEGREE OF TREATMENT ATTAINABLE
  The effluent quality obtainable is a function of the raw waste-
water concentration, the operating conditions (length of idle time,
temperature, etc.) and the specific treatment process employed. A
summary of the effluent quality achieved, in terms of commonly
used measures, is  presented in Table 2.

                          Table 2
             Summary of Achievable Effluent Quality

Sample Identification
Raw Wastewater
B1o eactor Effluent ^
1 React cycle/wk
1 React cycle/2 wks
1 React cycle/3 wks
15°C
1 react cycle/wk
• 5°C
1 react cycle/wk
• 1000 mg/1 PAC
1 react cycle/wk
t 2000 mg/1 PAC
1 react cycle/wk
( 5000 mg/1 PAC
1 react cycle/wk
• 1 React cycle/wk
GAC column polishing
Temp
(°C)
..

25
25
25
15

5

25

25

25

25

COD
(mg/1)
1900-2700
.
360-410)51
385-410 ,{
370-4401''
450-550

600-750

150-200

110-140

80-90

100

TOC
(TO/1)
1020

140
128
128
__

.-

58

37

24

--

BODc
(mg/n
1050-1130

12-22
17
--

--

1-5

2-4

1-2

--

TSS
(mg/IJ
-.

62
99

262

69

65

57

--

   20* breakthrough v°'

Notes:
(1) For bioreactors operated at 50% fill.
(2) Equilibrium COD.
(3) Bioreactor effluent contained 465-580 mg/1 COD.
COST OF ON-SITE TREATMENT
ALTERNATIVES
  Based on the treatability study results, seven treatment alter-
natives were identified for cost analysis. The central process in
each of these schemes is biological treatment in a batch reactor.
  These seven alternatives vary primarily with respect to the size
of the activated carbon column units incorporated and the quan-
tity of PAC used. These alternatives are identified below. For our
purpose, carbon breakthrough was defined as the time at which
the COD concentration in the carbon column effluent reached
100 mg/1.

Treatment Processes
• A batch biological reactor followed by a granulated  activated
  carbon (GAC) system sized for 30 min retention time at a flow
  rate of 100 gal/min (System I-A). (Carbon breakthrough ap-
  proximately every 4th batch  treated.)
                                                                                TREATMENT OF HAZARDOUS WASTES    145

-------
• A batch biological reactor followed by a GAC system sized for
  30 min retention at a flow rate of 27 gal/min  (System I-B).
  (Carbon breakthrough approximately every batch treated.)
• A batch biological reactor followed by a GAC system sized for
  60 min retention at a flow rate of 100 gal/min  (System I-C).
  (Carbon breakthrough approximately every 8th batch treated.)
• A batch biological reactor with 1,000 mg/1 powdered activated
  carbon (PAC) addition and a GAC system sized  for 30 min re-
  tention at a flow rate of 100 gal/min (System II-A). (Carbon
  breakthrough approximately every 9th batch  treated.)
• A batch  biological reactor with 1,000 mg/1 PAC addition and
  a GAC system sized for 30  min retention at a flow rate of
  11  gal/min (System II-B). (Carbon breakthrough approxi-
  mately every batch treated.)
• A batch biological reactor with 1,000 mg/1 PAC addition and a
  GAC system sized for 60 min retention at a  flow  rate of 100
  gal/min  (System  Il-C).  (Carbon breakthrough approximately
  every 18th batch  treated.)
• A batch  biological reactor with 5,000 mg/1 PAC addition and
  no GAC system (System III).

Cost Estimates
  Cost estimates were based on the following assumed operating
conditions:
• Bioreactor volume (approximately  50,000 gal) is  twice the feed
  volume to accommodate a 50:50 ratio of mixed liquor to  waste-
  water feed.
• One batch (24,000 gal) of wastewater is processed each  week.
• Groundwater/leachate is pumped directly to  the bioreactor so
  that the bioreactor is filled and ready for the reactor cycle to
  begin when the operator arrives.
• Twelve hours aeration per batch.
• Two hours settling time per batch.
• Discharge time varies according to GAC system.
• A treatment system operator is available locally  on an on-call
  basis.
• Minimum effluent quality required is 100 mg/1 COD.
• GAC column breakthrough is  20<%  of COD influent (500 mg/1
  COD reduced to 100 mg/1 COD)  for systems not  using PAC
  (Systems I-A, I-B and I-C).
• GAC column breakthrough is  40%  of COD influent (250 mg/1
  COD reduced to  100 mg/1 COD) for systems  using  1,000 mg/1
  PAC addition (Systems II-A,  II-B and II-C).
• The treatment  system is housed in a temperature-controlled
  building.
  Information regarding reactor size, sludge production, carbon
requirements and oxygen  requirements that affected capital and
operating costs was determined on the basis of the  treatability
study results. Capital costs were derived on the basis of vendor
quotes and annually published cost manuals.4'5

COST OF OFF-SITE ALTERNATIVES
  The cost of contract hauling and disposal  at  a  commercial
facility is a function of the  particular firm  that  is  contracted.
Costs for this alternative were based  on quotes from  three com-
mercial waste management firms located within 180 miles of the
PAS site.
  No basis could be established for deriving the cost of treatment
of PAS groundwater/leachate at the local POTW. The city does
not have  a  surcharge system  for  treatment  of  high-strength
wastewater. Furthermore, this POTW has a history of operating
problems.  To alleviate this problem, the city  has established a
policy that no "scavenger wastes" will  be accepted at the plant.
Due to this circumstance, no further consideration was given to
treatment at the local POTW as an off-site alternative.
COMPARISON OF ALTERNATIVES
  A summary of the capital and annual operation/maintenance
(O&M)  costs for each  on-site and off-site treatment alternative
considered is presented in Table 3. These costs were developed on
the  assumption  that  PAS  groundwater/leachate would  be
recovered and treated at a rate of one 24,000 gal batch per week.
The actual  rate at which the groundwater/leachate will  be
recovered is not known.

                           Table 3
             Cost Summary of Treatment Alternatives
Treatment
Alternative Description
Biological Treatment &
GAC Column Pol ishing:
A I-A
B 1-8
C I-C
Capital
Cost (S 1986)(1)

564,700
500,000
707,700
Operating
Annual Operating Cost Per
Cost ($ 1986)1Z) Gallons (S 1986)13'

155,800
179,000
163,800

0.125
0.143
0.131
     Biological Treatment with
     1,000 mg/1 PAC Addition J
      GAC Column Polishing:

D           II -A
E           II-B
F           II-C

G    Biological Treatment with
      5,000 mg/1 PAC Addition

H    Contract Haul ing/Of f-si te
            Disposal
                            589,100
                            506,200
                            723,900
                            515,100
115,500
120,900
127,200
                                   499,200 - 1,123,000
                                        (549,0001
                                                 0.40 - 0.90
                                                  (0.44)
Notes:
(1) Cost for facilities to treat wastewater in 24,000 gal batches.
(2) Cost to treat 24,000 gal/week.
(3) Cost/gal based on 1,248 million gal/yr (i.e., 52 wks x 24,000 gal/wk).


  Consideration of the quantity of groundwater/leachate to be
treated with time is essential to a comparison of the costs of the
various treatment alternatives, since these costs are a function of
both the quantity  of wastewater to be treated and the length of
time over which treatment is to be performed. The total quantity
of groundwater/leachate requiring treatment has been estimated
to be 3 million gallons based  on hydrogeological data collected
during the RI/FS and the Remedial Design Study. To address the
various rates at which this wastewater may be recovered,  four
scenarios  have been developed (Table 4).

                           Table 4
Hypothetical Rates of Groundwater/Leachate Recovery and Treatment
                                       Remaining years
                  (1)
Scenario
1
2
3
4
Year
1
1
1
1
Quantity
1 MG
1 MG
1 MG
1 MG
Cycles
41.7
41.7
41.7
41.7
Period
8.8 days
8.8 days
8.8 days
8.8 days
Years
2-3
2-5
2-7
2-10
Ouan
1 MG
0.5
0.33
0.22
tity
/YR
HG/YR
MG/YR
MG/YR
Per
41.
20
13
9
Year
.7
.8
.9
.2
Per-
8.8
17.5
26.3
39.4
iod
days
da v*.
davi
flays
(1) Assumes 24,000 gal per cycle.
  A present worth analysis of  each  treatment  alternative as a
function of the four recovery rate scenarios is given in Table 5.
The following observations are based on a review of this analysis:
• All on-site treatment alternatives (A through G) are more cost-
  effective than the off-site alternative (H).
• On-site alternatives which incorporate the smaller sized GAC
  columns  (B,E)  are more cost-effective than alternatives  in-
  corporating larger columns  (A,C,D,F).
 146
       TREATMENT OF HAZARDOUS WASTES

-------
                          Table 5
       Present Worth Analysis of Treatment Alternatives
(1) (2)
Treatment
Scenario Alt.
1 A
B
C
0
E
F
G
H
2 A
B
C
D
E
F
G
H
3 A
B
C
D
E
F
G
H
4 A
B
C
D
E
F
G
H
(3)
Years
of
Treatment
3
3
3
3
3
3
3
3
5
5
5
5
5
5
5
5
7
7
7
7
7
7
7
7
10
10
10
10
10
10
10
10
(4)
Capital
Cost
(S*10>)
564.7
500.0
707.7
589.1
506.2
723.9
515.1
t
564.7
500.0
707.7
589.1
506.2
723.9
515.1
B
564.7
500.0
707.7
589.1
506.2
723.9
515.1
0
564.7
500.0
707.7
589.1
506.2
723.9
515.1
f>
(5)
1st Yr.
Operating
Cost ($xlOJ)
125.0
143.0
131.0
92.5
96.9
102.0
87.4
400.0
125.0
143.0
131.0
92.5
96.9
102.2
87. 4
440.0
125.0
143.0
131.0
92.5
96.9
102.0
87.4
440.0
125.0
143.0
131.0
92.5
96.9
102.0
87.4
440.0
(6)
Remaining Yrs
Operating
Annual Cost
Ax(ixlO')
125.0
143.0
131.0
92.5
96.9
102.0
87.4
440.0
62.5
71.5
65.5
46.2
48.4
51.0
43.7
220.0
41.2
47.2
43.2
30.5
32.0
33.7
28.8
145.2
2;. 5
31.5
28.8
20.4
21.3
22.4
19.2
96.8
(7)
Total
Present
Uorth
(5x10=)
875.8
855.9
1033.8
819.3
747.4
977.6
732.7
1095.9
858.3
835.9
1015.4
806.3
733.7
963.5
720.4
1033.9
841.5
816.9
997.9
794.0
721.0
950.0
708.6
975.5
822.1
794.8
977.4
779.9
705.7
933.8
695.0
906.8
Notes:
Column 1: Scenario (Refer to Table 7-2)
Column 2: Treatment Alternative (Refer to Table 7-1)
Column 3: Total number of years that a treatment system is operated on-site
Column 7: Calculated on basis of:
  Present worth of capital and annual O&M at year n
  = (P/A, i, n) x (O&M) + Capital

  = (1 + i) n  - 1
              x O&M + Capital
     i (i + i)n
 where i = interest rate per year
     n = number of years
• Biological treatment enhanced by a 5,000 mg/1 dose of PAC
  (Alternative G) is more cost-effective than any of the alterna-
  tives (A through F) which include GAC columns  for effluent
  polishing.

CONCLUSIONS
  Based on laboratory treatability studies, biological treatment
with activated carbon polishing was determined to be the most ef-
fective treatment method. These studies  also demonstrated the
feasibility of using a batch  bioreactor operated intermittently.
The disadvantage of lower COD removal rates when the bioreac-
tor was operated at low temperature and increased idle periods
was overcome by the  addition  of powdered activated carbon.
Results of a present worth analysis showed that on-site treatment
will be much more cost-effective than off-site treatment/disposal
at a commercial facility.

REFERENCES
1. "Site Investigations and  Remedial Alternative  Evaluations at the
   Pollution Abatement Services (PAS) Site in Oswego, New York,"
   Draft Final Report, prepared for New York State  Department  of
   Environmental Conservation by URS Company, Inc., Jan. 1984.
2. "Engineering Design Report for Site Remedial Measures Pollution
   Abatement Services Site,  Oswego,  New York,"  prepared for New
   York State Department of Environmental Conservation by  URS
   Company, Inc., May 1985.
3. "Evaluation of Alternatives  for Treatment of Groundwater/Leachate
   at the  Pollution Abatement Services (PAS) Site in Oswego, New
   York," Final Report, prepared for New York State Department of
   Environmental Conservation by URS Company, Inc., Oct. 1985.
4. Building Construction Cost Data 1985, published by the Robert Snow
   Means Company.
5. Dodge  Manual for Building Construction, Pricing and Scheduling,
   published by McGraw-Hill Information Systems Co., Vol. 2.
                                                                                    TREATMENT OF HAZARDOUS WASTES    147

-------
                         Optimization of  Free  Liquid  Removal
                Alternatives  in the  Closure  of  Hazardous Waste
                                      Surface  Impoundments
                                                David K. Stevens
                                                 Black & Veatch
                                              Kansas  City, Missouri
ABSTRACT
  A flowchart is presented to serve as a guide in determining the
optimal alternative for removal of free liquids during the closure
of hazardous waste surface impoundments. The flowchart ad-
dresses each part or element  within a closure alternative as a
discrete decision process. This flowchart allows separate evalua-
tion of each phase of removal such as preclosure engineering,
removal and treatment and ultimate disposal.  Criteria used to
select  the best  option within  an element include technological
feasibility, environmental soundness, regulatory acceptance and
unit cost. Once the best elements for each phase of free liquid
removal are determined,  they are recombined to form the optimal
removal alternative.
  Important factors such as in situ reduction and provisions for
mixed  wastes are discussed. Removal equipment and  treatment
processes are presented as a function of the waste type. Ultimate
disposal options such as discharge to surface water vs. discharge
to sewer, on-site vs. off-site treatment and disposal and process
reuse  and recovery are  included. To illustrate the use of the
flowchart, three  case studies of actual  surface impoundment
closures are  presented. Optimal free liquid removal alternatives
are developed for impoundments containing heavy metals, heavy
metals and degreasers, and oily wastes.

INTRODUCTION
  There are tens of thousands of hazardous waste lagoons, pits or
ponds  in  the United States.' These  types of storage facilities or
surface impoundments are used for continuous or intermittent
storage of process water, stormwater and spent products from the
chemical and manufacturing industries. RCRA  or CERCLA can
require closure of an impoundment when soil and groundwater
contamination is discovered or if the impoundment poses an en-
vironmental  risk. Stricter liner requirements for impoundments
under RCRA are also forcing an increased number of closures.
Given the trend toward greater environmental protection, increased
enforcement  and stricter regulations, it is certain that  impound-
ment closures and their associated remedial actions will continue
at a rapid pace.
  Regardless of the RCRA or CERCLA regulations  governing
closure, removal of free liquids is required as one of the first steps
in an impoundment closure. The removal process can present a
number of technological and regulatory  problems as  a  large
volume of liquid hazardous waste must be removed from the site
which frequently  has no suitable direct outlet. The free liquids
often  have a high solids content, exhibit extremes of pH and
viscosity, have concentration gradients and are stratified mixtures
of both inorganic and organic chemicals. These  technological
problems combined with the regulatory pressures for tight closure
schedules and strict air and water discharge limits require careful
engineering of the free liquid removal processes. Recognizing the
inherent problems in dealing with liquid hazardous wastes in sur-
face impoundments, the flowchart directs the user in the selection
of the optimum free liquid removal alternative.

FLOWCHART DESCRIPTION
  In assembling the  flowchart elements into a system  of logical
steps, the universe of processes used for free liquid removal were
identified. Once the available processes were identified, they were
categorized into three elements: preclosure engineering; removal
and treatment; and ultimate disposal.  Figure 1 shows the basic
outline of the flowchart. To determine the optimal removal alter-
native, each process within an element is subjected to a set of deci-
sion criteria developed from recent CERCLA guidance1  and ac-
tual closure experience.  Through evaluation of each process ac-
cording to the developed decision criteria, the best option for that
process can be identified.  By  combining the best options within
each element, the optimal removal alternative is determined.

Flowchart Elements
   The flowchart contains  three  major  elements:   preclosure
engineering; removal and treatment; and ultimate disposal (Fig. 1).
Pre-closure engineering addresses the  processes  for flow diver-
sion, waste characterization and the removal alternative's impact
CLOSURE
DECISION
PRE-CLOSURE ENGINEERING
  •  Flow Div0nion
  •  Waste Characterization
  •  Facility Cloaure Plan
                   REMOVAL AND TREATMENT
                    •  In Situ Reduction
                    •  Pumping and Dredging
                    •  Treatment
                     ULTIMATE DISPOSAL OR
                       REUSE/RECOVERY
                         •   On Site
                         •   Off Site
                            IMPOUNDMENT
                              READY FOR
                            FINAL CLOSURE
                         Figure 1
             Overview of Flowchart for Determining
              Optimal Liquid Removal Alternatives
148   TREATMENT OF HAZARDOUS WASTES

-------
 on the overall facility closure plan. The removal and treatment
 element is composed of three processes: in situ reduction, liquid
 removal and treatment prior to ultimate disposal. In situ reduc-
 tion considers whether the hazard or volume of the free liquid
 should or can be reduced while in the impoundment before actual
 removal.  The removal section addresses the equipment used in
 removing  the waste, such  as  pump  type  and construction
 material. Treatment alternatives include physical and chemical
 treatment of the free liquids for reduction or elimination of the
 hazard from the waste before ultimate disposal. The treatment
 process and the ultimate disposal option often are synergistic, re-
 quiring concurrent evaluation.  The ultimate  disposal  element
 reviews the options for final deposition of the liquids such as sur-
 face water, publicly owned treatment works or an RCRA facility.
   Careful consideration always should be given to those processes
 which allow reuse or  recovery of the liquids.

 Decision Criteria
   The decision criteria employed for evaluation of each closure
 element are expressed in both cost and non-cost factors. Cost fac-
 tors  include  the  cost  for  present  value of  operating  and
 maintenance costs and the capital expenses. Non-cost factors in-
 clude  technological  feasibility,  environmental  soundness  and
 regulatory acceptance. Technological feasibility requires that the
 technology under consideration be proven at the design scale or
 that pilot-scale results show a good chance for success in the field.
 Environmental soundness is a measure of the exposure of the
 chemical constituents to the environment during the removal
operation. This decision can be a judgmental one as in some cases
environmental exposure occurs only for a short time and does not
persist once the removal is complete. Regulatory acceptance is the
third criterion. This factor indicates whether the closure processes
under consideration are acceptable to the governing  regulatory
agency.  Acceptance or  non-acceptance depends on the pertinent
regulations and the perceived  environmental risk.

PRECLOSURE ENGINEERING
  During development  of an RCRA closure plan or a CERCLA
feasibility study,  several preliminary concerns must be addressed
before the  actual free  liquid  removal alternatives can be fully
developed. Among these concerns are options for flow diversion,
requirements for complete waste characterization and the re-
moval alternative's impact on the overall impoundment closure
plan.  In the flowchart, Figure 2, these concerns are taken into
consideration in the preclosure engineering element. Carefully ad-
dressing each of these concerns is required for optimization of the
free liquid removal alternative.

Flow Diversion
  The initial step in implementation of an impoundment closure
plan is to eliminate the source of incoming waste. This is not a
major concern for inactive impoundments or impoundments used
for low-volume intermittent waste disposal. At active impound-
ments receiving  continuous or  semi-continuous waste streams
such as  process water or contaminated stormwater, careful con-
sideration is required as dedicated treatment plants and long-term
operation become potential design factors.
CLOSURE
DECISION
     PRE-CLOSURE ENGINEERING
                                                                        GENERATED SLUDGE DISPOSAL



HAZARD


REDUCE
HAZARD
- Biological
• Clu-oncnl
• Physical


REMOVAL AND
VOLUME
i 1
REDUCE
J VOLUME
• PhvMC.ll
1

TREATMENT
REMOVE
LIQUIDS
j
PR E TREATMENT


CHEMICAL
PHYSICAL
TREATMENT





ULTIMATE D
ONSITE


DISPOSE ONSITE
• Reuw; —
• Recovery
• Land Apply


ISPOSAL
DISPOSE
• POTW
• Surfiicu WuU-r
• RCRA Facility
	 1
                                                                                                                   CONTINUE
                                                                                                                   CLOSURE
                                                           Figure 2
                                                 Free Liquid Removal Flowchart
                                                                                TREATMENT OF HAZARDOUS WASTES    149

-------
Waste Characterization
  Accurate information on the chemical constituents in the im-
poundment is essential, including identification of any variances
in the composition as a function of depth. Each layer can require
special consideration and different removal procedures; for exam-
ple, a floatable oil layer can cover a relatively clear aqueous layer
which in turn covers a gelatinous metal hydroxide sludge. The im-
portant physical parameters include pH, viscosity and solids con-
tent. The pH influences pump selection and treatment processes.
Viscosity and solids  content influence the pump type and the
volume of liquid available for removal by pumping alone. Sludges
with solids content greater than 8-10%  must be removed either by
first mixing with  the  other impoundment  liquids  to  lower the
solids content, or by  leaving the high solids content sludge in
place and dewatering by evaporation or by the  addition  of a
solidifying agent.

Impact on the Overall Closure  Plan
  Consideration of the free liquid removal alternative's impact on
the overall closure plan is imperative for optimizing the impound-
ment closure.  Very often other  remediation  activities at  the facil-
ity can  influence selection of the  free liquid removal alternative.
These considerations  are  site-specific and are best illustrated
through examples. Table 1 shows examples of the impact of the
selected closure plan on the free liquid removal alternative.

                           Table 1
                 Impact of Facility Closure Plan
               on Free Liquid Removal Alternatives
Waste Type
Heavy Metals
Oily Wastes
PCB Oils
Facility Closure Plan
Groundwater Treatment
Program
Reclamation of Wastes
Soil Incineration
Free Liquid
Removal Alternative
Same Treatment
Process Employed
Free Liquids Re-
claimed by Same
Process
Free Liquids
Incinerated
REMOVAL AND TREATMENT
In Situ Reduction
  In situ reduction is a process within the treatment and removal
element (Fig.  2). It involves the reduction of the hazard or the
volume of the waste present in the impoundment before actual
removal. Such techniques can be used to facilitate final removal
and disposal of  the waste.
  In dealing with very corrosive wastes such as spent pickle liquors
with pH C2 or caustic washes with a pH > 12, it is advantageous
to add a neutralizing agent to the wastes in the impoundment. Ad-
dition  of bases also can be used  to remove toxicity such as the
alkaline precipitation of heavy metals at pH 9-10. In small acidic
ponds,  a 50% solution of sodium hydroxide is best as it  is easily
pumped at moderate ambient temperatures and offers a  high
degree of neutralizing power.  For larger acidic ponds or when
cost is an important factor, calcium hydroxide and calcium car-
bonate  should be examined. Multi-stage treatment with several
neutralizing materials has been investigated.' For alkaline ponds,
concentrated sulfuric acid is generally best as it is inexpensive, of-
fers a  high degree of neutralizing ability  and does  not pose  a
serious  material  handling problem. Other possibilities for in situ
hazard  reduction are the use of oil booms, reduction/precipita-
tion techniques  and activated carbon or  polymer addition for
organics adsorption.
  The other in situ alternative is to reduce the volumetric quantity
of free liquids  in the impoundment. The two primary mechanisms
are evaporation and  seepage.  Metal-containing  aqueous wastes
are usually more amenable to evaporation or seepage than organic
wastes. Although a limited number of sites actually experience a
high net  evaporation rate,  judicious scheduling  of the  liquid
removal step during periods of low rainfall and high temperature
can effectively reduce the total volume of free liquids. For exam-
ple, at  a West Coast manufacturing facility the volume of water in
a stormwater retention pond  varied from 3.5 million gallons dur-
ing January  to 2.5 million gallons during August. Determining
whether seepage is a viable method of volume reduction is  a less
straightforward  calculation  than determining  the evaporation
potential.' The use of impoundments as seepage basins for con-
centrating wastes has been documented to be very effective with a
limited impact on ground water quality.'

Free Liquid Removal
  Physical removal of the free liquids  is accomplished best by
pumping. If the waste has a high solids content,  mechanical, hy-
draulic or pneumatic dredging  may be required.  This  paper
discusses only pumping of free liquids.  Selection parameters in-
clude the type, construction materials  and size  of pumps. The
pump  type and construction  materials usually  depend on the
waste  material and its  physical  characteristics. Pump size  is
governed by the closure schedule and treatment system capacities.
Table 2 delineates the generally available pump types and con-
struction materials as a function of the waste type.

                            Table 2
             Pump Types and Construction Materials
                  as a Function of Waste Type
                                                                   Physical Description of Waste
                             Pump T)pe
Aqueous
Vi.scous or Oily
Sludge. 1-5% solids
Sludge. 5-8% solids

Chemical Description
of Waste

Aqueous
Oils, phenols
Solvents
Heavy metab
Basic pH >I2
Acidic pH < 2
                                                                                                Centrifugal
                                                                                                Rotary lobe, progressive cavity
                                                                                                Centrifugal
                                                                                                Posiihe displacement


                                                                                                Construction Material
Cast iron
Cast iron, possibly bronze fittings
Bronze fitted cast iron
Stainless steel
SS ASTM 743. CN-7M
SS ASTM 743. CN-7M (3«*o Mo)
Treatment
   At this stage in the evaluation process, the free liquid treatment
and disposal processes are developed. Selection of the treatment
process  is  based on  both  the waste  characteristics  and  the
availability  of ultimate disposal mechanisms. Relationships  be-
tween the disposal processes and treatment requirements are sum-
marized in Table 3.
                            Table 3
 Relationship Between Disposal Processes and Treatment Requirements
 Disposal Processes

 Off-site disposal
 On-site disposal
 On-sile reuse or recovery

 Publicly owned treatment works
 Process reuse

 Surface water
 Groundwatcr rcinjection
   Treatment Requirements

   Usually requires no treatment
   except suspended solids removal


   Suspended solid removal often
   required

   Treatment to drinking water
   standard required
 150    TREATMENT OF HAZARDOUS WASTES

-------
  Since the treatment process and the ultimate disposal method
represent the largest single capital cost items involved in free liq-
uid removal,  all alternatives need to be critically reviewed. Con-
sideration should be given to alternatives which provide a poten-
tial for reuse or combination with other treatment needs.
  Processes available for the treatment of aqueous hazardous
waste are numerous.6 The major limitation to the use of many of
these processes is that many are not cost-effective or not proven at
the field scale. New technologies and innovative processes are best
explored when a long-term use treatment plant is required, such
as in groundwater remediation programs or surface and process
water treatment systems. In other cases, leasing of mobile treat-
ment units or  low-cost  temporary equipment  should be con-
sidered. Table 4 shows the technologies available for treatment as
a function of the waste type.

                           Table 4
        Treatment Technologies as a Function of Waste Type
Waste Type
Acidic pH<2
Basic pH >12
Heavy metals
Low level
organic
solvents
High level
organic
solvents
Oil
Dedicated
On-site
Equipment
Alkali addition
Acid addition
Precipitation/
decollation
Ion exchange
Adsorption
Air stripping
Oxidation
Carbon adsorption
Incineration
Separators
Incineration
Filtration
Mobile
Equipment
alkali addition
Acid addition
Precipitation/
flocculation
Adsorption
Air stripping
Oxidation
Carbon adsorption
Incineration
Separators
Incineration
Filtration
Temporary
System
alkali addition
Acid addition
Precipitation/
flocculation
Adsorption
Carbon adsorption
Separators
Filtration
ULTIMATE DISPOSAL
Direct Off-Site Disposal
  In direct off-site disposal of free liquids, the liquids are taken
directly from the impoundment to an off-site facility. Potential
off-site facilities  include publicly owned  treatment works and
RCRA treatment facilities. If the concentrations of the hazardous
compounds are within sewer discharge limits, the liquids can be
pumped directly  into the sewer system. If the sewer line is not
within  economical  pumping   distance,  alternative  means  of
transportation to the treatment plant must be explored.  A most
practical means is to employ vacuum trucks or similar tanked
vehicles. Transportation costs  can be as much as $1.50 per
unloaded mile and $3.50 per loaded mile.  Sewer discharge rates
vary widely depending on  waste type,  quantity and treatment
facility. When the pollutant concentrations exceed the influent
pretreatment standards and direct off-site  disposal is a potential
alternative, a list of available facilities should be compiled. Many
available directories list RCRA  facilities, and the RCRA hotline
also can provide a list of currently permitted facilities. Transpor-
tation costs are similar to the amounts given above. Disposal costs
can  be from $0.10/gal  for low level contaminated  aqueous
streams to over  $l/gal for liquid injection of  highly  organic
wastes. Deep well injection, generally for  acids and bases only,
costs approximately $0.20/gal.

Treatment Followed  by Off-Site Discharge
  Employing a pretreatment process expands the number of off-
site disposal alternatives  to include surface water discharge and
groundwater injection. Groundwater injection is predominantly
used only in conjunction with other groundwater remediation
programs. Discharge to surface water requires an NPDES permit
as well as very low contaminant concentrations. For metals, some
type of filtration will be required to attain the discharge limit.

On-Site Disposal
  The alternatives to on-site disposal include construction of an
on-site RCRA permitted facility. For most closures at industrial
manufacturing  facilities,  the closed  system with  reuse and
recovery is the most advantageous. At uncontrolled sites, on-site
disposal of the liquids usually is confined to treatment processes
such as landfarming, evaporation or incineration.

GENERATED SLUDGE DISPOSAL
  Final consideration for free liquid alternatives is given to the
disposal of sludges generated  during a treatment process. Sludge
is generated during alkaline precipitation of metals, filtration or
solidification/stabilization processes.  Potential  disposal  alter-
natives include off-site disposal in an RCRA permitted landfill,
on-site disposal in a specially constructed RCRA permitted land-
fill and disposal in the closed impoundment.

CASE STUDIES
  Three case studies  of actual surface impoundment closures are
presented here to illustrate some of the aspects in determining the
optimal alternative for free liquid removal. The following cases
are discussed:
• Stormwater retention pond  containing hexavalent and trivalent
  chromium
• Process water storage lagoon contaminated with chlorinated
  solvents and nickel
• Earthen pit used for oily waste disposal

Chromium Contamination
  A manufacturing facility employed an earthen pond to store
stormwater runoff from its product storage area. The pond was
being closed because of its ineffective liner and the presence of
chromium in the underlying groundwater.

Preclosure Engineering
  The pond currently was being used for storage of stormwater,
and  an alternative mechanism Tor handling future runoff was
needed. Since manufacturing and storage operations were to con-
tinue at the site and no practical method of eliminating the con-
taminated runoff existed, a permanent treatment and disposal
solution was required. An ion-exchange treatment system with
surface water discharge was determined to be the best alternative
for addressing the runoff. Ion-exchange produced a very low level
concentration of chromium in the effluent and allowed reuse of
the recovered chromium. Ion  exchange also added the advantage
that no hazardous waste sludge was generated during treatment as
would have  been  produced using traditional chemical reduction
and precipitation.
  Results  of the pond sampling activities  showed that approx-
imately 400,000 ft 3  of free  liquids containing  5-10 mg/1 hex-
avalent chromium and 1-5 mg/1 trivalent chromium were in the
impoundment. Underlying the free liquids were 1000 ft3 of a
chemical  floe  sludge containing 1000 to  5000  mg/1 total
chromium. The-low chromium concentrations in the free liquids
permitted  use of the  current ion-exchange system for free liquid
treatment. The remaining sludge could be solar dried and fixed in-
place using a solidification agent.  Use of the on-site treatment
system was an attractive solution  for treatment as there was no
negative impact on the other closure activities and no additional
permitting was required for surface water discharge.
                                                                                 TREATMENT OF HAZARDOUS WASTES    151

-------
Removal,  Treatment and Ultimate Disposal
  Given the overwhelming advantages  in using the on-site ion-
exchange treatment system, the free liquids were removed using
an available contractor's pump, treated to NPDES requirements
and  disposed  via the force main to the nearby surface water
stream. Judicious closure scheduling during the summer months
reduced the volume of free liquids from 400,000 ft3 to 260,000 ft3.
This timing reduced pumping time from 14 days to 9 days and
saved approximately $20,000 in treatment expense.
  During the alternative review phase, contractor proposals were
solicited for removal and treatment of the free liquids. The con-
tractor proposed to treat the liquids using sulfite reduction fol-
lowed by  alkaline  precipitation.  For comparison purposes, the
decision criteria evaluation of the two treatment alternatives  is
shown below:
lowered  below  discharge requirements.  Different  treatment
schemes  were developed and compared using the four decision
criteria. The results are summarized below.
Alternative
Contractor
Bid
Use of On-site
Treatment
System
Technology
Feasibility
Feasible
Feasible
Environmental
Impact
Sludge
Generated
Process Reuse
Regulatory
Acceptance
Acceptable
More Favorable
Acceptance
Unit Cost
S/1000 gal
22.33
5.50
Degreasers and Contamination
   A metal plating and fabrication facility used an earthen im-
poundment for disposal of process rinsewaters.  Groundwater
contamination  by the degreasers was forcing closure of the im-
poundment.

Preclosure Engineering
   Due to the continuation of fabrication operations at the plant,
a permanent treatment solution was needed for the waste process
rinsewaters. Through water-use reductions and process modifica-
tions,  the  rinsewater flow rate  was reduced  from 300  to  50
gal/min. A 50 gal/min treatment system was put on line early dur-
ing the closure plan development. Treated rinsewater was dis-
charged to nearby surface water per the NPDES permit require-
ments.
   The impoundment contained approximately 130,000 ft3 of free
liquids, 8,000 ft3 of pumpable sludge and 2,000 ft3 of unpump-
able sludge. The nickel and degreaser concentrations in the free
liquids ranged from 30 to 60 mg/1 and 0.2 to 0.9 mg/1, respective-
ly. Both sludge layers contained higher levels of contaminants
than the free liquids, with the unpumpable sludge layer contain-
ing the bulk of nickel and degreaser.
   Due to the depth of degreaser contamination in the soil, the im-
poundment was closed with contaminated soils in place. Labora-
tory studies were conducted on both sludges to determine if they
were  amenable  to  stabilization.  If adequately  stabilized,  the
sludges could be disposed of by leaving them in  the impound-
ment. Tests showed that  the addition of portland cement effec-
tively formed a stable and solid mixture. Using the stabilization
process, only the free liquids needed to be removed and disposed;
the pumpable and unpumpable sludges were stabilized and left in
place.

Removal, Treatment and Ultimate Disposal
  The volume of free liquids in the impoundment precluded the
use of the process water treatment system; the time required for
treatment using the system would exceed the required 180-day
schedule. An alternative mechanism was required. Due to the low
level of contaminants in the free liquids, consideration was given
to the use of a  temporary on-site system consisting of filtration
and possibly  pH adjustment. Laboratory tests confirmed that
with pH adjustment  and settling, the contaminant levels were
Alternative
Coagulation
& Settling
Pressure
Filtration
Gravity
Filtration
Technology
Feasibility
Feasible
Feasible
Feasible
Environmental
Impact
Minimal
Minimal
Minimal
Regulatory
Acceptance
Acceptable
Acceptable
Acceptable
Unit Cost
5/1000 gal
11.25
8.75
7.50
  It was decided that gravity sand filtration would produce the
best quality effluent at the lowest cost. To enhance operation of
the filtration equipment, care was taken to maintain a low level of
suspended  solids in  the  free liquids.  A  floating intake was
employed to minimize disturbance of the  underlying sludge.  In
addition, when the suspended solids concentration exceeded 400
mg/1, the pump speed was decreased to further minimize resus-
pension. Backwash  water was  held  on-site, and the settled
material was returned to the impoundment after all the free liq-
uids were removed.

Oily Wasles
  An earthen pit located on a farmer's  property had been used
for the disposal of spent cutting  and hydraulic oils. Closure was
proceeding as a generator-directed and funded mitigation action.
The site functioned with little control, and  many accounts of un-
authorized  dumping had been  recorded.  However, one major
generator had  taken prime responsibility for closure instead of
waiting  for a potential enforcement action.

Preclosure Engineering
  The impoundment was receiving intermittent waste streams only,
and additional waste inputs were eliminated by banning all future
waste disposal practices. Results of the pond sampling activities
showed  that three distinct layers were present: an oil layer of ap-
proximately 10,000  ft 3, an aqueous  layer of 60,000 ft3 and a
viscous oily sludge layer of 4,000 ft3. The aqueous layer contained
between 0.2 and 0.5% total hydrocarbons  and could not be dis-
posed as a non-hazardous waste. However, analytical analysis of
the oily  layer showed that it potentially could be reclaimed.

Removal, Treatment and Ultimate Disposal
  In this closure, several site-specific constraints precluded many
of the normally available options; no surface water or sewer line
was readily available. Due to the uncontaminated nature of the oil
it allowed the potential for recovery. The oil layer was removed by
vacuum trucks and taken to an oil recovery  plant for recox cry and
reuse. Since no outlet  for  the treated  aqueous layer was readily
available,  an  off-site  disposal  alternative  was  required.  The
aqueous layer was removed by vacuum  trucks and taken to the
generator's manufacturing facility where it was treated  in their
own industrial wastewater treatment plant. Although the trucking
costs were substantial, the oil recovery and generator treatment of
the aqueous waste offset a portion of this cost. It was determined
that disposal of the viscous sludge was accomplished best by land-
filling rather than by incineration, since incineration was an order
of magnitude in cost higher than landfilting.

CONCLUSIONS
  A flowchart has been presented which guides in determining the
optimal alternative for removal  of free  liquids from  hazardous
waste surface impoundments. The flowchart addresses each ele-
ment of the closure process as a separate  entity, whereas in-
 152    TREATMENT OF HAZARDOUS WASTES

-------
dividual decision criteria are applied to each element before com-
bining the elements to form the optimal alternative. The major
elements include: preclosure engineering; removal and treatment;
and ultimate disposal.
  General trends identified from the case studies presented in this
paper are:

• Identify site-specific constraints early during closure plan  de-
  velopment. The combination of free liquid treatment and dis-
  posal options with soil and groundwater remediation programs
  can substantially reduce the cost for free liquid removal.
• Use recovery and reuse whenever possible; this alternative is at-
  tractive environmentally and to regulatory agencies.
• Employ temporary on-site treatment measures when levels of
  contaminants are low. This process is less expensive than off-
  site disposal at RCRA facilities.
• When  available, use  on-site process water or  groundwater
  waste treatment equipment to treat the free liquids.
REFERENCES
1.  Wyss, A.W., et al., "Closure of Hazardous Waste Surface Impound-
   ments," prepared by Acurex Corporation for the U.S. EPA, Sept.
   1982.
2.  "Guidance on Feasibility Studies Under CERCLA," prepared by JRB
   Associates for the U.S. EPA, June 1985.
3.  Hale, F.D., Murphy, C.B. and Parrat, R.S., "Spent Acid and Plat-
   ing Waste Surface Impoundment  Closure,"  Proc.  Third National
   Conference on Management of Uncontrolled Hazardous Waste Sites,
   Washington, DC, Nov. 1982, 195-201.
4.  McWhorter,  D.B., "Seepage in the Unsaturated Zone: A Review,"
   Proc. Seepage and Leakage from  Dams and  Impoundments, May
   1985, 200-219.
5.  Looney, B.B., "Surface Impoundment Legacy: Field Studies," paper
   presented at the  Hazard  Materials Management Conference, June
   1985.
6.  Sittig, M., "Pollutant Removal Handbook,"  Noyes Data Corpora-
   tion, Park Ridge, NJ,  1973.
                                                                                 TREATMENT OF HAZARDOUS WASTES     153

-------
                Assessment  of Chemical  Treatment  Technologies
                     and  Their Restrictive Waste  Characteristics

                                             Hamid Rastegar, Ph.D.
                                                 James Lu, Ph.D.
                                                   Chris Conroy
                                        Jacobs Engineering Group, Inc.
                                               Pasadena, California
ABSTRACT
  The applicability of a particular  treatment technology  to a
given liquid waste stream depends upon the physical, chemical
and biological characteristics  of the waste. Waste stream char-
acteristics which  restrict a waste's  applicability to a particular
method of treatment are presented.  With this information, the
most effective treatment technology for a waste stream can be
determined. The  technologies evaluated include chemical oxida-
tion  and reduction, chlorinolysis, chemical fixation,  ion ex-
change, neutralization and precipitation.

INTRODUCTION
  The land disposal of many hazardous wastes may be banned
within the coming years under the RCRA amendments. For those
wastes which will be banned  from landfill disposal, it is neces-
sary to determine alternative methods of treatment and disposal.
  A  wide variety of treatment technologies exists today which
can be used to treat such wastes.  Treatment technologies can be
classified as either chemical, physical, biological or thermal treat-
ment. In this paper,  the authors evaluate the applicability of
chemical  treatment technologies  for  hazardous liquid  waste
streams based on  the wastes restrictive characteristics.
  This analysis is part of an extensive study by the U.S. EPA to
determine the applicability of alternative waste treatment  tech-
nologies. The most promising types of each chemical treatment
technology are discussed. The major restrictive waste characteris-
tics determining  treatment applicability  for each also  are pre-
sented. Finally, the final product or residue from each treatment
technology and the limitations of each are evaluated.

CHEMICAL TREATMENT TECHNOLOGIES
  Several chemical treatment technologies are  widely used for
waste treatment.  The most effective technologies discussed here
include:

  Chemical Oxidation
  Chemical Reduction
  Chlorinolysis
  Fixation
  Ion Exchange
  Neutralization
  Precipitation

  The major restrictive waste characteristics as well as the type of
applications for each of the above technologies are given below.
A summary of the major restrictive waste characteristics is  given
in Table 1.
Chemical Oxidation
  Chemical oxidation is a process which raises the oxidation state
of chemical species, converting the wastes into less-hazardous or
non-hazardous  forms.  The oxidation  is accomplished by  the
addition of an oxidizing agent which is reduced.
  Chemical oxidation has been used most commonly to treat cya-
nide wastes from electroplating operations. The most widely used
oxidants  for this application are chlorine gas and calcium and
sodium hypochlorite.
                          Table 1
   Chemical Treatment Technologies Restrictive Waste Characteristics
Chemical Treatment
Technology
Major Restrictive Waste
Characteristics
Chemical Oxidation

Chemical Reduction

Chlorinolysis

Fixation

Ion Exchange


Neutralization
Precipitation
Physical form, oil and grease content, sus-
pended solids content, viscosity.
Physical form, oil and grease content, sus-
pended solids content, viscosity.
Total solids content, oxygen, sulfur, and
organic content, water content.
Physical  form,  organic  content,  ionic  com-
position, sulfate content.
Suspended solids content, metal content,
organic content, oxidizing agents, physical
form.
Dissolved solids content, physical form.
Physical form, viscosity, metal solubility.
  Other oxidants for cyanides include ozone, hydrogen peroxide,
potassium permanganate and chlorine dioxide. A wide variety of
organics and inorganics also can be treated by oxidation. The
end products of cyanide oxidation are nitrogen gas and carbon
dioxide. The end products of the oxidation of organics are either
simpler organics or carbon dioxide and water; oxidized metals
form metal hydroxides which can be precipitated out of solution
by pH adjustment.
  The  major restrictive characteristics of a waste which are used
to determine the applicability of chemical  oxidation include the
physical form, oil and grease content, suspended solids content
154    TREATMENT OF HAZARDOUS WASTES

-------
and viscosity of the waste. Chemical oxidation is most effective
for dilute aqueous wastes, although certain gaseous wastes also
can be treated. Since oxidizing agents will indiscriminantly attack
all oxidizable material, the oil and grease content and the oxidiz-
able  suspended solids content should be low,  preferably below
1%.  For liquid solutions,  viscosities much greater than that of
pure water are restrictive  because of the mixing problems en-
countered with highly viscous solutions.

Chemical Reduction
  Chemical reduction lowers the oxidation state of chemical spe-
cies,  converting  them  into less-hazardous  or non-hazardous
forms. The reduction is accomplished by the addition of a reduc-
ing agent which is oxidized.
  Chemical reduction is used most  commonly to  convert hexa-
valent chromium into trivalent chromium, which then is precip-
itated out of solution as chromium hydroxide. Mercury also can
be removed from solution by reducing mercury (II) to its elemen-
tal state. Sulfur dioxide and sulfite salts are the most widely used
reducing agents  for chromium. Sodium borohydride is an effec-
tive reducing agent for mercury and lead.
  The major restrictive waste characteristics for chemical reduc-
tion include  the wastes' physical  form, oil  and grease content,
suspended solids content and viscosity. These are the  same re-
strictive characteristics as for chemical oxidation, since chemical
oxidation and reduction are similar processes.

Chlorinolysis
  Chlorinolysis involves the reaction between chlorine and liquid
chlorinated hydrocarbons  at temperatures of 500 °C or greater
and high pressure to form carbon tetrachloride, hydrogen chlor-
ide and other by-products. In the Chlorinolysis process, the waste
first is  pretreated to remove all solids  and  any unwanted com-
pounds which could result in the formation of unwanted reac-
tion by-products. Then, the purified waste undergoes reaction
with chlorine, and the reaction products are separated and recov-
ered by distillation.
  The waste to be treated by  Chlorinolysis ideally should con-
tain only liquid chlorinated hydrocarbons. The presence of other
compounds may produce toxic reaction by-products, which may
include carbon tetrachloride, hydrogen chloride and phosgene.
The  restrictive waste characteristics include the solids oxygen,
sulfur organic and water content. As mentioned previously, the
presence of compounds other than  chlorinated hydrocarbons is
unwanted.

Fixation
  Chemical fixation is a process in  which hazardous substances
are stabilized and solidified in a form which is considered non-
hazardous and safe for ultimate disposal. Various types of ma-
terials can be mixed with a waste to form a solid material which
effectively immobilizes  the waste. The process is intended to be
permanent, allowing the safe disposal of the solidified mass to a
landfill or other permanent disposal facility.
  Fixation is used for both inorganic and organic solids and
sludges. It is possible to treat liquids, although the effectiveness
of this is less certain. Fixation also is used for low-level radioac-
tive waste. Fixation of inorganic wastes  is accomplished by solid-
ifying the waste in a cement, pozzolanic or lime based material.
Fixation of organic wastes usually requires the use of a glassifica-
tion, organic polymer or thermoplastic technique.
  The major restrictive  waste characteristics which determine the
applicability  of  fixation  include  the  wastes'  physical   form,
organic content, ionic composition and  sulfate content. Fixation
is most effective on solids,  and sludges.  The greater the moisture
content of the waste, the more difficult  fixation is, although
liquids can be treated. A high organic content may limit or cause
problems when fixating an inorganic waste, because the organics
may interfere with the solidification of the solid product. Simi-
larly, high ionic compositions and high concentrations of sulfate
interfere with stabilization and solidification.

Ion Exchange
  Ion exchange is the reversible interchange of ions between a
liquid and a solid phase. Ions in a liquid waste stream and ions on
the surface of an ion exchange resin are exchanged, purifying the
waste stream while concentrating the waste constituent on the
resin.
  Ion exchange resins are either cationic  or  anionic. Cationic
resins exchange positive ions such as H+  or Na+  for negative
ions in solutions such as SQ-3. Anionic resins  exchange negative
ions such as OH - and Cl -  for positive ions in solution, such as
metal cations.
  After an ion exchange  resin  has  reached its capacity for ex-
changing ions, it is  regenerated so that it can be used again.
Cationic ion  exchangers are regenerated by passing dilute acid
solution over the resin; anionic resins are regenerated with a weak
base. During the regeneration  process,  a solution containing a
high concentration of the  original waste constituent is produced.
This concentrated waste then must be treated or disposed of prop-
erly.
  The restrictive  waste characteristics for  ion exchange include
suspended solids, metal content and organic contents, oxidizing
agents concentration and physical  form.  Ion  exchange  is most
effective for very dilute aqueous wastes. High concentrations of
solids, organics and oxidizing agents interfere with the exchange
of ions between the solution and the  resins.

Neutralization
  Neutralization involves  the addition of an acid or  a base to a
solution to adjust the pH, usually to between  6 and 9. Neutral-
ization often is performed  in connection  with other  treatment
processes, such as oxidation or reduction. It can be either a final
treatment or a pretreatment step.
  Alkaline neutralization  is  commonly accomplished using lime,
sodium hydroxide or soda ash. Acidic  neutralization  utilizes
hydrochloric, sulfuric or nitric acid.  Neutralization is most applic-
able for acidic or alkaline aqueous wastes, although it also can be
used  for certain  organic  liquids. Neutralization is  a relatively
simple but important waste  treatment process  and most  often is
followed by precipitation  of metals  from solution. This technol-
ogy can be used  effectively for most aqueous  wastes, although
wastes with very high dissolved solids content may result in the
formation of complexes which are difficult to remove from solu-
tion.

Precipitation
  Waste removal  by precipitation involves the addition of a
chemical substance to alter the equilibrium affecting dissolved
and suspended solids solubility. Precipitation  normally requires
pH adjustment, followed by the addition of a precipitating agent,
resulting in flocculation and  coagulation of the solids in solution.
  Precipitation is used most commonly to remove heavy metals
from aqueous wastes. The precipitation process produces a sludge
composed of metal hydroxides, metal carbonates or metal sul-
fides  as well  as the precipitating agent used. In some instances,
precipitation can be used for organic-based liquids, although this
application is very limited due to sedimentation problems in vis-
cous media.
                                                                                  TREATMENT OF HAZARDOUS WASTES    155

-------
   The major restrictive waste characteristics limiting the use of
precipitation include voscosity, physical form and metal solubility
constant. As mentioned before, highly viscous  solutions, such as
many organic solvents, may  not be applicable for treatment by
precipitation because of difficulties in  settling the metals or other
particles out of solution. For highly soluble species, pH adjust-
ment is needed to minimize solubility and enhance precipitation.
CONCLUSIONS
  The applicability of seven chemical treatment technologies for
hazardous  liquid waste treatment has  been examined.  It was
found that  the  proper choice of treatment depends  upon  an
understanding of  the  restrictive  waste  characteristics  for each
technology. Once the waste characteristics which restrict the use
of a particular treatment technology are  known, proper pretreat-
ment or other treatment methods can be chosen.
156    TREATMENT OF HAZARDOUS WASTES

-------
                Treatment Technologies  for Hazardous  Materials
                                                 Jeffrey M. Thomas
                                                    DETOX, Inc.
                                             San Francisco,  California
                                                  Phillip T. Jarboe
                                                    U.S. Ecology
                                                Louisville, Kentucky
ABSTRACT
  The nature and quantities of hazardous materials and the
various technologies for treating them have grown rapidly in re-
cent years. The efficiency and cost-effectiveness of technologies
such  as  physical/chemical  treatment,  metal  precipitation,
biological treatment  and carbon adsorption, vary enormously
depending on the application requirements. In most instances,
these treatment processes have been utilized to treat large volumes
of mixed matrix wastes.
  Transferring these technologies for  effective use in ground-
water treatment, with relatively low flow volume and a range 'of
contaminants from a single substance to a complex mixture, re-
quires special considerations in the design and operation of  a
treatment program. The process  design for these  remedial pro-
grams must utilize the optimum mix of available technologies to
generate the highest degree of treatment at a reasonable cost.
  This paper evaluates the process design  requirements and
economic considerations for application of these processes to
groundwater  treatment. It  also examines the unique design and
operating conditions encountered in groundwater treatment. The
concept of "life-cycle"  design is addressed to demonstrate the ef-
ficiency of optimizing  the  particular attributes  of the different
technologies available.

INTRODUCTION
  There is a  broad range of potentially applicable technologies
for the elimination of  hazardous materials.  However, for each
material a specific review must be performed to select the most
appropriate form of treatment. During this review, consideration
must be given to the type of the contamination, organic versus in-
organic; the extent of the contamination, surface soil contamina-
tion alone, vadose zone, groundwater  or any  combination of
these; also to be considered are the physical and chemical char-
acteristics of the contaminants, the relative concentration of the
contaminants and the interactive effects the contaminants may
have with each other. In most instances after careful evaluation, it
will be found that a mix of appropriate technologies can be imple-
mented to produce the most efficient, cost-effective solution to
treat the contamination.
  Most of the conventional treatment technologies are intended
for use on waterborne contaminants present in fairly constant
concentrations. One of the characteristics of most groundwater
contamination problems is their relatively short term nature (2-5
year treatment programs). These treatment programs initially will
be treating a relatively high concentration of contaminants. With
time,  however, the concentration  of  these contaminants will
decrease.  This concentration decrease is due to the removal of
chemicals as a result of the treatment, as well as the dilution effect
of clean groundwater passing through the zones of contamination.
  As contaminated water is withdrawn from the ground (in either
the aquifer or vadose zone) clean water will either be drawn in
from the surrounding areas of the aquifer or, in the case of the
vadose zone, clean water used for flushing the contaminants will
displace the chemical compounds present in the interstitial spaces.
The net effect is a continuous, sometimes dramatic, decrease in
contaminant concentrations. This changing chemical composition
creates difficult design requirements for remedial actions.  The
treatment programs implemented must be flexible enough to ac-
commodate the  high concentrations  initially encountered as well
as successfully treat the lower concentrations near the completion
of the program. In fact, most programs will see the highest con-
centrations of contaminants withdrawn over a relatively short
time period during the initiation of the treatment program. This is
a most important  consideration  as the efficiency  and  operating
costs of the treatment process will be seriously affected.

TREATMENT TECHNOLOGY REVIEW
  There are several types of treatment processes which can be ap-
plied to the remediation of groundwater contamination. These
can be divided into three categories:  physical/chemical treatment
of inorganic constituents; physical/chemical treatment of organic
materials and biological treatment of organic compounds.  It is
important to understand the difference between the use of in situ
treatment as a technique for decontaminating groundwater and
soils and the use of process systems for the same purpose. When
using in  situ  techniques,  materials  (oxidants, microorganisms,
etc.) are introduced into the treatment area via wells, borings or
percolation beds. Under these conditions, there is  relatively little
control over the reactions or the environmental conditions under
which they are occurring. Analysis of the performance of these
techniques is difficult to ascertain and most certainly involves a
time lag between the start of a program and any result which may
be reached.
  In our approach to treating contaminated groundwaters and
soils, the contaminants and contaminated waters  are withdrawn
from the ground and treated in surface reactors. This is the case
for all the technologies discussed  in this paper. In most instances,
there is a requirement for pumping liquid into  or out  of the
ground. Since the surface reactors allow  a much greater degree of
control over the  process and better measurement of the results, we
find this to be much more appropriate for the treatment of  con-
taminated materials. In some cases, the combination of both tech-
niques will yield the most benefit and should be employed.

PHYSICAL/CHEMICAL
TREATMENT—INORGANIC
  The  inorganic compounds  most  frequently encountered in
groundwater contamination are  heavy  metals. These materials
                                                                              TREATMENT OF HAZARDOUS WASTES    157

-------
usually  are not  particularly mobile in the environment. Most
metals exist as a cationic species and have a tendency to adsorb
onto soil particles. The characteristics of some soils also allow an
"ion exchange" process  to occur, thus  removing the metals from
groundwater passing through a formation.

Hydroxide Precipitation
  The metals which are found in extracted groundwater can be
treated most effectively through precipitation processes.  Most of
the heavy metals are insoluble as a metal hydroxide or as a metal
sulfide.  If they are not in a chelated form, the metals can quite
readily  be removed by  using pH adjustment and  creating the
metal hydroxide. Typically, the minimum  solubility of most
heavy metals is in the pH range of 7.5 to  9.5.
  The steps necessary to determine metal removals involve ex-
perimentally determining the solubility of a metal (or metals) at
various  pH levels by neutralizing a sample of the contaminated
water using lime or caustic soda. An insoluble precipitate will be
formed  which can be physically removed, and the soluble concen-
trations of the metal(s) can be directly measured at each pH level.
  In  a  full-scale treatment process,  the  groundwater  will  be
withdrawn and neutralized to the optimum pH level. The metal
hydroxide precipitate will be removed from the liquid stream via
flocculation and sedimentation. The clarified, metal-free effluent
then can be discharged.  In many instances, this process  must be
performed prior to processing for organic compound removal.

Sulfide  Precipitation
  The process of sulfide precipitation of metals works in a similar
fashion. The minimum  solubility of any  given species will vary
somewhat from that  exhibited  during hydroxide precipitation;
however, the soluble concentration of the metals  after sulfide
precipitation will be considerably lower, as much as two  or three
orders of magnitude lower than with hydroxides. For most treat-
ment program requirements, the levels  which can be achieved via
hydroxide precipitation are more than  adequate.

Chrome Removal
  The process design for the removal  of  hexavalent chrome re-
quires an additional step  prior to the  precipitation process.
Chrome,  in  the hexavalent  form, will  not  precipitate as a metal
hydroxide. It must first  be reduced to  the trivalent state using a
strong reducing agent such  as sulfur dioxide or sodium bisulfite
under acidic pH conditions. The pH  then can  be  raised using
either lime or  sodium hydroxide and precipitation of the metal
will be effected.

System Design
  The design criteria  used in these  treatment programs usually
will follow those which have  long been  applied  in the  metal
plating and processing industries. The systems applied to ground-
water treatment should be designed with ease of installation and
operation as major considerations. Once the treatment process is
developed and implemented, there is generally little in the way of
process variables which need to be controlled. Automatic pH con-
trol systems can be used to control neutralization; fiocculant feed
rates need to be monitored and adjusted  periodically, but once
these systems are on line, there usually is  little which can be ad-
justed. The use of lamella separators for  liquid-solid separation
can substantially reduce the space required for this operation and
allow easy installation and operation of the system.
  The use of metal precipitation is the first step in processing con-
taminated waters containing both heavy metals and organic com-
pounds.  It is important to remove  the metals at  this point to
achieve  and  maintain high  degrees of treatment efficiencies in
subsequent groundwater treatment processes.
PHYSICAL/CHEMICAL
TREATMENT—ORGANIC
Downhole Recovery
  The category of physical/chemical treatment of organic com-
pounds is a broad one. The characteristics of groundwater con-
tamination are such  that in many instances  the compounds are
present in very high concentrations or as a pure material floating
on or settling below the groundwater. The first type of treatment
process to investigate would be recovery of pure product. Every
effort should be made to remove as much pure product as pos-
sible  using  either  downhole  recovery  systems  or  surface
separators. The nature of most  organic materials renders them
only  partially soluble  in  water,  and  many  compounds will
physically separate from water quite readily. There are a number
of downhole systems now available for recovering hydrocarbon
products from both the top and the bottom of a contaminated
aquifer.
  The use of simple separation equipment at the surface also can
be implemented. This may consist of rope-type skimmers or even
separator tanks using tubes and plates. The cost of the treatment
program is substantially reduced when relatively pure product can
be recovered as opposed to having to process the material. In
many instances, the recovered product will have some economic
value and may partially offset some of the treatment costs.
  There are also physical/chemical treatment processes for use on
low concentrations of organics in the liquid stream. Most notable
among these are air stripping,  carbon adsorption, peroxide oxida-
tion  and UV/ozone  treatment. These are the traditional  tech-
niques for  removing  (or  destroying) organic constituents  from
contaminated groundwater.

Air Stripping
  Air  stripping  of  organic  compounds  uses the  volatility
characteristics of certain  organic  species, such as  chlorinated
solvents, to eliminate them from an  aqueous stream. The con-
taminated groundwater is pumped to the top of a column which
contains an inert packing. The water flow is downward from the
top of the column against a countercurrent air flow from the bot-
tom of the unit. The air exhausted from the column contains the
volatile organic compounds and is simply exhausted from the air
stripping column via  exhaust  ports at the top of the column.
  Air stripping is  an  effective method for removing low concen-
trations of highly volatile compounds. The units are quite inex-
pensive to operate and require a minimum of operating attention.
There are, however, potential problems associated with their use
in groundwater treatment programs. Certain  compounds do not
have particularly high volatility characteristics and do not readily
strip from water.  In  some instances, this problem can be over-
come by increasing the temperature of the contaminated water,
thus increasing the volatility  of the  organics, but heating can
become quite expensive.
  Another consideration is the potential for fouling the internal
surfaces of the air strippers. The groundwaters which are heated
in the  strippers  generally will  contain  bacteria capable of
degrading the compounds present. Under the environmental con-
ditions found in an air stripper, it is possible to create substantial
biological growth  on  the internal packing surfaces as well as on
the mist eliminating packing at the top of the  towers. In some in-
stances,  this growth  can  lead  to excessive  maintenance  re-
quirements for the operating units.
  Another  major consideration when evaluating the use of air
stripping is the ultimate fate of the stripped compounds.  In the
past,  these  compounds have merely been  vented  to the  at-
mosphere. This is not truly a treatment technique for the elimina-
tion  of the compounds but rather a relocation of the problem
from  the  water to the air. There currently is a move by the
158    TREATMENT OF HAZARDOUS WASTES

-------
regulatory agencies in many areas to substantially reduce or elim-
inate the air emissions from these units. This control by air pollu-
tion authorities will  mean  implementing  complex and  expensive
emission control technology on the air streams exiting these units.

Activated Carbon
  Carbon adsorption has been used to purify water for over 100
years. It is based upon the natural process of adsorption—that is
the natural affinity of a gas or liquid to be attracted to and held at
the surface of a solid. This attractive characteristic is due to the
surface tension of the solid holding the molecules in water at its
surface. Activated carbon particles are characterized by very high
surface area to volume ratios. The carbon particle has a structure
with a series of pores whose diameters decrease as the pores ex-
tend into the particle so that the small molecules will migrate fur-
ther into a carbon particle  than a large molecule.
  The adsorption of the molecule onto the carbon surface is a
physical process consisting of three steps: (1) diffusion of the
molecules through the liquid phase to the carbon particle; (2) dif-
fusion of the molecules through the macropores to the adsorption
site; and (3) adsorption  of the molecule  to  the site. The
characteristics of the adsorbate molecule will determine the rate
of adsorption and, ultimately, the efficiency of the process.
  Carbon adsorption has gained a very favorable reputation as a
means of removing  a broad range of organic compounds  from
water streams. It is quite effective at removing low level concen-
trations  of  organics, particularly  chlorinated solvents. The  ef-
ficiency of this process varies enormously, from as low as 3<7o to
as high as 25% utilization  efficiency. Activated carbon is an ex-
tremely effective method for polishing treated effluents to remove
the last traces of contaminants.

Hydrogen Peroxide
  Hydrogen peroxide is used in waste treatment applications as
an  oxidizing agent. It is a moderately strong oxidizer of organic
compounds and is  quite capable of oxidizing a broad range of
compounds at almost any concentration. The reaction involves
the formation of a free radical hydroxyl group. Hydrogen perox-
ide is a highly reactive chemical which reacts quite readily with
most organic materials (except saturated alkanes). Most reactions
involving the use of peroxides require the use  of an iron catalyst
to effect the reaction.
  Peroxides can be  difficult to handle from a safety perspective
and are by nature unstable. Also, the reactions involved are quite
exothermic and can  create  significant hazards when used to treat
higher concentrations (>500 mg/1) of organic concentrations.
The most appropriate use for peroxide oxidation may be for the
treatment of highly biologically refractory materials, highly inter-
mittent waste flows  or extremely low concentrations of organics
in a polishing operation.

UV/Ozone Treatment
  The last major category of physical/chemical treatment tech-
niques  for organic compounds is UV/ozone. UV/ozone  treat-
ment technology seems to  offer a great deal of potential for the
treatment of  low  concentrations  of halogenated solvents. The
ozone is the source of a highly reactive free radical oxygen species.
This highly  reactive  free radical readily cleaves the double bonds
found in aromatic and unsaturated aliphatic compounds.
  UV  light  increases the reactivity of the target molecules, par-
ticularly those bearing carbon-halogen bonds. The wavelength of
the ultraviolet light is in the absorption wavelength of this bond
and excites  the bond to the point of breaking and cleaving the
halide from the carbon atom.
  Some of  the  limitations of this technology are:  the cost  in
generating ozone; developing efficient UV light systems; reducing
the necessary contact times in the reactors and thus the number of
lamps  and the electricity to operate them; and improving the
transfer of the generated ozone gas into the water stream.
  The  combination of UV and ozone is an excellent method for
treating extremely low levels of organics in water. This technology
is still  in its development phase and has not been used in more
than a few pilot applications.  One outstanding characteristic of
this technology and of the biological processes is that they both
are destructive technologies, converting hazardous organic com-
pounds into non-hazardous mineralized products of carbon di-
oxide and water.

BIOLOGICAL TREATMENT—ORGANICS
  The  use of the biological  process  in waste treatment involves
the creation  of a growth environment in which  bacteria are
capable of metabolizing  organic compounds for energy for cell
maintenance  and reproduction.  Essentially, this process is the
conversion of a soluble organic compound into  an  insoluble
organic (cell protein) which can be  physically  removed from  a
water stream. This technology is truly a destructive process, as the
end products of the biological process are carbon dioxide, water
and additional bacterial cells (sludge).
  In order to accomplish this reaction, it is important to provide a
proper  growth environment with  consideration of inorganic
nutrients (nitrogen and phosphorous), pH, dissolved oxygen (for
aerobic processes) and several other process  variables. The design
and operating conditions for a system of this type can be quite ex-
pensive using conventional activated sludge technology. Recent
developments in bioreactor and  process  technology have given
this process the opportunity to be one of the most  useful tech-
niques  for the treatment of hazardous materials.
  The use of biological treatment for degrading organic chemicals
has been in practice for some decades now.  The use  of this tech-
nology on hazardous materials, and in particular on groundwater
contaminants, has  not been  particularly  extensive until  very
recently.
Activated Sludge
  The  process  of activated  sludge  treatment  is an operator-
intensive  process which is  sensitive to variations   in  influent
characteristics and operating conditions and requires close, careful
operator attention. One of the key elements in activated sludge
treatment is  the  concentration  of   large  numbers  of
microorganisms in a reaction vessel that effectively use an organic
substance  for a  food source. In  order to concentrate these
organisms, liquid-solid separation must be  effectively and care-
fully controlled. The use  of fixed-film biological reactors instead
of the suspended growth activated sludge process has eliminated
the requirement for extensive operator attention and reduced the
critical nature of liquid-solid separation as  part of the process.

Fixed Films
  The  use of a fixed  medium surface allows  the  microbes  to
become  attached to  a  surface  over which their   food source
(substrate) passes. The organisms form a biological  film into
which  their substrate,  nutrients and oxygen  diffuse. In this
fashion, the organisms merely grow to a film thickness consistent
with the availability of substrate.  As the  concentration  of
substrate diminishes, the rate of  microbe growth decreases.
  In a submerged fixed-film bioreactor (rotating biological con-
tactor), the release  of  air in a  diffusion  system  beneath the
medium in the tank provides the means of  both oxygen transfer
and mixing energy. This mixing  energy creates a relatively high
velocity of air and  water moving past the  biological film. This
movement provides a means for shearing  the accumulated bio-
logical growth from the fixed-film and for preventing bridging of
biological growth and plugging within the medium.
                                                                                 TREATMENT OF HAZARDOUS WASTES     159

-------
DIFFUSION OF
k;) OXYGEN
;:7 NUTRIENTS
•../ SUBSTRATE

TO MEDIA SURFACE
j] THROUGH FILM
* \
] SLIME LAYER TOO
A THICK
i . OXYGEN AND
'•H SUBSTRATE DO
';;,.! NOT REACH INNER
j I/ FILM
•A ANAEROBIC STATE
'•''\ CAUSES SLOUGHING
** AT MEDIA FILM
f




b
\
}
(
\
)
n
SLOUGHED
\\ t. INTERFACE / PARTICLE TO
\NAEROBICBIOFILM \ / EFFLUENT
         PLASTIC MEDIA
                              ANAEROBIC LAYER
                  — INCREASING GROWTH	»

                           Figure 1
              Biological Film Growth and Sloughing
  This mechanism of sloughing (Fig. 1) of the biomass is essen-
tially self-regulating.  As the film becomes  thicker due  to  the
growth  of the microbes,  its ability  to  remain attached  to  the
medium is reduced. Since the velocity of the air and water mixture
remains constant,  the shear forces are constant. When the film
becomes  thick enough, the shear forces are greater than  the
adhesive forces and the film sloughs  off the  medium surface.
  The effective concentration of microorganisms within the reactor
tank would be the equivalent of 6,000-8,000 mg/1 of mixed liquor
suspended solids in an activated sludge system. These biotreaters
can be placed within a fairly small reactor volume and are quite
readily preassembled and installed at a treatment location.
  A minimum of operator expertise is required for these systems,
and  they are  becoming  increasingly  popular as  methods  of
destroying the hazardous materials. The biological process tends
to complete the degradation reaction, yielding carbon dioxide.
Perhaps 99%  of all known organics  can be effectively degraded
using biological treatment.

Slightly Contaminated Streams
  Recent  developments in  biological treatment technology now
enable low concentration waste streams to be biologically treated.
In the past, all biological reactors were designed using the growth
characteristics of the biological populations as the limiting factors
to their effectiveness.  It  was not considered possible to  use
biological processing on many wastes at a concentration contain-
ing less than 50 mg/1.  This limitation  is due to the nature of
biological growth dynamics, the effect of biological washout from
the reactors, loss of settling properties and other engineering con-
siderations. Low concentration reactors now have been developed
using acclimated populations for removal of low concentration
( <$ mg/1) contaminants to low fig/\ (1-3 /ig/1) effluent  levels.
These reactors are portable, require almost no operator attention
and appear capable of operating several months on the low in-
fluent concentrations before needing to be  reacclimated. This
technology probably will displace activated  carbon  in many in-
stances for the polishing of low level organic compounds.

Soil Treatment
  The use of  the biological process also has been  successfully
adapted to the treatment of contaminated soils. It is possible, using
the proper engineering designs, to create a growth environment
for microorganisms in a soil environment. This process requires the
same environmental conditions as the treatment of a wastewater
stream, but they can be created and maintained quite readily.
  The availability of commercial bacterial inocula has substan-
tially increased the ease with which the soils can be  treated,  and
these techniques offer substantial potential for eliminating a wide
range of soil contaminants such as pentachlorophenol, creosote,
hydrocarbons, pesticides and  many other types of compounds.
The economics of soil treatment are also very favorable, and once
again this type of treatment is truly a destructive technology.
  The current state-of-the-art  in groundwater  treatment technol-
ogies allows us to use these technologies now. However, the use of
these technologies as part of a remedial program must be carefully
scrutinized to ensure that they are consistent with the economics
and operating considerations of each individual program. For in-
stance, the use of a carbon adsorption  process for treatment of
organics  at  a  very high  concentration  (>100 mg/1) can create
more problems than it will solve. There are instances where highly
biodegradable compounds  present in  a  carbon system have cre-
ated high degrees of biological activity within the columns and
have subsequently generated substantial amounts of sulfides in
the carbon column effluent. These sulfides then required an addi-
tional peroxide process to eliminate the odor and corrosion prob-
lems associated with them.

ECONOMICS
  The economics of the various treatment technologies cover a
considerable range. The cost of any particular process can also vary
considerably depending on the degree of treatment required and the
time frame of the treatment program. The technologies ranked by
cost in descending order from most to least expensive are: peroxide
oxidation, activated carbon, biological and air stripping.
  The costs vary as a function of contaminant concentration; as a
rule the cost will be higher  for higher  concentrations (on a $/gal
treated basis); the efficiency of each process (i.e., an easily stripped
 compound) will cost  less to treat per gallon than a less volatile
compound. The duration of the treatment program also will have
a significant effect on the economics of treatment since a longer
term program will allow for amortization of the equipment over a
longer lifespai. Also,  some process equipment will have a higher
residual value ai the end of the program and can be sold or reused
in  other  applications.  The  following  example compares the
economics of treating a waste stream via  three different processes:
air stripping, carbon adsorption and biological/carbon adsorption.
Design Basis
    Flow = 5 gal/min
    Organic concentration =  300 mg/1 methyl ethyl ketone
    Operating period =  2  years

  The economic estimates  are based on the process achieving 1
mg/1 MEK and 100 ng/\ total other organics in the effluent. The
cost for all capital equipment, operations and maintenance are in-
cluded in this analysis.

    Carbon adsorption  =  $0.065/gal
    Biological/carbon = $0.042/gal
    Air stripping = $0.032/gal

  The economics involved  in the treatment of any  material can
vary enormously depending on the specific requirements  of the
application but,  in general, the above  figures are relatively ac-
curate. As the concentration of the contaminants to be treated in-
creases, the biological process becomes more economical. As the
volatility of the materials increases, the cost of air stripping im-
proves somewhat;  however, this cost  is difficult  to lower much
further.
  Other economic  analyses have been performed on a  $/lb re-
moved basis and have  been reported to be as low as $0.40-0.50/lb
removed  biologically to as high as $7.50-$10.00/lb removed with
carbon adsorption. A true economic evaluation of the actual costs
can be performed only on a specific case-by-case basis.
160    TREATMENT OF HAZARDOUS WASTES

-------
  The most important consideration when  considering the ap-
propriate technology is to  optimize the cost advantage of each
process. As can be seen from the economics, there are many
possible combinations of technologies which can be put together
to produce a highly efficient, cost-effective treatment process.

TECHNOLOGY TRANSFER
  The treatment requirements  for  hazardous  materials  in  a
groundwater contamination problem are  unique  for  several
reasons:

• Relatively low and/or variable flow conditions
• Varying influent concentrations
• Short time frame projects
• Ease of operation and simplicity of design
• Stringent discharge regulations
The design and operating requirements for  these treatment
systems have to reconcile all these factors and at the same time of-
fer an efficient cost-effective system. The most important con-
sideration in designing these systems is to insure the systems will
provide the necessary degree of treatment throughout the lifetime
of the project while adjusting to the changing influent conditions.
  Initially, the flow volume in these programs is generally quite
low. They rarely exceed 100-150 gal/min and quite often are in the
10-20 gal/min range. The actual flow over the course of the treat-
ment program may vary considerably due to  seasonal conditions,
the transmissivity characteristics of the aquifer and  percolation
rates of the overlying soils.
  The concentration of the contaminants will  also vary with a
general trend to decreasing concentrations. There  periodically
may be increased contaminant levels  as pockets of material are
freed, but the concentration will decrease over the life  of the treat-
ment program. It is important to  use a process design that will
cover the entire life-cycle of the project.
  The time frame  of most groundwater cleanup programs  is
generally between 2-5 years. Most  treatment systems for perma-
nent operation are  designed on a lifespan of 10-20  years. It  is
much easier to justify the costs of installing a major process on a
long-term system than on a relatively short-term system. It is most
important to minimize the amount of capital required to install a
short-term remedial project.
  The systems  being  used in these  treatment  situations are
generally unfamiliar to many of the industries and individuals in-
volved in  cleanup programs. There are many industries which,
until recently, merely discharged their wastewaters to public treat-
ment facilities and,  as a result, have limited  experience in waste-
water processing and the actual operation of wastewater treat-
ment facilities. In many instances,  there is a  minimal amount of
manpower available for operating a treatment system,  or the loca-
tions of the treatment facilities are  remote  and not frequently
attended.
  Taking all these factors into account, it becomes  apparent that
the design of a treatment facility to be used in these situations  is
indeed quite different than  a typical wastewater treatment plant.
It is important to optimize  the benefits of each technology. For
instance, it can be quite costly to provide a carbon adsorption
system to treat 50 gal/min of groundwater contaminated with 300
mg/1 of adsorbable solvent which will ultimately contain only a
few mg/1 of organics near the end  of the treatment.
  Preferable is a combined system using either air stripping or
biological  treatment before the carbon system to  reduce the
organic loading on the carbon and hence reduce the amount of
carbon which will be used. Both air stripping and biological treat-
ment are less expensive than the use of activated carbon  for the
removal of high concentrations of organics.
  In most  cases, using modular, life-cycle designs will tend to
lower the costs and increase the efficiency of the tretment process.
For instance, the use of two fixed-film bioreactors to handle a
given flow and loading is preferable over using one large unit. As
the concentrations decrease, one of the units can be taken off line
and the operating costs  and maintenance requirements will  be
decreased substantially. Similarly,  a carbon tank which is sized
for a certain flow and organic  concentration may become an
operational problem if the flow and concentrations in the influent
drop substantially. Problems with carbon treatment include: (1)
channeling  and carbon bed  may develop biological growth which
can cause anaerobic conditions and subsequent sulfide problems.
When properly designed for  life-cycle  operation, the treatment
systems can provide a highly efficient, cost-effective process.

CASE HISTORIES
Phenol Contaminated Groundwater Treatment
  An example of life-cycle design is the following treatment pro-
gram on the U.S. Gulf Coast. At the initiation of treatment, the
influent organic  concentration was approximately 1,300 mg/1 as
TOC with phenol concentration at 400 mg/1. Over the lifespan of
the treatment program, it is known that the concentration of the
influent stream will decrease substantially until  it is below 100
mg/1 TOC. It would  be  quite difficult to design and operate  a
system that functions as effectively on 1,300 mg/1 as on 100 mg/1.
Obviously,  activated  carbon  treatment  alone  would be pro-
hibitively expensive and would still require the disposal of large
quantities of spent carbon. The  use of an air stripping process
also is not possible since the organic compounds present are not
readily strippable. In order to meet all the needs of this particular
program, the  life-cycle design as  shown  in Figure  2 was im-
plemented.

                       I 300 ma/I TOC
                                                Carbon
                                                Adsorption









Aeration Aeration—
V

arifier
Fixed-Film
	 h
                         Blowers I
                       < 900 mq/1 TOC
                       < 300 mo/1 TOC
                                                 Carbon
                                                 Adsorption
                       « 100 ma/1 TQC
                                                  Carbon
                                                  Adsorption
                           Figure 2
  Life-Cycle Design for Phenol Contaminated Oroundwater Treatment
                                                                                 TREATMENT OF HAZARDOUS WASTES    161

-------
  In this installation, the aeration tanks were converted storage
tanks which, when coupled with a clarifier, were used as an ac-
tivated sludge process. As the concentrations decreased only one
tank was used for the treatment program. Eventually, that tank
was also removed from service and the fixed-film biotreater was
the remaining biological unit in operation. Throughout the dura-
tion of this program, the process design and equipment used were
changed to meet  the prevailing conditions.
  The cost of operating the system and the attention required for
its operation also diminished with time. Throughout the design of
the treatment process, those technologies used were applied to
provide the most  economical operation. The activated sludge pro-
cess (using existing equipment) provided bulk organic removal;
the fixed-film unit provided additional treatment (below about
300 mg/1 TOC the activated sludge  process would not  be  par-
ticularly effective) to reduce the organic loading to the carbon ad-
sorption system;  and, ultimately, the carbon was used to remove
the lowest level of contaminants.
  The nature of this program required that the effluent  be treated
to a TOC concentration less than 18 mg/1 so that it could be used
for reinjection to force additional contaminated water toward the
withdrawal wells. Once biological activity was established in the
system, TOC was consistently biologically reduced by 75-85%.
Effluent phenol  levels were consistently  reduced by 99%. This
biological oxidation of organics substantially reduced the  amount
of carbon required to attain the 18 mg/1 TOC  effluent  require-
ment. The influent to the system has decreased to  less than 800
mg/1 TOC, and  parts of the activated sludge system have been
taken off line. There is  now a requirement for the removal of
arsenic from this particular waste. An arsenic precipitation system
is being added so  that the arsenic can be co-precipitated with iron.
This system is being  added at a minimal cost.
  The economics involved in this particular case were based on
the use of carbon adsorption alone as compared to the cost of us-
ing biological processing. Assuming  23,000 gal/day flow at a
TOC of 1,300 mg/1,  249  pounds of organics need to be removed
from the water. Using a carbon cost of $0.75/lb and an operating
efficiency of 1 Ib organics removed/10 Ib carbon, the  operating
cost for just carbon is $l,875/day. This cost does not include the
cost of capitalizing the equipment or other costs of operation such
as electricity and  personnel.
  The anticipated cost for the operation of a biological system
alone, including capitalization of the equipment, electricity, but
not personnel, is approximately $0.46/lb  of organic material
removed. This is  $115/day plus the cost of carbon  for polishing
the effluent (assuming 50 Ib/day TOC is removed via carbon pol-
ishing) of $370/day.  The total costs of operating this system are
$485/day. This system has been in operation for over two years,
resulting in a cost savings of $834,000. The actual costs of daily
operation should  continue to diminish over the remainder of the
program.

Chemical Disposal Facility Leachale
  The optimization of several technologies in combination has
been investigated  to treat the leachate from a chemical disposal
facility.  It  has been  determined that a plume of high TOC
material was migrating off-site and required treatment. A  process
design using carbon  adsorption  alone and in conjunction with
biological treatment was evaluated in a pilot program.
  Initially, 300 gal of sample were collected from the well at the
site. Analysis revealed 2,000 mg/1 TC (to compensate  for  in-
organic  carbon, this  value was reduced by 10% to compute a
TOC value). N,N- dimethyl formamide and propionic acid were
specifically identified.
   A series of four carbon columns, each designed for a surface
loading of 2 gal/min ft2, processed the waste liquid. The utiliza-
tion efficiency of the carbon was found to be 25% (1 Ib organics
removed/4 Ib activated carbon used). The apparent attainable ef-
fluent level was 90 mg/1 as TC. The carbon was highly efficient in
its removal capabilities in this instance and was quite able to lower
the contaminant levels in the effluent. Even at this high efficien-
cy, the  cost for treatment in replacement carbon costs alone is
about $4/lb organics removed.
   A second  series of evaluation was performed to determine the
viability of using biological treatment in conjunction with the ac-
tivated  carbon in order to attain similar effluent levels at  a  re-
duced cost and, in addition, provide a destructive treatment of the
contaminant  compounds. A simple biological reactor  was de-
signed to provide degradation of the bulk of the organics present.
This reactor was followed by carbon adsorption for final polish-
ing and  removal of refractory organic compounds. The biological
reactor  was able to remove an average of 76% of the organic con-
centration of the  liquid. The carbon adsorption system removed
the remainder of the TOC down to a level of 40-70 mg/1 with the
carbon  system providing a utilization efficiency of about 16%.
The economics for the treatment of this waste stream break down
as follows:
Treatment System
Activated Carbon
   1 gal/min
  2 gal/min
Biological/Activated Carbon
   1 gal/min
  2 gal/min
Cost
(S/lb TOC removed)
8.50
6.90

4.50
3.50
These costs include both capital and operating expenses for the
treatment program.
  On the basis of this economic analysis and the needs of this par-
ticular  treatment program, a combination of  biological/carbon
treatment likely will be used at this facility.
CONCLUSIONS
  The adaption  of conventional treatment technologies for the
treatment of hazardous materials in groundwater situtations is
clearly possible. However, the use of these technologies requires a
design basis unique to each application as well as an economic
justification unique to each application. Each application must be
evaluated independently to  determine which technology or com-
bination  of technologies  is most appropriate for  producing the
desired level of treatment on a cost-effective basis.
BIBLIOGRAPHY
Envirosphcre Co.,  "Evaluation  of systems  to stabilize waste piles or
deposits," Dec. 1983.
Nyer,  E.K., Groundwater Treatment Technology', Van Nostrand Rein-
hold Company, New York, NY,  1985.
Wagner, K.  and Kosin, Z., "In Situ Treatment," Proc. of the 6th Na-
tional  Conference on Management of Controlled Hazardous Waste Sites,
Washington, DC, Nov. 1985, 221-230.
162    TREATMENT OF HAZARDOUS WASTES

-------
                                  Update on the  PACT Process

                                                Harry W.  Heath, Jr.
                                  E.I. du Pont de Nemours & Company,  Inc.
                                                  Chambers Works
                                               Deepwater, New Jersey
ABSTRACT
  This paper is an updated discussion of a 40 million gal/day
plant which uses the  du Pont developed PACT* process to pro-
vide combined secondary/tertiary treatment to industrial waste-
water.
  Reasons for selecting the PACT process are discussed. Special
features of the plant are reviewed including a 40 ton/day multiple
hearth furnace to regenerate powdered activated carbon and an
on-site, double-lined,  secure hazardous  waste landfill  where
primary sludge  from the plant is deposited. Operating data are
presented which demonstrate the effectiveness of the PACT pro-
cess including data on removal of priority pollutants.
  Because of under-utilization of the plant in recent years, a suc-
cessful outside wastewater treatment business has developed. This
aspect is discussed as an example of the financial, technical and
environmental advantages of using  a large,  advanced  treatment
facility to treat  industrial wastewaters from a variety of sources.

INTRODUCTION
  This paper is an updated case study of a 40 million gal/day
design industrial wastewater treatment plant (WWTP); several
papers discussing the initial years of operation of this plant were
presented in the early 1980s. This WWTP was built to treat the
waste from du Font's large Chambers Works site located on the
Delaware River estuary in southern New Jersey. The Chambers
Works site produces  a large variety of organic intermediates,
mainly substituted aromatic compounds; tetraethyl lead; Freon
fluorocarbons;  textile  treating  chemicals;  fuel additives;  and
miscellaneous other products. During the initial years of WWTP
operation, the site also made dyes, dye intermediates,  sulfuric
acid and isocyanates. In addition, the WWTP treated acidic waste
from  du Font's  adjacent Carney's Point nitrocellulose plant,
since shut down.
  Because the WWTP was greatly under-utilized by 1983, the site
began  to actively  seek outside  aqueous wastes for treatment to
help reduce the  fixed cost burden of the WWTP on the remaining
operations on the site. The last section of the discussion reviews
various aspects  of this very successful venture.
  The PACT process was  invented  specifically to  treat the
original Chambers Works-Carney's Point waste stream. It added
powdered  activated carbon to a conventional  activated sludge
aerator and achieved a higher degree of treatment than could be
obtained with activated sludge alone. Table 1  summarizes the
composition of the waste stream expected to be treated during the
design stages of the WWTP.

*PACT is an acronym for Powdered Activated Carbon Treatment. The
PACT technology is now owned by Zimpro, Inc., which is making the
process available under license.  "PACT System" is  a registered
trademark of Zimpro, Inc.
                          Table 1
       Design Load (Average) for Chambers Works WWTP
Flow
Soluble BOD
Color
Acidity
Dissolved Organic Carbon
Total Dissolved Solids
Total Suspended Solids
26,400 gal/min
88,700 Ib/day (280 mg/1)
1000 APHA
454,000 Ib/day* (1430 mg/1)
65,000 Ib/day (205 mg/1)
2000 to 5000 mg/1
80,000 Ib/day** (258 mg/1)
*Acidity expressed as CaCo^ equivalent throughout this paper.
**lncluded 56,000 Ib/day by product solids from lime neutralization.

  This wastewater was salty, acidic  (pH between 1  and 2) and
highly colored. In terms of organic concentration, the wastewater
was only medium strength; because of the high flow, the absolute
pounds loading was very large. Similarly, the total amount of acid
to be neutralized was high. The loading of dissolved organics was
highly variable, both in terms of composition and amount, in part
due to the batch nature of many Chambers Works processes. The
organic compounds to  be  treated  included many not normally
amenable to biological treatment.
  The first and third sections of the discussion review the factors
leading to selection of the PACT process and the subsequent per-
formance of the plant. The second section is a description of the
equipment, including the multiple hearth regeneration furnace for
activated carbon, to our knowledge the only use of this equipment
for powdered carbon regeneration in the world; the double-lined,
secure hazardous waste landfill for storage of primary sludge is
also described.
TECHNICAL AND ECONOMIC CONSIDERATIONS
IN SELECTING THE PACT PROCESS
  The PACT process developed from technical studies of treat-
ment processes for Chambers Works-Carney's Point wastewater.
Compared to a conventional activated sludge process, the PACT
process  with  a dosage  of  150 mg/1 virgin carbon at an 8-day
sludge age:
• Gave  consistent BOP removals of over 95%
• Increased DOC removal  from 62% to 85%
• Reduced color
• Minimized  foaming during aeration
• Improved sludge settling and filtration properties
• Protected the microorganisms from shock loadings of organics
  Data from typical continuous laboratory unit tests are shown in
Table 2. Table 3 shows similar data demonstrating the effect of
increasing carbon dose on system performance. The carbon used
in these tests  had an iodine number between 800 and 1,000. Sig-
                                                                              TREATMENT OF HAZARDOUS WASTES    163

-------
nificantly different performance would have been obtained with
carbon of different quality,  although  any  PACT unit generally
will outperform a parallel activated sludge  unit. Note in Table 2
that color actually was increased by activated sludge treatment.
  By  1970, when design work on  the Chambers Works WWTP
started, it had become obvious that proposed limits on DOC and
color  would require some form of tertiary  treatment. Review of
available technology showed the  most promising alternatives to
PACT was granular carbon columns located either before or after
a conventional biological reactor.  These options were studied in
detail.

                           Table 2
          Typical Laboratory PACT Unit Performance*

Feed Concentration
  BOD5                     296 mg/l
  DOC                      203 mg/l
  Color                     1960 APHA
Effluent	        PACTJJnll       Biological Unit

BOD5: Cone.
BOD5: Removal
DOC: Cone.
DOC: Removal
Color: Cone.
Color: Removal
Operating Conditions
Sludge Age, days              8.0             8.0
Carbon Dose, mg/l          148                0
Temp. Range, °C           15-24            15-24
Aerator Hydraulic
  Res. Time (hr)              7.4             7.3
* Data for time period 1/22/77  thru 5/6/77.
                           Table 3
         Laboratory Data Showing Effect of Carbon Dose

                                    Effluent
11.5 mg/l
96.1%
29.1 mg/l
85.3%
395 APHA
79.9%
26.1 mg/l
91.2%
91. 2 mg/l
62.1%
2280 APHA
-16%

Soluble BOD5, mg/l
BOD Removal
DOC, mg/l
DOC Removal
Color, APHA
Color Removal
Feed
111
86
820
Activated
Sludge
7.5
93.2%
31.9
63%
700
20%
PACT
25 mg/
Carbon
Dose
6.6
94.1%
21.7
75%
320
63%
PACT
100 mg/l
Carbon
Dose
6.1
94.5%
13.1
85%
170
80%
Data represent three months' operation of continuous, laboratory 7 5 I units. Temperature range
for all units was 18 to 25°C, and hydraulic resident lime was 8 hr
  In laboratory studies, neither process with carbon columns per-
formed as well as PACT. When carbon columns were placed after
the activated sludge unit, in the more conventional arrangement,
the biological  system constantly experienced upsets caused  by
shock loads of organics. With carbon columns before the biologi-
cal  unit, the  carbon was used inefficiently. Much of its com-
paratively expensive adsorption capacity was occupied by easily
biodegradable  compounds.  Further, as the composition of the
feed changed, it had a chromatographic effect on the carbon col-
umn. Thus, there  were  surges in the  concentration of specific
organics in the effluent from the column, and the biological unit
did not receive the expected protection from organic shocks.
  Investment costs  for the PACT system were  lower than for
either of the carbon column systems. Very generally, the invest-
ment difference was caused by the added expense for separate car-
bon columns as well  as the expense for twice the clarification
capacity in the biological portion of the carbon column systems.
Because of the superior settling properties of PACT  sludge, the
PACT system requires less  clarifier capacity than conventional ac-
tivated sludge systems.
  PACT process economics were compared with a conventional
activated sludge process for the  Chambers Works, even though
the latter was not a viable alternate in this case. Cost estimates for
PACT secondary/tertiary treatment vs. secondary treatment by a
conventional activated sludge plant  were within 10% of  each
other. The extra  PACT  investment for carbon handling and
regeneration facilities was mostly offset by the additional clarifier
capacity  and sludge disposal  facility needed  for the activated
sludge plant.
  In effect, the PACT process at Chambers Works was expected
to provide tertiary treatment quality effluent at a cost close to that
for secondary treatment alone. In  one piece of equipment, the
aerator, the PACT process destroyed most of the organic wastes
relatively cheaply by biological action, while using expensive car-
bon efficiently to remove only difficult to biodegrade or normally
non-biodegradable substances.
DESCRIPTION OF THE WWTP
  Figure 1 is a flow diagram of the WWTP. It is easiest to con-
sider the plant  in two sections:  (1) primary section, where the
acidic wastewater is neutralized, most heavy metal removal occurs
and primary solids are settled,  filtered and disposed of in the se-
cure landfill; and (2) secondary/tertiary PACT treatment section
where color  and dissolved organics are removed.


Primary Treatment
  The acidic water is  neutralized with lime in a single stage
neutralization   using   three,   stirred  200,000-gaI  reactors
operated in  parallel. The WWTP receives powdered lime by
rail car or truck. Powdered lime is stored in 4 large silos and is
slaked to 8-10% concentration as needed. Lime slurry is fed to
the neutralizes by an automatic control system that maintains
pH at any preset level. The WWTP  currently consumes ap-
proximately  60  tons/day of dry lime to treat 100 tons/day
acid. The plant has enough capacity to treat 240 tons/day acid.
The highest monthly average load treated to date is 220 tons/-
day.
  The neutralizes overflow to four primary clarifiers, each 1
million gallons  in size. The rectangular clarifiers operate in
parallel and are 230 ft  long, 55 ft wide and 12 ft deep. Solids in
the wastewater feed, as well as byproduct solids formed during
neutralization, settle and are removed as a 6-10% slurry in the
clarifier underflow. After filtration to a 45 to 50% solids cake,
the solids are hauled to the  secure landfill. Two or  three
(depending on the load) high-pressure (210 lb/in2 at end of cy-
cle) large recessed chamber filter presses make approximately 7
tons of wet press cake per 30 to 60  min cycle.
  The WWTP  currently  generates from  100,000 to 200,000
Ib/day (dry basis)  solids,  equivalent to 43,000 to 85,000
ydVyear of  landfill volume. The solids are mainly inorganic,
primarily calcium and magnesium salts, as well as silica com-
pounds  from river water  silt.  These solids also contain small
amounts of heavy metals and a variety of organic compounds.
The  solids are an inevitable byproduct of the primary treat-
ment process, and because they are primarily inorganic com-
pounds, the  only feasible disposal method is landfilling.
164    TREATMENT OF HAZARDOUS WASTES

-------
                            Figure 1
       Du Font's Chambers Works Wastewater Treatment Plant
                                                                                       KEY:
                                  RR TRACK
                                                      CARBON
                                                      UNLOADING
                             11,  iRRrrc.Ki, i.
                             111111111.1 h
                                  ©©©
                         AB =
                         AS =
                          C
                       CAW
                          F -
                         FP
                         FS
                         FT -
                         HP
                          L
                        LST =
                         NE -
                          P
                        RAS =
                       1°ST =
Afterburner
Acid Storage
Carbon Slurry Tank
Carbon Acid Wash Tank
Flocculator
Filter Press
Flow Splitter
Fuel Storage Tank
200 PSIO Filter Feed Pump
Lime Storage Silo
Lime Slurry Tank
Neutralization Tank
8,000 GPM Waste Water Feed Pump
Recycle Activated Sludge Pump
Primary Sludge Hold Tank
                                            CLARIFIER
                                        1,000,000 GAL
  A diagram of the landfill is shown in Figure 2. It is located on
the Chambers Works site, and primary sludge is hauled there by
truck.  A  double  liner  of  a  chlorosulfonated  polyethylene
material, Hypalon, covers the entire bottom and part of the sides
of the landfill. Collection pipes between the liners serve as leak
detectors and drain to sumps outside the landfill. The sumps are
sampled on a regular basis for any signs of contamination. Above
the top liner is a similar collection system for leachate which is
pumped back to the WWTP. The landfill ultimately will become a
70-ft high pyramid with a 15-acre base.
  As outer sections of the landfill are filled, the edges and top are
covered with a 2-ft layer of essentially impervious, permeability of
less than 1  X 10-7 cm/sec, clay and 12 in. of top soil. Newer sec-
tions of the landfill also may be required to have a Hypalon liner
under the clay cap.
  A controlling factor in the  operation of the primary filter
presses is that the filter cake must meet rigid soil stability criteria.
These solids have  to  support the weight of heavy earth-moving
equipment. Further, the allowable slope of the landfill sides,  and
hence the volumetric  capacity of the landfill,  is closely regulated
to assure a large safety factor against any possible slippage along
the slopes.
  As backup protection,  there are 26 monitor wells located
around the  landfill; and in the very unlikely event  of  a double
liner failure, interceptor wells would be utilized to  prevent  any
harm to the environment while the problem was corrected.
PACT Secondary/Tertiary Treatment
  PACT  treatment starts  at the  flow  splitter  which  feeds
neutralized primary effluent to the aerators. The activated carbon
is added here as a 1 Ib/gal aqueous  slurry.  The  three, 185-ft
diameter, 4,000,000 gal aerators operate in parallel and provide 5
to 9 hr hydraulic resident time.
  Air  is used  to  suspend the mixed  liquor suspended solids
(MLSS) which are an approximate 50/50 mix of biomass and ac-
tivated carbon.  Initially,  there were approximately 1150 static
mixers in each aerator to improve  oxygen transfer. These aerators
have since been removed, leaving  just air holes on the bottom of
the  injection air laterals. The BOD loading to the WWTP was
never high enough to require the  added gas transfer claimed for
the  mixers. At low BOD loadings more air was required than
needed for the bacterial metabolism alone; the extra air flow was
needed to keep the MLSS well mixed,  as the mixers appeared to
hinder gross solids circulation.
  After aeration, water  is  fed  through two  parallel, 172-ft
diameter,  2,500,000 gal  clarifiers.  Feed  enters through a
centerwell, flows through a flocculating zone and then beneath a
ring into  the annular area bounded by the clarifier  wall. The
treated wastewater overflows  into a circumferential trough and
then to the plant effluent trench.  It goes to a settling basin where
it mixes with approximately 1 Vi times  its volume of non-contact
cooling water before discharge to the Delaware River estuary.
                                                                                   TREATMENT OF HAZARDOUS WASTES    165

-------
                                                                                 V TOP SOIL

                                                                                 2-CLAY
                           TOP LINER
                                 BOTTOM LINER
1
\
• c
J<
7
LEACHATE SUMP
fs^

                                                          Figure 2
                                                       Secure Landfill
   Rim drive bridges support rakes which push settled solids to the
 center of the clarifiers; from there the solids flow via underground
 lines to  the recycle activated sludge (RAS)  pumps. Two 78-in.
 diameter screw pumps of 10,000 to 12,000 gal/min capacity each
 are provided to pump the clarifier underflow or RAS back to the
 aerators. Solids may be wasted from either the RAS stream or the
 more dilute aerator discharge.
   The WWTP can regenerate activated carbon from the wasted
 MLSS. When the plant  is  regenerating carbon, the wasted slurry
 is pumped through a thickener. Underflow from the thickener  at
 5 to  12%  solids  concentration is pumped to a 250 lb/in2 filter
 press. The 64-in.  diameter press has 112 recessed chambers 1 in.
 deep. Each press  cycle produces approximately 12,000 Ib of 38  to
 45%  solids press cake. This press is  similar  to the two other
 presses used in primary service. In fact, one  of the other two
 presses is a "swing" press which also can be  used to filter PACT
 sludge. When not regenerating carbon, all three presses may be
 used in primary service.  The filtered PACT sludge is conveyed by
 a series of drag flite, screw and belt conveyors to the top of a 26-ft
 diameter, 40-ft high multiple hearth regeneration furnace.
   There are five hearths in the furnace. Hot gases flow up the fur-
 nace  countercurrent to  the  sludge. The furnace has a rotating
 centershaft with  two  rabble arms extending over  each  hearth.
 Filter cake drops  onto the top hearth and is raked from hearth to
 hearth down  through the furnace.  From  the fifth (and last)
 hearth, the incandescent regenerated carbon drops into a water-
 filled quench tank.
  Some of the furnace's unique design features reduce upward
gas velocity and minimize  entrainment of carbon particles. The
general principle of operation is that water is evaporated on the
first and  second hearths at  gas temperatures of 480 to 700°C. On
the third hearth,  biological solids  and adsorbed organics arc
volatilized by 750 to 870 °C gas. On the bottom two hearths, at gas
temperatures of 870 to  1020°C, the powdered carbon is thermally
regenerated in the presence of water vapor.
  Off-gas from the  first hearth passes through a two-stage water
scrubbing system and then to an  oil-fired  afterburner which
destroys odors and any organics at temperatures of 650 to 760°C.
  Continuous water flow to the product quench tank is adjusted
to give a product slurry concentration of approximately 5%. The
regenerated carbon  is washed in acid to remove  inorganic  ash.
Washings from this  step are  returned to primary treatment where
removed ash precipitates and is trapped in the primary sludge. The
washed,  regenerated carbon  slurry is transferred  to a  60,000-gal
storage tank prior to recycle to the aerators. Three of these tanks
are provided to hold virgin and/or regenerative carbon.
  In recent years, organic loading to the WWTP  has decreased;
therefore, less activated carbon is needed. It has become more
economical to use only virgin carbon on a throw away basis. Cur-
rently, waste  MLSS are returned to the influent steam to primary
treatment. The carbon and  biomass are removed in underflow
from the primary clarifiers, making up about  10% of the filtered
solids deposited  in the  secure landfill.
WWTP PERFORMANCE
  During the first years of operation, the PACT process exceeded
expectations in terms of BOD and color removals and approx-
imately equalled design in DOC removal. Data for early years are
shown in Table 4; in this period, the system was  most heavily
loaded.  Performance in subsequent years has been comparable,
but inlet DOC and BOD concentration have been significantly
lower. The  flow rate also has decreased to approximately 20,000
gal/min.
166    TREATMENT OF HAZARDOUS WASTES

-------
                           Table 4
             PACT System Performance, 1978 & 1979
Flow (gal/min)
BOD
 Inlet Soluble BOD5, mg/1
 Effluent Soluble BOD5, mg/1
 Removal (%)
Color
 Inlet Color, APHA
 Effluent Color, APHA
 Removal (%)
Aerator
 Aerator MLSS (mg/1)
 Sludge Age (days)
Carbon
 Carbon Dose (mg/1)
  Virgin (°7o)
  Regenerated (%)
24-Month
Avg.

25,000

   171
    6.7
    96

 1,530
   483
    66
25,000
   41
   120
   53*
   47*
Design
Flowsheet

26,400

  280
   21
   93

 1,000
  500
   50

 8,000
   170
   30
   70
 *The low ratio of regenerated to virgin carbon reflects start up problems with the filtration and
 regeneration systems. In 1980 and 1981, the ratio was 70% regenerated/30% virgin at similar
 total carbon dose.
   The initial BOD load was less than expected because the mix of
 organics to be treated shifted and became less biodegradable than
 anticipated. The design ratio of BOD/DOC was 1.3 to 1.4, which
 was the then current ratio in Chambers Works waste. However,
 extensive waste reduction  efforts prior to WWTP  startup re-
 moved a disproportionate amount of biodegradables. As can be
 seen in Table 4, the BOD/DOC ratio in 1978 and 1979 was ap-
 proximately 1.0.
   The activated carbon in the aerator allowed the WWTP to treat
 this difficult waste. One result of the lower BOD load was that the
 minimum air flow to the aerators was set by  the amount needed to
 keep the biomass and carbon well suspended. This was about 13
 ftVmin per static mixer.  It had been expected that minimum air
 flow would be set by the need to maintain  a minimum  dissolved
 oxygen concentration  of 0.5 mg/1 in the aerator.  As was men-
 tioned earlier, this was the reason the static  mixers were removed
 from the aerators in 1981. This allowed the air flow to be reduced
 from 15,000 ftVmin per aerator to 8,000 ftVmin or lower, while
 still maintaining 0.5 mg/1 oxygen levels.
   The influent was more highly colored than was assumed during
 design. However, color removal efficiency was better than antici-
 pated, and the site was able to meet permit standards for color.
   Average DOC removal was equal to design, so effluent quality
 was slighly better than predicted, reflecting  the lower  influent
 concentration.
   In terms of meeting NPDES permit limits, variation in perfor-
 mance is often more critical than long term average. Most permits
 specify a daily maximum limit as well as a monthly average, and
 consistently meeting the  daily maximum is often the more de-
 manding task for the process. The consistency of the PACT pro-
 cess at Chambers Works is  outstanding.  Figures 3 and 4 are
 histograms showing the distribution of feed and effluent DOC
 and soluble BOD respectively in 1978 and 1979. Note the narrow
 distribution of effluent concentration, particularly of BOD.
  Analyses for the removal of U.S.  EPA-designated  priority
pollutants are in Table 5. Volatile organics  and acid  extractable
compounds generally  are removed very well; base  neutral com-
pounds are removed with a lesser amount of  success. Some metals
removal occurs across PACT treatment, although the process is
not designed for this purpose.
  N        = 721
  Avg      = 170
  Std D6V  =  506
  Max      = 278
  Mm      =  80
                                                                                       90% CL
                                                                                     Log Normal
10
110      160
  mg/llter DOC


4->
Q)
3
£
111





/
/
/.
/I
ill
, 90% CL
/Log Normal
y /
k /
V
i 	 1 	 1 	
10 60 110 160
mg/liter DOC
N
Avg
stci Dev
Max
Mm



— i 	
210

= 721
= 312
= 99
= 106
= 108



	 1 	 1
260

                                     52%
                                 60%
                 68%      76%      84%      92%     100%
                  % DOC Removal

                         Figure 3
            DOC Histogram (1978 and 1979 Data)
                                      A major reason for selection of PACT was the protection af-
                                    forded the biosystem from upsets caused by shock loadings  of
                                    organic wastes. Because of the large number of batch processes at
                                    Chambers Works, it was not impossible for the WWTP to see
                                    20,000 to 50,000-lb slugs of various organics in a 1- to 4-hr period.
                                    The biosystem handled documented spills  to  the WWTP  of
                                    50,000 Ibs of an aromatic diamine and  30,000 Ib of o-toluidine
                                    with no ill effects. No equalization is provided before the WWTP
                                    because pilot plant studies showed no significant improvement in
                                    performance with equalization.
                                      The demise of dyes manufacture on the  site  ended a major
                                    source of organic shock loads. However, the increasing amounts
                                    of trucked in outside wastes actually have made the overall prob-
                                    lem more severe. A 5,000 gal truck is discharged into the 20,000
                                    gal/min influent stream within 30 min. The PACT process has
                                    tolerated the resultant shock loads when wastes with relative high
                                    DOC content are received.
                                      Initially, the WWTP  operated  at higher sludge ages, 20 to  60
                                    days, than the 8 to 15 days specified in design. Organic removals
                                    were  improved at higher sludge  age,  with the  result that the
                                    WWTP operated at an average carbon dose of 120 mg/1.  This
                                    compares with a design  dose  of 170 mg/1. Because  of lower
                                    organic loadings, since 1983 the plant has operated at sludge ages
                                    of 8 to 30 days with virgin carbon doses of 10 to 50 mg/1.
                                      Because of the initial high solids concentrations in the aerators,
                                    10,000 to 30,000  mg/1,  the secondary clarifiers were  heavily
                                    overloaded. They were designed for a solids flux rate of 63 Ib/day
                                                                                 TREATMENT OF HAZARDOUS WASTES     167

-------
                   200      300      400
                      mg/llter BOD
                          500
             90% CL
           Log Normal
                      N
                      Avg
                      Std  Dev
                      Max
                      Min
                                                      600
543
  67
  75
 50
  01
200      300      400
  mg/llter BOD
                                              500
                                   500



n
0
Q>
5
N
Avg
Std Dev
Max
Min

2 60
= 543
= 961
= 36
= 999
= 66

68





1
76
                                90% CL
                                Gamma
                                      84
                   % BOD Removal
f
100
£. , ** ~ LJ HI JL U
4-Nitroph
Phenol
METALS
Antimony
Arsenic
Beryllium
Cadmium
Chromium
Copper
Lead
Mercury
Nickel
Selenium
Zinc
                           Figure 4
              BOD Histogram (1978 and 1979 Data)
                                                                                           Table 5
                                                                          Priority Pollutant Removals Across PACT System
VOLATILE ORGANICS*
Benzene
Carbon Tetrachlorlde
Chlorobenzene
Chloroechane
Chloroform
Ethylbenzene
Methyl Chloride
Tetrachloroethylene
Toluene
1,1,1-Trlchloroe thane
Trichloroethylene
Trichlorofluoromethane

BASE-NEUTRAL EXTRACTABLES*
1,2-Dlchlorobenzene
2,4-Dinttrotoluene
2,6-Dinitrotoluene
Nitrobenzene
1,2,4-Trichlorobenzene

ACID  EXTRACTABLES*
2-Chlorophenol
2,4-Dinitrophenol

Fetd
0>g/D
105
94
1720
280
201
41
1770
24
519
13
41
155
259
1900
1640
454
523
11
161
1020
489
24
19
1.4
5.6
26
72
112
.02
77
29
602
Concentration
Effluent
0-g/D
0.9
1.4
30
12
21
1.7
Nil
1.7
1.7
0.6
1.9
3.0
120
243
575
2
169
1.6
5.0
10
38
24
21
1.0
5.4
15
75
54
.01
59
23
360

Removal
(*)
99
99
98
96
90
96
99
93
99
95
95
98
54
87
65
99
68
85
97
99
92
Nil
Nil
29
4
42
Nil
52
50
23
21
40
                                              The values in Table 5 arc averages of data whtch sho» significant day-to-day variations in
                                             removals. Results were obtained by GC MS analyses of plant sample* and are subject to analytical
                                             error. The figures represent only estimates of average percent remcnab for these materials.
ft2  at an  underflow  concentration  of 3.6%.  They  actually
operated at solids  flux  rates from  200 to 350  Ib/day  ft2 at
underflow concentrations of 4 to 8%. The carbon in the sludge
greatly enhanced the thickening abilities  of the clarifiers.
  Part of the reason for such heavy solic4, loading were initial
problems with the PACT sludge filter prf ^s and, surprisingly, the
conveying system to carry  filtered slurge into the regeneration
furnace.  During 1978  and  1979,  the inability  to  remove solids
from the liquid train by normal wastage methods forced solids to
exit the system as total suspended solids (TSS) in the treated ef-
fluent overflowing the secondary clarifiers. Effluent TSS  aver-
aged about 100 mg/1 vs.  a  design level of 30 mg/1. These solids
settled in the polishing basin and ultimately were dredged to the
primary treatment section of the WWTP.
  From 1978 through March 1982 when  the regeneration furnace
was  shut down, there was a steady series  of mechanical im-
provements in the PACT  sludge filtration and conveying systems.
Not all portions of the system reached design rates, and some added
and/or redesigned equipment had to be installed.
  The expected improvement in sludge filterability with the addi-
tion  of carbon to the  biomass was achieved. Filter cake solids
averaged 45% vs. a predicted 35% to 38% level.  Press capacity
was below forecast because  of: (1) lower dosage of carbon needed
with longer sludge age operation, and hence less  carbon in the
sludge to aid filtration, and (2) operating problems with the filter
press, particularly with cloth life, cloth  changing procedure and
high pressure feed pumps. The original steel press plates are being
                                              replaced with lighter weight fiberglas or  plastic plates that are
                                              easier to dress and are expected to dramatically improve PACT
                                              sludge filtration if the furnace resumes operation.
                                                As anticipated, the regeneration furnace incurred  relatively
                                              high  maintenance costs.  There have been several failures of  he
                                              brick hearths.  The combination of the above, plus numerous
                                              design problems  of the type often associated with  a pioneering
                                              chemical processing venture, limited furnace in time to just over
                                              50%  in the first two years of operation. In time of furnace opera-
                                              tion improved from 55% in 1978-79 to 66% in 1980-81. Much of
                                              the furnace downtime was caused by problems with the feed con-
                                              veyors and not with the multiple hearth furnace itself.
                                                When operating, the furnace produced  an  average of 24,000
                                              Ib/day regenerated carbon in 1978-79. In 1980-81, production was
                                              raised to 38,000 Ib/day.  Recovery of virgin  carbon properties
                                              averaged just over 60% for the high activity carbons needed for
                                              Chambers Works waste.  Weight  yield recovery was over 80%
                                              after acid washing. As expected, the acid washing step was needed
                                              to prevent buildup of inorganic ash on the carbon with successive
                                              regenerations. However, during the last months of furnace opera-
                                              tion, ash levels increased in regenerated carbon. Revisions to the
                                              acid  washing system  to better remove iron and silica were sche-
                                              duled when the decision  was made to shut down the furnace.
                                                Despite  significant equipment improvements,  it  appears the
                                              solids train will  continue  to incur relatively  high  maintenance
                                              costs. As a result, alternate technologies  have been considered.
168    TREATMENT OF HAZARDOUS WASTES

-------
One which shows great theoretical promise is wet air oxidation. It
would eliminate both filtration and sludge handling and replace
the furnace with a high pressure reactor. Whether this high pres-
sure/lower temperature  liquid phase treatment can  regenerate
Chambers Works spent carbon to the quality necessary to treat
this type of waste is not certain. Experience at other locations sug-
gests it may be able to do so, however there has not been suffi-
cient justification yet to do the large scale pilot studies necessary
to test the process for Chambers Works.

OUTSIDE WASTE BUSINESS
  Operating experience at the Chambers Works WWTP over the
past 4 years has demonstrated  the  advantages of transporting
aqueous wastes to a large, central, advanced wastewater facility
for treatment.  In the early 1980s,  du  Pont began aggressively
seeking outside wastes for treatment as it became apparent that
the WWTP was going to be under-utilized because of declining
production  at  the site.  A  marketing study showed this  large
PACT plant had unique advantages  in treating outside wastes.
  Because of the great volume of wastewater being treated, the
WWTP had essentially  an  unlimited  volumetric  capacity for
trucked waste. Four tankers simultaneously discharging 5,000 gal
each in only 20 min increase influent flow less than 5%. Highly
acidic wastes are no problem. The WWTP has treated 2,500 gal of
almost pure chlorosulfonic acid in only 4 hr by direct discharge in-
to the influent trench with no release of fumes.
  The tremendous dilution provided by over 20,000 gal/min flow
is a major protection against organic shock loads from a tank
truck  discharge.  The  PACT system was  selected specifically
because of its ability to handle varying loads of industrial waste.
Therefore, this plant  is almost  ideally  suited to treat batch
discharges of aqueous wastes containing organics, especially those
which would pass through or be inhibitory to a conventional
biological treatment system.
  The Chambers Works  already had logistics in place to handle a
large  number of truck deliveries  each day and excellent interstate
highways leading to the site. This was vital because of the volume
of incoming tank trucks. Deliveries of over 50 trucks a day are
routine.
   Because of other activities on the  site, there were "in-house"
environmental  and technical groups to support this activity.  A
major technical effort is needed to evaluate wastes proposed for
treatment both in terms  of their suitability to the PACT process
and the estimated costs  of treatment. A strict protocol must be
established to approve a waste for  treatment,  and subsequent
deliveries  must be monitored closely to  assure that the waste ac-
tually received is what was contracted for. Because of the amount
of hazardous waste treated, there are frequent contacts with state
and federal regulatory  authorities regarding operation of the
WWTP. A state inspector is at the site every 1 to 2 weeks, and it is
important that a competent environmental staff is available  to
represent the plant.
  From a regulator's point of view, it  is advantageous to treat
wastes at a central site. The disposition  of a large number of in-
dividual wastes can be monitored conveniently by a single inspec-
tor. The single, large plant usually will be better controlled than a
variety of small separate facilities, often operated by those less
than expert in waste treatment. The quality of the treated effluent
can be assured  by a detailed analytical evaluation of a single ef-
fluent stream; such analyses might be impractical and prohibitive-
ly costly for a large number of individual sites.
  Another advantage of  this site is  the receiving water. Chambers
Works main outfall discharges into an estuary, so salt content of
the stream is not critical. There  are no downstream withdrawals
for drinking water purposes. The volume of flow at the Delaware
River estuary is immense, and so for a given absolute quantity of
residual waste the effluent concentration will be very low. Any
aqueous  waste treatment facility  will discharge  some  residual
material. It is an obvious advantage if the receiving water flow is
high; then the resultant concentrations of any residual pollutants
are low, and there is no adverse impact on downstream water use.
  Since 1980, the outside waste business has grown rapidly. Cur-
rently, over 15% of the organic load to the WWTP is from non-
Chambers  Works  wastes.  Over  50% of the outside  waste
originates in  New  Jersey; 35%  comes  from  Pennsylvania,
Delaware and Maryland. However, it has been cost-effective to
ship  aqueous wastes surprisingly long distances. There are routine
contracts originating  as  far away as  Maine, Florida and Col-
orado. The freight-logical area has been extended by offering rail
car service to transport wastes, and the potential exists for barge
deliveries.
  Wastes from a variety of sources have been treated. Some are
listed in Table 6. No one type of waste makes up more than 20%
of the total.

                           Table 6
     Typical Industrial  Wastewaters Treated at Chambers Works

• Tank Truck & Tank Car Washings
• Pharmaceutical Wastes
• Water from Oil-Water Separation Processes
• Textile Treating Wastes
• Metals Treating Wastes
• Bio-sludges from Industrial Wastewater Treatment Plants
• Electronics Industry Waste
• Landfill Leachates
• Lagoon Cleanups
• Miscellaneous Chemical Process Wastes
• Latex Wastes
• Chemical Cleaning Rinse Waters
• Waste Acids or Bases
• Food Processing Wastes
• Paint & Dye Wastes
  Initially, treatment of outside waste was limited to streams
which could be discharged directly into the WWTP influent as
rapidly as they flowed from the truck. This limitation precluded
certain aqueous wastes. For instance, wastes containing sulfide or
cyanide could not be accepted because of the possibility of releas-
ing toxic H2S or HCN gas when the waste mixed into the acid in-
fluent.
  The WWTP has  limits on most of the  heavy metals. It  is
capable of removing those present in the original wastewater feed
to meet  NPDES permit limits. Initially, some outside wastes
which were evaluated had to be  rejected because they contained
metals in amounts such that permit limits might have been ex-
ceeded had the waste been added directly to the WWTP influent.
  Recently,  heavy metals pretreatment facilities  have been in-
stalled and are now being expanded to allow the WWTP to accept
aqueous  waste streams containing higher  amounts of  heavy
metals. Similarly, the plant is developing treating techniques for
waste containing higher concentrations of sulfide  and cyanide.
  Facilities also are being installed for oil-water separation, with
the aqueous phase to be discharged to the WWTP and the oil to
be handled separately. Other expansions are being studied which
involve   separate  pretreatments  which  themselves  generate
aqueous  waste that could be handled by the WWTP. This might
give du Pont  a competitive advantage over other stand alone
facilities.
                                                                                 TREATMENT OF HAZARDOUS WASTES    169

-------
  This outside waste treatment has been successful enough that
du Pont is expanding the business beyond the Chambers Works
WWTP.  In  recognition  of their  expanded charter,  the outside
waste business group was recently  renamed du Pont Environmen-
tal Services and is actively seeking other waste treatment oppor-
tunities.

CONCLUSIONS

• PACT is an excellent, advanced treatment process for many
  wastewaters containing normally non-biodegradable or slowly
  biodegradable materials. The process can treat highly colored
  wastes. At du Font's  Chambers Works chemical  complex, a
  PACT plant removes over 96% of the BOD and over 80% of
  the DOC from 30 to 40 million gal/day of wastewater.
 • A PACT process is resistant to toxic upset, can tolerate shock
  loads of organic compounds and is tolerant of rapid changes in
  waste composition.
 • The presence of carbon in the mixed  liquor suspended solids in
  a  PACT system dramatically increases settling in  the clarifier
  and improves filtration rate. Solids concentrations of approxi-
  mately 40% have been achieved with PACT sludge in a conven-
  tional pressure  filter.
• Conventional wastewater treatment equipment can be used for
  a PACT system with a minimum amount of design changes.
• Powdered activated carbon has  been successfully regenerated
  from PACT sludges in a multiple hearth regeneration furnace
  at a rate of 20 tons/day. Maintenance costs for the entire fil-
  tration, conveying and regeneration unit are high, and to date
  the best system in time has been about 65%.
• The Chambers Works PACT plant is successfully treating a
  large variety of off plant  wastes delivered by tank  truck or
  railroad car. Over 30% of the current organic load to this plant
  is from outside wastes. Particular, demonstrated advantages of
  a large central wastewater treatment are:
  (1) An advanced treatment process is used.
  (2) High  quality "in house"  high quality technical and en-
     vironmental staffs can be afforded.
  (3) It is sited on optimum receiving waters.
  (4) It is  easier for  regulatory  authorities to monitor the  dis-
     posal of the wastes.
  (5) It can  treat  aqueous, organic waste  streams  which  might
     upset conventional municipal plants.
170    TREATMENT OF HAZARDOUS WASTES

-------
                         Treatment  of  PCB-Contaminated  Soil
                               in a  Circulating Bed Combustor

                                               D.D. Jensen, Ph.D.
                                                    D.T. Young
                                              GA Technologies Inc.
                                      Process & Energy Systems Division
                                              San Diego, California
ABSTRACT
  A trial burn of PCB-contaminated soils was completed in GA
Technologies' 16-in. Pilot Plant  Circulating  Bed Combustor
(CBC).  More than 4000 Ib  of soil  containing 1% PCB were
treated in three identical 4-hr runs at  1800°F. The results showed
excellent compliance with TSCA requirements. Destruction and
removal efficiencies  (DREs) were  greater  than 99.9999%, and
PCB in combustor ash  was  less than 200 ppb. No chlorinated
dioxins or furans were detected in the stack gas, bed ash or fly
ash. In addition, no significant concentrations of other Products
of Incomplete Combustion (PICs) were detected.  Combustion
efficiencies were  greater than  99.9%,  with CO concentrations
less than 50 ppm and NOX concentrations less than 75 ppm. Par-
ticulate emissions were generally below 0.08 grain/dscf, and HC1
emissions were maintained below 4.0 Ib/hr  by introducing lime-
stone directly into the combustor.
  These results demonstrate that the CBC is an environmentally
acceptable means of treating contaminated  soil containing PCB
and other organic wastes. In addition, the high thermal efficiency
of the  CBC, the absence of afterburners or scrubbers  and  the
use of simple feed systems make CBC treatment competitive with
soil removal and transport to landfills and other potential treat-
ment/disposal options.

INTRODUCTION
  Of the many hazardous chemicals found in waste sites around
the country, perhaps none has received the scrutiny conferred on
polychlorinated biphenyls (PCBs). This group of  209 synthetic
chlorinated organic compounds found wide use as a dielectric
fluid in utility transformers and capacitors and as a high-tempera-
ture heat transfer medium.' However, because of their exception-
al resistance to degradation in the biosphere and apparent toxic-
ity, the manufacture and sale of PCBs  were banned in 1976  for
virtually all purposes. The control,  treatment and disposal of
PCBs  were  mandated   by  TSCA and currently  are  handled
through the U.S. EPA's Office of Toxic Substances.
  Until recently, it has been common practice to remove PCB-
contaminated soils for burial in a secured landfill. However, this
option  is becoming less desirable since landfill  costs are escalat-
ing, the number of available landfill sites has decreased and gen-
erators or potential responsible parties (PRPs) retain the liability
associated with the contaminated soil, even in a secured landfill.
Treatment of PCB-contaminated soil by incineration in the CBC
can eliminate or significantly reduce the potential liability of gen-
erators or PRPs at a cost competitive with current landfill prices.
  CBC units are designed to burn a wide variety of fuels such as
coal, peat, wood, municipal wastes  and oil while treating haz-
ardous  wastes. Over 25 units are operating or under construction
worldwide. Three units are currently in operation in the United
States. In 1983 GA began concentrating its efforts on the applica-
tion of GBC technology to  incineration of hazardous wastes.
Table 1 presents examples of wastes that have been burned in the
CBC.  It was the successful treatment of this diversity of wastes
that provided assurance that PCBs could be destroyed in a CBC
at a lower temperature than used in conventional incinerators.


                         Table 1
 Circulating Bed Test Results1 for the Destruction of Hazardous Wastes
Waste
Carbon Tetrachlorlde
Freon
Malathlon
Dichlorobenzene
Aromatic Nltrlle
Trichloroethane
Form
Liquid
Liquid
Liquid
Sludge
Tacky
Liquid
Destruction
Efficiency, J
99.9992
99.9995
>99.9999
99.999
solid >99.9999
99.9999
HC1
Capture, %
99.3
99.7

99

99
Ca/Cl
Ratio
2.2
2.14

1 .7

1 .7
1 Results obtained in GA pilot plant CBC.

CBC DESCRIPTION
  The CBC is a new generation of incinerator that uses high
velocity air to entrain circulating solids in a highly turbulent com-
bustion  zone. This design allows combustion along the entire
length of the reaction zone. Because of its high thermal efficien-
cy, the CBC is ideally suited to treat low heat content feed, in-
cluding contaminated soil. Figure 1 shows the major components
of a CBC for soil treatment.
  Soil is introduced into the  combustor loop at the loop seal
where it immediately contacts  hot recirculating soil from the hot
cyclone. Hazardous materials  adhering to soil are rapidly heated
when introduced into the loop  and continue to be exposed to high
temperatures throughout their residence time in the CBC. Upon
entering the combustor, high velocity air (14 to 20 ft/sec) en-
trains the circulating soil which travels upward through the com-
bustor into the hot cyclone. Retention  times in the combustor
range from 2 sec for gases to approximately 30 min  for larger
feed materials (^1.0 in.).
  The cyclone separates the combustion gases from the hot solids
which are returned to the  combustion chamber via  a  proprie-
tary non-mechanical seal. Hot flue gases and fly ash pass through
a convective gas cooler  and on to a baghouse filter where fly
ash is removed. Filtered flue gas then exhausts to the atmosphere.
Heavier  particles of purified soil remaining in the combustor
lower bed are removed slowly by a water-cooled ash conveyor
system.
                                                                            TREATMENT OF HAZARDOUS WASTES    171

-------
                                    COMIUSTOR
                          LIMESTONE
                          FEED
                     SOIL
                     FEED
                                                          COOLING
                                                          WATER
                                                          Figure I
                                     Schematic Flow Diagram of Circulating Bed Combustor for
                                                       Soil Treatment
                     ASH
                     CONVEtO*
                     SYSTEM
  As a consequence of the highly turbulent combustion zone,
temperatures around the entire combustion loop (combustion
chamber, hot cyclone, return leg) are uniform to within  ±SO°F.
The uniform low temperature and high solids turbulence in the
CBC also help avoid ash slagging encountered in other types of
incinerators.
  Acid gases formed during destruction reactions are rapidly cap-
tured by limestone added directly into the combustor. The reac-
tion of limestone and HC1, released during PCB incineration,
forms dry, benign calcium chloride. The rapid combustion and
quick neutralization of the  acid gases within the combustion
chamber eliminate the need for afterburners and add-on scrub-
bers to complete destruction and acid gas capture. Emissions of
CO and NO, are controlled to low levels by excellent mixing, rela-
tively low temperatures (14SO to 1800°F) and staged combustion
achieved by injecting secondary air at higher locations in the com-
bustor.  Because of its efficient combustion and highly turbulent
mixing, the CBC is capable of attaining required DREs for  both
hazardous  wastes (99.99%) and  toxic  wastes (99.9999%) at
temperatures below those used  in  conventional  incinerators
(typically > 2000 °F).

TEST DESCRIPTION
  A variety of requirements are imposed  prior to and during a
PCB trial burn.' The key target of a trial burn is to ensure that
PCB DREs are ^99.9999%  at the operating conditions chosen
for the incinerator. In addition, the concentration of PCB in ash
from the unit must not exceed 2 ppm. The potential formation of
PICs also is carefully evaluated, with particular attention given
to polychlorinated dibenzo-p-dioxins (PCDDs)  and polychlori-
nated dibenzo-p-furans (PCDFs). The combustion efficiency of
the unit must be » 99.9fo,  and particulate emission must not ex-
ceed 0.08 grain/dscf.
                     Figure 2
Pilot Plant CBC and Feed Preparation/Handling Equipment
172    TREATMENT OF HAZARDOUS WASTES

-------
   The CBC trial burn was carried out in GA's 16-in. pilot plant
 unit shown in Figure 2. This is the smallest CBC offered by GA
 for commercial application. The glovebox in the foreground was
 used to prepare contaminated soil prior to transport to the feed
 system and CBC shown behind the glovebox.  Soil treated in the
 test was obtained from a former chemical processing site known
 to contain pockets  of PCB up to 6000 ppm as well as other
 organic and inorganic wastes. To ensure that the CBC would be
 permitted to treat all likely site concentrations of PCB,  uncon-
 taminated soil from the site was "spiked" with liquid  PCB to
 10,000 ppm. Spiking was carried out by blending a 50:50 com-
 mercial mixture of PCB "1248" and trichlorobenzene with a rib-
 bon blender in 1000 Ib lots. Approximately 4000 Ib of soil  were
 spiked for the three  burns  required by  the  TSCA trial burn
 permit.
   While the CBC was maintained at 1800T using natural gas
 as the auxiliary fuel, several barrels of clean  soil from  the site
 were treated in the combustion system prior to the addition of
 spiked soil. During this time, all  operating parameters and sys-
 tem components  were confirmed to be in the required operating
 ranges. Process parameters monitored included:
 • Temperature around the loop
 • Pressure drop across the loop
 • Soil feed rate
 • Primary air flow
 • Secondary air flow
 • Loop seal air flow
 • Total air flow
 • Methane flow
 • CO concentration
 • CO2 concentration
 • Excess oxygen level
 • NOX concentration

   Spiked soil was pneumatically  transported  to a bunker  and
 screw  feeder. Soil  feeding,  limestone addition and stack gas
 monitoring were  started simultaneously. A U.S. EPA Modified
 Method 5 sample train3 was used to sample stack gas emissions.

                           Table 2
          PCB Trial Burn Operational Data and Test Results
Parameter
Teat Duration, hr
Operating Temperature, °F
Soil Feed Rate, Ib/hr
Total Soil Feed, Ib
PCB Concentration in Feed, ppm
DRE, J
PCB Concentration
' Bed Ash, ppm
Fly Ash, ppm
Dioxin/Furan Concentration
Stack Gas, pp
Bed Ash, ppm
Fly Ash, ppm
Combustion Efficiency, J
Acid Gas Release, Ib/hr
Particulate Emissions,
grain/dscf
Excess Oxygen, J
CO , ppm
C02. >
NOX, PPB
TSCA
Requirement 1
-1 1
1800
328
1592
11 ,000
>99.9999 99.999995

<2 0.0035
<2 0.066

ND
ND
ND
>99.9 99.91
<1.0 0.16
<0.08 0.095(b)
'
>3.0 7.9
35
6.2
26
Test Number
2
1
1800
112
1321
12,000
99.999981 99

0.033
0.0099

ND
ND
ND
99.95
0.58
0.013

6.8
28
6.0
25
3
1
1800
321
1711
9,800
.999977

0.186
0.0032

ND
ND
ND
99.97
0.70
0.0021

6.8
22
7.5
76
 In addition, a separate Volatile Organic Sampling Train (VOST)4
 was used to sample for volatile organic PICs. Feed, bed ash and
 fly ash samples also were gathered throughout the test (see Fig. 1
 for sample port locations). Three identical tests of spiked soil (4
 hrs each) were carried out over two days in late May 1985. Each
 test was observed and/or audited by U.S. EPA personnel or rep-
 resentatives. All feed, ash and stack gas samples were subsequent-
 ly analyzed for PCBs, PCDDs and  PCDFs. Fly ash, bed ash and
 stack gas samples also were analyzed for other PICs (both vola-
 tile and semivolatile). Stack gases were analyzed for fly ash and
 chloride release as well.

 RESULTS
  Table 2 presents a summary of the trial burn operational data
 and test  results gathered  during the  tests. Near-identical con-
 ditions were maintained for each test. In each case, PCB DREs
 were well in excess of the  U.S. EPA-required 99.9999%.  PCB
 concentration  in the bed ash and the fly ash  was well below the
 required 2 ppm and, in fact, did not exceed 200 ppb. No PCDDs
 or PCDFs were detected in the stack gas, bed ash or fly ash. Com-
 bustion efficiencies were greater than 99.9%, and acid gas release
 was well below the required 4 Ib/hr.  Paniculate emissions  were
 generally less than the required 0.08 grain/dscf. Only the grain
loading from the  first test, obtained from a 2-hr makeup test
 after the completion of Tests 1 through 3, showed a value slight-
 ly higher  than  the limit. This higher than normal value is attrib-
uted to off-normal process conditions for the baghouse (i.e.,
excessive blowback air pressures along with  a higher-than-normal
number of blowback cycles). Nitrogen oxides and CO levels re-
mained low as a result of the staged combustion utilized in the
CBC and  the  relatively low combustion temperature (1800T).
               STACK
                                            FLUE GAS
                                            COOLER
                                                                                                                    8AGHOUSE
                                                                 COMBUSTOR
                                                                 ASH REMOVAL
                                                   FORCED
                                                   DRAFT
                                                   FAN
ND  Not detected.
(^Derived from 2-hr makeup test.
                         Figure 3
      Isometric of Site-Assembled Circulating Bed Combustor
                                                                               TREATMENT OF HAZARDOUS WASTES    173

-------
These results demonstrate that the CBC is an effective means of
destroying PCBs contained in a soil matrix without the need for
high temperatures, afterburners or wet scrubbers. In particular,
the absence of undesirable combustion byproducts (i.e., dioxins
and furans) is important in ensuring that  effective treatment of
soil can be accomplished in an environmentally acceptable man-
ner.
   These results  confirm the design of GA's transportable CBC
shown in Figure 3.  The combustion and all other plant  com-
ponents are designed as modular units which can be transported
by truck  or rail. These units are assembled at the site into an
operating unit  in 4 to 6 weeks. The major components of this
CBC plant include the combustor loop, feed system, pollution
control and air  induction equipment. GA's 36-in. transportable
CBC is capable  of processing up to 18,000 Ib/hr of dry soil on a
24-hr basis requiring an operating crew of only two persons per
shift. Soil treatment costs may be as low as  $100/ton at a large
site. For  smaller sites or sites having unique treatment require-
ments, costs may approach $400/ton.
 CONCLUSIONS
   The results of the PCB soil trial burn in GA's CBC demonstrate
 compliance with  TSCA requirements: stack  emissions are well
 within regulatory requirements and bed ash and fly ash contain
 PCBs  well below  the  regulatory  requirements. The superior
 thermal efficiency,  high throughput and small  staffing require-
 ments  of the CBC  provide a soil treatment option that is cost-
 competitive with landfill disposal while at the same time reducing
 overall liability of the generator or PRP.

REFERENCES
I.  SCS Engineers, Inc.,  "PCB Disposal Manual," Electric  Power Re-
   search Institute, Palo Alto, CA, Report No. CS-4098, June 1985.
2.  "Polychlorinated Biphenyls (PCBs) Manufacture, Processing, Distri-
   bution in  Commerce and Use Prohibition," 40 CFR 761.70.
3.  "Test Methods  for Evaluating Solid  Waste," U.S. EPA Report
   SW-846, 2nd Edition,  1984.
4.  "Proposed Sampling  and Analytical  Methodologies for Addition to
   Test  Methods for Evaluating Solid Waste: Physical/Chemical
   Methods  (SW-846,  2nd  Edition)." U.S. EPA Report PB85-103026,
   1984.
174    TREATMENT OF HAZARDOUS WASTES

-------
                        Treatment of Hazardous  Waste  Leachate

                                                  Judy L. McArdle
                                               Michael M. Arozarena
                                            William E.  Gallagher, P.E.
                                                PEI Associates, Inc.
                                                  Cincinnati, Ohio
                                                Edward J.  Opatken
                                     U.S. Environmental Protection Agency
                              Hazardous Waste Engineering Research Laboratory
                                                  Cincinnati, Ohio
ABSTRACT
  The nature of hazardous waste leachate varies greatly from site
to site and over time. Thus, leachate treatment systems must be
designed on a case-by-case basis. This paper presents profiles of
seven unit treatment operations which can be applied in varying
combinations  to effect the  desired degree of contaminant re-
moval: (1) equalization (to dampen influent flow and concentra-
tion fluctuations); (2) air stripping (to remove ammonia and vola-
tile organic compounds); (3) precipitation/flocculation/sedimen-
tation (to remove soluble heavy metals and  suspended solids);
(4)  neutralization (to adjust pH); (5) activated sludge treatment
(to remove aerobically biodegradable organic matter); (6) carbon
adsorption (to remove dissolved organics and certain inorgan-
ics); and (7) reverse osmosis  (to remove dissolved inorganics and
high molecular weight organics). These technology profiles em-
phasize process  applicability and limitations, treatment effective-
ness and capital and operating costs.

INTRODUCTION
  Leachate is generated by the percolation of water (from precip-
itation, groundwater flow or liquid wastes) through a waste dis-
posal site. The leachate can be expected to contain soluble toxic
components of the waste as well as soluble chemical and biochem-
ical reaction products. Leachate collection systems are designed
to collect this contaminated  liquid and to channel it away from
the disposal site before it can contaminate the surrounding soil,
groundwater or surface water. The collected leachate then can be
treated to reduce the level of contaminants prior to its discharge
into a stream, to groundwater  recharge or to a municipal  or in-
dustrial wastewater treatment system.
  The composition and volume of leachate vary highly from site
to site and over time. Factors contributing to this variability in-
clude: the nature of the waste;  the age of the land disposal unit;
the amount of precipitation; and the porosity, permeability and
adsorption characteristics of the soil. Table 1 presents leachate
characterization data from three land disposal sites.
  The high strength and widely varying nature of hazardous
waste leachate complicate its treatment.  Under contract to  the
U.S. EPA's Hazardous Waste Engineering Research Laboratory,
PEI Associates, Inc., Cincinnati, Ohio, is preparing a compre-
hensive handbook to guide in the planning and design of  leach-
ate treatment systems. Technology profiles emphasizing process
applicability and limitations, treatment effectiveness  and capital
and operating costs are being developed for both demonstrated
and potential technologies. Six highly developed unit treatment
operations (equalization, air stripping, precipitation/flocculation/
sedimentation, neutralization, activated sludge tretment and car-
bon adsorption) and one potentially applicable technology (re-
verse osmosis) were selected for presentation here from the more
than 20 technologies to be profiled in the handbook.
                          Table 1
    Leachate Characterization Data: Concentration of Contaminants
Parameter
Biochemical oxygen demand
(5-day)(BOD5)
Chemical oxygen demand (COD)
Total organic carbon
Total suspended solids
Total dissolved solids
pH, s.u.
Alkalinity (as CaCOj)
Hardness (as CaCOj)
Total Kjeldahl nitrogen
Ammonia-nitrogen
Nitrate-nitrogen
Nitrite-nitrogen
Total phosphorus (as P)
Phosphate (as P)
Sulfate
Sulfide
Chloride
Calcium
Magnesium
Sodium
Potassium
Cadmium
Chromium
Copper
Iron
Nickel
Lead
Zinc
Mercury
Love
Canal*
NRd
5,900-11.500
1,800-4,300
200-400
15,700
5.6-6.9
NR
NR
5
1
<0.1
<0.1
<0.1-3.2
<0.1
240
<0.1
9,500
2.500
NR
1,000
NR
0.01
0.27
0.54
30-330
0.24
0.3-0.4
0.48
<0.001
Stringfellpw
Acid Pits
<60
3,600
NR
6,300
24,900
3.37
NR
NR
30
10
14.0
<0.5
NR
2.5
NR
NR
300
400
540
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
G.R.O.W.S.
Landfillc
4,500-13,000
11,200-21,800
NR
500-2,000
11,200-14,200
6.6-7.3
4,800-5,700
3,100-5,100
600-1,700
700-2,000
NR
NR
NR
2.6-3.0
110-680
NR
3,100-4,800
650-900
250-650
1,200-1,500
950-970
0.04-0.10
0.16-0.43
0.32-0.44
180-380
0.55-2.0
0.4-0.8
8.7-31
0.005-0.012
Range
'60-13,000
3,600-21,800
1,800-4,300
200-6,300
11,200-24,900
3.3-7.3
4,800-5,700
3,100-5,100
5-1,700
1-2,000
<0. 1-14.0
<0.5
<0.1-3.2
<0. 1-3.0
110-680
<0.1
300-9,500
400-2.500
250-650
1,000-1,500
950-970
0.01-0.10
0.16-0.43
0.32-0.54
30-380
0.24-2.0
0.3-0.8
0.48-31
<0. 001-0. 012
aReference 1.
bReference 2.
cReference 3.
dNR = not reported.
eAll data except pH are in mg/L
                                                                             TREATMENT OF HAZARDOUS WASTES   175

-------
                                     FLOATING AERATOR
 "Vf
                   77/777
                          Figure 1
                  Equalization Basin Geometry


UNIT TREATMENT OPERATIONS

Equalization
  The objective of equalization is to dampen influent flow and
concentration fluctuations  to improve performance of  down-
stream processes.  Because the composition and volume of leach-
ate from a given landfill fluctuate with time, equalization should
be  considered in  the planning and design of all leachate treat-
ment facilities.
  Equalization basins may be of concrete, steel or (lined) earthen
construction.  With in-line  equalization, the entire  daily flow
passes through the basin; with off-line  equalization, only  that
portion of the flow outside of an acceptable concentration range
is diverted into a basin. When the  goal is equalization  of the
pollutant mass loading (as in the treatment of hazardous waste
leachate), the in-line arrangement must  be  used. Typical basin
geometry is illustrated in Figure 1.
  Mixing and aeration of equalization basins should be provided
to prevent both the deposition of solids  and the onset of septic
conditions.  As with any process  using  aeration, volatile com-
ponents can be stripped from the wastewater, which results in air
emissions of potentially toxic substances.
  Table 2 presents capital and operating and maintenance (O&M)
equalization costs. These costs are based on the use of a concrete
basin with a 24-hour retention time. In general,  capital costs of
equalization systems are moderately high because of the size of
the basins. For large basins, power costs for operating the aera-
tion system contribute significantly to the  overall O&M costs.

                           Table 2
                   Leachate Treatment Costs*
                        (1985 dollars)
Unit operation
Equalization
Air stripping
Prec/floc/sed
Neutral 1zat1on
Activated sludge
Carbon adsorption
Reverse osmosis
25 gal/mln
Capital
101,000
34,500
115,000
10,000
137,000
79,400
582,000
04MD
0.422
0.472
2.66
0.09
0.792
2.35
6.92
50 gal/m1n
Capital
118,000
76,000
140,000
15.000
183.000
104,800
885,000
04H
0.243
0.236
1.62
0.08
0.517
1.72
3.60
100 gal/min
Capital
144.000
127,000
191,000
23,000
242.000
202,800
1,342,000
OiM
0.172
0.118
1.03
0.07
0.328
1.18
2.10
aReference4.
bS/lOOOgal based on 300 day/yr operation.
Air Stripping
  Air stripping is a mass-transfer operation that uses forced air to
remove pollutants from a liquid phase. Applications of this pro-
cess include the removal of ammonia from wastewater and, more
recently, the removal  of volatile organic contaminants  from
drinking water. In the treatment  of leachate, air stripping can
be used ahead of biological processes (to reduce toxic concen-
trations of ammonia) or ahead of adsorption processes (to reduce
the organic loading and thereby extend the life of the sorbent).
  Among the various aeration  devices that can  be used in air
stripping are diffused aerators,  mechanical aerators, spray tow-
ers and countercurrent  flow packed towers. Packed towers pro-
vide  the greatest gas-liquid interfacial area for mass transfer but
the shortest contact time. Removal efficiencies are dependent on
the height of the tower and the air-to-liquid ratio.
  Figure 2  is a schematic diagram of a packed-tower air strip-
ping unit. The unit consists of a vertical column  with randomly
dumped packing on  a  packing  support. Common packings in-
clude the Raschig ring,  Berl saddle, Intalox saddle and Pall ring.
Liquid is distributed  uniformly on the top of the packing with
sprays  or distribution trays and  flows downward by gravity. Air
is blown upward through  the packing and flows countercurrently
to the  descending liquid. Depending on the concentration and
volatility of the contaminants, air-to-water volumetric ratios may
range from 10 to 1  up to  300 to I.' Air stripping  of ammonia is
carried out at elevated pH  (above 11).
                                      AIR PLUS VOLATILES
     LEACHATE
OEMISTER


LIQUID DISTRIBUTOR




PACKING



LIQUID REDISTRIBUTE




PACKING




PACKING SUPPORT


EFFLUENT
                                                                          BLOWER
                                                                                          Figure 2
                                                                 Schematic Diagram of Air Stripping in a Countercurrent Packed Tower
  The volatile organics or ammonia stripped from the leachate
during treatment in a packed tower subsequently may be removed
from the off-gas to prevent air emissions of toxic compounds
or, possibly, for economic recovery. It also may be necessary to
pretreat the leachate for removal of suspended solids and dis-
solved metals that are oxidized to an insoluble form when aerated
to prevent fouling of the packed bed.
  Capital and O&M costs for air stripping, presented in Table 2,
are based on the use of a countercurrent packed tower and in-
clude influent pumping and an ammonia recovery section. Costs
for labor, materials and power comprise the O&M costs of this
system.  Because air requirements,  and hence  power  require-
ments, increase at  lower temperatures,  operation of packed
towers at temperatures below freezing may be cost-prohibitive.
176   TREATMENT OF HAZARDOUS WASTES

-------
 Precipitation/Flocculation/Sedimentation
   The most common method of removing suspended solids and
 soluble  heavy metals  (including arsenic,  lead,  cadmium  and
 chromium) from leachate is by combined precipitation/floccula-
 tion/sedimentation. Precipitation involves the addition of chem-
 icals to leachate to form insoluble precipitates from soluble spe-
 cies. Flocculation promotes agglomeration of suspended solids,
 which makes them easier to remove.  Sedimentation is the removal
 of suspended particles by gravity settling.
   The processes of precipitation, flocculation and sedimentation
 can be carried out in separate basins, as shown in Figure 3, or in a
 single basin (e.g., an upflow solids-contact reactor-clarifier) with
 separate zones for each process. Precipitation requires rapid mix-
 ing  (10 to 60 sec) to disperse the chemical, whereas flocculation
 requires slow and gentle mixing (15 to 30 min) to promote particle
 contact. Ancillary equipment requirements for these processes in-
 clude mixers/paddles, chemical storage tanks and chemical feed
 pumps. Packaged plants are available for low flow rates (10,000
 gal/day to 2 million gal/day), but  these may require extensive
 modifications to enable them to handle the high level  of precipi-
 tated solids characteristic of hazardous waste leachates.
       CHEMICAL
      PRECIPITANTS
     CHEMICAL
    FLOCCULANTS
         RAPID-MIX TANK
                            M    M
                                     PAOOLES

L>

43D-

-D


i

                         FLOCCULATION CHAMBER
                                           SEDIMENTATION TANK
                           Figure 3
    Schematic Diagram of Precipitation/Flocculation/Sedimentation
   Metals can be precipitated from leachate as hydroxides or sul-
 fides through the addition of lime or ferric sulfide. Hydroxide
 precipitation with lime at high pH is the most commonly used
 method. The presence of complexing agents such as ammonia or
 cyanide  in the  leachate, however, will  inhibit precipitation  of
 some heavy metals as hydroxides. Flocculants such as alum, ferric
 chloride or polyelectrolytes often are added to the rapid-mix tank
 along with the  chemical precipitant  to  reduce repulsive forces
 between particles and  bring  about  particle  aggregation and
 settling.
   Precipitation/flocculation produces  large amounts  of wet
 sludge that may be hazardous because of its heavy metals con-
 tents. This sludge, which is settled in  a sedimentation unit, must
 be treated further (e.g.,  by dewatering or fixation/stabilization)
 before it is iandfilled.
   Fluctuating leachate quality requires frequent jar  testing to de-
 termine appropriate  chemicaraosages and removal efficiencies.
 Available data indicate that precipitation/flocculation/sedimen-
 tation can provide good removal of suspended solids (80 to 90%)
 and  moderate removal of BOD5 (40  to  70%) and  COD (30  to
 60%),' The additional treatment required will depend on effluent
 discharge limitations.
  Table  2 presents costs of a packaged precipitation/floccula-
tion/sedimentation system that includes a granular filtration unit.
Operation of these systems is labor-intensive, as reflected by the
moderately high  O&M costs.
 Neutralization
   Neutralization involves the addition of an acid or a base to an
 aqueous waste stream to adjust its pH to the desired level (usual-
 ly between 6.0 and 9.0). Neutralization may be required as a pre-
 treatment step to optimize the performance of downstream treat-
 ment, to protect  pH-sensitive processes (particularly biological
 teatment operations) or as a final step to meet effluent criteria.
 The technology is inexpensive, highly developed and widely used
 in the treatment of hazardous waste leachate.
   Typically, neutralization is carried out in completely mixed
 corrosion-resistant tanks as shown in Figure 4. The tanks may be
 operated in batch or continuous mode; however, the latter is only
 suitable for flow rates greater than about 100,000 gal/day. The
 neutralization process usually is controlled automatically by feed-
 back, feedforward or multimode controllers. Common bases used
 for neutralization include lime, calcium hydroxide, caustic,  soda
 ash  and ammonium hydroxide; common  acids include sulfuric
 acid, hydrochloric acid and nitric acid. Reagent selection is based
 on cost, speed of reaction and the reaction byproducts formed.
   Equalization usually precedes neutralization in the treatment
 process train to control fluctuations in influent flow and  con-
 centration. The neutralization process can produce heat and toxic
 gases (ammonia, hydrogen sulfide and hydrogen cyanide) if the
 chemicals are not properly added to and mixed with the waste.
 Treatment to remove precipitated solids (flocculation/sedimen-
 tation or filtration)  often is required after neutralization. If the
 removed solids are hazardous, secure disposal will be required.
   Table 2 presents costs of a neutralization system consisting  of
 an agitated tank with a 3-min retention time, a metering pump
 for acid or caustic and a pH control loop and valve. Both capital
 and O&M costs are low in comparison with other unit treatment
 operations.
                     pH CONTROLLER
ACID OR CAUSTIC-
                  2rl
                  FEED
                  PUMP
        LEACHATE	
                                                                                                                • MIXER
                                                                                                                 	" EFFLUENT
                              MIXING TANK
                           Figure 4
               Schematic Diagram of Neutralization

Activated Sludge
  Activated sludge is a biological treatment process in which an
active mass of microorganisms (bacteria,  protozoa, rotifers and
fungi) is used to  stabilize biodegradable organic matter under
aerobic conditions. Activated sludge can reduce concentrations of
a wide variety of organic  compounds, including many toxic and
hazardous compounds. It is widely used in municipal and indus-
trial wastewater treatment applications (typical BODj concentra-
tion of 200 mg/1) and can effectively treat leachate with organic
concentrations one to two  orders  of magnitude higher (up to
10,000 mg/1 BOD5."
  Conventional activated  sludge systems,  illustrated in Figure 5,
include  two stages. The first stage  involves aeration of the waste
in an open tank and maintenance of an active biomass. Aeration
is accomplished by mechanical-surface, diffused-air or sparged-
turbine aerators. The second stage entails separation of the solids
in a secondary clarifier. A  portion of the solids is recycled back to
the aeration basin to maintain the desired concentration of organ-
isms, and the remaining portion is wasted.
                                                                                 TREATMENT OF HAZARDOUS WASTES     177

-------
                   AERATOR
LEACHATE




i 	
C
>
, 	 I
D
AERATION TANK
RECYCLED SLUDGE ._
                                                     EFFLUENT
                           Figure 5
        Schematic Diagram of Conventional Activated Sludge
  Variations of the conventional activated  sludge system have
been developed to provide greater tolerance  for shock loadings,
to improve  sludge settling characteristics and to achieve higher
BOD5  removals.  Process modifications  include complete-mix,
step aeration, extended aeration, contact stabilization and the use
of pure oxygen.
  Because of the sensitivity of biological systems,  pretreatment
requirements for high-strength leachate are extensive and may in-
clude equalization to buffer hydraulic and  organic load varia-
tions;  sedimentation/flotation to remove suspended solids, oil
and grease;  neutralization to adjust the pH to near neutral; and
nutrient  addition to provide adequate levels of nitrogen, phos-
phorus and  trace elements. Acclimation of the biological system
to the influent waste stream is necessary prior to full-scale opera-
tion of the  process. The presence of refractory or biologically
toxic compounds (e.g., ammonia or heavy metals) in the leachate
may necessitate the use of other physical/chemical processes in
conjunction with biological treatment.
  Process residuals from activated sludge treatment of leachate
include waste activated sludge, which may be high in metals and
refractory organics,  and air  emissions of volatile organic com-
pounds that are stripped from the waste during aeration.
  In general, biological processes such as activated  sludge treat-
ment  are the most cost-effective means  for removing organics
from high-strength leachates. Table  2 presents costs of a conven-
tional activated sludge system consisting of an aeration tank with
a 6-hr retention time and a mixed liquor suspended solids con-
centration of 2000 mg/1, a secondary  clarifier and a  sludge re-
cycle pump. The O&M costs of activated sludge treatment  are
comprised principally of labor and materials.

Carbon Adsorption
  Carbon adsorption is a separation technique for removal of dis-
solved organics and certain inorganics (e.g., cyanide and chrom-
ium) from aqueous waste streams.  This  well-developed process
has numerous  full-scale applications including treatment  of
domestic and industrial wastewaters,  cleanup of spilled  haz-
ardous wastes. In leachate treatment applications, carbon adsorp-
tion can  be  used  in the pretreatment,  intermediate or polishing
steps.
  Treatment of leachate by carbon adsorption  involves passing
the waste stream through beds of granular activated carbon. Acti-
vated  carbon is an amorphous form of carbon  characterized by
a large internal surface area. Contaminants are adsorbed from the
waste onto the carbon  surface by physical and  chemical forces.
When the adsorptive capacity of the carbon has been reached,
regeneration or disposal of the spent carbon is required.
  Most carbon adsorption systems use cylindrical pressure vessels
arranged in series or parallel and operated in  a downflow, up-
flow or pulsed-bed mode. The downflow mode can handle a high-
er concentration of influent suspended solids than can the other
modes; however, frequent backwashing of the carbon beds may
be required to prevent excessive pressure drop. A schematic dia-
gram of an alternating, two-column, downflow system is pre-
sented in Figure 6.
                                                                 PRETREATED
                                                                  LEACHATE
 COLUMN IN
 ADSORPTION	«•
   STAGE
                                                  BACKWASH PLUS
                                                  CONTAMINANTS
               EFFLUENT
                           Figure 6
     Schematic Diagram of Alternating, Two-Column, Downflow,
                   Carbon Adsorption System
  Carbon adsorption  is applicable  to  the treatment of many
toxic and refractory organics; consequently, it often is used in
combination with biological treatment. Because the carbon beds
are susceptible to clogging, leachate must be pretreated  to re-
move high concentrations of suspended solids, oil and grease. In
general, influent concentrations  are limited to 10,000 mg/1 total
organic carbon,  1,000  mg/1 total inorganics,  2,000 mg/1  sus-
pended solids (downflow),  50 mg/1  suspended solids (upflow)
and 10 mg/1 oil and grease.' With proper pretreatment, removal
efficiencies greater than 99"% can be obtained.'
  Table 2 presents capital and O&M costs of a carbon adsorp-
tion system. The  O&M costs of this system are governed by the
organic loading, which is assumed to be 10 Ib organics per 100 Ib
of carbon. These costs do not include the costs of disposal or re-
generation of the spent carbon.  For plants using less than 200
Ib/day of carbon (or,  equivalently, for plants treating less than
800,000  gal/day  of leachate),  on-site  regeneration  of carbon
probably  is not economical.' Most leachate treatment facilities
will fall into this range.

Reverse Osmosis
   Reverse osmosis is  a membrane separation  technique which
primarily is utilized for the demineralization of water. Poten-
tially, this technique could be used as the final polishing step in
the treatment of leachate to separate dissolved solids [inorganic
salts and high (>120) molecular weight organics] from  secon-
dary treatment effluent. With current technology, reverse osmosis
is not practical for the  treatment of raw leachate because of rapid
fouling of the membrane and poor selectivity for low molecular
weight organics.'
   In reverse osmosis, water is separated  from dissolved solids in
solution by filtering it through  a semipermeable membrane at
pressures of 200  to 1200 Ib/in1.' Basic components of a reverse
osmosis unit are  the membrane,  a membrane support structure,
a containing vessel and a high-pressure pump; these are illus-
trated schematically in Figure 7.
   All  commercially available membranes are structured asym-
metrically, with a  thin (0.1 to 1.0 /un)  dense  surface ("skin")
supported by a porous substructure.  This design promotes high
178    TREATMENT OF HAZARDOUS WASTES

-------
water transport across the membrane.8 Cellulose acetate and aro-
matic polyamides are common membrane materials.
  Membrane support structures are of four types: tubular, spiral-
wound, hollow-fiber and plate-and-frame. The tubular configur-
ation, which consists of a rigid-walled porous tube lined on the
inside with  a membrane, is recommended for use  with waste-
water effluents.6 Reverse osmosis units can be arranged in parallel
(to increase hydraulic capacity) or in series (to effect the desired
degree of separation).
  For optimal performance, feed water to the reverse osmosis
unit must be pretreated  to remove gross amounts of solids and to
prevent fouling by scaling or biological growth inside the device.
In addition, compounds that are incompatible with the mem-
brane must be removed during pretreatment. Because the com-
position and strength of hazardous waste leachates vary widely,
pilot testing will be necessary to determine pretreatment require-
ments and separation performance. A major disadvantage of this
technology is that it produces a highly concentrated reject stream
that requires further treatment or disposal.
  Table 2 presents costs of a reverse osmosis unit operating at
800 lb/in2. The O&M costs of this unit do not include the costs of
disposing of the concentrated reject  stream. Regardless, reverse
osmosis is an extremely expensive technology and should be con-
sidered  only for leachate treatment applications when an effluent
of extremely high quality is required.

                            PRESSURE  VESSEL
             HIGH-PRESSURE
                PUMP
                                                   -»• EFFLUENT
                                        'SEMI PERMEABLE
                                           MEMBRANE
                     CONCENTRATED
                     REJECT STREAM
                           Figure 7
              Schematic Diagram of Reverse Osmosis
CONCLUSIONS
  The highly variable, complex nature of hazardous waste leach-
ate often requires that several  unit treatment operations be
applied  in combination to effect  the  desired degree of contami-
nant removal. Equalization is used to buffer influent flow and
concentration fluctuations  and to improve the performance of
more costly downstream operations. Air  stripping,  neutraliza-
tion and chemical precipitation often are used ahead of biolog-
ical treatment (activated sludge) to remove biologically toxic com-
pounds  such as ammonia and heavy metals and to adjust the pH
to near  neutral. Carbon adsorption and reverse osmosis are used
as polishing steps to remove refractory organics and to meet final
effluent criteria. Unit treatment costs range from $0.07/1000 gal
for neutralization to $2.10/1000 gal for reverse osmosis.
REFERENCES
1.  Shuckrow, A.J.,  Pajak, A.P. and Touhill, C.J.,  "Management of
   Hazardous Waste Leachate," U.S.  EPA  Publication No. SW-871,
   1982.
2.  Copa,  W.M., et al.,  "Powdered Activated  Carbon  Treatment
   (PACT™) of Leachate From the Stringfellow Quarry," Proc. of the
   Eleventh Annual  Research Symposium on Land Disposal, Remedial
   Action, Incineration and Treatment of  Hazardous  Waste, EPA-
   600/9-85-028, Sept. 1985, 52-65.
3.  Steiner, R.L., Keenan, J.D.  and Fungaroli, A.A., "Demonstrating
   Leachate Treatment: Report on a Full-Scale Operating Plant,"  U.S.
   EPA Publication No. SW-758, 1979.
4.  U.S. EPA, Handbook, Remedial Action at  Waste Disposal Sites,
   EPA-625/6-82-006, 1982.
5.  Metcalf & Eddy,  Inc., Briefing: "Technologies Applicable  to  Haz-
   ardous Waste," prepared for the U.S. EPA Hazardous Waste Engi-
   neering Research Laboratory, Cincinnati, OH, May 1985.
6.  Metcalf & Eddy, Inc., Waste-water Engineering: Treatment, Disposal,
   Reuse, 2nded., McGraw-Hill, Inc., New York, NY, 1979.
7.  Chian, E.S.K. and DeWalle, F.B.,  "Evaluation of Leachate Treat-
   ment, Vol. II: Biological and Physical-Chemical Processes," EPA-
   600/2-77-186b, 1977.
8.  Applegate, L.E., "Membrane Separation Processes," Chemical Engi-
   neering, 71, 1984,64-89.
                                                                                  TREATMENT OF HAZARDOUS WASTES     179

-------
             Manuscript Withdrawn at the Request
                    of The PQ Corporation
180  BARRIERS & WASTE SOLIDIFICATION

-------
Manuscript Withdrawn at the Request
       of The PQ Corporation
                            BARRIERS & WASTE SOLIDIFICATION  181

-------
             Manuscript Withdrawn at the Request
                    of The PQ Corporation
182  BARRIERS & WASTE SOLIDIFICATION

-------
Manuscript Withdrawn at the Request
       of The PQ Corporation
                            BARRIERS & WASTE SOLIDIFICATION  183

-------
             Manuscript Withdrawn at the Request
                    of The PQ Corporation
184  BARRIERS & WASTE SOLIDIFICATION

-------
Manuscript Withdrawn at the Request
       of The PQ Corporation
                             BARRIERS & WASTE SOLIDIFICATION   18~5

-------
                      Innovative  Techniques for  the  Evaluation
                        of Solidified Hazardous  Waste  Systems

                                                 Harvill C. Eaton
                                               Marty E. Tittlebaum
                                               Frank K. Cartledge
                                       Hazardous Waste Research  Center
                                           Louisiana State University
                                             Baton  Rouge, Louisiana
ABSTRACT
  Hazardous  wastes  solidified  or stabilized  by mixing with
cementing materials are described as complex composiie materials
with microslruclurcs and microchemisiries which determine their
properties. The behaxior of these materials and their possible
long-term effect on the environment can be determined  by  ex-
amining them  with the tools common to  studies  in materials
science and engineering. Several of the microscopic and  micro-
analytical techniques useful in these investigations are described
and selected results are shown. These techniques, together with
measurements  of bulk physical properties, currently are being
utilized by the present authors to determine the mechanisms of
waste stabilization. The results  are being used to design effec-
tive methods  to handle real-world organic and  inorganic haz-
ardous industrial wastes.

INTRODUCTION
  Certain types of hazardous wastes  can be converted  into a
manageable form by solidification. For  example, aqueous sludges
containing inorganic metal ions can be combined  with various
cementing mixtures to form solids which are often disposable in
a secure landfill. Several advantages are potentially afforded by
solidification, including little or no free-standing  liquids and re-
duced hydraulic pressures which might cause failure of the liner
system. If the  permeability of the solidified waste is low, then
the leachate from monolithic structures may be free of toxic ions
after their initial depletion from exposed surfaces.
  In the most desirable situation, the metal ions react chemically
with the matrix thereby forming a complex which is insoluble in
aqueous media and which is stable when leached by solutions with
a pH near those common in the environment or in a landfill.
Therefore, under the ideal situation, the waste is rendered effec-
tively non-hazardous. Consequently, solidification/stabilization
is potentially an important management alternative which may be
attractive for many hazardous wastes.
  Unfortunately, little is known about  the mechanisms of waste
solidification and even less about stabilization.  Until recently,
system design was essentially empirical and based on  industrial re-
search which usually remained proprietary. The relevant informa-
tion in the open literature dealt with nuclear waste stabilization'
or cement admixtures2 and, although useful, it more often than
not concerned  the interaction of cementing matrices with com-
pounds which  are not important to hazardous chemical waste
management.
  Recently, however, it has been recognized that advances can be
made toward understanding the mechanisms of solidification and
stabilization by regarding the products as "complex composite
materials" with a  "microstructure".1  Mechanistic information
which might result from this viewpoint is essential to establishing
u fundamental basis for the a priori design of a solidified or stabil-
i/cd waslc.
  A material science point  of view suggests that the measurable
properties (e.g., unconfined compressive strength and toxic ion
leachability) of the waste/matrix systems depend on the micro-
structure and  the microchemistry. The  experimental techniques
commonly used by  the materials scientist offer ways to examine
the waste/matrix  systems which may provide information about
the mechanisms of solidification and/or stabilization.
  In this paper, the authors discuss the techniques that they have
found useful in their recent  studies of solidified inorganic sludges
and of solidified organics.' The methods  can  be  conveniently
grouped into those which provide morphological information and
those  which provide chemical information. They are briefly de-
scribed and representative results are presented.

SAMPLE PREPARATION
  The authors' investigations primarily have involved mixtures of
hazardous chemical wastes and Type I Portland cement. Portland
cement was chosen  because it is a  common component  of many
solidification formulae which have been proposed and  of many
which currently are being  used by industry. Mixtures of pure
Portland cement and hazardous wastes are unlikely to be environ-
mentally, technically or economically practical, but they do offer
some of the laboratory control that is important  to a basic study.
Other, more complicated, matrices presently are  being examined
in the authors' laboratories as  part of a study of  interference
mechanisms. The work is funded by the U.S. EPA and  the U.S.
Army Corps of Engineers.
  A second application of the methods is part of an investigation
of test methods for evaluating real solidified wastes. The latter is a
cooperative study involving the U.S. EPA, Environment Canada
and several vendors in  the United States,  Canada and abroad.
  Weighed amounts of waste and water were mixed with 10 g of
Portland cement, contained in capped vials and stored at room
temperature in a darkened room to avoid photochemical reaction
of the organics. Samples were allowed to cure for various times
ranging from 1 week to over 1 year.
  Three organics  have been investigated:  p-bromophenol,  p-
chlorophenol  and ethylene  glycol. The  reasons for  selection of
these compounds and  detailed descriptions  of their morphology,
microchemistry and leaching behavior have been described else-
where.5'6

MICROSTRUCTURE
  The microstructure of a solidified waste depends on the follow-
ing factors:
186   BARRIERS & WASTE SOLIDIFICATION

-------
• The chemical composition of the waste
• The chemical composition of the solidifying matrix compon-
  ents (cement, fly ash, clay, etc.)
• Relative amounts of waste and matrix components
• Amount of water
• Extent of mixing
• Temperature of mixing
• Time and temperature of set
  In the studies  performed by  the authors, these  factors have
been considered and maintained constant when appropriate, even
though in actual  practice most of the factors can vary consider-
ably.  It will be important,  therefore, for researchers to estab-
lish how strongly these factors affect the gross properties of the
final product.

Optical Microscopy
  Microstructure can be observed at several microscopic levels.
For example, entrained air can form bubbles ranging in size from
0.1 to 1 mm in diameter. These bubbles may be important in the
development of mechanical strength in the solidified product. Air
bubbles also can be the source of carbon dioxide to the matrix.
Carbonation can affect the hydration of cement paste by promot-
ing the formation of calcium carbonate which then decomposes to
a mixture of calcium carbonate and hydrated silica.7 Microstruc-
tural features of this size are  resolved easily with the optical
microscope.
   There are several types of optical microscopes. They are con-
veniently categorized according to the branch of science which
they  serve  (e.g., the biological  microscope,  the  petrographic
microscope and the metallographic microscope).  The petro-
graphic microscope and, to a lesser extent, the metallographic
microscopes are the most useful for examining microstructures in
solidified and stabilized wastes.
                                                    '.'.
                                                     /"^   -..;
                                                 VfeP
                                                 100^, "Vf
                                                 	; «>w
                                                       '  O
                          Figure 1
   An Optical Micrograph of a Solidified P-Bromophenol Sample. A
               Bimodal Grain Structure is Revealed.
  Thin sections can be prepared, using methods standard to geo-
logical practice, so that light can be transmitted directly through
a small piece of solidified material. Using this method, the grain
structure can  be observed. Figure  1 is  an optical micrograph
which reveals the polycrystalline structure of solidified p-bromo-
phenol.  The Portland cement matrix material is apparent in the
microstructure with small grain-like hydration  phases.  There is
an equiaxed grain structure which is clearly visible at 100X mag-
nification, and the presence of several mineral phases is apparent.
If polarized light is used, the crystalline and noncrystalline com-
ponents of the  solidified system can be distinguished.  It also is
possible to do phase analysis of the crystalline components us-
ing index of refraction liquids. It is, therefore, in principle, pos-
sible to determine the compounds formed due to the presence of
the hazardous waste. In practice, this determination can be very
difficult and the researcher likely will more successfully identify
than locate the waste containing phases.  This identification is
illustrated in Figure 2 where the arrow indicates a mineral, i.e.,
crystalline, phase which contains hazardous p-bromophenol.
                                                    u.
                           Figure 2
  A Higher Magnification Optical Micrograph of the Material Shown in
    Figure 1. The Arrow Points to a Phase Which is Concentrated in
                       P-Bromophenol.

  When only morphological information is required, such as in
studies of air entrainment, the metallurgical microscope is use-
ful. With the metallurgical microscope,  light is directed down the
microscope tube and is reflected off of the specimen  surface.
Reflected light travels back up the tube and through the ocular to
be detected by the operator's eye or to be recorded on a photo-
graphic plate or film. The  chief advantage  in  the use of the
metallurgical microscope over the petrographic microscope is in
specimen preparation. Since light is not transmitted through the
specimen, only a polished surface is required.

Scanning Electron Microscopy
  Microstructure on a smaller scale is apparent when the several
minerals  formed upon hydration of Portland cement are  con-
sidered. For example, calcium hydroxide crystallites are formed
and often are observed as hexagonal plates several microns in
extent. Even smaller crystallites of ettringite are  observed in
young cement pastes. Ettringite usually appears as rod or needle
shaped, single-crystals. They contain sulphur and are only a few
microns long.  The ettringite group  of minerals form the AFt
(A  = Al, F = Fe and t =  three moles of CaO.SO3)  phases.'
Both the calcium hydroxide and the AFt phases are observed best
by scanning electron microscopy (SEM).
  In the scanning electron microscope, a primary electron beam
is focused onto a specimen surface and rastered over a small area.
Primary electrons interact with the waste and matrix atoms con-
tained in the specimen producing additional electrons. These elec-
trons are released from the specimen atoms themselves and can be
                                                                                  BARRIERS & WASTE SOLIDIFICATION    187

-------
released through several  mechanisms. The two electron genera-
tion mechanisms particularly useful for waste studies are secon-
dary electrons and backscattered electrons. A secondary electron
detector or a backscattered electron detector placed near the spec-
imen can be used to produce an image of the variations in their
generation as  the primary beam is rastered across  the specimen
surface. The utility of the SEM is related to the following char-
acteristics of its use and image contrast mechanisms:
•  Magnification variable from 10X to ca. 100X with a resolution
   of up to 5 nm or less
•  Magnification variable
•  Large depth of field
•  East of specimen preparation
   Specimens can be prepared by grinding and polishing in a man-
ner identical to that used for the preparation of specimens for
the petrographic microscope or for the metallurgical microscope.
The authors, in fact, frequently use the same specimen for both
types of examination. If three mutually orthogonal sections are
examined, the researcher usually can eliminate the  possibility of
mistaking textured for uniform structure. Polished sections must
be carefully studied, however, to eliminate artifacts caused by the
preferred  removal of certain phases during grinding and polish-
ing. Fractured surfaces can be useful in assuring that this is not
occurring. Figure 3 is an example of a scanning electron micro-
graph of solidified ethylene glycol.
                          Figure 3
    A Scanning Electron Micrograph of a Solidified Ethylene Glycol
                          Sample
Transmission Electron Microscopy
  When Portland cement is a component of the solidification or
stabilization matrix, a major hydration product is calcium sili-
cate hydrate (CSH). CSH is a poorly crystalline material which,
like the AFt phases, is comprised of several different compounds
which are mineralogically  similar.  When viewed by the optical
microscope or the SEM, CSH-appears as the dominant or matrix
phase of a solidified hazardous waste. The SEM sometimes  is
able to resolve Its morphology but  frequently reveals it as a fea-
tureless, but sometimes nodular  or grainy, material in which well
developed crystals like calcium hydroxide or AFt grow.
  The transmission electron microscope (TEM) is the instrument
of choice  for studying the morphology of CSH and other  very
small structures. The  TEM  is capable of resolving features as
small as a few angstrom units in cementitous materials. Unfor-
tunately, specimen preparation usually is more difficult than for
either optical or scanning electron microscopy.
   A stationary electron beam is focused onto a specimen which
intersects the optical axis of the instrument. The transmitted
beam is scattered and produces an image with contrast which is
primarily a function of the mass-thickness of the noncrystalline
components.  Crystalline  components scatter  the  specimen in
more complex ways due to diffraction. Figure 4 is a transmission
electron  micrograph of a small  crystalline phase in a solidified
waste.
                                                                                            Figure 4
                                                                   A Transmission Electron Micrograph of a Ca(OH)j Crystal (the large
                                                                  dark structure at (he center of (he micrograph) in a Solidified Ethylene
                                                                                         Glycol Sample
MICROCHEM1STRY
  Bulk  analytical instruments are an important and necessary
part of any investigation of the solidification and stabilization of
a hazardous chemical waste. If mechanistic information is desir-
able, it  is useful to have chemical analyses of the small phases
present  in the solidified material. Consequently, special methods
must be used.

Electron Diffraction
  The high resolution images  obtainable from the transmission
electron microscope and the ability to photograph diffraction pit-
terns make the TEM very useful for solidification/stabilization
research. Chemical  identification of mineral components of the
solidified waste is possible by analysis of the electron diffraction
patterns.
  Figure 5 is an example of the results which can be obtained
from examination of a solidified hazardous waste by transmission
electron microscopy. The dark structure indicated by the arrow is
a calcium hydroxide crystal as revealed by the diffraction pattern.

Energy Dispersive and Wavelength Dispersive
X-Ray Analysis
  Specimens examined in the SEM emit x-rays in addition to elec-
trons. The x-ray energies are characteristic of the elements from
which they come. This means that they can be used to determine
the elemental composition of the specimen. The method is known
as energy dispersive x-ray analysis (EDS). Most EDS spectromet-
ers  will  detect  elements with atomic number 11 (Na) or greater.
 188    BARRIERS & WASTE SOLIDIFICATION

-------
                           Figure 5
    A Transmission Electron Micrograph of a Ca (OH)2 Crystal in a
 Solidified P-Bromophenol Sample. The Diffraction Pattern was Used to
             Confirm the Composition of the Crystal.
Those spectrometers equipped with a windowless detector can de-
tect even lighter elements and are particularly useful for carbon
and oxygen determination. A similar technique is commonly asso-
ciated with electron probe microanalysis (EPMA) and is known as
wavelength dispersive x-ray analysis (WDS).
  It is obvious that EDS and WDS would be useful in studies of
inorganic metal ion stabilization, but in certain cases they also can
be used in studies of organic waste solidification and stabiliza-
tion. The authors have used the techniques to study the partition-
ing of selected organics to  the hydration products of Portland
cement. Two organics, p-bromophenol and p-chlorophenol, were
selected because they were molecules which contained a heavy ele-
ment  which could be detected by EDS and WDS. The crystalline
phase shown  by the arrow in Figure 2 was determined by WDS to
contain large amounts of Br and, therefore, to be rich in organic.

X-Ray Power Diffraction
  X-ray powder diffraction (XRD) is very useful for studies of
waste solidification. A bulk analytical technique, XRD often can
be used  to identify the exact chemical compounds present in the
solidified system. This technique is generally  familiar, so it will
not be discussed in this paper.

CONCLUSIONS
  Studies of hazardous waste solidification/stabilization are com-
plicated by the fact that cementing mixtures are used as cementing
matrices.  Upon  hydration, a complex composite material is
formed. The microstructure of this material has structure at levels
which vary from that visible to the eye to that which can only be
resolved by electron microscopy. The properties of the solidified
system are determined by the microstructure and the microchem-
istry.
  Incorporation of hazardous wastes into the matrix by mixing
prior to hydration introduces the possibility that the wastes them-
selves are chemically  involved with the hydration reactions and
therefore are stabilized by the process. In addition, there is a pos-
sibility that the wastes are physically contained by the matrix in
pores or interstitial spaces. These sites of physical entrapment
may be so small that they are difficult to locate even by micro-
scopic methods.
  These characteristics suggest the use of microscopic and micro-
analytical methods for studying the mechanisms of waste solidifi-
cation and stabilization. The various methods used by the present
authors have been discussed. Some of the results of these studies
have been shown in the form of photomicrographs which reveal
selected characteristics  of cement/organic waste  solidification
mixtures.

ACKNOWLEDGEMENTS
  The authors would like to thank the U.S. EPA,  Environment
Canada, the  U.S. Army Corps of Engineers and the LSU Haz-
ardous Waste Research Center for financial support of this work.
Special  thanks go to  the following individuals for their sugges-
tions,  conversations  and guidance:  Carlton  Wiles,   Charles
Meshni, Phil Malone, Jerry Jones, Trevor Bridle, Pierre Cote and
Roger Seals. Special  thanks  go to  our students: Marie Walsh,
Donna Skipper, Devi Chalasani and Amitava Roy.

REFERENCES
1. Cooper, J.A., Cousens, D.R., Lewis, R.A., Myhra, S., Segall, R.L.,
  Smart, R. St. C.,  Turner, P.S. and White, T.J., "Microstructural
  Characterization of Synroc C and E by Electron Microscopy,"  /.
  Am. CeramicSoc., 68, 1985,64-70.
2. Ramachandran, V.S., Concrete Admixtures Handbook: Properties,
  Science, and Technology, Noyes Publications, Park Ridge, NJ, 1984.
3. Cartledge,  F.K., Chalasani, D., Eaton,  H.C., Tittlebaum, M.E.
  and Walsh, M., "Combined Techniques to Probe Chemical Inter-
  actions of Organics with Solid Matrices," Proc. of the Meeting of the
  American Chemical Society, Miami, FL, May 1985.
4. Eaton, H.C., Walsh, M.B., Tittlebaum, M.E., Cartledge,  F.K. and
  Chalasani,  D., "Microscopic Characterization of the  Solidification/
  Stabilization of Organic Hazardous Wastes," Proc.  of the Energy-
  Sources and Technology Conference and Exhibition, Dallas, TX,
  Feb.  1985,  American Society of Mechanical Engineers Paper, No.
  85-Pet-4.
5. Walsh, M.B., Eaton, H.C., Tittlebaum, M.E., Cartledge,  F.K. and
  Chalasani,  D., "The Effect of Two Organic Compounds on a Port-
  land Cement-Based Stabilization Matrix," See these Proceedings.
6. Chalasani,  D., Cartledge, F.K., Eaton, H.C., Tittlebaum,  M.E. and
  Walsh, M.B., "The  Effects of Ethylene Glycol on a Cement-Based
  Solidification Process," See these Proceedings.
7. Suzuki,  K., Nishikawa, T. and Ito, S., "Formation and  Carbona-
  tion of C-S-H in Water," Cement and Concrete Research,  15, 1985,
  213-224.
8. Ramachandran, V.S., "Cement Science," in Concrete Admixtures
  Handbook: Properties,  Science, and Technology, V.S. Ramachan-
  dran, Ed., Noyes Publications, Park Ridge, NJ, 1984, 8.
                                                                                   BARRIERS & WASTE SOLIDIFICATION     189

-------
                  Site  Characteristics  and  the  Structural  Integrity
                           of  Dikes  for  Surface Impoundments

                                         Jey K. Jeyapalan,  Ph.D.,  P.E.
                                                  Ernest R. Hanna
                              Wisconsin Hazardous Waste Management Center
                                             University of Wisconsin
                                                Madison,  Wisconsin
ABSTRACT
  Before a land disposal  facility  is developed for hazardous
wastes, the site characteristics need to be determined. The infor-
mation obtained for the geological and hydrogeological features
is  carefully used in the design  of the  surface impoundment.
Special design provisions need to be chosen if the site conditions
are poor.
  The damage due to the failure of dikes containing the wastes in
surface impoundments poses severe risk of contamination of sur-
rounding areas. The requirements of RCRA seek to ensure the
structural  integrity  of dikes by requiring certification  of  these
structures  by registered engineers.  The evaluation of site  char-
acteristics and the structural integrity of dikes is a complex sub-
ject. For the benefit of those involved in dike certification and
review of permit applications, the essential steps involved in these
evaluations are summarized in this paper.

INTRODUCTION
  A surface impoundment design requires adequate data on the
site characteristics,  geometry of the dike and properties of the
soils forming the dike. In addition, information on groundwater
conditions is required to complete the design. Thus, for proper
water disposal techniques, the identification of various features at
the site is an important step. Subsequent  to the determination of
site characteristics, a detailed laboratory testing program needs to
be undertaken to determine the remaining engineering properties
of soils.
  The geometry of the problem is defined based on specifications
or field survey results,  and the  structural integrity of dikes is
determined using the appropriate analysis procedure. The struc-
tural adequacy needs to be ensured for different stages of the life
of the surface impoundment and shear  strength properties ap-
propriate for these stages need to be used in the analyses. Thus,
adequate understanding and recognition of all fundamental prin-
ciples of geotechnical engineering and hydrogeology are necessary
to successfully complete the above steps of engineering analyses.

SITE CHARACTERISTICS
  The first step is the site reconnaissance in which all pertinent in-
formation is obtained. An efficient and economical way to  begin
the site investigation is to review all available literature. This pro-
cess involves a review of past and present land uses, review of site
aerial photography  and  assessment of  the  site topography,
cultural landmarks, surface hydrology and surficial geology. The
location, accessibility, general area topography, past and present
ownership  and/or use  of a site are determined from maps ac-
quired from the local assessor's office and/or from the U.S. Geo-
logical Survey (USGS). The USGS and the Soil Conservation Ser-
vice (SCS) may be utilized as sources of preliminary information
regarding  the surficial geology of the site under investigation.
Site-study reports,  as well as previous plans and specifications,
may be available from area contractors, consultants and govern-
ment  agencies.  Any  construction activity  in  the project  area
should be visited. A site survey should be performed to familiarize
the engineer with the area. High water marks and vegetation can
indicate the nature of the soil. Recent changes in topography also
can be helpful in design procedures.
  Geological conditions  and project objectives will dictate the
type of drilling needed to meet the job requirements. It is not
always possible or cost-effective to use a single drilling method to
complete a project. This is one reason why, prior to specifying a
drilling method, an analysis  of the project requirements is done.
Selection of a drilling  method most suited to the particular job
can be based on the following factors:
• Type of formation
• Depth of drilling required
• Availability of drilling  equipment
• Special and/or other specific  requirements

  The use of a backhoe or shovel  is a very simple and quick
means of determining  shallow  soil conditions  and  shallow
groundwater conditions.  Advantages are the low cost, mobility
and ease of operation. The main disadvantages are the depth
limitation and sample disturbance.

Surface Geophysical Techniques
  The application of geophysics to hazardous waste management
combines proven technology with  new technology to form an in-
tegrated systems  approach.  This  process is efficient and  cost-
effective. Several types of geophysical exploration techniques are
now available for a rapid evaluation  of subsoil characteristics.
Definite interpretation of the results requires  correlation with
standard subsurface exploration methods. Geophysical methods
can be divided into two categories: (1) surface techniques and (2)
downhole  techniques. A  few of these surface geophysical tech-
niques are described below.

Ground Penetrating Radar
  Ground penetrating radar (GPR)  uses high  frequency  radio
waves to  acquire subsurface information. Energy  is radiated
downward from a small antenna which is moved slowly across the
surface of the ground. The energy is reflected back to the receiv-
ing antenna,  where variations in the return signal are continuously
recorded. Radar responds to changes such as bedding,  cementa-
tion, moisture, clay content, voids, fractures, intrusions and man-
made objects. Depth of penetration  is site-specific.
190    BARRIERS & WASTE SOLIDIHCATION

-------
  GPR is limited in depth by attenuation. Better overall penetra-
tion is achieved in dry, sandy or rocky areas. Poor results are ob-
tained in moist,  clayey or conductive soils. Radar penetration to
30 ft is common.

Electromagnetic Measurements
  The electromagnetic method provides a means of measuring the
electrical conductivity of subsurface soil, rock and groundwater.
Electrical conductivity is a function of the type of minerals in the
formation, its porosity and permeability and the fluids which fill
the pore space. Often the  conductivity of the pore  fluids
dominates the measurements.
  Natural variations in subsurface conductivity may be caused by
changes in soil moisture content, pore fluid conductance, depth
of soil and thickness of formation layers. The absolute values are
not necessarily diagnostic; it is the variations that are significant.

Seismic Refraction and Reflection
  Seismic refraction and reflection techniques are used to deter-
mine  the thickness and depth of geologic layers by using the
velocity of seismic waves which is an indication of the soil and
rock's engineering properties. Reflection can be used effectively
to depths ranging from about 50 ft to  several thousand feet.
Structural and formation boundaries can be detected.  Refraction
commonly is applied to shallow investigations up to a few hun-
dred feet. Depth to bedrock and the layer thickness of the overly-
ing unconsolidated material can be determined.
  Seismic waves transmitted  into the subsurface travel at dif-
ferent velocities in various types of soil and rock. The waves travel
in all directions  from the source and are  reflected and refracted
when they impinge on an interface. The combination of pattern
and velocity will affect travel time. An array of geophones on the
surface measures the travel times of the seismic waves from the
source to the geophones. The refraction and reflection techniques
use the travel times  of the waves to provide data on  subsurface
layers, including the number of layers, the thickness of the layers
and their depths.

Resistivity Method
  The resistivity method is used to measure the electrical resistiv-
ity of the section which includes the soil, rock and groundwater.
The method may be used to determine lateral changes and vertical
cross-sections.
  The procedure requires that an electrical current be injected into
the ground by a  pair of surface electrodes. The resulting potential
field  is  measured at the surface  between a second pair of elec-
trodes.  The subsurface resistivity can be  calculated by knowing
the separation and geometry of the electrode positions, applied
current and measured voltage.
  In  general, most soil and rock minerals are  highly resistive,
hence the flow  of current is conducted  primarily through the
moisture-filled pore  spaces within the soil and rock.  Therefore,
the resistivity of soils and rocks  is predominantly controlled by
the porosity and the permeability of the system. Electrical sound-
ing is used to reveal the  variations of apparent resistivity with
depth. Horizontal profiling is used to determine lateral variations
in resistivity.

In Situ Measurement Techniques
  The determination of soil properties by in situ measurement
techniques also has advanced greatly. A few of these methods are
described below.

Standard Penetration Test
  The standard  penetration test is the oldest and most widely ac-
cepted method for obtaining shear strength properties. This in-
volves counting the number of blows required of a 140-lb hammer
dropped from a height of 30 in. to move a split spoon sampler 12
in.  through the soil.  The resultant number is termed the  "N"
count and is at present the most widely used index of subsurface
soil conditions. Advantages of the standard penetration test in-
clude economy of use, simplicity of procedure and widespread ac-
ceptance.

Cone Penetration
  Cone penetration devices are useful because of simplicity of
testing, reproducibility of results and ease of correlating results to
other test results. The drawback of the test is that no soil sample is
obtained.
  The test procedure involves advancing vertically, at a constant
rate, a penetrating cone of about 10 cm2 base area and about 60°
apex angle. The force needed to penetrate  a given distance in a
soil is used in measuring shear strength.

Vane Shear
  In contrast  to methods which determine strength parameters
from empirical correlation,  the  vane  shear  test attempts  to
measure undrained shear strength directly.  The test procedure is
to advance  a vane configuration to a desired  soil  depth and
measure the applied torque as the vane is rotated at a constant
rate.
  Shearing resistance  is considered to be mobilized on a cylin-
drical failure surface of rotation, corresponding to the top, bot-
tom and sides of the vane assembly. The vane shear test possesses
the advantages of economy of use and reduced soil disturbance.

Pressuremeter
  The pressuremeter  is used by lowering the apparatus into a
borehole or, if the machine is self-boring,  by drilling to the re-
quired depth. Increments of hydraulic pressure are applied to the
test cell which acts against the borehole wall. The resulting defor-
mation  of the borehole  wall is determined by  monitoring the
volume  change of the  test cell.
  The relationship between pressure increment, volumetric ex-
pansion  and  time   is  examined  to determine   strength
characteristics. No soil  sample is obtained if  the  self-boring
pressuremeter is used.

Iowa Borehole Shear Test
  The Iowa borehole  shear test is performed by lowering a shear
head consisting of two opposing horizontally grooved shear plates
to the required test position in an uncased borehole. The two
shear plates are expanded until seated in the borehole walls at a
selected pressure. After a time  period allotted to allow for con-
solidation to occur, the  shear  head  is either pulled  upward  or
pushed  downward at  a steady rate of 2 mm/min. The required
forces for shearing are measured, and the shearing stress is plotted
against the normal pressure. By performing a number of tests at
different seating pressures,  the soil strength can be calculated.
  Soil strength may be expressed in a number of ways, and the
particular strength parameters to use in design will depend on the
nature of a specific problem.

SHEAR STRENGTH OF COHESIVE SOILS
  The shear strength behavior of cohesive soils differs consider-
ably from  that of cohesionless soils. This section presents the
details of the shearing resistance of cohesive soils under various
loading and drainage  conditions.  In  order  to obtain meaningful
results from  a laboratory  test program for the shear  strength
parameters of cohesive soils, it is necessary to test good quality
samples representative of the in situ soils or compacted soils pro-
posed for the project. For in  situ soils,  the samples to be used
                                                                                  BARRIERS & WASTE SOLIDIFICATION
                                                                                                                           191

-------
should be undisturbed as far as possible and must be tested at the
in situ water content and dry density. For compacted soils, it is
necessary to match the following three factors  between  the
laboratory samples and the soils used in the field:

•  Method of compaction
•  Dry unit weight, 74
•  Water content, u
   It is  necessary to relate  shear  strength to compaction at an
elementary level in this section. A typical compaction curve show-
ing the dry unit weight-water content relation for a cohesive soil is
given in Figure  l(a). The variation of shear  strength with water
content is shown in Figure  l(b). Although a  cohesive soil can be
compacted at the same dry unit weight at points A and B in Figure
l(a), the water content at these points would differ considerably.
Thus,  the shear strength of the compacted  soil at  A would be
higher  than that  at  B.  Therefore, it  is necessary  to  prepare
samples in the laboratory at values of water content  and dry unit
weight comparable to those of the field compacted soils. Further-
more, the structure of a compacted clay is highly dependent upon
the method  of compaction, and the structure has a significant in-
fluence on the shear strength and drainage properties of the clay.
Therefore, it is also important to use  a laboratory compaction
procedure that would yield compacted test samples with a struc-
ture similar to that of the field  compacted soils.
                                               Compaction
                                               Curve
                    Water  Content , oo


                          Figure 1
                      Compaction Curve
Shearing Resistance, Cohesion
  The shearing resistance of cohesive soils differs  from that of
cohesionless soils due to the following differences in soil factors:
 • Particle size
 • Particle shape
 • Permeability
 • Angle of internal friction
 • Plasticity

 The sizes of particles  forming cohesive soils are significantly
 smaller than those of cohesionless soils. Cohesionless  soils are
 made of angular to rounded shapes of particles, whereas cohesive
 soils are made of platey shaped particles. Cohesive  soils drain
 much more slowly than cohensionless soils due to the fact that the
 coefficient of permeability of cohesive soils is lower than that of
 cohesionless  soils. The angle of internal friction of cohesive soils
 generally is slightly lower than that of cohesionless soils.  Further-
 more,  the small  particle sizes of  cohesive soils have attractive
 forces  giving the plastic behavior  for cohesive soils;  this factor
 does not apply to cohesionless soils. Due to the above factors, the
 shear strength behavior of cohesive soils is expressed by the equa-
 tion:
     t =  c +  crtan 
(1)
where c is cohesion and  is the angle of internal friction.
  The Mohr-Coulomb failure envelopes for clays and sands are
shown in Figure 2 for comparison. The value c indicated on the
shear stress, also referred to as the "Cohesion Intercept," is due
to the attractive forces between the platey shaped particles form-
ing cohesive soils such as clays.
                                                                         Mohr-Coulomb
                                                                         Envelope  for  Sands-
                                                                                                     Mo hr -Coulomb
                                                                                                     Envelope  for Sands
                                                                              Normal Stress, o
                                                                                           Figure 2
                                                                                      Strength Envelopes
Pore Pressure Parameters
  The low permeability  of  clay causes pore water pressure to
build  up during laboratory  or field tests and  control the shear
strength  properties measured  for  use  in design computations.
Therefore, it becomes necessary to understand the mechanisms of
pore water pressure changes in cohesive soils before making an at-
tempt to understand the shear strength behavior of cohesive soils.
  When  a sample of clay is loaded  in a triaxial test initially by an
equal  all around confining pressure, 0$, followed by a deviator
stress, a, -<73, a total pore pressure change, Au, occurs as shown
in Figure 3, under undrained conditions. The above two types of
192    BARRIERS & WASTE SOLIDIFICATION

-------
                  cr 3 + (a j -a 3 )
         Pore
           Au
                                      Total
                                    Pressure
                                               Change
                 Au
                A u3
                          O
                         •=0=
                    Pore  Pressure
                      AU3=B03
                Change
                       (a,
a3 )
                Au
Pore Pressure CMange

        = A(  o"!-  0-3 )

     Figure 3
   Pore Pressures
loading independently cause two increments of pore pressures,
Au3 and Aui3, given by:
      Au = Au3 +
where:
and
        J3 =
                                   (2)
                                                        (3)
                                   (4)
      Au13 = A(ff] - CT3)
as shown in Figure 3. In Equations 2 and 4, B and A are referred
to as the "Pore Pressure Parameters." By knowing these pore
pressure parameters, the excess pore water pressures developed in
a triaxial test can be estimated.  Conversely, by  measuring  the
changes in pore pressures, Au3 and Aui3, in a laboratory^ test for
known values of a3 and (al  - a3), the values of B and A can be
computed.
  The pore pressure parameter, B, is a function  of the coefficient
of volume compressibility, mv, of the soil sample shown in Figure
4(a) and the degree of saturation, S, indicated in Figure 4(b). The
coefficient of volume compressibility increases as the stiffness of
the soil skeleton decreases and,  as a result,  the value of B in-
creases for the same value of saturation as shown in Figure 4(b).
When the sample is dry due to air being far more compressible
                                           than the soil skeleton, the pore water pressures developed are zero
                                           and, therefore, the value of B is regarded as zero. Since the pore
                                           water will  carry the entire confining pressure  under undrained
                                           loading conditions  for a saturated  sample of clay,  the pore
                                           pressure  parameter,   B,  becomes  1.0.  The pore  pressure
                                           parameter, A,  is a  function  of stress ratio,  oj/cr;, as  shown in
                                           Figure 4(c). For normally consolidated clay,  A is approximately
                                           1.0; for overconsolidated clays, it is between  -0.3 and 0.8.
                                                                                                        Au
                                                                               1.0
                                                                              0.5
                                                      a)
                                                      10
                                                      ^
                                                      to
                                                      a,
                                                                                               50          100
                                                                                         Saturation,  S(%)
                                                                              1.5
                                                                          \<
                                                        -0.3
                                 024
                                     Stress Ratio,  ai  /o 3

                                        Figure 4
                                 Pore Pressure Parameters

              Unconsolidated-Undrained Test
                The Unconsolidated-Undrained, U-U, Quick or Q Test usually
              is performed to obtain strength parameters, c and , correspond-
              ing to a fast loading of the soils at sites where drainage is poor. In
              this test,  the soil sample is trimmed and placed in a thin mem-
              brane as shown in Figure 5(a). The confining pressure, a3, is ap-
              plied to simulate the confinement available to the sample in the
              field while the drainage value is closed. The sample is sheared
              while the drainage  valve is closed, not allowing any excess pore
              pressures built up in the sample  to dissipate, as shown in Figure
              5(b).
                Since  it is impossible to  saturate the sample without con-
              solidating, pore pressures are never measured in the U-U test. The
              Mohr-Coulomb envelope for a saturated sample of clay in a U-U
              test is horizontal, as shown in Figure 6(a), since the deviator stress
              at failure is independent of the confining pressure applied to the
                                                                                 BARRIERS & WASTE SOLIDIFICATION    193

-------
                                                                                                  O 1
                             Sample of  Soil
                                  o 3
                                   Valve  Closed
                       o i
                                  0 3
                                  Valve  Closed
                                  (Pore  Pressures
                                   are never
                                   measured)
                       Figure 5
               Unconsolidated-Undrained Test
sample. For a saturated sample of clay yields the U-U test gives
the following strength parameters:
      c = Su

       = O
(5)
(6)
where Su is the undrained shear strength of the clay.
  Since the deviator stress  is independent  of  the confining
pressure, it is possible to perform a test with a confining pressure
of zero and obtain the undrained shear strength,  Su> of the soil
from the diameter of the failure Mohr's circle as shown in Figure
6(b). This test is referred to as the "Unconfined Compression
Test," where the  sample is sheared by a compressive load, a^
without the use of a confining cell. The diameter of the Mohr's
circle qu, in Figure 6(b)  is called the "Unconfined Compressive
Strength" of the sample.
  When a partially saturated sample of clay is sheared in a U-U
test, the size of the Mohr's circle increases with increasing confin-
ing pressure as shown in Figure 7. Once the confining pressure
reaches a high enough value  to cause the pore air to dissolve in
pore water, the sample becomes saturated. Beyond this value, the
sample yields a horizontal Mohr-Coulomb envelope as shown in
this figure. Thus,  in order to obtain the strength parameters, c
and , corresponding to a confining pressure, a3 (of interest in the
partially saturated range), it becomes necessary to draw a tangent
at <73  = a'j on the envelope and read values for c* and ' as shown
in Figure 7.
  The strength parameters obtained from a U-U test are  useful
for analyzing geotechnical engineering problems where the soils
did not consolidate appreciably and the drainage conditions were
poor. An  example of such an application is  the slope stability
evaluation for the dike for the end-of-construction loading condi-
tion.
                                                               in
                                                               to
                                                               0)
                                                               1-1
                                                               4-1
                                           0
           Confining
           Cell
                                                                          Unconsolidated-Undrained  Envelope
                                                                                    c = S  .  « =  0
                                                                        r   ZOLA
                                                                             Normal Stress, a
                                Mohr  Circle  for the
                                 Unconfined  Compression
                                 Test  with  O3  =  0
                                                                                     Normal Stress,o
                               Figure 6
                           Strength Envelopes
I/)

ID

in
              Soil  is
              Pa_rt ially
              Saturated
Soil  is Saturated
                                              *  for  03   03 *
        O 3   03
                             A     A
                                 Normal Stress,  o
                  03-03
                               Figure 7
                           Strength Envelopes
194   BARRIERS & WASTE SOLIDIFICATION

-------
Consolidated-Undrained Test
  The Consolidated-Undrained,  C-U or R Test usually is per-
formed to obtain strength parameters, c and , corresponding to a
fast loading of the soils at a site  where drainage is poor and the
site has undergone consolidation.  In this test, the drainage valve is
left open during the application of the initial confining pressure,
'under drained loading conditions. The rate
of loading in this test is slow enough to dissipate the excess pore
 w
 10
 0)
 1-1
 4-1
 CO
Total  Stress
Envelope
                                                                                                       Effective  Stress
                                                                                                       Envelope
                                                                               Effective Normal Stress,a


                                                                                           Figure 9
                                                                                      Strength Envelopes
water pressures as they build up during the application of the
shear load. The strength parameters obtained from this test pro-
vide the data necessary for long-term  stability calculations of
slopes and embankment dams.

EVALUATION OF STRUCTURAL
INTEGRITY OF DIKES
  It is necessary that the dike be stable for the following loading
conditions:
• End-of-construction or short-term
• Steady seepage or long-term
• Rapid drawdown
• Seismic
  The factor of safety against any form of slope failure usually is
determined using the following:
• Slope geometry
• Shear strength of soil
• Method of analysis

  There are more  than  30  methods of slope stability  analysis.
Among  these,  Bishop's modified method, Ordinary Method of
Slices and Spencer's method are most commonly used.  Bishop's
Method and the Ordinary Method of Slices are used with circular
failure surfaces, and the Spencer's Method is used for noncircular
slope surfaces. These methods can be applied for dike  stability
evaluation using:
                                                                                BARRIERS & WASTE SOLIDIFICATION    195

-------
• Hand computations
• Computer programs
• Stability charts

Hand Computations
  The hand computations are almost never used in practice at the
present time for slope stability calculations. These are tedious and
time-consuming.

Computer Programs
  Computer programs for IBM personal computers and  main
frame computers are available from the authors for performing
dike stability evaluation. These computer programs have various
desirable options to handle even the most difficult site conditions.

 Stability Charts
   The stability of slopes  can be analysed efficiently using the
 stability charts shown in the manual by Duncan and Buchignani.'
 Although the charts assume simple slopes and uniform soil condi-
 tions, they can  be used to obtain reasonably accurate answers for
 most complex problems if irregular slopes are approximated by
 simple  slopes and average values of  unit weight, cohesion and
 friction  angle are used.
 MINIMUM FACTOR OF SAFETY
   When detailed analyses of slope stability are performed using
 the  computer programs based on  either Ordinary Method  of
 Slices,  Bishop's  Modified  Method  or Spencer's  Method, a
 number of slip surfaces must  be examined  to  locate the most
 critical failure surface with the lowest factor of safety. The slope
 stability charts yield minimum factors of safety when properly ap-
 plied. Under most conditions,  the  uncertainties due  to approx-
 imations and assumptions in the  method of analysis  are smaller
 than the uncertainties due to inaccuracies in measuring the shear
 strength. Approximations in the analysis usually amount to 15%
 or less, but the margin for error in evaluating shear strength may
 be considerably greater. The minimum allowable value of factor
 of safety for a slope in a dike depends on  several factors; some
 guidance is given in Table 1 for static and seismic conditions.
                            Table 1
         Recommended Minimum Values of Factor of Safety

                            Uncertainly of Strength Measurements
Costs and Consequences
of Slope Failure                     Small1
Cost of repair comparable to cost
of construction. No danger to human   1.25
life or other property if slope fails.     (1.2)*

Cost of repair much greater than
cost of construction, or danger to      1.5
human life or other valuable          (1.3)
property if slope fails.
Large'

1.5
(1.3)
2.0
or greater
(1.7 or
greater)
 1 The uncertainly of ihc Mrcnglh measurement* is smallest when ihc soil conditions arc unilorm
  and high quality .strength lest data provide a conslslcnt, complete and logical picture of the
  strength characteristics.
 ' The uncertainly of (he strength measurements is greatest when the soil conditions are complex
  and when the available strength data do not provide a consistent, complete, or logical piclure
  of the strength characteristics
 * Numbers without parentheses apply for static conditions and Ihosc with parentheses apply lor
  seismic conditions
 (After Duncan and Buchignani')
                  EVALUATION OF BEARING
                  CAPACITY OF DIKE FOUNDATIONS
                    If the foundation of a  proposed dike  is composed of low
                  strength  soils, the factor of safety against any  type of bearing
                  capacity  failure should be evaluated. The bearing capacity consid-
                  eration will control the base width of the dike, especially when the
                  foundation contains a scam of weak material, as shown in Figure
                  10.  In this figure:

                      q'w=7H.                                            0)
                    Using stress influence charts, the influence factor, I, is obtained
                  corresponding to the dimensions, a and b.
                    Using  I, the stress, qw, applied to the seam of weak soils can be
                  calculated as:
                            Iq «
                                                           (8)
                    The ultimate bearing capacity of the weak material is evaluated
                  using the equation:
    qu  _ cNc
                                      BN- t-  Z"»-NC:
(9)
                  where Ne, N,f and Ng are bearing capacity factors obtained from
                  bearing capacity charts corresponding to the angle of friction, ,
                  of the soil.  In this equation, B and y refer to the width of the
                  stress distribution as shown in Figure 10 and the unit weight of the
                  soil, respectively.  The  factor of safety  against bearing capacity
                  failure is calculated using:
                       F
                       ' * ~ q»
                  and Fs should be higher than 2.5 to 3.
                                                                            (10)
                              1   1    1   i    1   1    i   1    1


WEAK LAYER
, B ,


                           Figure 10
                    Bearing Capacity Failure
CONCLUSIONS
  The RCRA regulations require that the surface impoundments
must  have dikes that are designed, constructed and maintained
with sufficient structural integrity to prevent massive failures of
the dikes. Thus, the engineers involved in certifying dikes need to
use  appropriate  engineering  procedures  to  determine  site
characteristics and  dike integrity. Some guidance is given in this
paper to accomplish these tasks undertaken by engineers involved
in the preparation and review of permit applications for surface
impoundments, waste piles and other land disposal units.
 196    BARRIERS & WA§TE SOLIDIFICATION

-------
REFERENCES
 1.  Bishop, A.W., "The Use of Slip Circle in the Stability Analysis of
    Slopes," Geotechnique, 15, 1955, 7-17.
 2.  Bishop, A.W. and Bjerrum, L., "The Relevance of the Triaxial Test
    to the Solution of Stability Problems," Proc. of the ASCE Research
    Conference on the Shear Strength of Cohesive Soils, Boulder, CO,
    1960.
 3.  Duncan, J.M. and Buchignani, A.L., "An Engineering Manual for
    Slope Stability Studies," U.S. Berkeley Geotechnical  Engineering
    Report, 1975.
 4.  Holtz, R.D. and Kovacs, W.D.,  An Introduction to Geotechnical
    Engineering, Prentice-Hall, NJ, 1981.
 5.  Hvorslev, M.J., "Subsurface Exploration and Sampling of Soils for
    Civil Engineering Purposes,"  Corps of Engineers, Waterways Ex-
    periment Station, Vicksburg, MS,  1949.
 6.  Janbu,  N., "Slope Stability Computations," Soil Mechanics  and
    Foundation Engineering Report," the Technical University of Nor-
    way, Trondheim, Norway, 1968.
 7.  Jeyapalan, J.K., Theory and Problems of Geotechnical Engineering,
    McGraw-Hill, New York, NY, 1985.
 8.  Lowe, III,  J. and Karafiath,  L.,  "Stability of  Earth Dams Upon
    Drawdown,"  Proc. of the First PanAmerican Conference on  Soil
    Mechanics and Foundation Engineering, Mexico City, Mexico, 2,
    1960, 537-552.
 9.  Mitchell,  J.K., Guzikowski, F. and Villet, W.C.B., "The Measure-
    ment of Soil  Properties In situ,  Present Methods—Their Applica-
    bility  and  Potential," National  Technical  Information  Service,
    1978.
10.  Morgenstern, N.R., "Stability Charts for Earth Slopes During Rapid
    Draw-down," Geotechnique, 13, 1983, 121-131.
11.  Navfac DM-7,  "Soil Mechanics Design Manual," U.S. Navy  Of-
    fice, 1981.
12.  Seed, H.B., "Considerations in Seismic Design of Earth and Rock-
    fill Drums,"  19th Rankine Lecture, London, 1979.
APPENDIX I — NOTATION
  The following symbols used in this paper:
U-U
C-U
D
c
0
S,,
H
7w
N
Qc
Unconsolidated-Undrained
Consolidated-Undrained
Drained
Cohesion
Angle of Internal Friction
Undrained Shear Strength
Effective Consolidation Pressure
Unit Weight
Dike Slope Angle
Dike Height
Unit Weight of Shear
Blow Count
Cone Resistance
Factor of Safety
                                                                                        BARRIERS & WASTE SOLIDIFICATION     197

-------
                          Use  of  X-Ray Radiographic  Methods
                                   in the  Study of Clay Liners

                                            Philip G.  Malone, Ph.D.
                                                  James H. May
                                            Geotechnical Laboratory
                            U.S. Army Engineers Waterways Experiment Station
                                              Vicksburg, Mississippi
                                             Kirk W.  Brown, Ph.D.
                                                James C. Thomas
                                      Soil and Crop Sciences Department
                                             Texas  A&M University
                                              College Station, Texas
ABSTRACT
  X-ray radiography has been widely used in soil investigations
to study the distribution of layers in soil cores and the effects of
changing conditions (loading or impact) on soil structure. X-ray
radiographic techniques also can be useful in  studying clays or
clay soils used in liners.
  Laboratory investigations were undertaken to demonstrate that
X-ray radiographic techniques could be used to detect density and
soil  structure changes that  usually accompany variations in
hydraulic conductivity of clay liners. An example of a real-time
test of a simulated bentonite and sand liner attacked with  acid
lead nitrate and examples of radiographic examination of clay soil
(non-calcareous smectite) samples that have been permeated by
lead acetate or lead nitrate are presented. The changes in density
and structure can be related to changes observed in hydraulic
conductivity during permeation.
  X-ray radiography easily can  be applied  to field samples of
soil or clay liner materials to detect density and structural changes
that occur as the liner and permeating fluid interact. X-ray tech-
niques have applications in both understanding failure mechan-
isms and forecasting liner performance.

INTRODUCTION
  X-ray radiographs of soil cores  have been employed in  geo-
technical investigations  for many years because radiographs  pro-
vide a  unique technique for detecting  changes in the physical
arrangement of soil particles without disturbing a soil core. In
many soil cores, X-ray radiographs will detect  subtle differences
in soil density that might have escaped notice  during the visual
examination  of core sections. The X-ray image displays varia-
tions in the absorption of X-ray photons in different parts of the
object examined.  Differences in absorption can be  related to
changes in sample thickness, density or chemical composition.1
When the sample geometry is controlled by preparing cylinders
or slices of cylinders for examinations, patterns observed in the
radiographs can be reliably related to the differences in sample
density and chemistry.
  This study examined the use of X-ray radiographic techniques
in identifying changes in the density and composition of samples
of simulated  clay  or other soil liners that  have been exposed to
permeants that could be encountered in  groundwater beneath
an industrial or hazardous  waste disposal site. Two approaches
were used: (1) examining the real-time invasion of a bentonite-
sand liner  using a lead nitrate-nitric acid solution; and (2) ex-
amining the  condition  of soil samples after the samples have
been permeated with simulated industrial waste solutions.
         PVCW€
         BENTOMTC
           AND  —
           SAND
      PVC COUPUNG -
s 	 83mm —
|> 	 75 mm 	 ^
/
	 . 	 _ . '
	 	 . "
/ 	 	 '
/ 	 "

"~ , _ , __ . _ . _ . /

/
•k ~"-"~ /
% -_ \f
W " ": " \
s - ^!_ ~ - __ - . /







IOC
1
I
T
1<
k.
*71 *.





40C

mm

X) mm

t



> rr




                         Figure I
    Cross-Section of the Test Cell Used in the Real-Time Liner Test
REAL-TIME STUDY OF LINER PERMEATION
  A  simulated  waste pond liner was prepared  by  mixing one
part of a commercially available bentonite (Aquagel from Baroid
Petroleum Services  of  N.L. Industries, Houston, Texas) with
four  parts by volume of coarse filter sand. The mixture follows
typical procedures for producing a bentonite liner.J The sand and
 198    BARRIERS & WASTE SOLIDIFICATION

-------
bentonite were mixed in a dry condition and water was gradually
added with continued mixing to assure that the bentonite was
completely hydrated. The wet mixture was permitted to settle for
24 hours, and excess water was decanted. A test cell was prepared
from a 400-mm length of 75-mm ID schedule 40 PVC pipe (Fig.
1).
  The bottom of the test cell was fitted with a plastic plug and a
hole was drilled through the plug so that a drainage line could be
attached. The bottom of the test cell  was packed with  Dacron
filter floss, and a 100-mm thick layer of coarse filter sand was
poured into the test cell. The bentonite and sand mixture was
poured in on top of the filter sand, and the test cell was gently
shaken to assure that the bentonite and sand layer settled and no
air bubbles formed in the simulated liner. Clean water was placed
on top of the bentonite and sand layer  to verify that the simu-
lated liner would not leak. To initiate the testing, the clean water
was decanted, and a simulated waste formed by saturating a 10%
nitric acid solution with reagent-grade crystalline lead nitrate was
added above the bentonite and  sand layer. The saturated acidic
lead nitrate solution was approximately 140 mm deep over  the
simulated liner.  This depth of  waste  is less than the 300-mm
depth permitted under disposal guidelines.3
  INVADED AREA-
                                                                  SUPPORT RING
                                                                     1 cm
                                                                                             Figure 3
                                                                    Print of Radiograph of Simulated Liner One Day after Application of
                                                                                     Acidic Lead Nitrate Solution
                            Figure 2
   Print of Radiograph of Simulated Liner Before Application of Acidic
                       Lead Nitrate Solution
  X-ray radiographs were  made of the test cell at one to  five
day intervals using a Norelco MG 3000 X-ray system equipped
with a tungsten target  tube. Kodak Industrial M film was used
with a 1 to 2 min exposure. The X-ray unit was operated at 290
kV with 11 mA current.
  Figure 2 shows the liner  prior to testing. Breakthrough of the
simulated waste was noted after 16 days. Figures 3 and 4 show the
progressive failure of the liner after a one-day and after a 25-day
exposure to the simulated waste. When the test cell was opened,
the presence of voids in the simulated liner easily could be ob-
served (Fig. 5). The X-ray radiographs had detected shrinkage
cracks forming in the  simulated liner and had accurately indi-
cated the mode of failure in the simulated liner. Lead nitrate, a
radio-opaque solution, was used initially in the tests to assure that
any invasion of the liner could be observed. The formation of the
voids in the clay reduced the usefulness of the lead solution be-
cause the lead salts lowered the contrast between the fluid-filled
voids and  the surrounding soil. Where flocculation and  void
formation are the expected  modes of failure, there is no benefit in
using radio-opaque permeants.
                                                                                    BARRIERS & WASTE SOLIDIFICATION    199

-------
 INVADED AREA
   1 cm
                           Figure 4
    Print of Radiograph of Simulated Liner 25 Days after Application
                     of Acidic Lead Nitrate
X-RAY RADIOGRAPHIC STUDY OF
PERMEAMETER SAMPLES
  The evaluation of the interaction of soil liners and simulated
or actual industrial wastes generally has been performed by forc-
ing samples of selected permeants (wastes) through packed soil
columns in  a  fixed wall or flexible wall permeameter.' Inter-
actions  are evaluated by determining if the permeability of the
packed  soil is changed by the permeant. The type of interaction
that has occurred between the soil and the permeant cannot be
determined  without examining  the  soil.  For  example,  the
hydraulic conductivity of the liner may decrease, but the reason
for the decrease (for example, swelling of the soil or the forma-
tion of precipitates) cannot be determined without measuring the
change in sample volume (for swelling) or looking for the forma-
tion of  new compounds in the clay (for precipitation). X-ray
radiographs can detect both volume changes and increased liner
density  due to the formation of precipitates. More than 60 soil
samples  from permeameter tests were examined to evaluate the
ability of radiographic data to assist in the  interpretation of
permeability changes.' The major effects observed have included
the breakdown of the clay structure (causing an increase in sample
permeability) and the formation of precipitates (with a decrease
in permeability).
                                                                                          Figure 5
                                                                Photo of Bentonile and Sand Mixture after Permeation with Acidic Lad
                                                                                       Nitrate Solution
  Samples  of permeated soils were prepared  using techniques
proposed for  testing potential  landfill liner materials.' In the
examples discussed here, six samples of the Lufkin soil (a non-
calcareous  smectite) from Brazos County, Texas, were packed
into standard-design fixed wall permeameters (Fig.  6).  Each
sample was compacted in three equal lifts using 25 blows per lift
from a 2.4l-kg hammer falling 300 mm. One set of three samples
was permeated with a 60% (by weight) lead acetate solution; the
other set was permeated with a 50% (by weight) lead nitrate solu-
tion in 0.1% HNOj.
  After a sufficient volume of permeant had passed through the
sample to establish a trend in changes  in hydraulic conductivity,
the permeameters were  opened and checked for swelling. The
samples permeated with lead acetate had increased in  height by
                        PRESSURE INPUT
                                              PERMEAMETER
                                                 BASE
           OUTLET
                                    POROUS STONE
                             TEFLON TUBING
                           Figure 6
     Cross-Section of Permeameter Used in the Study of Soil-Waste
                         Interaction
200    BARRIERS & WASTE SOLIDIFICATION

-------
 1  cm
                          Figure 7
   Print of Radiograph of Non-Calcareous Smectite Soil Permeated
                  with Lead Acetate Solution

maximum of 2.7%. Maximum swelling in the samples permeated
with lead nitrate was 1.6%. The samples were extruded and wrap-
ped  in foil. X-ray radiographs were prepared  using a  Norelco
MG 300 Industrial X-ray System. The unit was operated at 290 kV
at 10-12 mA. Exposure times varied from 1 to 3 min using Kodak
Industrial M film.
   Permeameter testing established that the lead acetate caused a
decrease in hydraulic conductivity by a minimum of 50%. X-ray
radiographs showed lead compounds were precipitating in the soil
(Fig. 7).
   The lead nitrate solution caused an increase in hydraulic con-
ductivity by a minimum of over  100%. The X-ray radiographs
showed irregular connecting voids  suggesting the clay had devel-
oped a flocculated or blocky structure similar to that  observed
in the real-time failure study (Fig. 8).

CONCLUSIONS
   X-ray radiographic techniques, previously applied to the study
of soils, can be employed usefully in the investigation of the inter-
action of wastes and clay (or clay soil) liners. Using suitable test
cells, real-time effects of waste (or simulated wastes) on sections
of liner can be observed. Soil samples that have been permeated
by simulated wastes can be X-rayed to examine the soils  for den-
sity or structure changes that indicate occurring liner-waste inter-
action.
   Field techniques for sampling soils to obtain undisturbed sam-
ples  are available, and it should be possible to use X-ray radio-
graphic methods along with standard soil description techniques
to examine the condition of liners at landfills or waste storage
areas. X-ray examination of soil cores is completely non-destruc-
tive, and altered sections observed in the X-ray radiographs can
be documented by sampling and examining anomalous soil inter-
vals.
  1 cm

                          Figure 8
   Print of Radiograph of Non-Calcareous Smectite Soil Permeated
                  with Lead Nitrate Solution
ACKNOWLEDGEMENTS
  The tests described and the resulting data presented herein, un-
less otherwise noted, were  obtained  from research conducted
under the Independent Laboratory Initiated Research Program,
Project No. 4A161102A91D Task Area 02 Work Unit 154, of the
United States Army Corps  of Engineers by the USAE Water-
ways Experiment Station. Permission was granted by the Chief
of Engineers to publish this information.

REFERENCES
1. Krinitzsky, E.L., Radiography in the Earth Sciences and Soil Mechan-
  ics, Plenum Press, New York, NY, 1970.
2. Matrecon, Inc., "Lining of Waste Impoundment and Disposal Facil-
  ities," U.S. EPA, SW-870, Washington, DC, 1983.
3. U.S. EPA,  "Hazardous  Waste Management System; Permitting
  Requirements for Land Disposal Facilities," Federal Register, July
  26, 1982.
4. Brown, K.W. and Anderson, D.C., "Effects of Organic Solvents on
  the Permeability of Clay Soils," EPA-600/2-83-016, U.S. EPA, Cin-
  cinnati, OH, 1983.
5. Malone, P.O., May,  J.H., Brown,  K.W. and Thomas,  J.C.,  "Use
  of X-ray Radiographic Techniques in the Evaluation of Soil Liners,"
  Miscellaneous Paper  GL-85-14, USAE Waterways Experiment Sta-
  tion, Vicksburg, MS, 1985.
                                                                                 BARRIERS & WASTE SOLIDIFICATION    201

-------
                 Closure Design and Construction  of Hazardous
                           Wastes Landfills  Using  Clay  Sealants
                                              John F.  O'Brien, P.E.
                                                  Lonnie E. Reese
                                                O.H. Materials Co.
                                                 Rosewell, Georgia
                                                 Ian Kinnear, P.E.
                                                  Dames & Moore
                                                Boca Raton,  Florida
ABSTRACT
  Clay sealants are employed at an increasing number of sites
securing land disposal of hazardous industrial wastes. Such seal-
ants are employed as topliners to limit infiltration and/or as bot-
tom liners to limit exfiltration and provide for leachate collection.
These sealants are employed both as primary liners and as added
security to synthetic liners. Liner design for hazardous waste land-
fills is well-regulated and quantitative analytical tools are  well-
documented. However,  the practical  implementation of  cost-
effective, buildable and  reliable clay liners requires  design in-
sight beyond the obvious. Key elements in the selection, design
and construction of landfill liners using clay sealants are discussed
herein. The case histories of two landfill closures are presented to
amplify the impact of the designers' assumptions on  implemen-
tation.

INTRODUCTION
  The national concern for environmentally secure disposition of
hazardous wastes  has led to increased sophistication in disposal
technology. In land disposal,  landfill design has been expedited
not only by expansive regulatory requirements, but also by ana-
lytical modeling  which  allows cost- and  time-effective  evalua-
tion of design alternatives.
  Ideally, landfills securing hazardous materials are designed
prior to disposal of the materials. Such planning includes design
to limit the development of leachate and preclude exfiltration of
leachate to the groundwater. However, as a result of years of un-
regulated and unenforced disposal, thousands of small to  large
"out back" sites at industrial  facilities exist across the nation. It
is the closure of these existing facilities which is the subject of this
paper.  These existing facilities  include  several  distinguishing
characteristics which impact closure design  and construction:

• The waste material often is poorly defined
• Stabilized or unstabilized, wastes are left in-place and closure
  includes emplacement of a top liner only
• Construction  of the design groundform often includes sub-
  stantial regrading of unknown materials, placing a premium on
  planning equipment, selection and safety.

  Two case histories will amplify this discussion.

• "Site A, Southern Florida": closure of a hazardous waste land-
  fill using an imported bentonitic clay as a topliner
• "Site B, South  Carolina": closure of a solvent contaminated
  landfill using locally available clays

Site A, South Florida
  The landfill at Site A  is located within an industrial facility in
South Florida. It consists of one 4-acre cell that was constructed
to a height of about 35 ft above existing grade. The landfill was in
operation from 1958 until the middle of 1984. The cell was util-
ized primarily for the disposal of paper cartons, boxes, construc-
tion debris,  cafeteria waste,  yard trimmings  and scrap metal.
However, in the early years of filling when operating  practices
and regulations were less stringent than today, unbagged asbestos
and drums containing solvents  and fuels were deposited in the
landfill.
  Below the cell, the groundwater is contaminated with VOCs;
the resulting contamination plume is slowly migrating eastward,
away  from the landfill. Associated with the landfill closure, a
groundwater recovery well system and  treatment plant will be
constructed around the perimeter of the landfill.
  The landfill is bounded closely on three sides by water. Uncon-
trolled filling over the  years left a very irregular  groundform
with slopes as steep as 1:1.

Site B, South Carolina
  Site B is located within an industrial facility in South Carolina.
Over at least 10 years of operation,  the site was used as a dump
for solvents  and drums. At the time of closure, the drums had
been removed and closure involved capping solvent contaminated
soils on the 5-acre site. The contaminated soils were encountered
largely in depressions in the land surface on this site.

MODELING LANDFILL PERFORMANCE
  Closure  design  of industrial  and hazardous waste landfills
usually involves the following generic evaluations:

• Determine the waste  composition and characteristics at the
  closure sites
• Determine  regulatory  and  design  dictated materials  and
  groundform requirements
• Determine liner materials alternatives based on  compatibility
  with the waste, availability, constructability and performance
• Complete analytical evaluations to determine optimum landfill
  design based on cost and performance

Regulatory
  The regulatory  requirements  for design of hazardous  waste
landfills are well-documented by the U.S. EPA under RCRA,
published July 26,  1982. The  Interim Final Rule on "Hazardous
Waste Management System: Permitting Requirements  for Land
Disposal Facilities" became effective Jan. 26,  1983.' These fed-
eral criteria do not dictate either liner parameter, but do include
the criteria for bottom liners for new (not existing) landfills.
  ...The liner must be constructed of materials that prevent
  wastes from passing into the  liner during the active life of
  the facility.1
202   BARRIERS & W/kSTE SOLIDIFICATION

-------
This requirement virtually requires the use of geomembranes in
lieu of any clay liner as the primary sealant of a new landfill.
  In the closure design of existing facilities, closure  requires a
cap to limit infiltration as well as groundwater monitoring. In the
usual case, closure of an  existing hazardous site allows the de-
signer some latitude in the selection of a landfill cap, either em-
ploying geomembranes or clay liners.
  The regulatory  environment of closure  design requires  con-
sideration. To this end, it is  imperative the owner and engineer
develop communication with the appropriate regulatory agency
early  in the  design process to assure compatibility of the design
with the regulatory expectations.

Analytical
  The hydrologic processes within landfills were originally mod-
eled using water balance techniques  developed by Thornwaite.3
The U.S. EPA developed the  first widely used landfill  model,
Hydrologic Simulation on Solid Waste Disposal Sites (HSSWDS),
in 1980.  Since that time,  the U.S. EPA has developed Hydro-
logic  Evaluation of Landfill Potential (HELP), the most sophis-
ticated analytical model now available for landfill systems. HELP
was  developed specifically for hazardous waste landfill evalua-
tions  as required by RCRA."
  HELP provides a quasi-two-dimensional hydrologic model of a
landfill section, applying section geometry, climatologic data and
material parameters to account for the movement of water across,
into,  through and out of landfills. Figure 1 presents a generic
landfill section indicating the solutions provided by the model.
The model allows the use of geomembranes in the section,  both
enhancing lateral drainage and limiting percolation.  Despite the
claims of some suppliers, geomembranes (synthetic liners) should
not be considered impervious. Hydrologic modeling of a landfill
performance may be completed using any of the above  tech-
niques:  Thornwaite,  HSSWDS  or  HELP.  Of these  models,
HELP certainly   represents  the  most sophisticated modeling
alternative.  Regardless of the approach used, some quantitative
evaluation of landfill performance should be undertaken.
                   PRECIPITATION
                                     EVAPOTRANSPIRATION
                                                RUNOFF
       ATIVE COVER      » » »             t t T T
               ^,J,x„,„,*,   *  n     L»*-,
               \'/'/'/'/ /'/'/'/'//\/\ V,'/'/'*'*'}•',''/'/'/'
               W'/'/'/.' if.-.'.-.   DISCHARGE
                                                      WASTE
                                                      SECTION
                 'LATERAL DRAINAGE LAYER •-.  '• .  .*;*.'*• ScJSSe'e
                                EXFILTRATION THROUGH
                               BASE OF LANDFILL SECTION
    (after Schroeder, et al)
                                           6" DRAINAGE PIPE
                                           TO DRAINAGE DITCH
         LINER ANCHOR TRENCH
                           Figure 1
             Modeling a Landfill Section Using HELP
                           Figure 2
  Typical Landfill Cap Employing Clay, Geomembrane and Geotextile
Sealant Alternatives
  Clay sealants may be used as, or as a part of, a landfill liner.
Figure 2 shows the implementation of a geomembrane and clay
liner as a cap following stabilization of a sludge pond. Whatever
the case, the designer will have two alternatives in clay selection:
• Naturally occurring and locally available clays
• Locallv available soils improved by admixtures
  Clays are chosen as sealants for  their low degree of pervious-
ness. In the design of topliners, a clay barrier with a coefficient of
permeability of 10~7 cm/sec or less normally will be required to
limit infiltration.
  Many  naturally  occurring clays  can readily achieve this low
permeability. However, the designer should not presume that any
available clay may be placed to this  permeability. Design assump-
tions should be correlated with careful laboratory testing. Repre-
sentative samples of the available  clay should be compacted in
laboratory molds and tested to determine permeability.
  In most  topliners  used  in the closure of existing hazardous
waste sites, the permeant will be rainwater. However, if this is not
the case,  a permeant representative of that which will infiltrate the
liner should be used to test the compatibility of the permeant
with the clay minerology.
  Permeability of  clays will vary with relative compaction. The
most commonly used compaction standard is the standard Proc-
tor, ASTM D-698. Laboratory molds compacted to varying  de-
grees of  relative compaction and thereafter tested to determine
permeability generally will show substantial variation. Clays com-
pacted to 95% of the maximum dry density achieved in the stan-
dard Proctor test may  show an order of magnitude decrease in
permeability over a sample of the same soil compacted to 90% of
standard Proctor. Similarly, a clay  compacted from wet material
of the optimum moisture content generally  will have a lower
permeability than the same soil at the same relative compaction,
but compacted from dry  material. Thorough testing to deter-
mine the mechanical properties of clays considered for use as lin-
ers is important not only to determine the suitability of a material,
but also  to determine the  range of performance to be expected
from field variations in installation.
  In many areas, the naturally occurring soils will not provide the
low permeability required for  liner  design. In  this  event, soil
admixes  may be used to lower permeability. Cement,  as well as
industrial wastes such as fly ash and kiln dust, may be added to
soils to lower permeability. However, the brittle behavior makes
these materials undesirable to seal structures as prone to irregular
settlement as  landfills.
  The  most common soil admix is sodium bentonite. Not com-
monly  encountered in nature, sodium bentonite is mined  largely
in the North Central United States. As a soil admixture, sodium
                                                                                 BARRIERS & WASTE SOLIDIFICATION    203

-------
       VEGETATIVE COVER
                                TOP SLOPE 4% (TTP)
                                SIDE SLOPE JM. IV
        IMPERMEABLE
        'SEAL  (4 INCHES)  •
IZ" SOIL BUfFER,
     REGRAOEO WASTE
                                    	   OAS CONVCTANCE TRENCH
                                    •f   12' WIDE • ?' OEEPI
                           Figure 3
        Typical Section, Florida Site. Including Gas Conveyance


 bentonite is attractive for its expansive characteristics. The addi-
 tion of 5-15% (by weight) sodium bentonite to soils will lead to a
 coefficient of permeability of 10 7 cm/sec or less. As with natural
 clays, the selection of sodium  bentonite as a soil  admix should
 include careful testing of representative materials.
   Figure 3 presents a section of the top liner used at Site A, South
 Florida. Clayey soils are almost totally absent in South Florida.
 Economic evaluations of geomembranes and  admix improved
 local soils led to the selection of the sodium  bentonite and local
 sand section. Laboratory testing indicated that the  sandy  local
 soils mixed with 13% by weight sodium bentonite and compacted
 to 95% of standard Proctor (ASTM D-698)  would achieve a
 permeability of 10~7 cm/sec or less.
   Figure 4 presents the topliner section used at Site  B, South Car-
 olina. At this site, the Owner's property included an area of thick
 clayey residual soil about 0.5 miles from the landfill. Laboratory
 testing demonstrated a  permeability of 10~7 cm/sec  for these
 clays compacted to 95% of standard Proctor.
    SOLVENT CONTAMINATED
RESIDUAL SOILS REGRADEO TO 6%
            SLOPE
                          Figure 4
              Typical Section, South Carolina Site B
Buffer and Cover Protection
   Protection of clay liners from environmental damages is an im-
portant design consideration. Clay liners should, at a minimum,
include a top soil buffer to protect the material from erosion. The
buffer also should be of sufficient thickness to preclude breaking
by roots and burrowing animals. A vegetated soil cover of about
1 ft is sufficient in most cases to adequately protect the clay.

Specifications and Quality Control
   Certainly the best closure planning and design is impotent with-
out carefully developed  specifications and good field quality con-
trol.
  The designer should be responsible for  the technical specifi-
cations. In the writers' experience, many landfill closure specifi-
cations include exacting criteria for material type, degree of com-
paction and permeability. Such tightly constructed specifications
open the owner to substantial costs in change orders.
   If   the  material  to be  used  has been  pre-determined  by
 laboratory testing, specifications for liner construction should be
 limited to degree of compaction and material type. Additionally,
 to assure good  field control,  a minimum compactive effort and
 soil moisture may be  specified (for example, "three passes of a
 kneading-type compactor weighing a minimum of 8 tons, com-
 pacting soils within 3%  of their optimum  moisture content").
   If  the specific  clay  material has  not been  predetermined,
 specifications should  be limited  to general material  type  and
 permeability requirements.
   Installation  of  clay  liners should  be  under  the  full-time
 surveillance  of  a representative  of the designer's office. At a
 minimum, this surveillance should be supplemented by testing for
 soil parameters  detailed  in the drawings and specifications to in-
 clude material type, thickness, degree of  compaction,  moisture
 and permeability.
   State regulated closure criteria for Site A in South Florida (Fig.
 3) dictated permeability requirements (1Q-7 cm/sec),  but  left
 materials selection to  the contractor. Quality control testing for
 the liner  installation included  thickness, moisture, relative com-
 paction and laboratory permeability of field samples.  Additional-
 ly, the contractor was required to  submit  his sodium bentonite-
 local  sand mix  design, which was  monitored in  the field. The
 average  permeability determined  from  field samples was 10~8
 cm/sec.
   At Site B, South Carolina, liner construction specifications de-
 tailed the borrow source and compaction requirements only. Pre-
 testing of  these  soils  had  well-established  the density  and
 permeability relationships of  these soils.  A full-time  represen-
 tative from the designer's office monitored compaction and liner
 thickness. Additionally, it was required that this 2-ft thick liner be
 placed and compacted in increments no greater than  6 in.

 CONSTRUCTION CONCERNS IN CLOSURE
   Closure construction of hazardous waste landfills requires con-
 siderable skill, equipment and experience. Such work should be
 undertaken only by a contractor experienced in the field.
   Construction  planning  for closure  involves  the following
 generic process:

 • Equipment scoping, based  on materials,  earthwork  and pro-
   ductivity requirements
 • Project safety planning
 • Decontamination planning
 • Contingency planning

   The contractor views the closure of a hazardous waste landfill
 differently than does the engineer. The good design engineer best
 serves his client, the owner, if designs and specifications consider
 the constructability questions  the contractor faces.

 Equipment Scoping
   Both sites A and  B present good examples of the constructor's
 response in equipment  and  approach to particular site  re-
 quirements.  Site A, in South  Florida, was characterized by con-
 strained work  space and relatively steep side slopes. Site B, in
 South Carolina, required effective  hauling of a large quantity of
 clay cover to the site.
   The constrained (bounded closely on three sides by water) site
 conditions at Site A, South Florida, required the landfill ground-
 form  to  be constructed with a side slope as steep as practical. The
204    BARRIERS & WASTE SOLIDIFICATION

-------
designer selected 3 horizontal to 1 vertical (3:1). Construction of
the topliner section required the importation of about 16,000 yd3
of soil. Permit requirements required construction to be complete
in an 8-week period. The physical constraints not only dictated
the landfill groundform, but also limited the  area available for
materials staging.
  The steep  side  slopes and  moderate  materials requirements
complicated earthwork at Site A.  All  soils were placed using
12-yd3 scrapers pushed by a dozer up the sides  of the landfill and
depositing  soils as they glided down-slope. Materials  deliveries
were carefully coordinated with placement productivity because
of the limited staging area.
  The sodium bentonite-local sand liner was mixed on-site using a
pugmill to  effect thorough mixing.  About half-way through the
clay liner construction, mechanical problems forced abandon-
ment of this mixing technique.  Thereafter, the clay and sand were
mixed using bulldozers to place the materials together  in a con-
trolled area. The scrappers then picked up and placed the mix.
  The 4-in. thick clay liner was difficult to place on the steep side
slopes. Measured thickness of  the liner varied  from 3.5 to 7 in.,
averaging about 5 in.
  A self-propelled, rubber tired vibratory compactor was used to
compact the sand-clay topliner at Site A. Again, because of the
steep side slopes, the compactor was pushed up-slope, then glided
and compacted moving down-slope. The thin liner section limited
equipment  alternatives. The contractor, concerned that a dozer-
towed vibratory compactor would damage the 4-in. liner with its
tread, elected to use the rubber tired compactor. Adequate com-
paction was  achieved in  a single  pass, which  alleviated the
awkwardness of compaction,
  Experience at Site A indicated that 3:1  slopes are  about the
steepest which may be practically constructed, limiting equipment
and earthwork  control. Indeed, flatter slopes are more  desirable
from a construction standpoint.
  At Site B,  South Carolina,  emplacement of the clay liner re-
quired moving about 30,000 yd3 of clay from the borrow source 1
mile (by road) from the site. To optimize placement, a  high pro-
ductivity earthwork program was established,  moving 3,000 yd3
per  day. Self-loading 12-yd3  scrapers  were used to  pick up,
transfer and place  the clay. Kneading compactors ("sheepsfoot"
compactors) towed behind bulldozers were used to compact the
clay to specifications.

Project Safety Planning
  Safety should be given the  highest priority during any work
with hazardous wastes. Personnel involved in cleanups at  both
Sites  A and  B were participants  in  comprehensive medical
monitoring programs—both that on-going by the contractor and
project  specific medical monitoring.
  No closure of a  hazardous waste  landfill should be completed
without a Site  Safety Plan (SSP) developed by a Certified In-
dustrial Hygienist (CIH). This  document not only advises person-
nel of hazards  and potential hazards caused by exposure to the
landfill  materials,  but also establishes site safety protocol.  At a
minimum, this protocol includes definitions of levels of protec-
tion required for different areas of the site as well as site monitor-
ing procedures and emergency equipment and procedures.
  Both Sites A and B required personnel to wear full face air-
purifying respirators and skin protection afforded by disposal
suits and gloves. Working in this wear (particularly in summer) is
quite  fatiguing and should be associated with increased breaks
and liquid intake. Project planning should consider the associated
loss in productivity.
  Both Sites A and B were attended by a full-time health and
safety officer. A portable  decontamination trailer,  including
showers, and contingency decontamination equipment  was util-
ized on each site. Daily "tailgate" safety sessions were conducted
throughout the projects to address safety concerns for each day's
planned work.

Equipment and Personnel Decontamination
  Equipment and personnel decontamination is required follow-
ing any exposure to hazardous materials. The best decontamina-
tion program is, of course, one which minimizes the contact of
equipment and personnel with the hazardous wastes. However, as
this exposure is not always avoidable, equipment and personnel
decontamination should be provided for.
  Equipment decontamination procedures should be such that no
decontamination is trafficked off-site. Equipment decontamina-
tion normally is undertaken on a specially constructed pad de-
signed to collect and store any wash water.
  A convenient personnel decontamination area should be pro-
vided. At this station, workers should have easy access to all safe-
ty equipment required for work on the site. It is here that workers
remove street clothes and put on protective clothing before enter-
ing the site.

CONCLUSIONS
  Landfill design has progressed to the point where a number of
liner alternatives may be evaluated quickly  and cost effectively.
While quantitative methods to model landfill performance are
available,  practical  implementation   of  cost-effective   and
buildable liners requires  design insight beyond the obvious.
Closure of existing hazardous waste facilities should be completed
by contractors experienced in the field. Experience not only in the
work  but also in the associated personnel protection and decon-
tamination concerns, should be closure imperatives.
REFERENCES
1. U.S. EPA, Federal Register, July 26, 1982, 32274-32388.
2. U.S. EPA, 40 CFR 264.301(a)(l).
3. Thornwaite, C.W. and Mather, J. R., "The Water Balance," Clima-
  tology, 8, No. 1, 1955.
4. Schroeder, et al., "The Hydrologic Evaluation of Landfill Perfor-
  mance (HELP) Model," Municipal Environmental Research Labora-
  tory,  Office of Research and Development, U.S. EPA, Cincinnati,
  OH, 1984.
                                                                                  BARRIERS & WASTE SOLIDIFICATION    205

-------
             Soil Liners  for  Hazardous Waste  Disposal Facilities
                                                  D.C. Anderson
                                       K.W.  Brown  and Associates, Inc.
                                              College Station, Texas
ABSTRACT
  The Hazardous and Solid  Waste Amendments of 1984 con-
tained a requirement for double liners containing both a flexible
membrane liner and compacted soil liner. The soil component can
be used either as the sole secondary  liner or as the lower com-
ponent in a composite secondary liner.  In either case, soil liners
must be at least 3 ft thick with a demonstrated field permeability
of less than or equal to 1 x 10- ' cm/sec.  Factors affecting field
permeability of soil liners  include soil  characteristics  and liner
construction methodology. Some soil characteristics which can
yield low permeability liners (e.g., high plasticity) also can give
rise to the potential for shrinkage and permeability increases.
  One  of the most important considerations in the selection of
methods for soil liner construction is the destruction of clod struc-
ture during compaction. After the selection of a specific construc-
tion methodology, the best way to achieve design specifications
is  by implementating a sound construction  quality  assurance
program.

INTRODUCTION
  Soil liners are a required component of double liner systems
for hazardous waste disposal  facilities (Fig. 1). This requirement
was established in the Hazardous and Solid  Waste Amendments
of 1984 and further defined in the Minimum Technology Guid-
ance (MTG) on Double Liner Systems.'  In this paper, the author
discusses ways to incorporate soil liners into these liner systems
and factors most greatly affecting the field permeability of these
liners.
  All double liner systems  are required  to have primary flexible
membrane liners (FML). The MTG provides two options for the
design of secondary liner systems, both of which include soil lin-
ers. Following is a discussion of the soil liner components to these
liner systems.

Soil Liner Component to
Secondary Composite Liners
  One of the designs suggested in the MTG incorporates a secon-
dary composite liner composed of an upper FML and a lower
compacted soil liner (Fig. 1A). In the configuration suggested in
the MTG, the FML and soil liner are sandwiched together to min-
imize the potential for lateral flow between the liners. It has been
suggested that this configuration would limit leakage through de-
fects in the FML to the rate at which leachate could flow into the
area of soil liner directly under the FML defect. If this theory is
correct, composite liners should be able to reduce leakage from
disposal facilities to a much greater extent  than either an FML or
soil liner alone.
                                            PRIMARY FML
                                           -SECONDARY FML
                   UMSATUSATED ZONE
      SECONDARY      PRIMARY
      LEACHATE      LEACKATE
      COLLECTION—,  (COLLECTION
                                       -PRIMARY FVL
                   UNSATURATEC 20NE

                         Figure 1
Double Liners for Hazardous Waste Landfills Must Have a Primary FML
     and Either an (A) Composite, or (B) Thick Compacted Soil
                      Secondary Liner
  Requirements for the soil component to secondary composite
liners are given in Table 1. The traditional requirement for a
permeability of less than or equal to 1 x 10-' cm/sec must be
verified in the field.' Laboratory permeability tests are no longer
considered adequate to verify actual field performance.

Secondary Soil Liners
  The other secondary liner design given in the MTG is that of a
thick compacted soil liner (Fig.  IB). The  liner  must be thick
enough and have a sufficiently low permeability to prevent the re-
lease of waste  constituents prior  to the end of the post-closure
period or greater than 30 years.' There is a considerable amount
of uncertainty in the regulatory community  as to whether such a
liner is feasible. A currently available model for evaluating the
time to breakthrough of moisture through unsaturated soil liners
indicates that such a liner would  need to be very thick.2 For in-
stance, the  model indicates that a 5-ft thick soil liner  with a
206
       BARRIERS & WASTE SOLIDIFICATION

-------
                             Table 1
   Requirements for the Soil Component to Secondary Composite Liners
   Total compacted thickness of > 36 in.
   Chemically resistant to the waste and leachate
   Permeability of < 1 x 10 -' cm/sec
   Permeability verified in field tests
   Compacted lifts of ? 6 in.
   Scarification between lifts
   Uniform moisture distribution
   Protection from desiccation/freezing
   Protection from damage due to traffic
   Adequate thickness to prevent failure from hydraulic pressures
   Minimization of clod size
   Removal of rocks, roots and rubbish
   No structural  nonuniformities which might increase the field perme-
   ability
   A comprehensive construction quality assurance plan
permeability of approximately 4.7 x 1.0-' cm/sec and only a
1-ft head of leachate would allow initial breakthrough in only
11 years (Fig. 2).
  The U.S. EPA has several reservations concerning the use of a
secondary soil liner, including the following:'
• There is no  standard method for measuring breakthrough in
  soil liners
• It is questionable as to whether  a soil liner of sufficient thick-
  ness would be economically feasible
• There is a great deal of uncertainty concerning the feasibility of
  constructing soil liners  with sufficiently low permeability to
  prevent constituent breakthrough during the operational life of
  the facility
  While use of a secondary soil liner is not ruled out by the U.S.
EPA, it may be very difficult to prove that a particular design
can meet the  no breakthrough requirement. Consequently,  the
course of least regulatory resistance and best economic feasibility
would appear to be the composite secondary liner.
                          I. INITIAL MOISTURE CONTENT OF 28%
                          2 PERMEABILITY = 4 7 < I0'9cm sec"'
                          3.SIGNIFICANT MOISTURE CHANGE AT BASE
                           OF THE SOIL LINER INDICATES INITIAL
                           BREAKTHROUGH.
                        2.0     E5      30

                        LINER DEPTH (FT)
                           Figure 2
Moisture Content as a Function of Depth in a 5-Foot Thick Liner Beneath
   a Landfill After 11 Years of Exposure to a 1-Foot Head of Leachate
                   (Modified from Reference 3)
FACTORS AFFECTING FIELD PERMEABILITY
OF SOIL LINERS
  While there  are  additional requirements in certain cases, all
soil liners must be at least 3 ft thick and have a permeability of
less than or equal to 1 x 10-7 cm/sec. The  U.S. EPA believes
that field permeability tests are needed to verify that the perme-
ability requirement  can be met using the  soil material,  equip-
ment  and compaction procedures proposed for use in the con-
struction of the soil liner.'
  Recent studies have indicated  that laboratory tests tend to
underestimate  the  actual permeability obtained  in the field.4"7
Several of  these studies discuss cases where very low permeabili-
ties were obtained in laboratory tests but much higher permeabil-
ities were detected in the field. It has become  increasingly appar-
ent that a thorough understanding of factors that affect perme-
ability is necessary to construct a soil liner which meets the field
permeability requirement. Following is a discussion of several of
these factors.

Soil Characteristics
  Characteristics of a soil which affect permeability include clay
content, plasticity,  water content  and density. Fine-grained and
high plasticity soils can be compacted to obtain low permeability
liners. However, high plasticity soils tend  to be susceptible to
large permeability increases in the field due to desiccation  crack-
ing. Highly plastic clay soils are widely used in  the construction of
soil liners  for  hazardous waste facilities  along the Gulf Coast.
During the hot, dry summer months, clay liners constructed from
these  soils can undergo significant desiccation  in a  matter of
hours. In a few days, desiccation cracks can extend downward
through much of a standard 3-ft thick soil liner. To prevent desic-
cation, a protective cover,  such as an FML, should  be placed
immediately following the completion of a soil liner.
                              OPTIMUM
                           WATER CONTENT
                                                                       SHRINKAGE (%)
                                                                      DENSITY (PCF)
                                                 KNEADING
                                                                                                                     VIBRATORY
                                                STATIC
                                                                                    104
                102
                                                                                      12    M     16    18    20   22   24
                                                                                              WATER COMTENT
                           Figure 3
  Shrinkage as a Function of Water Content and Type of Compaction
                      (From Reference 9)
                                                                                      BARRIERS & WASTE SOLIDIFICATION    207

-------
  Low permeabilities can be obtained by compaction of soils at
water contents a few percent  above optimum.' However, soils
compacted wet of optimum generally have a greater potential for
shrinkage (Fig. 3). An increase  in shrinkage potential may well co-
incide with an increased potential for cracking if a soil liner is
allowed to desiccate.
  Permeability generally decreases with increases  in the density
of a soil. There can, however,  be orders of magnitude differences
in the permeability of soil samples compacted to the same dens-
ity." This possibility is  especially important  to note due to  the
widespread reliance on density tests for construction quality con-
trol testing of soil liners. In  addition to testing the  density, water
content, particle size distribution and plasticity of a soil liner, a
construction quality control program should include laboratory
and field permeability tests.

Liner Construction Methodology
   A wide variety of equipment and  procedures are used to con-
struct soil  liners.10'" Few studies have been conducted, however,
to evaluate how these construction factors affect field permeabil-
ity. Examples of these factors include precompaction processing
of soil, characteristics  of compaction and construction  quality
assurance.
   Precompaction processing of soil includes such activities as the
removal of off-specification materials, reduction  in the  size of
clods and the adjustment of the  moisture content of  soil.  Re-
moval  of  off-specification  materials is important because their
presence can result in zones of high permeability due to either in-
 terference in the compaction  process (e.g., rocks, gravel, indur-
ated soil) or formation of channels through the liner (e.g., roots,
 trash, sand lenses).  It  is also important to minimize clod  size
because clods can both prevent the uniform distribution of mois-
 ture and result in the  occurrence of large pores between  clod
 remnants. One study found that increases in maximum clod size
caused increases  in the resulting permeability of a  compacted
soil.6 The ultimate solution, however, is to use compaction equip-
ment and procedures  which destroy all clod structure in the soil
liner.
  The compaction equipment used  will greatly affect the perme-
ability of a soil liner. If, for instance, a smooth wheeled roller,
tracked  vehicle or rubber tired roller is used, there may not be
sufficient shearing stress and remolding at depth in a lift to de-
stroy clod structure. Even padfoot rollers with short feet may not
completely remove clod remnants. Sheepsfoot rollers  with feet at
least as long as the loose lift thickness  are the best units in wide-
spread use. Such  a compactor, if sufficiently ballasted, can effec-
tively remove clod structure and the  associated large interclod
pores throughout the thickness of a lift.
  Procedures for compaction which can reduce field permeabil-
ity include the use of thinner lifts,  more compactor passes  and
immediately  covering the liner after compaction.  Thinner  lifts
assure effective compaction throughout the lift thickness. When
thin lifts are  installed with sheepsfoot  rollers, clod  remnants are
removed and the occurrence  of high horizontal permeability
zones between lifts can  be  minimized. More compactor passes
(i.e., the application of a greater compactive effort) generally re-
sults in a lower permeability liner.1  Finally, the immediate cover-
age of a soil  liner after compaction can preserve the low perme-
ability by preventing  either the heaving and ice lense formation
associated with freezing or the cracking associated with  desicca-
tion.
  One of the most underrated factors affecting field permeability
of soil liners is construction quality  assurance (CQA). A properly
conceived and implemented CQA program will ensure that a soil
liner is constructed in a way that meets the design specifications
(e.g., 3  ft thick  and  field permeability of less than  or equal to
 1x10-' cm/sec). The U.S. EPA"  has developed guidance which
recommends that CQA programs be implemented in three phases
                                   -AT LEAST THREE  SIX-INCH THICK LIFTS OF COMPACTED SOL
                                    A DRAINAGE LAYER OR UNDERORAiNAGE COLLECTION SYSTEM
                                 L-  DISTANCE REQUIRED FDR CONSTRUCTION EQUIPMENT TO REACH NORMAL
                                    RUNNING SPEED

                                 W-  DISTANCE AT LEAST FOUR TIMES WIDER  THAN THE WIDEST  PIECE  OF
                                    CONSTRUCTION EQUIPMENT
                                r7\
                                22'  AREA TO BE USED FOR  TESTING

                                                             Figure 4
                                                       Schematic of a Test Fill
                                                        (From Reference 13)
 208    BARRIERS & WASTE SOLIDIFICATION

-------
(i.e., pre-construction, construction and post-construction). Pre-
construction activities associated with soil liners include inspec-
tion of the soil and construction of a test fill (Fig. 4). The test fill
should be designed to validate the materials, equipment and pro-
cedures planned for use in the construction of the soil liner. While
a number of index property tests should be conducted on the test
fill," the most important of these is field permeability. An in-
filtrometer capable of measuring field permeabilities of less than
1 x 10 - ' cm/sec has been described by Anderson, et al.'"
  After completion of the pre-construction phase of the CQA
program for soil liner construction, there  should be  few remain-
ing uncertainties as to how to achieve the design specifications.
The construction phase CQA activities simply should be directed
toward assuring  that the soil liner  is constructed with the ma-
terials, equipment and procedures validated in  the pre-construc-
tion activities. The post-constructive CQA  activities should in-
clude a final inspection of the soil  liner followed by immediate
placement of a cover sufficient to protect the soil liner from
desiccation, freezing and other types of damage.

CONCLUSIONS
  All double liner systems to hazard9us waste  disposal facilities
must have a compacted  low permeability soil  component. The
most important design specification for soil liners is the require-
ment for a demonstrated field permeability  less than or equal to
Ix 10-'cm/sec.
  Factors impacting  the  field permeability of soil liners include
characteristics  of the soil and methods  used in the construction
of the liner. Soil characteristics which can result in a low perme-
ability liner (e.g., water content above optimum, high clay con-
tent and high plasticity) may also yield a  liner  which has a high
shrinkage potential. If such a soil liner is allowed to desiccate,
shrinkage cracks  may form, resulting in  large permeability in-
creases. Methods of construction which can result in a low perme-
ability liner include removal of off-specification materials from
the soil, reduction in clod size, minimization of lift thickness, use
of sheepsfoot rollers with feet as long as  the loose lift thickness
and implementation of a properly conceived CQA program.
REFERENCES
 1. U.S. EPA, Minimum Technology Guidance on Double Liners Sys-
   tems for Landfills and Surface Impoundments, U.S. EPA, Wash-
   ington, DC, EPA/530-SW-85-014, Second Draft, 1985.
 2. U.S. EPA, Procedures for Modeling Flow Through Clay Liners to
   Determine Required Liner Thickness (Draft Technical Resource Doc-
   ument for Public Comment), U.S.  EPA, Washington, DC, EPA/
   530-SW-84-001.1984.
 3. Johnson, R. and Wood, E., Unsaturated Flow Through Clay Liners.
   Report to the U.S. EPA,  GCA Corporation, Bedford, MA, EPA
   Contract #68-01-6871,1984.
 4. Griffin, R.A., Hughes, R.E., Follmer, L.R., Stohr, C.J., Morse,
   W.J., Johnson, T.M., Bartz, J.K., Steele, J.D.,  Cartwright,  K.,
   Killey, M.M. and DuMontelle, P.B., Migration of Industrial Chem-
   icals and Soil-Waste Interactions at Wilsonville, Illinois, In: Proc.
   of the Tenth Annual Research  Symposium  on Land Disposal of
   Hazardous Waste, U.S. EPA, 600/9-84-007, 1984.
 5. Gordon, M.E. and Hueber, P.M., An Evaluation of the Perfor-
   mance of Zone Saturation Landfills in Wisconsin,  Presented at the
   Sixth Annual Madison Waste Conference, September  14-15, Uni-
   versity of Wisconsin, Madison, WI, 1983.
 6. Daniel,  D.E., Predicting Hydraulic Conductivity  of Clay Liners,
   J. Geotech. Eng., 110, 1984, 285-300.
 7. Boutwell, G.P. and Donald, V.R.,  Compacted Clay Liners for In-
   dustrial Waste Disposal, Presented at the ASCE National Meetings,
   Las Vegas, NV, Apr. 1982.
 8. Mitchell, J.K., Hooper, D.R. and Campanella, R.G.,  Permeability
   of Compacted Clay, J. Soil Mechanics Foundations, ASCE91(SM4),
   1965,41-65.
 9. Seed, H.B. and Chan, C.K., Structure and Strength Characteristics
   of Compacted Clays, /. Soil Mechanics and Foundations, ASCE 85
   (SMS), 1959, 87-128.
10. U.S. EPA, Covers  for Uncontrolled Hazardous Waste Sites, U.S.
   EPA, Cincinnati, OH (in press).
11. U.S. EPA, Lining of Waste Impoundments and Disposal Facilities,
   U.S. EPA, SW-870, Washington, DC, 1983.
12. U.S. EPA, Construction Quality Assurance  for Hazardous Waste
   Land Disposal  Facilities  (Public Comment  Draft),  U.S.  EPA,
   EPA/530-SW-85-021, 1985.
13. Anderson, D.C.,  Sai, J.O. and Gill, A., Surface Impoundment Soil
   Liners, Report to U.S. EPA by K.W. Brown and Associates, Inc.,
   EPA Contract #68-03-2943, 1984.
                                                                                     BARRIERS & WASTE SOLIDIFICATION    209

-------
                         Slurry  Wall  Economical  in Dewatering
                                 of Sydney  Mine Disposal  Site
                                                   Bruce J. Haas
                                                 Mark  R. Nielsen
                                                Norman  N.  Hatch
                                                   CH2M HILL
                                               Gainesville,  Florida
ABSTRACT
  The Hillsborough County, Florida, Department of Solid Waste
used the Sydney Mine site for disposal of septage tank and grease
trap wastes and waste oil products. When hazardous constituents
were found in the ponds and groundwater, remedial action was
begun to remove and incinerate the pond contents. Conventional
dewatering techniques would have resulted in excessive incinera-
tion quantities and  costs. A temporary slurry wall used for de-
watering during site cleanup resulted in substantial cosl savings.

INTRODUCTION
  The Hillsborough County, Florida, Department of Solid Waste
deposited septage tank and grease trap wastes and waste oil pro-
ducts in two shallow ponds excavated in tailing sands left from
phosphate mining operations at Sydney Mine. The tailing sands,
approximately 10 to 25 ft thick, are underlain by very soft plastic
clay slimes, also  a byproduct of mining operations. A  perched
groundwater table occurs in the tailing sands near the ground sur-
face.
  When hazardous  constituents were found in the ponds and
groundwater, a remedial action was begun to remove and incin-
erate the pond  contents. If groundwater flowed into the ponds
during  removal of  the pond contents, then excessive  incinera-
tion quantities and costs would have resulted. Conventional de-
watering using well  points could have required the treatment of
approximately 36 million gallons of groundwater at  an esti-
mated cost of $1.8 million, assuming treatment costs of approx-
imately $0.05/gal.
  This  paper relates how a temporary slurry wall was used to
facilitate dewatering during site cleanup, thus  resulting in sub-
stantial  cost savings. By  installing a slurry wall to isolate  the
ponds,  the groundwater could be lowered at a  slower rate using
fewer well  points.
  Traditionally, slurry walls have been used to facilitate dewater-
ing of excavations.  Since the early 1980s, they have been used
for containment of  hazardous wastes and contaminated ground-
water and  for decreasing recovery volumes  for groundwater
restoration. The slurry wall constructed  at the  Sydney Mine site
combined these various applications for a cost-effective site clean-
up. Unique site conditions made possible additional cost savings
in wall construction  and dewatering.

SITE HISTORY
  The Sydney Mine Waste Disposal site is located east of Tampa
in Hillsborough  County, Florida (Fig.  1). The land was strip-
mined for removal of phosphate ore during the 1930s and 1950s
and then abandoned. Hillsborough  County leased the 9.5-acre
site in 1973 for disposal of grease trap waste, septic waste and
waste oil; the county utilized the site for waste disposal activities
until November 1981.
  The primary disposal facilities on-site consisted of two surface
impoundments, a 1.5-acre septage pond and a 0.6-acre oil  pond.
Each impoundment is about 4 to 6  ft deep. A  review of initial
site investigations';1 and additional intensive sampling' indicated
the presence of toxic organics in the disposal ponds and  in the
perched groundwater beneath the ponds.
                         Figure 1
           Sydney Mine Waste Disposal Site Location

Feasibility Study
  Using data obtained during the site investigations, a feasibility
study was conducted to identify and evaluate site closure alterna-
tives.' This study addressed both cleanup of the waste ponds and
restoration of the contaminated  groundwater. The general ap-
proach was to select technologies  for source removal or  control
that  were technically feasible, environmentally sound and cost-
effective. Following  source  removal  or control,  the selected
groundwater restoration technologies would  be implemented.
Groundwater cleanup therefore would be undertaken only after
the source of the contamination  was removed. This approach
would provide effective implementation of the remedial program
as County funding became available.
 210    BARRIERS & WASTE SOLIDIFICATION

-------
  Twenty-two alternatives were evaluated for the closure of the
waste ponds, with three categories of alternatives recommended
for further consideration by the County:
• On-site Containment. Waste materials would be left at the site,
  and features to impede off-site migration of contaminants from
  the source materials would be constructed. This solution would
  have involved construction of a permanent slurry containment
  wall all around the ponds and a clay or synthetic cap over the
  ponds.
• Off-site Disposal. All contaminated liquids and solids would
  be removed from the ponds and transported  to an approved
  out-of-state facility for disposal. No licensed facilities for dis-
  posal of such wastes currently are located in Florida.
• On-site Destruction.  All contaminated wastes would be re-
  moved from the ponds and incinerated on-site to destroy the
  organic contaminants.
  County officials studies each option to determine which would
be the most environmentally responsible and  yet cost-effective.
On-site containment was eliminated because the  potential future
migration of contaminants from  the site remained  an undesir-
able risk.  Also, leaving contaminated material at the site would
limit future use of the site.
  Because either of the remaining options  was acceptable, the
County retained CH2M HILL as  prime contractor to conduct a
detailed comparison  of off-site disposal and on-site destruction
and  to implement the site restoration.  Proposals were solicited
from potential subcontractors for  surface cleanup and closure of
the site. The responsible proposals showed that on-site destruc-
tion was less expensive than off-site disposal because of high costs
associated with transporting the wastes out-of-state. The County
therefore selected on-site destruction for implementation.
  A mobile waste incinerator  was found to be the least-cost
alternative for on-site destruction of organic contaminants within
the pond liquids, sludges and bottom sands. The mobile inciner-
ator would be transported to the site, set up, operated and then
removed when all the on-site waste materials had been inciner-
ated. Environmental Services Company (ENSCO, Inc.) of Little
Rock, Arkansas, in a joint venture with Tyger Construction Com-
pany of  Spartanburg, South Carolina, was selected to  remove
and incinerate the wastes. ENSCO manufactured the modular in-
cineration unit at its Pyrotech  Systems, Inc.  facility located in
White Bluff, Tennessee.


SURFACE CLEANUP
  The method of excavation of pond  wastes and underlying con-
taminated soil was not detailed in the surface  cleanup proposal.
ENSCO-Tyger proposed to pump out impounded liquids for in-
cineration, then excavate using bulldozers or backhoes to remove
sludges and soils.
  Upon closer examination of the cleanup  proposal, the feasi-
bility and cost  of removing materials  in the ponds became  a
source of concern. The groundwater table was at approximately
the same level as the liquid level  in the ponds.  A thin layer of
grease and oil covered the bottom of the ponds and provided an
impediment to vertical seepage from the ponds.  However, if the
liquid levels in the ponds were to be lowered, this thin layer would
not provide a barrier to groundwater flow in the reverse direc-
tion.
  The volume of groundwater seepage into the ponds would de-
pend on actual groundwater levels, rate of waste removal inside
the ponds and the permeability and thickness of the tailing sands.
An estimate of the groundwater flow suggested that more than
200,000 gal/day of groundwater might seep into the ponds during
source removal.  Without groundwater recovery  or control con-
current with source removal, this would have resulted in costs of
approximately $60,000/day  to  incinerate the groundwater.  In
reality,  the incineration facility would not have  been able  to
handle or treat that volume of  flow; ultimately, the pond con-
tents could not have been removed.
  Dredging to remove solids and contaminated soils seemed to be
a viable alternative since the majority of the contaminated wastes
(having the highest BTU value) could have been excavated and
incinerated without altering  groundwater levels. The remaining
impounded liquids would have been somewhat dilute and could
have been treated using a  separate groundwater  recovery and
treatment system; only a limited volume of liquid would  be in-
cinerated. However, the cleanup permit negotiated with the State
of Florida stipulated that  contaminated soils were to be exca-
vated to remove  all  "visible contamination." Dredging  below
the water level in the impoundments would not permit visible in-
spection of the pond bottom following removal of the contam-
inated  soils and solids. Furthermore, dredging raised the  possi-
bility that soils without  visible contamination could  become
contaminated during  excavation by coming into  contact with
pond liquids.  Therefore, it would  not have been  feasible  to
demonstrate removal of the contaminated soil.
  Installation of a dewatering scheme using a series of well points
around the ponds to lower groundwater levels prior to waste ex-
cavation was an acceptable possibility.  A groundwater treatment
system would have had to be constructed to treat the ground-
water from this dewatering system. Groundwater treatment using
air stripping and carbon adsorption would have been substantially
cheaper than incineration. At a treatment cost of approximately
$0.05/gal, treatment of the groundwater from dewatering would
have cost approximately  $10,000/day. If the  surface cleanup
lasted for 6 months, the treatment costs would have amounted to
more than $1,800,000. Although the groundwater treatment sys-
tem would be implemented  ahead of schedule, it could be used
during final groundwater restoration following source removal.
  Dewatering would have resulted in the removal of approx-
imately 200,000 gal/day of groundwater. This rate of ground-
water removal was substantially reduced by installing a temporary
slurry wall upgradient of the dewatering well point system. Tra-
ditionally, slurry  walls used in dewatering applications have
been useful where the required groundwater removal rates are so
large that drawdown cannot effectively be achieved. The Sydney
Mine site, having shallow, fine sand deposits and less than 8 ft of
drawdown to achieve  suitable dewatering, would not normally
warrant the use of a slurry wall for temporary dewatering.  How-
ever, high costs of groundwater treatment  associated with the
dewatering system made the relatively high capital  costs of a
slurry wall economical, even for short-term dewatering use.
  Estimates of probable construction costs indicated a slurry wall
would  cost on  the order of $250,000 and  would reduce total
groundwater treatment costs to approximately $300,000 for a  6-
month surface cleanup period. This represented a total potential
savings of more than $1,250,000. If surface cleanup continued for
a longer period of time, even greater savings could be realized.
Therefore, it was recommended that a contaminated  slurry con-
tainment wall be used in conjunction with a temporary well point
dewatering system and associated groundwater treatment system
for the Sydney Mine site.

DEWATERING DESIGN DEVELOPMENT
  Slurry walls have been used successfully in several hazardous
waste  remediation applications. These applications have  in-
cluded physical waste containment as well as a hydraulic barrier
to reduce groundwater extraction quantities. Most of these appli-
cations  are permanent or long-term installations  where  waste
compatibility and wall integrity are critical to the reliability of the
total remedial action.
                                                                                 BARRIERS & WASTE SOLIDIFICATION    211

-------
  At the Sydney Mine waste disposal site, however, the slurry
wall represented a short-term means of reducing groundwater re-
covery volumes. Contaminant concentrations in the upgradient
groundwater were relatively low; therefore, the effect of possible
contaminant degradation of wall backfill materials was not con-
sidered  a  serious threat to wall performance. In addition,  be-
cause the wall was not intended to serve as a sole method of con-
tainment or groundwater barrier, a loss of integrity of the wall
resulting in the presence of "windows," seams or gaps would
not represent an imminent threat to contaminant migration.
  Previous  subsurface explorations and  investigations  at  the
Sydney Mine waste  disposal  site provided information  on site
hydrogeologic conditions and  groundwater quality.2'1 However,
detailed soil characteristics and geotechnical information, such as
grain size, density, strength, permeability and variations in condi-
tions across the site, were not available. Because of the strict time
constraints, design and construction of the groundwater contain-
ment had to be completed prior to the start of waste incineration.
Therefore, the existing information was supplemented by con-
ducting a program of soil test borings and laboratory testing dur-
ing construction of the slurry wall.

Hydrogeological Setting
  A general understanding of the hydrogeological conditions
beneath the site was developed by evaluating well logs. These
logs revealed a unique combination of natural and man-made
conditions.
  Natural soil conditions in the area typically consist of approx-
imately 30 ft of silty  sands overlying stiff phosphatic clays of the
Bone Valley and Hawthorn formations.  These formations pro-
vide an effective upper confining layer for the porous limestones
which  comprise the  Floridan  aquifer,  an artesian  aquifer and
principal source of drinking water in the area. A perched ground-
water table is present in the overburden material above the phos-
phatic clays.
  Strip-mining  of   the  phosphate-rich  deposits  significantly
altered  these natural conditions, as shown in Fig.  2. The phos-
phatic clays were mined to a depth of approximately 60 ft;  the
former  strip mine area was eventually reclaimed  by backfilling
with clay wastes from phosphate processing. These wastes, called
clay slimes, are very  soft unconsolidated clays having a relative-
ly low permeability.
  Settling  ponds were constructed for the disposal of the clay
slimes by forming perimeter dikes using mine overburden ma-
terial. The dike itself, therefore, consists of moderately permeable
silty and clayey sand deposits. The Sydney Mine waste disposal
site is located at the southern edge of a  former dike separating
two settling ponds.
  The clay slimes were subsequently covered with a thin  layer of
tailing sands.  These tailing sands  are  fine-grained, uniformly-
graded, washed sands also obtained as a  byproduct of phosphate
processing. The thickness of the tailing sands varies, but is gen-
erally thickest at the  perimeter dikes and thinnest in the center of
the settling ponds. The tailing  sands reach a maximum thickness
of approximately 25 ft near the settling pond dike.
  Consolidation and settlement of the  soft -clay slimes  under-
lying the tailing sands  has  not been uniform. Next to the dike
where the thickness-of the clay slimes decreases because the dike
slopes, the surface of the clay slimes has developed a pronounced
"lip" which effectively separates the tailing sands from the dike
deposits. The elevation of  this "lip" is  near  the former water
level within the  settling ponds.
  A perched water table, recharged by rainfall, has developed
in the tailing sands. This perched water table is not used for'water
supply and  varies in thickness. North of the dike, the  perched
   Conumininl Flow
                          Figure 2
    Hydrogeologic Cross Section of the Sydney Mine Disposal Site

groundwater  table is present at a  level of approximately 20 ft.
South of the settling pond dike,  the perched  water table has a
limited outlet for groundwater flow; the "lip"  of clay slimes acts
as a weir, trapping the perched groundwater  within the tailing
sands. During rainy periods, a rising water table allows ground-
water to flow over this clay weir into the separate perched ground-
water within  the dike.  The groundwater table  south of the dike
is generally near the ground surface, approximate elevation 90 to
100 ft. The groundwater table within the dike deposits is present
at a level which is about 20 ft lower than the perched groundwater
within the tailing sands.

Wall Design
  These unique geologic conditions substantially influenced the
design of the groundwater containment system  for cleanup of the
Sydney  Mine waste disposal site.  Because the perched ground-
water within the dike was well below the level of the ponds, it did
not serve as a source of recharge or groundwater flow to the area
of the ponds. Therefore, both the slurry wall and the dewatering
well point system were constructed around only three sides of the
existing  waste impoundments, as shown in Fig.  3. In this way, the
slurry wall containment system would serve its  intended function
as an effective temporary means of dewatering the ponds, but
would not function as  a reliable long-term containment system.
The depth and length of the wall were therefore much less than
would have been needed for site containment,  providing further
cost savings.
  Design of the slurry wall specified a minimum width of 2 ft and
a minimum embedment into the clay slimes of 2 ft. The backfill
was to consist of a mixture of bentonite and the excavated tail-
ing sands and clay slimes. Conventional Wyoming bentonite
(high-swelling sodium  montmorillonite) was to be added  to the
backfill  mixture in dry powder form to comprise a minimum of
6% by weight of the completed backfill. The mixture could then
be sluiced with bentonite slurry from the trench to produce a
backfill  having a slump between 3 and 6 in. and a permeability
less than 1 x  10"7) cm/sec. Compatibility testing of the specified
mix was not performed because: (1) the contaminant concentra-
tions in  the area of the wall were generally low, (2) the wafl was
not intended  to serve as a long-term containment barrier~and (3)
only limited time was available for testing prior  to construction.
  The dewatering system to be installed inside the enclosed area
between the ponds and  the slurry wall consisted of a series of weB
points about  75 ft apart for a total of 20  wells. Each well point
was  to consist of a 2-in. diameter PVC  standpipe  with a 2-ft
slotted tip.
212    BARRIERS & WASTE SOLIDIFICATION

-------
        Scale in Feel

        0   100  200
                                    •x   r"n j—• —L
                   •   •   •  •   •  /*   I] |   |
                   O   O  4 O  O   6          Lj*!

                 Slurry Walt           Groundwater  -X
                                   Treatment —""^
                           Figure 3
            Plan of Slurry Wall and Dewatering System

  The well  points, which  were to be connected  to  a vacuum
header pipe, were to be installed by jetting to a depth  just above
the top of the clay slimes. Specifications called for the completed
dewatering system to be capable of delivering a minimum dis-
charge rate of 150 gal/min at a discharge pressure of at least 20
lb/in.2.
  Contract documents for the groundwater containment system,
including the slurry wall  and well point dewatering system, were
prepared and distributed  to four potential subcontractors for bid
in early September  1984. The successful bidder was Riverside
Construction Company of Tampa, a subsidiary of Moretrench
American Company of Rockaway, New Jersey.
  A groundwater treatment system was installed under a separate
subcontract and is only briefly described here. The treatment
system consisted of a 100,000-gal flow equalization  tank followed
by  air stripping, iron removal and carbon adsorption. Treated
groundwater effluent was  stored  temporarily  in  a  second
300,000-gal  holding  tank prior to spraying the effluent on the
tailing sands south of the  site.

SLURRY WALL CONSTRUCTION
  Initial site preparation  was begun on schedule in October 1984.
A 65-ft wide level pad was constructed around the perimeter of
the site to accommodate trenching and backfill mixing  equip-
ment. The pad was constructed initially using tailing sands avail-
able on-site; however,  the tailing sands were found to be exces-
sively uniform and non-cohesive. An imported silty material was
therefore brought in to  provide a  stable working surface in a
portion of the trench area.
  Soil borings were drilled along the  slurry wall alignment at a
spacing of approximately 100 ft. The borings were to be drilled to
a depth of 4 ft below the top of the clay slimes, although the sub-
contractor elected to drill even deeper to verify a minimum thick-
ness of the clay slime stratum. Split spoon samples  were taken
continuously at the anticipated level of the clay slimes.
  The soil borings indicated that the level of the clay  slimes was
higher than interpreted from earlier hydrogeologic studies. The
decreased depth to the clay slimes reduced the overall surface area
of the slurry wall, resulting in minor construction cost savings.
The clay slimes were found to be present to a depth more than 2
ft below the bottom of the slurry trench, thereby assuring that
the  clay slimes stratum would not be penetrated during well con-
struction.
  Laboratory testing specified in the Contract Documents in-
cluded grain-size analyses  of the tailing sands for selection of
appropriate well point screen slot size. A minimum of four trial
mixes of the proposed backfill mixture was to be tested using a
triaxial cell apparatus with nominal confining pressure and driv-
ing head of 5 to 7 lb/in2.
  After reviewing the initial soil boring results, the subcontractor
proposed  an alternative to the specified soil-bentonite backfill
mixture. The clay slimes were found to be very soft and malle-
able and could be mixed relatively easily with the tailing sands to
produce a homogenous mixture. The resulting material required
little added  slurry to achieve the specified slump. The principal
benefit, however, was that additional dry bentonite did not need
to be  added to  the mixture to obtain the required degree of im-
permeability.
  A series of permeability tests was conducted on various back-
fill mixtures to  identify an acceptable ratio of clay slimes to tail-
ing sands. Results showed that a mixture of approximately 35%
clay slimes and 65% tailing sands could achieve a permeability
less than 1 x 10"7 cm/sec. Consequently, the depth of the wall
was increased nearly 50%; the total depth of the wall was 1.55
times  the depth  to the top of the clay slimes. In two small sections
of the wall where  insufficient clay slimes  were available, the
specified dry-mixed bentonite backfill was used.
  The modified wall backfill material resulted in a cost savings of
$12,200, or 8.5% of the wall installation cost. The unit price for
the wall was reduced from $4.96/ft2 to $4.54/ft2. Measurement of
the wall area for payment purposes was determined based on
actual soundings of the depth to the top of the wall and based on
a theoretical depth of the wall below the clay slimes of 2 ft, as
specified.
  Construction activities  began  in  early November  1984.  A
temporary holding pond  was constructed for mixing a  5.65%
bentonite slurry. The slurry was pumped continuously from the
holding pond to the  portion of  the trench  being  excavated.
Trench excavation was accomplished using a backhoe with a 2-ft
wide bucket. Mixing the clay slimes and tailing sands for backfill
of the trench was accomplished using a D-7 bulldozer. Mixing was
continued until a visually homogenous mixture was  obtained.
Fig. 4 shows the slurry trench excavation and backfilling opera-
tions in progress during construction of the cutoff wall.
  The tailing sands, being very loose and non-cohesive, were sen-
sitive  to variations in the slurry level within the trench. Occa-
sional caving of the trench sidewalls occurred when slurry levels
dropped more than 2 ft below the level of the construction pad.
  A geotechnical engineer was present while the slurry wall was
constructed  to ensure that the backfill mixture was homogenous
and that no sand inclusions were present  because  sand inclu-
sions  could  impair the integrity of the wall. Health  and safety
procedures were strictly adhered to during trench excavation. At
several locations along the trench  alignment, construction work-
ers were required to use  respiratory protection because of high
levels  of organic vapors detected in the breathing zone.
  After the  slurry wall was backfilled, the top of the trench was
covered with a geotextile  fabric and soil mat to support occa-
sional or inadvertent pick-up truck traffic across the completed
trench.  The geotextile fabric  consisted  of  a 12-ft wide  woven
polypropylene fabric. Native tailing sands were then placed to a
depth of 12 in. over the geotextile, and the surface was  hydro-
seeded.
  The proposed well point dewatering system also was modified
based on suggestions offered by the subcontractor.  Because the
top of the clay slime stratum was irregular and varied up to 10 ft
in elevation, operation of the dewatering system could have been
complicated by  the need to continuously monitor throttling of the
well points to avoid cavitation or loss  of vacuum in shallower
wells as the groundwater table dropped.
  The accepted alternative well point detail shown in Fig. 5 con-
sisted of jetting a 6-in. diameter temporary  casing to  a uniform
depth of 25 ft, which was typically below the level  of the clay
                                                                                   BARRIERS & WASTE SOLIDIFICATION    213

-------
                          Figure 4
         Photographs Showing the Slurry Wall Construction

slimes. A 2-in. diameter PVC well point with 5-ft slotted screen
section at the level of the clay slimes was then inserted  inside
the temporary casing. Coarse filter sand backfill was added slow-
ly around the well point while  the temporary casing was with-
drawn. Finally, a 1.25-in. diameter open-ended PVC drawdown
tube was inserted into the well point to a depth of 25 ft.  Vacuum
was applied to the drawdown tube;  cavitation  was avoided by
setting the tip of the drawdown tube  below the maximum depth
of drawdown possible using  vacuum pressure.  The completed
well point therefore acted like a sump-type of installation.
  Construction,  which began in early November, remained on
schedule and was substantially completed prior to the Christmas
holidays. Waste removal and incineration could therefore pro-
ceed without delay.

PERFORMANCE
  Quality control testing included a series of slump tests, grain
size tests and triaxial permeability tests on the completed backfill
mixture. Results of this testing verified that the  required perme-
ability of 1 x 10  7 was obtained using the clay slimes-tailing sands
mixture.
  Startup testing of the well point dewatering system verified that
the 20 wells were capable of delivering the  specified 150 gal/min
at 20 lb/in2. As the groundwater table within  the contained area
dropped, recovery  from the dewatering system declined  substan-
tially. Therefore, flows to the groundwater treatment facility were
supplemented by installing 20 additional well points outside of
the wall.  This increase in number of wells allowed  recovery of
the contaminated groundwater within  the tailing sands outside the
containment area to proceed ahead of schedule.
                           Figure 5
               Typical Dewatering Well Point Detail
  Water levels in the  containment area dropped at a fairly uni-
form  rate for 3 to 4  months before leveling off, as shown  in
Fig.  6.  Heavy  rains,  such as those associated  with hurricane
Elena in September 1985, resulted in a sudden  rise  in ground-
water levels but did not jeopardize waste excavation.
  Waste excavation from the impoundments proceeded rapidly
in early March 1985. Most of the septage pond, including under-
lying  visibly contaminated soils, was excavated and placed in a
temporary, lined holding basin.  The excavation was dry and
stable and backfilled immediately.
  Difficulties in the incineration process resulted in a series of un-
avoidable delays. As of Dec. 1, 1985, however, most of the waste
material has been removed and incinerated, and the  former im-
poundments have been  backfilled to match surrounding grade.
The  extended waste removal  period will therefore be approx-
imately 12 months. The slurry wall containment system  pro-
vided a vital  means of controlling groundwater levels through-
out this  period. Dewatering  using the slurry wall represents a
substantial savings in treatment costs over  dewatering without
use of the wall, particularly  in view of unpredictable delays  in
site remediation.
                             Figure 6
     Groundwater Level Measurements at (he Sydney Mine Disposal Site
214    BARRIERS & W^STE SOLIDIFICATION

-------
CONCLUSIONS
  Slurry wall containment systems  can be an economical solu-
tion in waste disposal site  remediation when used for temporary
construction dewatering on small sites, even if not  economical
for long-term containment (encapsulation).
  Advantages afforded by unique site conditions can  be incor-
porated into slurry wall design to earn additional cost savings. At
the Sydney Mine waste disposal site, these conditions permitted a
three-sided  wall layout  for containment  and  offered an eco-
nomical alternative in backfill materials using native clay slimes.
  Flexibility in  design and construction can result in savings in
both time and cost. Site remediation is non-traditional  construc-
tion with unforeseen conditions and imposed constraints; cooper-
ation between  contractors, designers and  owners can identify
economical alternatives during construction.
REFERENCES
1.  U.S. EPA Region IV,  "Groundwater and Surface Water Investiga-
   tion, Sydney Mine Site, Hillsborough County, Florida." Surveillance
   and Analysis Division, Jan. 18, 1980.
2.  Seaburn and  Robertson,  Inc.,  "Hydrogeological Evaluation of the
   Sydney Mine Waste Disposal Site, Hillsborough County, Florida,"
   Oct., 1980.
3.  P.E. LaMoreaux and Associates, Inc., "Hydrogeological Impact of
   Liquid Waste Disposal with Recommendations for Remedial Action
   at  the Sydney Mine Site,  Hillsborough County, Florida," Mar.,
   1982.
4.  CH2M HILL, "Sampling and Analysis Report, Sydney Mine Waste
   Disposal Site," Gainesville, FL., Feb., 1983.
5.  CH2M HILL, "Groundwater Restoration and Site Closure Alterna-
   tives Evaluation for the Sydney Mine Waste Disposal Site," Gaines-
   ville, FL, Mar., 1983.
                                                                                     BARRIERS & WASTE SOLIDIFICATION    215

-------
                        Utility of  Soil  Barrier Permeability  Data

                                                Walter E. Grube,  Jr.
                              Hazardous Waste Engineering Research Laboratory
                                     U.S. Environmental  Protection Agency
                                                   Cincinnati, Ohio
ABSTRACT
  This report discusses the limitations of singular expressions of
permeability data for soil barriers designed to restrict the migra-
tion  of chemical contaminants from  landfills or  contaminated
soil sites  into surrounding groundwater.  Also described are the
additional quantitative data needed for more accurate interpreta-
tion of stated permeability values. These  supplementary data in-
clude: (1) results of soil blend analyses, (2) results  for additional
tests of soil characteristics, (3) information about sample prepara-
tion  before laboratory  tests, (4) data on soil physical state as
tested, (5) descriptions of test time periods and test termination
criteria, (7) information on statistical data treatment and (8) ob-
servations about test post-mortems.

INTRODUCTION
  Engineered soil structures are accepted by design engineers and
regulatory agencies as viable means of controlling liquid  move-
ment in hazardous waste treatment, storage and disposal facilities
and in uncontrolled  hazardous  chemical waste sites. Compacted
clay soils are being used both as part of landfill liner systems and
as an infiltration barrier in cover designs for both landfills and
uncontrolled dump sites cleaned up under Superfund authority.
Engineered soils also form  vertical  groundwater barriers con-
structed by slurry trench techniques. The primary design and con-
struction  specification used to  characterize soil-based hydraulic
barriers is hydraulic conductivity. The  soil barrier's degree of
hydraulic conductivity is commonly  referred to as the perme-
ability, since  the factor K in D'Arcy's equation  usually is ex-
pressed in the same units as hydraulic conductivity (volume cubed
per volume squared per unit of time).
  The hydraulic conductivity (or  corresponding  coefficient of
permeability) usually is  cited as a numerical criterion that de-
scribes the liquid migration barrier properties of these engineered
soil structures. This report discusses the limitations of such singu-
lar expressions of permeability data,  and it describes the addi-
tional quantitative information  that will permit more accurate in-
terpretation of stated permeability values. These supplementary
data are as follows:

  Results of soil blend analyses
  Results from additional tests of soil characteristics
  Information about sample preparation before laboratory tests
  Data on soil physical state as tested
  Descriptions of test cell features and techniques
  Specification of test time periods and test termination criteria
  Information on statistical data treatment
  Observations about test post-mortems
  Documentation of these parameters provides a basis for com-
paring permeability test results from different laboratories, in-
vestigators and soil and permeant liquid combinations. Such in-
formation also strengthens the credibility of hydraulic conductiv-
ity data by quantifying parameters that  significantly  affect the
accuracy and practical applicability of permeability test results.

BACKGROUND
  In the late 1970s,  the U.S.  EPA proposed technical perfor-
mance specifications for landfill construction that would require
compacted clay liners for  hazardous waste treatment,  storage or
disposal facilities (TSDFs) to have hydraulic conductivities that
did not exceed 1  x 10- ' cm/sec.' Since that proposal,  regulatory
approaches have changed  somewhat,-' but strong support  con-
tinues for low hydraulic conductivity. Since compacted clay soils
also are used for cover systems  at hazardous waste storage facili-
ties or at uncontrolled chemical dumps, the hydraulic conductiv-
ity of the clay cap has been used as a design and performance cri-
terion for the completed structure.

Permeability Test Methods
  The permeability of compacted clay soils can be tested in the
laboratory using  any of four general types of apparatus: A  rigid
wall cell, a triaxial cell, a consolidation cell or a constant flow rate
cell. The permeability of clay compacted in the field can be meas-
ured using: (1) lysimeter pans installed underneath an area of the
compacted clay, (2) ring infiltrometer systems installed on the sur-
face of the compacted clay or (3) shallow well slug tests in cylin-
drical excavations into the compacted clay.

Slurry Trench Cut-Off Walls
  CERCLA  has  specified  the application of available engineer-
ing technologies  to reduce the  adverse environmental impact of
uncontrolled  hazardous chemical dump sites. Both  regulatory
agencies  and engineering  companies have accepted  the slurry
trench cut-off wall as  a  structure that  can  restrict  the lateral
groundwater flow into and out of a subsurface land mass contam-
inated by hazardous chemical wastes.1
  A slurry trench cut-off wall  is built by constructing a vertical
trench that is kept open by clay suspensions during excavation. A
barrier material then is backfilled into the trench excavation. The
material may be a mixture of soils formulated by design  engi-
neers, a soil-cement-bentonite mixture, synthetic plastic sheets or
metal sheet pilings.
  For a backfill  barrier made  of soil, permeability is a primary
characteristic for describing the lateral movement of groundwater
216    BARRIERS & WASTE SOLIDIFICATION

-------
through it. Design specifications for slurry trench cut-off walls at
uncontrolled hazardous chemical waste sites usually require that
the backfill hydraulic conductivity not exceed 1 x  10-' cm/sec,
as is the case with compacted clay soil liners and surface caps.
  The permeability of slurry  trench  backfill mixtures  can be
tested in the laboratory using  any of the apparatus  applied  to
compacted clay soils, but sample preparation techniques differ.
The permeability of an installed backfill can be evaluated by in-
stalling monitoring well systems on both sides of the barrier or by
extracting samples of backfill materials from the barrier and sub-
jecting them to laboratory permeability tests. No simple field tests
have been reported to measure  the actual hydraulic performance
of a slurry trench cut-off wall after installation.
  Although both native soils and those amended with clays, fly-
ash, cements or other materials  are characterized by other labora-
tory tests  (grain-size  distribution,  plasticity, color, exchange
capacity, pH,  etc.), the hydraulic property of the soil after con-
struction is the main parameter of interest to a regulatory agency
and thus to a design engineer.

Existing Data Base on Soil Permeability
  Numerous studies have been  conducted in the past 6 or 7 years
to determine how hydraulic conductivity tests in soil are affected
by  variables such as soil type, liquid type and test  technique.
The results of these studies have been reported in technical jour-
nal publications, agency reports and proceedings of environmen-
tal research conferences and symposia.4'5'6'7 Briefly, this  research
has documented the effects of various organic liquids on the
permeability, of many different  soil types when the organic liquid
is substituted for water in a laboratory permeability test. Similar
studies have included both dilute and highly saline inorganic liq-
uids such as those from mining, metal  plating or drilling  mud
waste disposal. In addition, numerous  tests  using compaction
mold, triaxial cell and consolidation cell permeameters have been
conducted  (including various operating technique modifications
of each) to quantify and improve test precision. Ultimately, a
report is developed stating that the soil has a hydraulic conduc-
tivity of some stated numerical value (typically near or lower than
1 x 10- ' cm/sec) when it is tested with the permeant liquid speci-
fied. Thus, a large data base that declares the permeability of in-
dividual soil materials to individual liquid systems exists.

Soil/Pollutant Compatibility
  A geotechnical testing laboratory should state clearly whether
the soil permeability test  has  been conducted  with  laboratory
water, relatively uncontaminated groundwater from the site or a
liquid containing known pollutants. If a test is being conducted to
document the stability of an engineered soil barrier after exposure
to a leachate,  contaminated groundwater or other  liquid, the
permeability test should be clearly defined as a measure of com-
patibility. This step is necessary because testing procedures, as re-
ferred to later in this paper, require particular selection and oper-
ational techniques when the primary goal of the tests  is to deter-
mine soil and pollutant compatibility.
  The U.S. EPA has published  a procedure' for determining clay
liner compatibility using  a rigid-wall permeameter. Further stud-
ies  are under way to develop  reliable procedures using triaxial
cell and consolidation cell apparatus.' Another device, the high-
pressure, constant flowrate cell, is being used to  determine the
water permeability of clays,10 but it has not been extensively used
to determine soil and liquid compatibility. Since there has  been
little experience with the constant flowrate cell and  chemically
aggressive liquids, and since reported uses of this type of appa-
ratus have  involved relatively small sample volumes, the cell will
not be discussed further in this paper.
SUPPLEMENTARY DATA NEEDED FOR
INTERPRETING PERMEABILITY VALUES
  To objectively evaluate  permeability data, it is  necessary to
have information about soil  properties and test procedures that
goes beyond the expression of the coefficient of permeability.

Soil Blend and Other Characteristics
  Although compacted clay  liner and surface cap components
have essentially common design, construction and testing charac-
teristics, soils prepared to serve as a slurry trench barrier are usu-
ally a fluid mass that contains a much higher initial water content
than a compactable clay soil. However, each of these applications
usually  requires a certain  specified minimum clay or fines con-
tent so that a minimum of pore space is present in the completed
structure. Thus, soil  engineering tests of candidate soil barrier
materials  routinely include  particle  size distribution data. A
noticeable shortcoming in these data is the usual omission of pro-
portions of gravel or other coarse fractions in the borrow (i.e., the
content to which materials coarser than  a prescribed maximum
are removed before actual barrier construction). The presence of
rock fragments more than  an inch in dimension may become a
significant inclusion  in soils  compacted  to  form clay liners or
caps.  Thus, the largest allowable soil or rock particle sizes in
barrier materials must be  specified in  the standard particle size
classification reported for candidate barrier soils.
  Before soil is characterized by laboratory tests for barrier con-
struction, the representativeness of soil sample(s) must be deter-
mined. Most clay borrow areas vary in properties, so the engineer
who is characterizing the clay source must measure and report on
the variability of the total candidate soil resource. Both research
reports and actual RCRA Part B Permit Application information
indicate that soil areal and volumetric variability are unaccounted
for parameters. Sometimes these sources do state that the  total
soil supply is uniform within the limits of variability accepted by
ASTM for particle size or engineering index test results.
  A design engineer also should insist on a valid statistical  sam-
pling of the total soil volume to be used for barrier construction.
Such methods as are standardized for coal stockpile or limestone
stockpile sampling easily could be adapted to clay reserves. For
in-situ soil sources, guidelines adopted by agricultural testing lab-
oratories for field sampling could be scaled to the soil source site.
  Any of these suggested approaches or even more detailed site-
specific sampling schemes  should  be  appended to permeability
data to  assure clients  of conscientious protocols applied preced-
ing laboratory tests.

Sample Preparation Before Laboratory Tests
  Sample preparation before the permeability test has a profound
effect on the actual results obtained. Daniel" has published data
from  a  small laboratory experiment that describes the effect of
soil clod size  on permeability tests in the laboratory. These data
have been widely cited to proclaim the importance of clod size
as a controlling factor in clay liner permeability. These data are
more significant as an indicator of the effects of laboratory  sam-
ple preparation. Rarely would a clay liner or cap be installed with
clay soil in which all the clods  were reduced to tiny fractions of an
inch in  dimension; nevertheless, it is prudent to expect that an
extrapolation of Daniel's data to larger clod sizes will result in
correspondingly greater permeability values.
  Many laboratory permeability data have been published for soil
that has been ground to pass a 40 mesh sieve. Such sample treat-
ment  drastically improves  the precision of  replicate laboratory
permeability tests conducted on the same soil. One must be alert
for such reports of careful sample pulverization and mixing,  since
the resulting data are good documentation of technician compe-
                                                                                   BARRIERS & WASTE SOLIDIFICATION    217

-------
tence rather than the acceptability of a proposed soil material as
a pollutant barrier.
  Other aspects of sample preparation should be documented as
part of a report of permeability tests. Particle sizes excluded from
the laboratory sample but present in the sample obtained in the
field should be reported in detail, even if departures  are made
from standard ASTM  procedures. Terms such as  "moisturiza-
tion" and "equilibration to wet of optimum  moisture" have ap-
peared in permeability test reports. The degree to which all of the
soil sample to be tested has been brought to  a uniform moisture
content before assembly in the permeability test cell may have a
significant effect on both actual values and precision of replicate
permeability results. Uniformity certainly is greater where the soil
has been pulverized to uniform small fragments and mixed thor-
oughly. Many testing laboratories do not take proper care in size
reduction and mixing because of economic pressures.
  If a client is to understand the  potential limitations of the data
he receives,  he must insist that sample preparation  details be in-
cluded within the permeability report. A  client  also needs to be
provided with a thoroughly documented moisture/density curve
that includes numerous data points.  Where soil material is  pro-
posed to be amended with additives such as  additional clay ma-
terial,  flyash, cement, various  industrial  process  wastes, etc.,
curves should be provided for the actual amended soil.

Soil Physical State as Tested
  The soil physical state as tested for permeability in the labora-
tory depends upon  both sample  preparation  and subsequent
handling.  Application of standard or modified Proctor compac-
tion procedures to cell sizes other than those  accepted by ASTM
methods should be documented by statements describing  the
actual apparatus used. After compaction, a density measurement
calculated from cell volume and weight is the only parameter pre-
sented to  indicate  attainment of the desired compaction.  Geo-
technical engineers accept the fact that soil structural units must
be molded to achieve the most impervious material after compac-
tion. No one has shown how impact compaction using standard
or modified Proctor  procedures achieves remolding  in  clayey
soils. Compaction  "wet of optimum" has been reported to de-
crease permeability in many laboratory studies. Of these reports,
many fail to describe the sample compaction procedure applied.
  Some commercially available laboratory machines are designed
to apply "kneading compaction,"  but few reports of clay liner
permeability have  included  the  use of this apparatus. Where
kneading compaction might be advantageous, studies have not
clearly documented what minimum cell size is needed in relation
to the compaction foot size and soil particle or clod size  to achieve
expected or  acceptable kneading compaction. This also leads one
to question  the utility of some of the smaller sizes of laboratory
permeability cells—for example, those only 1  to 2 in.  in  size.
Where the soil sample has been compacted and later extruded
into a test cell (such as for triaxial confinement) the  apparatus
used should be clearly described.
  Smearing  or polishing the sample end surfaces has been shown
to lower clay permeability. Side polish or smearing caused by ex-
trusion may not have any significant effect, but it may mask ob-
servations of actual small clod structure on  the sample surface.
The client needs to be aware of the potential for such results, par-
ticularly so  he will not be surprised  if later  tests by a different
laboratory yield different results showing that the liner candidate
soil does not meet specifications.
  Several  procedures have been published8'12'11  detailing labora-
tory sample preparation and introduction of prepared samples
into permeability apparatus. These may provide guidelines,  but
since each laboratory is likely to have different apparatus and in-
house procedures, deviations from published techniques should
be documented as part of the permeability test result.

Test Cell Features and Techniques
  The particular features of laboratory test cells and the mode of
conducting a test need to be described. A report that permeability
was determined with a falling head test should not  be accepted.
The inflow and/or outflow permeant liquid level may have been
determined by two different readings on  a burette or graduated
sight tube.  But where hydraulic gradients  of several tens or more
are imposed to reduce testing time, it is likely that the liquid level
change  in  the burette represents an insignificant  head change
compared to the actual gradient imposed  during the test. Thus it
is imperative to include actual  apparatus dimensions and oper-
ating parameters within a test report.
  Statements of  sensitivity,  resolution  and readability of all
gauges,  burettes and digital or analog readout devices should be
available to the data recipient. Raw data parameters may include
such items  as examples of actual burette reading at actual time
and volume increments obtained during  testing of samples re-
ported.  Examples cited should consist of actual raw data, because
laboratory  notebook pages may have to be reproduced for data
quality assurance audit. This information  provides both the client
and the permit reviewer with confidence  that indeed the labora-
tory test was carried out in a professional  manner consistent with
the best known practice of such tests. Confidence in test results
needs to exceed the norm for structural engineering tests because
of the unique social, political and environmental  sensitivity of
data characterizing pollutant barriers.

Test Time Periods and Test Termination
Criteria
  Time  required for obtaining  a soil  permeability test result is
commonly  included  in graphic reports of the test  results. The
hydraulic conductivity is plotted as time  passes during the test.
Time  required to perform a permeability test of compacted soil
materials that are expected to be only slightly pervious is a func-
tion of several operational parameters that can be controlled by
the test  technician to perform tests at an  economical rate. Thus
the question of when to terminate a lest because acceptable values
have been obtained also needs to be clearly answered. A widely
accepted standard for a permeant liquid other than  water (e.g., a
known chemical liquid, solvent or waste stream) is that several
pore volumes of the permeant liquid should be passed through the
sample  to ensure adequate exposure of  the soil sample to the
liquid to cause an observable effect. A minimum of two pore vol-
umes has been used as a criterion by numerous laboratories. Such
tests as this may be more accurately called  liner/leachate compati-
bility tests, although hydraulic conductivity is the primary indica-
tor of compatibility. This case simplifies the question of minimum
testing time.
  A related question concerns the frequency (either on a real time
scale  or in terms of permeant  liquid inflow  and/or  outflow
volume) with which appropriate hydraulic  flow  rate measure-
ments may be validly calculated to indicate an accurate hydraulic
conductivity of the sample being tested.  Clearly, the closer the
apparatus readings, the greater  the chance of obtaining readings
that consecutively agree, thus indicating achievement of a stable
hydraulic flow and therefore indicating that the calculated perme-
ability accurately depicts the sample properties! The frequency of
acquiring the time and volume of data also must be considered
with respect to sample size, hydraulic gradient applied and other
factors. These factors should be included as  part of the test pro-
cedure,  as suggested in this paper. The procedure should describe
the criteria applied to determine the time and volume frequency
 218     BARRIERS & WASTE SOLIDIFICATION

-------
of data collection as well as the criteria used for ending the labor-
atory test. This information should be included with test results
even if decisions were unique to individual samples and not based
on standardized methods. Where a published procedure has been
applied,  any deviations  from the published method should be
noted.

Statistical Data Treatment
  Soil permeability data for materials proposed as clay liners,
clay claps or slurry wall backfills normally are reported as a few
distinct data points. Where a slurry trench cut-off wall is under
construction, data  may originate from daily sample determina-
tions. This is likely to result from batch mixing of trench back-
fill soils.  In  the case of clay soils proposed for liner or cover use,
individual datum statements are more common than summaries
of numerous tests. Thus, the application  of statistically valid and
accepted designs of soil borrow or clay liner test schemes is at an
early stage of development. Some private firms appear more sup-
portive of statistical approaches than others. The client who buys
permeability tests for pollutant barrier materials would be served
best by a clearly described statistical scheme for  sampling  soil,
conducting tests and reporting results.  Such an array of informa-
tion would allow the client to provide this information to review-
ing agencies with a high  degree of confidence that he  has taken
the best known approach to documenting the validity of test re-
sults for barrier material characterization.
   Agricultural literature abounds  with statistical approaches  to
soil testing. Engineering approaches to  assuring that soil struc-
tures are competently designed and constructed should be taken
as the minimum requirement with respect to statistical sampling
and data interpretation. A recent Draft Quality Assurance docu-
ment issues by the U.S. EPA'4 treats  statistical considerations at
some length and should be noted as  directed toward pollutant
containment structures.
   Irrespective of the  site-specific statistical approach applied  to
permeability testing,  the methods used should be included  with
data reports. Data  reports, either graphic or tabular, should use
parameter units that  are standard or well recognized in  ASTM,
ASCE or similar publications. Authors  should carefully assure
that all significant digits in reported data  are actually meaningful.
Where local terms (e.g., tap water, energy gradient, effective con-
fining stress, etc.) are used to test descriptions, they should be de-
fined so  that they are understandable to a broad range of tech-
nical reviewer audiences. Citation of published authorities can
easily account  for  standardized use  of  terms for common pa-
rameters.

Observations About Test Post-Mortems
   Where the hydraulic conductivity of a  compacted clay soil or a
slurry trench backfill  mixture is determined specifically as a com-
patibility test using  a site-specific leachate or other liquid  as
permeant, it is usually fruitful to carefully examine the soil mass
after removal from the  laboratory permeability cell. Although
specific procedures for such examinations or evaluations of ob-
servable  features are  not standardized, many geotechnical engi-
neers apply their own experience and that of consultants to note
any significant sample changes. The testing laboratory should  be
encouraged to  examine  permeated soil  samples and report the
post-mortem observations. Where index tests on the permeated
soil have been applied, these data should be reported. Some engi-
neers have examined  the soil fabric and noticed clear changes in-
duced by non-aqueous permeant liquids.
   Some  laboratories  routinely collect periodic  samples  of out-
flowing permeant liquid and publish (from chemical analyses)
changes in permeant liquid composition as solute breakthrough
curves. Such data provide useful documentation  of the barrier
soil's  attenuation capacity for potential pollutant compounds.
Where slurry wall backfills are being tested for permeability, the
actual slump and moisture content after completion of sample
consolidation and hydraulic testing would be useful to properly
interpret  the  permeability  results.  Whatever   post-mortem
approach is applied, the testing laboratory should be encouraged
to conduct these examinations and report observations for inclu-
sion with permeability data reports.
CONCLUSIONS
  Inclusion of supplementary data with singular expressions of
the hydraulic conductivity of a clay soil liner, clay cap or slurry
wall backfill can only improve the credibility  of  permeability
data. In addition, any data reviewer is  provided with  a better
opportunity to compare data from among sites, soils, laboratories
or test  methods. Since nearly all soil-based  pollutant barriers
represent unique site-specific installations, it is  essential that a
maximum degree of barrier performance comparability be avail-
able to  both owners of multiple sites and agency staffs respon-
sible for reviewing data.  Some of the specific  examples men-
tioned are not directly usable in every site-specific case, and it is
likely that  practicing  design engineers, geotechnical testing  lab-
oratory  staffs  and researchers may have additional factors  and
parameters that need to be considered. This paper is intended to
highlight shortcomings in present data reporting and  encourage
the geotechnical laboratory testing industry  to  incorporate the
results of supportive soil tests into permeability test reports.
ACKNOWLEDGEMENT
  This paper  has been reviewed in accordance  with the U.S.
EPA's peer and administrative review policies and approved for
presentation and publication.
REFERENCES
 1.  U.S. EPA, "Proposed Rules, 40 CFR Part 250 Subpart D, Section
    3004—Standards Applicable to Owners and Operators of Hazardous
    Waste Treatment,  Storage,  and Disposal  Facilities,"  Federal
    Register, 43,  Dec. 18, 1978, 59010.
 2.  U.S. EPA, "Minimum Technology Guidance Document on Double
    Liner Systems for Landfills and Surface Impoundments—Design,
    Construction and Operation," (Draft), EPA/530-SW-014, Office of
    Solid Waste,  U.S. EPA, Washington, DC.
 3.  U.S. EPA, OERR and ORD, Handbook, Remedial Action at Waste
    Disposal Sites (Revised), EPA/625/6-85/006, Oct. 1985.
 4.  Brown, K.W. and Anderson,  D.C., "Effects of Organic Solvents
    on the Permeability of  Clay  Soils,"  EPA  600/2-83-016,  1983,
    NTIS, Springfield, VA, PB-83179978.
 5.  U.S. EPA, Proc. of the Eleventh Annual Research Symposium on
    Land Disposal of Hazardous Waste, EPA/600/9-85/013, Apr. 1985.
 6.  U.S. EPA, Proc. of the Tenth Annual Research  Symposium on
    Land Disposal of Hazardous Waste, EPA-600/9-84-007, Apr. 1984.
 7.  U.S. EPA, Proc. of the  Eighth Annual Research  Symposium on
    Land Disposal of Hazardous Waste, EPA-600/9-82-002, Mar. 1982.
 8.  U.S. EPA,  OSWER, "Lining of Waste Impoundment and Dis-
    posal Facilities," SW-870, Mar. 1983, NTIS PB-81-166365.
 9.  Truesdale, R.S., Goldman, L.J., Cox,  E.G., Peirce, J.J., Witter,
    K. and Peel, T.A., "Laboratory Methods for Testing the Perme-
    ability and Chemical Compatibility of Inorganic Liner  Materials,"
    Draft Report of Contract No.  68-03-3149-10-6 to OSW, U.S.  EPA,
    July 1985.
                                                                                   BARRIERS & WASTE SOLIDIFICATION    219

-------
10. Olsen, H.W., Nichols, R.W. and Rice, T.L., "Low Gradient Perme-      13.  Daniel, D.E., Trautwein, S.J., Boynton, S.S. and Foreman. D.E.,
    ability Measurements in  a Triaxial  System,"  Geotechnique, June          "Permeability Testing With Flexible-Wall Permeameters," Ceoiech.
    1985, 145-157.                                                         Testing J.. ASTM, 7. Sepl. 1984, 113-122.
11. Daniel, D.E., "Predicting Hydraulic Conductivity of Clay Liners,"      14.  U.S. EPA, "Construction Quality Assurance for Hazardous Waste
    J. Geolech. Engi., ASCE. 110, 285-300.                                  Land Disposal Facilities, Public Comment Draft,"  EPA/530-SW-
12. Haji-Djafari, S. and Wright, J.C., Jr., "Determining the Long-Term          85-021, U.S. EPA, Cincinnati, OH.
    Effects  of Interactions  Between  Waste  Permeanls and  Porous
    Media," ASTM STP 805, 1983.
220    BARRIERS & WASTE SOLIDIFICATION

-------
              Quality  Assurance  and Quality  Control  Procedures
                    for Installation  of  Flexible Membrane  Liners

                                             James R. Woods, P.E.
                                       Wehran Engineering Corporation
                                            Grand Island,  New York
                                         Salvatore V. Arlotta, Jr., P.E.
                                       Wehran Engineering Corporation
                                             Middletown, New York
ABSTRACT
  The  significant  number of strengthening regulatory require-
ments governing the design, construction and operation of waste
management facilities has brought about a rapid growth in the
technology employed in the industry. Among these advancements
in technology is the use of flexible membrane liners (FMLs) for
lagoon and landfill lining systems. However, the rapidly growing
popularity of FMLs has resulted in a situation where FMLs are
being specified or used by owners, contractors and engineers with
little or no previous FML installation experience. The purpose of
this paper is to present the fundamental procedures relevant to
controlling the installation of an FML.

INTRODUCTION
  During the  past decade our experience inspecting the installa-
tion of flexible membrane liners has led us to conclude that there
are four primary aspects of FML installation control. As out-
lined below,  two of these primary aspects should be performed
prior to the commencement of any actual installation and prefer-
ably prior  to arrival of the material at the construction site. The
four primary  aspects  of controlling FML installations and the
appropriate times to implement such aspects are as follows:
• Prior to  Commencement of Installation
  -Material Assurance
  -Installation Planning
• During Installation
  -Constant Observation
  -Seam Testing
MATERIAL ASSURANCE
  Among the FMLs commercially available and commonly used
for lining waste management facilities are high density polyethy-
lene (HOPE), chlorosulfonated polyethylene (CSPE), polyvinyl
chloride (PVC) and chlorinated polyethylene  (CPE). The selec-
tion of the proper material for a given application is considered a
design function  and will not be discussed in this paper. This
paper will, however, recommend procedures for attaining a max-
imum quality installation of the specified material.
  One of the most fundamental aspects of providing control over
an FML installation is making certain the material delivered to the
site complies with the specifications. Most FMLs suitable for con-
sideration as liners in waste management facilities are not com-
prised of a single polymer; rather, FMLs consist of a combina-
tion of two or more materials that are produced independently
and shipped to a common manufacturing plant for production of
the specified FML.
  Among materials that may be added to the basic raw material
(polymer) are: reinforcement fabrics (such as nylon or polyester
scrims) to improve tear strength; pigments (such as carbon black)
to provide ultraviolet radiation resistance; and plasticizers to in-
crease flexibility and additives (such as calcium carbonate) to im-
prove surface smoothness characteristics. To further complicate
efforts aimed at assuring proper FML  manufacturing, the basic
raw materials often are specified  via  a generic name (such as
polyethylene) which could include specific polymers (resins) un-
able to yield a specification  complying  product. For this reason,
a proper FML specification always should include a listing of one
or more specific manufactured raw products (i.e., resin  number
and manufacturer)  that would be acceptable for the project. In
addition, the specification should include similar details for any
materials that are to be added to the basic raw product to com-
prise the final FML. Finally,  the specification should include a
list of minimum physical property criteria for the manufactured
FML.
  As stated above,  an  FML is manufactured from one or more
raw materials that are shipped to a manufacturing plant. The first
step in providing control over an installation should be to deter-
mine where the material is going to be manufactured and the spe-
cific raw products intended for use.
  Just as there are many FMLs  to choose from, there are many
FML suppliers and installation contractors in the industry. Many
firms in the business of installing FMLs are also in the business
of manufacturing FMLs. Other installers and/or suppliers pur-
chase the raw products and subcontract  with manufacturing plans
for production  of  the sheeting. Aside from the  fact that the
former case has  some advantage in providing a greater certainty
of warranty, both of the above procedures  can and  have  been
used with success. In any case, the plant identified for manufac-
turing the FML should  be qualified by verifying its experience in
producing the specific FML specified for the project.
  In addition, the manufacturer of each raw product used in the
production of the FML should be similarly identified and qual-
ified. For each material component used during the manufacture
of the FML, it is recommended  that a quality control certificate
be obtained from the manufacturer of the raw product. In addi-
tion, the transport of the approved material  to the manufactur-
ing plant should be verified to  make certain the approved ma-
terials are used in the manufacture of the project.
  Upon completion of  the manufacture, random samples of the
FML material should be tested both by the party providing the
quality control over the installation and  by the manufacturer. The
manufacturer's results  should be submitted to the former party
for comparison and verification. The specifications for the pro-
                                                                              BARRIERS & WASTE SOLIDIFICATION    221

-------
ject should contain a list of minimum physical properties to be
verified.  We  recommend, as a  minimum, that elongation,  ten-
sile strength and thickness always be verified. The frequency of
random sampling would depend upon the method of production
and the volume of FML  produced. Samples should be obtained
from  each representative mode of  production. For  example,
should the FML be manufactured over a several day period with
shutdown during the evenings, samples from each day's produc-
tion should be tested. Conversely, if the material is manufac-
tured continuously over  a multi-day period, samples should be
obtained from each shift of production workers.

INSTALLATION PLANNING
   Perhaps the most crucial component  for  controlling  the  in-
stallation of FMLs is the planning which  is necessary before the
material arrives  on the site. Just as  there are numerous manu-
facturers who provide a varied degree of control, there are num-
erous installers in the business. The qualifications, experience
and knowledge of the firm selected to install the FML must be
verified. As  will  be explained in  detail  below,  there are many
seaming and  testing techniques available and each  of these tech-
niques is highly dependent upon the skills of those using them.
   It is necessary to  look beyond the firm itself and determine
which individual  people  will be performing the  installation. All
seamers should have adequate experience in the type of seam to
be used. A foreman should be identified as a liaison between the
person providing the control and the production crews. The num-
ber of crews, the number of people  in each crew and the min-
imum skills of each person in each crew should  be identified. A
panel diagram should be provided by the installer indicating the
intended  alignment of the panels. The  following items  should
be discussed with the installer:

•  Panel  Alignment and  Seam Location:  Seams should be min-
   imized  and situated at least  5  ft  from toes of embankment
   slopes. Panel width should be optimized.
•  Treatment  of Project Irregularities: Penetrations, sumps, em-
   bankment  abutments, etc., must be addressed. Patching  tech-
   niques  should  be  identified on this diagram. Compensation
   folds for temperature fluctuations should also be discussed and
   agreed upon.
•  Operational Procedures: The direction of installation operation
   giving consideration to slope of subgrade and prevailing winds
   should  be agreed upon. Specific techniques to protect  against
   wind uplift and rainwater intrusion between the FML and sub-
   grade should be identified and  agreed  upon.  The shift hours
   and number of crews should  be determined to establish man-
   power needs for quality assurance and quality control.
•  Treatment  of Anchor Trench: A detail should be provided for
   the location and cross-section of the trench itself. The contrac-
   tor should agree to excavate the trench daily in sufficient length
   for the day's scheduled installation. The trench should be back-
   filled in sections immediately upon  installing each panel of the
   FML in the trench to protect against wind uplift.

   In addition, before approving the installer for the project, a set
of protocols  should  be  established.  This  agreement should in-
clude  an area set aside outside of the work area for smoking. It
should be understood that no smoking will take place during the
act of installing the membrane or upon  sections of completed
membrane. Coordination,  respective responsibilities  and com-
munication protocol  with other on-site  contractors  should be
thoroughly reviewed to avoid costly installation  delays and con-
flicts of interest between  contractors. This agreement should in-
clude  addressing  responsibility for  subgrade preparation  and
maintenance  of subgrade from  the time of acceptance to FML
installation.
  The specific seaming technique should be reviewed as well as
any proposed non-destructive testing to be implemented by the
installer himself.  If the party charged with controlling the in-
stallation is not totally familiar with either the seaming technique
or the non-destructive testing method, an arrangement should be
made—prior to approving the installer for the project—for test
seaming and demonstrations.
  The contractor should be made aware of the quality control
requirements that will be implemented during the installation and
agree to cooperate, be bound by the determinations of the quality
control effort and perform all patching or test seaming that may
be required. Other  aspects of the project which should be re-
viewed prior to any material arriving on  the site include on-site
storage of the  material,  the equipment  that will be used  and
cleaning and overlapping requirements at seams.

CONSTANT OBSERVATION
  The most crucial part  of the quality control process is con-
stant observation  of the installation by a qualified inspector. The
initial responsibility of the inspector is to verify that the material
delivered to the site is the  same material which was previously
approved for use under the material assurance aspect of the qual-
ity control program. This assurance  is made by comparing the
identification slips accompanying the material with the identifica-
tion appearing on  the submitted quality control certificate.
  The inspector should be responsible  for approving the sub-
grade preparations in conjunction with the installer. We recom-
mend that both the installer and the inspector certify the accep-
tability of the subgrade in  terms  of  firmness and texture regu-
larity (smooth and free from sharp objects such as stones).
  The inspector should carefully observe the handling of the ma-
terial, noting if any mechanical grips are used and subsequently
checking for any  damage caused  by  such mechanical grips.  He
should watch during the unfolding or unrolling of the  material
to note any material irregularities or concentrated stressing dur-
ing the handling operation.  He also should carefully observe the
placement of the material to confirm that the specified overlap is
attained and satisfactory cleaning of the FML (or an existing
FML in the case of an extension) is accomplished.
  All seaming should be done only upon the performance and
acceptance of a test seam at the beginning of each crew shift and
immediately following any stoppage of work for lunch, etc. The
test seam should be performed outside of the limits of the work,
but in conditions representative of the work area.  Test seams
should be carefully observed by the inspector and then be checked
by hand or mechanical stressing in tension  and in peel. Should
the test seams prove inferior, no seaming operations should com-
mence until the cause of the inferiority is determined, corrective
action is implemented and acceptable test seaming is conducted.

SEAM TESTING
  All seaming should be observed by  the inspector and tested
using a non-destructive technique  by the installer under  the con-
stant observation  of the inspector. Literally every inch  of seam
should be inspected using  one or more non-destructive tech-
niques.
  There are numerous  non-destructive techniques available  and
suitable depending upon  the type and thickness of the specific
FML. Among the test methods available are:  ultrasonic pulse
echo which utilizes a principle similar to sonar; air lancing which
involves detecting unbonded areas via air jets; vacuum chambers
which utilize a vacuum pump to detect leaks; mechanical point
stressing which involves blunt instrument  probing of the seam to
detect  unbonded  areas; and dual-seam pressurization which is
limited to specific seams that result in a twin bond. Information
is available in the literature  with regard to which non-destructive
222    BARRIERS & WASTE SOLIDIFICATION

-------
techniques are appropriate for specific FMLs of a given thick-
ness and material type.
  As  discussed above, proper installation planning would have
resulted in a prior agreement  with the contractor regarding the
specific nature of the non-destructive technique or techniques to
be employed. The inspector should observe every inch of non-
destructive seam testing and document the testing. Any areas of
failure should be identified and patched in accordance with the
previously agreed upon patching technique.  Any suspect areas
should be patched, cap-stripped (is appropriate for the specific
FML) or identified for subsequent destructive sampling.
  Despite the most rigorous field observation and non-destruc-
tive testing effort,  our experience  has  shown that destructive
sampling  is imperative. Field seaming is a  labor-intensive/de-
pendent operation and, despite the most conscientious efforts on
behalf of the installer and the  inspector,  destructive sampling
sometimes reveals unknown and unexpected occurrences that po-
tentially could impact the quality of the installation.
  Destructive sampling involves  obtaining a sample  along the
length approximately 18 in. square (centered on the seam)  and
cutting out test  specimens for laboratory testing in accordance
with the appropriate ASTM Standards for the specific FML  ma-
terial. We recommend that a field  test be done from each de-
structive sample cut in the field to  determine whether it  is pos-
sible for a seam to pass the field test and not the laboratory test.
Minimum laboratory testing should be peel adhesion testing  and
tensile testing to determine elongation and tensile strength.
  Since all sampling must  result in a patch in  the completed pro-
duct, destruction sampling should be minimized in cases where
destructive sampling has not revealed unexpected causes for con-
cern. We  recommend that  the location  of destructive sampling
be prioritized as follows:
• At all areas identified as suspect, but not otherwise remed-
  iated, during the non-destructive sampling.
• A minimum of one sample from each seamer (both personnel
  and equipment).
• A minimum of one sample from each different seaming con-
  dition (for example, relatively "flat" floor areas, toe areas,
  grade steps, slopes, etc.).
• A minimum of one sample from each representative working
  conditions (for example, extremely hot or cold days, seams
  during or following precipitation events, etc.).
  Should  samples representative of all of the conditions out-
lined above  be obtained and tested yielding satisfactory results,
no  further destructive testing would be warranted. However,
should the test results be unsatisfactory or otherwise reveal  unex-
pected occurrences,  further destructive testing might  be war-
ranted to  identify the problem and determine the  appropriate
remedial measure. The quantity of destructive sampling should
be the  minimum necessary to provide a representative cross-sec-
tion of the various types and locations of seams.

CONCLUSIONS
  If the installation is properly planned and adequate measures
are taken to  assure the proper material is provided, full-time con-
struction observation and seam testing can assure an environmen-
tally sound FML installation.
                                                                                   BARRIERS & WASTE SOLIDIFICATION    223

-------
             Manuscript Withdrawn at the Request
                    of The PQ Corporation
224  BARRIERS* WASTE SOLIDIFICATION

-------
Manuscript Withdrawn at the Request
       of The PQ Corporation
                             BARRIERS & WASTE SOLIDIFICATION  225

-------
              Manuscript Withdrawn at the Request
                     of The PQ Corporation
226  BARRIERS & WASTE SOLIDIFICATION

-------
Manuscript Withdrawn at the Request
       of The PQ Corporation
                             BARRIERS & WASTE SOLIDIFICATION  227

-------
              Manuscript Withdrawn at the Request
                     of The PQ Corporation
228  BARRIERS & WASTE SOLIDIFICATION

-------
                                      Illinois  Plan for  On-Site
                              Incineration  of Hazardous  Waste

                                                  James F. Frank
                                                Robert Kuykendall
                                      Division of Land  Pollution Control
                                   Illinois Environmental Protection Agency
                                                Springfield, Illinois
ABSTRACT
  The Illinois Environmental Protection Agency (IEPA) has been
a leader in promoting the use of alternative treatment technolo-
gies for hazardous waste treatment and disposal.  One of these
technologies is  the Transportable Thermal  Destruction Unit
(TTDU) or mobile incinerator. IEPA has prequalified a list of
companies and put a request for proposal out for bids on the
cleanup of two state-funded hazardous  waste sites utilizing a
TTDU. IEPA also has worked with U.S. EPA in the creation of a
task force to study ways to make mobile incineration more attrac-
tive at NPL sites. The House of Representatives has adopted an
amendment in its version of CERCLA that will make it easier for
Illinois and other states to use mobile incineration at non-NPL
sites under State statutes.

THE ILLINOIS HAZARDOUS WASTE
PROBLEM AND PROGRAM
  Illinois generates 526,227,923 gal of hazardous waste annual-
ly. This amount is second only to the state of New Jersey. Due to
the state's long industrial history, there are hundreds of hazard-
ous waste sites within its borders. There are nearly 1,000 known
or suspected hazardous waste sites in Illinois. IEPA has evaluated
in a preliminary manner 750 of  these sites. The remaining 250
sites will be evaluated in  1986. It is anticipated that a total of at
least 50 of these sites will be  placed on the federal Superfund list
during the next 5 years.  There are currently 22 sites on or pro-
posed for the federal list. Another 50 sites will be placed on the
state cleanup list over the next 5 years including the 19 which are
currently on the list or proposed. In addition, the state is current-
ly conducting cleanups classified as immediate removals at 18 sites
per year. The agency is currently involved in the oversight of 20
voluntary cleanups being financed by responsible parties.
  IEPA is currently  involved in  approximately 80 hazardous
waste  studies, oversight  or  cleanup  activities. Over the next 5
years the agency may be involved in as many as 290 such cleanup
activities. The primary focus of the state-funded  Clean Illinois
program in its first 2 years has been to remove the source of con-
tamination from the sites and secure waste on-site so that it no
longer migrates into the groundwater or leaves the site. This pro-
gram has resulted in the excavation and removal from the sites of
6,510 tons, 10,730 yd3, 16,077 55-gal drums and 194,815 gal of
hazardous waste from state-funded cleanups. These wastes typ-
ically were contaminated  with PCBs, volatile organic solvents and
heavy metals. It is estimated the  combined federal, state, local
and responsible party costs to support the cleanup program in Illi-
nois may exceed $1 billion between now and the year 2000.
IS THERE A BETTER WAY?
  As states go, Illinois is blessed with hazardous waste treatment
facilities including three hazardous waste landfills, two commer-
cial stationary hazardous waste incinerators and numerous aque-
ous waste treaters, solvent  recovery operations and recyclers.
However,  from the spring of 1984 through the spring of 1985,
IEPA began to realize the  tremendous amount of hazardous
waste that was being landfilled as the result of State-funded clean-
ups. The potential problems  of future state liability at these sites
really become evident when manifests signed by agency employ-
ees  as generators are returned to the agency and appear in com-
puterized monthly and  annual reports. To make this problem
more acute, some  of the cleanup contractors were proposing in
the  lowest cost bids to dispose of the waste in out-of-state land-
fills. These developments caused the  agency to take the follow-
ing  steps.
  The agency would begin paying a premium for cleanup plans
that land disposed the least amount of hazardous waste and incin-
erated, recycled or treated  the greatest amount of hazardous
waste. The Illinois General Assembly already had passed a law
prohibiting the disposal of liquid hazardous waste in landfills as
of July 1, 1984 and prohibiting the land disposal of all hazardous
waste by Jan. 1, 1987, unless there is no other alternative to land-
filling. By these actions, the  state was trying to place less depen-
dence on landfilling of hazardous waste and more reliance on
alternative treatment technologies such as incineration.
  Even though the state has  two fixed incinerators, IEPA found
it difficult to obtain cleanup plans with serious incineration al-
ternatives  for thousands of  cubic yards of contaminated soils.
Cleanup contractors found it difficult to obtain firm incineration
prices over time. In addition, the incinerators did not want to
accept large  volumes  of soils, and if they did accept soils they
had to be  packed in 15-gal fiber drums. These drums had to be
hand-filled and sometimes posed leakage problems in transporta-
tion. In addition,  Illinois has a hazardous waste treatment and
disposal siting law which allows local governmental jurisdiction
to veto  new  facilities that are to accept waste outside of their
political jurisdiction. For these reasons the agency began looking
at mobile incineration options.

DOES ANYBODY HAVE A MOBILE
INCINERATOR FOR SALE?
  In March 1984,  IEPA created  an in-house work group which
included representatives from its various organizational divisions
including Air Pollution Control, Land Pollution Control and the
Office of Chemical Safety; legal staff were included. IEPA also
                                                                                                   INCINERATION    229

-------
obtained assistance from the U.S. EPA Region 7 and the Cin-
cinnati based  ORD incineration research group. This panel of
nine people reviewed the equipment of six companies that claimed
they had  or soon would  have mobile incineration capabilities.
The work group was sufficiently encouraged by this review to
identify more interested companies  that  had or could build a
mobile incineration system  for lEPA's use.  By the summer of
1984, the IEPA had, through  the placement of advertisements
and direct mailings, developed  a list  of 78 persons or companies
interested in providing TTDU services to IEPA. The IEPA hosted
a meeting in Springfield in June which 35 companies attended to
discuss a future Request for Statement of Qualifications (RFQ).
In  September 1985,  IEPA put out an RFQ  which had the pre-
amble given in Table 1.
                           Table I
         Preamble lo the IEPA RFQ on a Mobile Incinerator

      Request for Statement of Qualifications for Cleanup Services
           Using a Transportable Thermal Destruction Unit

  It is the intention of Illinois Environmental Protection Agency, Divis-
ion of Land Pollution Control ("1EPA/DLPC" or "agency"), to utilize
Transportable Thermal Destruction Unit (TTDU) technology at a haz-
ardous waste site that is now on the State Remedial Action Priority List
(SRAPL) or National Priority List (NPL).
  IEPA/DLPC is seeking qualified firms  who will agree to  own and
operate a TTDU in conjunction with their hazardous waste cleanup serv-
ices on an Illinois hazardous waste site to be designated by IEPA/DLPC.
Typical contaminants and wastes found at the site will be PCBs at levels
above 50 ppm volatile organic solvents, non-volatile organics, halogen-
ated organics, pesticides, oils and other petroleum products. These con-
taminants will be in several different forms, including liquid wastes,
in soils or in various objects (for example, 55-gal metal drums).
  IEPA/DLPC desires proposing firms to provide the following services:

1. Own and operate a TTDU at the specified cleanup site.
2. Design and build a feed  system that will handle contaminated soil,
   liquids, sludges and various sized metal or fiber drums.
3. Perform all acts necessary  for a hazardous waste cleanup, includ-
   ing excavating, staging, processing and feeding waste material to the
   TTDU.
4. Remove treated wastes from the TTDU for appropriate management
   and/or disposal.
5. Test air emissions and obtain necessary permits.
6. Prepare the delisting petition for treated wastes.
7. Provide all necessary equipment and appurtenances to make system
   functional.
8. Provide all necessary personnel to operate all systems.

  The purpose of this Request for Statement of Qualifications (RFQ) is
to select from the universe of prospective bidders those bidders who can
best  demonstrate to the agency that they possess the financial strength
and  resources, state-of-the-art technology,  management strength, rele-
vant experience and satisfactory plan for financing the  program's capi-
tal expenditures to successfully implement  all  of the services previously
described.
   Twelve firms responded to the RFQ. One was considered non-
 responsive due to incomplete  information.  In  December 1985,
 IEPA put out a Request for Proposal (RFP) covering two sites;
 the Beardstown PCB site and  the Ability  Drum site. Responses
 on these two sites are due by Jan. 27, 1986.  The IEPA has been
 informed that  two of the 11 companies do not wish to bid.  That
 leaves nine firms  on the pre-qualified bidders list  for these two
 sites. Those firms  from whom we anticipate bids arc:
1. Ensco
   Pyrotech Division
   Franklin, Tennessee
2. Sitex Corporation
   Clayton, Missouri
3. Rollins Environmental Services
   Wilmington, Delaware
4. John Zink Co.
   Tulsa, Oklahoma
5. IT Corp.
   Knoxville, Tennessee
6. Roy F. Weston, Inc.
   West Chester, Pennsylvania
7, G.A. Technologies
   San Diego, California
8. Detoxco Inc.
   Walnut Creek, California
9. Mid-America Environment Services, Inc.
   Riverdale, Illinois

BEARDSTOWN PCB SITE
  The Beardstown  site  was a metal salvage yard occupying 10
acres located 3 miles west of the  Illinois River at Beardstown,
Illinois in Mason County. The site, located  on very  fine sandy
soil, had taken in transformers from local utilities and salvaged
the copper from them.  The PCB-contaminated oil from several
hundred transformers was dumped on the ground over the years.
The sandy soil was contaminated to a depth of 10 ft, with the ma-
jority of contamination  occurring in the top 2 ft. An initial clean-
up of the site resulted in the removal of 4,125 yd' of clean scrap
metal, 320 transformer  cases, % capacitors and 120 yd1 of con-
taminated soil. The second phase of the cleanup identified 4,900
yd1 of soil  contaminated at levels exceeding 5 ppm PCBs, the es-
tablished cleanup objective. The third phase of the cleanup will
use the TTDU to incinerate  the 4,900 yd5 of contaminated soil
and treat to a level of 5  ppm  or less. The treated material will be
placed back on the site and will not  be capped.

ABILITY DRUM SITE
  Ability  Drum, located in Tazewell County near Washington,
Illinois, operated from 1964 to 1984 as a metal drum cleaner and
reconditioner. As a byproduct of treating thousands of  drums
from the heavily industrialized Peoria area, hazardous waste was
generated. This  material was spilled and dumped on the plant
property and placed in metal  separation tanks and in a 1 acre la-
goon. The lagoon was drained later and the soil berm was pushed
in on top  of several feet of contaminated  sludge. IEPA removed
over 100 drums of hazardous  waste from the site. Also discovered
at the site were 70 buried drums filled  with hazardous  waste.
These were secured in a building at the site. What remains to be
handled by the TTDU is 12,000 yd1  of contaminated  sludge and
soil and 70 metal drums  and their contents.

SELECTING THE BEST EQUIPMENT
AND COMPANY
   IEPA has created a review  panel to assist in the selection of the
best qualified company to provide  mobile incineration services
along with waste excavation services. In addition to in-house re-
viewers,  the  IEPA  has contracted with  CH2M  Hill  to provide
expert engineering,  advice and technical assistance in  the evalua-
tion and selection of a contractor.  IEPA intends to award a con-
tract to construct,  mobilize,  own  and operate a TTDU  at  both
the Beardstown and Ability sites.  In deciding which company to
select, IEPA intends to look  at the stage of development of their
 230
        INCINERATION

-------
unit, permitting status, mobilization time frames, prior operating
experience, financial strength and cost of treatment. The pay
items in the contract will be tons of soil and sludge treated meas-
ured as dry treated tons after processing and drums and gallons
as inputted to treatment. IEPA hopes to have started treatment
at one site by the end of 1986.

CAN WE USE A TTDU AT A NATIONAL
PRIORITY LIST (NPL) SITE?
  If Illinois is to fully accomplish its goal of the use of TTDUs
at private and  NPL sites, there is a lot more work to be done.
IEPA has taken several actions to gain acceptance of mobile in-
cineration technology at NPL sites.
  In April of 1985, IEPA asked Region 5 U.S. EPA to adopt in a
Federal Record of Decision that the Acme Solvent site use on-
site incineration as the means to detoxify thousands of yards of
soil, sludges and drums. In October  1985, the Regional Adminis-
trator signed a Record of Decision which included incineration.
Since then the  consultant retained by the potentially responsible
parties (PRPs)  at Acme has estimated that on-site incineration of
approximately  12,000 yd3 of Acme waste will cost approximately
$10 million and off-site incineration will cost approximately $30
million. While IEPA does not agree with these on-site cost fig-
ures, the ratio is instructive  of the  cost effectiveness of on-site
incineration.
  On Sept.  12, 1985 the Director of  IEPA, Richard Carlson,
wrote to Lee Thomas,  Administrator of U.S. EPA, to  request
that a task force on the use of TTDUs at hazardous waste sites
be formed. Mr. Carlson's letter reads in part:
  ..."Over the past few years Illinois EPA has been a vig-
orous proponent of alternative treatment technologies.  I
know you share our concern for reducing dependence on
the landfilling  of hazardous waste. Over the past year we
have been  evaluating the  procurement of Transportable
Thermal Destruction Unit (TTDU) that could be utilized at
National Priority List (NPL) sites as well as sites in our
State clean-up  program....My staff and several incinerator
manufacturers have been in contact with U.S. EPA em-
ployees in various regional offices and at headquarters at-
tempting to identify what federal requirements must be met
to operate a TTDU at a hazardous  waste site. Due to the
newness of this issue and the plethora of programs and
regulations that are involved, it has been difficult to obtain
consistent, clear answers to questions about relevant regu-
latory  requirements.  The  regulations  at issue  involve
RCRA, TSCA, CERCLA  and the Clean Air Act.  There-
fore, I am requesting that a task force  be formed by U.S.
EPA to establish a uniform regulatory policy regarding the
use of TTDUs  at cleanup sites....! believe it would be very
helpful in reaching our mutual goal of decontaminating
soils and destroying hazardous waste at NPL sites if there
was a clear understanding of a reasonable regulatory struc-
ture that could be articulated to prospective users. I look
forward to your reply."
  In November 1985, Winston Porter, U.S. EPA Assistant Ad-
ministrator, responded to Mr. Carlson's letter by informing him
that a TTDU task force had been formed and inviting Mr.  Carl-
son to participate. Since that time, staff of IEPA and U.S. EPA
have conversed about the objectives of the task force. An initial
meeting of U.S. EPA  members to the task  force was  held in
December 1985. IEPA  is hopeful and optimistic  that this task
force can evaluate the  many federal programs and regulations
that  impinge on  the  technology  of mobile  incineration  and
develop a simple but effective regulatory structure to speed the
use of this technology at hazardous waste sites.
  On Nov. 20, 1985, final publication of the National Contin-
gency Plan goes part of the way in demonstrating U.S. EPA's
willingness to  consider on-site mobile incineration as an  inno-
vative technology that  may  be applicable at some NPL  sites.
Several wording changes from the original NCP show change in
attitude and policy on U.S. EPA's part. The NCP amendment
also included exemptions from having to obtain federal and state
permits at NPL sites.
  The  State of Illinois  supports  this  provision and  similar
language in the  House and Senate versions  of the CERCLA
amendment, as long as the substantive state standards are con-
sidered and  the state is consulted prior to a final remedy  being
selected.  In  fact, IEPA thought this permit exemption at  NPL
sites  was such  a good idea that we proposed an amendment to
the House Superfund Reauthorization bill that accomplishes the
same thing for states that have done certain things in relationship
to non-NPL sites. At Section 121, (J) (13) (M) Permits For On-
Site Cleanups Under State Authority, the House CERCLA ver-
sion  would exempt State-operated mobile incinerators from ob-
taining a RCRA Subtitle C permit, if certain conditions are met
by the State. We believe by reducing the time normally required
to obtain an RCRA Subtitle C permit, mobile incinerators will
become most cost-effective because they can be kept in more con-
stant service. The State of Illinois intends to meet all applicable
state and federal regulations while operating  its TTDUs  when
exercising the permit waiver.

TTDU USE AT FUTURE NPL SITES
  IEPA  is considering the use of a  TTDU at a portion of the
Waukegan Harbor NPL site. This proposal would involve on-
site dredging, dewatering and incineration of 12,000 yd3 of PCB-
contaminated silt and sand. The treated material would be  land-
filled off-site.
  IEPA is the state lead on the LaSalle Electric Utilities site. This
site manufactured and reconditioned transformers for approxi-
mately 40 years in LaSalle, Illinois. As part of its operations, it
sprayed waste  oils  containing PCBs on  adjacent parking lots,
soils, on city streets and alleys; some PCBs reached  residential
lawns. IEPA is considering proposing the use of a TTDU on the
plant property to decontaminate tens of thousands of yards of
PCB-contaminated soils.

PRIVATE USE OF TTDUs
  A  great many of the hazardous waste sites on and off the NPL
are going to be remediated by responsible parties. IEPA has been
encouraged that PRPs at the Acme Solvents site have put  out a
bid request to  incinerate 12,000 yd3 of sludge, soil and contam-
inated drums on-site. The IEPA also has been encouraged that
Monsanto has  chosen to exhume and incinerate 5,000 drums of
hazardous waste buried on its property. This waste is being incin-
erated off-site  by Rollins Environmental  Services. In addition,
the IEPA has been encouraged by the dozens of inquiries  from
potential TTDU vendors, PRPs and other prospective users in
the private and governmental sector including various agencies of
the Federal government and other states.

CONCLUSIONS
  The handling of  contaminated soil at hazardous waste sites is
one of the most vexing problems to be dealt with in site clean-
ups.  There is a tendency to cap, cover or contain on-site with a
slurry wall or transport off-site for disposal in an RCRA approved
landfill. IEPA believes that in many situations this is a short-term
approach which will not permanently solve the problem. Under
CERCLA, the state assumes  the responsibility for long-term
operation and maintenance. Under the uniform manifest system,
                                                                                                     INCINERATION    231

-------
the state ends up signing  as generator at  many state  cleanups
and at all state-led NPL sites. This is a very disconcerting situa-
tion which the state is attempting to remedy. IEPA believes the
availability of mobile incineration technology is one way to really
cleanup sites more permanently—not merely postpone the prob-
lem until later.
  IEPA is proud of its accomplishments in serving as a catalyst
in promoting the use of mobile incineration  as an alternative
treatment technology whose time has come. IEPA has been  will-
ing to match its conviction  with substantial amounts of state
monies. We do not mind being the "guinea pig" if it serves the
public purpose of furthering a promising technology. Our goal is
to have this technology provided by many different companies in
a competitive market place situation. This will allow the State of
Illinois to have its choice among many qualified providers so
that we are able to landfill less  hazardous waste and incinerate
more hazardous waste as the IEPA strives to achieve more perm-
anent environmental  solutions in our hazardous waste manage-
ment program.
232    INCINERATION

-------
           An  Introduction  to EPA's New Trial Burn Data Book

                                                   M.P. Esposito
                                               PEI Associates, Inc.
                                                  Cincinnati, Ohio
                                                   N.J.  Kulujian
                                    U.S.  Environmental Protection Agency
                               Center  for Environmental Research Information
                                                  Cincinnati, Ohio
ABSTRACT
  A trial burn data book has been prepared by the U.S. EPA's
Center for  Environmental  Research  Information, Technology
Transfer Staff, for use in the permitting and testing of hazardous
waste  incinerators regulated  under RCRA.  Test results from 23
hazardous  waste  trial burns conducted  in the United  States  at
various types of full-scale stationary incinerators are summarized.
In addition to the incinerator trial burn data, the book  also con-
tains results of U.S. EPA Hazardous Waste Engineering Research
Laboratory tests at 11 U.S. lime, cement and aggregate  kilns and
nine industrial boilers. This paper discusses only the data derived
specifically from trial burns at incinerators.

INTRODUCTION
  RCRA requires that hazardous waste incinerators adequately
destroy hazardous organic compounds while maintaining accep-
table levels of particulate and chloride (HC1) emissions. Owners
and operators of these incinerators must demonstrate the accep-
table performance of the facility by means of a trial burn. Conse-
quently, industry and control agency personnel have become in-
volved in planning for, conducting and interpreting the results
from trial burns as an integral part of the RCRA permitting pro-
cess.
  In an effort to assist the RCRA permitting process, the U.  S.
EPA has prepared a trial burn data book as a reference document
for use in the permitting and testing  of  hazardous waste in-
cinerators and other thermal treatment  devices that are now  or
soon may be regulated under RCRA. The  trial burn data book
summarizes the test results from hazardous waste trial burns con-
ducted at 23 full-scale stationary incinerators in the United States.
Eight  of these trial burns were designed and conducted by the
U.S.  EPA  and  its  contractors  as part  of  the U.S.  EPA's
Regulatory Impact Analysis of the RCRA incinerator regulations.
The other 15 were  conducted separately  and individually by
private industrial concerns and their contractors as part of their
Part B application requirements for obtaining full operating per-
mits under RCRA.
  This paper discusses incinerator performance only;  however,
the report also contains data from hazardous waste test burns at
11 lime, cement and aggregate kilns and  nine industrial  boilers in
the United States. The U.S.  EPA conducted these tests as part of
an overall research program aimed at determining the efficiency
of these thermal units for cofiring (and thereby destroying) hazar-
dous wastes as fuel supplements or replacements. The practice of
substituting hazardous wastes for fossil fuels in kilns and boilers
is currently exempt from RCRA regulation, but this situation may
change as more is learned about the overall effect of typical kiln
and boiler operating conditions on various waste-oriented perfor-
mance  parameters,  such  as waste destruction and removal effi-
ciencies, chloride and particulate  emissions and the formation of
unwanted byproducts of combustion.
  This  is the first time that a trial burn data book containing
results  from a wide variety  of combustion tests has been assem-
bled. The authors sincerely hope the document will be a beneficial
source  of information for those involved in the planning, execu-
tion  and evalulation of hazardous waste trial/test burns. This
data book is intended to  be used  in conjunction with other U.S.
EPA guidance documents on hazardous waste incineration.

DATA BOOK CONTENT
  The  trial  burn data  book gives introductory material on the
purpose, scope and intended use of the document. Major types of
incinerators, boilers and  process  kilns now in use in the United
States are discussed, and  the names and addresses of incinerator
manufacturers/vendors are identified. Schematic diagrams are in-
cluded  to help the reader visualize each type of unit. The design
information presented herein  is meant to give the  reader  a
technical overview of these  processes.
  The results of the  trial and test burns conducted at incinerators,
boilers  and  kilns are presented and analyzed. The discussion in-
cludes specifics of the types of units tested, operating conditions
during the tests, emission test results and any notable trends in the
data. Basic  data derived  from  the trial and test burn reports at
each incinerator, boiler or kiln are presented on summary forms
similar  to that shown in Figure 1.


ANALYSIS OF INCINERATOR
PERFORMANCE DATA
  As noted earlier,  this  book contains avai'able trial burn test
data from 23 separate incinerators located throughout the United
States.  These test data were taken from 15 trial burn reports' sub-
mitted  by private concerns to the U.S. EPA over the past several
years as part of their RCRA permitting requirements  and from
the U.S. EPA's most  recent R&D performance study of eight
operating units.2 Most of the trial burns consisted of multiple
tests or runs in which one or more POHCs were monitored under
varying operating conditions (e.g., varying feed concentrations or
rates, temperatures  or residence times). As a result, the trial burn
data discussed and  analyzed in this report include a total of 534
POCH/test combinations. Table 1  presents a broad overview of
the test results.
                                                                                                  INCINERATION
                                                                                                                    233

-------
                    INCINERATOR TRIAL BURN
   Da_te__g_f Trial Burn,
   Run So.   	

   Incinerator  infonnjt

       Type of umf
       Capacity:  	
       Pollution control syilem:
       Residence tin*:
   Tr 1 j 1 _gurn Cond 1 1 1 on*
       Haste feed data
                                 Private/industrial
       Type of waste(s) burned
       Length of bum: __	
       Total amount of waste burned
       Waste feed rate: 	
                  Name
                                        Concentration
       Btu conten
       Ash conten

       Operating
Chlorine content.
Moisture content-
                                Average
               iel used:
       Cftcess air:
       Other:
       Monitoring Kethods:
       POHt's: _______
       Cl:
       Paniculate:
       Other:
          and D«E Retultt
       POHC'l
       Cl. 	
       Paniculate:
       THC- 	
       CO:  	
       Other:
       PIC's:
                           Figure 1
                 Example Data Summary Form
                                                                                                 Table 1
                                                                                    Overview of Incinerator Test Results

Item
Sites
Individual POHC's
POHC/tPSt combinations

Tested
23
57
534
No. of failures
DRE
9*
22
1W
Paniculate
10


HC1
3

-
a All of (hew »H« also had at leas) some POCH ORE* greaicr t
b AtmoM all of ihcwr POHCs also had at least some succeuful (greater than 99 99r* DRE) te«
  runs  Exceptions were dieihyl phthalaie. hexachloro butadiene,  m-dichloroberuxnc and 1.2.4-
  I nchlorobcnzcnc

DRK
  Tables  2  and 3 show more detailed summaries of the POHC-
DRE test data. Table 2 lists the test data by  site, giving informa-
tion on the type of  incinerator,  type of wastes,  temperature,
POHCs and ORE results. The table shows that seven of the eight
units tested by the U.S. EPA experienced periodic DRE failures
(as  well as successes), whereas only one of the 15  privately spon-
sored study units (Site 4) reported a  DRE value  of less than
99.99%. One would expect  the trial burn data submitted to the
U.S. EPA in support of a permit application to indicate DREs of
99.99% or  greater,  but the  operating incinerators tested by the
U.S. EPA also should perform consistently at this  level. One con-
clusion  that can be drawn from this data pattern is that  the DRE
is obtained  during "tweeked"  conditions (i.e., when operating
parameters  are  adjusted  to their  optimum  performance levels)
during trial  burns often does not exist during  extended day-to-day
operations.
  The relationship between incinerator type  and performance as
measured by DRE successes and  failures can be  studied by ex-
amining Table 2. The data show that  four of seven rotary kilns
tested experienced DRE failures. None of the single-chambered
liquid or liquid-gas incinerators had DRE failures, whereas three
of the five double- or triple-chambered  liquid-solid incinerators
did. Although no configuration emerges from the data as a clearly
superior system, those which burn liquids or gases  (not solids)
tend to  get  consistently better  DRE performance  results. This is
consistent with acknowledged difficulties in achieving complete
combustion efficiency when  burning solids.
  Table 3 shows POHC statistics in terms of DRE successes and
failures. Some POHCs such  as carbon tetrachloride, tetrachloro-
ethylene, trichloroethylene, chlorobenzene, toluene, chloroform,
trichloroethane and a few others have been tested more frequent-
ly.  Most tests have  been successful, but there also have been a
significant number of  failures. On average, better  than one of
every five POHC tests has failed. The table clearly demonstrates
that no POHC is guaranteed  to get  4-9's (i.e., 99.99%) DRE.
POHCs  which have been more extensively tested tend to  ex-
perience a 10  to  30% failure rate. For those POHCs for which
only three or four test results are  available,  the true failure rate
probably not known.

Paniculate and Hydrogen Chloride Emissions
  Emissions of paniculate matter and HC1 are limited by 40 CFR
264.343 as follows:

  Paniculate matter —  0.08 gr/dscf* corrected to 7% O2

  HCl — 4 Ib/hr, or an HCI removal efficiency equal to or greater than 99^

•gr/dscf = grains/dry standard cubic fool of gas
234    INCINERATION

-------
                                             Table 2
                                Average DREs by POHC and Test Site
Site
NO.
1

2
T




4



5










6


7


9











10
11

12



13









H

15
16








17
















Sponsor
Private

Private
EPA




Private



EPA










Private


Private


EPA











Prlva'e
Private

Private



EPA









EPA

Private
EPA








EPA
















Type of Incinerator
Rotary kiln with secondary chamber

Vertical cylinder
Single-chamber




Rotary kiln with secondary chamber














Acid regeneration furnace


Rotary kiln with secondary chamber


Rotary kiln with secondary chamber











Single-chamber
Flu1dfzed-bed

Double-chamber



Double-chamber









Single-chamber

Single-chamber
Unknown








Rotary kiln
















Type of Waste
Solid, liquid

Solid, liquid
Liquid




Liquid, solid














Liquid


Solid, liquid

Sol 1d 1 iould
Solid, liquid











Liquid, gas
Sludge, liquid

Solid, liquid



Liquid









Liquid, gas

Liquid, gas
Unknown








Solid, liquid
















Approximate
Temperature,
•f
1880-2030

1620-1830
1160-1240




1800














1830


1800-1860

1830-1910
2640











1700-1730
1310

1800



1930-2050









2090

2350
Unknown








2040-2110
















POHC
1,1,2-Trlchloroethane
Carbon tetrachlorlde
:ormaldehyde
Aniline
>1phenylam1ne
m-Dlnltrobenzene
fanonHrobenzene
Phenylene dlamlne
^hlorobenzene
texachloroethane
Methyl benzene
Tetrachloroethene
Carbon tetrachlorlde
Chloroform
Dlchloro benzene
Hexachlorobenzene
Hexachloroethane
Hexachloroethene
Hexachl orocyc 1 opentadl ene
Pentachloroethane
Tetrachl oroethane
Tetrachloroethene
Trlchloroethane
THchloroethylene
1,1,1-Trlchloroethane
Benzene
Carbon tetrachlorlde
1,1,1-Trlchloroethane
Carbon tetrachlorlde
Tr1 chl orobenzenes
Dlchloromethane
1,1,1-Trlchloroethane
Benzyl chloride
Carbon tetrachlorlde
Chloroform
C1s-d1chlorobutene
Dlchloromethane
Hexachloroe thane
Naphthalene
Tetraehl oroethyl ene
Toluene
Trans-dlchlorobutene
Trlchloroethylene
Formaldehyde
Naphthalene
Phenol
l.l.l-Trlchloroethane
Carbon tetrachlorlde
Tetrachl oroethyl ene
Trlchloroethylene
Benzene
B1s(ethylhexyl)phtha1ate
Butyl benzyl phthalate
Carbon tetrachlorlde
Methyl ethyl ketone
Naphthalene
Phenol
Tetrachl oroethyl ene
Toluene
Trlchloroethylene
Dlchlorodlfluorome thane
Trl chl orofluorome thane
01 chl orofluoroe thane
Butyl benzyl phthalate
Carbon tetrachlorlde
Chloroform
Dlethyl phthalate
Naphthalene
Phenol
Tetrachl oroethylene
Toluene
Trlchloroethylene
1,1,1-Trlchloroethane
1.1,2-Trlchloroethane
2,4-Dlmethylphenol
Aniline
Butyl benzyl phthalate
Carbon tetrachlorlde
Cresol(s)
Dlchloromethane
Methyl ethyl ketone
Methyl pyrldtne
N,N-d1methylacetaw1de
Naphthalene
Phenol
Phthallc anhydride
Tetrachl oroethyl ene
Toluene
Trlchloroethylene
Average
ORE, I
(No. Values
99.9973 (10)
99.9988 (10)
99.993777 (9)
99.999918 (4)
99.999133 (3)
99.99 (1)
99.99991 (1)
99.9984 (3)
99.99916 (5)
99.9958 (5)
99.99856 {5}
99.992 (5)
Man (2\
•JO \t}
99.966 (5)
99.99 (5)
99.99 3
99.99 (6)
99.99 (3)
99.99 (6)
99.981666 (6)
99.99 (3)
99.99 (2)
99.986 (S)
99.99 (1)
99.99 (l)
99.999979 (4)
99.999995 (4)
99.999979 (4J
99.997 (2)
99.9975 (2
99.9935 (2
99 999851 (7)
99i999642 (7)
99.932 (1)
99.999533 (3)
99.99985 (3)
99.990733 (3)
99.999953 (3J
99.999103 (3)
99.99 (3)
98.16(666 3)
99.999486 3)
99.999883 (3
99.999906 (3)
99.99798 (3)
99.996666 (3)
99.998 (3)
99.993333 (3)
99.999992 (4)
99.999957 (4
99.997555 (4)
99.999855 (4)
99.903 (2)
99.995833 2}
99.986666 3)
99.994375 4)
99.991675 4)
99.975333 3)
99.998153 3)
99.9929 (1)
99.96075 (4)
99.98897S (4)
99.99 (2)
99.99985 (2)
99.998142 (7)
99.9687 (3)
99.90636 (5)
99.362 (5)
99.959666 (3)
99.862333 (3
99.981333 (3)
99.975516 5
99.991306 5)
99.9026 (5)
99.999173 (3)
99.999994 3)
99.9992 (3
99.998 (3)
99.998866 (3)
99.996133 3
99.999133 (3)
99.978333 (3)
99.99943 (3)
99.998 (3)
99.999866 (3)
99.993 (3
99.994 (3
99.99 (3)
99.998473 (3)
99.998513 3
99.997676 3)
Ho. of ORE
Values Less
Than 99.991
0
0
0
0
0
0
0
0
0
0
0
1
1
2
0
0
0
0
0
2
0
0
1
0
0
0
0

0
0
0
0
0
1
0
0
1
0
0
0
3
0
0
0
0
0
0
0
0
0
0
0
2
0
1
1
2
3
0
0
4
2
0
0
0
1
4
5
3
3
3
2
2
S
0
0
0
0
0
0
0
3
0
0
0
0
0
0
0
0
0
(Continued)
                                                                                     INCINERATION    235

-------
                                                       Table 2 (continued)
Site
NO.
te
19

?0














i\



21












23




Sponsor
Prlvi'e
Privi'.e

EP«














Prlvl'e



CP»












EP«




Type o' Inclneritor
RoUv kiln with tecondiry clumber
S1r,gle-ch**ber

Doublp-chtmber














ThrM-ch«i»ber



MorUontll cylinder












Double-Ch«nbtr




Type or wtue
Solid, liquid
Liquid

Solid, liquid














Solid, liquid



Liquid, gtt












liquid




ApprttilMte
Tt«p«r«ture,
•r
2120
1620-1710

uio-2080














1600-1900



7.040












1550-1660




POMC
Kl
CMorofora
Kethy1beni.ru
Tltr«chloro«th«nc
1,1,1-TrtcMorMtklM
Nnier,.
Hl(tthylh«iy1)p»lUliUte
Cirbofl tttricMorlte
ClilordlM
Chlorobtnttnt
Chlorofom
01 braaoae thine
DIchloroMtJunt
rfeiichlorobuUdlene
HeiacMoroc/cloptntidtent.
HiOllllUleiW
TetricMorotthyUnt
Tolucnt
Trlchlor«tliylt«t
1.2-OlcfclorobtMtrw
Chlorobeniene
HtiicMorMthtnc
Tetrichloro* thy len*
1 .1, 4-Tr1cMoro6*nitnt
tall In.
Ill(flhylh4iyl)p»t>ullt«
Cirbon tltrtehlorld.
Chlorobvnien*
CMorowtlUM
Chloroplxtiyl lucyuutt
•-Dlchlorobtfiltfl*
o-OI chl orobciuent
p-DlchlorobtRUM
Phtnjrt liocyifuU
PSoiotne
TrlcMonxtkyltM
Cirbox Utrichlorldt
Chlorobtaztftt
DIchloroBttKinc
Tolutiw
Trlchlorocthyltn*
>*tri«
OM."
(Do. Hiluei)
n.t*t;«2 (4)
M.MMt (]
M.MU43
W.WWM
1,
1)
M.IMi (1)
M.MMM (1)
N.t»7S (4)
M.W7UI
M.iHIU
M.MUI7
M.4S5S (1
N.M1S01
M.7M! (S
1
3)
1)

1)

»«.»• (1)
W.W4(4)
M.tMU)
H.H042I (1)
M.M«716
M.MSIM
1
I)
M.SM704 (U)
«f.*N3««
M.MMM
"I
11}
M.M471 (12)
M.3U113 (3)
M.M2IM (3)
»».»7 (3)
N.IM1M
tt.WBS (t
M.M71 ()
n.*ni It
M.tlMM
«.M7 (J)
N.*f7tM
M.WM13
M.XS7S
M.*W*2
3)



3)

3)
3)
2j
1)
W.M1327 (4)
M.MMi (4)
M.fM (1)
M.M30S (4)
M.mS (4)
«o. of roue.
Vllutt L«>l
Tlun M.M
0
0
0

t
3
4
0
0
7
1
4
8
1
0
0
7
0
1
0
0
0
0
3
1
3
0
1
0
0
3
0
0
0
0
0
1
0
1
1
1
Although these emissions are generally a function of the ash and
chloride contents of the  waste,  the outlet  concentration  also
depends on  the  exhaust  gas  control system.  Because  control
systems varied from site to  site, correlating  the paniculate and
HC1 emissions with input concentrations is impossible. Although
the available data do not permit the development of such a rela-
tionship, they do indicate that, in general, the emission limits are
achievable.
  Not all of the reports were complete enough to determine with
certainty whether both the HC1 emission limits were met, but the
data indicate that HC1 emissions exceeded 4 Ib/hr at only two sites
(Sites 5 and 13). One site (No. 21) reported HC1 removal efficien-
cies  of less than  99%, but  the information  was insufficient to
determine whether emissions from this site were within the quan-
titative limit of 4 Ib/hr.
  Ten sites (representing both  the privately sponsored trial burn
group and the U.S. EPA's recent eight-plant study) experienced
periodic problems in limiting  paniculate emissions to the 0.08
gr/dscf required by Federal  Regulation 40  CFR 264.343. Six of
the sites in the eight-plant U.S. EPA study exceeded the 0.08
gr/dscf (corrected to 7% O?) during one or more of the test runs.
Four sites (Sites 4, 5, 13 and 16) were particularly deficient  in con-
trol of paniculate matter.
REFERENCES
 1. Part B Trial Burn Reports submitted to U.S. EPA Regions 2, 3, 4, 5,
    6 and 7, 1983-85.
 2. Trenholm, A., et a/., "Performance Evaluation of Full-Scale Haz-
    ardous  Waste Incinerators," Volumes I through V, Midwest Re-
    search Institute, U.S. EPA Contract No. 68-02-3177, 1985.
 3. Branscome, M., et a/., "Evaluation of Waste Combustion in Dry-
   Process Cement Kiln at  Lone Star Industries,  Oglesby, Illinois,"
   prepared for U.S. EPA  by Research Triangle Institute and Engi-
   neering Science under Contract No. 68-02-3149.  1984.
 4. Branscome, M., "Summary Report on Hazardous Waste Combus-
   tion in Calcining Kilns," prepared for U.S. EPA, Cincinnati, OH,
   by Research Triangle Institute, 1985.
 5. Day, D.R. and Cox. L.A., "Evaluation of Hazardous Waste Incin-
   eration in an Aggregate Kiln: Florida Solite Corporation," prepared
   for U.S. EPA by  Monsanto Research Corporation under Contract
   No. 68-03-3025. 1984.
 6. Day, D.R. and Cox, L.A., "Evaluation of Hazardous Waste Incin-
   eration in a  Lime Kiln:  Rockwell Lime  Company," prepared for
   U.S. EPA by Monsanto Research Corporation under Contract No.
   68-03-3025. June 1984.
 7. Higgins, G.M., Helmstetter, A.J., "Evaluation of Hazardous Waste
   Incineration  in a  Dry Process Cement Kiln, In:  Incineration and
   Treatment of Hazardous Waste," Proc.  of the  Eighth Annual Re-
   search Symposium, EPA-600-9-83-003, Mar. 1982.
 8. Jenkins, A.C., et  al., "Supplemental Fuels Project, General Port-
   land,  Inc., Los  Robles Cement Plant,"  Report  C-82-080. State of
   California Air Resources Board,  1982.
 9. MacDonald,  L.P., et al., "Burning Waste Chlorinated Hydrocar-
   bons in a Cement  Kiln," Water Pollution Control Directorate, En-
   vironmental Protection Service, Fisheries  and Environment Canada,
   Report No. EPS 4-WP-77-2, 1977.
10. PEI Associates, Inc., "Guidance Manual for Co-Firing Hazardous
   Wastes in Cement  and Lime Kilns" (Draft), prepared for U.S. EPA
   under Contract No. 68-02-3995, 1985.
II. Peters, J.A., et al., "Evaluation of Hazardous  Waste Incineration
   in Cement Kilns at San Juan Cement Company," prepared for U.S.
   EPA by Monsanto Research Corporation  under Contract No 68-03-
   3025,  Aug. 1983.
236    INCINERATION

-------
                             Table 3
    ORE Successes and Failures Arranged Alphabetically by POHC
                        Table 3 (continued)
POHC
Anil ine
Benzene
Benzyl chloride
Bisfethylhexyl )phthalate
Bromodi chl orome thane
Butyl benzyl phthalate
Carbon tetrachloride
Chlordane
Chlorobenzene
Chloroform
Chloromethane
Chlorophenyl isocyanate
Cis-dichlorobutene
Cresol (s)
Dibromome thane
Di chl orodifl uoromethane
DicMorobenzene
1,2-Di chlorobenzene
Dichlorofluoroe thane
Dichloromethane
Diethyl phthalate
2, 4-Di methyl phenol
Diphenylamine
Formaldehyde
Hexachlorobenzene
Hexachlorobutadiene
Hexachloroethane
Hexachloroethene
Hexachl orocycl opentadi ene
No. of
Tests
10
14
3
10
2
9
62
3
31
24
3
1
3
3
8
2
3
12
7
22
3
3
3
12
6
1
23
6
10
No. of
Successes
(>99.99%)
9
9
3
3
1
7
54
3
22
10
3
1
3
3
4
2
3
12
7
10
0
3
3
12
6
0
23
6
8
No. of
Failures
1
5
0
7
1
2
8
0
9
14
0
0
0
0
4
0
0
0
0
12
3
0
0
0
0
1
0
0
2
(continued)
POHC
m-Dichlorobenzene
m-Di nitrobenzene
Methyl ethyl ketone
Methyl pyridine
Methyl benzene
Mononitrobenzene
N.N-Dimethylacetamide
Naphthalene
o-Dichlorobenzene
p-Dichlorobenzene
PCB
Pen tachloroe thane
Phenol
Phenol isocyanate
Phenylene diamine
Phosgene
Phthalic anhydride
Tetrachloroethane
Tetrachloroethylene
Toluene
Trans -dichlorobutene
Tri chl orofl uoromethane
1,2,4-Trichlorobenzene
Trichlorobenzenes
Trichloroethane
1,1,1-Trichloroethane
1,1,2-Trichloroe thane
Trichloroethylene
Totals
No. of
Tests
3
1
7
3
8
1
3
16
3
3
4
3
12
3
3
2
2
2
48
27
3
2
3
2
1
22
13
35
534
No. of
Successes
(>99.99«)
0
1
5
3
8
1
3
13
3
3
H
3
9
3
3
2
2
2
37
20
3
2
0
2
1
13
13
26
415
No. of
Fai lures
(c99.99%)
3
0
Z
0
0
0
0
3
0
0
0
0
3
0
0
0
0
0
11
7
0
0
3
0
0
9
0
9
119
12. Research Triangle Institute and Engineering Science (RTI and ES),
   "Evaluation of Waste Combustion in Cement Kilns at General Port-
   land, Inc., Paulding, Ohio," prepared for U.S. EPA under Contract
   No.  68-02-3149, Mar. 1984.
13. Weitzman,  L., "Cement Kilns at Hazardous Waste Incinerators,"
   Environ. Prog., 2, (1983), 10-14.
14. Wyss, A.W., Castaldini, C.  and Murray, M.M., "Field Evaluation
   of Resource Recovery of Hazardous Wastes,"  prepared for  U.S.
   EPA by Acurex Corporation under Contract No. 68-02-3176, 1984.
15. Castaldini, C., Mason, H.B. and DeRosier, R. J., "Field Tests of In-
   dustrial Boilers Cofiring Hazardous Wastes," In:  Proc. from the
   Tenth Annual Research Symposium, EPA-600/9-84-022. U.S. EPA,
   Industrial Environmental  Research  Laboratory, Cincinnati,  OH,
   1984.
16. Castaldini, C., Unnash, S. and Mason, H.B., "Engineering Assess-
   ment Report   Hazardous Waste Cofiring in Industrial Boilers,"
   prepared by Acurex Corporation for U.S. EPA under Contract No.
   68-02-3188, Cincinnati, OH,  1984.
                                                                                                                INCINERATION    237

-------
                         Low Temperature Thermal  Stripping
                                     of Volatile Compounds

                                           John W. Noland, P.E.
                                             Nancy P. McDevitt
                                             Roy F. Weston, Inc.
                                         West Chester, Pennsylvania
                                             Donna L. Koltuniak
                           U.S. Army Toxic and Hazardous Materials Agency
                                   Aberdeen  Proving Ground, Maryland
ABSTRACT
  The U.S. Army Toxic and  Hazardous Materials  Agency
(USATHAMA),  located  in the Edgewood area of Aberdeen
Proving Ground, Maryland, has Army responsibility for installa-
tion, restoration and demilitarization of unserviceable or obsolete
chemical agent munitions. It also serves as the lead agency within
the U.S. Army Materiel Command (AMC) for pollution abate-
ment and environmental control technology development. In this
role, USATHAMA routinely conducts generic research and
development (RAD) studies  with wide application to current
U.S. Army environmental problems. The low temperature ther-
mal stripping of volatile organic compounds from soil is an ex-
ample of one of the many successful R&D efforts USATHAMA
has conducted.

INTRODUCTION
  Contamination of soils from past  operations involving vola-
tile organic compounds (VOCs) has become one of the predom-
inant environmental concerns at several U.S. Army installations.
Trichloroethylenc (TCE)  is the most  frequently found contam-
inant; however,  other VOCs such  as  dichloroethylene,  tctra-
chloroethylene and xylene also have  been found. If allowed to
remain in the soil, these compounds can migrate to contaminate
underlying ground waters.
  The U.S. Army Toxic  and Hazardous Materials Agency cur-
rently is developing a number of technologies to treat VOC-con-
taminated soils. One technology is low temperature thermal strip-
ping. This innovative process utilizes a heated screw conveyor to
volatilize the organics from the soil and an afterburner  for
thermal destruction of the volatilized organics. Bench-scale stud-
ies of the thermal stripping concept identified the method as a
feasible treatment process. A pilot-scale field demonstration of
the process, using actual VOC-contaminated soils,  became the
next phase in the development of the technology.
  In May 1985, USATHAMA contracted Roy F. Weston, Inc.
to conduct a pilot test of the low temperature thermal stripping
technology using soils from an actual VOC-contaminated site.
Weston's responsibilities fro the project included:
  Design of the pilot system
  Preparation of test and safety plans
  Environmental permitting
  Equipment selection
  Equipment installation and startup
  Performance of the test program
  Demobilization and site closure
  Preparation of a technical report
OBJECTIVES
  The primary objective of the pilot investigation was to deter-
mine the feasibility of low temperature thermal stripping as • re-
medial action technology for VOC-contaminated soils. Secondary
objectives included:
• Detennination of the impact on system performance of varying
  design parameters (e.g., operating temperature, residence time,
  air temperature, etc.)
• Detennination of the optimum range of operational parameters
  for the pilot system
• Identification of full-scale design criteria
• Evaluation of the off-gas pollutant levels to determine m
  pollution control devices required for full-scale implementation
  of the process
• Preparation of engineering cost estimates for application of the
  technology
• Identification of any future research needs for system optimiza-
  tion

PROJECT DESCRIPTION
  The pilot-scale  thermal processing equipment  (Fig.  1)  was
transportable and was delivered to the site on a single flatbed
truck. The entire system was '"T**11**1 and fully operational within
10 days after delivery.
                        Figure 1
           Installed Pilot Thermal Processing System
238   INCINERATION

-------
  During the 28 days of rigorous testing, more than 15,000 Ib of
soil were processed. The four primary elements of the testing pro-
gram were excavation and materials handling,  thermal process-
ing, sampling and analysis.

Excavation and Materials Handling
  The VOC-contaminated soil was excavated by backhoe from a
site  near the processing  area (Fig. 2). Excavation operations be-
gan early each morning of the testing period. As excavation pro-
gressed, a portable gas chromatograph was used to identify con-
taminated soils for the test. VOC levels in the soils were as high
as 20,000 ppm. Soils chosen for processing were placed in closed
containers to minimize release of contaminants prior to process-
ing. The soils typically  were processed within  several hours of
excavation.

Thermal Processing
  The heart of the treatment  system  is the thermal processor
which heats the soils sufficiently to volatilize the organics (Fig. 3).
Once volatilized, the organics are destroyed in the afterburner.
                       Figure 2
Typical Excavation Operations of VOC-Contaminated Soil
     HOT OIL
   RESERVOIR
                                                             AIR CONTAINING
                                                             STRIPPED VOC'S
                                                               COMBUSTION AIR
                                                                   BLOWER
                            Figure 3
      Schematic Illustration of the Low Temperature Thermal
                        Stripping System
                                                                                                                       Soil feed hopper. Rotary
                                                                                                                       valve design provided positive
                                                                                                                       feed control and ensured air-
                                                                                                                       tight operation.
                                                                                                                       Thermal Processor. Internal
                                                                                                                       view of Holo-Flitc9 Processor
                                                                                                                       leased from the Joy Manufac-
                                                                                                                       turing Company.
                                       Holo-FIite® Screw. Hot oil
                                       entered the first flight of each
                                       screw conveyor, traveled the
                                       full length of each .screw, then
                                       returned back through the
                                       center of each shaft continu-
                                       ously throughout system
                                       operation.

                                       Trough jacket. Trough walls
                                       provided additional heat trans-
                                       fer surfaces. The hot oil moved
                                       down and back the length of
                                       the jacketed trough.
                                                                                                                       Oil heating system. Electric
                                                                                                                       resistance oil heater provided
                                                                                                                       infinite control of oil inlet
                                                                                                                       temperature up to iOO'C.
                                                                                                                       Off-gas emission monitor-
                                                                                                                       ing. Continuous emission
                                                                                                                       monitoring of VOC concentra-
                                                                                                                       tions was conducted for each
                                                                                                                       of the three off-gas lines.
                                       Afterburner. The air con-
                                       taining stripped VOC's was
                                       combusted in an afterburner
                                       which provided a residence
                                       time of greater than two
                                       seconds at a temperature of
                                       IOOO°C to ensure complete
                                       destruction of the organics.

                                       Stack testing. Stack testing
                                       was conducted to determine
                                       atmospheric emission rate, if
                                       any, of VOC's, panieulates, and
                                       HCI.
                                                                                                                    INCINERATION    239

-------
  An indirect heat transfer fluid, in this case oil, was used to heat
the thermal processor. The hot oil was not contaminated by the
soil and was recycled to obtain maximum thermal efficiency.
  The soil was conveyed from the feed end of the thermal pro-
cessor to the discharge end by the twin screws. Both the screws
and the screw conveyor trough were heated by the circulating hot
oil.
  The continuous movement of the screws conveyed and thor-
oughly mixed the  VOC-contaminated soils. This intermixing
action caused break-up of soil lumps and allowed soil particles to
come into frequent contact with the surfaces containing the circu-
lating hot oil, providing good heat transfer and soil/air interface.
  For the pilot-scale system,  the processed soil discharged into a
55-gal drum for containment and storage.
  The air entering the thermal processor was either ambient or
heated, depending on test conditions. The air containing the vol-
atilized organics was sent to the afterburner.
  The afterburner provided a residence time of greater than 2 sec
at a temperature of 1,000 °C to ensure complete destruction of the
organics. The  VOC-contaminated air served  as the combustion
air  for the burner flame. Therefore, the VOCs were exposed to
the high temperatures and turbulence within the flame vortex,
which ensured  complete thermal destruction of the organics.

Sampling
  The field demonstration program determined the VOC con-
taminant removal efficiency under several varied operating con-
ditions, including:
• Soil discharge temperatures (50 °C to  150 °Q
• Soil residence times (30 to 90 min)
• Air inlet temperatures (ambient to 90 °Q
• Heating oil conditions (100 °C to 300 °Q
  Samples were collected for analysis  from soil and air. Figure
4 provides a thermal processor instrumentation and sampling/
analysis diagram.
  Soil Sampling—Composite soil samples were collected for each
test run at these locations:
• The excavation site
• Prior to feed into the thermal processor
• After discharge from the thermal processor
           VOC («xi-ninlKm.
                                              VtK LufHrnirjiiiin*
                                                •i. MniMun
                                                TiHjl
          VOC (jKKi-mrjIHXB       Hroxiri iiNnimrni
            •I.MiiiKuri-
  This procedure allowed: (1)  the estimation of contaminants
violated during  soil excavation and handling operations; and
(2) determination of the VOC removal efficiency of the thermal
processor.
  Air Sampling—Air VOC concentrations and corresponding re-
moval efficiencies were determined at three separate locations:
ambient air entering the  thermal processor; off-gas leaving the
thermal processor; and stack gas downstream of the afterburner.
  In addition, paniculate and HC1 emissions were sampled in the
afterburner stack to determine  air pollution control devices re-
quired for full-scale implementation of the process.

Analysis
  A portable field gas chromatograph (GQ was used to provide
continuous real-time total VOC emission monitoring in each of
the three air discharge lines from the thermal processor  during
actual testing. A mobile  gas chromatograph/mass spectrometer
(GC/MS) also was utilized at the site during I week of testing to
provide additional qualitative and quantitative VOC data at the
same locations. All  soil and air samples collected during  testing
were analyzed using a GC/MS (Fig. 5). These analyses  provided
complete and accurate identification  of all  VOCs present in the
samples.
  The testing program consisted of 28 test runs in a matrix format
designed to facilitate statistical  analysis  of  the results.  Analysis
of the test data provided important  information  regarding opti-
mization of system  performance and development of  full-scale
design criteria.
                          Figure 4
 Thermal Processor Instrumentation and Sampling/Analysis Diagram
                                                                                           Figure 5
                                                                                   GC/MS Analyses of Samples
RESULTS
  Eighteen days of formal testing were completed within 22 con-
secutive calendar days. During this period, more than 10,000 Ib
of contaminated soils were processed.
  Upon completion of formal testing, 10 additional days of test-
ing were conducted to optimize system performance. During this
period, more than 5,000 Ib of contaminated soils were processed.
  A comparison of the VOCs measured in the feed soil to VOCs
measured  in the processed soil and stack gas yielded the follow-
ing destruction and removal efficiencies (Table 1):
240    INCINERATION

-------
                       Table 1
 Summary of Soil VOC Concentrations and Maximum VOC
                 Removal Efficiencies
         Summary of Soil VOC Concentrations and
           Maximum VOC Removal Efficiencies

                   Feed Soil Concentrations    Maximum
VOC                  Average    Maximum     Removal
                     (ppm)      (ppm)      Efficiency
Dichloroethylene
Trichloroethylene
Tetrachloroethylene
Xylene'
Other VOC's
Total VOC's
83
1,673
429
64
14
2,263
470
19,000
2,500
380
88
22,438
100%
100%
100%
100%
100%
> 99.99%
'Xylene is not classified as a VOC since Its boiling point is approximately
 140"C. However, it was Included In this study to evaluate the effectiveness
 of this technology on higher boiling point semi-volatile compounds.
• Greater  than 99.99% removal of VOCs from the soil was
  attained
• No VOCs were detected in the stack gas, indicating a destruc-
  tion and removal efficiency (DRE) for  the overall system  of
  100%
  Stack emissions were in compliance with all Federal and state
regulations (including VOCs, HC1, CO and particulates).
  After regulatory approval, the processed soils were disposed of
on-site as backfill.

CONCLUSIONS
  Based on the unqualified success  of this field demonstration
program, USATHAMA will proceed with the following phases  of
low temperature thermal processing technology development:
• Develop optimized design and performance criteria for full-
  scale implementation of the  low temperature thermal stripping
  process
• Develop life-cycle costs for comparison of the thermal stripping
  process  to alternative treatment/disposal options applicable  to
  VOC-contaminated soils
• Prepare  detailed thermal processing equipment bid specifica-
  tions for full-scale applications of the technology
                                                                                                    INCINERATION    241

-------
           Comparisons Between  Fluidized  Bed and Rotary Kiln
       Incinerators  for Decontamination of PCB  Soils/Sediments
                                        at CERCLA Sites

                                              Henry Munoz
                                     U.S. Army Corps of Engineers
                                          Kansas City,  Missouri
                                        Frank  L. Cross, Jr., P.E.
                                        Joseph L.  Tessitore, P.E.
                                   Cross/Tessitore & Associates, P. A.
                                             Orlando, Florida
ABSTRACT
  Currently, there is a tremendous number of sites at which con-
taminated soils will need to be detoxified by on-site incineration
or taken for off-site disposal in a secure landfill.
  A total of 787 hazardous waste sites are currently on the U.S.
EPA National Priorities List, designated and proposed. On Mar.
26, 1985, 26 new sites were recommended for addition to the list.'
  In  October 1984, the U.S. EPA estimated that,  nationwide,
there were 54,644 landfills and 180,000 industrial park lagoons
containing potentially dangerous amounts of hazardous waste.
Some 18,000 of these sites presently are listed as posing potential
health risks to the surrounding populations.'
  The U.S. Congress, Office of Technology Assessment, esti-
mates that at least 10,000 hazardous waste sites nwill have to be
cleaned up over the next 50 years, involving over $100 billion in
Federal funding.1
  Since 1950, possibly 6 billion tons of toxic waste have been
disposed of in or on the land, creating a tremendous residual
backlog in waste management. Incremental tons of contaminated
soil, which have leached contaminants from hazardous landfills,
have been created over the years.
  Presently, the Corps of Engineers and the U.S. EPA are very
interested in  on-site incineration for the detoxification of con-
taminated soil, so that  the decontaminated material can be used
for  site reclamation. Since landfill costs  range from $200 to
$400/ton and incineration costs for on-site cleanup may range
from $400 to $600/ton, incineration may appear more expen-
sive.2'3 However, considering the liability involved with landfilling
and any possible leakage, costs may exceed Sl.OOO/ton over a
period of 10-20 years. Considering the Liability situation, on-site
incineration may  look more attractive. In addition, costs for
transportation are reduced, and the costs of fill materials for
                                                         Ash Handling
                                                           Svstem
                                                                     Durao Stack
            Drum Feed Conveyor
            (Solids, Sludges,etc.
                                  Incinerator
                                           Burner
                                        Tor Liquids
                              Boiler  Water
                              Treatment
                                                                            Waste  Heat
                                                                             Boiler
                                                Stack  Fan
           Scrubber/Packed Tower
                   Quencher
                Fuel  Storage
                                                   Scrubber
                                                 Water Treatment
                                                      System
                    NOT TO SCALE
                                                   Figure 1
                                    Integrated Hazardous Waste Incineration System
242    INCINERATION

-------
reclamation are practically eliminated since incinerated soils may
be available for reclamation.
  More importantly,  under U.S.  EPA regulations, release  of
PCB-contaminated material into the environment at concentra-
tions of or exceeding 500 ppm is considered improper PCB dis-
posal. The U.S. EPA requires that PCBs in these concentrations
be cleaned up and disposed of properly. Thus, this requirement
eliminates landfilling as an ultimate disposal option and limits
cleanup to decontamination by incineration.

INTRODUCTION
  Traditionally, rotary kiln incinerators have been used at com-
mercial facilities for the destruction of hazardous wastes. The
once emerging technology of fluidized bed incineration is now a
demonstrated technology and is being considered for the destruc-
tion of hazardous wastes and the decontamination of soils/sedi-
ments at industrial and CERCLA sites.
  For  either type of incinerator system, a systematic approach
should be followed and an integrated system should be developed
for the excavation, storage,  mixing, combustion and ash disposal
as shown in the schematic in Figure 1.
  Assuming that the system basically would be the same for any
given site, the focus of this paper should be on the  combustion
unit (Fluidized Bed Incinerator and Rotary Kiln Incinerator).

ROTARY KILN UNIT
  The  Rotary Kiln Unit has the potential for handling a wide
variety of solid wastes including sludges, soils and slurries. The
system also can handle liquid and gaseous wastes if required.
  The  Rotary Kiln Unit consists of a horizontal cylinder which
is lined with refractory; this cylinder rotates about its horizontal
axis with  its angle to the horizontal being normally less than 2 or
3%. As the kiln shown in Figure 2 rotates, it continually ex-
poses the waste material surfaces to the heat and  oxygen in the
flowing gas. The speed of rotation is variable, normally being in
the range of 0.25-1.5 revs/min. The peripheral speed  of the kiln
(outer surface) normally is within the range of 1 to 5 ft/min.
  Another unique property of the Rotary Kiln Unit is the differ-
ence in retention time between solid materials within the kiln and
combustion gases traversing the kiln. The retention  time of the
solids within the kiln is a function of the kiln speed, the rake and
the physical parameters of the unit; the combustion gas reten-
tion time is based on temperature, excess air utilization and kiln
diameter.
  In the system shown in Figure 2, the material is ram fed into
the kiln. As the kiln rotates, the waste burns to ash which is dis-
charged to an ash bin. Burners can be  mounted in either end of
the kiln to provide wither startup  or supplementary heat as re-
quired.
  Both liquids which will burn in suspension and gases can be in-
jected into the kiln from the front or  rear face, although these
nozzles  normally are placed at the kiln entrance. Liquid and
gaseous waste also can be injected  into the afterburner for their
destruction.
  Sludges, slurries  and solids are  dropped on the kiln hearth.
For some materials, the rake is set at zero or at a negative value to
increase retention time.
  The  main advantages and disadvantages of the Rotary  Kiln
Unit are summarized below.

Advantages of Rotary Kilns
• Incinerates a wide variety of liquid and solid wastes
• Incinerates materials passing through a melt phase
• Capable of receiving liquids and solids  independently or in
  combination
• Capable of accepting drums and bulk containers
• Adaptable to a wide variety of feed mechanism designs
• Characterized by high turbulence and air  exposure of solid
  wastes
• Has continuous ash removal which does not interfere with the
  waste oxidation
• Has no  moving parts inside the kiln (except when  chains  are
  added)
                                                                                        MAIN STACK
                         ASH/ NON-COMEUSTIBLE
                               DISPOSAL
                                                                                                    SLUCC-c
 GRINDER
X
                                                                               SCRUBBER          "BLEED
                                                                                 WATER          DISCHARGE
                                                                              TREATMENT SYSTEM
                                                LIQUID WASTE STORAGE
                                                          Figure 2
                                         Hazardous Waste Rotary Kiln Incineration System
                                                                                                      INCINERATION    243

-------
• Adaptable for use with a wet gas scrubbing system
• Retention or residence  time  of the  non-volatile  components
  can be controlled by adjusting the rotational speed
• Waste can be fed  directly into the kiln without any prepara-
  tion such as preheating, mixing, etc.
• Can be operated at temperatures in excess of 2,500 °F (1,400 °C),
  making them well suited for the destruction  of toxic com-
  pounds that are difficult to thermally degrade.

Disadvantages of Rotary Kilns
• High capital cost
• Operating care  is  necessary to prevent refractory damage;
  thermal shock is a particularly damaging event
• Airborne particles may be carried out of the kiln  before com-
  plete combustion
• Spherical or cylindrical  items (such as disposable  drums) may
  roll through the kiln before complete combustion of materials
• Frequently requires additional makeup air  to make up for air
  leakage through the kiln end seals
• Drying  or ignition grates, if used prior to  the kiln, can cause
  problems with melt plugging of grates and grate mechanisms
• High paniculate loadings
• Relatively low thermal efficiency

FLUIDIZED BED INCINERATOR
  The Fluidized Bed Incinerator consists of the fixed bed,  bub-
bling bed  and circulating fluid bed. The units  which  are presently
being used for soils decontamination are of the circulating  fluid
bed type, as shown  in Figure 3. The circulating fluid unit con-
sists of the combustion vessel with a refractory lining, a hot cy-
clone, a heat exchange device and a particulate and gaseous emis-
sion control system. The combustion chamber or vessel also in-
cludes a preheat burner assembly.
  The Fluidized Bed Incinerator operating in the circulating bed
mode operates at higher  super surficial  gas  velocities than the
conventional mode.  In this mode, the  bed material and  fly ash
are captured in the hot cyclone, collected in the cyclone bottom
and reintroduced into the bed.
  The hot combustion gases and particulate matter also leave the
combustion vessel and enter the hot cyclone.  Heavier particulate
matter  is  returned to the  combustor with the recycled bed ma-
terial, while lighter particulate matter is routed with  the hot gases
through the heat exchanger and then through the particulate and
gaseous pollution control systems.
OPERATING CONDITIONS
   A comparison of operating conditions of the two combustion
units is presented in Table 1. In general, the rotary kiln/after-
burner system is characterized by higher temperatures than the
fluidized bed  system. This condition is due primarily to the in-
creased turbulence and mixing in the fluidized bed system which
results in increased combustion efficiency.
   Also, lower gas  flow velocities in the fluidized bed may provide
for a longer retention time than in the rotary kiln. This longer re-
tention time also results in increased combustion efficiency.

                           Table 1
   Comparison of Operating Parameters for Fluidized Bed and Rotary
    Kiln Incinerators for Decontamination of Soil at CERCLA Sllet
      PARAMETER
                                   tin Of IHCIHERATOR
                              ROTARY KlLJi       fLUlOlICO BCD
                              1200 - 1600


                              2000 - 2400


                               0.5   1.5
                                               1150 - 1500
   Not
Applicable
                                               0.75   2.5
      dcnc« TlM/S«cortd«ry
       H«c>
       BTU/hr-ft'

  Airflow Requirements
       ( in HjO w.C. T
  Velocity
       (ft/>ecl
                                               Applic«bU

                            25.000   40.000    20.000   15.000*
                            SO  2001  «xcet»
                                     • ir
                              -0.5 to -t.O
15% e>cen


  -1.0


  15 - 18
"Based on limited data. In actuality, expected to be higher because of high turbulence con-
ditions,

COMPARISON OF PCB DESTRUCTION
EFFICIENCY
  TSCA requires that destruction of PCBs in contaminated solids
meet the following criteria:

(1)  99.99% destruction removal efficiency
(2)  Exhaust stack PCBs fed to incinerator.

  Although operational destruction efficiency data for PCB-con-
taminated soils are  limited, some measured destruction efficien-
cies for the Rotary Kiln incinerator and the Fluidized Bed Incin-
erators exist (Tables 2 and 3).
  In general, both types of incinerators meet the required TSCA
destruction efficiency criterion of 99.9999%.
                           Figure 3
             Circulating Fluidized Bed Combustor Unit
                                                                                             Table 2
                                                                                  Rotary Kiln Performance Results
                                                                           (PCB-Contamlnaled Soil Destruction Efficiency)
                                                                      INCINERATOR
                                                                                                 PCB DESTRUCTION EmCIENCY Ml
                                                                      Rolllni/D««r Park

                                                                      Enico
                                    • 99.5

                                     99.99994}
COMPARISON OF COST AND SPACE REQUIREMENTS
   A comparison of cost and space requirements of the two types
of combustion units is presented in Table 4. In general, this table
shows that,  because  of  lower temperature  requirements and
greater heat  release  rates, the Fluidized Bed  Incinerator should
have lower operating costs and should require less space. How-
ever, the above conclusions are based on  limited Fluidized Bed
Incinerator data and also the effect  of scale-up factor  has not
been established for the Fluidized Bed Incinerator.
244    INCINERATION

-------
                            Table 3
                Fluidized Bed Performance Results
         (PCB-Conlaminated Solil Destruction Efficiency)
TEST NUMBER
CONDITIONS

Soil feed rate, Ib/hr
P
combustion chamber, ft/sec
Excess oxygen, %
1
11,000
325
1,800
18.7
7.9
2
12,000
410
1,800
18.7
6.8
3
9,800
325
1,800
18.1
6.8
RESULTS
     Destruction and removal
      efficiency, (DRE), %

     PCB in  bed ash,  ppm

     PCS in  fly ash,  ppm

     Dioxin/furan in ash, ppm

     Combustion efficiency, %
                                    >99.999
                                             >99.999
                                                      >99.999
0.0035    0.0330     0.1860

0.0660    0.0100     0.0320

	Not Detected	

 99.94     99.95      99.97
                                                                Table 4
                                           Features of Fluidized Bed vs Rotary Kiln Incinerators
                                                                           PARAMETER
                                                                          Space
                                                                          Requirements
                                                                           Instrumentation
                                                                          Operating
                                                                          Costs
                                                                          Capital
                                                                          Costs
                                     Operator
                                     Skill
Maintenance
Cost
Scale-up
Problems
                                                                                                       COMMENTS
FBC should require  less space
than RK due to greater heat
release at same waste feed
rate.   Also fBC has no
afterburner.

No appreciable difference
between FBC and RD

FBC should have lower cost due
to lower fuel consumption
because of lower temperature
requirement

FBC should have lower cost due
to smaller size and no
afterburner.

No appreciable difference
between FBC and RK

FBC should be slightly lower
due to  smaller size and  no
afterburner

More data available for  RK
while FBC scale-up not
established
CONCLUSIONS
  It appears that either of the technologies for hazardous waste
incineration discussed in this paper can be used for contaminated
soil  cleanup.  The waste preparation and  operating  conditions
will determine the resulting PCB destruction efficiency.
  It appears that the circulated Fluid Bed  Incinerator may have
decided cost advantages and space benefits and  could be moved
more easily from site to  site as a transportable unit. However,
this conclusion is based on limited operational and test data.
                                  REFERENCES
                                  1.  "Waste Management Systems," PEDCo, Inc., Cincinnati, OH, 1985.
                                  2.  Martin, J.F., "Use of a Mobile Waste Incineration System for a Site
                                     Cleanup," ENSCO/Pyrotech, White Bluffs, TN, July 30, 1985.
                                  3.  Santoleri, J.J.,  "Energy Recovery From Industrial Waste Incineration
                                     Processes," 1985 Industrial Pollut. Control Symp., Feb. 1985,49-56.
                                  4.  Frankel, I., Sander, N., Vogel, G. and Lee, C.C., "Profile of the Haz-
                                     ardous Waste Incinerator Manufacturing Industry," Incineration and
                                     Treatment of Hazardous Waste: Proc. of the Eighth Annual Research
                                     Symp., U.S. EPA, Mar. 1982, 1-13.
                                                                                                                 INCINERATION    245

-------
              New Requirements  for  Underground Storage Tanks

                                            Anna O.  Buonocore, P.E.
                                                  Gerald F. Kotas
                                    U.S. Environmental  Protection Agency
                                    Office of Waste Programs Enforcement
                                                Washington, D.C.
                                            Kevin G. Garrahan, P.E.
                                    U.S. Environmental  Protection Agency
                                      Office  of Research and Development
                                                Washington, D.C.
ABSTRACT
  The Hazardous and Solid Waste Amendments of 1984 became
law on Nov. 8, 1984 and, in addition to reauthorizing and mak-
ing numerous changes to RCRA, a new Subtitle 1 was added that
requires the U.S.  EPA to  establish a comprehensive program
for the regulation of underground storage  tanks. Subtitle  I
addresses underground storage in tanks  of all substances regu-
lated under CERCLA (except for hazardous waste storage tanks
which already are regulated under Subtitle  C of RCRA), as well
as petroleum products. It  is estimated  that  over  one  million
underground tanks are subject to regulation under the new pro-
gram.
  Congress enacted the legislation  for regulating underground
storage tanks due  to the increasing number and severity of en-
vironmental contamination  incidents resulting  from  leaking
underground tanks. There is growing evidence  that leaking tanks
are becoming a major cause of groundwater  contamination.
  The major provisions of Subtitle I include notification by own-
ers of underground tanks to designated State  or local agencies,
an interim  prohibition on the  new installation  of  unprotected
tanks, the development of technical standards for new and exist-
ing underground tanks including corrective action requirements
for leaking tanks and tough new enforcement authorities to com-
pel compliance with the new regulatory requirements. Subtitle I
also allows  the U.S. EPA to promulgate financial responsibility
regulations  for taking corrective action and compensating third
parties for bodily injury and property damage caused by leaks
from tanks. In addition, the U.S. EPA is required to authorize
State programs regulating underground storage tanks in lieu of
the Federal  program provided that the State's program is no less
stringent and provides for adequate enforcement.

INTRODUCTION
  On Nov.  8, 1984, President Reagan signed into law the Haz-
ardous and Solid Waste Amendments of 1984 which,  several days
earlier, had  been approved by Congress. In  addition to reauthor-
izing and making numerous changes to RCRA, a  major new sec-
tion was added to the  RCRA statute requiring the U.S. EPA to
establish a comprehensive program for the  regulation of under-
ground storage tanks.  This new  section,  Subtitle I, addresses
underground storage of most hazardous  substances and petrol-
eum products. It is estimated that over one  million underground
tanks will be subject to regulation under this new program. This
paper describes the major statutory  provisions of Subtitle I, the
the U.S. EPA plans to implement these provisions and,  finally,
the effect this program may have on the regulated community.
PAST PROBLEMS
  Congress enacted the legislation  for regulating underground
storage tanks due to the increasing number of incidents and sever-
ity of environmental contamination resulting from leaks  from
underground storage tanks (USTs). Leaks from underground
tanks can create safety,  health and environmental hazards. Leaks
also can result in fires and explosions, even in structures remote
from the tanks. Underground storage tank leaks are insidious—
they can contaminate soil, groundwater and surface water, and
yet the contamination usually is not identified until after signif-
icant damage has  occurred. Oftentimes,  the cost of cleanup
greatly exceeds the cost of prevention in the first place.
  There is growing evidence that leaking tanks are becoming a
major cause of groundwater  contamination. This contamina-
tion is of special concern since over one-half of the U.S. popu-
lation relies upon groundwater for its drinking water supply. This
increased endangerment to groundwater users is illustrated by the
recent update of the U.S. EPA's National Priorities List (NPL)
of hazardous wastes sites eligible for federal funding of remedial
cleanup activities under  CERCLA.
  The October 1984 update to the  NPL proposed the addition
of 248 new sites to the  existing list of 538 sites. Of the 248 new
proposed sites,  19 were sites containing leaking storage tanks lo-
cated  in the Silicon  Valley of California, an area  previously
thought to contain the  pollution-free computer industry. Leaks
from these tanks,  primarily chlorinated organic  solvents, have
been linked to  potential adverse health effects. Of the total of
786 sites on the NPL, at least 57 now are known to include chem-
ical contamination in  the environment due to leaking under-
ground storage tanks. Since petroleum  products were exempted
by Congress under CERCLA, there are many problem sites which
presently cannot be cleaned up using "Superfund"  monies or
CERCLA enforcement authorities.
  Gasoline and other petroleum products stored in underground
tanks (particularly  those installed in the 1950s and 1960s) are in
fact becoming increasingly identified as responsible for  numer-
ous safety, health and environmental  contamination incidents
around  the country.' In one example, 30,000 gal of gasoline
leaked  into a New York  community's groundwater  supply in
1978 causing odor, safety  and  potential health problems for 27
families. The involved company's costs for cleanup and damages
were estimated  to be between 5 and  10 million dollars as of May
1984.
  Gasoline from a leaking underground  tank in Colorado in
1980 posed a continuing threat to nearby residents, thereby caus-
ing the courts to order the responsible company to purchase 41
246   UNDERGROUND LEAKING TANKS

-------
houses at over twice  their appraised value, an estimated total
cost to the company of 10 million dollars. In a third example, 23
private water wells in Massachusetts were contaminated by oil
leaking from nearby  underground storage tanks  in  1981.  The
company responsible was required to supply bottled water to the
affected residences.

SCOPE OF NEW RCRA SUBTITLE I
  The requirements for underground storage tanks are specified
in the new Subtitle I of  RCRA.The major provisions in this
statute include notification by owners of underground  tanks to
designated State or local agencies, an interim prohibition on the
new installation of unprotected tanks, the development of tech-
nical standards for new  and existing underground tanks includ-
ing corrective action requirements for releasing tanks and tough
new enforcement authorities to compel compliance with  new reg-
ulatory requirements. Subtitle I also allows the U.S.  EPA to
promulgate financial  responsibility  regulations  for taking  cor-
rective action and compensating third parties for bodily injury
and property damage  caused by leaks from tanks. In addition,
the U.S. EPA is required to authorize State programs regulat-
ing underground storage tanks in lieu of the Federal program pro-
vided that the State's  program is no less stringent and  provides
for adequate enforcement.
  Subtitle I is divided  into ten parts, Sections 9001-9010. Table 1
provides a brief description of each section of the statute.
                           Table 1
                   Scope of RCRA Subtitle I
Section    Description
9001       Definitions and Exemptions
9002       Notification
9003       Release Detection, Prevention, Correction Regulations;
          Financial Responsibility; and Interim Prohibition
9004       Approval of State Programs
9005       Inspection, Monitoring and Testing
9006       Federal Enforcement & Penalties
9007       Federal Facilities
9008       State Authority
9009       Study of Underground Storage Tanks
9010       Authorization of Appropriations
  Section 9001 of Subtitle I contains the definitions of key terms
and also identifies exemptions to the statute. The reader should
refer directly to the statute for the exact wording in order to de-
termine the applicability of certain requirements to himself or his
client.  However,  a few significant definitions and exemptions
are briefly noted below.
  The  term "underground storage tank" has been  defined as
one or more tanks (including connected  underground  pipes)
used to contain an accumulation of regulated substances  with a
volume which is 10% or more beneath the surface of the ground.
Specifically exempted from this definition are:
• Farm or  residential tanks  with a capacity less than 1,100 gal
  used for storing motor fuel oil for non-commercial purposes
• Tanks used for storing heating oil for consumptive use on the
  premises
  Septic tanks
  Regulated pipeline facilities
  Surface impoundments
  Storm water or waste water collection systems
  Flow-through process tanks
• Liquid trap or gathering lines associated with oil or gas produc-
  tion operations
• Storage tanks situated in an underground area if the tanks are
  situated upon or above the surface of the floor.

  The term "regulated substance" has been defined to include:
• Any substance defined as a hazardous substance under Section
  101 of CERCLA (Superfund) but not including any substance
  already regulated under RCRA Subtitle C as a hazardous waste
• Petroleum, including crude oil or any fraction thereof, which
  is liquid at standard conditions of temperature and pressure
  The term "release" has been defined to mean any spilling, leak-
ing, emitting, discharging, escaping, leaching or  disposing from
an underground storage tank into groundwater or  subsurface
soils.

SUMMARY OF MAJOR PROVISIONS

Notification
  Every owner of an existing underground storage tank holding
petroleum or hazardous substances currently in use or taken out
of operation after Jan. 1,1974 (but left in the ground) must notify
a designated State or local agency by May 8, 1986. Notification
must  specify the age, size, type, location and uses of the tank.
Owners of tanks taken  out of operation after Jan. 1, 1974 must
also notify the appropriate government agency, the date the tank
was taken out of service, the age of the tank on that date and the
type and quantity of substances left stored in the tank. Exempted
from  this requirement are owners of tanks which have  been re-
moved from the ground, owners of tanks taken out of operation
on or before Jan.  1, 1974 and owners who previously notified
the U.S. EPA pursuant to Section 103(c) of CERCLA.
  Any owner who brings into service a new underground storage
tank containing regulated substances after May 8,1986 is required
to notify a designated State or local agency within thirty days of
bringing the tank into use.
  The U.S. EPA, in consultation with designated State and local
officials and after public comment, prescribed on Nov. 8, 1985,
the form  of notice and the information it  would contain.  The
U.S. EPA requires States to use the prescribed form unless  they
develop comparable forms that contain, as a minimum,  the same
information. A list of the designated State or local agencies to
receive the notifications as well as a list of States whose form(s)
complied with the notification requirements was published in the
Federal Register with the form.
  In addition to the above notification requirements for owners,
Subtitle I also imposes obligations on persons who deposit regu-
lated  substances in underground storage tanks and tank sellers.
Beginning Dec. 8,  1985 and lasting for a period of 18  months,
any person who deposits substances in an underground storage
tank is required to inform the owner and operation of such  tank
of the owner's notification obligations. Furthermore, beginning
30 days after the U.S.  EPA publishes standards for  new tanks,
tank sellers are required to inform tank purchasers of their noti-
fication obligations.

Interim Prohibition
  During the interim period, between May 7, 1985 and the effec-
tive date for new tank standards, Section 9003(g) of Subtitle I
prohibits the new installation of any underground storage  tank
for the purpose of storing regulated substances unless it meets
three  criteria:
• It will prevent releases due to corrosion or structural failure for
  its operational life
                                                                                     UNDERGROUND LEAKING TANKS    247

-------
•  It is cathodically protected against corrosion, is constructed of
   non-corrosive material, is steel clad with a noncorrosive ma-
   terial or is designed to prevent the release or threatened release
   of any stored substance
•  The material used in the construction or lining of the tank is
   compatible with the substances to be stored
   A limited waiver is provided dealing with corrosion preven-
tion. The statute specifically allows for tanks to be installed with-
out corrosion protection if the soil has a resistivity value of 12,000
ohm/cm  or more as measured  in accordance with ASTM  Stan-
dard G57-78.
   At the  time of this writing, the U.S. EPA has prepared an in-
terpretive rule on the interim prohibition. An interpretive rule
simply construes the language of the statute and does not impose
any additional obligations. Its purpose is to provide clarification
of the interim  prohibition provisions necessary for implementa-
tion and enforcement purposes. The U.S.  EPA also is preparing
a guidance document that will discuss available methods and tech-
nologies which the U.S.  EPA believes fulfill the interim  prohi-
bition requirements.

Technical Standards
   Also under Section 9003 of Subtitle I, the U.S. EPA is required
to develop regulations for the detection, prevention and correc-
tion of releases from tanks, as necessary to protect human health
and the environment. These regulations are applicable to all own-
ers and operators of underground storage tanks. In promulgating
these regulations, the  U.S. EPA may distinguish between tanks
of different types, classes and ages.  In making such distinctions,
the U.S. EPA will consider many factors including: location of
the tank,  soil and climate conditions, uses of the tank, history of
maintenance,  age of  the  tank,  current  industry  recommended
practices, national consensus codes, hydrogeology, site-specific
depth  to  water table, size of the tank, quantity  of substances
stored, technical  capability  of  tank owners and operators and
the compatability of the stored substances with the tank material.
   The statute specifically states that, at  a  minimum, the release
detection, prevention  and  correction regulations  shall include
requirements for:
•  Maintaining a leak detection system, inventory control system
   or comparable system to identify releases
•  Maintaining  records of any leak  detection  or  inventory con-
   trol system
•  Reporting releases and any corrective action taken
•  Taking corrective action in response to a release from an under-
   ground storage tank
•  Closing tanks to prevent future releases
  The U.S. EPA also is required by Subtitle I to develop and
issue performance standards for new underground storage tanks.
These  performance standards must cover  design, construction,
installation, release detection and compatibility requirements.
  Finally, the statute  specifies schedules and  deadlines that the
U.S. EPA must meet in issuing technical standards and regula-
tions. These deadlines differentiate between regulations for new
and existing tanks as well as for  petroleum tanks and non-petrol-
eum tanks. Table 2 provides a summary of the deadlines imposed
on the U.S. EPA to develop technical standards.

Financial Responsibility
  At the Administrator's discretion, the U.S. EPA is allowed to
develop regulations containing requirements for maintaining evi-
dence of financial responsibility as deemed necessary for taking
corrective action and  compensating third parties for bodily  in-
jury and property damage caused by releases from an under-
                           Table 2
           Summary of Subtitle 19003 Statutory Deadline*

Tank Regulations Governing:    Statutory Deadline
New and Existing Petroleum     Promulgate by February 1987
  Tanks                     (Effective May 1987)
New Hazardous Chemical Tanks Promulgate by August 1987
                            (Effective November 1987)
Existing Hazardous Chemical    Promulgate by August 1988
  Tanks                     (Effective November 1988)
ground  storage tank. These requirements would apply to both
sudden and nonsudden accidental releases.
  The financial responsibility requirements may be established by
any one or combination of the following mechanisms:

  Insurance
  Guarantee
  Surety bond
  Letter of credit
  Qualification as self-insurer
  The U.S. EPA presently is planning to perform a study to de-
termine the most  effective  mechanisms to maintain financial
responsibility.

Federal Enforcement Authority
  Section  9006 of Subtitle I establishes tough new federal  en-
forcement authorities to compel compliance with the regulatory
requirements for underground  storage tanks.  According to  the
statute, whenever the U.S. EPA determines that any person is
in violation of any requirement of Subtitle I, the U.S. EPA may
issue an administrative order requiring compliance or may com-
mence  a civil action  in federal district court for appropriate re-
lief, including a temporary or permanent injunction. If a viola-
tor  fails to comply with a compliance order, he shall be liable for
a civil penalty of up  to $25,000 for each day of continued non-
compliance.
  Any tank owner who fails to notify or falsely notifies the U.S.
EPA of his underground storage tanks shall be subject to a pen-
alty of up to $10,000  for each tank. Additionally, any lank owner
who fails to comply with any of the technical standards developed
by the  U.S. EPA or  an approved State  program or violates  the
interim  prohibition provisions shall be subject to civil penalties of
up to $10,000 per tank for each day of violation.

DEVELOPMENT OF THE U.S. ERA'S
UNDERGROUND STORAGE TANK PROGRAM
  While Congress was  busy  debating  various provisions  for
underground storage tank legislation, the U.S. EPA already was
preparing to implement what would likely be a new mandate for
comprehensive regulation of underground storage tanks. In antic-
ipation of extremely short deadlines for the development of regu-
lations, the U.S.  EPA's Office of Solid Waste and Emergency
Response  began formulating its own strategic plan for imple-
menting new regulations for underground storage tanks. As a
result,  several  weeks before the Hazardous  and Solid Waste
Amendments of 1984 became law on Nov. 8, 1984, the U.S. EPA
already had prepared an "Organization  and Strategic Plan" for
implementing the new Subtitle 1 of RCRA.
  The  Strategic Plan specified the formation  of several Work-
groups to address the major provision of Subtitle I. These Work-
groups  were directed by a Steering  Committee of senior U.S.
248   UNDERGROUND LEAKING TANKS

-------
EPA managers. The goal of each Workgroup was to write a
development plan for its respective topic. Each development plan
was required to address major issues, recommended and alterna-
tive approaches, scheduling and resource needs.
  The development  plans are now officially in final draft and
have been consolidated as chapters of the U.S. EPA's Develop-
ment Plan on the Regulation of Underground Storage Tanks.
General goals and objectives were identified during preparation
of the development plans and are reflected in the chapters. These
goals and objectives also will be  reflected in any modifications
to the plans as well as the implementation of the plans. These
goals and objectives are as follows:
• The U.S. EPA will place high priority on regulation develop-
  ment in order to meet Congressional deadlines.
• In order  to develop regulations as efficiently  as possible, a
  common target date will be used for all regulations.
• Innovative approaches will be emphasized.
• Underground  tank  regulations  will  reflect the U.S. EPA's
  groundwater protection strategy to the extent possible.
• An extensive public out-reach and consultation program will
  be developed to involve the various sectors of the interested
  public.
• High priority will be given to State participation in the program
  development.
• Special considerations will be given to the problems of small
  businesses during development of the program.

INSPECTION AND ENFORCEMENT
  A three-phased approach will be utilized to implement inspec-
tion and enforcement activities:
• Compliance promotion
• Compliance monitoring
• Enforcement response actions

  This approach will be applied toward the provisions of the in-
terim prohibition  of non-protected tanks, notification require-
ments and all the regulatory requirements (such as tank standards
and corrective action provisions) which are promulgated.
  Compliance promotion activities will consist of an educational
program implemented mostly through the extensive public out-
reach and consultation program which is being developed for the
overall UST regulatory program.  This communication program
will  involve various sectors  of the interested public including
trade associations, the Tank Coalition, the U.S. EPA's Small
Business Ombudsman, trade journals and local government agen-
cies. A section of the out-reach and consultation program will be
tailored specifically toward promoting inspection and compliance
activities. In addition to promoting the technical requirements of
the Underground Storage Tank (UST) regulatory program, com-
pliance promotion activities will point out the existence of tough
penalties for noncompliance cited in Section 9006 of the statute.
  The compliance monitoring program will discover noncompli-
ance  through innovative  enforcement activities as  well as tra-
ditional enforcement inspections. One such activity may be to
develop a program which crosschecks local construction or build-
ing permit applications. Additional activities may include the use
of self-monitoring reports, the existing RCRA hotline  or a sep-
arate hotline for  private citizens to report releases  that may be
threatening public health  or the environment.
  The targeting of inspections will be necessary due to the size of
the regulated community. A targeting strategy will be developed
using information  from past  problems and new data generated
from the communications program. Of particular importance will
be inspection of new tank installations, since improper installa-
tion appears to be the cause of many leaks. The UST inspection
program also may be integrated with existing environmental in-
spection programs (such as the Spill Prevention and  Counter-
measures [Controls program or the Mobile Sources gasoline sta-
tion inspection program) to incorporate institutional experience
and resources.
  The final and  most important enforcement phase in imple-
menting UST inspection and enforcement activities is the estab-
lishment of a highly visible and effective enforcement presence
through the active and successful use of enforcement  response
actions. This presence  will be established through the active use
of compliance orders, tough penalties and judicial actions in fed-
eral court.
  Section 9006 provides the U.S. EPA authority to enforce the
provisions of Subtitle I through administrative orders  and civil
judicial actions for injunctive relief or penalties.  When  a signifi-
cant violation such as failure to have a monitoring system is dis-
covered, an administrative order will be issued promptly and an
appropriate penalty will be levied. If compliance is not  attained,
further penalties will be levied in accordance with the U.S. EPA's
RCRA penalty policy and the enforcement action  will be esca-
lated to a judicial action.
  The following penalties are specified in Subtitle  I  for certain
violations of Subtitle I: up to $25,000 for each day for the viola-
tion of an administrative order issued under Section 9006; up to
$10,000 for each tank each day a violation continues  for any re-
quirement or standard promulgated under section 9003 (or sec-
tion 9004 in approved  states);  up to $10,000 for each tank each
day it  remains in  violation of the interim prohibition of section
9003 (g).
  Section 9006 orders also will be  used to require compliance
with corrective action provisions of regulations to be issued under
section 9003. This type of order will enable enforcement person-
nel to require responsible parties to perform complete  cleanup of
releases. Depending on corrective action requirements, the order
can be phased using compliance schedules.
  In situations  where there  exists an imminent and  substantial
endangerment to human health and the environment, other en-
forcement authorities such as RCRA 7003 and CERCLA 106
and 107 will be considered in determining enforcement responses.


CONCLUSIONS
• The regulatory program for underground storage tanks is, per-
  haps,  the largest and potentially  most costly  regulatory  pro-
  gram the U.S. EPA has had to develop to date.
• The tough provisions for federal enforcement make non-com-
  pliance a very risky and potentially costly alternative.
• Given the technical difficulties, high costs and legal  liabilities
  of cleaning up releases from underground storage tanks, the
  best strategy  for an owner or operator is to perform everything
  in his power to prevent leaks or releases. In the long run, this is
  clearly the best approach.
REFERENCES
1. Aim,  A.L., U.S. EPA Memorandum to  L.M. Thomas  regarding
  Responsibility for LUST Program, Sept. 1985.
2. Garrahan, K.G., U.S. EPA Memorandum to P.  Hansen  regarding
  Revised Draft Development Plan for Inspection & Enforcement of
  Underground Storage Tanks, Jan. 18, 1985.
3. Lehman, J.P.,  U.S. EPA Memorandum to A. Montrone  regarding
  LUST Phase I Planning, Oct. 31, 1984.
4. Moore, J.A., U.S. EPA Memorandum (draft) to W. Ruckelshaus
  regarding Leaking Underground Storage Tanks: Initiation of Regula-
  tory Action—Action Memorandum, Sept. 25, 1984.
                                                                                     UNDERGROUND LEAKING TANKS    249

-------
5. Thomas,  L.M.,  U.S.  EPA Memorandum to A.L. Aim  regarding      7.  U.S. EPA, Regulation of Underground Storage Tanks, RCRA Sub-
   OSWER's Strategic and Organizational Plans for the Congression-         title I, Development Plan (Draft), Mar. 18, 1985.
   ally-mandated Regulation of Underground  Storage Tanks (RUST)      g  y s EPA| Underground Storage Tanks (UST) Project Update, Nov.
   Program, Oct. 19, 1984.                                                6_ 1984.
6. U.S. Code of Federal  Regulations, 40 CFR Part  280, 1984,  Haz-
   ardous and Solid Waste Amendments. Resource Conservation and
   Recovery Act Subtitle 1 Regulation of Underground Storage Tanks.
250    UNDERGROUND LEAKING TANKS

-------
                Cost-Effectiveness Evaluation of Leak Detection
                              and Monitoring Technologies for
                          Leaking Underground  Storage Tanks

                                             James Lu, Ph.D., P.E.
                                               Wayne Barcikowski
                                        Jacobs Engineering Group, Inc.
                                              Pasadena, California
ABSTRACT
  In recent years the contamination of groundwater by leaking
underground storage tanks (LUST) has become a major environ-
mental concern.  Local,  state  and federal governments  have
responded by passing legislation and promulgating regulations for
the control of LUST. These government regulations are ap-
plicable to both new and existing underground storage facilities.
Tank owners/operators are having to replace or retrofit existing
facilities and have little information and guidance  available to
them on the cost and effectiveness of leak control options.
  Under most UST regulations, owners of UST facilities typically
have a variety of leak prevention, detection/monitoring, contain-
ment and cleanup options available. In particular, there are many
options for leak detection and monitoring. The cost-effectiveness
of leak detection monitoring techniques has been examined in the
course  of several U.S. EPA and private client sponsored studies
on UST.
  This evaluation consists of the following analysis:
• Type of detection/monitoring techniques
• Applicability of detectors/monitors
• Cost comparisons
• Effectiveness analysis
• Cost-effectiveness analysis
  A comprehensive survey was conducted to investigate commer-
cially available detectors and  monitoring equipment. Data of ac-
curacy, response time, detection limits, sensitivities, application
ranges, costs and constraints  were analyzed  and  compared.
Models were developed to evaluate the effectiveness based on the
extent of leakage/contamination possible before detection can be
achieved.  Capital, O&M  and associated soil and groundwater
cleanup costs were analyzed under a variety of scenarios.

INTRODUCTION
  With the recent increased  concern for protection of ground-
water from contamination by leaking underground storage tanks
(LUST) being translated into new governmental regulations, both
tank owners and government agencies charged with overseeing
the problem have to decide what leak detection/monitoring tech-
nologies are appropriate for their needs. There is much informa-
tion available on detecting and monitoring for contamination in
the vadose zone and in the groundwater,  but there still  is not
much published literature analyzing either the effectiveness or the
costs of various  technologies  for LUST. Analyses and  com-
parisons of vadose zone and groundwater monitoring techniques
are available for leak monitoring  technologies.1A3 There is also
some information available on effectiveness of leak detection
technologies.4 At this time, analyses of the costs of technologies
and costs of cleanup in the event of a leak are needed by those
having to set up leak detection/monitoring programs for under-
ground tanks.
  To address this need, we have compiled and analyzed informa-
tion  from the  literature and information on  costs  and
characteristics  of  leak  detection/monitoring  technologies
gathered from vendors as part of a U.S. EPA sponsored study on
LUST. The effectiveness of a technology and associated costs are
reviewed and summarized. Costs of soil and groundwater con-
tamination also are arrived at for each technology with the use of
liquid and vapor transport models for the vadose  zone and
saturated zone.

EXISTING TECHNOLOGIES
  For identifying leaks from underground storage tanks, there
are a wide variety of technologies available. These technologies
can be classified into two categories for assessing effectiveness
and for identification purposes. One category,  Leak Detection,
includes methods which,  when applied to the tank itself, will
determine  if the tank if presently leaking. A second category,
Leak Monitoring, includes methods  for use outside the tank and
will identify both present and past leaks. Leak detection includes
inventory reconciliation, interstitial  monitoring of the space be-
tween primary and secondary containment (e.g., monitoring the
anular space of a double-walled tank) and the use of tank "tight-
ness  testing" for leak identification. Leak monitoring in the en-
vironment outside of the tank and  any secondary containment
systems includes visual monitoring  (of the tank and soil), soil
monitoring, air monitoring, vadose zone monitoring (gas and liq-
uids) and groundwater monitoring.
  Some devices used in the above  mentioned  methods  are ap-
propriate for both leak detection and leak monitoring. Various
sensors for detecting the presence of specific liquids  or gases are
utilized for interstitial, vadose-zone  or groundwater  monitoring.
These  detection devices and other  practical  leak  detec-
tion/monitoring methods are listed in Tables 1,  2 and 3.
  A  summary of advantages and disadvantages for each method
is included in these tables. Table 1 includes various tank integrity
(tank tightness tests), inventory reconciliation and visual observa-
tion for the above ground portion of a tank. Table 2 summarizes
the characteristics of gas/vapor detection  techniques for air or
vadose zone monitoring. Table 3 presents various liquid detection
techniques including visual observation of the tank and of the
soil/backfill around the tank.
                                                                                UNDERGROUND LEAKING TANKS    251

-------
                                                                 Table 1
                                                 Comparisons of Some Tank Integrity Tests
              Technique/
                Device
                                                  Advantage
                                                       Oil advantage
            Inventor)
            Reconciliation
            Pneumatic Tests
             J-Tube
             Manometer
             ARCO HTC
             Hunter Leak
             Lokator
             (Sunmark)
             Heath Petrotite
             (Kent-Moore)
             Horner Ezy-Check
             Ainly Tank
             lntegrit> Tester
Quick and relatively Inexpensive
Quantifiable leak rate
In very small tanks—sensitive to small
leaks
Wide applicability
No down time
No damage potential
Fair reliabilit)
Few restrictions

Ver>  sensitive to small leaks
A reliable test
Very sensitive
Quantifiable leak rate
Wide applicability
Low damage potential
Few restrictions

Ver> sensitive test
Good quantification of leak rate
Wide applicability
Low damage potential
Few restrictions
Reliable
Temperature compensated

Very sensitive
Good quantification of leak rate
Wide applicability
Reliable
Detects leaks in tank vs. pipe
Few restrictions
Corrects for temperature

Good sensitivity
Good quantification of leak rate
Wide applicability
Reliable
Few restrictions
Corrects for temperature

Good sensitivity
Good quantification of leak rate
Wide applicability
Reliable
Detects teaks in pipe vs. tank
Few restrictions
Corrects for temperature

Good sensitivity
Good quantification of leak rate
Wide applicability
Reliable
Few restrictions
Corrects for temperature
With larger tank is less sensitive to 8 given
leak rate (e.g. 0.05 gph)
Sensitive to ground water fluctuations in
location of tank
Can't locate leaks
Sensitive to vapor "pockets"
Can damage tank/piping
Can worsen leak
Tests only for portion of tank/piping isolated
for pretsurlzlng/depmsurization
Significant down time
Can't locate leaks

Local groundwater level* can effect test
Sensitive to temperature changes and
temperature distribution in tank
Down time 1-JO hrs
Can't locate leaks in piping versus tank

Sensitive to local groundwater fluctuations
Only test tank at  75%-82% full
Down time fc-8 hrs.
Sensitive to vapor pockets
Overfill technique
Potential damage due to overfill
Downtime 2-5 hrs.
Sensitive to vapor pockets
Overfill technique
Potential damage due to overfill
Down time 3-9 hrs.
Can't locate leaks
                                                                                   Sensitive to vapor pockets
                                                                                   Overfill technique
                                                                                   Potential damage due to overfil
                                                                                   Down time 1-3 hrs.
Sensitive to vapor pockets
Pressure/overfill technique
Potential damage due to pressure/overfill
Down time 1-3 hrs.
Can't locate leaks
                                                               (continued)
APPLICABILITY OF DETECTORS
AND MONITORS
  Tables 1  through 3 illustrate a number of points regarding the
application of existing leak detection/monitoring technologies to
underground  storage tanks.  A  particular  technique may  be
reliable, have  good sensitivity (and low limit of detection), have
low  potential  for  damaging the tank and yet not  be widely ap-
plicable for a  number of reasons. The  technique may be  useful
only for a  limited number of compounds  and/or  require  direct
                                 contact with  the leaking  liquid for  leak  detection. For some
                                 facilities, specific tank integrity tests may be inappropriate due to
                                 the amount of time the tank would have to be out of service (the
                                 "downtime"). Some tank integrity tests, in addition, are not able
                                 to detect leaks in all sections of the tank and/or-discriminate  be-
                                 tween  leaks in  the tank  and associated  piping.  A particular
                                 method or  device also may be limited by  environmental condi-
                                 tions. "Background"  contamination or local groundwater levels
                                 may  preclude the use of a technique.
252    UNDERGROUND LEAKING TANKS

-------
                                                        Table 1 (continued)
             Technique/
               Device
                                               Advantage
                                                    Disadvantage
           Mooney Tank
           Leak Detector
           Helium Sniffer
            Other Tracer
            Techniques
            Acoustic Techniques
            Monitoring Level
            in Tank
            Monitoring for
            Water in tank
Good sensitivity
Good quantification of leak rate
Wide applicability
Reliable
Detects leaks in tank vs. pipe
Few restrictions
Corrects for temperature

Very sensitive
Can pinpoint location of the leak
Wide applicability
Few restrictions
Insensitive to temperature changes
Few restrictions
Wide applicability
Very sensitive
Temperature insensitive
No down time

Sensitive
Widely applicable
Not affected by temperature
Fair to poor sensitivity
Instantaneous response
Fair quantification of leak rate
Widely applicable
Reliable
No down time
No potential for damage

Fair to good sensitivity
Widely applicable
Reliable
No down time
No potential for damage
Few restrictions
Insensitive to temperature
                                           Sensitive to vapor pockets
                                           Overfill technique
                                           Potential damage due to overfill
                                           Down time 2-4 hrs.
Expensive
Some down time
Only rough guess for leak rate
Pressure technique
Possible damage to tank/pipes due to pressuri-
zation

Rough guess for leak rate
Some are pressure techniques
Can't locate leak
Some down time
No quantification of leak rate
Some techniques sensitive to groundwater level
Some possibility of damage due to depressuri-
zation/pressurization
Can't locate leak

Temperature sensitive
Sensitive to local groundwater levels
Sensitivity depends upon size/dimensions of
the tank
Can't locate leak
Limited to organic compounds/halogenated HC's
Sensitive to groundwater level
Sensitivity depends on tank size
Not quantifiable
Can't locate leak
   Tank integrity tests and inventory reconciliation in general have
 wide applicability in terms of range of compounds,  size of tanks
 and environmental conditions. Some tightness tests are limited,
 however, to detection of leaks in a tank which is less than 100%
 full (Table 1). These tests may not be used to detect leaks in the
 upper portion of a tank and possibly in the associated  piping.
   Techniques for detecting the volatilized portion of  a leaking liq-
 uid in the vadose zone can be very sensitive in addition  to having
 wide applicability. Methods which entail taking a vapor  sample
 for subsequent  laboratory  analysis are appropriate for  a wide
 range of compounds,  environmental conditions and monitoring
 situations.
   Additional techniques for vadose zone vapor monitoring in-
 clude the placement of a sensor in  the ground  or  sampling of
 several  vapor wells with a portable gas  detection instrument
 (Table 2). There  are  many types of sensors and  instruments
 available, and all vary in the range of compounds they will detect,
 their sensitivity  and reliability. Typically, there is a "tradeoff"
 between the sensitivity (or limits of detection) and the  reliability
 of the instrument.
  Monitoring for liquid product in the vadose zone or in ground-
 water is more difficult because placement of the sampling points
may not intercept  all possible leaks. Fluctuations in groundwater
level may limit the use of some methods for laboratory monitor-
ing in the vadose zone. Except  for methods utilizing liquid tracers
and/or laboratory analysis, most techniques rely on devices which
                                are limited to use with specific classes of compounds or which are
                                sensitive to environmental conditions (groundwater level or back-
                                ground contamination).
                                  Most techniques used for leak monitoring also can be used for
                                leak detection in the interstitial space of tanks between primary
                                and secondary  containment. In  this technology,  placement of
                                monitoring points is less critical than for environmental monitor-
                                ing, and specific sensors or laboratory analysis can be tailored for
                                each tank.
                                COST-EFFECTIVE METHODOLOGY
                                  Two approaches are used for the cost-effectiveness analysis of
                                underground tank detectors and monitors. One approach is based
                                on the technology itself (i.e., technical effectiveness, implementa-
                                tion costs and O&M costs); the other is based on the technology
                                evaluation plus the potential liability (i.e., cleanup costs) if leaks
                                occur. Evaluation of the latter case is considered necessary owing
                                to the inherent problems of some of the detectors/monitors. For
                                example,  vadose zone or groundwater monitors will not  detect
                                problems  until a leachate plume reaches the sensors. Tank tight-
                                ness tests  can only detect leaks at the moment of the test. Inven-
                                tory control (reconciliation) may never detect leaks if leak rates
                                are  lower than 0.2 gal/hr (as in most cases). Due to the complexity
                                of the second  approach, only case examples are  given in this
                                report.
                                                                                          UNDERGROUND LEAKING TANKS    253

-------
                                                                     Table 2
                                                Comparisons of Vapor/Gas Sensing Technologies
               Technique/
                 Device
            Advantage
                                                          Disadvantage
              Odor Detection
              Adsorption
              Sensitive Resistor
              (sensor probe)
              Metal Oxide
              Semiconductor
              Gas Detectors
              (probe)
              Synthetic Filament
              Probes (e.g.,
              pol>prop> lene)
Quick and inexpensive
No down time
Sensitive for some compounds
Reliable
Insensitive to temperature
No damage potential for tank

No down time
Wide applicability (or volatile compounds
fair to good sensitivity
No damage potential
Fxcellent reliabilit)
Few restrictions
Adjustable for "background" contamination
Temperature compensated

No down time
Ver> good sensitivit)
Wide applicabillt) for volatile compounds
No potential for damage to tanks
Reliable
Few restrictions
Adjustable for "background" contamination
Temperature compensated

No down time
Inexpensive devices
Fair sensitivit>
Reliable
No potential for damage to tanks
Few restrictions
Temperature insensitive
Not suitable for all compounds 'only volatile)
Leak rate is not quantifiable
Restrictions due to Health 4 Safet) concerns
Can't locate leaks
Limited to volatile compounds and gases
Leak rate is not quantifiable
Can't locate leaks
Limited to volatile compounds and gases
Leak rate is not quantifiable
Can't locate leaks
Limited to volatile hydrocarbons (e.g., gasoline)
Leak rate is not quantifiable
Can't locate leaks
Sensitive to background contamination
              Photo-lonization
              Gas Detector
              (survey instrument)
              Flame-lonization
              Gas Detectors
              (survey instrument)
              Combustible
              Gas Meters
              (varous types)
              (survey instrument)
              Tracer Gas with
              Intermittent
              Sampling Program
 No down time
 Excellent sensitivity
 Reliable
 No damage potential
 Few restrictions
 Temperature insensitive
 Wide applicability for volatile compounds
 Can compensate for background contamination

 No down time
 Excellent sensitivity
 Reliable
 No damage potential
 Few restrictions
 Temperature insensitive
 Can compensate for background contamination
 Wide applicability for volatile compounds

 No down time
 Good sensitivity
 Very reliable
 No potential for damage  to tank
 Few restrictions
 Temperature insensitive
 Can compensate for background contamination
 Wide applicability for volatile compounds

 No down time/no damage to tank
 Wide applicability
 Very sensitive
 Temperature insensitive
 Few restrictions
 Insensitive to background contamination
 Reliable
Limited to volatile compounds and gases
Leak rate is not quantifiable
Can't locate leaks
Limited to volatile compounds
Doesn't detect inorganic gases
Leak rate is not quantifiable
Can't locate leaks
 Limited to combustible vapors »nd gases
 Leak rate is not quantifiable
 Can't locate leaks
 Leak rate is not quantifiable
 Can't locate leaks
              Air Sample Traps            No down time/no damage to tank
              with Lab Analysis            Wide applicability
              or "Draeger" Tubes           Good to excellent sensitivity
                                          Temperature insensitive
                                          Few restrictions
                                          Reliable
                                          Can compensate for background contamination
                                                Leak rate is not quantifiable
                                                Can't locate leaks
254     UNDERGROUND LEAKING TANKS

-------
                                                                  Table 3
                                                 Comparisons of Liquid Sensing Technologies
               Technique/
                Device
                                                   Advantage
                                                                                                Disadvantage
             Synthetic Filament
             Devices (e.g.,
             polypropylene)
             Psychometers or
             Tensiometers
             Electrical
             Conductivity
             Sensor
             (Electrodes)
             Thermal
             Conductivity
             Sensors
             Contact Sensitive
             Coated Wires
             (Electrical
             resistance
             change)
             Geophysical
             Techniques
             (e.g., resistivity
             of soil/groundwater)
             Visual Observation
             of Tank and Soil
             Lysimeter Sample
             with Lab Analysis
                or
             Soil Core Sample
             with Lab Analysis
                or
             Tracer with
             Lab Analysis of
             Soil/Groundwater
             Samples
No down time
Reliable
Fair sensitivity
No damage potential for tanks
Few restrictions
Temperature insensitive

No down time
Fair sensitivity
No damage potential for tank
Few restrictions
Temperature compensated

No down time
Reliable
Fair sensitivity
No damage potential for tank
Few restrictions
Adjustable for background contamination
Insensitive to temperature

No down time
Reliable
Fair sensitivity
No potential for damage to tank
Few restrictions
Adjustable for background contamination
Compensates for temperature
Wide applicability

No down time
Reliable
Rough guess of leak location with a
network of wires
Fair sensitivity
No damage to tank
Few restrictions
Temperature insensitive
Wide applicability

No down time
Fair-Good Sensitivity
Fair-Good applicability
No damage potential for tank
Few restrictions
Temperature compensated

Quick and relatively inexpensive
Sensitive to small leaks
Applicable to a wide range of compounds
No down time
No potential for damage to tank
Good reliability
No restrictions

No down time
Fair sensitivity-excellent  sensitivity
No damage potential for tank
Temperature insensitive
Wide applicability
Compensates for background contamination
Reliable
Few restrictions
Limited to specific classes of compounds
Requires contact or close proximity to leaks
Sensitive to background contamination
Can't determine leak rate
Can't locate leak source
Require contact or close proximity
to leaks
Affected by local groundwater fluctuations
Can't determine leak rate
Can't locate leak source

Limited for use with hydrocarbons
Leak rate is not quantifiable
Can't locate leak source
Can't differentiate between H^O, some
halogenated hydrocarbons or other compounds
Requires contact with liquid
Leak rate is not quantifiable
Can't locate leak source
Requires contact with liquid
Limited to specific classes of compounds
Requires contact with liquid
Sensitive to background contamination
Can't determine leak rate
Problem with background contamination
Can't determine leak rate
Can't determine leak location
Affected by groundwater conditions
Expensive and requires expertise


Limited to parts of tank with access
Only a rough guess for leak rate
Requires contact with plume or contaminated
aquifer
Affected by groundwater conditions
Can't determine leak rate
Can't locate leak source
Evaluation Based on Technology
  The  cost-effective  evaluation of  leak  detection/monitoring
technologies is based on a two-level rating scheme which can be
expressed  by  the three equations in  matrix  form  as shown in
Figure 1.
  In these relationships, Ay is the first level rating factor which
grades the  relative rating of cost or effectiveness of detection/
monitoring alternatives, "i", under different factors, "j", affect-
                                  ing the rating. As will be discussed later in this report, Ays for ef-
                                  fectiveness analysis range  from 0 to 3 with 3 being the best. Ays
                                  for cost analysis are based on the following equation:
                                     Ay cost rating  =  log
                                                               105
                                                     (4)
                                  where x  = Annualized cost for leak detection/monitoring.
                                                                                                 UNDERGROUND LEAKING TANKS    255

-------
                  All   A12  Al>

                  A21   A22  A2J
                       Ai2   Ai3
(1)
                                                     (2)
    and
                            A X R
                                                    -(3)
                          Figure 1
     Two-Level Rating Scheme for Cost-Effectiveness Evaluation

  The second level rating factor, TJ, represents the relative impor-
tance of each different factor, j. After multiplication of A,  R
[i.e.,  Equations (1) and (2)] and a conversion  factor,  1/Kj, the
results,  Pj, will represent the weighted  overall  cost-effectiveness
of a certain leak detection/monitoring option, i, that is
                                                         (5)
Where Kj is equal to the maximum value of Ay, in most cases, Kj
 = 3.  A comparison among PJS will give a relative rating of leak
detection/monitoring   technologies  in  terms  of  their  cost-
effectiveness at controlling leaking underground storage tanks.

Evaluation Based on Technology and Liability
  For the second type of cost-effectiveness evaluation, the future
liability should be considered. The future  liability is associated
with the cleanup costs required to decontaminate soil and ground-
water. The soil and groundwater contamination volumes are used
to calculate the costs of cleanup for a particular  leak. If a
technology is based upon sampling at regular intervals, then it is
assumed that the leak just reached the sampling point at one
sampling time and is detected at the next sampling time (i.e., the
worst-case condition).
  For the determination of contaminated volume of soil due to
vapor migration, a vapor diffusion model is used. The model used
is an analytical diffusion equation for a continuous point source
in an infinite volume.3'3 For the model, it is assumed that the soil
is homogenous and isotropic and that 5% of the volume of a leak
(at 0.05 gal/hr) is volatilized. Other model inputs include liquid
density and the effective diffusion coefficient  for the compound
of interest.  The diffusion coefficient of a compound in air is cor-
rected for the available soil porosity (assumed to be 30%) and for
tortuosity.3 The model was run  for hexane, octane and  xylene
(major components of gasoline) and also  for trichloroethylcne
(TCE). The detection times for hexane, octane and xylene were
averaged and used to represent the detection times for gasoline at
various  limits of detection. The detection  times for TCE were
within 30% of the average value of hexane, xylene and octane.
  Due to  the complexity  of the subsurface  environment  and
unknown nature of the pure or concentrated  organic  solvent
migrating in the vadose zone, rational approaches were used to
estimate  the extent of soil contamination due to leachate migra-
tion. Two types of rational approach were selected:
• Modified Darcy's Law
• Solution Routing Technique
  To simplify the calculation,  the  following assumptions were
made:
• Soil beneath the LUST is anisotropic and only one type of soil
  is involved
• Two extremes of leak patterns, i.e., one leak opening and mul-
  tiple leak openings (i.e.,  assuming leaks occur throughout the
  entire horizontal projection area of the tank)
• The horizontal extent of migration is calculated based on the
  ratio of horizontal permeability to vertical permeability
• For organic solvents  lighter  than water (e.g., gasoline),  the
  vertical extent  of contamination is limited by the groundwater
  table

Modified Darcy's Law
  Darcy's law (i.e., v =  ki, where v, k and i are flow velocity,
permeability and hydraulic gradient, respectively) needs  to be
modified for the unsaturated condition of the vadose zone. The
equation can be modified by relating the degree of saturation to
the relative permeability (ratio of effective permeability to  ab-
solute or saturated permeability). Degree of saturation is defined
as the solvent occupied soil pore volume  to the total soil pore
volume. In tht subject case, two types of solvents are involved-
water moisture and leaking organic (e.g., gasoline). Since  the
water moisture in the  vadose zone, below the root zone area, is
mainly affected by infiltrating  precipitation,  fluctuation of the
degree of saturation may occur. If there is no pavement on the
surface of the ground, water moisture fluctuation is assumed to
be between "wilting point" under  dry  weather  conditions and
"field capacity"  under wet weather  conditions. If there  is pave-
ment above the tank,  a moisture content equivalent to the mid-
point between field capacity and wilting point is assumed. The ex-
tent of soil contamination is derived as follows:
  For one leak opening:
    Volume of Contaminated Soil
     = VD
     = Conic volume
        T
                      (Kht)2(Kvt)
                                                         (6)
           where:
    vt < h;
     Kv  =
                       vertical permeability of leaking solvent
                Kf,  =  horizontal permeability  of leaking solvent  (usually
                       Kh =  1 to 4 Kv)
                 h  =  distance  between tank  bottom and  groundwater
                       table
                 t  =  leak period
             (For gasoline, assume Kv = Kn)

           Solution Routing Technique
             The solution routing technique is based on the premise that the
           infiltrating solvent fills the voids in  a  layer of the  soil to field
256    UNDERGROUND LEAKING TANKS

-------
capacity before migrating to the layer beneath. It is again assumed
that, under the no pavement condition, the moisture content in-
the soil fluctuates between field capacity and wilting point. With
the existence  of pavement,  the  available  solvent  absorption
capacity can be calculated as follows:
     A  =  Absorption Capacity (volume %)
         =  F - (F + W)
                    2
         where
      F  =  field capacity (in %)
     W  =  wilting point (in %)
(7)
Therefore:
     VF  =
             vt
(8)
             A
         where:
     VF  =  volume of contaminated soil
      V  —  leak rate (in terms of volume/time)
       t  =  leak period

  The above three equations (Eqs. 6 to 8) are  applicable only
when the vertical extent of migration is above the groundwater
table. For floatable chemicals (e.g., gasoline), the extent of soil
contamination, VG, will be limited by the depth  of groundwater
table as shown  below:
  For one leak opening:
      G  =
                 kv
                                                         (9)
  Since either the  leak  source  or the soil characteristic (i.e.,
permeability) may become the limiting factor for controlling the
rate of migration,  the volume  of the contaminated soil was
selected based on the smallest value among VD, VF and VG.
  For groundwater contamination, it  is assumed that the con-
taminant is less dense than water and will float on the surface of
the groundwater. Furthermore, the spread of the contaminant  on
the groundwater surface is radial and equal in all directions. The
thickness of the contaminant on the top of the groundwater is as-
sumed to be equivalent to 2 in. of product. The radial spread of
this layer then is used to determine the point in time at which the
leak can be detected by groundwater monitoring.


EFFECTIVENESS ANALYSIS
  Effectiveness is a combination of qualitative and quantitative
non-cost measures. Different methodologies of preference can be
applied in assessing effectiveness.  Our approach was based  on
evaluation of the following 13 non-cost measures:

• Accuracy 0 =  01)
• Detection  Limit/Sensitivity (j = 02)
• Observation Frequency (j  = 03)
• Application Range (Types of Materials can be measured)
  (j = 04)
• Leak Rate Quantification (j = 05)
• Leak Location/Source Determination (j  =  06)
• Early Warning Capability  0 =  07)
• Down-time to the Tank (j = 08)
• Reliability/Completeness of Coverage (j  =  09)
• Environmental Effects (e.g., temperature,  evaporation, pres-
  sure, vapor pockets,  groundwater  conditions,  geologic/soil
  conditions, contamination) (j = 10)
• Health Effects (j  = 11)
• Constructability (e.g.,  ease of implementation,  space limita-
  tion, installation time) (j  =  12)
• Special Limitations (e.g., weather, damage to tank/pipe,  re-
  stricted to certain tank types or locations and any other limi-
  tations not included above (j = 13)
  Results of the effectiveness evaluation of underground tank de-
tectors/monitors are shown in Table 4. Accuracy is defined as the
deviation of the test results from the true leak or non-leak condi-
tions.  Among  the control technologies  evaluated,  interstitial
monitoring (i.e., i  = 7-9), tightness test by liquid-level measure-
ment (i.e., i  = 3), tracer techniques (i.e.,  i  = 6),  vadose zone
vapor monitoring (i.e., i = 12-19) and vadose zone liquid sam-
pling and laboratory analysis (i.e., i =  29, 30) are believed to be
better techniques in terms of accuracy. Detection limit/sensitivity
is defined as the lowest leak rate or the smallest detectable change
which can be detected by the detector/monitor. Again, all the in-
terstitial monitoring techniques, i.e., i =  7-9)  and  some of the
tightness tests (i.e., i = 3,6) and vadose zone vapor monitors (i.e.,
i  =  15-17, 19) can achieve desirable levels (e.g., less than 0.05
gal/hr or 1 mg/1) of results.
  Observation frequency represents, practically, how often the
technique can detect  leaks. If the technique can be  applied con-
tinuously and automatically, a highest rating (i.e.,  AJJ  =  3) is
given.
  Application range refers to the types of stored material that can
be measured. If there is virtually no material limitation, a highest
rating is assigned (as  shown in Table 4).
  Leak rate quantification is rated based on the following results:
0 =  leak rate determination impossible, 1 = very rough guess of
leak  rate, 2 = leak rate quantification based on rough guess by
calculation and 3 = quantitative measurement of leak rates. Leak
location or leak source determination (i.e., j = 06) is defined by
the following: 0 =  impossible, 1  = only part of tank and/or pip-
ing,  2  =  all of tank versus all of piping and 3 = exact location.
None of the detector/monitors evaluated  could  be rated 3  in
terms of leak location/source determination. Some liquid-level
tightness tests (i.e., i  = 3) are the only type of technique that can
differentiate between tank and pipe leaks (i.e., AJJ = 2).
  Early warning capability (i.e., j = 7) is one of the  most impor-
tant  factors for effectiveness rating. Interstitial monitoring tech-
niques are  found to be the best techniques for early warning of
tank  leaks. Downtime  to the tank (j =  08) during testing,
although not directly related to the leak detection, is considered
because of its effects on normal tank operation.
  Tightness tests are found to be the only type of technique which
require significant downtime (i.e., a few hours to a day). Reliabil-
ity and completeness of coverage (i.e., j  = 09) include downtime
of the technique, level of confidence and radius of distance the
detecting equipment can cover. Lower reliability/completeness
ratings were found for most of the techniques applied outside the
tanks.
  Environmental effects (i.e., j  = 10) which may affect the test
results also were  evaluated. Temperature changes,  evaporation
during testing, barometric  pressure changes, existence of vapor
pockets inside tanks, groundwater conditions, geologic/soil con-
ditions  and  contamination  are major  environmental  effects
evaluated.  Results have shown  that interstitial monitoring and
tightness tests are less affected by environmental effects.
  Health effects (j  =  11)  was  also selected for evaluation,
because,  consideration of this  factor is required for  the use of
cleanup related control technologies by paragraph  300.68(h) of
the NCP. Almost all techniques,  with the exception  of sensory
monitoring (i =  10,  11)  evaluated to not pose  any  significant
health effects.
  Constructability (j =  12), which refers to ease of implementa-
tion, space limitation and installation time required, is roughly
rated equal for most of the techniques evaluated.
                                                                                      UNDERGROUND LEAKING TANKS   257

-------
                                                              Table 4
                                Effectiveness Evaluation of Underground Tank Leak Detectors and Monitors
                                              (Values of the First-Level Rating Factor, A,j)




TYPE Of DETECTOR/MONITOR

INVENTORY
CONTROL


TIGHTNESS
TESTING






INTERSTITIAL
MONITORING



SENSORY
MONITORING










VADOSE &
AMBIENT
AIR MONITORING










VADOSE
LIQUID
MONITORING

Manual

Automatic
Liquid-Level
Measurements
Acoustic
Measurements

Pressure
Tests
Tracer
Techniques
Liquid
Monitoring
Vapor
Monitoring
Pressure
Monitoring
Visual
Monitoring

Sense of
Smell
Adsorption
Sensitive
Resistors
Combustible
Gas
Detectors
Synthetic
Filaments

Photo-ionization
Gas Detectors
Flame-lomzation
Gas Detectors
Infrared
Analyzers
Draeger
Tubes
Sample Traps
with Lab Analysis
Synthetic
Filament Devices

Electrical
Conductivity
Sensors

)" i = i = Is ) =
01 02 05 04 05
1112)1

222)51/2
5 ) ) 1 ) 3

422150


521/2150

652/5120

7)5550

85552/50

953550

10 1 1 2 1 0


11 1 1 2 1 0


12 2/3 1 2/5 2/5 0


12 2/3 2 2/3 2/3 0

14 2/5 1 2/52 0


15 2/3 2/3 2/3 2/3 0

16 2/3 2/5 2/5 2/5 0

17 2/3 2/3 2/3 2/3 0

IB 2/3 2 1 2/50

19 2/3 2/31 3 0

20 2 1 3 2 0



21 2 2 3 2/50

j =
06
0

0
1/2

0


0

0/1

0

0

0

0/1


0/1


0/1


0.1

0/1


0/1

0/1

0/1

0/1

0/1

0/1



0/1
Alr
) = 1 =
07 08
1 3

2 3
1/2 1

1/2 1


1/2 1

1 2

3 3

) 5

5 5

0/1 5


0/1 5


1/2 3


1 3

1/2 3


1 5

1 5

1 3

0/1 3

0/1 5

1/2 3



1/2 3

i = i = ) = 1 =
09 10 11 12
2133

21)3
2/3 2/3 3 2/3

2/3 2/3 3 J


2/3 2/3 3 2

2 2/3 2/3 2

3)55

35)3

3352

1225


1223


12)2/3


12)2/5

1252/3


1232/3

1232/3

1232/3

1232/5

12)2/3

1 2 3 2/>



1132/3

j =
1)
2

3
2

2


2

2

3

}

2

1


1


2


2

2


2

2

2

2

2

2



2


TOTAL
23

28.5
29.5

23.5


21

23.0

3)

32.5

31

18


18


24


24.5

23.5


25

25

25

22.5

23.5

23.5



24
                                                             (continued)
258    UNDERGROUND LEAKING TANKS

-------
Table 4 (continued)


j - i - i -
TYPE OF DETECTOR/MONITOR 01 02 03
Thermal
Conductivity
Sensors 22 2 2 3
Contact
Sensitive
Coated Wires 23 2 2 3
Psychrometers 24 2 2 3
VADOSE Tensiometers 25 2 1 3
LIQUID
MONITORING Neutron
(Continued) Moderation
Method 26 2 2 1
Soil
Resistance 27 2 1 1
Gamma Ray
Transmission 28 2 1 1
Vacuum-Pressure
Lysimeter
Sample/Lab
Analysis 29 2/3 1 1
Core Sample
Lab Analysis 30 2/3 2 1
In-Situ
GROUNDWATER Measurements 31 2 2 3
MONITORING
Sampling/
Lab Analysis 32 2 2 1
* i Evaluation Factor (see text for details)
01 Accuracy
02 Detection Limit/Sensitivity
03 Observation Frequency
04 Application Range
05 Leak Rate Quantification
06 Leak Location/Source Determination
07 Early Warning Capability
08 Down Town to the Tank
09 Reliability/Completeness of Coverage
10 Environmental Effects
11 Health Effects
12 Constructability
13 ^pBcisl Limitations





Other limitations 0 = 13) mainly refer to the effects of any fac-
tors which are not covered by the above listed factors (i.e., j = 1
to 12). Sensory monitoring techniques (i = 10, 1 1) are lowly rated
for j = 13 because they are unscientific and subjective. Ratings
for monitoring techniques applied outside the tanks are also
relatively low (Ajj = 2) in terms of other limitations (j = 13)
because weather effects such as precipitation and wind could eas-
ily change readings for some of these monitors.
13
The overall second-level rating factors (i.e...E rp were subjec-
tively assigned as 70% . That means the relative importance of ef-
fectiveness and cost is at a ratio of 7:3. The second-level rating
factors for the 13 factors evaluated above are selected as follows:
Aij
i— i— i— i— 1— i— i— i— i— i —


J- )- )- )- I- J- J- J- J- J -
04 05 06 07 08 09 10 11 12 13 TOTAL


2/3 0 0/1 1/23 1 2 3 2/3 2 25


2/3 0 0/1 1/23 1 1 3 2/3 2 24
1/2 0 0/1 1/23 1 2 3 2/3 2 24
1/2 0 0/1 1/23 1 2 3 2/3 2 23



200/10312322 20.

2 0 0/10/13 1 2 3 2/3 2 20.

200/10312322 19.



30/10/10312322 21.

31/20/10312322 23.

2 0 0/1 0/1 3 1 1/23 2/3 2 23


3 0 0/10 3 1 1/23 2/3 2 21.












j I";
J J
01 6%
02 8%
03 4%
04 3%
05 6%
06 4%
07 16%
08 4%
09 10%
10 4%
11 1%
12 3%
13 1%

Total 70%
Based on the above TJ values and AJJ values listed in
Pi values for effectiveness rating can be calculated, an
are shown in Table 5 .











5

5

5



5

5




5





























Tal
idtl

                             UNDERGROUND LEAKING TANKS   259

-------
                           Table 5
           Results of Effectiveness Evaluation (P| Values)

          Type of Detector/Monitor        P,
                   (i)                 (%)
1
2
3
4
5
6
7
B
9
10
11
12
13
14
IS
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
34.4
47.0
50.8
3B.6
35.9
37.6
60.0
59.5
SB. 7
25.8
25.8
3b.4
36.3
35.9
37.7
37.7
37.7
31.7
33.6
35.6
37.3
38.7
37.3
37.7
34.6
27.0
27.6
24.4
27.4
32.0
29.6
25.2
   Results  of effectiveness  evaluation  show that  interstitial
monitoring may be the best technique in terms of effectiveness.
Tightness test by liquid-level measurement is found to be the se-
cond choice for underground tank leak detection. The top  five
types of leak detection/monitoring techniques are shown below:

(1) Interstitial monitoring (all types of monitoring equipment)
(2) Tightness test by liquid-level measurements
(3) Automatic inventory control
(4) Thermal conductivity sensor for vadose  zone monitoring, or
    tightness test by acoustic measurements
(5) Vadose zone vapor or liquid monitoring by ionization, infra-
    red methods or  psychrometers, or tightness test by tracer
    techniques

COST COMPARISONS

Survey of Technology Costs
   For the estimation  of the capital costs of various leak detection
or monitoring  technologies,  vendors from all regions of the
United States were  surveyed.  Average cost and range  were
calculated based upon the survey. In the case where only  1 vendor
was found for a particular technology, the cost from that vendor
 was assumed to be typical for that particular technology. Opera-
 tion and maintenance (O&M) costs were calculated based upon
 best professional judgment. Tables 6 and 7 summarize the typical
 capital and annual costs of each control technology (based on
 1986 prices) for vapor detection, liquid detection and tank integri-
 ty testing. When  using these techniques for interstitial monitoring
 (between primary  and secondary  containment),  an additional
 capital cost will be incurred for the secondary containment. This
 cost is typically  in the range of $3,300 to  $6,000 for a  10,000
 gallon tank. The  costs  from this table are used to calculate the an-
 nualized costs for each control technology.

 Cost Ratings
   Cost comparisons for leak detection/monitoring are conducted
 in two ways. One is based purely on the cost required for the im-
 plementation and O&M  of the technologies. The  other includes
 the potential cleanup costs or costs which might result because of
 the inherent disadvantages of the technologies. In this section, the
 former type of cost comparison is presented. The latter  type of
 cost comparison  is site-specific and will be discussed in the  Cost-
 Effectiveness Analysis section below.
   Table  8  presents  results  of  cost  evaluation  based on
 methodology  described  previously. Assumptions  used for the
 calculation of annual costs are shown at the end of the table. As
 mentioned previously, a  30"%  rating was assigned to the second-
 level rating factor (rp. As can be expected, manual or sensory
 methods, without the  use of equipment, have higher P;  values.
 For non-continuous detection/monitoring techniques (e.g., tight-
 ness tests) costs are greatly affected by the measurement frequen-
 cies. In order to compare the effects of measurement frequencies,
 weekly,  monthly, quarterly, semi-yearly and yearly frequencies
 are selected.

 CONCLUSIONS OF COST-EFFECTIVE
 ANALYSIS STUDY

Evaluation Based on Technology
  Table 9 shows  the results of the cost-effectiveness evaluation
for various detectors and monitors. Based on this evaluation, it is
found that the interstitial monitoring techniques appear to be the
best alternatives for underground tank leak detection/monitor-
ing. Tightness test by liquid-level measurements may be the se-
cond most cost-effective alternative approach if test frequencies
are longer than  bi-yearly. Automatic inventory control is the next
alternative for  selection. Neutron moderation  and gamma ray
transmission techniques are not cost-effective and should not be
considered for underground tank leak detection/monitoring.

Evaluation Based on Technology and Liability
  A "case study" is presented for a number of leak detection/
monitoring technologies for the evaluation of cost-effectiveness
based on both technology and future liability. The assumptions
are that a 10,000 gal gasoline tank begins leaking at a point in time
such that the leak is detected 10 years after installation. The ex-
tent of soil and groundwater contamination were determined by
the methods discussed previously. Table 10 presents the results of
our case study where a  leak is detected 10 years after installation.
Cleanup  and  annual costs  are assumed to be constant  for the
10-year  period (expressed in 1986 dollars). Table  11 presents a
summary of information  used to produce the 10-year cost totals
in Table 10. This summary is based upon  information from the
nationwide survey of technology costs. This exercise suggests that
interstitial space leak detection and vadose zone vapor monitoring
(for  volatile  compounds only)  are  the most cost effective
technologies over  this representative time span.
260   UNDERGROUND LEAKING TANKS

-------
                                                  Table 6
              Costs of Liquid Detection and Monitoring Techniques and Tank Tightness Tests
Principle of Device/Technique
Electrical Conductivity Sensors
Thermal Conductivity Sensors
Contact-Sensitive Coated Wires
Contact Sensitive Filaments
(e.g. polypropylene)
Tensiometers
Survey Instrument
Psychrometers (Soil probe)
Survey Instrument
Geophysical Techniques
(Soil Resistivity)
Vadose Liquid Sampling and
Lab Analysis (Lysimeter)
Groundwater Sampling and
Lab Analysis (bailer)
Vadose Zone Well
(One 15 ft. well, 2" dia.)
Groundwater Well
(One 60 ft. well, 2" dia.')
Tank Tightness Tests
Measurement Device/
Sensor/Sampler^ TecHnique*
540 1,500
(450 - 800) (1,400 - 1,600)
900 1,450
230 1,400
300

40
(20 - 155)
35 1,930
1,930
(3,300 - 8,000)
500 142/test
(325 - 1,000) (45 - 550 per test)
150 142/test
(50 - 200) (45 - 550 per test)
400
(300 - 700)
1,400
(1,100 -2,100)
600/test
(400 - 800 per test)
Annual Costs
(O&M)2
400
400
400
400

450

450
400
450 (+142/test)
600 (+142/test)
„

„

600/test
1.  Based on average or typical prices from vendor survey. Price range is in parentheses.
2.  Costs assigned.
                                                  Table 7
                         Costs of Vapor/Gas Detection and Monitoring Techniques
Principle of Device/Technique
Odor Detection
Adsorbtion Sensitive
Resistor Probe
Metal-Oxide
Semiconductor Probe
Synthetic Filament Probe
(e.g. polypropylene)
Photoionization Detector
Survey Instrument
Flame lonization Detector
Survey Instrument
Combustible Gas Meter
Survey Instrument
Vapor Sampling and
Lab Analysis (Lysimeter)
Vadose Zone Well
(One 15 ft. well, 2" dia.)
Measurement Device/
Sensor/Sampler^ Technique*
.-
325 1,500
(100 550)
600 1,350
(480-760)
300
3,950
(2,900 - 5,000)
5,200
1,100
(800 - 1,600)
500 142/test
(325 - 1,000) (45 - 550 per test)
400
(300 - 700)
Annual Costs
(P&M)2
400
400
400
400
400
400
400
450 (+142/test)
—
1. Based on average or typical prices from vendor survey. Price range is in parentheses.
2. Costs assigned.
                                                                                  UNDERGROUND LEAKING TANKS     261

-------
                                                             Table 8
                                                Results of Cost Evaluation (Pj Values)*
TYPE OF
INVENTORY
CONTKOL



TIGHTNESS
TESTING











Annualized
i Cost An = KJ ri
DETECTOK/MONITOK ($) (x) log(lbVx) (in%)
Manual 1 400 2.40 J JO
(Weekly)
Automatic 2 620 2.21 J JO
Liquid-Level
Measurements
(Monthly) 3A 7,200 1.14 3 JO
Liquid-Level
Measurements
(Quarterly) JB 2,400 1.62 3 JO
Liquid-Level
Measurements
(Bi-yearl>) 3C 1,200 1.92 5 50
Liquid-Level
Measurements
(Yearly) }D 600 2.22 5 JO
Acoustic
Measurements
(Monthly) iiA 7,200 l.U 5 JO
Acoustic
Measurements
(Quarterly) 48 2,400 1.62 3 JO
Acoustic
Measurements
(Si-yearly) 4C 1,200 1.92 J JO
Acoustic
Measurements
(Yearly) 4D 600 2.22 5 30
Pressure
Tests
(Monthly) S A 7,200 1.14 J M)
Pressure
Tests
(Quarterly) ^8 2,400 1.62 J 30
Pressure
Tests
(Si-yearly) SC 1,200 1.92 3 JO
Pressure
Tests
(Yearl>) SD 600 2.22 3 JO
Tracer
Techniques
(Monthly) 6A S.270 1.28 3 JQ
Tracer
Techniques
(Quarterly) 68 1,910 1.72 J 3Q
Tracer
Techniques
(Bi-yearl>) 6C 1,070 1.97 J JQ,
Tracer
Techniques
(Yearly) 6D 6SO 2.19 3 jrj
pi
(in%)
24.0
22.1
11.4
16.2
19.2
22.2
11.4
16.2
19.2
22.2
11.4
16.2
19.2
22.2
12.8
17.2
19.7
21.9
         * Sec Icxl for details.
                                                          (continued)
262    UNDERGROUND LEAKING TANKS

-------
Table 8 (continued)

Annualized
i Cost Ai; = Kj r:
TYPE OF DETECTOR/MONITOK ($) (x) log (ITP/x) (in %)


INTERSTITIAL
MONITORING
(Continuous)



SENSORY
MONITORING
(Weekly)




VADOSE &
AMBIENT
AIR MONITORING























VADOSE
LIQUID
MONITORING





Liquid
Monitoring 7 620 2.21 3 30
Vapor
Monitoring 8 620 2.21 3 30

Pressure
Monitoring 9 420 2.38 3 30
Visual
Monitoring 10 400 2.40 3 30

Sense of
Smell 11 400 2.40 3 30
Adsorption
Sensitive
Resistors 12 1,700 1.77 3 30
Combustible
Gas
Detectors 13 800 2.10 3 30
Synthetic
Filaments 14 820 2.09 3 30
Photo-ionization
Gas Detectors
(Monthly) 15 1,240 1.91 3 30
Flame-Ionization
Gas Detectors
(Monthly) 16 1,440 1.84 3 30
Infrared
Analyzers
(Monthly) 17 1,200 1.92 3 30
Draeger
Tubes
(Monthly) 18 830 2.08 3 30
Sample Traps with
Lab Analysis
(Monthly) 19 910 2.04 3 30
Synthetic
Filament
Devices 20 820 2.09 3 30
Electrical
Conductivity
Sensors 21 1,830 1.74 3 30

Thermal
Conductivity
Sensors 22 2,020 1.69 3 30
Contact
Sensitive
Coated Wires 23 1,600 1.80 3 30
Psychrometers 24 800 2.10 3 30

(ink)

22.1

22.1


23.8

24.0


24.0


17.7


21.0

20.9


19.1


18.4


19.2


20.8


20.4


20.9


17.4



16.9


18.0
21.0
   (continued)
                              UNDERGROUND LEAKING TANKS    263

-------
                                                                Table 8 (continued)
Annualized
i Coat An =
TYPE OF DETECTOR/MONITOR ($) (x) log (lbs/x)


VAOOSE
LIQUID
MONITORING
(Continued)


GROUNDWATER
MONITORING
Tensiometer§ 25 710 2.15
Neutron
Moderation
Method
(Monthly) 26 20,000 0.70
Soil
Resistance 27 S.J50 1.47
Gamma Ray
Transmission
(Monthly) 28 20,000 0.70
Vacuum-Pressure
Lyslmeter Sampling/
Lab Analysis
(Monthly) 29 1,120 1.95
Core Sample/
Lab Analysis
(Monthly) 30 1 , 400 1.85
In-Situ
Measurements 51 2,000 1.70
Sampling/
Lab Analysis
(Monthly) 32 1,650 1 . 7H
Kj ri PI
' (in k) (in %)
J 30 21.5
J JO 7.0
J JO 14.7
J JO 7.0
J JO 19.5
J JO 18.5
J JO 17.0
3 JO 17.8
          Assumptions:
                1.    Useful life of detectors/monitors is 1*> years.
                2.    Interest rate for amortization = 12%.
                3.    Tour in-situ monitoring wells are used for calculation for both vadose and groundwater zones monitoring.
                4.    four monitoring wells or monitoring points are used for tracer techniques.
264     UNDERGROUND LEAKING TANKS

-------
              Table 9
Results of Cost-Effectiveness Evaluation
TYPE OF
INVENTORY
CONTROL



TIGHTNESS
TESTING











DETECTOR/MONITOR
Manual
(Weekly)
Automatic
Liquid-Level
Measurements
(Monthly)
Liquid-Level
Measurements
(Quarterly)
Liquid-Level
Measurements
(Bi-yearly)
Liquid-Level
Measurements
(Yearly)
Acoustic
Measurements
(Monthly)
Acoustic
Measurements
(Quarterly)
Acoustic
Measurements
(Bi-yearly)
Acoustic
Measurements
(Yearly)
Pressure
Tests
(Monthly)
Pressure
Tests
(Quarterly)
Pressure
Tests
(Bi-yearly)
Pressure
Tests
(Yearly)
Tracer
Techniques
(Monthly)
Tracer
Techniques
(Quarterly)
Tracer
Techniques
(Bi-yearly)
Tracer
Techniques
(Yearly)
i
1
2
3A
3B
3C
3D
4A
4B
4C
4D
5A
5B
5C
5D
6A
6B
6C
6D
Effectiveness
Pi
(in %)
34.4
47.0
50.8
50.8
50.8
50.8
38.6
38.6
38.6
38.6
35.9
35.9
35.9
35.9
37.6
37.6
37.6
37.6
Cost
Pi
(in%)
24.0
22.1
11.4
16.2
19.2
22.2
11.4
16.2
19.2
22.2
11.4
16.2
19.2
22.2
12.8
17.2
19.7
21.9
Overall
Pi
(in%)
58.4
69.1
62.2
67.0
70.0
73.0
50.0
54.8
57.8
60.8
47.3
52.1
55.1
58.1
50.4
54.8
57.3
59.5
             (continued)
                                          UNDERGROUND LEAKING TANKS    265

-------
                                                  Table 9 (continued)
TYPE OF

INTERSTITIAL
MONITORING
(Continuous)
SENSORY
MONITORING
(Weekly)

1
DETECTOR/MONITOR
Liquid
Monitoring 7
Vapor
Monitoring fl
Pressure
Monitoring 9
Visual
Monitoring 10
Sense of
Smell 11
Adsorption
Sensitive
Resistors 12
VADOSE 4 Combustible
AMBIENT Gas
AIR MONITORING Detectors 13







VAOOSE
LIQUID
MONITORING


Synthetic
Filaments 14
Photo-ionization
Gas Detectors
(Monthly) IS
Flame-Ionization
Gas Detectors
(Monthly) 16
Infrared
Analyzers
(Monthly) 17
Draeger
Tubes
(Monthly) 18
Sample Traps
with Lab Analysis
(Monthly) 19
Synthetic
Filament Devices 20
Electrical
Conductivity
Sensors 21
Thermal
Conductivity
Sensors 22
Contact
Sensitive
Coated Wires 2»
Psychrometers 24
Effectiveness
(ink)
60.0
59.5
58.7
25.8
2^.8
36.4
36.3
35.9
37.7
37.7
37.7
31.7
33.6
35.6
37.3
38.7
37.3
37.7
Coat
(ink)
22.1
22.1
23.8
24.0
24.0
17.7
21.0
20.9
19.1
18.4
19.2
20.8
20.4
20.9
17.4
16.9
18.0
21.0
Overall
(inb)
82.1
81.6
82.5
49.8
49.8
54.1
57.3
56.8
56.8
56.1
56.9
52.5
54.0
56.5
54.7
55.6
55.3
58.7
                                                     (continued)
266   UNDERGROUND LEAKING TANKS

-------
                          Table 9 (continued)


TYPE OF






VADOSE
LIQUID
MONITORING
(Continued)









Effectiveness Cost
i PI Pj
DETECTOR/MONITOR (in %) (in %)
Tensiometers 25 34.6 21.5
Neutron
Moderation
Method
(Monthly) 26 27.0 7.0
Soil
Resistance 27 27.6 14.7

Gamma Ray
Transmission
(Monthly) 28 24.4 7.0
Vacuum-Pressure
Lysimeter Sampling/
Lab Analysis
(Monthly) 29 27.4 19.5
Core Sample/
Lab Analysis
(Monthly) 30 32.0 18.5
In-Situ '
GROUNDWATER Measurements 31 29.6 17.0
MONITORING




Sampling/
Lab Analysis
(Monthly) 32 25.2 17.8
Overall
Pj
(in%)
56.1



34.0

42.3



31.4



46.9


50.5

46.6



43.0
                                Table 10
          10-Year Cost-Effectiveness Case-Study of Technologies

                                                          Ten-Year
       Technology                                    Cumulative Cost1

 Manual Inventory Reconciliation with
 Annual Tank Integrity Testing                               25,873

 Monthly Tank Integrity Testing                              75,573
 6-Month Tank Integrity Testing                              20,455

 Interstitial Space Liquid Detection (Liner System)
    -   Contact Sensitive Coated Wire                        11,117
       Electrical Conductivity Sensor                        11,587
       Thermal Conductivity Sensor                          11,897

 Interstitial Space Vapor Detection (Double-Walled Tank)
       Adsorption Sensitive Resistor  Sensor                   11,797
       MOS Combustible Gas Sensor                          11,922

 Vadose Zone Liquid Monitoring
       Electrical Conductivity Sensor                        16,810
    -   Lysimeter with Lab Analysis                          15,770
       (3-Month Interval)

 Vadose Zone Vapor Monitoring
       MOS Combustible Gas Sensor                           9,405
       Photoionization Survey Instrument                     11,405
       (Monthly)

 Groundwater Monitoring
       Thermal Conductivity  Sensor                          180,968
       Well Bailing and Lab Analysis                          193,078
       (3-Month Interval)
1. Cumulative costs based upon complete payment of capital cost in first year and constant an-
  nual costs. Model systems for 10,000 gallon tank and a 0.05 gph leak rate.
                                                               UNDERGROUND LEAKING TANKS     267

-------
                                                                    Table II
                                                Costs of Technologies and Contamination Cleanup
                                                                       Cost of
                                                                    Technology ($)
                                                                  Capital   Annual

                                                                             1,000
                 Volume of
             Contaminated Soil
           before Detection (yd*)

                   66.8
                                                                                                Co»t of Tank and
                                                                                                Soll/GroundwBter
                                                                                               5.5
                                                                                               33.4
                                           15,873
                                            2,573
                                            7,455
Technology

Manual Inventory Reconciliation (Weekly)
(With Annual Tank Integrity Test)

Tank Tightness Test (liquid level)
     Monthly Testing                                  --    7,200
     6-month Testing                                 --    1,200

Interstitial Space Liquid Detection'"
(Liner System with Carbon Steel Tank)
     Contact Sensitive Coated Wire                 5,567      400             Backfill                  1,610
     Electrical Conductivity Sensor                 5,977      400             Backfill                  1,610
     Thermal Conductivity Sensor                   6,287      400             Backfill                  1,610

Interstitial Space Vapor Detection^'
(Double-Welled FRP Tank)
     Adsorbtion Sensitive Resister  Sensor            7,797      400
     Combustible Gas Sensor                        7,922      400
     (i.e., metal-oxide semiconductor, MOS)

Vadose Zone Liquid Monitoring
(Two 20  ft. wells)
     Electrical Conductivity Sensor                 2,940      400             47.2                     9,870
     Lysimeter Sample with Analysis                1,400      450             47.2                     9,870
     (at 3-month Intervals)

Vadose Zone Vapor Monitoring
(Two 20  ft. wells)
     Combustible Gas Sensor (MOS)                 2,850      400              5.4                     2,555
     Photoionization Survey Instrument (Monthly)     4,850      400              5.4                     2,555

Ground Water Monitoring
(One 60  ft. well)
      Thermal Conductivity Sensor                   3,440      400             68.1                   173,528
      Well Bailing and Lab Analysis                  1,550    1,200             68.1                   173,528
      ( At 3-month  Intervals)


I  The cost of liner system and a backfill monitoring well is included in the capital cost (SO37)
2 The additional cost of a double-walled Fibreglass Reinforced Plastic (FRP) tank is included in the capital cost (55.972)
3  Groundwaler level is assumed to be 30 feel below (he bottom of the lank
4 Leak rate is 0.05 gph (average rale) and the soil is Sandy Soil (Because inventory reconciliation is only sensitive to a leak of about 0 1 to 0 2 gph.
  it is assumed that yearly tank integrity testing will detect a leak 2 years after it has started and the leak is leu than 0 5 gph for the first year.)
REFERENCES
1. Barcellona, M.J.,  Helfrich, J.A.,  Garske, E.E. and Gibb, J.P., "A
   Laboratory  Evaluation  of Groundwater Sampling Mechanisms,"
   Ground Water Monitoring Rev., 4, Spring 1984, 32-41.
2. Nielsen,  D.M.  and Yeates,  G.L., "A  Comparison  of Sampling
   Mechanisms Available for Small Diameter Ground Water Monitor-
   ing Wells," Ground Water Monitoring Rev., S, Spring  1985, 93-98.
3. Robbins, G.A. and Gemmell, M.M., "Factors Requiring Resolution
   in Installing Vadose Zone Monitoring Systems," Ground Water Mon-
   itoring Rev., 5, Spring 1985, 75-80.
4. Askenaizer, D.J., Barcikowski,  W.,  Jennings,  K.V.B.  and Sarna,
   J.E.,  "Development  of A Compliance Program for Underground
   Tanks Containing Hazardous Substances,"  Environmental  Science
   and Engineering Report No. 85-61, UCLA, Los Angeles. CA, Mar.
   1985.
5. Crank,  J., The Mathematics of Diffusion, 2nd edition. Clarendon
   Press, Oxford, England, 1975.
  268    UNDERGROUND LEAKING TANKS

-------
                      Underground  Storage System  Assessment,
                                     Testing  and  Remediation

                                               Scott J. Adamowski
                                            Angelo J. Caracciolo, III
                                            G. David Knowles, P.E.
                                        O'Brien & Gere Engineers, Inc.
                                               Syracuse,  New York
INTRODUCTION
  It is estimated that groundwater supplies 24% of the Nation's
domestic, agricultural and industrial water. Reliance on ground-
water has increased greatly over the past few decades with over
50% of the U.S. population utilizing it  for its potable water
supply.
  Groundwater reserves are threatened by a variety of contami-
nant sources. A major, though often overlooked, source of con-
tamination is leaking underground storage tanks. The U.S. EPA
estimates that nearly 11 million gallons of gasoline alone may be
leaking from underground storage  tanks.  Gasoline, however, is
only one of the substances stored in underground  tanks. Other
substances include hazardous materials and wastes and  nearly
every type of petroleum product. Most of  these storage tanks
are constructed of bare steel, which the American Petroleum In-
stitute estimates to have only a 15-year service life before leaking
will occur.
  To address this situation, governments at the state, local and
federal levels have begun to implement legislation to govern new
and existing underground storage systems. These programs im-
plement a registration system for existing tanks to define the ex-
tent of underground storage. Design standards also are being im-
plemented for new underground facilities, as well as a  tank in-
tegrity testing program for existing underground storage systems.
  In this paper, the authors discuss the regulatory aspects of the
underground tank issue  and the method  by which O'Brien &
Gere conducts underground storage system evaluations, includ-
ing site assessment and priority modeling,  leak detection, recov-
ery of product and site remediation. The paper will begin with an
overview of existing and proposed  federal legislation governing
underground storage of petroleum products, hazardous materials
and wastes.  The paper then will discuss the methodology for pre-
dicting  leakage from, and for the testing and monitoring of,
underground storage systems.
  O'Brien & Gere's approach to conducting a site assessment be-
gins with the gathering of tank and site-specific data and, using a
computer, modeling and prioritizing the tanks. Once an assess-
ment and priority list have been established, an action plan is
developed and system testing may begin. Various aspects of the
subsequent tank testing, using the Petro-Tite® method, will be
discussed in  detail.


REGULATORY REQUIREMENTS
  As the problem of leaking underground  storage tanks became
more apparent, legislation was adopted to prevent and control it.
In the late 1970s, the federal government and many state legisla-
tures adopted oil spill prevention laws which regulated oil storage
facilities. These regulations stated that facilities  having greater
than 42,000 gal of underground petroleum storage must develop
plans to prevent and control oil spills. During the early 1980s, the
federal government and several state agencies realized the wide-
spread problem  associated  with  the  underground storage  of
petroleum and other hazardous chemicals. Federal legislation was
subsequently developed governing petroleum storage of 1,100 gal
or more at both new and existing systems. The Hazardous and
Solid Waste Amendments  of 1984 (HSWA), which amended
RCRA, contain provisions for the regulation of underground
storage systems. The U.S. EPA has also proposed changes  to
Subtitle J of RCRA, which governs underground storage of haz-
ardous waste. The following sections describe those aspects of the
federal regulations which apply to underground storage tanks.

HAZARDOUS AND SOLID WASTE
AMENDMENTS OF 1984
  The HSWA of 1984 enacted Subtitle I—Regulation of Under-
ground Storage Tanks (RUST) as a part of  RCRA. Subtitle I
(RUST Program) regulates all petroleum products and all haz-
ardous substances as defined in CERCLA that  are stored  in
underground tanks. For the purpose of this regulation, an under-
ground tank is defined as any tank with 10% or more of its vol-
ume, including all attached piping, located below the ground sur-
face. For those  systems under the jurisdiction of RUST, the U.S.
EPA has  authorized several programs dealing with underground
storage systems. These programs include:

• Registration/Inventory: This underground tank registration/in-
  ventory program calls for registration of new and existing tanks
  with the designated state and/or local regulatory agencies. This
  has resulted in the proposed rule 40 CFR 280—Notification
  Requirements for Owners of Underground Storage Tanks. This
  rule describes the registration forms and the information which
  must be contained.
• U.S. EPA Regulations: The U.S. EPA will issue regulations
  regarding all underground storage  tanks.  These regulations
  shall contain the elements necessary to protect human health
  and the environment  and include, but not be limited to, re-
  quirements for:
  -leak detection or inventory control and tank testing
  -record keeping and reporting
  -corrective action
  -financial responsibility, as necessary or desirable, for correc-
   tive action and third party liability
  -closure
• Interim Standards: Interim standards for  the installation of
  new underground tanks will  be established. These standards
  will require that a new tank for the storage of regulated sub-
  stances may not be installed unless the tank:
                                                                                  UNDERGROUND LEAKING TANKS    269

-------
  -will prevent releases due to corrosion or structural failure for
   the operational life of the tank
  -is cathodically protected  against  corrosion, constructed  of
   non-corrosive material, steel clad with a non-corrosive ma-
   terial, or designed in a manner to prevent the release or threat-
   ened release of any stored substance
  -is constructed or lined  with material that is compatible with
   the substance to  be stored; however, a storage tank may  be
   installed without corrosion protection  if the soil resistivity  is
   12,000 ohm/cm or more
• Performance Standards: The  U.S. EPA  must  issue perma-
  nent performance standards for the installation of new under-
  ground systems. These standards shall include, but not be lim-
  ited to, design, construction, installation, release detection and
  compatibility standards.
• Stale Programs: RUST establishes that states may  administer
  the underground tank program in lieu of the U.S.  EPA. The
  states' requirements must be no less stringent than those estab-
  lished by the U.S.  EPA.
• Further Study: The U.S. EPA must conduct studies of petrol-
  eum and other regulated substances contained in underground
  tanks.
• Enforcement: RUST  provides  for the enforcement of  the
  established laws and regulations.

RCRA AMENDMENTS
  In addition to HSWA, the U.S. EPA proposes to amend regu-
lations under RCRA for control of tank systems storing or treat-
ing hazardous  waste.  These regulations provide for secondary
containment  of all  new underground tanks  storing  hazardous
wastes.
  The secondary containment systems must:

• Be designed, installed and operated to prevent waste or liquid
  from escaping to soil or water for the life of the tank
• Be capable of detecting and collecting waste or leakage until
  removal of material
• Be constructed or  lined with materials compatible with the
  waste and be of sufficient strength to prevent failure from pres-
  sure, climate, traffic and daily use
• Have an adequate base or foundation to resist settlement, com-
  pression and uplift
• Have a leak detection system capable of detecting leaks within
  24 hours of occurrence
• Be sloped or drained  to remove leaks, spills and precipitation
  with  provisions for accumulation  to be removed as soon as
  possible, but within  24 hours of release
• Be designed or operated to contain 110% of the design capac-
  ity of the largest tank within the containment boundary
• Prevent run-on or infiltration of precipitation unless the collec-
  tion system has excess capacity  over the 110% to contain pre-
  cipitation from the 25-year, 24-hour rainstorm

  Existing tanks can avoid the secondary containment require-
ments by developing a groundwater monitoring system combined
with a scheduled tank testing program.
  The above regulations clearly show that the federal govern-
ment has implemented a major program to control the problem
of leaking underground storage tanks. In addition to these federal
constraints, states also have begun to administer regulations for
underground storage systems. In most instances, these are more
stringent than the federal regulations.

UNDERGROUND STORAGE SYSTEM  ASSESSMENTS
  O'Brien & Gere has developed an  approach for evaluating the
integrity of an underground  storage system which combines the
use of a mathematical model for predicting environmental haz-
ard with the procedures for the testing and monitoring of under-
ground storage tanks. When this approach is utilized with back-
ground information relating to the design of storage systems and
the development of environmental remediation programs, the re-
sult is a complete program for assessing underground storage sys-
tems. The approach utilized consists of four basic tasks which
are described in detail in the sections below.

Task 1—Facility Survey:
  A survey of an underground storage system consists of obtain-
ing site-specific data and using these data to determine a system's
potential to corrode and,  hence, leak.  The survey is conducted
utilizing a questionnaire. The questionnaire is  structured to be
easily modified and used at facilities in various stales. The  data
pertain  to three major areas of coverage: (1) regulatory infor-
mation, (2)  tank data and (3) environmental data. The regula-
tory aspects of the questionnaire include the elements identified
by the U.S. EPA in the proposed rule of 40 CFR  Part 280—Notif-
ication Requirements for Owners of Underground Storage Tanks.
In instances where requirements stipulated by the state in which
the facility is located exceed the part 280 requirements, these slate
provisions also are addressed.
  Technical information to be obtained for each tank includes
tank design  features, such as dimensions and capacity, materials
of construction, location, depth of cover and  a  description of
pipes, pumps and other appropriate appurtenances. Information
regarding the  tanks' history such as age,  material stored,  past
leaks, repairs and previous testing also is obtained.
  To facilitate the prioritization, which is accomplished in Task
2, available  environmental information also is requested. These
data include site hydrogeologic and soil characteristics, distance
to nearest water supply, distance to nearest residence, adjacent
population densities  and other issues relating to  environmental
sensitivity. A site plan,  identifying the approximate location of
each tank, also is developed.
  To ensure consistency in data generation, a  set of user's in-
structions has been developed to accompany the  questionnaire.
These instructions include a concise explanation  of each entry on
the questionnaire and recommended terminology. The question-
naire then is completed by either O'Brien & Gere or facility per-
sonnel.  As the data are obtained, they are collated, entered into
a computer data file and a master list of underground tanks and
information is developed. Following completion  of  the master
list, the next task, ranking the tanks in  a prioritized order, is in-
itiated.

Task 2—Rank Tanks in Priority Order:
  Utilizing the information obtained and compiled in Task 1, the
storage systems are  ranked using a site-specific mathematical
model.  Based  upon the results of this prioritizaiion, recommen-
dations for further action  on each tank are identified within this
ranking. The priority ranking saves time by providing a concise
listing of those tanks which are likely to represent  existing or po-
tential problems.
  The  information necessary to rank the tanks is supplied on the
survey forms completed in Task 1. These data  are composed of
three basic types: (1) tank data,  (2) environmental data and (3)
leak hazard data. These tank data consist of physical data related
to the tank such as tank age, the tank's materials of construction
and any existing protective coatings or devices. Environmental
data include the relative location of groundwater and its users,
the relative location of surface water bodies and their users and
the types of soil surrounding the tank. The leak  hazard data con-
sist of the past history of the tank concerning  leaks and/or re-
270    UNDERGROUND LEAKING TANKS

-------
                            Table 1
             Underground Storage System Assessment
            Priority Ranking Model Data Descriptions

Data Parameters                 Description
Tank Data:
1. Tank age
2. Materials of construction

3. Internal protection

4. External protection

5. Cathodic protection
6. Secondary containment


Environmental Data:
1. Depth to groundwater

2. Distance to nearest ground-
  water well
3. Distance to nearest surface
  water body
4. Soil type surrounding tank


Leak Hazard Data:
1. Tank status

2. Tank repair frequency

3. Previous leak testing

4. Previous leaks detected
5. Material stored
-The API estimates that 50% of bare steel
 tanks leak after 15 years. Therefore, this
 is the most important factor for steel
 tanks.
-i.e., carbon steel, fiberglas, reinforced
 plastic, stainless steel.
-Coating and lining, such as glass, zinc or
 plastic.
Coatings and coverings such as FRP,
 asphalt or epoxy
Sacrificial anode or impressed current.
Such as double-walled tanks, concrete
 vaults or lined excavations.


-Depth of average groundwater surface
 below the ground surface.
-Distance to nearest potential groundwater
receptor.
-Potential lakes or streams to be impacted
 by a leak.
-Porosity and resistivity affect movement
 of leaked material and corrosion of tank.


-Is tank currently active or inactive,
 empty or full?
-History of past tank repairs and methods
 used.
-Test methods used (i.e., Petro-Tite® or
 static heat test) and results.
-Frequency of past leaks.
-Hazard ratings of the different materials
 were ranked according to such things as
 toxicity, carcinogenicity, flammability,
 etc.
 pairs, the hazard rating of materials stored in the tank and the
 status  of the tanks,  either active or  inactive. Table 1  lists the
 parameters  which are used to rank the tanks, descriptions and
 examples.
   The information for each tank regarding the data identified in
 Table 1 will be gathered through the use of the survey forms de-
 veloped during Task 1. For each parameter, various specific input
 variables are possible. The mathematical model assigns a value to
 each specific variable based on whether it increases or decreases
 the tank's potential  for leaking. The separate data points are
 weighted according to their importance in determining a tank's
 leak potential.
   To develop the weighting factor, tanks are divided  into two
 categories:  steel or  fiberglass  reinforced plastic  (FRP).  The
 weighting of tank parameters differs for these two types. For
 steel tanks, the age of the tank and the corrosion protection meth-
 ods employed are the most important parameters. For FRP tanks,
 which do not corrode due  to soil moisture, the installation meth-
 od, location of the tank with respect to traffic and groundwater
 or frost upheaval and compatibility with the material stored be-
 come the most important factors. After the weighting factors
 are applied, each input variable is assigned a specific value which
 is combined numerically to generate a total tank ranking value.
   The total tank ranking value represents a certain hazard rating
 for the tanks. Ranges of ranking values have been developed for
 each hazard rating based on empirical data and modified to facil-
 itate ease of use. The hazard ratings are divided into the follow-
 ing four categories:
 • Remove Tank
 • High Hazard Rating
 • Medium Hazard Rating
 • Low Hazard Rating

  Tanks previously tested and  information about other tanks
which O'Brien & Gere has worked with have been ranked using
the mathematical model. The ranking scores which were gener-
ated were  then combined with their leak  history to develop the
ranges  for the above four ratings. The model is used  on each
tank tested, and the outcome of  the test is used to refine the sen-
sitivity of the hazard rating ranges.
  The  mathematical model is incorporated into the computer
data base file developed in Task 1. In this way, as the informa-
tion gathered during Task 1 is entered into the computer file, the
program automatically assigns a  ranking score and hazard rating
to the tank. Using these individual ratings, a prioritized schedule
for the testing of the tanks is developed.

Task 3—Site and Tank Assessment:
  Based upon the finalized testing schedule established in Task 2,
an assessment of each site relative to its potential for having tank
testing performed is conducted. Following the assessment and any
necessary modifications, the tanks are tested following the sched-
ule developed in Task 2.
  Currently, several methods exist for testing underground tank
systems; however, there are only a limited number which meet the
stringent requirements of NFPA. Six methods which propose to
meet these requirements are:
• Petro-Tite®  (formerly Kent-Moore)
• ARCO-HTC
• Leak-Lokator
• Tank Auditor
• Homer EZY-CHEK
• Ainlay Tank Tegrity Tester
  The testing method used by O'Brien & Gere is the Petro-Trite®
method. A brief description of the method is given below.

Petro- Tite®  (formerly Kent-Moore) Method:
  The Petro-Tite®   method of testing underground storage sys-
tems consists of a hydrostatic test that compensates for tempera-
ture, pressure and viscosity variation and enables the detection of
leaks as small as O.t)l gal/hr. The Petro-Tite® test apparatus ex-
erts a  static pressure head on the tank using a standpipe filled
with the same liquid stored in the tank. A pump is used to circu-
late the liquid  and produce a uniform temperature throughout
the tank, and a thermal sensor monitors temperature changes to
account for expansion and contraction of the liquid. Volumetric
changes are then calculated due to these changes in temperature.
The Petro-Tite®  test accurately measures the amount of product
added to or removed from the standpipe to maintain a constant
head. By comparing the product added or drained with the volu-
metric variances anticipated due to temperature changes, it is pos-
sible to reliably detect a leak as small as 0.01 gal/hr. The method
compensates for tank-end wall deflection by elimination of de-
flection. This is accomplished by a pressure reduction induced
during the test. Due to the intricate nature of the test and the need
for a  skilled technician, the  Petro-Tite®  test requires  several
hours to complete.  In addition, the tank must be out of service,
and full, in order to conduct the test.
  To  perform the Petro-Tite®   test,  the following conditions
must be met:

• The tank must be  out of service and have one testing access
  port per  10,000 gal of tank capacity.
                                                                                         UNDERGROUND LEAKING TANKS    271

-------
• The access ports must be straight down into the tank, must be
  3 to 4 in. in diameter and must not contain a drop tube.
• The depth of groundwater in contact with the tank  must be
  known.
• The tanks must be able to be isolated from piping.
• Power supplies must be located within 100 ft of the tank.  The
  power available must consist of a  110 volt AC service with a
  20 amp breaker for tanks of 12,000 gal or less, or two separate
  110 volt AC circuits, one having a 20 amp breaker and one hav-
  ing a  30 amp breaker, for  tanks of  12,000- to  30,000-gai
  capacity.
• The viscosity of the material  must be such that the liquid can
  be pumped.
• No more than  1 in. of water is allowable in the bottom of the
  tank (unless testing with water). Sludge on the bottom of the
  tank may not be greater  than 1 in. in depth.
  Prior to initiating the tank testing activities, a survey of the site
will be conducted to determine if these conditions are met. After
the tanks have been  retrofitted  to meet the above criteria, the
tanks would be tested according to  the schedule determined in
Task  2. The initial test  utilized is  a  system test (tank and pipe-
lines) which  tests the entire tank system.  Should the results of
testing indicate  the system is not  leaking (less than the NFPA
0.05 gal/hr rate), the tank  system will be certified as non-leaking.
If the testing shows leakage, it would not be defined from  this
initial test whether the leak is in  the tank or  the pipeline. At this
time.  Task 4 would be initiated for those systems which are leak-
ing. Task 4,  which is described in the following section, is con-
ducted concurrently with Task 3, only upon detection of a tank
leak.  This coordination of tasks precludes  any interruption of
service  and ensures project progression in  the most economic
manner.

Task  4—Leak Characterization and Initial
Environmental Testing:
  The location of a leak identified in the initial testing of Task 3
will be determined immediately following its detection. There are
two methods  by which  this will be  accomplished. The first in-
cludes isolating the tank  from the  pipelines by installation of
valves or  capping the tank and pipelines and  retesting  the
affected tank or pipelines. This method would produce a numer-
ical result identifying the  leakage rate in  both the tank and the
pipeline. The second method includes  excavating the soil from
over the pipelines and retesting the affected  tank while visually
observing the piping for  leakage. Each method  offers distinct
advantages and  disadvantages  depending on  the  situation at
hand. O'Brien & Gere would, depending on the actual field situa-
tion,  recommend the most cost-effective method with which to
proceed.
  When the  location of the leak has been determined, a sam-
pling  program is developed for each facility to determine if there
has been environmental contamination. This sampling program
could include the collection of surficial soil samples and subse-
quent analysis, installation of soil borings  with continuous split-
spoon sampling of soils and subsequent  analysis,  the  installa-
tion of groundwater monitoring  wells and  the collection of
groundwater samples for subsequent analysis. All sampling  and
analysis is conducted utilizing U.S.  EPA-approved procedures.
Samples generally are analyzed  in  the laboratory for  petroleum
fractions including benzene, toluene, xylene  and total hydrocar-
bons.
  At  the completion of the initial environmental assessment,  a
report is prepared which presents the results of the leak charac-
terization  and environmental testing.  This  report provides an in-
dication of the magnitude of contamination, recommendations
for tank or line remediation and  recommendations for further
environmental investigations to define the full extent of contam-
ination. The implementation of these recommendations and sub-
sequent activities associated with  remedial alternatives and de-
sign are discussed in the following sections of this paper.
  Those systems found to be leaking in Tasks 3 and 4 of the
above-described evaluation are scheduled for replacement and/or
repair in accordance with available regulations and technologies.
Those facilities identified as contaminated are then further inves-
tigated  and remedial alternatives are developed, evaluated and
selected.
  Replacements and/or repairs should be conducted in accor-
dance with guidelines established by API, NFPA, Underwriters
Laboratories (UL)  and the Steel  Tank  Institute (STI), among
others. Construction should proceed with structurally sound ma-
terials compatible with the product to  be  stored. The system
should intrinsically, or by design, prevent corrosion from occur-
ring and include safety features such as overfill prevention de-
vices, secondary containment and leak monitoring.
  Depending on the severity of the situation, system repair may
be as simple as replacing a valve or fitting, or as involved as re-
lining an entire tank. Regardless, it must be understood that these
types of repairs are only temporary and should be considered in-
terim until a suitable replacement is financially feasible.
  If necessary, a full investigation to determine the extent of en-
vironmental contamination is conducted concurrent with tank
and line rehabilitation. A program would be developed that could
utilize the following:
• Additional soil sampling and analysis
• Additional soil boring installation
• Additional groundwater monitoring well installation
• Additional groundwater sampling and analysis
• Geophysical investigations
• Site Water Budget Analyses
• Contaminant Mass Transport Modeling and
• Risk Assessment
  The information gathered from these investigations is  utilized
to identify the most feasible remedial alternative, including no
action,  product recovery or in-place containment. This alterna-
tive then is designed and implemented utilizing the best available
technology.
REFERENCES
1.  "Leaking  Underground  Storage Tanks Containing Engine Fuels,"
   Contract No. 68-01-8271, May 1984.
2.  "More About Leaking Underground Storage Tanks: A Background
   Booklet For the Chemical Advisory;" U.S. EPA, Washington, DC,
   1984.
3.  "Technology  for the Storage of Hazardous  Liquids: A State-of-
   thc-Art  Review;"  New  York  State Department of Environmental
   Conservation, Albany, NY, 1983.
4.  "Escalante, E.. ed., "Underground Corrosion," ASTM STP 741,
   November 1979.
5.  National Fire Codes,  Volume I, National Fire Protection Associa-
   tion, 1985.
6.  "Subtitle  I—Regulations of Underground  Storage Tanks;"  Haz-
   ardous and Solid Waste Amendments of 1984, PL 98-616.
7.  "Storage of Flammable and Combustible Liquids,"  R 29.2301 et.
   seq. Michigan State Fire Safely Board, Pursuant  to Section 29.3c of
   the Michigan Compiled Laws.
8.  "Leaking  Underground  Storage Tanks:  Solving the  Problem,"
   O'Brien. & Gere Engineers, Inc., paper  presented at the Michigan
   Industrial Hazardous Waste Conference, May 1985.
272    UNDERGROUND LEAKING TANKS

-------
                Case Study  of Product  Detection  in  Groundwater
                                          Pratap N. Singh, Ph.D., P.E.
                                           Miller Consulting Engineers
                                               Sheboygan, Wisconsin
ABSTRACT
  This study describes the systematic procedures taken to assess
groundwater contamination in the vicinity of a gasoline station
having five underground storage tanks installed 15 to 20 years
ago. A gasoline odor in the basement of an adjacent office
building was detected. Because of its proximity, it was believed
that the source  of gasoline in the  groundwater was the nearby
gasoline station. Leaking pump and storage tanks were the sus-
pected sources. Tests on water samples from a sump pump in the
office  building confirmed high  concentrations  of  benzene,
toluene and xylene in the groundwater.
  A groundwater monitoring program was developed consistent
with the geologic and hydrogeologic conditions of  the locality.
The program consisted of installing six observation wells 20 to 30
ft deep. Two monitoring wells in the vicinity of two underground
tanks and  one well adjacent to a leaking gasoline pump showed
evidence of high concentrations  of gasoline. The two under-
ground tanks then were tested for leaks using the Ainlay Tank
Tegrity Testing Method. Test results were evaluated in terms of
soil and groundwater conditions. Discussion of a product recov-
ery well designed to recover the product is also presented.
 INTRODUCTION
  Bringing underground storage tanks into compliance with the
 1984 amendments to RCRA will have a significant effect on the
 regulated community. Subtitle I of the amended RCRA provides
 a comprehensive regulatory program for underground tanks that (
 contain petroleum and other regulated substances. The U.S. EPA
 has estimated that between 100,000  and 400,000  underground
 tanks in the United States are leaking and are thought to be a
 source of a substantial number  of groundwater contamination
 cases.
  Leaking underground tanks and piping systems used for the
 storage of liquid chemicals and fuels is an important issue for the
 U.S. EPA and industry to address over the next few years. Tank
 leakage problems have many similarities to hazardous waste sites
 in their requirements for remedial action. Groundwater and soil
 contamination can be very difficult and very expensive to correct.
  For tank owners, insuring against non-sudden accidental pollu-
 tion occurrences such as tank leakage  is becoming extremely dif-
 ficult. The  condition  and operating history of  underground
storage and piping systems is coming under much closer scrutiny
in plant/facility environmental control systems.
  The emergence of  legislation  and concomitant regulations
regarding registration,  monitoring of existing tanks and design
and installation of new tanks has also  increased interest and con-
cern for the status of underground tank installations. Finally, the
development of national and state groundwater policy legislation
and regulations has placed further  focus  on the problem of
underground tanks. These factors, coupled with the aging and in-
evitable failure of tanks, makes the management of the problem
unique.
  This paper presents a systematic evaluation of groundwater
contamination in the vicinity of a gasoline station. The program
for this investigation  included background data collection, a
review of site geology and hydrogeology, water quality testing, a
review of applicable groundwater quality standards, tank testing,
assessment  of adverse impact on  groundwater  and product
recovery from soil and groundwater.

BACKGROUND
  The gasoline station is located in a suburban area. An office
building is located immediately west  of the gasoline station. In
early 1985, a gasoline odor in the basement of the office building
was detected. The basement has two sump pumps located in the
northwest and  southeast  corners.  During  high  groundwater
periods, these sump pumps operate frequently. These sumps were
the source of the  odor. Because of the proximity of the office
building to the gasoline station, it was believed that the source of
gasoline was the station. The Wisconsin  Department of Natural
Resources directed the owner of  the station to conduct  an in-
vestigation to determine the  source  of  gasoline  in the nearby
building.
  Investigation indicated that the super unleaded pump was leak-
ing; it was repaired on Jan. 28, 1985. Two samples of water were
collected from the sumps on  Mar. 29,  1985. Test results con-
firmed the presence of petroleum products in the sump water. The
concentration of4he product was  highest in the northwest sump
pit.
  Records indicated that there had  been a small farm house
located close to the southern edge of the office building, parallel
to the gasoline station property. It was learned that a drain tile
was buried about 6 ft below the existing grade at the farm  house.
The location and flow  direction of the drain tile was not known.
Based  on the groundwater flow direction at the site, it was an-
ticipated that the drain tile was laid in a  north-south direction.
  The layout of sanitary and storm sewers and water supply lines
is shown in Figure 1.  The sanitary sewers were installed in 1964
approximately 6 to 8 ft below the existing ground surface. Water
supply lines were placed at the same depth in the same trench.
These pipes were embedded in granular backfill.
  The piping from the underground  storage tanks to the filling
station (Fig.  1) was laid in a trench filled with granular fill.
                                                                                  UNDERGROUND LEAKING TANKS    273

-------
                                _UNDERCROUND
                       i. J* L j"   STORAGE TANKS
                         MW1  J
                          Figure 1
           Gasoline Station and Monitoring Well Layout
METHODOLOGY
  The investigation included installing six test borings from 20-30
ft deep, installing six observation wells and conducting field and
laboratory testing. The well locations are shown on  Figure  1.
Standard field sampling techniques were  employed in accordance
with applicable standards. Drilling was performed using 8-in.
diameter hollow stem augers.  Final boring logs were prepared;
these logs included boring methods, sample  method, depth of
samples, amount of recovery and description of subsoil types and
groundwater levels.
  Observation  wells  were  installed  in  accordance  with the
guidelines established by the Wisconsin  Department of Natural
Resources  (WDNR).'  Monitoring  wells  were developed  and
sampled in accordance  with  the  standards  set  forth  by the
WDNR.2"4 Water  samples were  tested for  benzene,  toluene,
xylene, pH and conductivity.
  Two  of the  station's  underground tanks  having  8,000 and
10,000 gal capacities were tested using the Ainlay Tank Tegrity
Testing method. They were found to be structurally tight.
  Groundwater quality standards,  as established under NR140,
were reviewed and an assessment was made of any adverse impact
on  groundwater.  Test results  were evaluated  in  view of site
geology and hydrogeology.

WISCONSIN GROUNDWATER
QUALITY STANDARDS
  WDNR  has  established groundwater quality standards for
substances detected in, or having  a reasonable probability of
entering, the groundwater resources of the State. The standards
include procedures for determining whether a parameter has been
exceeded and procedures for establishing points of standard ap-
plication.
  The groundwater quality standards for substances of public
health concern are  listed  in  a WDNR  publication.5  For all
substances that have carcinogenic, mutagenic or teratogenic prop-
erties or interactive effects, the preventive limit is 20% of the en-
forcement standard for all other substances  that are of public
health concern.
  For each substance of public health concern, the preventive ac-
tion limit is 50<7o of the established enforcement  standard. For
each indicator parameter, the preventive action limit is established
based on a change of water quality with respect to the background
concentration of the parameter. Background water quality means
groundwater quality near a facility which has not been affected by
the facility.  If, based on the monitoring results from a point of
standard application, a preventive action limit or enforcement
standard for a substance of public health or welfare concern is ex-
ceeded, appropriate additional sampling and testing is required by
the WDNR.

SITE CHARACTERIZATION
  In Eastern  Wisconsin, consolidated rocks consist mainly of
sandstone, dolomite, shale and possibly some limestone, all of the
Paleozoic Age. These rocks are covered largely by soil deposits of
Pleistocene and Recent Age.
  There is evidence of the advance and retreat of at least four ma-
jor ice sheets  in the northern  United States. Although all four
overrode the subject area, only the most recent one, designated as
the Wisconsinan Stage, left soil deposits in quantities of engineer-
ing significance. The Wisconsin ice sheets generally flowed south
through Wisconsin  and the neighboring states. They deposited
lithologically uniform sheets of tills separated by  discontinuous
lacustrine and outwash deposits. The tills were deposited beneath
the ice sheets and are classified as lodgement tills. The outwash
and lacustrine sediments were deposited in proglacial  lakes.
  The individual till sheets have been grouped into three till units
on the basis of their lithologic characteristics. They are referred to
as Till 1, Till 2 and Till 3;  Till 1  being the oldest.6 The general
characteristics of the soil units are summarized by Singh, et a/.7
  The geology of the site described here is dependent on data ob-
tained from the six soil borings and several water supply wells in a
2,000 ft radius of the facility. Also, several publications were re-
viewed. Based on the water supply well logs,  I concluded that the
bedrock elevation changed from about 802 to 606 ft USGS eleva-
tion; the depth to bedrock ranges from 70 to 200 ft; the bedrock
surface  slopes to the east with a  gradient of  about 50 to 100
ft/mile.
  The bedrock underlies a sand and gravel layer.  The sand and
gravel layer  in turn  underlies an older soil unit characterized as
Till 2, which is a  fine grained soil and ranges from sandy silt to
silty clay.  It contains numerous cobbles and  boulders. The Till 2
soil is overlain by lacustrine sand and silt and clay layers,  which
are, in turn, masked by a silty clay layer characterized as Till 3.
The Till 3 soil at the site is overlain by 2 to 14 ft of fill  which
ranges from sand to gravel. The top 4 in. are asphalt  pavement.

SUBSURFACE CONDITIONS
  Construction  of  commercial  buildings  and   utilities  have
significantly changed the subsoils at the site which is typical of an
urban environment. The changes in subsoil generally are limited
to the top 2 to 4 ft of soils which consist of coarse-grained sand
and gravel fill. In general, the fill thickness is about 2 to 4 ft and
approximately 10 ft or more at the locations of the underground
storage tanks.
  The two underground tanks in the vicinity  of Monitoring Wells
MW-2 and MW-3 are 9 ft in diameter and buried about 4 ft below
the existing grade. These two tanks  are 17 and 21 ft long. The re-
maining three underground storage tanks  in the  vicinity of
Monitoring Well MW-1 are 7 to 9 ft in diameter and 14 to 17 ft in
length. These tanks  are also approximately 4 ft below the existing
grade and are embedded in granular backfill.  The trenches for
piping from the storage tanks to the service station also are filled
with  granular backfill. These trenches are approximately 4 ft
deep.
274    UNDERGROUND LEAKING TANKS

-------
      MW6
             MW5
                     -GROUND SURFACE
                                                MW3 MW2
      GW
c  9°—
   80-
   TILL 3
(SILTY CLAY)
                          CL
    70
                                      - FILL_
                                     GW

                                    -^
                               *
                                                 SM

                                                 ML
                                  LACUSTRINE _^
                                  SAND &  SILT
                                      TILL 3
                                   (SILTY CLAY)
                           Figure 2
                   Soil Profile — Section A-A
   Reconstruction of the office building has altered the subsoil
 conditions at the site. The fill thickness at the building is approx-
 imately 4.5 ft. Backfill at the perimeter of the deepest  building
 foundation extends down to  approximately  10 ft below existing
 grade.
   The subsoils underlying the fill are rather uniform and consist
 of very stiff to hard reddish brown silty clay.  Occasional cobbles
 were noted. This soil is characterized as Till 3; it is encountered
 throughout the site, except where excavation for utilities was
 made. The  silty clay (Till 3) is underlain by discontinuous layers
 of lacustrine sand and silt as shown in the soil profiles in Figures 2
 and  3.  These subsoils are water-bearing and moderately
 permeable.  This layer is underlain by silty clay soil of very still to
 hard consistency. This layer has a low permeability, estimated to
 be on the order of  10 ~7 cm/sec.

 SITE HYDROGEOLOGY
   The topography of the area in the proximity of the gasoline sta-
 tion  ranges from slightly rolling to generally flat.  The site is
 located on the crest of a ridge with an elevation of approximately
882 ft (USGS) which extends in a northerly direction with a rise in
elevation to about 900 ft. The ridge acts as a groundwater divide
for surface water and also for groundwater. The surface runoff
drains to the east and west from the site.
  Water supply wells installed at the Silurian dolomite and basal
sand and gravel aquifer indicate that groundwater is approximate-
ly 10 to 75 ft below grade. The variation in groundwater level is
attributed to a highly variable drift thickness and considerable
relief in the bedrock surface. The groundwater flow direction in
the bedrock is easterly  from the ridge  which  agrees with  the
regional   groundwater   flow  direction   in  the   consolidated
sediments.
  In general, groundwater flow in the unconsolidated sediments
appears to divide westerly and easterly based on the topography
of the area.  In proximity  to the site, groundwater  flow in  the
shallow aquifer is northwesterly to westerly (Fig. 2). Groundwater
elevations in the six monitoring wells at the site are approximately
3 to 10 ft below existing  grade.
  Based on the data gathered, I concluded  that the basal sand and
gravel aquifer and the shallow aquifer are not hydraulically con-
nected. These two aquifers are separated by thick layers of tills.
  The sensing zone for the monitoring  wells was  the discon-
tinuous layer of  lacustrine  sand and silt and Till  3  layers.
Permeability of silty sand and sandy silt was estimated to be 10 ~5
cm/sec, which is typical  of such soils. The permeability of silty
clay soils was estimated to be on the order of 10~7 cm/sec or less.
These estimates of permeability for clay and low hydraulic gradi-
ents indicate that movement of groundwater is very slow.


WATER QUALITY ASSESSMENT
  Water samples from the sump in the building were tested  for
gasoline, benzene, xylene, toluene and ethylbenzene. Product
concentrations ranged from 820 to 3,100 /tg/1 for gasoline; 230 to
860 fig/1 for xylene; 33 to 140 /tg/1 for toluene; and 1.3 to 67 /*g/l
for benzene. The  enforcement limits for  benzene,  toluene and
xylene in groundwater are  0.067, 68 and  126 /tg/1, respectively,
under WDNR regulation  140(5). Concentrations of product in the
water samples significantly exceeded  the enforcement standards.
The decision was made first to determine  the source of product
and second the transport mechanism of product from the source
to the building.
                                       MW4
                                                               MW3  MW2
                                   SPILL SOURCE

                       GROUND SURFACE--.
                           100-
                                              ISLAND  EXCAVATION

                                                   •PRODUCT
                            90-
                            80-
                                       CL
                                       SM

                                       CL
                             /
                                                                  GROUNDWATER
                                                   TILL 3
                                                (SILTY CLAY)
                                                                            CL
                     •TANK EXCAVATION
                                                                                         SP
                                                                                             TANKS
                                                                                         CL
                            70-
                                                            Figure 3
                                               Product Transport Path — Section B-B
                                                                                       UNDERGROUND LEAKING TANKS    275

-------
  To assess  the  source  of product, three Monitoring  Wells
(MW-1, MW-2 and MW-3) were placed in the vicinity of the five
storage tanks. Since the storage tanks were built in the late 1960s,
it was possible that they were leaking. Also, granular bedding for
the tanks and piping may have provided a channel for product
transport from the source of leak to the  tanks. Monitoring Well
M-4 was located in proximity to the gasoline station where the
owner reported surface leakage. Monitoring Wells  MW-5 and
MW-6 were placed to evaluate both the groundwater movement
and the  transport of product  in the groundwater.  Monitoring
Well MW-6  was installed in  proximity to the nearby office
building.
  Three  rounds of testing over a five-month period were con-
ducted  to determine the  concentration of product  in  water
samples. In addition, samples of gasoline from super unleaded,
regular and regular  unleaded tanks were also tested  for com-
parison purposes. The test results are summarized in Table 1.

                          Table 1
             Summary of Water Quality Test Results
Gasoline
(•R/l )
Benzene
("g/1 )
Ethyl
Benzene
(•g/l>
Toluene
<•*/!>
Xylene
pH
Conduc-
tivity
10/02
6/26
5/28
10/02
6/26
5/28
10/02
6/26
5/28
10/02
6/26
5/28
10/02
6/26
5/28
10/02
6/26
5/28
10/02
6/26
5/28

-------
 CONCLUSIONS
   An  investigation was conducted to determine  the  source of
 petroleum products  and its  transport mechanism in  soil. The
 following conclusions were reached:
 • Tanks in the vicinity of Monitoring Well MW-1 have not shown
   any evidence of leaking.
 • Although  the  gasoline  tanks adjacent to  Monitoring  Wells
   MW-2 and MW-3 are considered structurally tight, they may be
   leaking at a rate of 0.02 gal/hr.
 • An inventory of gasoline in regular and unleaded storage tanks
   is necessary. Any unaccounted-for loss in  gasoline will con-
   firm leakage from tanks.
 • Existing monitoring wells were monitored periodically  for
   product thickness and for spread of contamination caused by
   product.
 • The transport of  product in the soil and groundwater was
   through the granular backfill in the subsoil.
 • A pump will be placed in the vicinity of regular and unleaded
   tanks in the sand and gravel fill. As discussed earlier, product
   present in the granular material can be pumped.
REFERENCES
1. WDNR "Guidelines For Monitoring Well Installation," prepared by
  Bureau of Solid Waste Management, Wisconsin Department of Na-
  tural Resources, Madison,  WI, 1985.
2.  U.S.  EPA "Procedures Manual  for  Groundwater Monitoring at
   Solid Waste Disposal Sites," EPA SW-6711, Office of Water and
   Waste Management, U.S. EPA, Dec. 1980, Washington, DC.
3.  USGS "Techniques of Water Resources Investigations of the United
   States Geological Survey, Guidelines for Collection and Field Analy-
   sis of Ground Water Samples  for Selected Unstable Constituents,"
   Book I, Chapter D2, U.S. Geological Survey, Washington, DC
4.  Illinois State Water Survey, "Procedures for the Collection of Repre-
   sentative Water Quality Data from Monitoring Wells," Cooperative
   Groundwater Report 7, Illinois State  Water  Survey, 1981,  Cham-
   paign, IL.
5.  WDNR "Groundwater Quality 1985," Chapter NR 140, Department
   of Natural Resources, Wisconsin Administrative Code,  Sept. 1985,
   No. 357, p. 524-37, 523-53.
6.  Mickelson, D.M., Acomb, L., Brouwer, N., Edil, T.,  Haas,  B.,
   Hadley, D., Hess, C., Klauk,  R.,  Laska, N.  and Schneider, A.F.,
   "Shore  Erosion Study Technical  Report," Shoreline Erosion and
   Bluff Stability along  Lake Michigan and Lake  Superior Shorelines
   of Wisconsin, Wisconsin Coastal Management, 1977.
7.  Singh, P.N.,  Tatioussian, S.U. and Flagg, C.G., "A Study of the
   Geotechnical Properties of Milwaukee Area Soils." A special publica-
   tion on Geological Environment and Soil Properties,  ASCE Geo-
   technical Engineering Division,  Houston, TX, 1983,  269-309.
                                                                                        UNDERGROUND LEAKING TANKS    277

-------
                                Underground Storage  Tanks—
                   Leak  Prevention,  Leak  Detection, and  Design

                                         Jey K. Jeyapalan, Ph.D., P.E.
                                            James B.  Hutchison, P.E.
                              Wisconsin Hazardous Waste Management Center
                                             University of Wisconsin
                                               Madison, Wisconsin
ABSTRACT
  Three to five million underground storage tanks containing
hazardous substances exist in the United States. Several thousand
of these tanks are leaking and causing contamination of ground-
water, and the U.S. EPA has been charged with the implemen-
tation of RCRA to control this problem.
  In response to this problem, leak prevention and detection pro-
visions are becoming an integral part of the design and construc-
tion of underground storage tanks. Although several standards
and codes are in existence to yield proper design, many of these
are consensus standards and perhaps may not ensure structural
safety for ail anticipated loading conditions.
  This paper  presents the details of leak prevention and detec-
tion techniques. In addition, the important aspects of structural
design of these tanks are discussed.

INTRODUCTION
  In the Silicone Valley of northern California,  leaking under-
ground storage tanks  containing trichloroethylene  led to the
promulgation of model  underground ordinances which  cover
overfill protection equipment, secondary containment and leak
monitoring systems for all new facilities." The neighborhood gas
stations in Rhode Island had storage tanks that leaked and con-
taminated the water supply. The owners of the gas stations were
found liable for the cost of cleaning up the water supply and sup-
plying clean water to area residences. A Massachusetts gas sta-
tion owner was found liable for the cleanup of the contaminated
groundwater caused by a leaking underground storage tank. In
Long  Island,  leaks from underground  storage  tanks contami-
nated 25 basements. The owner of the tanks was forced to pur-
chase the 25 homes affected by the spill.
  Oliveira and Sitar" have reported the details of over 100 cases
of underground storage tank leaks and spills involving  organic
chemicals  in the Santa Clara Valley of California.  In most of
these cases, the groundwater was contaminated. These case his-
tories indicate only the tip of the iceberg.
  Three to five million underground storage tanks containing
hazardous substances exist in the United States. Of these,  100,000
are presently leaking and another 350,000 are expected  to start
leaking within the next 5 years.10 The users of these underground
storage tanks are listed in Table 1.

FEDERAL LEGISLATION
  RCRA was passed to control this growing problem. The act
covers a wide range of items which include  the storage of haz-
ardous materials in underground storage tanks.  RCRA  was re-
vised by the Hazardous and Solid Waste Amendments of 1984.'4
These amendments define underground storage tank as any one
                                                                                Tablt 1
                                                                     User* of Underground Storage Tank*
                                                       Airports
                                                       Auio Dealers
                                                       Auto/Truck Rental
                                                       Auto Repair Shops
                                                       Banks
                                                       Can* ashes
                                                       Cemetaries
                                                       Churches
                                                       Colleges
                                                       Commercial/Industrial
                                                         Office Buildings
                                                       Construction Companies
                                                       County & Local Governments
                                                         Fire Department
                                                         Police Department
                                                         Prisons
                                                         Sanitation Department
                                                         Public Bus System
                                                         Water Treatment Plant
                                                         Municipal Building
                                                         Highway Department
                                                       Convenience Store
                                                       Delivery Services (UPS,
                                                         Department Stores, Emery, etc.)
                                                       Distribution Companies
                                                       Elementary & High Schools
                                                       Farms
                                                       Federal Government
                                                         Dams
                                                         Federal Highway Department
                                                         Military Bases
                                                         Office Buildings
                                                         Post Office Department
                                                         Prisons
                                                       or a combination of tanks (including underground pipes con-
                                                       nected thereto) used to contain an accumulation of regulated sub-
                                                       stances where the volume (including the volume of the under-
                                                       ground pipes connected thereto) is lOVo or more beneath the sur-
                                                       face of the ground.
                                                         The amendment also requires the following items  to be in-
                                                       cluded in the regulations:
                                                       • Requirements for maintaining a leak detection system, an in-
                                                         ventory control system together with tank testing or a compar-
                                                         able system or method designed to identify releases  in a man-
                                                         ner consistent with the protection of human health and the en-
                                                         vironment
Home Owners
Hospitals
Hotels
Independently Owned
  Service Stations
Grocery Stores
Jobber Bulk Terminal
Major Oil Bulk Terminal
Major Oil Service Stations
Manufacturing Plants
Marinas
Mining Companies
Motels
Nursing Homes
Recreational Facilities
Residential Apartment Buildings
Restaurants
Slate Governments
  Prisons
  Highway Department
  State Office Buildings
School Bus Garages
Shopping Centers
Tire Stores
Transportation Services (Taxi,
  Limousine, Bus Lines)
Truck Stops
Trucking Firms
Utility Companies
278
UNDERGROUND LEAKING TANKS

-------
• Requirements  for maintaining records of any monitoring or
  leak detection system or inventory control system or tank test-
  ing or comparable system
• Requirements for reporting releases and corrective action taken
  in response to a release from an underground storage tank
• Requirements  for taking corrective action in response to a re-
  lease from an underground storage tank
• Requirements  for the closure  of tanks to prevent future re-
  leases of regulated substances into the environment
  The promulgation of any regulation takes time. In order to
minimize potential leaks of tanks which currently are being con-
structed, an interim  prohibition of underground storage tanks
was immediately imposed by the  Hazardous and  Solid Waste
Amendments of 1984. The interim prohibition states:

  "(1) Until the effective date of the standards promul-
       gated by  the administrator under  subsection (e)
       and  after one hundred and  eight days after the
       date of the enactment of the Hazardous and Solid
       Waste Amendments of 1984, no person may install
       an underground storage  tank for the purpose of
       storing  regulated  substances unless such  tank
       (whether of single or double wall construction)—
       (a) will prevent releases due to corrosion or struc-
          tural failure for the operational life of the tank;
       (b) is  cathodically protected  against  corrosion
          (Fig.  1), constructed of noncorrosive material,
          steel  clad with  a noncorrosive material,  or de-
          signed in a manner to prevent  the  release or
          threatened release of any stored substance; and
       (c) the material used in the construction or lining of
          the tank  is compatible with the  substances
          to  be stored.
    (2) Notwithstanding paragraph (1),  if soil  tests con-
       ducted in accordance with ASTM Standard G57-
       78, or another standard approved by the Admin-
       istrator, show that soil resistivity in  an installation
       location is 12,000 ohm/cm  or more (unless a more
       stringent standard is prescribed by the Adminis-
       trator by rule), a  storage tank without  corrosion
       protection may be  installed in that location during
       the period referred  to in paragraph (1)."
                                             WATER
                          Figure 1
                  Phenomenon of Corrosion

  The  amendment  also  establishes  civil  penalties of  up to
$10,000/tank/day for any owner or operator who fails to comply
with either state or federal regulations regarding  underground
storage tanks.
  These regulations are the initial steps taken to reduce the threat
of leaking  underground  storage tanks contaminating ground-
water. The implementation of these RCRA-UST regulations is in
progress and is continuing to gain momentum. As part of the
RCRA regulation promulgation, a proposed rule (The Hazardous
Waste Management  System; Standards for Hazardous Waste
Storage and Treatment Tank Systems) was published in the Fed-
eral Register on June 16, 1985.'
  Statistics provided  by the proposed rule on leaks due to cor-
rosion cause concern about the integrity of existing tanks and the
requirements for future tanks. "Seventy to eighty-five percent of
the leaks were due to tank and/or piping  failures from subsur-
face  corrosion," states a study of over 2,000 cases  of leaking
underground storage  tanks by the American Petroleum Institute
(API). A study conducted for API entitled "Underground Un-
protected  Steel Tank  Study-Statistical Analysis  of  Corrosion
Failure" indicated that 50% of all unprotected steel tanks will
develop a leak by the time they are 16 years old.  The National
Association of Corrosion Engineers (NACE) reported that  after
5 years, the accumulated number of leaks in underground pip-
ing increases by a factor of 10 every 5 years.
  The U.S. EPA and those involved in the design of underground
storage tanks are convinced that any tank will leak and a secon-
dary confinement is the most cost-effective way to address this
problem. Secondary confinement will determine the cause of a
leak, ensure detection of the leak and perhaps contain the leak-
ing material to avoid contamination of groundwater.
                                                                                          Figure 2
                                                                                   Rigid Double Wall Tank
SECONDARY CONTAINMENT
  A concept which has become an important ingredient in leak
prevention and monitoring required by new RCRA is secondary
containment. A  general definition of secondary containment is
the construction  of a barrier to resist free flow between the tank
and the environment. The barrier may be another tank wall or
shell (double walled tanks or jacketed tanks),22'23 as  shown in
Figures 2 and 3.  The shell may be a concrete pit (Fig. 4), a liner
composed of soils (Figs. 5 and 6) or a layer of soil sealant such as
                                                                                    UNDERGROUND LEAKING TANKS    279

-------
                HOLDOOWN CUPS
LEAK DETECTION AND MONITORING METHODS
  The federal legislation which is being implemented places em-
phasis on keeping leaks in underground storage tanks to a min-
imum. There is also the need to detect leaks early so that immed-
iate remedial action can be taken. To understand how leaks can
be eliminated best, it is beneficial to explore the causes of leaks in
underground storage tanks.
                                                                                                     Tible 2
                                                                                         Corrosion Type*; Causes & Corn
                                       V-
                           SECTION A-A


                            Figure 3
                    Flexible Double Wall Tank
soil cement, bentonite,  engineered  low  permeability soil sur-
rounding the tank. Secondary containment systems are designed
to be installed at the time of initial tank installation.  Secondary
containment systems can be retrofitted to existing tanks, but this
usually means excavating and removing the existing  tank to install
the system.
   Leak  detection and  monitoring are easier  on systems which
have secondary containment for two reasons: (1) the leaking ma-
terial must penetrate two barriers to get to the environment and
therefore takes more time; and (2) the space between the two bar-
riers provides a good location for the placement of leak moni-
toring devices, since material which leaked through the first bar-
rier (the tank wall) would accumulate in this space  prior to leak-
ing through the second barrier. Because of these reasons, secon-
dary containment systems always have a leak sensing system con-
nected to the space between the two barriers. Pipes also may have
a secondary confinement system as shown in Figure 7.
                                                                        Corrosion
                                                                        Types	

                                                                        Galvanic
 Stray Currents
Stress
Corrosion
 Biochemical
 Direct
 Chemical
 Attack
                ClUKi
                                              Com
                I. Occur* when dissimilar metals
                are connected by an electrolyte.
                2. Occurs when structural pipe
                or lank joins to different soil
                types.
                3. Dissimilarities on the surface
                of the  pipe.
4. Differential aeration of the
toil around the pipe.

5. Backfilling with mixed strati-
fied layers (local corrosion).


1. Direct current is discharged
from the pipe wall. (Other
sources of current near under-
ground structures with the
ground as pan of their circuit.)

I. Points of concentrated stress
near areas of uniform stress in
the presence of an electrolyte.


1. Bacteria in the soil affecting
the concentration of ions in the
electrolyte.
1. Chemicals directly attacking
the structures (i.e.. road salt).
I. Insulate
metals one from another.
2. Insulate structure
from soils.

3. Close adherence to
correct manufacturing
procedures.

4. Thoroughly tamp
backfill to one ft above
the pipe.
5. Control backfill
material.  (Backfill with
homogenous material.)

I. Must be defined for
each specific instance.
I. Reduce number and
magnitude of concen-
trated stress areas in the
tank and pipeline

1. Thoroughly tamp
backfill to one ft above
the structure. Remove all
deleterious materials
from the backfill.

I. Seal the surface
from infiltration.
                        MTCMCIO     ACCtt* TO TAH«
                                                                 Figure 4
                                                          Tank in Concrete Vault
280    UNDERGROUND LEAKING TANKS

-------
                    DRAINAGE
                     OtTCH
      LINEN ON
    WAaTEPROOP
    COATINQ ON
EN   CONCNCTI
                                                                            LEAK
                                                                         MOMTORMO
                                                                            powr
SEE 'TYPICAL OWE »Y«TEU-
    DRAWING FOH DETAIL*
                                                       •OIL COVEN
                                                     FOB EROaiON AND
                                                       ULTRA VIOLET
                                                       PROTECTION
                                                                                 LINEN ON
                                                                               WASTEPROOF
                                                                               COATINQ ON
                                                                                CONCRETE
                                                            Figure 5
                                                       New Inground Tank
                           LINEN
                         TURHIAOK
                                                                     MONITORINO POINT-
                                                                     aUITABLE PON LEAK
                                                                     DETECTION AND WITH-
                                                                     DRAWAL OP ACCUMU-
                                                                     LATED WATER
                                                  ,PAVEMENT
                                                             COLLAN TO.
                                                             CONNECT PIPE
                                                             TRENCH TO TANK
                                                               LINER
                                                                     • TRENCH
                                                                    TOP LINER
                                                                                          PIPE
                                                                                    FOR LEAK
                                                                                  DETECTION AND
                                                                                   WITHDRAWAL
                                                                                    OP WATER
                                                             Figure 6
                                                 Synthetic Membranes Around Tanks
Causes of Leaks
  External underground corrosion is by far the most common
cause of leaks in  steel  tanks.  Underground corrosion involves
chemical reactions and the flow of electrical current. Therefore,
corrosion occurs through an electrochemical process. The soil,
with its water content,  acts as an electrolyte. Corrosion occurs
when metal atoms go into solution and are transformed into com-
pounds as shown in Figure 1."
  It may be helpful to visualize the process of external corrosion
of an underground storage tank by viewing the tank/soil system
as an electrolytic cell. The corroding area of the tank is the mode,
the soil is the electrolyte and another area of the tank forms the
cathode. An anode is the electrode of an electrolytic cell at which
direct current is discharged and corrosion occurs.24 The cathode is
the electrode of an electrolytic cell at which reduction or preven-
tion of corrosion occurs. When an area of a tank wall corrodes to
the extent that the useful  wall thickness cannot carry external
pressures or internal pressures, a leak develops.
                                     The different types of corrosion, their causes and  cures are
                                   listed in Table 2.24'26-27 Cathodic protection is an important meas-
                                   ure taken to prevent corrosion. A sacrificial metal is placed near
                                   the tank and acts as the anode. Current flows from the sacrificial
                                   metal through the soil to the tank which acts as the cathode. The
                                   tank shall be protected from corrosion as long as the  sacrificial
                                   metal can transmit a current to the tank.
                                     Internal corrosion accounts for  10-29% of corrosion induced
                                   failures.' Internal corrosion has the same causes and cures as ex-
                                   ternal  corrosion, plus the additional factors of dynamic impact
                                   loads of the gauging stick and the loads  caused when  filling the
                                   tank. Installing a striker  plate below the fill pipe eliminates the
                                   possibility of scratching the liner with the gauging stick; it also
                                   diverts the incoming hazardous material, thus minimizing the im-
                                   pingement at the bottom of the tank.''
                                     Inadequate installation and design procedures  are another
                                   cause of leaks. Improper handling of coated steel tanks may cause
                                   scratches in the lining which expose the steel to corrosive activity.
                                                                                       UNDERGROUND LEAKING TANKS    281

-------
                                                          PAVIUIHT
                                         WELL COMPACTED NON CORH08I
                                               (CLIAN WAiH«0  IAHO OK •MAVtl)
                       LINIM ON WAIT1PMOOP
                       OOATIN* ON COHOMITf
                                                           Figure 7
                                               Underground Lined Trench for Piping
Neglecting hold-down straps or underdesigning them may cause
uplift of the tank when groundwater surrounds the empty tank.
Poor quality backfill or improper backfilling procedures may
cause excessive strains in flexible tanks. Filling the tank prior to
backfilling also may cause rupture of the tank due to lack of lat-
eral support from the backfill.
  Poor operating procedures  can  create leaks in underground
storage tanks. Puncture of the tank wall by the gauging stick or
filling/discharge pipe may cause a leak. Overfilling causes spills
and unexpected pressure head within the tank. Proper operational
practices such as the use of a level sensing device, level indicating
device, high level alarm, automatic shutoff control system, by-
pass prevention of overfill  prevention system and interlocking
loading process with overfill prevention system  so loading cannot
occur  without the overfill  prevention  system  being activated,"
will prevent spills which account for a large portion (± 40%) of
contamination.*

Leak Detection
  Tanks cannot be designed to be  virtually leak proof even at a
significant cost. Thus, vendors take the approach  of designing
tanks to last 30 to 50 years. Since  tanks will inevitably leak due
to one of many causes, it is best to design the tank with leak de-
tection and monitoring. Early detection and remedial action can
save millions of dollars.
  Testing for leaks can be classified into four general groups:"

• Volumetric leak tests (quantitative)
• Non-volumetric leak tests (qualitative)
• Inventory control and reconciliation analysis (trends)
• Effects monitoring (perceivable environmental changes)

Volumetric Testing
  Volumetric leak tests provide data which can determine if a
tank is leaking and at what rate it is leaking. This is perhaps the
most helpful information a tank owner or regulator can obtain.
Knowing the rate of the leak helps  to determine the needed level
of corrective action.
  The  rate of leak  is found by determining the  volume of de-
crease  of stored product by measuring properties associated with
  0.09


  0.08


  0.07


  0.06


_ 0.05
 o
 o>
«-i
°. 0.04


  0.03


  0.02


  0.01
                           J_
                                       I      I     I     J	1
                2000
                          4000       6000
                            Tank Size (90!)
                                                8000
                                                          10.000
                           Figure 8
  Change in Temperature of Gasoline Needed to Produce a 0.05 Gallon
             Volume Change as a Function of Tank Size
the change in volume. The volume of the stored material is mon-
itored constantly by a float inside the tank or by a pressure sen-
sor in the bottom of the tank. Since the temperature of the stored
material has an important effect on the volume of the material, it
must be measured. A small change in temperature, particularly
with a large volume as seen in Figure 8, may change the volume of
the material significantly. Vapor pockets within the tank may in-
fluence the measured volume of the material (for tests that re-
quire a full tank) because air or  vapors expand and contract at
different rates than fluids.
282
       UNDERGROUND LEAKING TANKS

-------
  Vibrations of vehicle traffic or mechanical machines can cause
a surface ripple in the stored substance causing the float to move
erratically.  Since the measurement of the depth of stored  ma-
terial must  be within fractions of an inch,  any vibration move-
ment should be eliminated.
  The presence of groundwater around  the tank will not affect
the volume of the stored substance, but will affect the rate  at
which the substance leaks.  This change in  rate is due to a de-
crease in differential pressure between the inside and outside  of
the tank. Evaporation and  condensation within the tank  may
cause errors in volumetric testing. These usually cause problems
only if the stored material is not sealed to atmospheric conditions
and can  be remedied by placing the cap back on the fill pipe.
Volumetric testing requires trained personnel and very accurate
measurements. These elements  make volumetric testing expen-
sive and time consuming.

Non- Volumetric Testing
  Data from  non-volumetric tests determine if a tank is leaking
or not but do not  determine the leak rate.  These data are valu-
able but not as useful as a volumetric test. Non-volumetric tests,
however, are  cheaper and easier to perform and may be used  to
determine which tanks, within  a field  of multiple tanks, need
further testing or may be used  as a  first alert for more serious
potential problems.
     Normal.
     After  40 minutes.
     Alarm  situation.
     After  90  minutes.
  Several methods are used for non-volumetric testing. Helium
leak detectors commonly are used: helium vapor pressurizes the
in situ tank and the amount of helium that diffuses through the
leak back to the surface is monitored. Although helium can dif-
fuse through asphalt and concrete, it takes time and, therefore,
boreholes may be cored for quicker detection.
  The vacuum/hydrophone tests involve creating a vacuum in a
sealed tank greater than the static head pressure of the liquid in
the tank.  The tank then is monitored with hydrophones for the
sound of bubbles being drawn in through the leaks. If the ground-
water table surrounds any of the tank, then the amount of water
within the tank should be determined before and after the test to
see if any water was drawn in through the leak.
  Direct buried cable sensing  is useful in long-term constant
monitoring situations.39 Two wires are placed within an insula-
tion jacket, as shown in Figure 9, which dissolves in the pres-
ence of petroleum; when the jacket dissolves, the two wires touch
and signal an alarm. Since the cable is  buried directly, it can be
placed in  areas where it will do the most good as shown in Fig-
ures 10 and 11.
                                                                                         Figure 10
                                                                        Direct Buried Cable Placement for No Water Table
                        Figure 9
             Direct Buried Cable for Leak Testing
                                                                                         Figure 11
                                                                        Direct Buried Cable Placement for High Water Table
  Vapor testing can be performed in the soil next to the tank sys-
tem within the space between the two  walls of a  double-walled
tank,  called interstitial testing, or within the space between  the
other  secondary  confinement structures.  This test  utilizes a
vacuum to draw the vapors through the monitoring device. Ob-
servation wells with either temporary or continuous monitoring"1
can be used.
                                                                                     UNDERGROUND LEAKING TANKS    283

-------
                         Figure 12
            Typical Service Station Tank Installation
   In the case of petroleum, continuous monitoring is done by
placing a float within the monitoring well as shown in Figures 12
and 13. The float is equipped with petroleum sensing equipment
which  can sense as  little  as 0.25 in. of petroleum floating on top
of the  water table. Direct sampling of the water can be done with
wells to determine the chemical make-up of the water.
   Interstitial flooding can determine a leak in one of the two walls
 of a double-walled tank. The interstitial space is flooded with a
 liquid  and the height of this liquid in  the fill pipe to the inter-
 stitial space is monitored as shown in  Figures 14 and 15. If the
 liquid suddenly goes down, this decrease in liquid level indicates
 that there is a  leak in either of the two walls.
   In the dye  method, a dye is  used in the stored substance. If
 this dye shows up in surrounding storm sewers, sanitary sewers,
 ditches, basements or any other low area, its presence indicates a
 leak in the pipe or the tank.
   It should be noted that these non-volumetric tests are versa-
 tile, and various tests may be combined or altered to collect more
 data. For example, a dye may be flooded into the interstitial space
 of a leaking tank to see if the inside wall is leaking, in which case
 the stored material will  carry the dye color, or if the outside wall
 is leaking, in  which  case the backfill will become dyed. The di-
 rectly buried cable also could be placed in the interstitial space
 to check the leaks.

 Inventory Monitoring and Reconciliation
   Symptoms of leaks can be observed if careful inventory mon-
 itoring of the stored substance is used." Leaks cannot be directly
 detected or measured by this method, but it can be a helpful tool
 in determining where there is a strong possibility of a leak. Symp-
 toms which appear when there is a leaking tank are an apparent
 loss or gain of stored substance when the substance has not been
 dispensed or filled, differences in the amount of substance dis-
 pensed and received,  a  hesitation in the delivery  of the material
 from a standard dispensing pump indicating a leak in the suction
 pipeline and odors of the stored material in places below ground.
   Factors which  influence the   inventory control  method are
temperature related  volume changes of the stored substance,
 faulty  or  erroneous  dipstick readings  and displays.  Computer
analysis of the data enhances the interpretation of the data, but
the data must first be as accurate as possible.
                                                                                  FH.Iol. ConUI Option*
                                   Conduit
                                            u vtholl Cow
                                                         Solid
                                                         flliw
                                                         Pip* (Not Suppll.cn
                                                  SlKvianJ Slotted Pip*
                                                  Monitoring Will VrMn
                                                  4' 8CM40PVCor
                                                  <• SttfnlM* SIM (No! Suppltod)
                             UNOcmnouNO INSTALLATION
                               (Typta* Ouollni Billion)
                                                                        AaOVCOflOUNO INSTALLATION
                                                            Figure 13
                                                        Monitoring Wells
284    UNDERGROUND LEAKING TANKS

-------
                                                                                                       XNATIVE
                                                                                                       '"-on
                                                                         •&&f"                   "•••'%
                                                                         ^l":--^   UrtLJ r*f\r*etf\at\it!    '7.W..
                                                           Figure 14
                                                     Double Wall Steel Tank
Leak Effects
  Monitoring  the environment immediately adjacent to  the
underground storage tank locations for any changes due to leaks
is another qualitative  method of finding leaks. A leak may be
discovered, but the quantity of leakage and the exact locations,
particularly if there is more than one tank at the location,  will
not be known.
                                       Liquid Media
                                       Reservoir
                                       Water Table

                                       Inner wall breach
                                       Liquid media goes into
                                       tank; prevents product
                                       from escaping.
                                       Outer wall breach
                                       Liquid media goes into
                                       tank excavation
Monitoring Methods
  Proposed legislation will require  proof of monitoring  and,
therefore, records will have to be kept. There  are  three basic
methods of compiling these records: (1) by hand, (2) by mechan-
ical means (or computer), or (3) by a combination of these two
means.
  Continuous monitoring almost always is compiled by mechan-
ical means.  Few continuous monitoring techniques yield  con-
tinuous data such as the direct buried cable, but will set off an
alarm whenever a leak is detected, even by a modem to the own-
er's remote location.
  Routine or scheduled compiling could verify such continuous
monitoring. This monitoring  would entail a daily or  weekly
checking of devices in place. Scheduled air samples or water sam-
ples could be taken and tested. Good record keeping will be the
owner's responsibility. Spot check monitoring with volumetric
tests should  be done when inventory reconciliation or leak detec-
tion alarms are triggered.
                          Optional Leak
                          Simulator*'
                                                           Figure 15
                                                  Interstitial Monitoring System
                                                                                      UNDERGROUND LEAKING TANKS    285

-------
       Reinforced concrete
       #5 rebar 8" OC both ways
       (2-layers)
                               -  24" x 24" manhole box
                                 with cover (not supplied)
  9V
                                             2" x4' x 3' mln or
                                             equivalent wood
                                             required to support
                                             riser
22S" sq
(opening in riser)
                                       1) Riser Is not load bearing
                                       2) Rebar mat must extend
                                         4 ft. in  all directions
                                         from center of riser
                          Figure 16
                      Manway Riser Pipe
         r
         l	j
                              Tank
                              top
                                        r
                                          Two filling
                                            option


                                                      |
                                              Plug
                                        I	I
                                       4	.	
                                       Fiberglas layup
                         Fiberglas
                         layup
                                         2" or 4" NPT
                                         half-couplings
* "X" Dimension on Double 2" Fittings Equals 5"
  "X" Dimension on Double 4" Fittings Equals 9"


                          Figure 17
                      Steel NPT Fittings
Tank Appurtenances
  In addition to requirements about leak detection devices, the
new tanks should  include several  types  of key appurtenances.
These include manway riser pipes, fittings, heating coils and lad-
ders for visual inspection as shown in Figures 16-19.
                                                                       Heating Coll
                                                                                        Material: carbon steel

                                                                                              28' dia  x  ',." thick
                                                                                                  24 - V holes on a
                                                                                                 /25^,." boll circle dia.
                                                                                                . 2" NPT couplings (optional)
                                                                            1".* coil inlet and outlet
                                                                                       Heavy-duty 22"
                                                                                       ID manway
                                                                                                               Varies
                                                                                                              with tank
                                                                                                              diameter
                                                                                                             4-0"
                                                                                                                8" min.
                                                                                           Figure 18
                                                                                      Helical Heating Coils
STANDARDS AND CODES
  One of the attempts tank  manufacturers have undertaken to
minimize leaks is the use of standards and codes for the design
of underground storage tanks. The following codes and standards
pertain to the design and performance  of reinforced fiberglass
and steel underground storage tanks:

• American Petroleum Institute Standard 6201
• American Society for Testing and Materials D 4021-18'
• National Association of Corrosion Engineers RP-01-69"
 286     UNDERGROUND LEAKING TANKS

-------
              22" manway on 6', 8', and 10' diameter tanks

                                     FRP slip lug
                          Figure 19
                        Tank Ladders
 • National Fire Protection Association  Codes 30 and 31 and
  Handbook29'31
 • Underwriters Laboratory Codes 58 and 131640'41
  There are three types of standards or codes for underground
 storage tanks. The first are the standards to be followed when de-
 signing these structures. The second are performance standards
 which list a series of tests or qualifications that must be passed by
 the manufactured underground storage tanks. The third are stan-
 dards or codes which specify how the manufactured tank should
 be installed and operated.


 Steel Tanks
  The standards which govern steel underground storage tank de-
 sign are UL 58,40 API 6202 and NACE RP-01-69. The UL 58 is
 the standard for steel underground tanks for flammable and com-
 bustible liquids. The  requirements  include:  capacities, dimen-
 sions, metal thicknesses, materials, shell seams, heads and head
 joints, compartment tanks, pipe connections, manholes, heating
 coils and hot  wells, manufacturing and production tests and
 markings.
  API 620 contains recommended rules for the design and con-
 struction of large, welded,  low-pressure storage tanks. The text
 includes: materials;  design  of walls, roofs,  internal and extern-
 al structural members, wall  openings, nozzle necks, bolted flange
connections, cover plates and joints (welds); inspection and tests;
and pressure and vacuum-relieving  devices. The  recommended
rule also contains a number of helpful appendices which include
suggested good practices regarding foundations, corrosion allow-
ance, attachment structures (internal and external), peening, in-
stallation, supporting  structures and determination of relieving
capacity required.
   NACE RP-01-69 is a recommended practice of control of ex-
ternal corrosion on underground or submerged metallic  piping
systems and pertains to underground steel structures. This stan-
dard is somewhat detailed in dealing with corrosion control. The
standard  deals with the following  list of thems: structural  de-
sign of corrosion resistance;  coatings for corrosion resistance
and their applicability; cathode protection and its criteria,  de-
sign, installation, operation and maintenance; control of inter-
ference currents; and corrosion control records.

Fiberglass Tanks
   The two codes listed for reinforced fiberglass tanks are both
performance standards. These codes are described below.
   The ASTM D 4021-81 contains  standard  specifications for
glass-fiber-reinforced polyester underground  petroleum storage
tanks and prescribes a  number of requirements the finished pro-
duct must fulfill.3 Tests  include the concentrated load test, ex-
ternal  hydrostatic  load test,  internal  pressure  tests,  fitting
moment test, fitting torque test, leakage test, internal impact test,
lifting lug strength test, negative pressure test, material property
tests and chemical resistance tests. Minimum thickness for shell,
pipe fittings and flange  also are given in this standard.
   The UL 1316 is  a standard for  glass-fiber-reinforced plastic
underground storage tank for petroleum products and this gives a
number of tests that must be fulfilled by the finished product.41
The test titles are virtually the same  as those in ASTM D 4021-81
standards, but the procedures are somewhat  different. The UL
1316 standard also  has additional marking requirements  along
with manufacturer installation requirements.
   Finally, the installation, usage and monitoring codes pertain to
all tanks regardless of the material used in making the tank. These
codes  are: NFPA  30  "Flammable and  Combustible  Liquids
Code,"24 NFPA 30 "Oil Burning Equipment  Code,"30  and the
NFPA "Fire Protection Handbook,"31 These  standards consider
the possibility of fires  and thus are written to reduce  the risk.
Methods of construction and abandonment  of existing under-
ground storage tanks also are detailed in these standards.
DESIGN CONSIDERATIONS
  A successful design of underground storage tanks should con-
sider the following:

• Loads
• Materials and properties
• Compatibility and life
• Tests
• Handling
• Installation
• Procedures for analysis
• Monitoring and management
  The design of underground storage tanks must take into ac-
count the various loadings that the tank will experience. Failure
to recognize these loads could result in collapse of structures as
shown in Figure 20. These loads include internal pressures (or
vacuum) caused by liquids and gases, the weight of the tank itself
and external loads. The external loads consist of traffic loads,
soil loads, reactions at the supports and pipe connections, wind
loads,  seismic loads, frost loads and hydrostatic  loads. Fatigue
loads caused by filling and  unfilling the tank also should be
addressed.
                                                                                     UNDERGROUND LEAKING TANKS    287

-------
                         Figure 20
                       Tank Failures
Loads
  The internal pressure loads depend on the properties of the ma-
terial being stored and the geometry and fiUing/unfilling pro-
cedures of the tank system. For example, if the material stored
is delivered hot and  evaporates, it may cause a vapor pressure
within the closed tank system.  If the filling pipe extends to an
appreciable height vertically above the tank  and  is filled to the
top, high pressure heads in the liquid  will develop. The weight of
the tank can be determined prior to the installation. This load
becomes important in empty tanks and tanks that are simply at
their ends.
  The external loads may be more difficult to determine than the
previously mentioned loads. The site at which the tank is to be in-
stalled must be characterized adequately to determine these loads.
Traffic loads are the dynamic loads  transmitted by moving ve-
hicles through the soil to the tank. These loads are a function of
the weight of the vehicle, size of the tire, the vertical and the hor-
izontal distance from the load to  the tank and the soil asphalt-
concrete structure between the load and the tank.
  Soil loads are a function of the type of soil surrounding the
tank, the depth of burial of the tank and the pressure of any load
producing structures such as foundations adjacent to the tank.
The weight of the soil above the tank minus any arching affect
the soil may possess, will load the tank. Dynamic loads such as a
foundation supporting vibrating machines or equipment may in-
duce additional loads on the tank and the piping system.
  Reactions at supports tend to localize the external loads on the
tank at the support areas. This increases the shear and moment
induced stresses and strains within the tank wall. The tank wall
material must be designed to withstand these stresses and strains.
  Wind, seismic and frost loads also should be taken into account
if they are applicable. Sites must be evaluated  for the severity
and likelihood of these loads.
  Hydrostatic loads are  loads imposed on the tank due to the
presence of water in the soil surrounding the tank. If the water-
table or the presence of static or flowing water is never encoun-
tered by the tank, the hydrostatic loads may be neglected. If the
tank is not designed with any hold down mechanisms and the
buoyant force is greater than the weight of the tank and the
stored  materials and the resistance from  the soil-asphalt-coo-
Crete above the tank, the tank will float out of the ground. Straps
and anchors should be designed approximately to withstand these
forces.
  Fatigue loads  caused by filling and discharging the tank also
should be considered in the design.  This consideration may in-
clude the design of a spill plate  below the filling pipe and extra
reinforcing around pump and pipe connections within the tank.
In addition, temperature effects and Poisson's effects need to be
considered.
  The various types of loads which need to be considered were
discussed in the previous section. The properties of materials con-
stituting the tank wall required to complete the design  are a*
follows:
• Strengths under tensile, compressive and shear loads
• Moduli under tensile, compressive and shear loads
• Stress-strain curves
• Poisson's ratio

Material Compatibility
  In order to ensure long life for the designed tank, the compat-
ibility of the stored  materials with the tank wall, stress  corro-
sion, strain corrosion and creep need to be accounted for in the
analyses, and proper allowances need to be made in the final de-
sign alternatives.

Handling WM! Installation
  Proper handling and installation of the tank are important for
the  tank to function adequately during its anticipated life. After
the  underground storage tank has passed a number of manufac-
turing tests (such as the 5 Ib/in.' air test and the x-ray test) and
has passed on-site delivery inspection, it is ready to be placed in
the  ground.  Care must be taken when handling  these tanks, es-
pecially the more flexible tanks,  to avoid unnecessary protective
coatings and other expenditures.  Manufacturers have recommen-
dations to follow to meet installation requirements.
  Once the tank is in place on its appropriate bedding material,
backfilling may begin. Most tank installation instructions prevent
filling the tank with the contents prior to backfilling. Because of
the  need to reduce the pressure of water and to provide adequate
lateral  soil support, it is necessary  to use a  granular material
which is backfilled and compacted in one foot lifts. Methods of
compaction to ensure uniform and complete compaction require
the  use of proper equipment to suit the types of backfill and tank
used. A  technique used  to  evaluate  how  well compaction is
achieved is to compare the field densities to standard or mod-
ified Proctor laboratory test data on samples of soil. If the field
288    UNDERGROUND LEAKING TANKS

-------
density is less than that required by the designer's specifications,
the backfill should be recompacted.
  It may be of interest that there is an optimum water content of
the soil where the soil compacts easiest. This value is obtained by
the Proctor test.  Field densities can  be determined by the sand
cone method or by a nuclear density test. Emphasis should be
placed on the importance of properly placed backfill to ensure
tank longevity.
  Once the tank has been backfilled, the ancillary equipment  and
piping system can be installed. The whole system should be con-
structed to achieve easy accessibility for  tank inspection (i.e.,
measuring stick,  etc.), tank filling and emptying. Access man-
holes  and vents should be located under low overhangs or adja-
cent to a wall.
  After installation, the 5 lb/in.2 air pressure (or water pressure)
test should be run on the underground storage tank and its pip-
ing system. The entire system is pressurized by air (or water),  and
the amount of pressure drop over a certain time period is meas-
ured.  If the leakage rate is too excessive, corrective measures must
be taken.

Analysis
  The soil  tank interaction  behavior also needs to be  analyzed
for groundwater induced seepage pressures, settlement, live load
induced  stresses, longitudinal stresses  and  circumferential
stresses. Finite element models provide reliable means of analyz-
ing even the most complex situations in  the  field. The senior
author has developed a series of analysis procedures for perform-
ing the above design computations. Further details are given in
Jeyapalan et a/.16~21  The analysis procedures also permit  the
following types of investigations:
• To  assess the structural adequacy of existing tanks
• To  assess the remaining life of existing tanks before these need
  replacement
• To assess the adequacy of new tanks sold by various vendors
• To provide backup analysis when outdated codes and standards
  are inadequate to ensure structural safety
  These procedures have been found useful when examining sev-
eral  underground  installations. Additional  information  can be
found in Jeyapalan, et a/.15~21

TESTING UNDERGROUND TANKS
  There are a number of non-destructive testing (NDT) methods
available to test the integrity of the underground storage tank
walls" (Table 3). Most of these methods require  access to both
sides of the tank wall, but two methods can be performed with
access to only one side of the tank wall. These methods may be
able  to test in situ tanks if inside (or outside) access to the tank is
safe and available.

Ultrasonic Testing
  The first method is ultrasonic testing.4'6'13 This method is used
to inspect voids, cracks and non-metallic inclusions and to deter-
mine the thickness of the tank wall. Residual strains within the
tank wall  can be measured, although material anisotropy makes
this measurement difficult. This method requires  a smooth sur-
face, a couplant between the probe face and the  material being
tested, point by point testing and skilled personnel to run the test.
Both steel and FRP tanks can be tested in this manner.

Acoustic Emission Testing
  The second method  is the  acoustic emission test using stress
waves.4'36  When a  defect such as a crack within a tank wall ex-
tends within a metal or FRP storage tank, it generates stress waves
which can be detected. A tank is tested under a loading to deter-
mine the extent of defects in the wall.  The results can be inter-
preted in terms of rate of crack growth, the quality of material
and the potential of failure of the structure.
                                                           Table 3
                                      Comparison of Common Nondestructive Testing Methods
Method
Ultrasonics
Radiography
Eddy Currents
Visual -Optical
Liquid
Penetrant
Magnetic
Particles
Characteristics
Detected
Changes 1n acoustic
impedance caused by
cracks, nonbonds, In-
clusions, or Interfaces.
Changes immaterial
density from voids,
inclusions, material
variations; placement
of Internal parts.
Changes in electrical
conductivity or mag-
netic permeability
caused by material
variations, cracks,
voids, or inclusions.
Surface characteristics
such as finish,
scratches, cracks, or
color; strain In trans-
parent materials.
Surface openings doe
to cracks, porosity.
seams, or folds.
Leakage magnetic flux
caused by surface or
near-surface cracks,
voids, inclusions, ma-
terial or geometry
changes.
Advantages
Can penetrate thick
materials; excellent
for crack detection;
can be automated.
Can be used to Inspect
wide range of 'materials
and thicknesses; versa-
tile; film provides re-
cord of Inspection.
Readily automated;
node rate cost.
Often convenient; can
be automated.
Inexpensive, easy to
use, readily portable,
sensitive to small sur-
face flaws.
Inexpensive, sensitive
both to surface and
near-surface flaws.
Limitations
Normally requires
coupling either by
contact to surface
or immersion in a
fluid. Orientation
can present problems
In detection or Inter-
pretation of defect.
Radiation safety re-
quires precautions; ex-
pensive; detection of
cracks can be diffi-
cult.
Limited to electrically
conducting materials;
limited penetration
depth. Interpretation
of defect signals can
be difficult.
Can be applied only to
surfaces, through sur-
face openings, or to
transparent material .
Flaw must be open to
surface. Not useful
on porous materials.
Limited to ferromag-
netic material ; surface
preparation and post-
Inspection demagnetiza-
tion may be required.
Example of
Use
Adhesive as-
semblies for
bond Integrity.
Detection of
creeks.
Pipeline welds
for penetration.
Inclusions, voids.
Verification of
parts 1n assemb! les.
Heat exchanger
tubes for wall
thinning and cracks.
Verification of
material heat
treatment. /
Paper, wood, or
metal for surface
finish and uni-
formity.
Turbine blades
for surface
cracks or porosity.
Railroad wheels
for cracks.
Detection of
weld defects.
                                                                                     UNDERGROUND LEAKING TANKS    289

-------
  Non-destructive testing methods which require access to both
sides  of the  tank wall are:  radiographic inspection (i.e., x-
rays),4-'3'5 eddy-current techniques which can be performed on
electrical conducting tanks,4'13'8 liquid  penetrant  which  detects
cracks on the surface of tanks, magnetic particle testing which
detects surface or near surface cracks, voids, inclusions and ma-
terial or geometric changes in ferromagnetic material and visual
inspection. Visual inspection may be an inexpensive initial step of
tank integrity investigation. Internal  inspection of a tank, either
through direct access or by the use of mirrors, may give the first
clues to the condition of the tank.
  New methods which eventually will be used by the industry in-
clude lasers and holographic methods.7

CONCLUSIONS
  The construction of many chemical industries, gasoline stations
and other facilities storing hazardous materials in large quantities
led to the installation of thousands of underground storage tanks
during  the  period from  1950 to  1970  throughout the  United
States.  Many of these tanks have exceeded their useful lives and
either are leaking now or soon will be leaking. The RCRA-UST
regulations introduced by the U.S. EPA to protect human health
and the environment from contamination are summarized in this
paper. In addition,  various leak detection and monitoring tech-
niques to avoid  groundwater contamination also  have been re-
viewed.
  The details of codes and standards  for steel and fiberglass tanks
have been presented. However, most of these are consensus stan-
dards and may not ensure structural design adequacy for all antic-
ipated loading and operating conditions. It is necessary to analyze
and design these storage  tanks using proper analysis techniques
and procedures of state-of-the-art design practices.
  The most important design considerations also were discussed.
The design principles applied for large diameter buried pipelines
during the last decade provide additional useful information for
the designers of new  tanks. The new tanks should  be designed
with  appropriate leak detection and  secondary containments.
They should  further have provisions for testing the integrity in
place. Proper tank management and operation procedure are re-
quired to ensure long life for these potentially hazardous under-
ground storage structures.


REFERENCES
 1.  Antonenko, E.V.,  "Stress-Strain  Conditions of Cylindrical  Con-
    tainers Sunk Into the Ground," Chem.  and Pel. Eng.,  June  1969,
    8-9.
 2.  API Standard  620, "Recommended Rules for the Design of Large,
    Welded,  Low Pressure Storage Tanks,"  API,  New  York,  NY,
    1956.
 3.  ASTM D4021-81, "Standard  Specifications for Glass-Fiber-Rcin-
    forced Polyester Underground Storage  Tanks," Annual Book of
    ASTM Standards, Vol. 08.04, ASTM, Philadelphia, PA, 1983.
 4.  Berger, H., "Nondestructive Evaluation," Pressure  Vessel and Pip-
    ing Technology-1985-A Decade of Progress, 1985, 591-599.
 5.  Berger, H., "Radiographic Inspection," Pressure Vessel and Piping
    Technology-1985-A Decade of Progress, 1985,633-647.
 6.  Birchon,  D., "Non-Destructive Testing," For the Design Council,
    the British  Standard Institute, and the Council  of Engineering In-
    stitute, Oxford University Press, 1975.
 7.  Birnbaum,  G.  and White,  G.S.,  "Laser Generated and Detected
    Ultrasound and Holographic Methods," Pressure Vessel and Piping
    Technology- 1985-A Decade of Progress, 1985,661-669.
 8.  Deeds, W.E. and Dodd, C.V., "Eddy-Current Technique," Pressure
    Vessel and Piping Technology-1985-A Decade of Progress,  1985,
    647-659.
 9.  U.S. EPA, "Hazardous Waste Management Systems; Standards for
    Hazardous Waste Storage and Treatment Tank Systems," Proposed
    Rules, Federal Register, 50, June 26, 1985.
10.  U.S.  EPA, "Leaking Underground Storage Tanks,"  Document
    EPA/530-SW-009, U.S. EPA, Washington, DC, Apr. 1985.
11.  Fitzgerald, III,  J.H.,  "Corrosion of Underground Storage Tanks
    ... Causes and Cures, "Plant Eng., 37. July 21, 1983,46-48.
12.  Callctly, G.D. and Pemsing, K., "On Design Procedures for the
    Buckling of Cylinders Under Combined Axial Compression  and
    External Pressure," J. of Pressure Technot. 106, May 1984, 134-142.
13.  Green, Jr., R.E., "Ultrasonic Methods," Pressure Vessel and Pip-
    ing Technology-1985-A Decade of Progress. 1985,601-611.
14.  The Hazardous and Solid Waste Amendments of 1984 (To amend
    the Solid Waste Disposal Act), Jan. 23, 1984.
15.  Jcyapalan, J.K.,  Proc. of the  International Conference of Ad-
    vances in  Underground Pipeline Engineering, Madison, WI, Aug.
    1985.
16.  Jcyapalan, J.K.  and Ethiyajeevakaruna,  S.W..  "JFLOW-IBM
    P.C.  Software for Flownct Construction and  Groundwater Seep-
    age Analysis."  Report of the  Wisconsin Hazardous  Waste Man-
    agement Center, Jan. 1986.
17.  Jeyapalan, J.K.  and  Elhiyajcevakaruna,  S.W., "JSTRESS-IBM
    P.C.  Software for Calculating  Stresses Induced by Surface  Sur-
    charge Loads and Traffic  Loads on Tanks and Other Buried Struc-
    tures," Report  of the Wisconsin Hazardous Waste Management
    Center. Jan. 1986.
18.  Jeyapalan, J.K.  and  Ethiyajeevakaruna, S.W.,  "JSETTLE-IBM
    P.C. Software for Calculating Settlement of Tanks and Pipes Under
    Loads," Report  of the Wisconsin  Hazardous Waste Management
    Center, 1984.
19.  Jeyapalan, J.K. and Ethiyajeevakaruna, S.W., "JLONG-1BM P.C.
    Software for Calculating Longitudinal Stresses and Strains in Under-
    ground Tanks and Pipes,"  Report of the Wisconsin Hazardous
    Waste Management Center, Jan. 1986.
20.  Jcyapalan, J.K. and Ethiyajeevakaruna. S.W.,  "JLAMINA-IBM
    P.C. Software for Calculating Stress and Strains in Composite FRP
    Under Combined Loads," Report  of  the Hazardous  Waste Man-
    agement Center, Jan. 1986.
21.  Jeyapalan, J.K. and Ethiyajeevakaruna, S.W.. "JSS1P-1BM P.C.
    Software  for Performing Finite Element  Analysis of Soil-Tank-
    Pipe Interaction Effects," Report of the Wisconsin Hazardous Waste
    Center, Jan. 1986.
22.  Joor, "Plasleel Composite Double Wall Tank" Product Literature,
    Escondido, CA, 1985.
23.  Karp, R.R., Duffey, T.A. and Neal, T.R., 1983, "Response of Con-
    tainment Vessels to Explosive Blast Loading," J. of Pressure Vessel
    Techno!.. 205. Feb. 1983, 23-27.
24.  Kinsey. W.R., "Underground Pipeline Corrosion," Paper presented
    at ASCE National Water Resources Engineering Meeting, Jan. 1972.
25.  Knopp, P.V., "Inventory Reconciliation," Proc. Underground Stor-
    age Tank  Management Seminar, Department of Engineering Pro-
    fessional Development, University  of  Wisconsin,  Madison,  WI,
    Dec. 1985.
26.  Myers,  J.R., "Causes and  Cures of  Corrosion in Fuel Tanks,"
    Air Force Civil Eng.. 14. Aug. 1973.
27.  Myers, J.R., "Fundamentals and Forms of Corrosion," Proc. Leak-
    ing Underground Storage Tank Seminar, Department of Professional
    Development, University of Wisconsin, Madison, July 1985.
28.  NACE  Standard RP-01-69, "Recommended Practice,  Control of
    External Corrosion on Underground or Submerged Metallic Piping
    Systems," National Association of Corrosion Engineers, Houston,
    TX, 1969.
29.  NFPA 30-1984, "Flammable Burning Equipment," National Fire
    Code, /, NFPA, Quincy, MA, 1984.
30.  NFPA 31-1984,  "Oil  Burning  Equipment," National Fire Code,
    /, NFPA, Quincy, MA, 1984.
290    UNDERGROUND LEAKING TANKS

-------
31. NFPA Standards Book, Fire Protection Handbook, ISthed., NFPA,
   Quincy, MA, 1981.
32. Oliveira, D.P.  and Sitar, N.,  "Groundwater Contamination from
   Underground Storage Tanks, Santa Clara,  California," Paper pre-
   sented at National Symposium  on Aquifer Restoration and Ground
   Water Monitoring, Columbus, OH, May 1985.
33. Owens-Corning, "Fiberglass Double Wall Tank," Owens Corning
   Fiberglass Literature, Toledo, OH, 1985.
34. Pollulert, "FD 103 Fluid Detection  System" Product Literature,
   Pollulert, Indianapolis, IN, 1985.
35. Rosenvasser, G.R., Ol'Mezov,  V.I. and Sanzharov, Y.A., "Normal
   Pressures on Walls of Subterranean Structures," Soil Mechanics
   and Foundation Eng., 17, May-June 1980.
36. Spanner, Sr., J.C., "Acoustic  Emission," Pressure Vessel and Pip-
   ing Technology- 1985-A Decade of Progress, 1985, 613-632.
37.  Steel Tank Institute Fact Sheet,  Northbrook, IL, 1985.
38. Tessitore, J.L., "General Guideline for Underground Storage Tank
    Management Seminar,  Department  of Engineering  Professional
    Development, University of Wisconsin, Madison, Dec. 1985.
39. Total Containment, "TC-3000 Lake Detection System," Corporate
    Literature, Charlotte, NC, 1985.
40. UL 58,  "Standard for Steel Underground Tanks  for Flammable
    and Combustible Liquids," Underwriter Laboratories, Northbrook,
    IL, 1981.
41. UL 1316, "Standard  for Glass-Fiber-Reinforced  Plastic  Under-
    ground Storage Tanks  for Petroleum Products," Underwriter Lab-
    oratories, Northbrook,  IL, 1983.
42. Wilcox,  K., Flora, J.D., Haile, C. and Gabriel, M., "Variables in
    Tank Testing," Proc. Underground Tank Management Seminar, De-
    partment of  Engineering Professional Development, University of
    Wisconsin, Madison, Dec. 1985.
43. Wong,  F.M.G., Craft, W.J. and East, Jr., G.H.,  "Stresses  and
    Displacements in Vessels due to Loads Imposed by Single and Mul-
    tiple Piping Attachments," J. of Pressure Technol., 107, Feb. 1985,
    51-59.
44. Wood,  P.H. and Webster,  D.E.,  "Underground Storage  Tanks:
    Problems, Technology, and Trends," Pollut. Eng., 16,  July 1984,
    30-40.
                                                                                           UNDERGROUND LEAKING TANKS     291

-------
             Performance  Evaluation  of Commercial Hazardous
                          Waste  Treatment  Facility Operations

                                                Ronald J. Turner
                            Hazardous Waste Engineering Research Laboratory
                                   U.S. Environmental Protection Agency
                                                Cincinnati, Ohio
                                                 Joan V. Boegel
                                              Metcalf & Eddy, Inc.
                                            Woburn, Massachusetts
ABSTRACT
  A number of the management processes employed by commer-
cial hazardous waste treatment facilities were evaluated under a
U.S. EPA research program. This program and the results of pro-
cess monitoring at two facilities are discussed.

TEST PROGRAM BACKGROUND
AND OBJECTIVES
  In  support of  RCRA and  the Hazardous and Solid Waste
Amendments of 1984 (HSWA), the U.S. EPA Office of Research
and Development (ORD)  is  seeking information on  the  ap-
plicability, effectiveness, capacity, cost and environmental impact
of existing  hazardous waste treatment technologies  which are
alternatives  to land disposal.
  The general waste types included in this study are several which
were identified  for  priority  action inthe  HSWA  legislation:
solvents, cyanides, metals,  corrosives and halogenated organics.
For solvents, the HSWA states that, effective 24 months after the
date of enactment, further land disposal is prohibited unless the
U.S.  EPA determines that such prohibition is not required in
order to protect human health  and the environment. If the Agen-
cy fails to  meet this deadline, these wastes  will  be banned
automatically from further land disposal.
  The  HSWA  specifically addresses the spent solvents  listed
under F001, F002, F003, F004 and F005. Some examples of other
waste types which are candidates for  land disposal restrictions are
F006 through F012 (metals and cyanides), D002 (corrosives) and
K015,  KOI6 and K029 (halogenated organics).
  A wide variety of established treatment processes, combina-
tions of processes and waste  streams are being considered for
evaluation under this phase of the ORD program which involves
commercial  off-site hazardous waste treatment  facilities. There
will be emphasis on the chemical and physical treatment methods
including cyanide oxidation, chromate reduction, neutralization
of corrosive wastes, heavy  metal  precipitation, evaporation,
steam  stripping and various forms of solvent  distillation. In-
cineration and other thermal treatment processes and the  waste
categories for dioxins (F020 to F023) and PCBs are being investi-
gated by others and are not included in this discussion.

APPROACH
  The primary criteria used  to select commercial facilities for
evaluation under this program are:

• Ability to treat hazardous wastes scheduled to be  banned or
  restricted  from  land disposal
• Well run facilities employing proven technology that could be
  representative of best demonstrated available technology
  A number of sources were used to develop a list of candidate
facilities, including the following:
• 1985  Industrial  and Hazardous  Waste  Management Firms
  Directory
• RCRA Part B computer listing
• State agencies
• Trade associations
• Equipment vendors
• Conferences and seminars
  Once  identified, the candidate  facilities were contacted about
the waste types they treated and the treatment methods used; we
also attempted to ascertain whether their company would be in-
terested  in participating in the program. With the cooperation of
the companies, arrangements were made to visit each site and to
discuss testing of specific operations. Based on these visits, several
facilities were selected for one week of sampling and monitoring.
The results of the first two field  tests are the subject of this paper.

PROCESS EVALUATIONS
  The treatment processes and  waste types of interest at two of
the hazardous waste treatment  facilities are given in Table 1.
                         Table 1
          Waste Treatment Processes and Waste Types
Facility         Process

   A      Distillation


   8      Fuel Blending


   B      Alkaline Chlorination
         Aqueous Waste Treatment
           Lime Precipitation
         Pressure Filtration
           Carbon Adsorption
  Waste Type (RCRA Codes)

Organic Solvents
(F001. F002. F003, F005)

Organic Liquids and Sludges
(0001, F003)

Cyanidp Liquids and Sludges
(F007. F008. F009, F010,
 F011, F012 and F019)

Metal-laden and Organics
(DOOZ. D007)
  Both facilities accept and treat corrosives, metal-laden aqueous
wastes,  aqueous  cyanide  wastes,  halogenated  and  non-
halogenated  spent solvents. The waste treatment processes are
described briefly and analytical results are presented.
292   SITE MANAGEMENT

-------
 Batch Distillation
   As illustrated in Table 2, Facility A processed several types of
 waste solvents from one or more generators during the test pro-
 gram. Wastes containing less than 40% of the major solvent com-
 ponent are  not  accepted by  this  facility due  to unfavorable
 economics for reclamation. Facility A places an upper  limit on
 suspended solids  content of the waste solvents  of about 10 to
 15%. Solvent wastes with higher solids content may be accepted,
 but the generator would be charged a premium for disposal of the
 excess solids.

                            Table 2
       Generator's Manifest Data for Waste Solvents, Facility A
Day  Batch  Generator  Industry Type   Major Components

          Electronics
                    Waste Code
           Electronics

           Biotechnology Research

           Electronics
           Plating
           Specialty Chemical
            Manufacturer
           Motor Rebuilder
1,1,1  trichloroethane   F001
freon  113
isopropanol

isopropanol            D001

methylene chloride      F002

methyl ethyl ketone     F005
   Figure 1 is a simple flow diagram for the batch distillation pro-
 cess. The feed is charged to the still pot and steam, in an internal
 coil, is applied to heat the liquid to boiling. Vapors rising from the
 still  pot are condensed by cooling,  and collected. Figure 2 il-
 lustrates the fate of the waste solvents processed during the test
 period. The major limitation  on batch distillation was the  140 °C
 maximum temperature achievable due to the capacity of the site's
 steam boiler. The facility plans to install a vacuum system to over-
 come this limitation.
   Data on the still bottoms from four batches of waste solvents
 are presented in Table 3.  The gas chromatograph data for the
 F001 waste was discounted because of the large discrepancy be-
 tween the reported Freon 113 concentration of the bottoms, the
percent chlorine and the process conditions. The principal organic
components analysis for the second waste (D001) accounted for
only 36% of the total mass as isopropanol, xylene  and ethanol.
However, the organic analyses accounted for 98% of the F002
batch and 70% of the F005 batch.

Fuel Blending
  Organic  waste  blending operations  constitute about 40% of
Facility  B's  business.  Incoming wastes for  the  process  are
segregated according to chlorine and moisture content, ground to
particle sizes of less than 3 mm  and placed in  agitated storage
tanks. The  final blend is homogeneous, easily pumpable and
suitable for burning in  a liquid injection incinerator.  The high
chlorine waste blends (up to 45% chlorine and free of PCBs) are
used by a manufacturer of low alkali cement. The  low chlorine
fuel blends (less than 6% chlorine) are sold to operators of kilns
or furnaces that are permitted to store and burn  U.S. EPA-listed
hazardous wastes. Figure 3 depicts the organic waste blending
operation.
  Two batches of organic wastes were processed during the test
period. The first waste was a low chlorine content waste drawn by
vacuum from drums of paint sludges (waste  code D001). The se-
cond waste was a mixture of varnish, paints, alcohols and xylene
(D001 and F003). As a result of the grinding process,  the suspend-
ed solids content  of the  treated organic  wastes increased, the
viscosity improved and a homogeneous material was produced.
Table 4 presents the results of tests on the two fuel blends. Ap-
proximately 80%  of the low chlorine waste  included in the first
blend was pumpable. The remainder of the material was solidified
with lime and transported to a land disposal facility.
                               IN-LINE
                               SCREEN
                                 t
                               SOLIDS
                                          SPENT FILTER
                                           BAG WITH
                                            SOLIDS
                       SAMPLE LOCATION
                                                            Figure 1
                                          Flow Diagram/Sample Points for Batch Distillation
                                                                                                     SITE MANAGEMENT    293

-------
                                  WASTE SOLVENT
                                  1543 GALLONS
                                                                    80JGALLONS
                                                                        52%
                   PRODUCT SOLVENT
                   FOR SALE
 604 GALLONS
63 GALLONS

    4%

55 GALLONS

   Ti%
19 GALLONS <
   i~2%'
                                                             Figure 2
                                                       Fate of Waste Solvent
                                                                                     REDISTILL	





                                                                                     FUEL



                                                                                     SOLIDIFY FOR LANDFILL


                                                                                     WASTEWATER TREATMENT
                            Table 3
              Still Bottoms Characterization, Facility A
Parameter
Viscosity (P25 C)
Paint Filter Test
Density
COO
Total Solids
Total Volati le Solids
Total Suspended Solids
Total Vol Sus Solids
Total Vol Dis Solids
Total Dissolved Solids
Silica
Metals
Copper
Lead
Chromi urn
Silver
Iron
Nickel
POC
Acetone
Butanol
Iso-Butanol
Dloxane
Ethanol
Freon 113
Isopropanol
Methanol
Methylene Chloride
Methyl Ethyl Ketone
1 ,1,1-Trichloroethane
Toluene
P-xylene, H-Xylene
0-xylene
TOX
EP TOX
TCLP
% Carbon
% Hydrogen
* Hater
BTU Content
Symbols:
LT Less Than
GT Greater Than
NS = Not Sampled
ND = Not Detected
Units
cSt
--
g/ml
mg/l
t
% of TS
mg/ t
mg/l
mg/l
mg/l
wet

mg/kg
mg/kg
mg/kg
mg/kg
mg/kg
mg/kg

mg/l
mg/l
mg/l
mg/ 1
mg/l
mg/l
mg/l
mg/l
mg/ 1
mg/l
mg/l
mg/ 1
mg/l
mg/ 1
wU Cl


wt£
wt$
wt*
Btu/lb

Batch
Day 1
(F001)
4.56
Fail
0.952
I
71
100
6700
5600
560000
560000
NA

NA
NA
NA
NA
NA
NA

NO
ND
ND
150000
ND
ND
ND
ND
ND
ND
37000U
ND
ND
ND
8.85
X
X
73.33
11.82
0.04
15722

NM Nol
NA = Nol
1
X
1 Batch
Day 2
(0001)

Fail
0.789
1
10
100
14
12
100000
1 00000
NA

NA
NA
NA
NA
NA
NA

ND
ND
ND
ND
2100
NM
360000
ND
NO
ND
ND
ND
ND
10000
0.071
X
X
66.73
13.98
2.86
15244

Measured
1 Batch 1
Day 3
(FOO?)

Fail
0.866
1
0.63
82
170
160
4100
4800
LT 0.005

2.3
NS
0.096
LT 0.05
1.4
NS

1300
7500
640
4000
14000
NM
98000
230000
ND
370000
NM
11UOOO
3100
LT 500
1.07
X
X
57.81
11.41
7.13
12339


Applicable Under Ihc Circum
Batch 2
Day 3
(F005)

Fail
O.R5H
I
6.7
98
1400
860
66000
66000
0.01

2.6
NS
0.65
0.052
44
NS

LT 500
11000
LT 500
6400
34000
NM
56000
14000
NO
530000
NM
18000
23000
4300
1.11
X
X
52.29
11.48
13.79
11555


siunccs
Interference Prevented Quantilalion
- Analytical Results Arc Nol Yet Available
                                                                   Alkaline Chlorination
                                                                     Alkaline chlorination is a well-established technology for the
                                                                   destruction  of simple  aqueous  cyanide mixtures.  Typically,
                                                                   cyanide oxidation is achieved by raising the pH of the wastewater
                                                                   to  8.5-9.0 and adding chlorine. In practice, many of the spent
                                                                   metal cyanide compounds require additional amounts of time and
                                                                   chlorine to complete the conversion of the complex cyanides to
                                                                   cyanates or to carbon dioxide, nitrogen and water.
                                                                     At   Facility  B,   cyanide-containing  aqueous   wastes   are
                                                                   segregated and treated in a single-stage process. Sodium hydrox-
                                                                   ide is used for pH adjustment if necessary, and sodium hypo-
                                                                   chlorite is used as the source of chlorine. Wastes containing more
                                                                   than  4%  cyanide are diluted to this  level  before  treatment to
                                                                   minimize the temperature effects on the reinforced fiberglas reac-
                                                                   tion vessels. Following the reaction, the waste is mixed with other
                                                                   aqueous wastes for further treatment (described in next section).
                                                                    The cyanide wastes treated  during  the sampling program at
                                                                   Facility B included cyanide rinse waters (F007-F012) and cyanide
                                                                   sludges (FOI9). Table 5 presents the results of the analytical tests
                                                                   performed on  these wastes. Based on  the data obtained during
                                                                   this test, the alkaline chlorination process effluent concentrations
                                                                   were  less than  5 mg/l.

                                                                   Aqueous Waste Treatment
                                                                    The aqueous waste treatment system at Facility B is comprised
                                                                   of a number of major unit processes arranged in a series: pH ad-
                                                                   justment, chemical reduction, lime precipitation, pressure filtra-
                                                                   tion,  neutralization  and granular activated  carbon adsorption
                                                                   (Fig.  4).  The aqueous wastes  were  characterized as  D002.
                                                                   However, they originated from several generators and some had
                                                                   received some previous treatment within Facility B.
                                                                    The effluent from the alkaline chlorination process (75<%) and
                                                                   non-cyanide aqueous wastes received at the site (25^o) are pumped
                                                                   to equalization tanks. The blended wastewaters then are pumped
                                                                   to pH adjustment tanks  for the addition of waste acid to  pH 2.
                                                                   This  acidification serves to  break  some  metal-organic chelate
                                                                  complexes, but primarily introduces metal wastes for removal in
                                                                   the next process, lime addition and precipitation of metal hydrox-
                                                                  ides. The resulting hydroxide slurry is dewatered by a plate-and-
                                                                   frame filter press  to approximately 45% solids. The sludge filter
                                                                  cake is trucked off-site for land disposal.
294    SITE MANAGEMENT

-------
            SEGRAGATED
            ORGANIC WASTE
            MATERIAL




OARSE TRASH
RINDER PUMP
UMP

F
C
P




INE
RINDER
UMP
.
*
                                                                         Figure 3
                                                            Organic Waste Blending, Facility B
                                                                                                        SEGRAGATED
                                                                                                        ORGANIC WASTE
                                                                                                        BLEND/STORAGE
                                                                                                        TANKS
                                                                                                                                    TO FUEL
                                                                                                                                    OR CHLORINE
                                                                                                                                    SUBSTITUTE
                                                                                                                                    OUTLET
                                                                          Table 4
                                               Analytical Results for the Fuel Blending Process, Facility B
                              Sample Period One
  Toxic Ketals

  Cadmium
  Total Chromium
  Copper
  Lead
  Nickel
  Zinc

  Other Analysis

  Total Organic Halide
  Total Solids
  Total Volatile Solids
  Total Suspended Sol ids
  Total Dissolved Solids
  Total Volatile Suspended Solids
  Total Volatile Dissolved Solids
  EP Toxocity
  TCLP
  * Carbon
  % Hyoroyen
  BTU Content
  % Water
                                   mg/kg
                                   mg/kg
                                   mg/kg
                                   mg/kg
                                   mg/kg
                                   mg/kg
                                   weight %
                                   ppm
                                   ppm
                                   ppm
                                   PPm
                                   ppm
                                   ppm
                                   we i gh t %
                                   weight %
                                   BTU/Ib
                                   weight %
          NS
          NS
          NS
          NS
          NS
          NS
       0.074
      69,000
      58,000
      26.00U
      43,000
      21,000
      37,000
          NS
          NS
          NS
          NS
       2,472
          NS
           NS
           NS
           NS
           NS
           NS
           NS
       NS
       NS
       NS
       NS
       NS
       NS
0.26
130,000
110,000
46,000
84,000
37,000
73,000
NS
NS
NS
NS
786
NS
0.78
120,000
100,000
70,000
50,000
55,000
45,000
NS
NS
23.13
10.71
2,179
80.56
     NS
     NS
     NS
     NS
     NS
     NS
                                                                                                             Sample  Period  Two
Pollutant
Units
Low Cl?
Organic
Li quid
Composite
Influent
Blend
Treated
Fuel
Blend
Organic
Liquid
On
S'
Non-1
                                                                                          Organic
                                                                                           Sludge
       220
       390
     3,900
    36,000
        62
     1,31)0
                                                                                                         Organic
                                                                                                          SIudge
                                                                                                       Non-Puropable
2.32
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
410,000
240,000
NS
NS
NS
NS
X
X
40.45
7.94
5,217
NS
         3,400
           260
         5,100
        70,000
           260
         8,500
                                                          NS
                                                     560,000
                                                     540,000
                                                          NS
                                                          NS
                                                          NS
                                                          NS
                                                            X
                                                            X
                                                        59.15
                                                        10.33
                                                       15,350
                                                          NS
                                                                               Organic
                                                                                Sludge
                                                                             Non-Pumpable
            5,000
              170
              510
            8,600
               31
              610
                                                           NS
                                                      530,000
                                                      260,000
                                                           NS
                                                           NS
                                                           NS
                                                           NS
                                                            X
                                                            X
                                                        40.75
                                                         6.94
                                                        8.443
                                                           NS
                                                                                  Organic
                                                                                   SI udge
                                                                                Non-Pumpable	
              30,000
               1,400
                 800
               5,900
                 650
              14,000
                                                                                 Treated
                                                                                   Fuel
                                                                                  Blend
                   NS
                   NS
                   NS
                   NS
                   NS
                   NS
NS
690,000
620,000
NS
NS
NS
NS
X
X
64.85
10.91
14,765
NS
0.79
390,000
350,000
330,000
60,000
280,000
70,000
NS
NS
58.17
10.15
11,601
11.14
  NS = Not Suspended
  X = Analytical Results Are Not Available Yet
                                                                          Table 5
                                  Analytical Results for the Cyanide Treatment Process at Facility B (F007-F012, F019)
                          Sample  Period One
                               Cyanide
                              Treatment
                              Influent
Po]J_utant	Units

Toxic  Metals

Cadmium               mg/1
Total  Chromium        mg/1
Copper               mg/1
Lead                  mg/l
Nickel                mg/1
Zinc                  mg/1

Other  Analysjj
Total Organic Carbon  mg/1       20,000
Total Organic Halide  weight %    0.30
Cyanide               mg/1        5.8UO
 Cyanide    Cyanide
Treatment Containing
Ef f 1 uent	51 udge
                                   45
                                  2.3
                                1,300
                                  2.8
                                  450
                                3,200
    3.5
     12
     42
    1.9
    6.2
  1,000
                                           5,700
                                            0.68
                                            LT 5
» 11,000
   I 180
I 34,000
    I 36
   * 750
   I 750
                  NS
                  NS
            I  96,000
  NS = Not Sampled
  I  = Interference
                                                  Sample Period Two

             Cyanide    Cyanide    Cyanide     Cyanide     Cyanide     Cyanide     Cyanide     Cyanide
            Treatment  Treatment  Containing  Containing  Containing  Containing  Containing  Containing
            Influent   Effluent     Sludge	Sludge      Sludge	Sludge  	Sludge	Sludge
  230
   24
6,200
   21
1,400
1,500
             20,000
              10.19
             11,000
                              #  = mg/kg
                              X  = Analytical Results Are Not Available Yet
 160
  19
,300
 8.5
 730
 920
          6,900
         LT  0.6
          LT  5
» 1,500
   * 54
  » 150
   * 67
  * 280
* 1,200
 I 1,300
I 14,000
* 15,000
 I 1,900
   I 490
» 13,000
    # 24
* 26,000
I 15,000
 0 8,500
   0 270
 I 4,700
  I 120
   * 73
  0 720
  f 130
* 1,300
I 9,700
 » 2,400
* 15,000
I 21,000
 I 2,600
   » 430
* 46,000
                                                                                                                                                  *  7
I 1
  t
  I
,200
 250
 600
,700
 200
 480
NS
NS
',000
NS
NS
1 9,000
NS
NS
» 8,000
NS
NS
» 1,900
NS
NS
* I
NS
NS
0 69,000
                                                                                                                           SITE MANAGEMENT    295

-------
               WASTE ACID STORAGE
                                                                                   VIRGIN
                                                                                HYDROCHLORIC
                                                                                 ACID TANK
   AQUEOUS WASTE
   RECEIVING TANKS
                                                      FILTER PRESS
                                                      SURGE TANKS
                          MULTI-MEDIA  CARSON
                          FILTRATION   ADSORPTION
                          TANKS       TANKS
                                                                                          OEWATERED SLUDGE
                                                                                          TO LANDFILL
                                                           Figure 4
                                                Aqueous Waste Treatment Facility
  Filtrate from the press is neutralized with hydrochloric acid and
passed through two multi-media filters in series followed by two
beds of granular activated carbon (9100 kg each) before discharg-
ing into the public sewer collection system. Facility B's effluent
discharge permit contains  limits on heavy  metals, chlorinated
compounds,  phenol and total organic carbon. Phenol  break-
through  is monitored as an  indicator of carbon exhaustion.
Breakthrough typically occurs after about 20 to 25 days of opera-
tion.
  The analytical results for the principal components found in 3
days of sampling the filtrate, filter press sludge and carbon col-
umn effluent are given  in Table 6. Heavy metal concentrations in
the final effluent  were  acceptable for discharge. The carbon col-
umns removed the organic compounds amenable to adsorption
(trichloroethane,  toluene,  total  xylenes  and   phenol),  but
moderately high  concentrations of acetone were present in the
GAC column effluent.  The TOC concentrations were reduced by
an average of 90°7o.
                                                           Table 6
                                     Analytical Results for Aqueous Waste Treatment, Facility B
Pollutants
Methylene Chloride
Acetone
Trichloroethane
Toluene
Chlorobenzene
Ethyl Benzene
Total Xylenes
Phenol
Cadmium
Total Chromium
Copper
Lead
Nickel
Z1nc
Cyanide
TOC
Units
ug/1
-ui/l
ug/1
ug/1
ug/l
«JL/'
ug/1
ug/1
mg/l
mg_/l
mg/l
-J59/1
pg/1
mg/l
mg/l
mg/l
Discharge
Permit
Limits
21,400
__
6,160
76S
765
__
..
6160
1.1
2.16
4.8
3.43
11.48
4.26
1.33
1500
Filtrate
220,000; 43.000;
130,000
68,000; 61,000;
69,000
4200; 2500;
1600
13,000; 7700;
3900
ND; ND; ND
NO; NO; ND
ND: 8,000: ND
19,000; 4.J60;
5100
1.2; 7.9; 50
2.1; 1.9; 0.7
160; 170; 380
1.4; 0.8; O.fi
2.2; 2.3; 22
24; 16; 46
NS
2500; 2400;
5000
Filter
Press
Sludge
ND; 61.000:
79.000
260,000:
NO; ND
GAC Effluent
ND; ND:
30,000
140,000; 160,00;
55,000
'400'. 000; '160. 006;""
110,000 ND; ND; 1600
1,400,000:
580,000; 290.000
15,000; 34.000;
11,000
110,000; 73,000;
55,000
57.000; 720,000;
300,000
ND; ND; NP
12,000; 6,700;
2900
	 rtT.W; 23,000;
2200
50,000; 43.000;
25,000
1.300: 79(1;
830
9,300: 6,400;
1400
35,000; 14,000;
4000
900; 400;
	 650 	
NS
ND; ND; ND
NO; ND; NO
ND; NO; NO
NO; ND; ND
ND; NO; ND
0.15; 1.5; 1.4
0.1; 0.12; 0.09
0.32; 0.9; 0.6
0.56; 0.7; 0.6
2.2; 2.0; 1.8
1.4; 1.6; 0.7
NS
IlrjuTTSOU
1400
296    SITE MANAGEMENT

-------
CONCLUSIONS
  The treatment technologies evaluated at Facility A or Facility B
were batch distillations, fuel blending and alkaline chlorination/
aqueous waste treatment.  If simple batch distillation is the sole
waste solvent treatment process, the recovered solvent must be ac-
ceptable for  use either as recycled to the process which originally
generated  the waste solvent or in some other application.  This
limits the types and concentrations of impurities that can be ac-
cepted in the influent and this limitation is specific to the end use.
  The waste segregation,  blending and grinding operation ap-
pears to be a practical approach to management of organic solids
such as  resins, latex, paint sludges and distillation bottoms for
ultimate disposal as a fuel or chlorine substitute suspension.
  Finally,  the  alkaline chlorination/aqueous  waste treatment
system was effective in treating cyanide and metal-laden wastes,
producing an effluent suitable for discharge to a publicly owned
treatment works.
  The data from these two facilities and other locations will pro-
vide process and residual  characterization information that will
contribute  to  the  determination of the adequacy  of available
treatment options.

DISCLAIMER
  Although the research described in this article has been funded
wholly or in part by the U.S. EPA, it has not been subjected to
Agency review and therefore does not necessarily reflect the views
of the Agency and no official endorsement should be inferred.
                                                                                                   SITE MANAGEMENT    297

-------
                           An Examination of  Siting Problems
                                          for Off-Site  TSDFs

                                             Douglas  B. Taylor, P.E.
                                               AeroVironment,  Inc.
                                               Monrovia, California
ABSTRACT
  The number of off-site Treatment Storage and Disposal Facili-
ties (TSDFs) in  the United States is decreasing. Because of the
social stigma associated with a facility which manages hazardous
waste and the high public visibility of an  off-site  facility, pro-
posed commercial TSDFs are finding it increasingly difficult to
get approval to operate. Meanwhile, several factors are working
to close currently existing facilities.  This decrease in  TSDFs is
going to lead to an increasing problem for waste generators—how
to manage their waste.
  The Solid and Hazardous Waste Management Committee of
the ASCE has been researching this problem over the last year. A
previous study confirmed that a decrease in sites has occurred in
the period 1980-1985. In addition, the revisions to RCRA have
forced many land disposal facilities to close and may have a sim-
ilar impact on other types of facilities in the next 3 years. Addi-
tional work has  been done to evaluate the reasons for the overall
decline. Specifically, a state survey is being conducted to identify:
(1) why  existing facilities have shut down during 1980-1986 and
(2) what  has caused the denial  of permits  for most of the pro-
posed new sites during the same period.

INTRODUCTION
  The siting of new off-site Treatment,  Storage and  Disposal
Facilities (TSDFs) for hazardous waste management has become
increasingly difficult over the past several years. In addition,  exist-
ing facilities across the country are feeling increased pressures to
cease operations.  Such  pressures are being exerted by various
organizations: local citizens  who fear the long- and  short-term
health effects of a TSDF in their town; regulatory groups who
identify  technical design or  operational problems which violate
existing regulations; and insurance brokers and company manage-
ment who see long-term financial risks.
  Off-site facilities are  defined in this paper as TSDFs which
accept waste from generators who pay the  facility to handle the
waste for them. Most clients of off-site  facilities are small gen-
erators for whom  a dedicated on-site TSDF is not  economically
feasible. In many cases, the clients include federal, state and local
governments. Because an off-site TSDF is a commercial enter-
prise, it is highly visible and thus subject to close public scrutiny.
Off-site  facilities that are hazardous waste landfills particularly
attract attention.
  Much of the opposition to off-site hazardous waste manage-
ment derives from poor public  relations. No number of logical
technical discussions can overcome past  horror stories. The mere
mention of any hazardous waste facility causes an immediate pub-
lic outcry in opposition.
  Many engineers and scientists who work in the hazardous waste
management field are concerned that the country will soon face a
crisis in the effective management of hazardous waste. All three
types of facilities are being affected: (1) treatment, (2) storage
and (3) disposal. Land disposal is being phased out out as a result
of the RCRA amendments passed in 1984 (however, off-site dis-
posal usually is considered as a cleanup method at every uncon-
trolled waste site).
  Treatment technology has not advanced  far enough so that:
(1) absolute destruction  of the waste can be  assured and (2) pro-
cess upsets will not occur. As a result, a release is possible and the
public does not want to take that chance in their towns.
  Storage  facilities for  hazardous waste are  only  a temporary
measure and do nothing to solve the real problem. Storage mere-
ly  serves as an intermediary step. Long-term  storage is nothing
more than aboveground landfilling before disposal or treatment.
As a result of these technical and public relations problems, it is
difficult to open or continue operation of an off-site TSDF.

PREVIOUS WORK
  The  Solid  and Hazardous Waste  Management Committee
(SHWMC) of the American Society of Civil Engineers has been
investigating the  problems of siting  off-site  TSDFs. A survey of
states was conducted from January through April of 1985  to de-
termine the status of new facility siting. In May of  1985 the results
of the  SHWMC survey were presented at  the Hazardous Ma-
terial Control Research Institute's (HMCRI)  Hazardous Waste
and Environmental Emergencies Conference.  That  paper, "The
Status  of Off-Site TSDF Siting," provided information  on the
then current number of TSDFs operating, the change since 1980
(RCRA implementation) and the success rates of  proposed facil-
ities.' The results of that survey are summarized below:

• As of May 1985, 566  off-site TSDFs operated in  this country.
  Approximately 50 of  these sites were disposal.  Ten states had
  no TSDFs; California had 133.
• A total  of  118 facilities were closed between November 1980
  and May 1985. (No reasons were given.)
• A total of 119 facilities were proposed in that same period.
• Of the 119 proposed TSDFs,  only 33 began operations (five
  disposal facilities).
• The  number of off-site TSDFs decreased by about 13% from
   1980 to  1985. In that  same period, only 28% of proposed sites
  have been approved.

  The  SHWMC study was done prior to the Nov. 8, 1985 dead-
line for land  disposal facilities to submit a Part B application.
That deadline has had  the impact of further  reducing the num-
298
       SITE MANAGEMENT

-------
                       Sites Which Existed
                       in November of 1980

                      RCRA Implementation
                            Add —
                       New Sites Approved
                      Nov 19X0 - Mar 1986
                                       Question: What about
                                       sites which were pro-
                                       posed but not approved?
                          Subtract —
                      Sites Closed Between
                      Nov 1980 - Nov 1985
                                       Question: What were
                                       the reasons for the
                                       closure of these sites?
                          Subtract —
                   Disposal Sites Closed Due to
                 RCRA Land Disposal Regulations
                           Nov 1985
                                       Question: What was the
                                       specific reason for
                                       closure of each site?
                        Existing Off-Site
                  Treatment Storage 4 Disposal
                           Facilities
                                       Question: How many more
                                       land disposal facilities
                                       will be closed before
                                       final pennit is issued?
                                       Also, how many other
                                       facilities will be closed
                                       by Part B calls in
                                       Nov 19S6 and 1988?
                         Original Number of
                       Land Disposal Facilities
Nov 8, 1985
         Subtract —
     Those Facilities Which
     Did Not Submit Part B
Application and Other Documents
Nov 8, 1985
   thru
Nov S, 1988
         Subtract —
   Those Facilities Which Are
      Later Found To Be
        in Violation of
 the Application They Submitted
      (During Inspection)
                                                                       By
                                                                   Nov 8, 1988
                           Subtract —
                     Those Facilities Which Are
                     Denied Their Final Permit
                   Ultimate Number of Facilities*
*Does not include facilities which
 open after 1985 or voluntarily close.


                            Figure 2
         Effects of Part B Deadline on Land Disposal Facilities
                           Figure 1
           Summary of Change in the Number of TSDFs
her of disposal facilities.  The specific impact of this deadline is
not yet known (Fig. 1). A certain number of facilities have not
submitted a Part B application and have stopped operation. Un-
doubtedly, some percentage of  those  who submitted certifica-
tions in November ultimately will be found to be out of com-
pliance. This decrease in  facilities only adds to the problems of
commercial sites still available for hazardous waste management
(throughout our work we do not discriminate "good" sites from
"bad" sites and we do not wish to imply that a  reduction in sites
is a bad change).
  The  problems  with  opening or  keeping open a TSDF apply
both on-site and off-site. However,  SHWMC  selected  to study
off-site facilities in particular because of  the direct impact off-
site TSDFs have on the majority of generators. Generators  who
use off-site facilities have very limited options  if their  TSDF is
closed.
  The  results of SHWMC's study are being used  as a basis for
further work. This will include an assessment of  the impact of the
land disposal permit deadline. Our followup study is described
below.
METHODS
  The followup study being conducted by the SHWMC is using
information provided by state contact people responsible for per-
mitting TSDFs. This list was provided by the U.S. EPA and has
been updated during previous contacts. The states will be asked to
provide information about  the  up-to-date  number of facilities
operating in each state and the problems TSDFs are experiencing.
The following questions will be asked:
• How many off-site TSDFs currently operate (accept waste) in
  your state?  If possible, break this number  down into treat-
  ment, storage and disposal.
• What were the reasons for the closure of off-site facilities which
  were reported in the  1985  survey? The reasons will be grouped
  into general categories: technical, financial, political, capacity.
• How many off-site disposal facilities did not submit their Part
  B application in November 1985? Do you suspect that more will
  be closed after permit review?
• What were the reasons for the rejection of off-site facilities
  which were reported in the 1985 survey?
  Written responses are being requested, and the responses will
be verified and/or clarified by telephone contact.
  After responses are received from each of the states, they will
be tabulated. The summary table will show the current number of
sites, the number of closed sites, the number of new sites and the
reasons for site closure and application rejection. The results of
                                                                                                      SITE MANAGEMENT    299

-------
 the state survey will be presented in the table just as they are pre-      a basis for establishing that the number of sites has been decreas-
 sented.  Only the responses related to the reasons will  be inter-      ing in the past 5 years. This survey will both update the informa-
 preted. As mentioned, this information will be grouped into cate-      tion generated  one year ago and explore more fully the reasons
 gories which best describe the specific situation at each site.            for this decline.
   The results of this followup survey are intended to provide in-
 formation to SHWMC and other interested members of the scien-      REFERENCES
 tific community on the current  status of TSDFs throughout the      ,  Tay,or  D c  .1Jhe Sta(us of off.SUe TSDF si(jng .. Proc  0/Hn
 United States and on the problems which are causing the recent        ardoM  Waste and Environmenlai  Emergencies Conference, Cin-
 decline in the number of facilities. The previous paper served as        cinnati. OH, May 1985, 95-97.
300   SITE MANAGEMENT

-------
              Spatial Data Research for  Hazardous  Waste Sites

                                             Timothy W. Foresman
                                   U.S. Environmental Protection Agency
                              Environmental Monitoring Systems Laboratory
                                               Las Vegas,  Nevada
                                             Lynn K. Fenstermaker
                        Lockheed Engineering  and Management Services Company
                                               Las Vegas,  Nevada
ABSTRACT
  Analytical and numerical models are used throughout the haz-
ardous  waste  site industry to  assist managers  investingating
groundwater conditions. An investigation based on geographic
information systems (GIS) was initiated in the San Gabriel Basin
in Los Angeles County, California. The objective of the pilot pro-
ject is to determine operational capabilities of a spatial-based sys-
tem to enhance management of the hazardous waste site data. An
initial goal of the research is to demonstrate feasibility of using
the GIS to run various groundwater flow and transport models. A
project  design  is presented which describes the  hazardous waste
spatial data research processes leading to input into two classes of
groundwater models. Anticipated strength of the GIS approach is
in the management of the spatial site data for multiple modeling
interfaces to offer increased  analytical options for the U.S. EPA
Site Offices.

INTRODUCTION
  Numerous analytical and numeric models are being used to
provide imput  for engineering and management decisions at haz-
ardous  waste sites. Groundwater models require  significant re-
sources to build  data  sets, often taking  months or years to
assemble. Resources in time and money once expended should
not require wasteful repetition due  to phase changes or manage-
ment adjustments of a site study. The ability to  switch  or choose
among different models is important to the success of a site eval-
uation.  However, the availability of data may curtail the use of
some models.  Efforts to satisfy the data requirements of mul-
tiple models easily can surpass the resources allocated for a site.
Geographic Information Systems  (GIS)  have  demonstrated
robustness and flexibility in natural resource applications over the
past few years. In particular, GISs permit maintenance of spatial
data sets and preparation of these data for multiple use analysis
programs.
  A research program was initiated by the U.S. EPA's Environ-
mental Monitoring Systems Laboratory (EMSL) in 1985 to assess
the utility of GIS technology in providing an interface  capability
between input data  and standard groundwater models for haz-
ardous  waste sites. The primary benefit to be gained for an en-
hanced  interface capability is the multidisciplary use of the mul-
tiple types of hazardous waste site data. Such a trend in capabili-
ties would provide cost sharing of data and enhance uses of the
data sets. To investigate these issues in a real world context, the
San Gabriel Basin Superfund site was selected for this pilot study.

SITE HISTORY
  The San Gabriel Groundwater Basin is comprised of approx-
imately 444 km2 (165 miles2) of groundwater bearing,  urbanized
                               LOS ANGELES COUNTY LINE _
                    OI020llll«i

                    0 10 20 30 Kllom«U

                  Approximate Sell*
                         Figure 1
    San Gabriel Basin Geographic Information System Pilot Study
land in east Los Angeles County (Fig. 1). Groundwater in this
basin is used extensively for public water supply with over 400
known public and private wells. Approximately 47 water pur-
veyors supply drinking water to over 400,000 residents in over 33
communities.' The geologic structure is composed primarily of
coarse alluvium lying along the southern base of the San Gabriel
Mountains; aptly described as a large "sand box." This area is al-
most completely urbanized.
  In 1979, the first  incident of groundwater contaminated by
chlorinated hydrocarbons was reported. A concentration of 1800
ug/1 trichloroethylene (TCE) was discovered in  the central por-
tion of the basin. The state and federal standard for safe drink-
ing water is  5  ug/1  TCE. Further sampling by the California
Department of Health Services revealed a widespread situation of
contaminated groundwater which included  unsafe levels of car-
bon tetrachloride (CTC) and perchloroethylne (PCE). The U.S.
                                                                                            SITE MANAGEMENT    301

-------
EPA became involved because of the toxicity of the contaminants
and the magnitude of resources required to mitigate and manage
this hazardous waste problem.1
  The San Gabriel Basin is atypical of CERCLA sites by virtue
of its geographic range. It is, in fact, a groundwater basin im-
pacted by numerous contaminant sources, some of which prob-
ably were originated decades ago by post World War  II indus-
tries.1 The complexity and size of this site demands the examina-
tion of approaches oriented toward large, spatial data base man-
agement systems. Spatial management tools with the flexibility to
potentially  interfere  with  existing groundwater  models  were
selected for examination in this study.

RESEARCH APPROACH
  An important value of managing hazardous waste site data us-
ing spatially-based information systems lies in  the organization
and storage of the disparate types of information or data. These
data are the essential descriptors of a site's physical,  chemical
and biological characteristics. By examining steps leading up to
the processing of groundwater models, other uses of these pro-
cessing capabilities for site operations should become evident.
  A flow diagram of the processing steps  for this project is pre-
sented in Figure 2. Most  technically mature CIS packages will
have  similar programming capabilities. This  research used  pro-
grams or algorithms developed for the ARC/INFO GIS  package
and will be noted  by parentheses in the  following  descriptive
text.4 The reader is referred to references5'7 for a more thorough
description of GIS components and capabilities.

DATA COLLECTION
   Data collection was based on a thorough review of existing data
from the San Gabriel area. Emphasis then was placed on defining
appropriate data to be used as source input to transport and flow
models. Use of transport and flow models was based on an assess-
ment of operational U.S. EPA and contractor practices.
DATA COLLECTION
• EXISTING MAPPED DATA
• EXISTING DIGITAL DATA
• EXISTING REPORT DATA

t
DATA ENTRY

CREATE MAP BASE
STANDARDIZE MAP DATA
PREPARE/DIGITIZE MAP DATA
PROCESS EXISTING DIGITAL DATA
EDIT COVERAGE/BUILD TOPOLOGY
ENTER ATTRIBUTE DATA
CREATE VERIFICATION PLOTS
t
DATA PROCESSING


• CONVERT MAP DATA TO UTM
• DEVELOP IRREGULAR POINT DATA MATRIX
• DEVELOP REGULAR GRID DATA MATRIX
t
HYDROLOGIC MODELLING
• FINITE ELEMENT (SUTRA)
• FINITE DIFFERENCE (MODFLOW)

                          Figure 2
                    Processing Flow Chart
  Additional data collection incorporated information necessary
to define: (1) the extent and characteristics of the aquifer; (2) the
extent of contamination by  various chemicals; (3)  spatial dis-
tribution of pumpage and  water levels over time; (4) hazardous
substance activities within a given area; and (5) population serv-
iced by water supply wells. Data were obtained from engineer-
ing contractors, local and  state water  agencies and several fed-
eral agencies.  To spatially  integrate all of the data sets into the
GIS, a  base map  at 1:62,500 scale was created. The data then
were entered  as either point, line  or  polygon (area) informa-
tion. Attribute data, such  as well water quality, were then re-
corded and referenced to a correlated point or area.

DATA ENTRY
  Data  entry of the  data sets required  the steps depicted in Fig-
ure 2. These steps were required to  establish and maintain qual-
ity  control standards for automating the data in terms of precis-
ion and accuracy consistent with National Mapping Standards.
Fifteen U.S. Geological Survey 7.5-min quadrangle maps were re-
quired for a base scale of 1:62,500. Standardizing the map data
entailed the precise integration of common  boundaries between
map themes such as  geology,  streams and soil using DIGITIZE.
This standardization eliminated the artificially created phenom-
enon of boundary slivers.  Preparation and  digitization of map
data required  the careful transfer of data from templates of ex-
isting mapped data,  such as potentiometric surfaces and annual
precipitation, into computer files with the use of digitizing tables.
Careful attention to ambiguities during this stage can eliminate
erroneous output in later site analysis. Existing digital data on
pumpage, water levels, chemical analysis and elevation data then
were integrated into the GIS (GENERATE).
  To remove numerous input errors, editing was performed and
all  data sets were attached to the proper spatial coordinates to
provide correct topologic structuring. All supporting file informa-
tion was entered to  these data sets (ADDITEM, DROPITEM,
JOINITEM). These  attribute data are descriptive files, such as
the chemical analysis of wellwater over time, which relate spatial
coordinates to data  base information. Displays then  were pro-
duced for the different information themes (for example, ground-
water boundary, population service areas, soil types and precip-
itation) to use in the determination of the accuracies of these ver-
ification plots.

DATA PROCESSING
  Once all the base map coverages and attribute data were coded
and verified, processing, analysis and creation of higher order or
derived information proceeded. The analytical processes required
to prepare data for hydrologic models are presented in Figure 3.
  Well locations were recorded in Universal  Transverse Mecator
(UTM) coordinates,  while other data sets were in units defined
by the digitizing process. To relate or match well information to
other data sets, a coordinate transformation algorithm was re-
quired (ARC TRANSFORM)
  The next analysis  step entailed the  linkage of water service
areas as defined  by  the water purveyors to specific  well loca-
tions and attribute data. This step allowed further correlation of
population data with water  service and site-specific well informa-
tion.
  The capability of the GIS  includes the manipulation of both
vector and raster (grid) data  formats. These formats  have spe-
cific limitations and uses in cartographic processing, storage and
retrieval. The reader  again is referred to references5"7 for detailed
information on uses of the two formats.
  A standard  grid spacing  and point of origin were required to
establish the location and relationship of cells in advanced spatial
processing stages. Grid spacing was designed at 0.25 mile cell size.
302    SITE MANAGEMENT

-------

r
R (LINE)
DATA

UTM
CONVERSION





I
WELL (POINT)
ATTRIBUTE
DATA
WAT
SERVI
AREA
                          Figure 3
          Groundwater Data Flow for Hydrologic Models

 The grid structure then was used to assign reference cells to both
 line and point data sets. Topographic data defined by elevation
 points were registered on this grid  structure. These grid registra-
 tion steps enable analysis of aquifer parameters, such as calculat-
 ing aquifer thickness and effective porosity. Pumpage rates could
 be assigned as aggregates within cells and used in further model-
 ing efforts. The  aquifer parameters  data sets  (e.g.,  hydraulic
 conductivity  and potentiometric  surface) were then  available
 through various GIS algorithms in both grid data matrix (GRID
 TOPO) and point data matrix formats (TIN and TIN GRID). At
 this stage of analysis, the hazardous waste site parameters could
 be  filed, reformatted if necessary and input to groundwater
 models.

 GROUNDWATER MODELING
  Testing the effectiveness of the GIS for support data input to
 groundwater models has to be completed. Two types of standard
 categories of models were selected for the modeling analysis:
 finite difference  (MODFLOW) and  finite element (SUTRA).
 Finite difference models use hydraulic parameters structured in a
 grid base format,  both regular grid and irregular grid. Finite ele-
 ment models use these same parameters but as discretely spaced
 points, as in the location  of wells in  the San Gabriel Basin.
  Application of algorithms for establishing gridded and discrete
 point networks are  major contributions of the GIS's  flexibility
 for model interfaces (GRID TOPO and TIN). These algorithms
 soon will be used to prepare irregular point data into finite ele-
 ment format to run the USGS. SUTRA groundwater  transport
 model.' The algorithms also will be used to prepare grid data for
 a finite difference format to run the USGS MODFLOW ground-
 water flow model.'

 DISCUSSION
  The use of a GIS provides potential improvements over conven-
 tional approaches in three general areas. These are: (1) data man-
agement and storage; (2) identification of site data spatial rela-
tionships; and (3) data  interface options for  models and users.
While these areas presently are difficult to quantitatively compare
in any  formal analysis, results of  user  oriented pilot  studies
demonstrate GIS strength for these  areas.10'12 The San Gabriel
Basin GIS demonstration should help focus  these strengths for
hazardous waste management. Operational use of GIS for haz-
ardous waste studies over the next few years  should provide the
necessary technology experience base to validate and quantify the
system's value to federal, state and private sector users.
  Central collection and spatial formatting of voluminous site
data is of high functional value to hazardous  waste site manage-
ment. A site can be probed and analyzed by various teams of
federal, state and contractor personnel and can experience major
changes in chemical or physical states. A spatial data manage-
ment system to handle dynamic inputs and facilitate the automa-
tion of analysis can save time for data aquisition, data analysis
and data update functions. These improvements also should have
significant economic benefits.
  Site data spatial relationships are important to proper environ-
mental assessments. The ability to correlate text information to
points, lines and areas of a hazardous waste  site is mandatory.
GISs are the only data base management systems built on a spatial
architecture which enables an automated  capability to perform
spatial assessments and proximity analyses. As a built-in  func-
tion, topology or spatial structure is  significantly better adapted
to toxic transport and fate issues than systems which must add on
such capabilities.
  Communication of site data to multiple users for multiple
analyses is a key aspect of efficient  site management. Final re-
sults of the pilot studies' hydrologic groundwater modeling inter-
face tasks should verify the design hypothesis that enhanced com-
munication amongst industry,  public and federal agencies  is de-
pendent upon  spatial data management. Capabilities of the GIS
could be used to develop internal groundwater modeling rou-
tines. However, the anticipated strength of the GIS is in the man-
agement of the spatial site data for multiple modeling interfaces,
thereby relieving the U.S. EPA Site Officers from restrictive ana-
lytical options.
  There always will be limits for many models in using parame-
ters and variables existing at a site.  This  limitation will not be
overcome  by  any  single  system.   However,  basic  hydraulic
parameters should be usable on a majority of industry approved
flow, transport and particle tracking models. This interface man-
agement ability in data  systems communications  is  within the
technical range of current state-of-the-practice GISs. The  U.S.
EPA's research effort will continue to assess the issue of range
and applicability of a GIS approach to hazardous waste site man-
agement.
REFERENCES
 1. California Department of Water Resources, "Planned Utilization of
   Groundwater Basin, San Gabriel Valley, Appendix A: Geohydrol-
   ogy," CDWR Bulletin, No. 104-2, Mar. 1966.
 2. Ecology and Environment, Inc., "Summary  of FIT Data Gather-
   ing Efforts and Source Investigation Workplan, Main  San Gabriel
   Basin,  Los Angeles County, California,"  TDD No. R-09-8303-1,
   Oct. 1983.
 3. Epstein, S.S.,  Brown, L.O. and Pope, C.,  Hazardous Waste in
   America, Sierra Club Book, San Francisco, CA, 1982, 593.
 4. E.S.R.I., "ARC/INFO  Users  Manual,"  Version  3.0 Redlands,
   CA, 1985.
 5. Clarke, K.E., "The Functional Capabilities of Geographic Informa-
   tion Systems," NASA, Ames  Research  Center, Ames, IA,  No.
   NCA2-OR 305-201, 1983.
                                                                                                   SITE MANAGEMENT    303

-------
  6. Dangermond, J., "A Classification of Software Components Com-
    monly Used  in  Geographic  Information Systems,"  Design  and
    Implementation of Computer-Based Geographic  Information  Sys-
    tems, Amherst, NY, IGU Commission on Geographical Data Sens-
    ing and Processing, 1983.
  7. Marble, D.F.  and  Peuquet, D.J., "Geographic  Information  Sys-
    tems and Remote Sensing," Chapter 22 in Manual of Remote Sens-
    ing,  Second Edition, American Society of Photogrammetry,  Falls
    Church, VA, 1984.
  8. Voss, C.I., "Saturated-Unsaturated Transport (SUTRA)," Water
    Resources Investigative Report 84-4369, U.S. Geological Survey,
    Washington, DC, 1984.
 9.  McDonald, M.G. and Harbaugh,  A.W., "Modular Groundwater
    Flow Model (MODFLOW)," Open File Report 83-875, U.S. Geo-
    logical Survey, Washington, DC, 1983.
10.  Marble,  D.F., Calkins,  H.W. and Peuquet, D.J.,  "Basic Read-
    ings In Geographic  Information Systems," SPAD Systems, Ltd.,
    Wiliiamsville, NY, 1984.
II.  Tomlinson, R.F. and Boyle, A.R., "The State of Development of
    Systems for Handling Natural Resources Inventory Data," Carlo-
    graphica, 18, No. 4, 1981, 55-69.
12.  Von Braun,  M.,  "Demonstration of  a Geographic Information
    System in the Analysis of a  Hazardous Waste Site,"  U.S. EPA-
    MERL, Cincinnati, OH, 1984.
304    SITE MANAGEMENT

-------
                         Sensitivity  Analysis of  Remedial  Action
                         Alternatives for Hazardous Waste  Sites

                                                Elio F. Arniella, P.E.
                                           Camp Dresser & McKee Inc.
                                              Hazardous Waste Group
                                                   Atlanta,  Georgia
                                           E. Lawrence  Adams, Jr., P.E.
                                           Camp Dresser & McKee Inc.
                                              Fort Lauderdale,  Florida
ABSTRACT
  This paper presents the sensitivity analysis methodology used in
the economic evaluation of remedial action alternatives at hazar-
dous waste sites. The authors base their presentation on actual ex-
perience  in  conducting sensitivity analyses in the economic
evaluation of alternatives for such sites. The paper recognizes that
in the  evaluation  of  remedial  action  alternatives,  certain
technologies have a larger degree of sensitivity than others and
establishes procedures for the evaluation of possible sensitivity
factors. The methodology gives decision-makers a range of unit
costs and statistically calculated levels of confidence associated
with each alternative.
  The objective of the costing exercise is to develop criteria for
obtaining a range of probable project costs. Once the range of
possible project costs is obtained, basic statistical principles are
incorporated to obtain levels of confidence of the cost estimates.
  The authors establish a baseline (present worth) cost for each
alternative,  including the following components:
  Mobilization and startup
  General site work
  Off-site hauling
  Contractor's labor, overhead and profit
  Equipment and materials
  Energy
  Monitoring and safety
  Legal,  license and permits
  Engineering and administrative costs
  Post-closure costs (present worth)
  Sensitivity factors for  each  of  the above categories  are
established and present worth costs are  calculated for each case
using a computer program. Combinations of sensitivity factors of
the principal cost component are added to establish the possible
project costs.
  A statistical analysis is incorporated to determine the baseline
price (mean) and other parameters, including the level of con-
fidence.

INTRODUCTION
  There is a degree of uncertainty in costs for remedial action
alternatives at hazardous waste sites. It is important to recognize
and evaluate this uncertainty in selecting the alternatives  on a
cost-effective basis.  This evaluation is  known as  a sensitivity
analysis.
  This paper presents  a methodology  for doing  a sensitivity
analysis.  The methodology  focuses on presenting the decision-
makers with a range of unit costs and statistically calculated levels
of confidence associated with «ach alternative.
  Factors basic to a sensitivity analysis, such as discount rate in a
present value analysis, are straight forward and are not discussed.
REASONS FOR SENSITIVITY ANALYSIS
  Not only is there a wide range of alternatives for remedial ac-
tion, but also within each alternative the actual costs may differ
significantly from the costs estimated in the feasibility study. Not
considering the potential difference in costs could result in the
selection of an alternative with actual costs more than estimated.
Moreover, the selected system could have ended up with a cost
higher than the actual costs of other alternatives even though their
estimated study costs were higher.
  The sensitivity analysis gives the decision-maker an opportunity
to judge the possible range of costs  for a certain alternative and to
decide whether to accept this range or select  another alternative,
with an upper range of costs less than the  first, but a greater
calculated "best estimate" cost.
  In doing the sensitivity analysis for an alternative, each cost
component must be carefully evaluated with respect to magnitude
and degree of certainty. A cost component with a high uncertain-
ty but with a very small range of possible costs compared to total
costs  may not need to undergo sensitivity analysis. Similarly, a
cost component with  a relative high proportionate cost of the
alternative may have little or no uncertainty and, hence, may not
need a sensitivity analysis.
  Examples of cost components and related  factors that may be
subject to a sensitivity analysis  are set forth below.

Project Schedule
  The actual project schedule can affect several items in the pro-
ject cost including:  energy, materials, resources and labor. An
early project finish may result in significant savings. However, a
late finish (due to weather delays, equipment malfunctions, unex-
pected  site problems,  etc.)  can increase  the project  cost
significantly.

Unexpected Site Complications
  Cleanup  of hazardous  waste  sites  containing buried  toxic
materials that have not been qualified or quantified in detail may
present unexpected project expenses or  savings during remedial
action implementation. For example, during a remedial investiga-
tion (RI)  phase at  a hazardous waste site, heavy metals are
detected with no documentation of waste disposal activities.  Elec-
tromagnetometer (EM) studies  reveal the presence of metals; in
addition, the  ground water  is  contaminated  with metals  and
volatile organics (VO). The remedial action  decision consists of
excavating the site, removing the metals andtreating the ground-
water. During-excavation, buried drums with unclassified solvents
are  unearthed. This situation will cause  additional expenses for:
monitoring and analysis, added  personnel safety equipment, time
delays and  an overall  increment in the excavation and disposal
costs."On the contrary, another possible  scenario may present no
complications and reduce the actual cost of excavation.
                                                                                                SITE MANAGEMENT   305

-------
Off-site Disposal
  There are only a few off-site RCRA-approved disposal facilities
in the continental United States. A cleanup remedy may consist of
disposal of hazardous wastes at a chemical landfill. Due to uncon-
trolled conditions, the facility planned for waste disposal may not
be available at the time of cleanup. This change in disposal plans
may result in hauling the material  to  another  facility located
several hundred miles further than originally planned. In addi-
tion, this new facility may have higher tipping fees, require new
transportation permits and demand  other disposal taxes which
may (depending on the  waste quantities and material to be dis-
posed of) increase the off-site disposal costs significantly.

Research and Development
  A feasibility study (FS) recommendation could  be based on the
satisfactory performance of pilot testing of waste treatment units.
The apparent preferred technology identified in the FS phase may
not prove  technically feasible when site-specific  conditions are
tested. Therefore,  another  option must  be evaluated  and re-
searched, resulting in a significant increase in the engineering, ad-
ministrative and capital  expenditures.

Legal Fees and Permits
  Unknown situations during implementation  of a remedy may
result in an increment of legal fees and permit application and ap-
proval  costs.  This condition  may affect  the  project  cost
significantly.

Post-closure Activities
  After implementation of a remedy, the post-closure monitoring
may be scheduled and budgeted for several years. However,  site
cleanup probably will exceed  the cleanup goals and, after consis-
tent verification, the site may be determined clean and require no
additional  monitoring. This positive development may result in
significant  savings  in  present  worth  costs  for post-closure
monitoring and analysis.
STATISTICAL ANALYSIS
  The sensitivity analysis  methodology  used in assessing  the
possible cost variations of  remedial action  alternatives uses  the
principles of the "Central Limit Theorem"  to calculate the con-
fidence limits that can be attributed to a range of probable project
costs. The "Central Limit Theorem," first  stated by Laplace in
1812', formulates that:
"For almost all populations the sampling distribution of
the mean derived from the  universe will be  approximately
normally distributed provided that  the associated sample
size is sufficiently large."2
  In other terms, this theorem states that if a random sample of
size n is drawn from a population with mean p. and  variance a2,
then as n  increases, the distribution of
                                        X - u
                                       ~~
                                             approaches  the
standard normal distribution.  It  has  been demonstrated em-
pirically2 that a sample size with 30 or more observations will be
"sufficiently large"  to  produce  an  approximately normally
distributed sample mean.
  Based on previous experience in performing sensitivity analyses
of remedial action alternatives, it is estimated that the number of
"what ifs" or uncertainties of different cost estimate components
may generate combinations and permutations of possible project
costs exceeding 30 or more probable cost products. Therefore,  it
is reasonable to presume that if the possible project costs may ex-
ceed 30 or more products, the analyst can evaluate the confidence
level of the data using the normal distribution equations. The nor-
                                                                  mal distribution formulas for unclassified (not grouped in cost
                                                                  class intervals) and classified data are presented below.
                                                                  Normal distribution formulas for unclassified data

                                                                                X  =   X/n                                  (1)
                                                                  and
                                                                                       1
                                                                                s -  [-    i. (X-X)  ]*
                                                                                       -
                                                                                                                           (2)
                                                                 where:
                                                                    X  = mean project cost or base cost
                                                                    X  = variable project cost obtained from  one or more pos-
                                                                         sible  variations of  cost components of the sampling
                                                                         population
                                                                   n   =  number of possible project costs (  30), and
                                                                    s   =  standard deviation

                                                                 Normal distribution formulas for classified data
                                                                    If the data have been classified into k classes,  then by defini-
                                                                 lion:
                                                                 where:
                                                                                        k                k
                                                                           Si    [-L  (  £   fi  XI - £  (  L   fi Xi)')]
                                                                                       1=0              1=0
                                                                   Si  = standard deviation for the classified data
                                                                   Xi  = the midpoint (class mark) of the ith class
                                                                    fi  = the number of observations in the ith class
                                                                   w  = incremental value of the class interval
                                                                    n  = the normal number of observations
                                                                                          I (fi Xi)
                                                                           Xi  - Xo + w  ('
                                                                                               fi
                                                                                                                          (4)
                                                                 where:
                                                                   Xi  =  the sample mean
                                                                   Xo =  value of the midpoint class mark at i = 0

                                                                 Confidence Limits
                                                                   The confidence limits of the confidence interval of unclassified
                                                                 or classified data evaluation are obtained as follows:
          xl 1 x 1
                                                                                         or
                                                                                                   X1
                                                                                                                          (5)
where:
  X and Xi are the unclassified and classified means, respectively,
         and s and si are the unclassified and classified standard
         deviation, respectively
  X,  =  lower confidence interval =  X -  zs or Xi -  zsi
  X;  =  upper confidence interval =  X  +  zs or Xi +  zsi
   z  =  one-half the area under the normal curve within a speci-
         fied confidence interval
  The value of z can be obtained from  integration of the area
under  the  normal curve or  from normal curve areas tables
available in most statistics books. Common z values used in the
sensitivity analysis  are shown in Table 1.

APPLICATION
  As previously discussed, the purpose of the sensitivity analysis
is  to  assess  the  effect  of  variations in  specific assumptions
associated with the cost estimates of remedial action alternatives.
The sensitivity analysis is specifically concerned with those factors
that could bring significant changes in the overall costs with only
a small variation  in  value.  In addition, factors  with  the most
uncertain values are chosen  for the analysis.
 306   SITE MANAGEMENT

-------
                             Table 1
        Normal Curve Areas for Certain Confidence Intervals
Confidence
Interval
(%)
95
90
50
30


z
1.960
1.645
0.675
0.385
Taken from Wide'


  An actual study of a hazardous waste site in Florida is used as
an  example to illustrate the  application of  sensitivity analysis
methodology. This study involved a site located over a sole source
drinking water aquifer. Heavy metals and PCBs were the basic
contaminants. Possible remedial alternatives included solidifica-
tion, solidification/stabilization, solidification/containment, par-
tial off-site hauling/stabilization/solidification, extraction/solidi-
fication, extraction/containment and incineration/solidification.
A sensitivity analysis example  performed  on the  incineration/
solidification alternative is presented in this paper.
  The preferred remedial action identified in the feasibility study
consists of incineration of about 2,400 yd3 of soils with 50 ppm or
higher of PCBs and solidification/stabilization of approximately
44,000 yd3 of soils with PCB concentrations below 50 ppm and
heavy metals.  Cost  estimates  for this  alternative include  the
following components.
• Mobilization and startup
• General site work
• Off-site hauling
• Contractor's labor, overhead and profit
• Equipment and materials
• Energy
• Monitoring and safety
• Legal, license and permits
• Engineering and administrative costs
• Post-closure costs (present worth)

  Once  the base costs  are estimated for each component,  the
analyst evaluated the assumptions made to establish the base costs
and asks  "what if"  the  presumed cost estimating conditions
                                                                 Table 2
                                         Possible Sensitivity Factors for Remedial Action Alternative




Sensitivity Factor
1.







PROJECT SCHEDULE
1.1 Early Finish
• Site Work
• Energy
* Labor
• Equipment
• Engineering 4 Admn.
• Monitoring 8 Safely

Probable Cost,
$1,000


821
12.2
800
1,100
6% a
283
Difference
from Base
$1,000


-341
-1.4
-500
-742
-4% a
-57


Justification for Cost


Unit cost of $9/yd' instead of $16.5 $/ydJ
Facilities used less than estimated
Work done with 60 percent of estimate
Less cost of rental equipment
No research needed; Pre-assembled equipment
Good post closure results
                            1.2  Late Finish

                                * Energy              18            +4.4
                                • Labor               1,396         +280
                                * Equipment           2,300         +458
                                « Engineering & Adirin.   17% a         +7% a
                                * Monitoring           200           +22
                                • Permits and Legal     13            +3

                        2.  SITE COPLICATIONS

                                • Site Work           2,000         +500
                                • Engineering & Admn.   17% a         +7% a
                                «Monitoring           500           +150
                                t Labor              .1,500         +200

                        3.  RESEARCH 8 DEVELOPMENT

                                t Engineering & Adm'n.   10% a         -0% a
                                * Engineering & Adm'n.   20% a         +10% a

                        4.  LEG9L FEES * PERMITS

                                »Legal               5% a          -5% a
                                • Permits and Legal     15% a         +5% a

                        5.  POST CLOSURE

                                • Monitoring           150           -108
                                                    258           0
                                                    458           +200
   Use 257,000 kwh @ $0.07
   Labor hours increased by 20%
   Higher cost for use of incinerator
   Extensive research and duplicated design
   20% more time at $l,000/week
   Extension of temporary permit
   Site excavation cost of $21.50 $/yd'
   Redesign based on actual site conditions
   Unexpected sources are discovered
   Additional cost due to unexpected situations
   No ccmplications during design
   Preferred alternative is not implanentable
   No need for litigation
   Litigation and unexpected permits
   5 years monitoring
   30 years monitoring
   30 years with more intensive monitoring
                           ) of capital cost
                                                                                                             SITE MANAGEMENT    307

-------
                            COST COMPONENTS
                       SUBJECT TO HIGHEST SENSITIVITY
                     A.  »lTt won*
                     B,  COMTRACTOM LAiOH AND Oaf
                     C. tOUIPMIMT 1 M«Tim»ll
                     0. MONfTOimO 1 I«HTY
                      E, ICO*!, LIClMtl »
                      .  INQINEtHlMO i AOMIMHTHATIV1
                      C. POlT-CLOiU*K COtTi
     •  PCftCCNT Of THf tUU OP A.I.C AND 0
     • • AMOUNT! M THOUSANDS UNIIM
        OTHCHWIM tPCCVKD
                           Figure I
          Cost Components Subject to Highest Sensitivity
change due to justifiable  reasons.  In addition, the lower  and
higher possible costs are established based on possible or expected
conditions. Table 2 summarizes the possible sensitivity  factors
established for the case study analysis. These possible sensitivity
factors are screened, and those that have the highest possibility of
influencing significant cost changes are selected for the analysis
(as shown in Fig. 1). In this particular case, it is determined  that
possible mobilization,  off-site hauling and energy cost variations
will not influence significant total  project cost variations (less
than 2% of the cost). The  remaining cost components, judged to
yield significant changes, are selected.
   This arrangement can product 210 cost permutations [n!/(n-r>!
or 7!/4!]. However,  this  analysis will generate many repeated
combinations  of  cost  products,  and for practical  purposes the
overall normal distribution parameters may not have significant
difference if  a smaller  sample  size  (larger  than  30) is used.
Therefore, the cost combination pattern shown in Table 3  was
used in this case study. This pattern  produced a sample size of 39
possible project costs summarized in Table 4.
   These data are evaluated by either the classified or unclassified
methods previously stated.
                             Table 4
       Possible Project Cosu for Remedial Action Alternatives
                                                                                          CoiU In Thouund Dollan
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
6.515
9,264
3.748
6.716
7.238
7,498
7.701
8,319
8.589
7.115
7.356
7,905
7,433
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
8,412
9.064
5.701
5.101
4.211
4.129
4,033
3.904
6.408
6.199
5.939
5.864
5,263
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
4.486
5,040
5,595
6,419
6,660
7,247
8,068
4,040
4,105
4.814
5.107
6,727
7,672
RESULTS
  The 39 possible cost products were classified in seven $1 million
class intervals ranging from $3 to $9 million, as shown in Table 5,
and  the frequency  histogram  is  presented  in  Figure 2. The
calculated  mean and standard deviations for the classified cost
data are $6.37 million  and $1.59 million, respectively. Likewise,
the sample mean and standard deviations for the unclassified data
are $6.30 million and $1.58 million, respectively. The results of
the evaluation and pertinent statistical parameters are summar-
ized  in Table 6.
                            Table 5
                    Analysis of Classified Data


Clm lnt*rwl
3.000-4.000
4,001-5.03)
5,001-6.000
6.001-7.000
7.001-8.000
8,001-9 .ODD
9,001-10,03]
TDTHL

Clm
*«..« F
3.500
4. SOD
5.500
6.500
1.500
8.501
9.500
-

n fvtknil
rt*»jcnc? fiv^gny
2 O.OSJ
7 0.1795
a 0.2DS1
7 0.17«
9 O.O09
4 1.10%
2 0.0513
J. 1

Cxulitlw
frWTC>
0.051)
0.2JOB
O.U»
0.6154
0^462
0.9<87
i.can

»
tkvtlttcn
fr^ ItV
Origin
0
1
2
3
4
5
t
21


fIJCI
0
7
16
21
36
20
12
112


rut
0
7
32
a
144
10)
72
418
                                                               Table 3
                                             Project Cost Combinations Used in the Analysis

A
B
C
U
E
F
G
1 2
1 2
1 2
1 2
1 2
1 2
1 2
1 2
3 4
3 1
3 1
3 1
3 1
3 1
3 1
3 2
5
1
1
1
1
1
2
2
6
1
1
1
1
2
2
2
7 8
1 1
1 1
1 2
2 2
2 2
2 2
2 2
9 10 11
1 2 2
2 1 2
2 1 1
2 1 1
2 1 1
2 1 1
2 1 1
12
2
2
2
1
1
1
1
13
2
2
2
2
1
1
1
14 15
2 2
2 2
2 2
2 2
2 2
1 2
1 1
16 17
3 3
1 3
1 1
1 1
1 1
1 1
1 1
IB
3
3
3
1
1
1
1
19
3
3
3
3
1
1
1
20
3
3
3
3
3
1
1
21
3
3
3
3
3
3
1
22
1
1
1
1
1
3
3
23 24
1 1
1 1
1 1
1 1
1 3
3 3
3 3
25
1
1
1
3
3
3
3
26 27
1 1
1 3
3 3
3 3
3 3
3 3
3 3
28
2
3
3
3
3
3
3
29
2
2
3
3
3
3
3
3D
2
2
2
3
3
3
3
31
2
2
2
2
3
3
3
32
2
2
2
2
2
3
3
33
2
2
2
2
2
2
3
34
3
3
3
3
3
3
2
35
3
3
3
3
3
2
2
36
3
3
3
3
2
2
2
37
3
3
3
2
2
2
2
38
3
3
2
2
2
2
2
39
3
2
2
2
2
2
2
         Nolc Refer 10 Fig I for actual values of A. B, C'. D, L. I- and G
             Numbers 1 ihrough 39 arc probable projcxi cosis.
308    SITE MANAGEMENT

-------
         s
         o
         o
         2
                             PROBABLE COST
                          ($ AMIS. IN MILLIONS)

                             Figure 2
                Histogram of Probable Project Costs
                                                                                                   Table 6
                                                                              Summary of Results for Economic Sensitivity Analysis
                                                                                        of Remedial Action Alternatives
                                                                      Statistical
                                                                      Parameters
                            Values for
                          Classified Data
    Values for
Unclassified Data
                                                                      Mean                            6.37
                                                                      Standard Deviation              1.59
                                                                      Coefficient of Variance"         0.25C

                                                                      Confidence Intervals
                                                      6.30
                                                      1.58
                                                      0.25°
-90%
+90*
-50%
+50%
3.74
9.00
5.30
7.45
3.70
8.90
5.23
7.37
                                                                      a In millions of dollars unless specified otherwise.
                                                                      b Coefficient of variance is defined as the ratio of the standard deviation(s) over the
                                                                        population mean (x).
                                                                      c Unitless.
CONCLUSIONS
  The alternative of incineration-solidification/stabilization of a
hazardous waste site shows that there is a place in the decision-
making process for sensitivity analysis of remedial action alter-
natives.
  The economic impact of various cleanup methods can be com-
pared in terms of levels of confidence of cost estimates using this
methodology.  By establishing possible variations  of cost com-
ponents for each alternative, the analyst can obtain a range  of
probable overall project costs. This range of project costs can  be
statistically analyzed to determine levels of confidence associated
with each alternative. The end results show that the apparent low-
est cost alternative may not necessarily be the one  with the least
economic risk.
REFERENCES
1.  Laplace, P.S.,  Theorie Analytique des Probabilities, 1st ed., Paris,
   1812.
2.  Parsons,  R.,  Statistical Analysis: A  Decision-Making Approach,
   Harper & Row Publishers, New York, NY, 1974.
3.  Camp Dresser  & McKee Inc., "Draft Feasibility Study Report for
   the Pepper's Steel and Alloys Site in Medley, Florida," Sept. 1985.
4.  Wine, R.L., Statistics for Scientists and Engineers, Prentice-Hall,
   Inc., Englewood Cliffs, NJ,  1964.
                                                                                                          SITE MANAGEMENT    309

-------
                     Design  of  a  Lateral  and  Vertical  Expansion
                            at an Existing  Interim  Status Landfill

                                                 Rodney T. Bloese
                                              Thomas G. Ryan, P.E.
                                        Foth & Van Dyke Industrial, Inc.
                                               Green Bay, Wisconsin
ABSTRACT
  The  Hazardous  and  Solid  Waste  Amendments  of  1984
(HSWA) have resulted in changes to the U.S. EPA's existing
hazardous  waste management program. Final rules  have im-
pacted lateral and vertical expansions of existing interim status
landfills. According to HSWA, the landfill expansion area must
comply with  minimum technological requirements of  RCRA
even if the facility no longer accepts or intends to accept haz-
ardous waste. An exception to the minimum technological re-
quirements would be possible if the expansion area were con-
structed and maintained as a separate and discrete  unit.  Within
the framework of the new rules,  proposed  remedial  work and
horizontal and vertical expansion at an existing zone of  satura-
tion interim status landfill site have been complicated.
  Control of seepage from in-place hazardous and solid waste
utilizing a groundwater containment structure was mandated by
the state regulatory agency (Wisconsin Department of Natural
Resources) as part of the existing  solid  waste landfill permit. A
program was developed and implemented to generate informa-
tion regarding containment design criteria and to define the pres-
ence of leachate at the site perimeter as well as the off-site migra-
tion potential of the leachate. The recommended design also is
impacted  by the HSWA; the potential exists that the  recom-
mended  containment structure meets  minimum  technological
RCRA requirements.
  The  proposed vertical and horizontal expansion design options
may have to be modified due to required separation of the exist-
ing hazardous waste unit from the  proposed solid waste  landfill
unit unless costly minimum technology RCRA designs, such as
double liner and separate leachate collection systems, are pro-
posed.  The design is further complicated by the lack of guidance
from the U.S. EPA as to what is an acceptable design for  separa-
tion of solid and hazardous waste units.

INTRODUCTION
  The  desire of  the  site operator was  to obtain a preliminary
indication from the state  regulatory agency as  to  the  potential
suitability of a lateral and vertical expansion to this existing in-
terim status landfill. The site operator desires an expansion that
will maximize refuse volume while producing final grades which
will conform  with the intended final use. In addition,  a con-
dition of the existing landfill operation  permit requires the con-
struction of a groundwater containment structure along two sides
of the existing site.
  Recent state and federal statutes have impacted the  design of
landfill expansions by limiting or complicating design options.
Presently, state statutes and guidelines are being revised.
  According to the U.S. EPA', a lateral expansion to an exist-
ing  interim status landfill must comply with the double liner and
leachate collection requirements. However,  this rule may only
apply to sites that continue to receive hazardous wastes.
SITE HISTORY
  The southwest portion of the existing landfill site (Fig. 1) was
used as a municipal dump in the early 1970s. Foundry sand and
municipal wastes fill this approximately 10-acre area. There are
no documented base grades  or leachate collection systems. Ac-
cording to a local contractor, a clay cutoff wall was constructed
along the western edge of the site. However, based upon incom-
plete construction documentation, the cutoff wall apparently did
not extend through a shallow sand seam.'
  The facility commenced operations as a sanitary landfill in the
mid-1970s. Four areas  were filled in a sequential manner. How-
ever, corridor 3 was constructed between corridors 1 and 2 by
excavating until the collapse of corridor 1 and 2 sidewalls pre-
cluded  further digging. These three corridors were constructed
without leachate collection systems and documented base grades.
Corridor 4 was excavated to a greater depth than the earlier dispo-
sal areas. Documentation of ultimate base grades or liner con-
struction was not performed. Installation  of a partial leachate
collection system was undertaken, but the system did not extend
to the base of corridor 4, and construction of the system was only
partially documented.'
  Corridor 5  was overexcavated by 5 to 10 ft. This excavation
resulted in backfilling the corridor with clay up to the approved
base grade. Documentation of the recompacted clay base and the
installation of the leachate collection system was performed.1
  The  area  between  the old municipal site and corridors  1
through 4 has been partially  filled with primarily foundry waste.
Again, the  base grade was not documented nor was a leachate
collection system installed.-'
                         Figure 1
                  Plan View Map of Landfill
310    RCRA SITE REMEDIATION & EXPANSION

-------
  Presently, corridor 6 has been constructed with  documented
base grades, 5 ft of documented compacted clay liner and a leach-
ate collection system. Approved for filling this corridor with solid
waste has been received from the state regulatory agency.
  The  site  operator was  licensed by the Wisconsin Department
of Natural Resources  (DNR)  to  receive hazardous waste from
1975 through 1982. On Nov. 19, 1980, Part A of the RCRA haz-
ardous waste facility permitting process was submitted to the U.S.
EPA. As a  result of submitting Part A and the Preliminary Notif-
ication on Aug. 18, 1980, the site received  "Interim Status" and
was allowed by the U.S. EPA to continue to accept hazardous
wastes. Since Nov. 19, 1980, various types of hazardous wastes
have been disposed with municipal solid waste in corridors 3 and
4. The hazardous wastes which have been  approved for receipt
and disposal under the federal Interim Status requirements in-
clude spent halogenated solvents, spent non-halogenated solvents
and spent plating bath solutions and wastewater treatment plant
sludges from electroplating operations.3 Hazardous wastes have
not been accepted at the site since Jan. 1,1983.

SITE GEOLOGY
  The  positions of the various strata  in the subsurface strati-
graphy are  illustrated in the generalized geologic cross-section in
Figure 2. The  strata are  comprised  of till  and proglacial fluvial
and lacustrine sediments of the Oak Creek Formation.4
  The  existing landfill and the  proposed expansion  area  are
located in  front of and  on the distal  slope  of the inner Lake
Border moraine. The moraine represents a  standstill position of
the ice margin. During glaciation, drainage to the  east was
blocked by the ice and  the north-south trending morainal  de-
posits and produced glacial ponding. These sedimentary units are
more likely to be continuous  parallel to the  former ice margin
(north-south orientation) rather than in an east-west direction.
The western part of the site area is dominated by lacustrine and
fluvial sediments while glacial and ice-marginal deposits are more
dominant in the eastern part of the site.4
  Redisposition of till and sorted sediment by mass-flow pro-
cesses is common in ice-marginal environments, and some of the
sediments  in the geologic section at the site probably resulted
from mass-flow.  Mass-flow at a ponded  ice-margin would re-
sult in mass-flow deposits interbedded  with fluvial or lacustrine
sediment.4  Because of the variety of processes involved in  the
deposition  of  sediments  at the site area,  the general geologic
cross-section (Fig. 2) may not be indicative  of true conditions
(i.e., the layers are probably not as interconnected or laterally
extensive).
                    EXIST  GROUND
                                  UNIT 6_	- - ~

                               UNIT 7
                          Figure 2
         Typical Geologic Cross-Section of Landfill Area
  The upper unit is composed of till. The lower units appear to
consist primarily of proglacial fluvial and/or lacustrine materials
with the exceptions of units 3 and 7, which may be in part till, and
unit 5, which may be a mass-flow deposit."
  Based  upon regional  data and nearby  water supply  well
records, the bedrock is at a depth of approximately 150 ft and is
composed of Silurian dolomite at the bedrock surface."
  In summary, the soils  beneath the site  are layered and non-
homogeneous. The geologic environment  is quite complex and
needs to be defined in greater detail.

SITE HYDROGEOLOGY
  The top of the zone of saturation (water table) exists generally
between 3 and 10 ft below the ground surface. The general trend
of groundwater flow is  toward the west, although the  water table
has been distorted by excavation and filling  within the  existing
site. The excavation within the present site has created a depres-
sion in the water  table  and an inward gradient toward the exca-
vation. However, along the south side and southwest corner  of
the  existing site, there is some question as to whether these inward
gradients are being  maintained due to possible leachate mound-
ing  within the filled areas.
  The shallow groundwater aquifers do not appear to be inter-
connected. The hydraulic conductivity  of the most shallow silt
and sand zone (unit 2) has a horizontal permeability range  of
several orders of magnitude (10~3 to 10~6 cm/sec). The grain size,
as measured by P2000 values, varies between 17  and  97%.4 The
fourth and sixth units also exhibit similar grain size and horizon-
tal permeability variations.

CONTAINMENT STRUCTURE
  As indicated previously in this paper, the NDR is requiring the
landfill owner to design and  construct a containment structure
around the south and west sides of the existing landfill site. The
purpose for requiring the containment structure is to ensure that
inward groundwater gradients are maintained into the landfill site
area. This design  will be intended to prevent the  outward migra-
tion of any contaminated water from the documented base areas
of the site.
  Three design options were developed for consideration by the
landfill owner for construction of the containment structure. The
first design  option  included the excavation of a trench  to con-
struct a soil/bentonite  slurry  wall. As part of the  design of the
slurry wall, the appropriate mixture of native soils and commer-
cially available bentonite and the compatibility of the soil/ben-
tonite slurry wall to leachate-contaminated groundwater would
be  determined. The leachate compatibility and soil/bentonite
mixture specifications would be determined in a series of labora-
tory tests and also would be based on the experience  of the geo-
technical contractor in similar applications. This option would in-
clude, in addition  to  the soil/bentonite slurry  wall, a second
trench constructed on the landfill side of the slurry wall to con-
tain a perforated groundwater  collection  pipe, which the state
regulatory agency requires to be installed. A  possible restriction
with this design is the sidewall stability of a narrow trench exca-
vated to a depth  of approximately 45 ft.  The possibility would
exist for the collapse of sidewall soils into the trench because of
high groundwater conditions  and layers of sandy and silty soils.
This  option would require all  pipe placement, bedding place-
ment and backfill to be accomplished by the contractor from the
ground surface.
  The second design option recommended a partial excavation
using scrapers  along the slurry wall followed by construction of a
soil/bentonite slurry wall and gravel filled trench similar to Op-
tion 1. The advantage of excavation along the containment struc-
                                                                              RCRA SITE REMEDIATION & EXPANSION    311

-------
                                                       OPTION  NO. 3
                                             EXCAVATION  WITH  PIPE PLACEMENT  AT  BASE
      POWER UNI
                                                           Figure 3
                                              Design Option—Containment Structure
ture to provide a workable base area for the slurry wall and pipe
trenching contractor would be to reduce the total depth of slurry
wall and trench excavation required. This design concept would
include the placement of a synthetic flexible membrane liner and/
or clay liner along the outward slope of the scraper excavation to
ensure that groundwater flow from the landfill would be inter-
cepted by the containment structure and routed to the collection
pipe.  The remainder of the excavated area would be filled with
earth  material or with approved solid waste to the original ground
surface.
  The third design option includes scraper excavation along the
containment structure extending to the design depth to install the
perforated  groundwater collection pipe. A cross-section view of
this option  is shown in Figure 3. The excavation would be an aver-
age of 45 ft deep over the 3700-ft  long containment structure.
Similar to option 2, a flexible synthetic liner or clay liner would
be constructed along the outward slope of the excavation to pro-
vide a low permeability barrier to seepage from the landfill area.
The landfill owner desires that the air space created by soil exca-
vation along the containment structure be approved  for filling
with solid  waste materials. In addition, solid wastes would  be
filled  above existing ground surface over this area, which would
result in a modified final grade for the landfill.
  The design  concept  of a refuse filled containment  structure
would result in revenue to the landfill  operator due  to the in-
creased waste  disposal  capacity  provided by the containment
structure and would modify the final side slopes on the south and
west sides of the site from an approximate 2:1 slope to a  final
slope  of 4:1. The landfill owner presently is under administra-
tive orders by  the DNR to modify the south and west slopes.
Therefore,  option 3  would provide a solution to issues associated
with the present 2:1  slope along the borders of the site and would
provide additional refuse fill capacity for the landfill owner. After
evaluating the advantages and disadvantages of each option, the
landfill owner has decided to prepare detailed plans and specifica-
tions for option 3 and to conduct the DNR relative to approval of
this design concept for the containment structure.
  The issue of minimum technical  compliance with hazardous
waste regulations for a  vertical and horizontal  landfill expansion
also will be addressed relative to  the proposed design of the con-
tainment structure. It is possible that the NDR may determine
that the proposed containment structure constitutes a lateral ex-
pansion to a hazardous waste landfill. The NDR also may deter-
mine  that previously  filled hazardous waste areas,  due to the
nature of closure and  lack of documented base areas or leachate
collection systems, do not constitute a separate and discrete haz-
ardous waste unit. If such a determination is made, it is possible
that the landfill owner may be required to comply with the min-
imum technology standards for a hazardous waste landfill for the
containment  structure area. This decision would require that a
double liner, equal to or equivalent with guidelines promulgated
in the HSWA, be compiled with for the design of the containment
structure. If this determination  is made, approval for the con-
tainment structure design would have to be obtained from the
hazardous waste section  of DNR as well as the U.S. EPA. In
addition, it is probably that the DNR will require the landfill own-
er to prepare the necessary feasibility reports, engineering plans
and specifications  for the landfill licensing process for the con-
tainment structure. In the State of Wisconsin, the landfill regula-
tory process can take from 3 to 5 years to complete. The land-
fill owner intends to prepare the necessary feasibility reports and
engineering plans to fulfill state licensing requirements for an ex-
pansion to an existing landfill site.

SITE EXPANSION
  Initial DNR reaction' to the  lateral solid waste site expansion
centered around the issue of separation of solid and hazardous
waste units. According to the State of Wisconsin regulatory agen-
cy interpretation, the  proposed lateral expansion (solid waste dis-
posal) must be physically and hydraulically separated from the
existing hazardous waste  unit.  Hydraulic separation  is required
due to the U.S.  EPA mixture rule which specifies that all leach-
ate from the existing site  would be considered hazardous. There-
fore,  a separate leachate collection  system  is recommended by
DNR for both the vertical and lateral expansion areas.
  There are six corridors remaining to be constructed and filled
in the existing landfill. A determination is needed as to whether or
not these corridors are considered to be part of the hazardous
waste unit under the HSWA. The remaining corridors have DNR
solid waste approval but have not been reviewed by the hazardous
312    RCRA SITE REMEDIATION & EXPANSION

-------
waste permit section. It was suggested by DNR that a clay wall of
"scraper width"  may be needed for physical separation of the
solid and hazardous waste units. However, redesign  of the re-
maining corridors within the existing permitted landfill area may
be required by DNR.
  A second possible alternative  was suggested by DNR—delist-
ing the site as  a hazardous waste unit. This procedure would in-
volve obtaining certification from the U.S. EPA that the haz-
ardous wastes  accepted at the site from 1975 through 1982 should
not have been characterized as  hazardous. Based  on the  waste
types disposed at the site, it is  unlikely that the recommended
certification could be made.
  Since the  DNR is not fully authorized under RCRA, the U.S.
EPA must determine what part(s) of the existing landfill  is the
hazardous waste  unit and what engineering  design options are
recommended to satisfy the separate and discrete unit issue.
  According to U.S. EPA guidance,1'6 a new unit or  lateral ex-
pansion of an existing unit operating under interim status may
continue to receive non-hazardous wastes without complying with
the minimum technological RCRA requirements. Apparently, the
minimum technological  RCRA  requirements  apply to  those
interim status  landfills that receive  hazardous wastes  after May
8, 1986.' In regard to placing waste over a partially or finally cap-
ped hazardous waste unit without complying with minimum tech-
nological RCRA requirements, the guidance6  suggests  that if the
new waste is effectively isolated from the waste within the cap-
ped unit, the new waste area will be considered as a new unit. In
addition, the U.S. EPA is concerned with closure  of  hazardous
waste units that no  longer accept hazardous wastes. Solid  waste
disposal over such a closed unit  would constitute a modification
of the closure  plan and would require U.S. EPA review and ap-
proval.
CONCLUSIONS
  A solid waste landfill adjacent to and over an existing hazard-
ous waste   landfill can  be engineered in an  environmentally
sound manner. A containment structure can be  designed to pro-
vide environmental protection along an existing  hazardous waste
unit and also be a source of revenue to the landfill owner.
  The U.S.  EPA and NDR must determine the applicability of
minimum technological RCRA requirements to solid waste land-
fill  expansions at interim  status  landfill facilities. Until  this de-
termination  is made,  it will be  difficult and perhaps costly to
develop acceptable design concepts.
  The containment structure design issue also is complicated by
the recent HSWA due to possible limitations on design concepts.
  Environmentally sound designs presently are available for land-
fill expansions and containment structures. However, compliance
with the recent HSWA may be expensive.

REFERENCES
1. U.S. EPA, "Hazardous Waste Management System; Final Codifica-
  tion Rule," Federal Register, 50, July 15, 1985, 28705-28711.
2. Wisconsin Department of Natural  Resources, "Determination  and
  Conditional Approval for a Plan of Operation Land Reclamation
  Sanitary Landfill," Dec. 1984.
3. Residuals Management Technology, "Feasibility/In-field Conditions
  Report Land Reclamation  Ltd.,"Mar. 1981.
4. FOTH & VAN DYKE Industrial, Inc., "Initial Site Report for the
  Land Reclamation, Ltd., Sanitary Landfill Expansion," 1985.
5. Meeting with Wisconsin  Department of Natural Resources Solid
  Waste Division, Dec. 4, 1985.
6. U.S. EPA, "Guidance on Implementation  of the Minimum Tech-
  nological Requirements of HSWA  of 1984, Respecting Liners  and
  Leachate Collection Systems," May 1985 (Draft).
                                                                               RCRA SITE REMEDIATION & EXPANSION    313

-------
                           Cost  Estimating for  RCRA/CERCLA
                                              Remedial Actions

                                                 Michael R. Morrison
                                              Gregory P.  Peterson,  P.E.
                                                     CH2M HILL
                                                   Corvallis, Oregon
  ABSTRACT
    Developing cost estimates for hazardous waste remedial actions
  is an integral part of the process of designing the remedial facility.
  This paper details the content of screening and feasibility cost
  estimates and their use in feasibility studies for remedial actions.
  Estimating terminology is defined, and key differences between
  "traditional" estimating and estimating hazardous waste projects
  are discussed.

  INTRODUCTION
    Estimates of the costs for remedial actions at hazardous waste
  sites are an essential, and often required, part of the feasibility
  study and  the process of determining the  most cost-effective
  method for site cleanup. During the feasibility process, techno-
  logical and cost criteria are weighed and compared several times
  during the  process of  refining  alternative  remedial actions
  (Fig.l).
    Initially,  comparative screening costs are generated to screen
  out disproportionately expensive alternatives. Subsequently, feas-
  ibility costs are developed and used in  the selection of the most
 cost-effective  alternative. The basic  procedures  for generating
 these two cost estimates are essentially identical, although greater
 accuracy is achieved for the feasibility costs through a more de-
 tailed evaluation of the alternatives and their related cost com-
 ponents.

 ESTIMATING DESIGN AND
 CONSTRUCTION COSTS
   The American  Association of Cost  Engineers (AACE) lists
 three levels of cost estimating accuracy to reflect the increasing
 definition of a remedy at three points in a project.  These three
 levels are: order of magnitude, budget and definitive. Feasibility
 studies are always order-of-magnitude estimates because  of the
 limited scope of work available at this stage in the project. AACE
 defines  order-of-magnitude estimates as approximate estimates
 made without detailed engineering data. Some examples  would
 be: an estimate  from cost capacity curves,  an estimate  using
 scaleup or down factors and an approximate ratio estimate.
 Order-of-magnitude estimates normally  are accurate within plus
 50% or minus 30%.
   Estimating ranges are not meant to be absolute limits, but in-
 stead imply that there is a high probability that the final project
 costs williall within a range. Contractors' bids often fall outside
 of these ranges, but cost estimates focus on predicting the selected
 low bid.
  When using cost estimating  terminology, it is important to dis-
criminate between the terms  "allowance" and "contingency."
The AACE defines allowances as incremental resources included

314   RCRA SITE REMEDIATION & EXPANSION
                          Figure 1
                   Project Interrelationships
  This Figure Diagrams the Inseparability of Scope, Schedule, and Cost.
 in estimates to cover known but as yet undefined requirements
 for individual accounts or subaccounts. An example of an allow-
 ance would be the inclusion of pipe supports and freeze protec-
 tion for a 3-in. above-ground pipe, estimated at $15/ft. The pipe
 supports and freeze protection are neither designed nor quanti-
 fied in the feasibility study; however, since good designers would
 include them in the completed design, an allowance of 70% could
 be included in the estimate for these items.
  Contingencies are defined as specific provisions for  unfore-
 seeable elements of cost within the defined project scope; partic-
 ularly important where previous experience has shown  that un-
 foreseeable events  which will increase costs are likely to occur. If
 escalation is included in the contingency, it should be a separate
 item, determined to fit  expected escalation conditions of the pro-
ject.

 HAZARDOUS WASTE ESTIMATING VS
TRADITIONAL ESTIMATING
  A defined scope  of work is the first requirement for developing
a cost estimate in the traditional manner. However, at the feasi-
bility stage, hazardous waste remedial actions do not have de-
fined scopes and the site often is incompletely characterized.

-------
  The feasibility analysis consists of comparing a number of op-
tions. These options may reflect different assumptions about the
characteristics of the site and a range of possible designs and pro-
ject scopes. For example, suppose a hazardous liquid is leaching
out of a municipal landfill. Remedial action options may include
excavating a range of soil volumes for disposal  at one of several
licensed disposal sites; a range of groundwater pumping and treat-
ment rates;  and a  range of on-site containment  options. The
geology, geohydrology and many other factors may be only par-
tially explored.
  There are two basic steps that can be taken to narrow this large
range of unknowns: (1) collect more data or (2) rely on the use of
good engineering judgment to make the necessary assumptions to
help quantify each alternative. In hazardous waste remedial work,
the collection of additional data can reduce the  range  of un-
knowns but  never will replace the use of good engineering judg-
ment in identifying a reasonable estimating scope. Since many of
the elements  of a remedial action involve subsurface conditions
and/or new  waste technology, a conservative viewpoint is essen-
tial for cost estimating.
  Furthermore, projects involving hazardous waste have certain
unique features that affect cost estimates. These may include:
• Reduced productivity due to personal protective gear
• Purchase and disposal cost of protective equipment
• Decontamination facilities
• Worker physical examinations and training
• Seasonal weather effects on construction costs
• Freeze protection for on-site facilities
• Nonconventional, multi-unit process treatment trains
• Extraordinary ancillary items
• High redundancy or conservative loads
• Disposal of treatment sludges and residues
  It is very important that all parties understand the inherent un-
certainty in feasibility cost  estimates. The estimated costs should
be used only in the light of the accuracy, context and scope under
which they were prepared. It is therefore necessary that all the
assumptions  about the scope be documented. This documenta-
tion, or at least the summary  of it,  should be  included in any
report, along with cautions about the limitations associated with
using any of the costs out of context.


FEASIBILITY STUDY COSTING
  Feasibility studies make  use  of screening costs and the some-
what more accurate feasibility costs. Screening  costs are used to
compare alternatives in order to eliminate those with higher costs
but no significantly greater environmental or public health ben-
efits. Prior estimates,  site cost experience and good engineering
judgment are needed to identify those unique  items in each al-
ternative which will control these screening estimates. Items com-
mon to all alternatives do not warrant substantial effort at the
screening stage.
  For alternatives  that survive screening, feasibility costs are
developed. The feasibility cost estimates are more detailed than
the screening cost estimates and include both capital and annual
operating costs. The annual operating  costs are viewed over a
predicted life and converted to life-cycle costs. A sensitivity analy-
sis  also can be done on the key items in the  alternatives.  The
latter steps are discussed later in this paper.

Capital and Annual Operating Costs
  Capital costs include only those expenditures that are initially
incurred to develop and install  a remedial action and major cap-
ital expenditures anticipated in future years. Both direct and in-
direct costs are included.
Direct Capital Costs
  The four types of direct capital costs discussed briefly below
should be line items in the estimate. These four line items may
include a number of other items, some of which are listed.
  Construction costs cover equipment, labor and materials re-
quired to install a remedial action. They also include site modifi-
cations, process  equipment, testing and monitoring equipment,
health and safety monitoring and site closure costs. Construction
costs  available from nonhazardous waste sites should be used
with care.  Normally, considerable modification of any of these
costs  is necessary. Labor costs will  include all fringe benefits,
workman's compensation and contractor fees.
  Land and site development expenses  include easements, devel-
opment of access roads,  access control (e.g., gates, fences) and
site preparation for equipment and buildings.
  Buildings and services  costs cover construction of temporary
and permanent buildings and establishing utilities.
  Construction of a hazardous waste remedial facility also may
require relocating affected populations. This may involve temp-
orary or permanent accommodations for nearby residents, mov-
ing allocations, temporary wage settlements and other associated
costs, as necessary.  Guidance for relocation costs is available
from  the  Federal  Emergency  Management  Administration
(FEMA).
Indirect Capital Costs
  Indirect  capital costs consist of engineering, financial,  super-
vision and other services necessary to carry out a remedial action.
They are not incurred as  part of actual remedial actions but are
ancillary to direct or construction costs. Indirect capital costs in-
clude allowances for design and engineering and contingency.
  All other costs, for example, monitoring and sampling, data
collection,  pilot testing, startup and  shakedown,  "debugging,"
operator training and initial field monitoring, should be listed
separately  in the estimate. For remedial  actions  involving an
extended (more than 6 months) construction schedule, an adjust-
ment can be made for interest and inflation during construction.
The assumed schedule and cost index date should be stated.
Allowances and Contingencies for Capital Costs
  Two categories of contingencies can  be added to  total  capital
costs for unforeseen circumstances which may result in additional
costs. Bid  contingencies  cover unknown costs associated with
constructing a given project scope such as adverse weather con-
ditions, strikes by  material suppliers,  geotechnical unknowns
and unfavorable market conditions for a particular project scope.
Bid contingencies may vary between 10 and 20% and are applied
against the construction subtotal for each alternative.
  Scope contingencies cover scope changes that invariably occur
during final design and  implementation. They are intended to
adjust the estimate so it can be used for budgetary purposes. Since
the site characterization is often incomplete and volume estimates
rudimentary, the scope contingency enables the use of engineer-
ing judgment to establish a reserve for project scope modifica-
tions. Scope contingencies will vary  between alternatives. They
may be 10  to 25 %  of the total capital costs for known technolo-
gies and sites with minimum elements of risk; they can be much
more for state-of-the-art technologies at  sites with higher  ele-
ments of risk. Even on well-defined conventional projects, order-
of-magnitude cost estimates tend to be 10 to 30"% below the actual
bid price. Below are some of the factors that determine the mag-
nitude of the scope contingency:
• A reserve of 6 to 8% of the total construction cost is desirable
  for change orders.
• Scope  changes during design  and/or  construction may be
  needed to meet performance-based criteria.
                                                                               RCRA SITE REMEDIATION & EXPANSION
                                                                                                                           315

-------
• Bidding competitiveness may be low; this, in turn, increases the
  bids.
• Vendors may provide over-optimistic performance data.
• Quantities may be inaccurately defined and listed.
• Performance histories for developing performance  specifica-
  tions may be lacking.
• Regulatory or policy changes may impact basic study assump-
  tions.
• Different  firms and agencies, with different operating  pro-
  cedures, can be involved from concept to cleanup.
Presentation of Total Capital Cost
  Superfund  site  cleanup  activities are classified  by  the U.S.
EPA as either planning costs (remedial investigation,  feasibility
study and design) or implementation costs (construction, con-
struction management and startup). Since funding for site clean-
up can come from more than one source, it is important  to format
the presentation of the estimate to  facilitate any cost sharing that
may  occur. For example,  the  cost  estimate required for the
Record of Decision and the fraction of the  cost that the state
shares, is a part of the implementation cost.
  Capital cost estimates for feasibility studies should  be struc-
tured as shown in Table 1, below. Note the compounding effect
of the calculation. If all calculated  costs shown were a percentage
of Item A only, a lower total would be calculated.
                           Table 1
        Formal and Method for Calculating Total Capital Costs
          Name

          Construction Subtotal
          Bid Contingencies
          Scope Contingencies
          CONSTRUCTION TOTAL

          Permitting and Legal
          Services During Construction
          Other Costs
          TOTAL IMPLEMENTATION COST

          Engineering Design Costs
          TOTAL CAPITAL COST
Method of Calc.

Prom estimate details
* of Item A
% of Item A
Sun of Items A 8 6 C
l of  Item D
% of  Item D
Varies with estimate
Sum of Items D E F & G

Usually % of D
Sum if Items H 6 I
Annual Costs
   Annual operating costs for a remedial action include the oper-
ating and maintenance (O&M) costs incurred following its con-
struction and/or installation, and, where  applicable, annualized
capital costs. Screening level annual costs  may be obtained from
such references as the Compendium of Cost of Remedial  Tech-
nologies at Hazardous Waste Sites' or from prior estimates. For
the detailed  feasibility analysis,  O&M cost estimates must be
based on  site-specific  information. Equipment vendors, similar
estimates  or  standard  costing guidance references may be con-
sulted. All estimates a year or more old should be updated with
current labor rates and price quotes.
O&M Costs
   The post-construction/installation activities necessary  to pro-
vide continued effectiveness of a remedial  action may involve the
following cost components:
•  Operating labor—all wages, salaries, overhead (which includes
   payroll  taxes, salaries,  etc.) and fringe benefits. Nonproduc-
   tive labor for vacations, sick leave, holidays and H&S training
   sessions can consume 16 to 18% of the standard 2,080-hr work
   year.
• Maintenance materials and labor—the costs for labor, parts
  and other materials required to perform routine maintenance
  of facilities and equipment.
• Auxiliary materials and energy—such items as chemicals and
  electricity needed for plant operations, water and sewer service
  and fuel costs.
• Purchased services—such items as  sampling costs, laboratory
  fees and other professional services for which the need can be
  predicted.
• Disposal—transportation and disposal of any waste materials,
  such as treatment plant residues, generated during the course
  of a remedial action.
• Administrative costs—all costs associated with administration
  of operation and maintenance not  included under  other cate-
  gories such as labor overhead. This item can be 10 to  \5% of
  the operating labor cost.
• Insurance, taxes and license—such  items as: liability and sud-
  den and accidental insurance; real estate taxes  on purchased
  land or right-of-way; licensing  fees for certain technologies;
  and permit renewal and reporting costs.
• Maintenance reserve and contingency costs—annual payments
  into escrow funds to cover anticipated replacement  or rebuild-
  ing of equipment and any  large unanticipated O&M costs, re-
  spectively.

Life-Cycle Costs
  Present-worth analysis is a method  of evaluating and compar-
ing costs that occur during different time periods by discounting
all future expenditures (i.e., replacement of worn parts) to a com-
mon base  year. The result is the total  cost, in today's dollars, of
the remedial action alternative.  The U.S. EPA has  indicated that
the current Office of Management and  Budget (OMB) guidance
should be followed for calculating life cycle costs. Many engineers
are familiar with life-cycle cost calculations, or a firm may employ
economists or statisticians to help with these calculations.

Sensitivity Analysis
  Errors in estimating key variables may have a large effect on
the accuracy of the  overall cost estimate. A sensitivity analysis
can assess the effect on  the cost of varying specific assumptions
associated with economic conditions  or with the design, imple-
mentation, operation and effective life  of the remedial action
strategy.
  Through good engineering judgment, one should  be  able to
identify those items  that could bring  about a significant change
in overall costs  with only a small change in the value of the item.
Results of the analysis can be used to identify "worst case" scen-
arios and to revise estimates of contingency or reserve funds.
  In addition, sensitivity analysis can be used to optimize the final
design of a remedial action alternative. This analysis is particular-
ly useful where design parameters are interdependent,  such as for
treatment  plant capacity versus time for treating  contaminated
groundwater.
  The  following factors are primary candidates for considera-
tion in conducting sensitivity analyses:
• The effective life  of the remedial action. In estimating replace-
  ment costs, use base period dollars; do not adjust for inflation.
• How  clean is clean?  The  public health assessment normally
  provides a range  of cleanup criteria dependent on variation
  between the different  health standards and cancer risks. There
  often can be  a 10- or 100-fold difference between alternate cri-
  teria.
• O&M costs
• Duration of cleanup activities
• Availability of off-site treatment, storage and disposal facilities
316
       RCRA SITE,REMEDIATION & EXPANSION

-------
• Uncertainty regarding site conditions. This item includes the
  volume to be treated, materials present and treatment used.

CONCLUSIONS
  If there is one message to be gained from this paper, it is that in
the development of remedial actions for hazardous waste contam-
ination, perhaps more than  for  any other type of engineering,
cost and scope  (and,  of  course, schedule)  are  interdependent
and inseparable. This paper outlines some procedures for devel-
oping feasibility costs, including discussions of line items peculiar
to hazardous waste work.
  Estimating costs for hazardous waste facilities has many in-
trinsic risks. The authors' experience in defusing some of these
risks has been included where possible. However, in the space per-
mitted here, there is no way to cover all of the problems the cost
engineer may encounter in estimating hazardous waste remedial
action designs.

REFERENCES
1.  American  Association of Cost Engineers, "Cost Engineer's  Note-
   book," American Association of Cost Engineers, Morgantown, WV.
2.  Environmental Law Institute,  "Compendium of Cost of Remedial
   Technologies at Hazardous Waste Sites," Environmental Law Insti-
   tute, Washington, DC.
3.  Morrison,  M.R. and Anderson, G.P.,  "CH2M  HILL REM/FIT
   Cost Estimating Guide," CH2M HILL, 1985.
                                                                               RCRA SITE REMEDIATION & EXPANSION    317

-------
                     A  Modular Computerized  Cost Model  for
                     Remedial  Technologies  at  Superfund Sites
                                                 William Kemner
                                              John  Abraham, P.E.
                                                   Jack  Greber
                                                  Jay  Palmisano
                                              PEI Associates,  Inc.
                                                 Cincinnati, Ohio
ABSTRACT
  The  purpose of the research reported in this paper was to
develop prototype cost estimation models for response technol-
ogies at uncontrolled hazardous  waste sites. This research was
conducted for the U.S. EPA, Hazardous Waste Engineering Re-
search  Laboratory in cooperation with the Office of Emergency
and Remedial Response.
  The  model is designed for operation on a personal computer
and can provide levels of detail in the estimate as a function of
input parameters available. The cost estimation model takes into
account site, waste, workers safety and regional variations. Out-
put formats and types of cost calculations are consistent with the
U.S. EPA Remedial  Action Costing Procedures Manual. De-
fault values for interest  rates, discount rates, planned life, con-
tingency allowances, etc., are  consistent with  the costing  pro-
cedures manual.
  The  model can be used by U.S. EPA Headquarters and Reg-
ional Staff,  State personnel and  contractors (REM/F1T Con-
tractors, cleanup contractors, etc.) for quick, easy and effective
calculations of costs for evaluating and comparing a variety of
different remedial measures.

INTRODUCTION
  Computerized cost  estimating models are finding  widespread
use in the building and construction fields because of the advan-
tages of uniform procedures and access to  cost data bases they
provide. The U.S. Army Corps of Engineers has perhaps the best
established program, i.e., the Computer Assisted Cost Estimating
system (CAGES). The Corps requires that this program be used
by architects and engineers to generate cost estimates on Corps
projects. Other government entities (U.S. DOE, VA, Air  Force,
Navy, etc.) and private firms have computerized systems and data
bases for various applications.
  Remedial action technologies at Superfund sites employ many
conventional civil engineering elements with well-documented
costs as well as completely new technologies  for which little or no
cost data exist.  Both construction and chemical process com-
ponents are involved in cleanup technologies. This is a unique re-
quirement not encountered in other computerized models.
  With the huge backlog of Superfund site work and the com-
plexity of cleanup technology, it was apparent that some form
of computer-assisted estimating was needed  by the U.S. EPA for
use by both contractors and Agency personnel.
  The  Land Pollution Control  Division of the Hazardous Waste
Engineering Research Laboratory contracted with  PEI to develop
such a  system. The project is  being done in phases. This paper
presents the  results of Phase I; the development of the overall
design  and objectives of the computerized  system and develop-
ment of several cost modules. Work is currently underway to ex-
pand the cost data base and algorithms. The system is designed
to be flexible and dynamic; it is anticipated that new data con-
tinually will be used to update the cost data base.
  The system is designed to operate as an interaction program
on an IBM PC/XT or AT. The intended users include:
  U.S. EPA research personnel
  U.S. EPA policy analysts
  Superfund Regional Office personnel
  Superfund contractors
  U.S. EPA enforcement personnel
  The system is intended to be used to:
  Develop scoping costs for budget planning
  Calculate specific remedial site costs
  Check estimates performed by others
  Compare alternative technologies
  Evaluate costs proposed by potentially responsible parties
  Perform value engineering analysis
  Costs can be calculated as specific estimates or include prob-
ability ranges based on uncertainly input by the user. Net present
worth analysis also is included in the program.
  The advantages of the system include:
• Detailed, documented cost  elements consistent with the  Na-
  tional Contingency Plan (NCP) and EPA Cost Guidelines
• Uniform and consistent format for costs
• Rapid calculation  of cost and accommodation of changes in
  specifications
• Ability to consider  uncertainty in estimates
• Ability to compare alternatives for a given site
PROGRAM STRUCTURE AND OPERATION
  The program is structured to a level of detail based on the data
available prior to detailed design specifications and drawings by a
contractor.  Generally, a line item approach is used for output
specifying line  item detail for those users interested in the de-
tail level.
  The program is designed to be flexible. Some of the flexibility
features include:
• Ability to include specific values for input parameters or use de-
  fault values in the data base
• Ability to add write-ins for unusual items
• Ability to use adjustment  factors for such things as inflation,
  geographic location, job difficulty, labor productivity due  to
  health and safety requirements, geological effects, etc.
318    RCRA SITE REMEDIATION & F:\PANSION

-------
• Ability to have various levels of printouts with more or less de-
  tail
• Ability to override individual computed items of cost
• Ability to store partial or complete runs for recall and modifi-
  cation at a later date

  The system is designed to separate the data base from the cal-
culation requirements as much as possible. The data base includes
room for expansion and the ability to add or update data items
without affecting the program. This enables expansion and up-
dating by personnel who are only casually  familiar  with com-
puters.
  The costing system requires the user to  understand the tech-
nologies being used because relatively detailed design input data
are required. Unlike some other approaches, this system does not
rely on historical costs of previous remedial actions. Rather, the
remedial technologies are broken down to levels of detail which
can be costed based on sound engineering cost data.
  The program is designed to be transportable and operate on an
IBM PC/XT or AT or equivalent with at least 5 Mb of available
hard disk storage capacity and 256K RAM. The operating sys-
tem is MS DOS 2.0 or higher.  The program is written in Micro-
soft FORTRAN. LOTUS 1-2-3 spreadsheet software is an option
to enable users to change values on intermediate output. The data
base is structured using dBASE III.
                                                                       Current Data File  in  use
                                                                                                     NO FILE  IN  USE
                          Figure 1
   Overall System Flow Diagram for the Computerized Cost Model
  Overall system structure and flow are shown in Figure 1. The
program is designed around a series of menus used to select spe-
cific technologies and components. The general program control
menu is shown in Figure 2. The user is guided through the menus
in a logical fashion; experienced users can access  a specific com-
ponent directly without stepping through the menu hierarchy by
entering  the component I.D. Once at a component level, the
user enters the design parameters which determine the cost of
each subassembly.
                    Main Menu
      1)  Begin a  New Data  File
      2>  Edit Current Data File
      3)  Save Current Data File to  Disk -for Future Use
      4)  Load in  Previously Saved Data File -from Disk
      5)  Calculate  Costs -from Currant  Data File and
           Print  Report
      6)  Exit Program

Enter  your selection (1-6)

                          Figure 2
                 General Program Control Menu
  Two general input screens are used to enter information such
as date of run, title and various cost adjustment factors. Figures
3a and 3b are example general input screens.
  The unit cost data base contains the unit costs or cost equa-
tion coefficients for each line item. The data base consists of two
files:  one for linear cost relationships (i.e., line item extensions)
and one  for exponential cost relationships (i.e., cost estimating
relationships for process equipment).  Many conventional con-
struction line item costs are excerpted  from references  1 and 2.
The default factors file contains default values for items such as
interest rate, contractor overhead, etc. If values are not specified
for these parameters by the user, the default factors are used.
  A job input file contains the input data for a past run and en-
ables  the user to recall  and adjust past runs without reentering
all the data.
                                                                                      I REMEDIAL  RESPONSE  CONSTRUCTION I
                                                                                      I   COST  ESTIMflTING  SYSTEM     I
                                                                     DOTE	) 1£-17-B3<—1

                                                                     ESTIMATOR	>JDHN DOE
                               ANALYSIS NO.	> 13-333  <—I

                                    (— I
                                                                     SITE NAME-
                                                                                    -)SENERIC SITE ONE
                                                                     SITE LOCBTION	) ANYWHERE, USA              <—I
                                                                                   —)                         <—I
                                                                     ZIPCODE	) 12343<—I

                                                                     CONTRACT NO.	>12-033GKDFA-43NMK4  <—I

                                                                     CONTRACTOR	> ACME COMPHNY               (—I

                                                                     COMMENTS/SCOPE OF WORK I
                                                                     >EXAMPLE OF SUBSURFACE DRAIN MODULE CALCULATION FOR A
                                                                     >HYPOTHETICAL SITE.
                                                                      IS THE INFORMATION CORRECT ?  (Y/N)
                                                                                             Figure 3a
                                                                                        General Input Screen
                                                                                          GENERAL PROJECT  COST DflTA
   GENERAL CONDITIONS   . 0*

  GEOGRAPHIC LOCATION FACTOR I

   LABOR t EQUIPMENT  18.0 »
   MATERIAL

  ESCALATION t

    LABOR

    EQUIPMENT

    MATERIAL

  SALES TAX
                                                                                      3. 0 <
           GENERAL CONTRACTOR I

            OVERHEAD  10.0 *

             PROFIT   10.0 *

             E,D t I  13.0 X

    PERMIT/LICENSE FEE   l.o  *

             BOND      1.0 X

            INSURANCE   0.0 X

ADDITIONAL LAND COST •     33000
         ARE THESE CORRECT ? (Y/N)
                          Figure 3 b
                General Project Cost Data Screen
                                                                                RCRA SITE REMEDIATION & EXPANSION    319

-------
                  Currant  Data File in use i  FILE3.BIN
                                                                           Current  Data Film  in  use i FILE3.BIN

Technology
Level

1.0.0.0.0
2.0.0.0.0
3.0.0.0.0
4.0.0.O.O
S. 0.0. 0.0
6.0.O.O.O
7.O.O.O.O
8.0.0.0.0
9.0.O.O.O
Major Program Bubdl visions
Technology Categories

Control Technologies
Treatment Technologies
Excavation
Transportation (OH Bite).
Disposal
Alternative Water Supplies
Relocation
Oeneral Site Work
Building/Structure Decontamination
O
Modules
Chosen

O
0
0
0
0
0
0
O
0
1. 3. 4.0.0
Technology
Leve 1


1.3. 4. 1.0
1.3.4.2.0
1.3.4.3.0




Leachate Control
Technology Categories


Subsurface Drains (T)
Drainage Ditches   99 to return to the Main Menu
Input a  Technology Level  1

                           Figure 4
                   Major Program Subdivision


    The technology-specific data entry is based on the "menu"
  approach because this is a  proven approach which works well
  from a programming and organization standpoint and has found
  wide acceptance by noncomputer-oriented personnel.
    The hierarchy of menus is organized  by a numbering system
  and proceeds from broad categories of technologies to greater
  levels of detail in  specification of individual components. Once
  the finest level of detail menu is reached, the user is prompted
  for the specific design parameters required to estimate the cost of
  that component. Following the design data, general cost informa-
  tion is required from the user such as contingency, subcontractor
  overhead and profit and adjustment factors. Defaults  are pro-
  vided wherever feasible.
    The menu hierarchy is summarized below:

  X.                         Major Program subdivisions
  X. X.                       Technology categories
  X.X.X.                     Specific technologies
  X.X.X.X.                  Technology subcategories
  Data input within X.X.X.X.  Subcategory  modules and  sub-
                               assemblies
    Figure 4 describes the major program subdivisions and is the
  highest level of the menu hierarchy. Figures 5 and 6 are examples
  of menus progressing through finer levels of detail for technology
  selection.
                                                                Enter  —)  0.  to  return to the  Main Technology  Menu.
                                                                Enter  —>  99  to  return to the  Main Menu

                                                                Input  a Technology  Level . 1

                                                                                          Figure 6
                                                                                Detail Technology Category Menu
                                                                     1.3.4. l.O
                                                                         i
                                                                                  Subsurface Drains  (T)
                                                                                  MODULE  NUMBER 1

                                                                                Drain Pipe  size I material
                                                                                          Material
Size
4 in
6 in
8 in
1O in
12 in
18 in
1 Asbestos
! Cement
I 1
! 2
! 3
i 4
1 5
1
: Por ou s
! Concrete
: 6
! 7
1 8
i -
i 9
1 10
Vitrified
Clay
11
12
13
14
IS
—
pvc
IB
17
IB
19
20
"
                                                                              Enter  Drain Pipe  choice (1-20)7  14

                                                                                 Is  this screen  O.K.  ?   Y

                                                                                          Figure 7a
                                                                                      Input Data Screen
                                                                     1.3.4. 1.0
                                                                         i
                                                                                  Subsurface Drains (T)
                                                                                  MODULE NUMBER  1
                                                                             Non-perforated  Pipe size  4  material
                                                                                            Material
                   Current Data File  in u.e t FILE3.BIN
1.0.0. 0. 0    Control Technologies


                  Technology Categories
Technology
Level
                                                   Modules
                                                    Chosen
 1.1.0.0. 0    Air Emissions Controls                       9
 1.2.0.6.0    Surface Water Controls                       0
 1.3.0. e. 0    Ground Water Controls                        B
 1.4.8.8.8    Contaminated Water and Sewer Lines           0
 Enter —> 0.  to return to  the Main Technology Menu.
 Enter —> 99  to return to  the Main Menu

 Input a Technology Level  .3
                           Figure 5
                   Technology Category Menu
Size
4 in
6 in
8 in
10 in
12 in
18 in
Asbestos
Cement
	
21
22
23
24
—
Porous
Concrete
	
23
26
27
28
29
Vitrified
Clay
30
31
32
33
34
35
PVC
36
37
38
39
40
—
                                                                        Enter  Non-perforated Pipe choice (21-40)? 33
                                                                                 Is this screen O.K.  ? V

                                                                                          Figure 7b
                                                                                      Input Data Screen

                                                                   Items marked with "(T)" (for technology component) are at
                                                                the lowest level of the menu hierarchy. Selection of one of these
                                                                leads to a data input screen  for user input of design parameters
                                                                for that technology component. Menu selection can be bypassed
  320    RCRA SITE REMEDIATION & EXPANSION

-------
 at any level by the experienced user selecting the specific menu
 I.D. number; e.g., 1.3.4.1 will take the user directly to the input
 screen for that particular subcategory component. Numerical de-
 sign data are not required until the user reaches the technology
 level.  At this point, the specific input varies depending on the
 component as illustrated in Figures 7a through 7f.
                                                                                        REMEDIAL RESPONSE CONSTRUCTION
                                                                                           COST ESTIMATING SYSTEM
             Subsurface Drains CT)
             MODULE NUMBER 1
                       Pipe A Trench depth
3 feet
5 feet
3)  6 feet
4)  8 feet
                        5)  10 feet
                        6)  IS feet
                                     7)  1 <. feet
                                     S)  16 feet
 9) ia feet
te> 20 feet
                Enter Pipe t Tronch choice (1-10)7  7
      Total length of drainage pipe trench,  (feet)   214.00

      Tot a 1 1 engt h of Non-per for at ed pip* t rench i ng. < feet)   900. 00

      Manholes or Junction boxes  t  Number required?   3
                              i  Interior coating required ? Y


                           Figure 7c
                      Input Data Screen
 1.3.4.1.0
    1
             Subsurface Drains (T)
             MODULE NUMBER 1
                         Met Well Pump(s)

      Pump Capacity,  Gallons/Minute        I  23 I  50 I 100 I £00

      Number of Pumps required            I   01   0 >   31   0

      Stainless/Lined Pump necessary  (Y/N)  I  N  I  N  I  Y  I  N



                   Leachate Holding Tank  Capacity

         Tank Capacity, Gallons   I  1000 I 2800 I 10000 I  20000

         Number of Tanks required I   0  I   0  I    01    2

                     Is this screen O.K.  (Y/N) ? Y
1.3.4.1.0
    1
                          Figure 7d
                      Input Data Screen
              Subsurface  Drains  (T)
              MODULE NUMBER  1

                    100 Gallon/Minute  Capacity  Pumps

              Distance from  pump  to  leachate holding  tank

             Pump * i   see. 00
             Pump » Z   700.00
             Pump * 3   400.00
             Pump It 4  l£0e. e0
             Pump * 5  1400.00

                      Is  this screen  O.K.  (Y/N>?  Y

                          Figure 7e
                      Input Data Screen
  Table  1 is an example of the input (or cost factor) compon-
ents for  several example modules. Each component and  sub-
assembly shown in Table 1 and included in the cost file is ref-
erenced to a component "fact sheet" which documents the bat-
tery limits of that  component, the design, cost data sources and
all items  included  in the cost. An example fact sheet for 1.1.1-
Pipe Vents is shown in Figure 8. The numerical data values used
in this paper are for illustrative purposes only and are not neces-
sarily final or accurate.
  There are two levels of output: a detailed output by technology
component and subassembly, and a summary output for the en-
                                                                                         Drains  
                                                                                 MODULE NUMBER 1
                                                                      SUBCONTRACTOR 1
                                                                       OVERHEAD  14.00X
                                                                       PROFIT     7.SOX

                                                                      GENERAL CONTRACTOR MARKUP
                                                                      ON SUBCONTRACTED WORKI
                                                                       OVERHEAD   4. 08*
                                                                       PROFIT     s. sat
                                                                                   COST DATA

                                                                                      HEALTH S SAFETY FACTOR  33.00*
                                                                                      GEOLOGY FACTOR         22.38*
                                                                                      DECONTAMINATION FACTOR   3. 00*
                                                                                      CONTINGENCY           18.25*
                                                                                      START-UP                . 90X
                                                                                      NET PRESENT WORTHI
                                                                                       INSTALLATION PERIOD  1  YRS
                                                                                       OPERATING LIFETIME(X)  20  YRS
                                                                                       DISCOUNT RATE      10. 00X
                                                                          ARE THE VALUES CORRECT? (Y/N) Y

                                                                                Figure 7f
                                                                            Input Data Screen

                                                     tire site scenario. The detailed output format for capital costs is
                                                     illustrated by an example shown in Figure 9. This is an example
                                                     for the subsurface drains module  only and does not represent a
                                                     complete remedial scenario. Using the spreadsheet software, the
                                                     user will be able to change any output value. A Lotus  1-2-3 tem-
                                                     plate is provided to do all necessary extensions. All output is de-
                                                     signed to fit on 8.5-in. x 11-in. paper.
                                                       The job input file contains  all user input data. The cost data
                                                     base and default factors files  are created and maintained using
                                                     dBASE HI but are used by the program as ASCII files. Figure 10
                                                     contains an example record from the unit cost data base file.

                                                     Calculation of Capital Costs
                                                       Capital costs are calculated  as the sum of labor, material and
                                                     equipment costs. These are defined as follows:
                                                     Labor      Labor to install  components  at  the  site. Off-site
                                                                 labor involved in fabricating delivery material is in-
                                                                 cluded in the material cost.
                                                     Material    Material such as piping, concrete, etc., used to build
                                                                 a technology, and preconstructed items such as in-
                                                                 cinerators, clarifiers, etc. (Note the latter are usually
                                                                 referred to as "equipment" in typical process cost
                                                                 engineering).
                                                     Equipment  Construction equipment cost, typically   rental  of
                                                                 bulldozers, cranes, etc.
                                                       The general cost equation
                                                     y = (axb +  c)d
                                                                                                                            (1)
                                                                  where a, b, c are coefficients and d is the number of items,  is
                                                                  used to calculate the component costs. This form of the equation
                                                                  can be used to represent many cost characteristics. For example,
                                                                  a simple line item cost such as 100 ft of 4-in. PVC pipe at $4.507
                                                                  ft can be represented  as y  =  [4.50(1)' + 0](100). A more com-
                                                                  plex CER such as for a liquid injection incinerator with 50 x 10'
                                                                  Btu/hr capacity may be represented as y  =  [1.6(50 x 10')055](1)
                                                                  where the coefficients 1.6 and 0.55 have been determined from a
                                                                  least squares fit of cost data.
                                                                    For most components, additional calculations are required in
                                                                  the program to convert user-specified parameters into component
                                                                  cost  drivers. For example, in excavation components, the cost
                                                                  driver is cubic yards. While some users could input this directly if
                                                                  known, it is easier to enter dimensions so the program can calcu-
                                                                  late volume.

                                                                  Calculation of Operating and Maintenance Costs
                                                                    Operating and maintenance costs include those costs required
                                                                  to operate and maintain the system after the initial installation is
                                                                  complete. The user specifies  the  time period  over which O&M
                                                                                  RCRA SITE REMEDIATION & EXPANSION    321

-------
                                                             Table 1
                                Cost Input Data Requirements by Subcalegory Component and Subassembly

1.0.
1.1.1














i.i. r




















Component
Pipe vents














Trench vents




















Subassembly
Backfill

Piping and laterals




PVC
»BS
Galvanlied Iron
Other
Eihaust system
butterfly valves)


Eicavatlon, backfill, spreading,
and struts
Piping and laterals
(includes elbows and tees)



PVC
ASS
Galvanized Iron
Corrugated polyethylene
(uses PVC risers)
Evhaust system

L 1 ner
* Hyp* Ion
0 4 In. gunfte and mesh
* Other
" 6 1n. clay
* Other

I
Diameter
(feel)
Diameter








Mow rale
1 ,r ff,}
\ 1C I"! 1


Depth
( f PC t )

Dlamrlrr
( tnchrs 1








Flo. rate
(icfm)





data Input

Da
?
Depth (feet)

Depth (feet)








Pressure
{fl in. used
If not
ipec if ted)
terujth
(fret 1

len9th
( f>M i








drop
(8 In. used
i* not
specified)
ub-Jisembly da1




U require^
3


Cap type
• U-top
(not app) ic .
If lateral*
are used)








Width
( feet )

Cap i/(*
* Mushroom^
* u-top
(not appl Ic ,
If lateral*








a input)





4


L*t. length














lat. length
"eet]

















i
Number

humtter








Hunter





hyrtwr


















HF
1

1








I



I

,









'








u
t














'




















u



































        Nolc- HF. GF. DF denote whether the adjustment factors arc applied to this cost clement; HF is health and safcl> (aclor. til IN Rcolog> factor, and DF is decontamination factor
costs are to be incurred. Uniform annual costs are assumed; i.e.,
no inflation and no increase due to deterioration of equipment.
The expected useful life of equipment is included in the data base
and printed out with the capital cost calculation. Salvage value is
not considered in the net present worth calculation.
   O&M costs  include labor, electricity, service water, fuel, chem-
icals, maintenance materials, water  treatment, area maintenance
and similar items. Unit prices for these items are maintained in the
default rates file which the user can  update  for each run if de-
sired:
   Net present  worth is calculated from the formula:
          NPW
X /(I
(2)
where NPW  =  net present worth in dollars
X,  = expenditures in year t
i    = discount rate (10% or entered by user)
t    = year

Costs are not inflated beyond the base year.
  For capital costs, the cost data base file is  used to provide in-
stallation time. The longest duration item is used as total installa-
tion time. Alternatively, the user can input an  installation time, in
which case the total cost is equally divided by  year. For O&M
costs, the user provides the number of years the site is to be main-
tained.
ADJUSTMENT FACTORS

General Conditions Factor
  This factor is used to adjust costs to account for qualitative
factors related to job difficulty. This factor allows a user to ac-
count for such  things  as  unusual or difficult site conditions,
weather, local labor conditions, local economy, contractor com-
petitiveness and other factors that cannot otherwise be accounted
for quantitatively. The  factor ranges from zero (no  adjustment)
to any percentage value the user  thinks  is reasonable. For ex-
ample, a value of 10% will increase the cost by 1.1. The factor is
applied to labor, material and equipment.

Geographic Location Factor
  The geographic location of the Superfund site where the tech-
nology is to be implemented is another important consideration.
Construction costs, material and  labor vary  by regions of the
country. The variation can be significant,  as much as 30% above
or below an average national cost for the same category of items.
Values for these geographic factors are available in  various cost
estimating publications such as R.S.  Means and Dodge. The user
can provide whatever adjustment is considered appropriate by en-
tering a percentage to directly adjust costs upward or downward.
If no factor is provided, costs are based on national average.
322    RCRA SITE REMEDIATION & EXPANSION

-------
                                       I.D.:   1.1.1

                                       Subassemblies:
             COMPONENT  FACT SHEET

                               Title:  Pipe vents

                          1 - Excavation
                          2 - Piping and materials
                          3 - Exhaust system
                                                                    BUTTERFLY
                                                                    VALVES
                                                          SRAVEL
                        (<) ATMOSPHERIC
                            VENT
                         MUSHROOM TOP
(b) ATMOSPHERIC
    VENT
   "U" TOP
                                                                       (d) VERTICAL PIPE VENTS CONNECTED TO FORCED
                                                                             VENTILATION MANIFOLD SYSTEM
            Exhaust System Subassembly Line items
            Fan: Belt-driven, utility mount, weather cover, corrosion-resistant coating mounted
             1 on  foundation/ready  to connect to on-site electrical supply HP =  0.0001575
              (AP)(SCFM)
            Flow meter: Mounted on steel frame
            Moisture trap: Trap and drain connection
            Butterfly valves:

            Piping and Laterals Subassembly Line Items
            Slotted pipe
            Coupling
            Solid  pipe (lateral piping diam. = riser diam.  + 2 in.), 100 ft lateral added to
              user input from fan to treatment
            Caps (mushroom or U-top), or tees

                                                              Figure 8
                                                         Pipe Vent Fact Sheet
                                          Excavation Subassembly Line Items
                                          Excavate holes or trench
                                          Spread burden
                                          Backfill with gravel (nut size)
                                          Cover with 6 in. dirt
                                          Source: SCS, 1981 (Ref. 3), Modified by PEI for detail.
 Health and Safety Factor
   This factor is applied to labor and equipment and represents
 increased costs  because of decreased productivity due to person-
 nel protection requirements in Superfund work as compared with
 conventional construction.
   Worker health and safety considerations or precautions can sig-
 nificantly  impact worker productivity and thus costs for some
 technologies at  some remedial action sites. The impacts  are ex-
 pected to  be more significant for control  technologies than for
 treatment  technologies. A literature review has revealed only one
 source of information that directly addresses the issue.3
   The reported  range of adjustments is extremely broad. For one
 technology, the reported percentage adjustment ranged from 58
 to 1990%.
   The user enters an adjustment percentage to increase labor and
 equipment costs based on judgment.
   Users of the model will have to be somewhat familiar with the
 nature of the waste at the site to be remediated and with the de-
 fined levels of personnel protection. Four levels of protection are
 defined. Suggested values are as follows:
Personnel
protection level
D
C
B
A
         Adjustment
         factor
                   1.0
                   1.1
                   1.3
                   1.5
• Level A—Requires full encapsulation and protection from any
  body contact or exposure to materials (i.e., toxic by inhalation
  and skin absorption).
• Level B—Requires self-contained breathing apparatus (SCBA),
  and cutaneous or percutaneous exposure to unprotected areas
  of the body (i.e., neck and back of head) is within acceptable
  exposure standards (i.e., below harmful concentrations).
• Level C—Hazardous constituents known; protection required
  for low-level concentrations in  air;  exposure of unprotected
  body areas (i.e., head, face, and neck) is not harmful.
• Level  D—No  identified hazard present, but  conditions are
  monitored and minimal safety equipment is available.

Geology Factor
  This factor is applied to selected components (equipment and
labor) which are affected by a site's geology. Example  compon-
ents affected are pipe vents, wells, trenches, etc.  Base  costs are
calculated for  "typical" geology.  This is defined as unconsoli-
dated silty sand with a water table  10 to 100 ft deep. The geology
factor is a direct multiplier of the affected components; i.e., an
input of 10% will increase costs by 1.1.

Decontamination Factor
  This factor is applied to equipment and labor  and represents
increased cost  due to the requirement to decontaminate equip-
ment. The decontamination of buildings and structures is a sep-
                                                                                  RCRA SITE REMEDIATION & EXPANSION    323

-------
          12-17-83

          JOHN  DOE
        U.S.  ENVIRONMENTAL PROTECTION AOENCY

REMEDIAL RESPONSE CONSTRUCTION COST ESTIHATINO SYSTEM
          PILE I  FILE3.BIN

         ANALYSIS NO.  12-333
                                                   GENERIC  SITE  ONE


                                                   ANYIHERE.  USA
                                                             CONTRACTOR IACNE COMPANY


                                                            CONTRACT NO. 12-«336ICDFA-43NMI(4
                                         COMMENTS/  EXAMPLE  OF  SUBSURFACE  DRAIN  MODULE CALCULATION FOR A
                                    SCOPE OF WORK I  HYPOTHETICAL  SITE.
          1.3.4.1 Subeurface Dralne IT)


          ITEM ID             DESCRIPTION
                 • i CAPITAL COSTS ••
     OUANTITY   UN    LABOR    MATERIAL  EBUIPMCNT
                                                                                                     HF
NODULE NUMBER 1


   OF        DF
                                                                                                                                  TOTAL
          •2318812M Trnoh Excavtng 4P»id
                     botta tank alpd .9-1
                     IBFdp 2CY Hrpr«ad*r

          »232953«5« Piping,  Subdrainig*,
                     Vtrfitd Cl«7 p*rf. 2F
                     Ingth* C-211 l**di»

          •2M2141M Catch Baalna-Nanhol*
                     Slab top* pracaat «'
                     thick 4Fdlaa aanhol*

          •2M2142M Catch Baaina-Hanhola
                     Su«p 4ft.
                     incio* dlaMttr

          •29M2*114CCatch BMln.-B.nhol.
                     Pracact 4FID rla«r
                     •27.79 pvr TLF 14Fdp

          •9M39M4* Kail coating acrylic
                     glaud coatinga aai.
                     For confined apace

          132e94M2KSuep Pucp  Capacity
                     1MDPH 19Fdeep veil
                     Trench Depth 12F, 14F

          132t94ta6eCPuep coat addition
                     Stnlee atl or pletlc
                            cap tol9Fdpth
           1919M2S«*BPlpe  plaetic  fitting
                      3  .Ibo.e
                      2'dlaceter

           191ft»e449eBPlplng
                      Valvea
                      2'dlaeeter

           19149«2eMBPlpe  Flberglaea
                      Reinforced
                      IMFplpe  2*dia>
           4217CY        1263         •      4M7      3379      1333       3*3     11M9
           2637CY
           2697CY
                            •      73*6
                            •      397B
           2637CY       27499
                                                                              I      3976
                                                •     13124      6*49       1374     31*46
            1933CY         397         •       31t       366       146        33       1212
            1333CT         714
            1236ST        2*39
                                              2*1       9*3       2*1        43       1664
                                                •      1121
                                                                            1B1      37*9
                                                          1236NSF      13311      7846      433*     1MM      4323       902     43612
                          387       123e        37       244        97        22      2*37
               7Ea
                           349       323        133       263        1*6         24       14*2
               TEa         179*       14M       669       1328       331        12*       3794
               7VLF       2468      49M        933       1871        748        17t      11»92
              73SF
                           161         67
                                                          98        39
                                                                                       391
               SEa        4373     21779
               9Ea
                             I     2999*
              13Ea         63*       12*
               9Ea
                            33       193
                                                 •      24K       962       218     29736
                                                                               •     2939*
                                                         346       138        31      1263
                                                                    12
            42NLF      22680*     77279         •    124748     49893     11348    49M34
                                                                       Figure 9
                                                                     Cost Output
324     RCRA SITE REMEDIATION & EXPANSION

-------
1330898028 Tank Fbrglaa Undgrnd
           20e00gai.cap.  unese
           •arway I it. tholddvna

0231608200 Bulk Excavating
           Backho* Hyd.  eravler
           .td.  ICY cao. '45CYhr

0230030001 Backfill  By  hand
           no compaction
           Light (Oil

823083885* Backfill
           Air taap,  add

8232202400 Grading Site excav.t
           fill 1/4 puih dozer
           Hand grading,  finiah

0284602200 Seeding
           Feacue 5. 5«/HSF
           Tall puah apreader

1310001000BTotala for 100Faaln
           Includea diach. pipe
           elbova & tee 4*aain

1510002008BAdd or aubtract for
           •ore or lea* than
           100F   6-aain

02303Se001ATrench & Backfill
           coat for trench 4.3F
           dpxlMFlongxS'vide

0230380002AGrade/Flnlah
           coat for trench 4. 3F
           dexlMFlongx8*vide
        ~~~~~~
   2Ea        5240     23440         0      2882      1132       262     34976
 720CY         604
 328CY        2902
  84CY         2S1
 124SY         204
                      S06       775       310        70      2565



                        0      1396       638       145      5281



                       62       188        75        17       623


                        0       112        44        10       370
 124HSF       1535        787       436      1084       433        98      4373
   5Ea        2420      1525
3700LF        5216      3365
  42Ea         966
                                            1331       532       121      5929
                                           2868      1147       268     14856
                                  1260       1224       489       111      4050
               630       210        42        369       147        33      1431
 HEALTH C SAFETY COSTS

 GEOLOGIC COHSIDERATIONS

 DECONTAMIHATION COSTS
167271

 66898

 15197
                                  7739

                                  3091

                                   698
 GENERAL  CONDITIONS
 GEOGRAPHIC CONSIDERATIONS
 SALES TAX
                                                             55351      19158      2561
                                                             55351       9575
                                                                        9575
                                                                                 2561
                                                                                                                          67487
                                                                                                                           9575
 Sy§IQTAL_DIRECT_CQSTS	

 SUBCONTRACTOR OVERHEAD AND PROFIT
 GENERAL  CONTRACTOR OHtP
  ON  SUBCONTRACTOR
 GENERAL CONTRACTOR OHtP
           _664216	229880	30732_
                                                            _924769_


                                                             386506
                                                                                                                          27585
 SUBTOTAL	

 PERMIT/LICENSE


 INSURANCE


 BOND


 ENGINEERING


 STARTUP
                                                                       _1258B60_

                                                                           12388


                                                                           62943


                                                                           12588


                                                                          202046
 CONTINGENCY
 ESCALATION
                                                                        _154902S_

                                                                          278624


                                                                           54834
 I9Iit-CAPIIAL.COST_I_!!gDULE_NyHBER.l	1B82683__


                                                    Figure 9 (continued)


                                                                                 RCRA SITE REMEDIATION & EXPANSION    325

-------
CODE
UM
OESCJ
DESC_2
DE5C_3
UNIT_MAT
UNIT LAB
UNIT_EQP
HEALTH_FLG
GEOLGY_FLG
DECONT_FLG
REFERENCE
COST_OATE
0252933040
L.F.
Piping, subdrainage.
Concrete porous wall
Cone, drain 8"diam
2.25
1.87
0.28
0
0
0
Means
6/84
(CSI code)
(Unit of measure)
(Description)
(Description)
(Description)
(Unit material cost)
(Unit labor cost)
(Unit equipment cost)
(Health factor flag)
(Geology factor flag)
(Decontamination factor flag)
(Cost reference)
(Cost date)
(CSI Construction Standards Institute)
                          Figure 10
               Example Record in Cost Data Base
arate component and is not covered by this factor. The base costs
do not include any decontamination. The user can input a per-
centage that is directly applied to the appropriate components.
Note that the above three factors are applied only to those items
flagged in the data base (Fig. 10). A zero flag means the item cost
is unaffected by the given factor.

Escalation Factor
   Costs are calculated in mid-1985 dollars. An escalation factor
is used to adjust costs for inflation. Separate factors are provided
for labor (also used for equipment) and material if the user de-
sires to use different factors. These values can be obtained from
many references as the user desires. References include the CPI,
CE Index, etc. The value is applied to total cost. An entry of
10% will increase costs by 1.1.

Contingency Factor
  The  contingency factor  is  a broad adjustment factor to in-
crease costs for any aspects of the estimate the user believes are
unaccounted for otherwise. In this program, the contingency fac-
tor can be used in the conventional manner (i.e., enter a per-
centage which will  increase all costs accordingly) or to implement
the more sophisticated probabilistic feature of the program. The
latter is done by entering "P" in the contingency field. When
this is done, an alternate series of input screens will be activated
which require not only entry of a given data value, but also min-
imum and maximum values and selection of an appropriate  fre-
quency (probability) distribution for the data value.

PROBABILISTIC ESTIMATING
  Uncertainty in the cost estimates for a given site can be due to
several different factors:
• Uncertainty in the scope of the job
• Uncertainty in the technology
• Uncertainty in the cost model
  Typically, cost models deal only with the last item, cost model
uncertainty. Costs  are loosely referred to as "budget estimates,"
"study  estimates." "detail  estimates,"  "site specific,"  etc.
These items imply some  level of certainty expressed  as plus or
minus some percent. In fact, the uncertainty is not really calcu-
lated, but an arbitrary error band is applied after a point estimate
is calculated, for example, ±30%. This percent is applied as a
qualitative indicator of the amount of detail considered in the
estimate. An accurate  estimate is obtained only after a  specific
job is engineered with design  drawings,  detailed takeoffs  and
vendor quotations to a bid specification.
  At this writing, the probabilistic cost estimating feature has not
been  implemented. An option  is being planned  structured to
accommodate the uncertainty of all three types referred to above.
Furthermore, each element of the cost  will have its own uncer-
tainty in order to calculate a composite error range (i.e., a joint
probability density function).
  The probabilistic approach takes  into  account the  fact that
some  elements of the estimate  have  much different uncertainty
than others, and, simultaneously, each element does not have
the same effect on the  final estimate. Clearly,  the objective is to
identify what factors have the greatest impact and to obtain the
most accurate estimate for those factors.
  The greatest uncertainty in Superfund estimates usually is re-
lated to uncertainty in  the scope of the job. The extent of con-
tamination is really only defined  after detailed site investigation
and  characterization.   One  advantage  of  the  probabilistic
approach  is that  it helps the user to  organize the estimate and
see where uncertainty has the greatest impact. Monte Carlo simu-
lation of the cost  estimating procedure can be used effectively in
this application. This is superior to simply assuming all variables
simultaneously are at their maximum or minimum, respectively.
  Practical limitations to the application of this approach are:
• User sophistication in such an approach
• User knowledge of uncertainty in the data
• Execution time  for the model, especially on a PC
• Storage requirements for the additional data and program sub-
  routines

CONCLUSIONS
  The large amount of resources committed to Superfund remed-
ial actions make it very important that  thoughtful and accurate
cost estimates be  performed.  Cost-effectiveness decisions are an
important aspect of resource allocation and budget policy analy-
sis. Computerized cost estimation programs have a proven record
in other areas (e.g., construction, process equipment) for improv-
ing  the speed, accuracy and consistency of cost estimates. This
program is designed to address the special problems associated
with costing hazardous  waste  cleanup. These  factors  include
special worker safety considerations, geological factors and vari-
ous degrees of maturity  of  remedial action  technology (e.g.,
drainage ditches vs. biological treatment).
  It is hoped that this program will be a useful tool for U.S. EPA
personnel and contractors  through the detailed design stage of
remedial actions.  Strong emphasis was placed on the program's
user interface. Experience with other programs has shown this to
be essential in ensuring widespread usage. A minimum amount of
computer knowledge is required to allow the user to concentrate
on the application.


REFERENCES
I. Means, R.S., Means Cost Data Books. 1983, 1984.
2. Richardson Engineering Services, Inc., Construction Cost Estimating
  Standards, 1984.
3. Walsh, J., Lippett, J., Scott,  M., SCS Engineers, "Costs of Remed-
  ial  Actions of  Uncontrolled  Hazardous Waste  Sites—Impact of
  Worker Health  and  Safety  Considerations,"  EPA-600/D-84-019,
  1983.
326    RCRA SITE REMEDIATION & EXPANSION

-------
                   Using Cost/Risk  Analysis  in  Waste  Planning:
                               The New England Pilot  Project

                                                Debora C. Martin
                                    U.S. Environmental Protection Agency
                                                Washington, D.C.
ABSTRACT
  In the fall of 1984, the Office of Policy Analysis at the U.S.
EPA began a Pilot Project aimed at portraying the flows of haz-
ardous wastes in and around one  region of the country and at
characterizing the impacts of future changes in that regional waste
flow system. Selecting the New England six-state region as the
subject of the Pilot Project, staff from the U.S.  EPA's Inte-
grated  Analysis Branch and Temple, Barker and  Sloane, Inc.
developed a computerized model tying together data on the vol-
umes and locations of industrial wastes in each state, estimates
of the constituents present in each waste, the location and method
of treatment, storage or disposal for each waste, representative
exposure characteristics for each  location and available health
and ecological effects data.
  Working with representatives from state, local, environmental
and industry interests, a waste flow data base and  computer
model were developed—analysis of waste planning strategies is
currently underway and a similar project is beginning in another
region of the country. This  paper describes the Project approach
and Phase I findings.
INTRODUCTION
  The Amendments to  RCRA require that the United States
move away from the practice of land disposal of hazardous wastes
to alternatives that minimize present and future threats to human
health and the environment and encourage the reduction of haz-
ardous wastes at the point of generation. The New England Haz-
ardous Waste Pilot Project is hopefully one way to assist regional
and state decision-makers in implementing these broad goals by
pulling together what is known about a particular "waste  sys-
tem" so that the best strategies for  shifting to  alternatives to
land disposal and the best candidates for source reduction  can
be identified, given  what we know now about potential risks to
public health and the environment. Thus, the two major objec-
tives of the Pilot Project are: (1) to paint a believable picture of a
regional waste management system and (2) to use it, in conjunc-
tion with other information, to help point to the best waste man-
agement strategies given the particular mix and location of wastes
in that region.
  The paper is arranged in three sections; the first explains the
nature of waste flow data base developed for the New England
Pilot Project and, equally important, the process used to develop
and refine that data base. The second highlights other critical
data elements and methods that were  tied together in Phase  I of
the Project. The third is an overview of the cost/risk model to be
used in Phase II of the Project and of the kinds of scenarios to be
evaluated.
  In reading this paper, the reader should remember that, al-
though there is little argument about the advantages of develop-
ing better information about the movement and effects of haz-
ardous wastes, the acceptable and unacceptable uses for cost/
risk analysis in waste planning still are being identified by Project
participants. While the concepts of quantifying risks and focus-
ing on exposures across all environmental medium are not new,
to our knowledge this is the first attempt at applying these con-
cepts to regional waste planning issues.

DEVELOPING THE WASTE FLOW DATABASE
  Our first steps in developing a waste flow data base for a region
are to determine which wastes should and should not be counted
and to set in place a series of data collection activities and data
confirmation  processes.  Hazardous wastewater regulated by
water permits,  Superfund removals, waste  oils, asbestos and
PCB wastes are not included in the pilot analysis. These exclu-
sions are made to realistically approximate the annual amount of
waste to be managed in the region—or, the amount that might
need to be managed differently if capacity shrinks due to facility
closures.
  Superfund and wastewater exclusions are made based  on the
assumption that their inclusion artificially inflates the amount of
waste to be managed on  an annual basis. In other words, for
Superfund,  removals conducted in 1 year might not occur on an
annual basis and, for wastewater, there are few economic incen-
tives for redirecting the large volumes of wastewater currently
managed on-site to other types of treatment facilities. (The waste-
water exclusion applied to wastewater that is treated  in a treat-
ment system permitted by the National Pollutant Discharge Elim-
ination System but is reported as hazardous waste under the Re-
source Conservation and Recovery Act programs; in one of the
states studied, inclusion of these wastes  would have increased the
 volumes of hazardous waste to  be managed  by almost  80%.)
 Similarly, it is assumed that waste oils, PCBs and asbestos require
 specialized treatments and, hence, there are few alternative treat-
 ment options to explore. These assumptions are easily changed
 and some are changing in the second regional analysis.
   The project team begins with the information that each state is
 required to submit reports to the federal government on a biennial
 basis and the annual reports required by some state governments.
 These reports, and the detailed data they are drawn from, often
 do not  provide all the information needed to  accurately reflect
 the flows of waste between generators and waste management
 facilities. Thus, meetings with various project participants, fol-
 lowup with state officials, generators, waste managers  and,  in
 some cases, the process of elimination play an essential role  in
 creating the instant picture of the regional waste flow system.
                                                                                                RISK ASSESSMENT    327

-------
  Two groups of project participants assist the New England
Pilot Project team by serving as reality checks on the data base
as it  develops. First, the  Regional Coordinating  Committee
(RCC), includes a small group of representatives from regional,
state, local, environmental and industry groups who have exper-
ience  or  strong interest in hazardous waste  management issues.
Individuals associated with the New England Council, the New
England  Environmental Network, the Vermonters Organized for
Cleanup, the Tufts Center for Environmental Management, the
New England Congressional Institute, the  New England Gov-
ernor's Conference, the New Hampshire  Governor's Office, the
Rhode Island League of Cities and Towns, the Connecticut Haz-
ardous Waste Management Services, the  Small Business Admin-
istration of New England are members of the RCC. Second, a haz-
ardous waste  staff in each of the states  provides assistance by
directing the project team to the best sources of data, reviewing
the data base for their state and helping  to reconcile differences
between  the Project data base and other data  sources.
   Much  is learned in the process of confirming waste flow dates.
For example, generator annual report information  often  is not
appropriate for determining what  types of treatments are being
applied to shipped wastes  once they reach their final destination.
This is especially true where transfer stations arrange  for land-
filling or incineration is actually occurring at the  transfer station.
Review of facility annual reports or personal  followup with waste
management facilities is necessary to determine if, in fact, the re-
ported treatment and disposal technologies exist  at the facilities.
Also,  facility  closures sometimes occur after annual or biennial
reports are submitted. In these cases the team  works with the
states  to determine, where possible,  which facilities the affected
generators are currently using.

Data Base Description
   There are almost 15,000 computerized waste  records describ-
ing the New England waste flow system. Each waste records con-
tains types of waste generated by the U.S. EPA waste code, the
U.S. EPA generator identification number,  the Standard Indus-
trial Classification (SIC) code for  the generator, the amount of
waste generated and treated on-site,  the amount  of waste gener-
ated and treated  off-site,  the U.S. EPA  TSD number for treat-
ment,  storage and disposal facilities receiving the waste, the type
of treatment (by U.S.  EPA Appendix I  Code) being performed
at the  facility and the community location for generators and re-
ceiving TSD facilities.
   After  exclusions, the current data base tracks 326 thousand
metric tons generated  by  industrial  sources  across the  six New
England states. Since this figure is based on required reporting
mechanisms, it does not pick up any generators who fail to report
their  wastes. Thirty thousand  metric  tons  are  generated from
waste  management facilities,  including  both wastes that pass
through  transfer stations  and residuals  from waste treatments
such as solvent recovery.
   The highest  volumes of wastes  tracked in New England are
various  wastes  from  electroplating  operations, solvents and
wastes reported as "otherwise  unlisted corrosive or ignitable"
wastes. One third of the waste is treated by chemical, physical or
biological treatment processes located inside or  outside the reg-
ion,  excluding solvent recovery operations  which account for
about 10% of the total waste treatment load.
   Regionally,  about one-half of the  waste is treated  on-site (in-
cluding the amount in storage at the end of the year)  and one-
half is treated off-site, with a little less than 50% of the waste that
is shipped to commercial facilities going outside of the region for
treatment,  most to facilities in New York  and New Jersey. Figure
1  provides state-by-state summaries of the  amounts of wastes

9O -
SO •
70 -
? S°-
§
3
O
!§. 40 -
30 -
20 -
1O -
T
^
/
^
^
/
'.
/
/
/
/
/
/
/
/
/
/
/








^
\
\
\
\
\
\
N
S
s

J

7-

'.
/
/
/
/
/




i
\ Yt « n-i TfAvm . , mRl frrfTTI 171 rTJtK
CT MA NM VT Rl UE
1771 ON-SITE [V3 OFF/IN g%Zl OFF/OUT PS53 OUT NE
                          Figure 1
                  Waste Flows in New England
                      (1983 Metric Tons)
treated on-site and treated off-site in the same state, in  one of
the other states in the New England region and in other regions
of the country.

PREPARING FOR STRATEGY ANALYSIS
  A computer model, designed to analyze planning strategies in
four ways, is developed to support the Hazardous Waste Pilot
Project through the strategy analysis phase. The model calculates
changes in the probability of health effects across the  region as
measured  by potential changes in  the concentrations  of waste
constituents transferred to air, surface water and ground water;
exposed populations; and  dose-response information for various
health  effects. In separate modules, it relates the same changes
in concentrations to environmental effects measures and relates
the changes in the distances that waste is shipped to transporta-
tion risk measures. Finally, it calculates changes in regional costs.
To perform these analyses, several data elements described below
are assembled.

Constituents, Concentrations, Costs and
Capacity
  The  constituents  and concentrations of constituents present
in each hazardous waste stream are estimated based on average
or typical  characterizations used in other projects conducted by
the U.S. EPA. The certainty associated with each waste stream
that is characterized varies.' For example, there is more certainty
about the constituents and concentrations of discarded off-specif-
ication product  wastes  than  with wastes that  are reported as
"ignitable, not otherwise listed."
  Costs for on-site treatment, storage or disposal are, for the
most part, based on the capital and  operating costs developed
for the various waste management methods in other studies con-
ducted by the U.S. EPA.' They are annualized and divided by
the annual operating volumes to determine cost per ton or cost
per gallon for various waste types. Prices rather than operating
costs are used for off-site facilities, generally based on Massa-
chusetts cost data.1
  A capacity file is assembled to help assure that strategies for
redirecting waste flows are realistic. In other words, if we move
something to another  facility  to reduce risk, is there really room
for it?  Most of the information gathered to date relates to larger
commercial facilities.
328    RISK ASSESSMENT

-------
Characterizing Potential Exposure
  The exposure  characteristics  identified in the model  are not
based on site-specific data in the strictest sense.  Rather, they are
designed to eventually assist in pointing to large differences  in
exposures between areas. If and when significant differences are
found, some check against real monitoring data will need to be
made. Currently, the location of generators and handlers of haz-
ardous waste is characterized according to the closest major com-
munity, the approximate latitude and longitude of the commun-
ity, population and general air, surface water and groundwater
conditions. These environmental factors are used with fate and
transport models and waste volume data to calculate the changes
in exposure (i.e., changes in concentrations) that may result under
various   waste  management  scenarios. Each  waste handling
method is assumed to have some potential route of environmen-
tal release and some probable amount of release. These potential
releases may be routine (as in the case of air releases from incin-
eration) or episodic (as in the case of spills around many types  of
treatment facilities).1
  Specific release routes for storage technologies include volatil-
ization into air routinely or as  a result of spills, migration into
surface  water from  overtopping or spilling, and seepage into
groundwater from contaminated leachate or spilling. The routes
for treatment and  disposal technologies are similar, but also in-
clude for thermal  treatment,  consideration of residual organics
and metals emitted into the air, scrubber discharge into surface
water and potential leaching from residues. For chemical treat-
ment,  potential equipment  failure is  included;  for  biological
treatment, effluent discharges to surface water are included.
  Although disposal via underground injection or ocean disposal
is not currently practiced in  New  England, these management
methods are modeled for use in future studies.  Possible well head,
case or  formation  failure are  considered for underground injec-
tion, as well as the obvious fate of ocean-disposed wastes.'
  To convert mass air  emissions into ambient concentrations,
conversion factors for point sources and area sources are used for
five concentric distances ranging from 1.05 km to 35 km.4 Storage
facilities, oxidation/reduction  facilities,  stabilization/fixation
facilities, biological treatment and  landfills are  modeled as area
sources while thermal treatment and solvent recovery are modeled
as point sources. All communities in the study are assumed  to
have similar air  environments described by an afternoon mixing
height of 1200 m and an average ambient temperature  of 7°C.
This assumption is made after reviewing data on  mixing height,
temperature,  wind velocity and precipitation for New England
weather star stations and finding insignificant differences.
  For surface water,  each community environment included  in
the analysis is assumed to contain  one of three types  of surface
water environments: a small coldwater stream with low flow less
than 3 m'/sec and a maximum temperature less than 25 °C; a med-
ium coldwater stream with a low flow between 3  and 30 mVsec
and a maximum temperature  less  than 25 °C;  or a large cold-
water river with a low flow greater than 30 mVsec and a max-
imum temperature less than 25 °C. Mass loadings are converted
to concentrations across five stream compartments for  three cate-
gories of pollutants including metals, high transformation pollu-
tants and low transformation pollutants.'
  Groundwater  environment characterization involves gathering
and reviewing  information  on hydraulic  conductivity,  linear
velocity and depth to groundwater. These data are matched  to
three representative flow field environments, all assuming a single
shallow aquifer. The first flow  field is characterized by an aver-
age linear groundwater velocity of 10 m/yr in the horizontal direc-
tion and 1 m/yr  in the vertical direction, the second by a velocity
of 100 m/yr in a horizontal direction and 10 m/yr in  a vertical
direction and the third by a velocity of 1000 m/yr in a horizontal
direction and 100 m/yr in the vertical direction. The model used
to convert loadings to concentrations in groundwater assumes
that exposure points are directly downgradient of the source of
contamination  and that aquifer thickness is constant.' Concen-
trations are calculated  at 600 m for total population exposure
and at 60 m to calculate a maximum individual exposure.
  The concentrations  that are calculated  represent an average
over the first 70 years after initial contamination. This assump-
tion is made to  deal with the timing problem associated with
comparing and summing potential risk across air, surface water
and groundwater. Concentrations of substances in air and surface
water occur within a short time-frame and thus can be measured
using annual averages. However, concentrations of substances in
groundwater may not appear for several years and, assuming that
they occur in the first year, may lead to an over-estimation of
potential  risk given current monitoring and corrective action
activities. Nevertheless, the current assumption and others easily
can be changed to determine worst case or best case scenarios.
  Finally, populations for each of the five  distances used to cal-
culate air concentrations  are stored in the computer model for
each community  environment, as well as an estimate, based on
the largest town   in the  35-km area,  of the  percentage of the
population drawing drinking water from groundwater sources.8
These percentages are 100%,  75%, 50%, 25%  or 5%, and the
balance of the drinking water is assumed to be drawn from sur-
face water sources.

Health Risk, Environmental Risk and
Transportation Risk
  Available data on potential health effects  from exposure to
various pollutants are obtained and used to define dose-response
relationships. These relationships eventually are employed to esti-
mate  incidences of health effects given  various exposure levels.
Health effects  are  classified into seven broad categories—can-
cer, reproductive effects, liver effects, kidney effects, fetal devel-
opment effects, neuro-behavioral effects and  other effects. The
sum of estimated  incidences across effects and pollutants for each
waste handling strategy is the health risk score for the strategy
in the analysis phase. Currently, dose response curves are avail-
able for approximately 30 constituents of various  hazardous
wastes, and others are being researched on an ongoing basis.
  The environmental effects component consists of an index of
"ecological integrity" based on dose response information devel-
oped  for aquatic populations only. These dose response relation-
ships are constructed after review of available data on the effects
of  pollutants on aquatic organisms and are used to estimate
potential impacts on finfish, vegetation zooplankton, macroin-
vertebrates and non-fish vertebrates resulting from increases or
decreases in chemical discharge to surface water bodies.
  For transportation risk, an average risk per  km of 0.45 per
million truck km  for a releasing hazardous waste accident is used
to assess changes  in transportation risk.' This calculation is based
on  the probability of a releasing accident on  a composite high-
way trip (interstate, state and urban).

ANALYSIS OF WASTE MANAGEMENT
STRATEGIES
  The project team has three objectives in Phase II of the Pro-
ject. In the broadest sense, the team is trying to determine to what
extent this type of  approach can benefit waste planning  decis-
ions as measured  by the reception of the analyses by project par-
ticipants  from local, state, regional, environmental and private
concerns.  Analytically,  the team is trying to determine if meas-
urable and significant changes occur when the waste system  is
                                                                                                     RISK ASSESSMENT    329

-------
changed in a variety of ways given what we know now about the
characteristics and effects of hazardous wastes in the environ-
ment.  Finally,  in  tying together all  available information, the
team is trying to determine where the biggest data gaps are and
how important or insignificant they are to waste management de-
cisions.
  The kinds of strategies  that the project team currently is test-
ing include a series of "what if" questions ranging from waste
reduction queries to treatment capacity issues.  For example, what
if all  electroplaters—identified by SIC code in the data base-
could implement a particular waste reduction initiative? Or, what
if the  ratio  of on-site  versus off-site treatment changes? Will
either  of these scenarios show significant increases or decreases
in potential risks and costs?
  The kinds of strategies  to be evaluated in the coming months
depend largely on the  interests  of projeci participants.  These
strategies will be analyzed  using the computer model.

The Computer Model
  All data elements and exposure algorithms are linked in a fairly
large  but conceptually simple computer model. In the  baseline
analysis, the computer  model relates volumes and locations of
wastes (broken down by constituent  concentrations) to the pos-
sible  environmental releases  and  exposure conditions around
waste  management facilities and to the health and environmen-
tal effects associated with each waste  constituent. Baseline trans-
portation  risk and baseline  costs  are computed  in separate
modules.  Thus, the model works as  shown in a simplified flow
chart (Fig. 2).
VOLUMES OF BASTES
OR DISPOSED BY
WASTE TYPE AND
CONSTITUENT


	 >

EXPOSURE ACROSS
ENVIRONMENTS ,
WASTED, AND
CONSTITUENTS


	 >

REGIONAL RISK
METHOD, WASTE,
AND CONSTITUENT

           MILES WASTE IS
           TRANSPORTED FRO*
           GENERATOR TO
           TREATMENT STORAGE,
           OR DISPOSAL FACILITY
            ANNUAL COSTS OF
            TRtATMENT, STORAGE,
            AND DISPOSAL
REGIONAL
COST
                           Figure 2
     Flowchart Describing Baseline Transportation Risks and Costs
  The first step  in analyzing a strategy is to identify the key
changes that need to be made to the baseline situation. For source
reduction scenarios, this means the percentage of waste and kinds
of waste that might be reduced on  an annual basis.  For shifting
wastes to alternative facilities, either existing or new, the amounts
and  kinds of waste to be shifted and  the alternative facilities,
facility locations and treatment operations must be identified.
These changes are entered into the model to compute new risk
scores, costs and transportation accident probabilities. These re-
sults  then are compared to the baseline results to determine if
significant changes occur.
  After the changes are  calculated, the next step is to "dissect"
strategies to determine why  the changes occur. The purpose of
this important step is to see  to what extent general  changes are
driven by volumes of waste, waste  characterizations, environ-
mental settings and/or amounts of available effects information.
 In this way, we can begin to determine the relative importance of
 the many variables involved and to highlight critical uncertainties
 and data gaps.
  There are some important questions that this kind of analysis
 cannot answer. While it provides a context for comparing, across
 environmental media, the risks and  costs of a variety of waste
 management options at various locations, it does not compare the
 risks  at waste handling  operations to the risks  associated with
 other industrial operations at the same locations. This is prob-
 ably a valuable piece of information to communities which might
 be asked to accept risks associated with new waste treatment facil-
 ities,  no matter how minimal they appear relative to other pos-
 sible locations.

 CONCLUSIONS
  The New England Hazardous Waste Pilot Project is one way to
 assist in implementing the goals of federal and state legislation.
 In Phase  I we've learned that there is, in particular, a need  for
 better understanding of waste flow systems from a state, regional
 and possibly national perspective. We have learned that  by  re-
 solving  definitional   issues and  involving  key interests  in  an
 attempt to quantify  amounts of waste can dramatically change
 perceptions of the magnitude of a particular hazardous  waste
 system.
  In Phase II, the team will work to develop ways of maintaining
 waste flow information on an ongoing basis and continue to work
 with  project  participants  to analyze  waste  planning  strategies.
 By  continuing to work with participants  from the New England
 states and U.S. EPA officials, we hope to refine both the data
 and methods  used in the approach and to determine acceptable
 and unacceptable uses for Projeci outputs.

 REFERENCES
 I. ICF Incorporated, "The RCRA Risk-Cost Analysis Model Phase III
   Report," Submitted to the Office of Solid Waste Economic Analysis
   Branch, U.S. EPA,  Mar. I, 1984.
 2. Massachusetts Department of Environmental Management,  "Haz-
   ardous Waste Management in Massachusetts Environmental Impact
   Report," 1982.
 3. "1983 Connecticut  Hazardous Waste Generation and Management
   Analysis,"  ERM-Northwest,  Incorporated, Feb.  1985. "Environ-
   mental  Impact  Report,"  Massachusetts  Department of Environ-
   mental  Management, 1984. Telephone contacts  with Slate environ-
   mental program staffers in Vermont.
 4. Temple, Barker &  Sloane, Incorporated, "Risk and Cost  Assess-
   ment of Hazardous  Waste Incineration Regulation," (Draft), "1984
   Methodology adopted from Inhalation Exposure Methodology
   Modeling of Hazardous Waste Incinerators,"  prepared by MidWest
   Research Institute, 1984.
 5. Temple, Barker & Sloane,  Incorporated, "Foundry Industry Method-
   ology," 1983, methodology and results prepared by Arthur D. Little,
   Incorporated using EPA EXAMS Model.
 6. Ocean dumping exposure and risk assessment is taken from GCA's
   earlier sludge work, "Ocean Disposal of Sewage Sludge, April  1985."
   This work is currently being re-done by Aqua Terra  Consultants and
   we will  incorporate the  new results  into the  Regional  Hazardous
   Waste Model as appropriate.
 7. Geraghty & Miller,  appendix B to "Liner Location Risk and Cost
   Analysis Model," 1984, groundwater dispersions adopted from  the
   Random Walk Particle Tracking Model of Prickett, el al.
 8. For populations, U.S. EPA GEMS Modeling System is used;  for
   drinking water assumptions, U.S. EPA Office of Drinking Water Fed-
   eral Reporting Data Systems (FRDS) was used.
9. Albowsty, el al.. "Assessing the Releases and Costs  Associated with
  Truck Transport of Hazardous Wastes," 1984.
330    RISK ASSESSMENT

-------
                 Assessment of Potential  Public  Health Impacts
                       Associated  with Predicted Emissions  of
                         Polychlorinated  Dibenzo-Dioxins and
                       Polychlorinated  Dibenzo-Furans  from a
                                  Resource Recovery Facility
                                             David Lipsky, Ph.D.
                                            Dynamac Corporation
                                             Fort Lee, New Jersey
ABSTRACT
  This paper provides summary results of a risk assessment per-
formed by Fred C. Hart Associates for potential dioxin-related
impacts associated with a proposed resource recovery facility to
be constructed in  New York City by the NYC Department of
Sanitation.
  With the assumptions used in this risk assessment, the max-
imum increased cancer risk due to the complex mixture of PCDDs
and PCDFs present in the emissions ranges from less than 0.24 to
an upper-bound limit of less than 5.9 x 10~6 for inhalation, in-
gestion and dermal absorption exposure pathways combined. In
our opinion, within the context of the assumptions used in the
risk assessment,  the methods used to estimate risk and consider-
ing current regulatory practice regarding acceptable versus unac-
ceptable health risks, a worst-case upper-bound excess cancer risk
less than 5.9 x  10-6 is below levels found by many regulatory
agencies to require additional  review and probable action to
reduce risk.

INTRODUCTION
  The facility,  known as the Brooklyn Navy Yard Resource
Recovery Facility (BNYRRF), will be a 3,000 ton/day mass burn
facility that will include four 750 ton/day combustion units and
will incorporate the Martin stoker system.  The BNYRRF will
have natural gas auxiliary burners located at the point where
secondary air will be injected to maintain a minimum secondary
combustion temperature. Emission control systems will include a
high-efficiency (greater than 99.5%) fabric filter to meet an outlet
particular emission loading of 0.015 gr/dscf at 12% CO2. This
particular emission rate is well below the Federal New Source Per-
formance Standard (40 CFR 60.52) and the New York State emis-
sion limitation (Part 222). Subsequent to our completion of this
risk assessment,  a dry acid gas scrubber system was added to the
proposed pollution control ensemble.
  The risk assessment was completed after an extensive examina-
tion of the literature focusing on the high temperature behavior of
dioxins in incinerators. Reports of laboratory studies as well as
sampling and analysis data on flue gases from actual incinerators
were included. As  part of the risk assessment, the predicted am-
bient air concentrations and daily intake of polychlorinated di-
benzodioxins  (PCDDs)  and  polychlorinated  dibenzofurans
(PCDFs) were compared to applicable standards and criteria and
concentrations of  PCDDs and  PCDFs  known or suspected to
cause a toxic effect. As discussed below, a number of conservative
assumptions were utilized  in the risk assessment.
EXPOSURE ASSESSMENT
  Emission test data from a number of incinerators have demon-
strated a wide range of PCDF and PCDD emissions spanning
several orders of magnitude. However, the lowest emissions were
observed from incinerators with  the  same  basic design as the
BNYRRF. Actual emission data  from operating municipal re-
source recovery facilities were used to  predict PCDD and PCDF
emission rates from the BNYRRF. In selecting the test data, the
objective was to utilize data from facilities similar  in design,
operations and waste types to the BNYRRF. Furthermore, test
data had to be obtained using sampling methods that are compati-
ble with methods approved  by  the  U.S.  EPA for  sampling
organics in flue gases under rigorous quality control procedures.
The emission data selected as the most representative were ob-
tained fromthe Chicago Northwest facility1  and the Zurich-
Josefstrasse  facility.2 Both use a  furnace design similar to that
proposed for the BNYRRF; however, both use electrostatic pre-
cipitators rather than fabric filters for fly ash emission control.
  If the PCDFs and  PCDDs are emitted wholly in gaseous form,
then the emission data obtained from similar systems could be ap-
plied to the BNYRRF, regardless of pollution  control  system
dissimilarities. If the PCDF and PCDD materials are wholly ab-
sorbed onto particulate matter, more  of these materials  will be
trapped in the fabric  filter than in a comparably sized electrostatic
precipitator. Measurement data show generally higher concentra-
tions of PCDFs and PCDDs on emitted fly ash than  on  fly ash
trapped in electrostatic precipitators, indicating  a relationship
that is influenced by particle  surface area.  By utilizing particle-
size data and an assumption that PCDF and PCDD materials are
absorbed onto fly ash in proportion to the available particle sur-
face area, calculations were made concerning the amounts col-
lected and emitted.
  Since the partition between gas and  solid forms is not known,
emissions were calculated in two ways: (1) assuming all PCDF and
PCDD materials observed at the test facilities are gaseous and (2)
assuming all PCDF  and PCDD  emissions observed at the test
facilities are particulate.  If the emissions contain both gaseous
and particulate PCDFs and PCDDs, then the emission rates will
fall between these values. These emissions,  corrected for the size
of the BNYRRF, are shown in Table 1 as mass rates per unit time.
  To  assess the impact  of  the  emissions  from the  proposed
BNYRRF on the general population, computer modeling was per-
formed on the emission data presented above to predict down-
wind concentrations of PCDFs and PCDDs in the ambient air,
soil, dust and dirt. The predictions of ambient air concentrations
                                                                                           RISK ASSESSMENT   331

-------
                           Table 1
        Summary of Predicted PCDF and PCDD Emissions
          Brooklyn Navy Yard Resource Recovery Facility
                     (Outlet of Fabric Filler)
      TH-CDF
      Tetri-COF
      Ptnti-CDf
      Hoi-CDF
      Hepta-CDF
      Octa-CDF

      Ton I-CDF
      Trl-CDD
      Tttrt-CDD
      F-tnla-COF
      H««a-CDO
      Hepta-CDD
      Octa-COO

      Total-CDO
      2.3,7,6 T«tra-C00
51.16
15.35
 4.43
10.56
 1 28
 0,10

82 92
 2.22
 1.07
 1.79
 2.73
 1.30
 0.43

 9.54
   Cast 2
All Partkulau
  Emissions
  (ua/itc)

    23.95
    7.18
    2.07
    4.94
    0.60
    0.05

    38 79
    1.03
    0.51
    0 83
    1.28
    0.61
    0.20

    4.46
Case 1:  assumes all PCDF and PCDD materials are gaseous, therefore no additional collection
occurs in fabric filler above that collected by an electrostatic precipitator.
Case 2:  assumes all PCDF and PCDD materials are adsorbed on paniculatcs, therefore additional
fly ash collection in fabric filter reduces PCDF and PCDD emission.
  The emission rales may vary between these values if there arc gaseous and paniculate fractions
of PCDF and PCDD emissions.
of PCDFs and PCDDs quantify potential population exposure
through  the inhalation  pathway, while the deposition analysis
provides a quantification of potential human  exposure through
ingestion of contaminated dirt and dust as well as through dermal
contact with these materials.
  The model used for analyzing air quality impacts resulting from
the BNYRRF stack  emissions was the Multiple  Point Source
Gaussian Dispersion  Algorithm  with  Optional Terrain Adjust-
ment (MPTER) modified  specifically for use in the urban en-
vironment. When using this dispersion analysis methodology and
the PCDF and PCDD emission calculations from Case 1 and Case
2 of Table 1, it is assumed that all PCDFs and PCDDs are emitted
and that they behave in the atmosphere as a gas. Application of
these emission rates to  the maximum normalized concentrations
results in the maximum annual average concentrations shown in
Table  2.  Since these concentrations of PCDDs are extremely
small,  the units  presented are picograms/m'  (pg/m*  or 10-'2
g/mJ).
   In Case 2, it was assumed that all dioxins entering the fabric
filter enter attached to particulates. Those particulates that escape
the fabric filter, if heavy enough, will deposit on ground surfaces,
mixing with dust, dirt and soil for potential uptake through inges-
tion or dermal  pathways. To determine the rate at which the par-
ticles  settle to the  ground, a  computer model  with specific
capability  to analyze particulate deposition rates was selected.
This  model  is the  Industrial  Source Complex  (ISC) model
developed  under the sponsorship of the U.S. EPA by a private
contractor.'
   The maximum concentration of PCDDs in soils was estimated
by assuming a one centimeter mixing zone and a soil density of 1.6
g/cm' and estimating a build-up over a 20-year  period. The
estimate is clearly conservative, with no losses due to volatiliza-
tion, degradation or physical clearing.
   PCDD levels in  home dust  and street dirt  on  a  surface area
basis as opposed to a weight basis also were used in the estimate of
risk.  By  comparing  indoor deposition  and  accumulation  to
measured  lead data,  the  maximum  deposition  on floors was
estimated at 25% of the outdoor rate with a 30-day period of ac-
cumulation."
   There is some evidence to suggest  that a 30-day accumulation
period is overly conservative. The data provided by Solomon, el
a/.,' if extrapolated, indicates 0.6 10  1 g dust/m^  in single family
urban middle class homes (gas heated) in Champagne-Urbana, Il-
linois.  His data,  if applied  to the PCDD and total particulate
deposition rates,  would indicate that steady state conditions for
dust buildup in homes would be reached in 3 to 5 days.
   For  outdoor deposition,  the  accumulation  period in urban
streets is unknown. The Swiss EPS, in its assessment of dioxin
emissions from incinerators, used an estimated half life equal to
14 days for rural environments.2 Consistent with  this Swiss EPA
assessment, we have assumed a 30-day period of accumulation as
a reasonable accumulation period.
   Using the deposition model, the concentration of PCDDs in
street  dirt was determined  by dividing the PCDD deposition rate
by the  average  particulate deposition rate   from non-dioxin
sources. The deposition rate of non-dioxin particulates in Brook-
                                                             Table 2
  Predicted Maximum Exposure Concentrations of PCDFs and PCDDs Due to  Emissions from the Brooklyn Navy Yard Resource Recovery Facility


Trl-COF
Ttlra-COF
Ptnta-COF
Haia-CDF
H>pta-CDF
Octa-CDF
Total-CDF
Trl-CDO
Titra-CDD
P«nti-C00
Htia-CDD
Htpta-COD
Octa-COD
Total -COD
2.3.7.8 Tatra-CDO
Can 1: >CDF and fCDO
lilnloni Art Castoui.
Hixiaua Conctntratfon
In Air (pg/ar)
1.2281
0. )(84
0.1063
0.2539
0.0)0?
0.0024
1.9900
0.05)]
0.0257
0.04)0
0.0(55
1.0112
0.0101
0. 2290
0.001(8


Nanlatu* Concanlrallo
(n Air (po/.'l
0.574?
0.1721
0.0497
0.1187
0.0144
0.0011
0.9)09
0.0241
0.0012
0.0120
0.0)06
0.0146
0.0049
0.08*2
0.000782
CM. 2; All KOF and

n Haxlstai Ajmuil
Dapoilllon Ilia (m
4.790
1.43«
0.414
0.990
0.120
0.009
7.759
0.208
0.101
0.1(7
0.25(
0 122
0.041
0.195
0.0065
KDO Eiliilant Art Adiortad



. In Sail A(Ur Concantratloni In )
!/•*) OM Y.ar («J/j) Slr.it Dirt (iu>/a) II
299
90

e«
g
O.fe
4B&
._
*
10

1

M
0.41
44

.
13

0.1
104
j
]
2

J
«fij
12
0.087


IHIUy Acci^litlo
toufti Dust (DQ/B )




I*
.3
1U


Jc
, 9

2.S
_jj
i«
0.135
         Notes: The concentrations presented in this table assume no chemical/biological degradation or transformation of PCDF and PCDD compounds.
         pg/m^ — picograms per cubic meter
         ng/m^ — nanograms per square meter                 Pg/12 ~ picograms per gram
         fs/g — fentograms per gram                       pg/irr — picograms per square meter
332    RISK ASSESSMENT

-------
lyn was measured at 74 g/mVyear." Table 2 shows the average an-
nual deposition rate for 2,3,7,8-TCDD equal to 0.0065 ng/mV
year. At any point in time, this material is being diluted by 74
g/mVyear  of particulates from non-dioxin sources. Therefore,
for the purpose of this assessment, we predict a maximum 2,3,7,8-
TCDD concentration to equal 8.8 x 10-14 g/g of street dirt. This
is  a highly  conservative estimate, since the dioxin containing fly
ash will be diluted not only by other airborne particulate matter
but also by lint,  dirt, gravel, etc. already present on the surface.
The  estimates are particularly conservative in that no degrada-
tion, volatilization or  physical cleaning (e.g., runoff to storm
sewers) of TCDD is assumed.
   On a weight to weight basis, TCDD concentrations in home
dust were assumed to be equal to the concentration in street dirt,
an assumption which is used by CDC in its risk assessment and
supported by some studies of lead contamination in indoor versus
outdoor environments.3

RISK ASSESSMENT
   A risk  assessment  was  performed  to  estimate the  risks
associated  with predicted maximum PCDD and PCDF concen-
trations in ambient air, soil, dust and dirt  downwind of the
BNYRRF.  The assessment compared the predicted ambient air
concentrations and estimated daily intake levels for PCDDs and
PCDFs with: (1) applicable standards and criteria; and (2) con-
centrations of PCDDs and PCDFs known or suspected to cause a
toxic effect.
   Several conservative assumptions were incorporated into our
assessment. They include the following:
•  Use  of worst-case assumptions for three potential pathways of
   exposure: inhalation of gaseous or particulate emissions,  in-
   gestion and/or dermal absorption of particulates deposited on
   outdoor  and indoor surfaces. Typically,  risk assessments of
   point source emissions assess risks only for the inhalation path-
   way.
•  Application of a toxic equivalency multiplier equal to 59. This
   number provides an estimate of the non-carcinogenic toxicity
   of the complex mixture of PCDDs and PCDFs compared with
   the concentration of 2,3,7,8-TCDD.  This  equivalency factor
   also was used to estimate carcinogenic equivalency as a worst-
   case assumption although such an application is  speculative.
•  Reductions in potential PCDD  and PCDF concentrations in
   the environment due to volatilization or degradation or losses
   due to such actions as rainfall were not considered.
•  Exposure assessments for ingestion and dermal pathways were
   done for both indoor and outdoor exposure; although build-
   ings may have a filtering effect on airborne particles, indoor
   dioxin concentrations were assumed to be as high as outdoor
   ambient  concentrations  for  the analysis by  the  weight ap-
   proach.
•  Risks were calculated based on  exposure for 24 hours a day,
   over a 70-year lifetime, at the point of maximum impact.
•  No threshold for a safe level of dioxin exposure was assumed;
   any exposure to dioxin was assumed to pose some health risk.
   Moreover,  of  the accepted  models for extrapolating the po-
   tential risk of very low level exposure, the  most conservative
   model factor was used.
•  Conservative assumptions for bioavailability and rates of in-
   gestions  and dermal contact were included to estimate worst-
   case daily intakes of PCDDs and PCDFs.

   The  results of our risk analysis  are summarized below. Max-
imum average annual ambient air  concentrations (Table 2) were
predicted for ground-level and elevated receptors assuming that
all PCDDs and  PCDFs are emitted as  a gas. The maximum
average annual  ambient air concentrations are predicted  to  be
well below existing standards and guidelines  typically less than
1%. Only the NYSDEC performance guideline for evaluating
combustion sources  is  approached  but at only  28% of  the
guideline.
  Risks associated with inhalation of PCDDs and PCDFs  at the
maximum  average  annual  concentrations were  determined  by
comparing a predicted maximum daily intake through inhalation
(DIinh) with Acceptable  Daily Intakes (ADIs) and  cancer  dose/
response extrapolations.  The assumptions used to calculate DIjnh
are provided in Table 3.

                           Table 3
 Upper-Bound Estimated Percentage of Acceptable Daily Intake (ADI)
  Attributable to Inhalation of PCDD and PCDF Emissions from the
         BNYRRF at Ground Level and Elevated Receptors

                          Acceptable Daily Intake (ADI)
                 Canada ...Netherlands,,,   NYSDOH ,,„,    CDC   ,..,
               10 pg/kg-day(B) 4pg/kg-dayti>) 2pg/kg-dayt1"' l.Spg/kg-day"1'
Upper-bound
Daily Intake-
Inhalation
Pathway

2,3.7,8-ICDO(a)
(2.7 > 10"4pg/kg-day)


TCDD'")
(4.2 x 10~3pg/kg-day)
 TCDO Toxic
   Equivalency^'
 (1.6 x 10"2 pg/kg-day)
                  0.0027X
                 0.16X
                           D.007X
                           0.11X
                          0.40X
                                      0.014X
                                      0.21X
                                    0.80X
                                                 0.23X
                                                0.89X
Assumes
  DIinh (g/day) = PCDD (g/m3) x ventilation rate (mVday) x particulate reten-
  tion (%) x bioavailability (%)
  75% of the inhaled particles are retained by the body12
  15 m3 of air are exchanged in one day"
  100% of the inhaled particles are bioavailable
  Maximum ambient air concentration of 2,3,7,8-TCDD equals 0.00168 pg/m3
  70 kg/person inhaling .00168 pg/m3 of 2,3,7,8-TCDD
(a) DIinh  (pg/kg-day) = DIinh  (g/day)/70kg = 1.9 x 10"2 pg/day/70 = 2.7 x
   10-4 g/day
(b) DIinh (TCDD) = DIinh (2,3,7,8-TCDD X  15.5)
(c) DIinh  (Toxic Equivalents) = DIinh (2,3,7,8-TCDD x 59)


As shown  in Table 3,  the  maximum daily intake DIinh of
2,3,7,8-TCDD, TCDD and TCDD toxic equivalents are predicted
to be well below (less  than 1%) any ADI promulgated by any
regulatory agency. These ADIs provide a margin of safety and
they identify a very safe dose, below which risks are insignificant
for non-carcinogenic toxic effects.
  An upper-bound to carcinogenicity risk was also determined by
comparing  DIinh with  three  different  cancer  dose/response ex-
trapolations. The most conservative dose/response extrapolation
is that used by the U.S. EPA (unit risk estimate equal to 1.56 x
10~7).7  As shown in Table 4, the upper-bound excess risk due to
exposure to 2,3,7,8-TCDD and HCDD, at the point of maximum
impact, ranges from less than .0019 to less than 0.055 X 10-6 or
less than six cases per 100 million people exposed to the maximum
concentration over a 70-year  lifetime.
  The sensitivity of the risk estimates  derived for the inhalation
pathway to the additional risks that might be  attributable to the
ingestion and dermal pathways was  tested. These two pathways
were not considered in the initial risk assessment of the inhalation
pathway due to the high degree of uncertainty in the  estimates of
ingestion, absorption and bioavailability rates and due to the dif-
ficulties in modeling PCDD concentrations in soil, dirt and dust.
                                                                                                     RISK ASSESSMENT    333

-------
                             Table 4
    Upper-Bound Estimated Increased Cancer Risk for Population
     of One Million People Exposed for a 70-Year Period to (he
         Maximum Daily Intake via the Inhalation Pathway

                       Dose for  1 x IQll Excess Cancer  Rlak
                       Klmbrough     Klmbrough      EPA
                       (1.4 pq/kq-    I .028 MAa±   I .0064 pq/Ra-
                         day)         day)          day I

                       Upper Bound Excesa Riik for 70 Kg man
        Daily Intake

        1.9 x 10-2
         pg/day
        (2.J.7.8-TCDD)
        0.15 po/day
        (1,2.3.6.7,1
        and
        1.2.3,7,8.9-HCDDI
A range of  daily  intake  estimates was calculated (DIing and
Dldertn)- Tne assumptions used to calculate DIing and DIdcrm are
provided in Tables  5 and 6.

                             Table 5
          Estimaled  Daily Intake of 2,3,7, 8-TCDD through
                      the Ingestion Pathway
                                 Source of Exposure
                              Ho«e Oust*
                                          Street Olrt
                                                     Soil'
A. Height Approach


Concentration of 2.3.7.8-ICDD    fl.8xlo"14o/g  S.SxlO"14 g/8   8 2xlo"15g/g


Mechanist.                             Daily Intake (g/day)
Ingestion of:
100  «g/day -  30X Bloavail
                                        2.6.10"
                                 .-14
410  «g/day - 801 Bioavail      2.9.10""    2.9.10"


B. Surface Area Approach


Concentration of 2.3.7.8 TCDD    1.3.1013g/"2  5.3«10"13g/»J


                         Dally Intake (g/day)
                                                        2.5.10"
                                                        2.7.10"
    HechaniSB
    Contact with:
    0.0033 «2/day - 30* Bloavail


    0.016 «2/day - BOX BioavalI
                         1 3.10"16    5 2.10"16    5 2.10"
                                    6.8x10"
                                                    •15
   ' 20-year accumulation period
   2 Assumes 30-day accumulation period al 25% of outdoor accumulation rate


   As summarized in Table 7, for the low end of this range, the
 data indicate that the predicted  additional 2,3,7,8-TCDD intake
 from  ingestion  or dermal absorption  of paniculate emissions
 from the BNYRRF would be negligible.  Using worst-case assump-
 tions regarding the maximum rates of exposure through ingestion
 or dermal in absorption, the total daily intake of 2,3,7,8-TCDD
 would  increase by  2.4 times  the amount from inhalation alone.
 Under this "worst-case" scenario, ingestion  becomes the major
 pathway of exposure.
   The general conclusions regarding the level  of risk posed by the
 BNYRRF are not significantly altered if the maximum total daily
 intake  of 2,3,7,8-TCDD increases 2.4 fold. This additional daily
 intake  of 2,3,7,8-TCDD equals only 0.036%  of the most conser-
 vative ADI (1.8  pg/kg-day)  and a toxic equivalent daily intake
 which is only 2.1% of this ADI.
                                                                                               Table 6
                                                                     Estimated Dally Inlake of 2,3,7,8-TCDD from Dermal Absorption

                                                                                                     Sourc* of Exposure	
                                                                                                Ham, Ou.t	Street Dirt   OT

                                                                      A  Weight Approach

                                                                      Concentration of 2,3.7.8-ICOO    8.8.10"" g/g 8 8.10"14 g/g  8.2.10   g/g

                                                                                                           Dally Intaka (g/day)

                                                                      055 g of dust -1* Absorption    4.8.10""    4.8»10~16     4.5.10"1

                                                                      0 55 g of dust -10)1 Absorption   4.8.10"15    4 8.10"      4.5«10

                                                                                                  	Source of Exposure	
                                                                      B  Surface Area Approach

                                                                      Concentration of 2.3.7.8-tCDO
                                                                      0 028 •  of surface
                                                                          • re« -U Absorption
                                                                                                    13 . 10"13 g/.2
                                                                                                                   Street Dirt"
                                                                                                                  i.3 . 10"U 9/m2
                                                                                                3 6 . 10
                                                                          0 14 .  of surface
                                                                              4fci -10t Absorption
                                                                                                          •15
                                                                                                            Dally Intake (g/day)


                                                                                                               1.5 x 10"16

                                                                                                                    -15
                                                                                                1 8 . 10""       ; 4 . 10

                                                                   ' PCDf) concentration* in M>iK on a Mjrfacc area bam arc a-«umed lo be identical wilh Hrecl din.
                                                                                               Table 7
                                                                         Calculation of a Total Daily Inlake of 2,3,7,8-TCDD for
                                                                          Inhalation, Ingeslion and Dermal Absorption Pathway

                                                                          Case 1   PCDOs and PCOfs inter fabric filter in Gaseous Phase

                                                                                 01 "»,„. ""ing""*™
                                                                                 01  1 9 x 10"" -0-0
                                                                              Cast 2  PCODs and PCDFi Enter fabric filter Attached to
                                                                                                                 4 . 10
                                                                                                                      "15
                                                                                     Dl  .  = 1 0 x 10
                                                                                     OIWI = 45. 10"   g/«ay
                                                                      The upper-bound carcinogenic risk estimates will also increase
                                                                   by a  factor of 2.4 assuming the most  conservative estimates for
                                                                   the ingestion and dermal absorption pathways. This would result
                                                                   in an upperbound increased cancer  risk for a population exposed
                                                                   to the maximum 2,3,7,8-TCDD and HCDD concentrations over a
                                                                   70-year lifetime ranging from 0.0046 x 10-6 to less than 0.13  x
                                                                   10-6.
                                                                      As a further measure of conservatism, we examined the poten-
                                                                   tial carcinogenic risks due to exposure to the complex mixture of
                                                                   PCDDs and  PCDFs emitted  from the BNYRRF for all three
                                                                   potential pathways of exposure. Because no actual animal studies
                                                                   have  been conducted to assess the carcinogenicity of most of the
                                                                   individual PCDD/PCDF isomers on the complex mixtures pre-
                                                                   sent in fly ash  emissions, another  conservative assumption was
                                                                   made:  that the  carcinogenic  effect   of these  compounds  is
                                                                   equivalent to their toxic effect  (i.e., that carcinogenic equivalency
                                                                   equals a  toxic equivalency). However, this assumption is highly
                                                                   speculative. With  these conservative assumptions, the maximum
                                                                   increased cancer risk due to the complex mixture of PCDDs and
                                                                   PCDFs present in the emissions ranges from less than 0.24 to an
                                                                   upper-bound limit of less than 5.9 x 10-6 for an these pathways
                                                                   combined.
                                                                      In  our opinion, within the context of the  assumptions used in
                                                                   the risk assessment, the methods used to  estimate risk and con-
                                                                   sidering current regulatory practice regarding acceptable versus
                                                                   unacceptable  health risks,  a  worst-case upper-bound excess
334    RISK ASSESSMENT

-------
cancer risk less than 5.9 x  10-6 is below levels found by many
regulatory agencies to require additional review and probable ac-
tion to reduce risk.

REFERENCES
 1.  Redford,  D.P.,  et al.,  "Emission of  PCDD  from Combustion
    Sources," International Symposium on Chlorinated Dioxins and Re-
    lated Compounds,  Arlington, VA,  Oct. 1981.
 2.  SFOEP (Swiss Federal Office for Environmental Protection), "En-
    vironmental Pollution Due to Dioxins and Furans from Communal
    Rubbish Incineration Plants," Schriftenreihe Unweltschutz, No. 5,
    1982.
 3.  Cramer, H.E.  Co., Inc., "Industrial Source Complex (ISC)  User's
    Guide," Vol. I. NTIS PB80-133044, 1979.
 4.  Nudelman, H., New York  City Department of Environmental Pro-
    tection, 1984.  Personal communication  to  Ben  Miller,  New York
    City Department of Sanitation, June 29, 1984.
 5.  Harrison,  R.M., "Toxic Metals in Street and Household Dusts,"
    Science Total Environ. 11,  1979, 89.
 6.  Solomon,  R.L. and Reinbold,  K.A., "Environmental Contamina-
    tion by Lead and Other Heavy Metals," National Science Founda-
    tion. Publ. No. NSF/RA 770681-5, 1977.
 7.  U.S. EPA, 1984, Ambient Water Quality Criteria for 2,3,7,8-Tetra-
    chlorodibenzo-p-dioxin. Office of Water Regulations and Standards,
    U.S. EPA, Publ. No. 440/5-84-007.
 8.  Harding,  D.H., "Chlorinated Dioxins and Chlorinated  Dibenzo-
    furans: Ambient Air Guideline," Health Studies  Service, Special
    Studies and Services Branch, Ministry of Labor, 1982.
 9.  Heigden, et al., "Royal Institute of Public Health,  Bilthoven, The
    Netherlands, Report DOC/LCM 300/292, 1982.
10.  Kim,  N. and Hawley,  J.,  "Revised Risk Assessment,  Binghamton
    State  Office Building, Albany, NY, New York State Department of
    Health, Bureau of Toxic Substance Assessment, Jan. 17, 1984.
11.  Kimbrough, R., et al., "Health Implications of 2,3,7,8-Tetrachloro-
    dibenzodioxin  (TCDD)  Contamination in Residential Soil," Center
    for Environmental Health, Centers for Disease Control, 1983.
12.  U.S.  EPA, "Interim Evaluation of Health Risks Associated with
    Emissions of Tetrachlorinated Dioxins from Municipal Waste Re-
    source Recovery Facilities," Office of the Deputy  Administrator,
    U.S. EPA, Nov. 1981.
13.  Brunekreef, B., et al.,  "Blood Lead Levels of Dutch City Children
    and Their Relationship  to  Lead in the Environment," JAPCA,  33,
    1983, 872-876.
                                                                                                          RISK ASSESSMENT
                                                                                                                                  335

-------
                  Improving the  Risk  Relevance of Systems  for
                               Assessing  the Relative Hazard
                                     of Contaminated  Sites

                                                 Ellen D. Smith
                                      Lawrence W. Barnthouse, Ph.D.
                                          Glenn W. Suter II, Ph.D.
                                            James E. Breck,  Ph.D.
                                       Oak  Ridge National Laboratory
                                      Environmental Sciences Division
                                             Oak Ridge, Tennessee
                                                Troyce D. Jones
                                       Oak  Ridge National Laboratory
                                    Health and Safety  Research Division
                                             Oak Ridge, Tennessee
                                           Dee Ann Sanders, Ph.D.
                  U.S. Air Force Occupational and Environmental Health Laboratory
                                        Brooks Air Force Base,  Texas
ABSTRACT
  The  ranking  system  (Hazard Assessment Ranking  Meth-
odology, or HARM) used to evaluate the relative hazard of con-
taminated sites during the records-search phase of the U.S. Air
Force Installation Restoration Program (IRP) has been revised to
priorities for later phases of the IRP. The resulting methodology,
HARM II, is designed to use data obtained in preliminary site in-
vestigations and is expected to provide better approximations of
relative site risk than could be obtained with the original HARM
or with the U.S. EPA's Hazard Ranking System. This paper
describes the  methodological  enhancements in HARM II  and
reports the results of preliminary testing of HARM II.

INTRODUCTION
  The U.S. EPA "Superfund" program and other federal and
state waste site remedial action programs set remedial action
priorities using various  methodologies that  rank contaminated
sites according to their relative hazard. Ideally, the site rankings
produced by these methodologies would closely correlate with the
actual health and ecological risks posed by individual sites;  i.e.,
the relative scores would  indicate relative potential for adverse ef-
fects, taking into account the relative probabilities of exposures to
contaminants  and  the relative consequences of such exposures.
We call this property "risk relevance."
  While risk relevance was a critical consideration in the design of
such widely used site-ranking systems as the U.S. EPA Hazard
Ranking System  (HRS)'  used in the  Superfund  program,
methodological simplifications needed to permit scoring of large
numbers of sites in a short time with limited site-specific data
often diminish  the risk relevance of the  resulting scores.
Simplified ranking systems are appropriate for preliminary iden-
tification of sites where cleanup may be necessary, but final deci-
sions on allocating remedial action resources among sites should
use methods  that relate more  directly to  actual health  and
ecological risks.
1. Research sponsored by the U.S. Air Force, under Imeragency Agrcemcnl No.
40-1413-83 with the U.S. Department of Energy, under Contract No. DE-AC05-
840R21400 with Martin Marietta Energy Systems, Inc.
  In this paper,  the  authors describe a modification of the
Hazard  Assessment  Rating Methodology  (HARM),  a site-
screening system developed by the U.S. Air Force for use in the
initial phase of its Installation Restoration Program (IRP). The
Air Force used HARM to evaluate sites of suspected contamina-
tion based on information obtained from a records search. The
new  site-rating  methodology  (Hazard Assessment  Rating
Methodology II, or HARM II) is intended to set  priorities for
detailed  site investigation and possible remedial action in later
phases of the IRP. It uses data obtained from preliminary field
surveys of sites identified during the  records  search as being
potentially hazardous. The primary  reasons for  modifying
HARM were to permit effective use of new field data and to im-
prove the methodology's ability to discriminate between sites on
the basis of actual risk. Although HARM is designed specifically
to meet  the  needs of the Air Force, features incorporated in
HARM II to improve its risk relevance may  be useful in other
decision-making problems involving hazardous materials  and
waste sites.

THE HAZARD ASSESSMENT
RATING METHODOLOGY
  The original HARM methodology is a close relative of the
HRS.' As in the HRS, sites are scored on the basis of potential for
contaminant  transport ("Pathways"), contaminant character-
istics ("Hazard") and potential receptors ("Receptors") along
each of several possible exposure pathways. Subscores for these
three  risk  categories are  obtained  by evaluating  the sites with
respect to several rating factors, multiplying factor scores by ap-
propriate factor weights and summing the weighted factor scores.
The resulting subscores then are averaged to arrive at a single
overall site score.
  Only hydrologic transport and exposure pathways are con-
sidered in HARM;-air exposures are not included.  Other dif-
ferences between HARM and HRS  include the use of standard
four-point (i.e., 0-1-2-3) factor rating scales throughout HARM,
inclusion of a few different rating factors (e.g., distance from the
site to a facility boundary) appropriate to Air Force applications,
assignment of different ranges of values (e.g., of local population
336   RISK ASSESSMENT

-------
                            Receptors
                                                                    Waste Characteristics
       START
DETERMINE
WASTE
QUANTITY/
HAZARD SCORE

"*"
APPLY
PERSISTENCE
FACTOR


APPLY
PHYSICAL
STATE
FACTOR



CALCULATE
WASTE CHAR.
SUB8CORE



                                                                                 Waste_Managernent
                                                                                      Practices
                                                             Figure 1
                                        Hazardous Assessment Rating Methodology Flow Chart
and waste quantity) to individual factor scores and use of additive
algorithms (instead of the multiplicative and root-mean-square
algorithms employed  in HRS) to combine subscores. Figure 1 il-
lustrates the structure of the original HARM.2
  The  structure of HARM II (Fig. 2) is  significantly different
from that of the original HARM. In the original HARM (Fig. 1),
three contaminant  migration routes (Surface Water, Flooding
and Groundwater) are evaluated in determining the subscore for
the Pathways category, but a single subscore was determined for
each of the three categories (i.e., Pathways, Hazard and Recep-
tors) regardless of the mode of contaminant migration. To ensure
that the factors considered in Receptors scoring are relevant to the
expected mode(s) of  contaminant migration and  that Hazard
scores reflect the  sensitivity of the exposed populations, HARM
            II requires separate calculation of scores for each  of four ex-
            posure pathways:  (1) Surface Water Migration/Human Health
            Receptors, (2)  Surface Water Migration/Ecological Receptors,
            (3) Groundwater Migration/Human Health Receptors,  and (4)
            Groundwater  Migration/Ecological  Receptors.  Subscores  for
            each of the three rating categories are aggregated to arrive at an
            overall pathway score, and the four pathway scores then are com-
            bined to  arrive at a single overall  site score. In HARM II, the
            Flooding migration route from the original HARM has been in-
            corporated into the  Surface Water migration  route, and  the
            Waste Containment factor from the original HARM  has been in-
            corporated into the Pathways scoring category. Other features of
            HARM II intended to  improve its risk relevance are described
            below.
                                  CALCULATE SW/HEALTH
                                   HAZARD SUBSCORE
                                      (SECT. 3.2.1)
CALCULATE SURFACE
WATER PATHWAY
SUBSCORE
(SECT. 3.1)


	 1






CALCULATE SW/ECOLOOICAL
ISECT. 322}
CALCULATE SW/HEALTH
 RECEPTOR SUBSCORE
    (SECT. 3 Jl
CALCULATE SW/HEALTH
  PATHWAY SCORE
    ISECT. 3.41
                                                       CALCULATE SW/ECOLOCICAL
                                                         RECEPTOR SUBSCORE
                                                            ISECT. 3.3)
                     CALCULATE SW/ECOLOO.ICAL
                        PATHWAY SCORE
                          ISECT. 3.4)
CALCULATE GROUND
WATER PATHWAY
SUBSCORE
ISECT. 3.11


	 1






CALCULATE QW/ECOLOOICAL
ISECT. 3.2.21
CALCULATE OW/HEALTH
HAZARD SUBSCORE
ISECT. 3J.1I


CALCULATE GW/HEALTH
RECEPTOR SUBSCORE
ISECT. 3.3)
                                                                               CALCULATE OW/HEALTH
                                                                                 PATHWAY SCORE
                                                                                   (SECT. 3.4)
                                                       CALCULATE OW/ECOLOOICAL
                                                         RECEPTOR SUBSCORE
                                                            ISECT. 3.3)
                     CALCULATE QW/ECOLOOICAL
                         PATHWAY SCORE
                          ISECT. 3.4)
                                                             Figure 2
                                             Flow Chart for HARM II Site Rating System
                                                                                                        RISK ASSESSMENT    337

-------
HARM II Contaminant Hazard Scoring
  The most important limitation in the contaminant hazard scor-
ing portion of the original HARM is the method's reliance on the
Sax Toxicity  Rating.5 The Sax method, which is widely  used in
site-ranking methodologies such  as  the  HRS,  assigns  toxicity
scores  on the  basis  of  a  substance's observed  biological
mechanism of action, independent of the  dose level at which the
substance is toxic. Also, Sax  ratings  do not indicate hazards to
nonhuman ecological receptors and sometimes are not relevant to
the modes of exposure that would be expected along environmen-
tal exposure pathways (e.g., a chemical whose fumes are highly
toxic  may not be very toxic when  ingested in drinking water). A
second problem with HARM that is common to most site-ranking
methodologies is the use of only the most  toxic chemical at a site
to characterize  the site  hazard. When  a  site's hazard rating is
based only on the most toxic  chemical, it is possible to assign a
high score to a site having a large quantity of low-hazard wastes
solely due to the presence of a  very small quantity of high-toxicity
wastes.
  The risk relevance of HARM II contaminant hazard scores has
been enhanced by replacing the Sax rating with  a set of toxico-
logical benchmarks and by scoring hazard on the basis of com-
parisons between  the benchmarks and measured  contaminant
levels. Separate sets of benchmarks are used, and separate scores
are calculated for human-health hazards and ecological hazards.
As in HARM, the score for site contaminant hazard is obtained
after  applying   contaminant  persistence and   waste  quantity
multipliers to the toxicity score.
  Human-health  hazard scores  are based on ingestion of con-
taminated water and consumption of  fish  that have accumulated
waterborne contaminants.  Ecological  hazard scores are based on
acute lethality to fish and aquatic invertebrates and effects on ter-
restrial plant  production due to contamination of growth media.
If data are available on contaminant concentrations  in  surface
waters  or groundwaters,  these  concentrations  are  used  to
calculate hazard quotients, which  are ratios of the estimated in-
take rates of the detected contaminants  to the toxicity bench-
marks of the contaminants. The specific formula used depends on
the existing or potential uses of the contaminated water;  the
human-health hazard contribution of a contaminant found in sur-
face  water that supplies drinking  water would  be calculated as
follows:
  Q = (Cw x Fw + Cr x Ff)/Bh
                                                        (1)
where:
   Q

  Cw
  Fw
  Cf
  Bh  =
         human-health hazard quotient for the contaminant of
         interest
         contaminant concentration in surface water 0*g/l)
         estimated individual drinking water intake (2 I/day)
         concentration in fish 0*g/g) = Cw x Bb, where Bb
         bioaccumulation factor (1/g) for the contaminant of
         interest
         estimated mean daily individual fish consumption
         (6.5 g/day)
         health effects benchmark (>ig/day) for the contaminant
         of interest
Where more than one contaminant has been detected, quotients
for the individual contaminants are summed. Hazard scores are
assigned on a scale of 0 to 6, based on the order of magnitude of
the sum of quotients.
  Procedures have  been specified' for obtaining contaminant-
specific benchmarks for  effects on humans, effects on aquatic
and terrestrial biota and potential bioaccumulation in fish.  For
regulated chemicals, the human-health toxicity benchmarks used
in HARM II are based on the U.S. EPA's drinking water quality
standards (40 CFR 141) or on the "permissible concentrations"
promulgated by the National Institute for Occupational  Safety
and Health. For unregulated contaminants, benchmarks are ob-
tained by evaluating test data available in such published compila-
tions  as the Registry  of Toxic  Effects of Chemical Substances
(RTECS),' using the relative potency technique described by
Jones, el at.' Similarly, benchmarks for ecological effects are ob-
tained from U.S. EPA water quality criteria for protection of
aquatic life or from the U.S. EPA's on-line data bases on toxicity
to aquatic  organisms  (QUIRE) and terrestrial plants (PHYTO-
TOX). Bioaccumulation factors for estimating the contribution
of contaminated food  chains to human intake, are obtained from
published compilations,710 or estimated from structure-activity
relationships, e.g., from the octanol-water partition coefficient.
  If contaminant migration has not been detected along  a par-
ticular migration route  or  if water  concentration  data are
unavailable, hazard scores are based on the characteristics of the
most toxic contaminant. A score from 0 to 6 is assigned, based on
the toxicological benchmarks of the most toxic contaminant. For
this case, the scale used in scoring  human-health hazards is based
on the range of potencies of 60 chemicals drawn at random from
RTECS,' and the scale used in scoring ecological hazards is based
on the range of toxicities for chemicals listed in the AQUIRE data
base.
  The calculations and data assessments required to implement
this portion of HARM II  are significantly more complex and
time-consuming than those involved in either the original HARM
or the HRS. However, because it is unlikely that HARM II users
will have the expertise needed  to determine appropriate contami-
nant benchmarks, Oak Ridge National Laboratory is determining
benchmarks for all nonradioactive contaminants that have been-
found at Air Force facilities. Their results will be published in a
user's manual for HARM II.  Given this tabulated information,
determination  of HARM  II  contaminant hazard  scores can
become a routine microcomputer spreadsheet application.

HARM II Pathways Scoring
  The "Pathways" portion of HARM  II rates the potential for
contaminants from  a waste  site to  enter surface waters via
overland flow routes or to enter groundwater. As in the HRS and
the original HARM,  if contaminants from a site have already
been detected in water, the maximum possible pathway score is
assigned. Thus,  the Pathways scoring procedures are applied only
if contamination has not yet been  detected.
  Factors in the Pathways scoring category in HARM II are listed
in Table 1.  The  HARM II surface  water pathways scoring system
is very similar to that used in HARM, with the notable exception
that the potential for floodwater transport of contaminants is in-
cluded in the surface  water pathway in HARM II. This change
was made to reflect the fact that floodwater transport is only one
of several routes for overland transport of contaminants to sur-
face waters, and factors affecting risk from floodwater transport
are similar  to those affecting risk  from other overland transport
routes. A few factor weights were adjusted to better reflect their
relative  importance to overland transport or overflow from im-
poundments.
  The scoring system  for the groundwater pathway was modified
slightly from HARM to make better use of the new data obtained
in site investigations.  The original HARM included two factors
related to the length of the pathway between the waste and the
groundwater: (1) depth to groundwater from the land surface and
(2) potential for subsurface flows (i.e., the water table) to in-
tersect the base of the waste. HARM II replaces these with  a
single new  factor: depth to seasonal high groundwater from the
base of the waste or contaminated zone. This new scoring factor
338
       RISK ASSESSMENT

-------
                             Table 1
     Listing of HARM II Pathways and Receptors Scoring Factors
     Category
                                Scoring Factors
              Surface water migration
                                      Groundwater migration
  PATHWAYS
  HtCEPTUKS
   Human health
              Distance to nearest surface
                Mater
              Net precipitation
              Surface erosion potential
              Rainfall intensity
              Surface penneability
              Flooding potential
              Containment effectiveness
                        Depth to groundwater from oase
                          of Maste or contaminated zone
                        Permeability of unsaturated zone
                        Potential for discrete features
                          to "short circuit" pathway to
                          the water table
                        Infiltration potential
                        Containment effectiveness
              Population serveo by surface
                water supplies within 3 miles
                (4.8 km) downstream
              Hater quality classification
                of nearest surface water
              Population within 1000 ft
                (305 m) of site
              Distance to nearest
                installation boundary
              Land use and zoning within 1
                mile (1.6 km) of site
   Ecological
                        Mean groundwater travel time to
                          nearest downgradient well
                        Groundwater use of uppermost
                          aquifer
                        Population served by affected
                          downgradient aquifers within 3
                          miles (4.8 km)
                        Population served by affected
                          aquifers not in downgradient
                          direction within 3 miles
                          (4.8 km)
                        Mean groundwater travel time to
                          downgradient surface waters
                          that supply domestic uses or
                          food-chain agriculture
                        Population within 1000 ft
                          (305 m) of site
                        Distance to nearest installation
                          boundary

Importance/sensitivity of      Mean groundwater travel time to
 biota/habitats in potentially    downgradient habitat or
 affected surface waters        natural area
Presence of "critical         Importance/sensitivity of
 environments" within 1 mile     downgradient habitats/natural
 (1.6 km)                   areas
                        Presence of "critical
                          environments" within 1 mile
                          (1.6 km)
 uses information that may not have been available before site in-
 vestigation, and it should enhance the ability of the method to
 identify sites where there is a real potential for contamination of
 groundwater. Another feature of this portion of HARM II that
 represents a change from the original HARM is the addition of a
 factor  called  "infiltration  potential,"  which  is based  on  an
 estimate of the amount of water available to cause waste leaching
 or infiltration. Infiltration potential is determined as a function of
 net precipitation (a Pathways scoring factor included in the waste
 hazard score in the original HARM). This new scoring factor is
 expected to be a better indicator of the hydraulic potential for
 waste constituents to reach groundwater.
   The waste containment effectiveness factor, which was applied
 to the overall site score in the original HARM, is  applied only to
 the Pathways scores  in HARM  II, because effective waste con-
 tainment reduces the  potential for contaminants to enter water
 but does not affect the other major risk categories considered in
 HARM. HARM II includes separate rating scales for waste con-
 tainment  effectiveness  for  surface  water and  groundwater
 pathways, permitting separate evaluation of factors limiting sur-
 face water or groundwater transport. The HARM II rating scales
 for containment effectiveness were derived from  the "waste
 management factors" in HARM2 and the "containment values"
 intheHRS.1

 HARM II Receptor Scoring
   The "Receptors" portion of HARM II evaluates the potential
 for human populations and  ecosystems to  be  exposed to con-
 taminants present in surface waters or groundwaters. This portion
 of HARM II  has been modified significantly from the original
 HARM, primarily to permit use of groundwater information ob-
 tained through site  investigation.
   Factors in the Receptors scoring category in HARM II are listed
in Table 1. The scoring factors  for human  health receptors af-
 fected by surface water migration are essentially unchanged from
HARM. The dominant factors in this category are those related
 to the  potential  for  humans  to ingest contaminated surface
 waters,  either as drinking water or in foods. Three other factors
 related to the site's proximity to human populations (i.e., popula-
 tion within 1,000 ft [305 m], distance to installation boundary and
 land use and zoning) are included because they indicate a poten-
 tial for  humans to come into contact  with contaminated waters
 through routes other  than  ingestion  (e.g.,  direct contact  by
 children playing near contaminated drainage ditches).
   The scoring  factors for human receptors affected by ground-
 water migration are significantly different from those used in the
 original HARM. This portion of HARM, like the analogous por-
 tion of HRS, is concerned primarily with distances to water sup-
 ply wells and does not consider the direction or rate of ground-
 water flow. Initial  field investigations should have supplied the in-
 formation needed to estimate groundwater velocities  and  flow
 directions,  so  HARM  II  receptors scores can  be based on
 estimates of the actual potential for contaminants to reach points
 of groundwater use. HARM  II  assigns more weight to receptors
 that are found to  be hydraulically downgradient from the waste
 site and incorporates a measure  of the urgency of a particular
 contamination  problem by including  scoring  factors based on
 estimated groundwater travel times to  sites  of groundwater use.
 Also, because surface  waters are often outlets for groundwater
 discharge, the human health receptors  category for groundwater
 migration includes scoring factors related to the uses and suscep-
 tibility of surface  waters hydraulically downgradient from con-
 taminated sites. These features of HARM II are intended to en-
 sure that scores are  based on consideration of the populations
 that are  actually at risk.
   The groundwater travel time parameter in HARM II is similar
 to  the travel time criterion adopted by  federal  regulators as a
 measure  of the  isolation  of  radioactive  wastes  in  geologic
 repositories.11'12 Mean groundwater travel time is calculated from
 the estimated hydraulic conductivity of the affected aquifer (k),
 the hydraulic gradient  in the affected aquifer (i), the distance to
 the well  (d) and the effective porosity of the affected aquifer (n),
using the following equation:
                                                            Travel time = dn/ki
                                                                                                                    (2)
                                                       This calculation will provide an estimate of the mean travel time
                                                       of a nonreactive contaminant,  neglecting  the effects  of disper-
                                                       sion. It is recognized that the data needed for this calculation will
                                                       not always be available. To promote consistency in evaluating this
                                                       parameter,  guidance  has been  developed"  for estimating  input
                                                       values when appropriate data are unavailable. For consistency in
                                                       applying HMAR II, "downgradient" is defined to mean a 90- to
                                                       120-degree  arc  containing  the best  estimate of the direction  of
                                                       groundwater flow at  its center. Where the  direction of ground-
                                                       water  flow  is highly  uncertain  or indeterminate,  all travel-time
                                                       scoring is based on an estimated travel time to the nearest well or
                                                       surface water in any direction.
                                                         The method  for scoring the  potential exposure of  ecological
                                                       receptors has been expanded  from  the  original HARM and  the
                                                       HRS,  both  of which emphasize the  proximity of the site  to
                                                       various critical environments.  HARM II also considers  the poten-
                                                       tial ecological importance or sensitivity or ecosystems likely to be
                                                       exposed to  a particular transport pathway but not necessarily
                                                       "critical environments." These  include all permanent  streams,
                                                       lakes,  wetlands and irrigated land areas.

                                                       Aggregation of HARM II Subscores
                                                         In the original HARM (Fig. 1), the overall site score is obtained
                                                       by simple arithmetic averaging  of  the Pathways,  Hazard and
                                                       Receptors subscores. This method of aggregating scores is simple
                                                       to apply, but its results are not well correlated with actual risks of
                                                                                                          RISK ASSESSMENT    339

-------
toxic exposures. Three factors must be present to result in an ex-
posure risk: toxic contaminants ("Hazard"), potential for migra-
tion ("Pathways") and a potentially exposed human population
or ecological resource ("Receptors"). If an additive algorithm
(such as arithmetic averaging) is used to combine subscores for an
exposure pathway, it is possible to obtain a relatively high score
for a pathway along which only two of these three factors are pre-
sent and for which risk is therefore rather low. For example, a site
with highly toxic contaminants and  a large receptor population
but very little potential  for aqueous transport of contaminants
poses relatively little risk, but could receive a high score. To avoid
this outcome, HARM II uses a multiplicative algorithm similar to
that in the HRS' to combine subscores into a  single pathway score.
The three subscores are multiplied together, and,  to normalize to
a 100-point scale, the product is divided by 10* (the product of the
highest possible category scores). With this algorithm,  an ex-
posure pathway with a very low or zero subscore for Pathways,
Receptors or Hazard will receive a very low  or zero score, consis-
tent with the low pathway risk.
  Several different approaches were available for obtaining  an
overall  site score  from  the four HARM  II exposure pathway
scores.  Arithmetic averaging was  not  used  for  this purpose
because this method assigns equal weight to high-  and low-risk ex-
posure pathways. A site with a high risk via one exposure pathway
is a high-risk site even if it  has a low risk  via another exposure
pathway.  Although such a site should receive a higher score than
a site with a moderate risk via both pathways, the two sites might
receive  the same  score if  the  pathway scores  were averaged
arithmetically. The objective of assigning greater weight to high-
risk pathways could be achieved by setting the overall site score
equal to the single highest exposure  pathway score, but this ap-
proach is  inappropriate because it does not  reflect the fact that a
site with high risks along two or more exposure pathways presents
greater risk than a site with high risk along  only one pathway.
  A weighted root-mean-square algorithm was selected for com-
bining the four HARM II exposure pathway scores  into a single
overall site score.  The HRS uses a root-mean-square algorithm
for combining exposure pathway scores because this approach
reflects the contributions of several pathways while it gives greater
weight to pathways with higher scores." Explicit factor weights
are also applied in combining HARM II pathway scores in order
to give  human health risks greater weight than ecological risks.
Thus,  the following algorithm is used to  obtain the final site
score:

  Sf = [5(S,ih)2 + (Ss,e)2 + 5(Sg,h)2  + (Sg,e)2] 1/2/3.464      (3)
where
    Sf   =  final score
  Ss h   =  score for human health risk via surface water pathway
  SSie   =  score for ecological risk via surface water pathway
  Sg h   =  score for human health risk via groundwater pathway
  Sg e   =   score for ecological risk via groundwater  pathway
The weight of 5 for the human health scores was selected because
the  original HARM assigned factors related  to human receptors 5
times the weight given  to factors related  to  ecological receptors.
The divisor 3.464 is the square root of 12, which is the sum of the
weights assigned to individual pathway scores.

TESTING OF HARM  II
  As a preliminary test of the methodology, HARM II was ap-
plied to a  representative Air Force  facility  that  contained nine
contaminated sites for  which both preliminary site investigation
data and original HARM scores  were available.' The nine sites
that were  assessed represented a  diverse  range  of  waste
characteristics,  waste quantities  and physical settings.  As in-
                                                                 dicated by the scores  in Table  2,  HARM  II produced a clear
                                                                 separation between a group of high-scoring sites (i.e., the 46-55
                                                                 range) and a group of  relatively  low-scoring sites (i.e., the 15-30
                                                                 range), whereas scores  from the original HARM had been spread
                                                                 fairly evening over a range  from 51 to  88. This improved score
                                                                 separation, which is largely due to the change from an additive to
                                                                 a multiplicative algorithm, should be helpful  in setting  and
                                                                 defending remedial action priorities. In general, scores assigned
                                                                 with HARM II were lower than  those assigned with the original
                                                                 HARM, largely because the change to a multiplicative algorithm
                                                                 resulted in significantly lower overall scores for sites that had low
                                                                 subscores  in one of the three rating categories. Since the HARM
                                                                 systems are tools for evaluating relative site hazards and are not
                                                                 correlated with absolute degrees of risk,  the lower range of scores
                                                                 obtained with HARM II does not mean that the sites pose any less
                                                                 risk now than they did  when the original HARM was applied.
                                                                   HARM  II  also produced a different ranking of sites  than the
                                                                 original HARM. The most dramatic changes in site rank were at-
                                                                 tributable  to new information obtained during site investigations.
                                                                 For example, in applying the original HARM, the chrome waste
                                                                 pit (Table  2) had been scored as having a small quantity of waste
                                                                 and a low potential for contaminant migration,  but it later was
                                                                 found  to contain a large quantity of waste and to have caused
                                                                 significant  groundwater  contamination. A variety of smaller
                                                                 changes in scores and site ranks could be attributed to new data,
                                                                 changes in individual  scoring factors  and  changes in scoring
                                                                 algorithms.4

                                                                                            Table 2
                                                                             Comparison of Scores for Sites Evaluated
                                                                                With Both HARM and HARM II
                                                                                                             KMH II score
                                                                                                               Una rinft
Undf 111 r 1
Ldndttll fj
tusle ol 1 pits
01(1 fuel Stor«g« ttnk 1 1 IV
Die ytrQ cheated! pits
f ire-lrttnlng Durn pit »6
Chroac M4SIC pit
Two other f Ire-lrainlnq (Kirn pill*
SS
«,
It
n
M
ss
55
si-se
ID
HI
(11
l«)
(5)
(61
18)
(6-9)
47
SS
«6
H
54
IS
54
11-30
(«)
(>>
(S)
(7)
(2)
(9)
U)
(6-8)
                                                                  Individual burn pus rated *ilh HARM ft could nol be directly correlated with specific burn pits
                                                                 rated with HARM
                                                                   This testing indicated that it is feasible to apply HARM II with
                                                                 the types of data available for contaminated sites on Air Force
                                                                 facilities and gave a preliminary indication that the method will
                                                                 provide meaningful distinctions between high- and low-risk sites.
                                                                 More extensive test applications and sensitivity analyses are now
                                                                 being performed to confirm these observations, to evaluate the
                                                                 methodology's effectiveness for comparing a wide variety of sites
                                                                 in  different  geographical  areas and  to  identify any  needed
                                                                 refinements in HARM II.

                                                                 CONCLUSIONS
                                                                   HARM  II is  expected to  provide  a  better approximation of
                                                                 relative risk than has been available previously in scoring systems
                                                                 suitable for wide application in setting remedial action priorities.
                                                                 Although HARM H was designed specifically to meet the needs of
                                                                 the Air Force, the concepts employed in  HARM II should find
                                                                 application in other decision-making problems involving hazar-
                                                                 dous materials and waste sites. Important  innovations in HARM
                                                                 II include estimation of contaminant hazard using actual risk ap-
                                                                 proximations and the use of field data in  determining the actual
                                                                 susceptibility of groundwater receptors.  Like the original HARM,
                                                                 the HRS and other similar systems, HARM 11 is not a formal risk-
                                                                 assessment methodology but is designed  for routine application to
340
       RISK ASSESSMENT

-------
large numbers of sites. Although the method is expected to pro-
duce reasonable approximations of relative risks, it is not possible
to objectively relate the numerical scores to absolute levels of
potential health and environmental hazard.
  Additional changes could be made in HARM II  to  produce
scores that would more closely resemble risk-assessment results.
For example, the Receptors scoring categories might be expanded
to include factors related to surface water dilution, aquifer disper-
sion or contaminant sorption in aquifers. However,  most such
refinements would imply greater data availability and reliability
than  can  be  realistically expected,  and any  potential  im-
provements in scores do not appear to justify the additional effort
that would be required to implement the system.
REFERENCES
 1. U.S. EPA, "National Oil and Hazardous Substances Contingency
    Plan," Appendix A: "Uncontrolled Hazardous Waste Site Ranking
    System: A User's Manual," Federal Register 47, 1982, 31219-31227.
 2. Engineering-Science, Inc., "Installation Restoration Program Phase
    I: Records Search, Tinker AFB, Oklahoma," Atlanta, GA, 1982.
 3. Sax, N.I., Dangerous  Properties of Industrial Materials, 5th  ed.,
    Van Nostrand Reinhold Co., New York, NY, 1979.
 4. Barnthouse, L.W., Breck, J.E., Jones, T.D., Kraemer, S.R., Smith,
    E.D. and Suter, G.W., II, "Development and Demonstration of a
    Hazard Assessment Rating Methodology for Phase II of the Installa-
    tion Restoration Program,"  ORNL/TM-9857, Oak Ridge National
    Laboratory, Oak Ridge, TN, in press.
 5. Lewis, R.J., Sr.  and Tatken, R.L., "Registry of Toxic Effects of
    Chemical Substances," U.S. Department of Health and Human
    Services, Washington,  DC, 1982.
 6.  Jones, T.D., Griffin, G.D. and Walsh, P.J., "A Unifying Concept
    for Carcinogenic  Risk  Assessments," J. Theor.  Biol. 105, 1983,
    35-61.
 7.  Callahan, M.A., Slimak, M.W., Gabel, N.W., May, I.P., Fowler,
    C.F., Freed,  J.R., Jennings, P., Durfee,  R.L., Whitmore, F.C.,
    Maestri,  B., Mabey, W.R., Holt, B.R. and Gould, C., "Water-Re-
    lated Environmental Fate of 129 Priority Pollutants. Vol. I: "Intro-
    duction and Technical Background, Metals and Organics, Pesticides
    and PCBs,"  EPA-440/4-79-029a, Office of Water Planning and
    Standards, Office of Water and Waste Management, U.S.  EPA,
    Washington, DC,  1979.
 8.  Callahan, M.A., Slimak, M.W., Gabel, N.W., May, I.P., Fowler,
    C.F., Freed,  J.R., Jennings, P., Durfee,  R.L., Whitmore, F.C.,
    Maestri,  B., Mabey, W.R., Holt,  B.R. and Gould, C.,  "Water-
    Related  Environmental Fate of 129 Priority Pollutants." Vol.  II:
    "Halogenated Aliphatic Hydrocarbons, Halogenated Ethers, Mono-
    cyclic Aromatics,  Phthalate Esters, Polycyclic Aromatic Hydrocar-
    bons, Nitrosamines, and Miscellaneous Compounds," EPA-440/4-
    79-029b,  Office of Water Planning and Standards,  Office of Water
    and Waste Management, U.S. EPA, Washington, DC, 1979.
 9.  Trabalka, J.R. and Garten, C.T., Jr., "Development of Predictive
    Models for Xenobiotic Bioaccumulation in Terrestrial Ecosystems,"
    ORNL-5869, Oak Ridge National Laboratory, Oak Ridge, TN, 1982.
10.  U.S. EPA, "Ambient Water Quality Criteria for Phthalate Esters,"
    EPA 440/5-80-067, 1980.
11.  U.S.  EPA, "Environmental Standards  for the Management and
    Disposal  of Spent Nuclear Fuel,  High-Level and Transuranic Radio-
    active Wastes (40 CFR 191)," Federal Register 50, 1985.
12.  U.S.  Nuclear  Regulatory Commission,  "Disposal of High-Level
    Radioactive Wastes in Geologic  Repositories (10 CFR 60)," Federal
    Register 48, 1983,  28222-28229.
13.  Caldwell, S., Barrett, K.W. and  Chang, S.S., "Ranking Systems for
    Releases  of Hazardous  Substances," Proc. of the First National
    Conference on Management of Uncontrolled Hazardous Waste
    Sites, Washington, DC,  Oct., 1981, 14-20.
                                                                                                         RISK ASSESSMENT    341

-------
                                     Soil Cleanup Criteria for
                                  Selected Petroleum Products

                                                 Sofia K.  Stokman
                                               Richard Dime, Ph.D.
                            New Jersey Department of Environmental Protection
                                   Hazardous Site Mitigation  Administration
                                               Trenton, New Jersey
ABSTRACT
  A large number of sites are contaminated with petroleum prod-
ucts which may pose a potential health hazard. Therefore, risk
assessment and acceptable soil cleanup levels for these products
are urgently  needed.  This  paper  compares alternatives  for
developing  cleanup objectives  for soil  contaminated  with
petroleum products. A soil cleanup objective based on the con-
centrations of total petroleum hydrocarbons versus a soil cleanup
objective based on the concentration of individual chemical con-
stituents of concern is investigated.
  A review of the chemical composition of crude oil, gasolines,
petroleum naphtha,  fuel oil No. 1 Get fuel, kerosene),  fuel oil No.
2 (heating oil, diesel oil), fuel oil No. 6 (Bunker "C"), lubricating
oil, fresh motor oil and used motor oils (over 5,000 and 10,000 km
usage) was undertaken  to  identify  individual chemical consti-
tuents that would be harmful to public health and/or the environ-
ment.  The  carcinogenic  polycyclic  aromatic  hydrocarbons
(CaPAHs) and benzene were identified as constituents of con-
cern. A lifetime soil exposure model based on pica tendencies and
inhalation of soil was used to estimate soil concentrations of these
constituents that would  result in a one in  a million (1 x 10 ~6)
cancer risk if exposure were to occur. These levels were compared
with residual concentrations of CaPAHs and benzene in soil after
cleanup to 110 ppm total petroleum hydrocarbons (PHs) for each
petroleum product studied. For those constituents remaining in
concentrations which would exceed a 1 x 10 ~6 if exposure were
to occur, a soil cleanup objective which reduces the health risk to
an  acceptable level  was  proposed.  In addition,  the proposed
cleanup objectives were compared to "normal" soil background
levels  of CaPAHs and benzene.
  With the exception of motor oils used over  10,000 km, using
100 ppm of total  PHs as the soil cleanup  objective results in
residual soil levels of CaPAHs and benzene not yielding a 1  x
10~6 cancer risk. Changing the soil  cleanup objective from 100
ppm to 60 ppm  will reduce CaPAHs from used motor oils over
10,000 km to an acceptable cancer risk level of one in a million.
"Normal" background soil levels of benzene are exceeded in  all
products for which data were available.
  Determination of soil cleanup objectives based on residual soil
levels  of CaPAHs and benzene cannot be further refined at this
time due to the limited information available on the concentration
of these compounds in petroleum products and in "normal" soil
background levels.

INTRODUCTION
  Large-scale terrestrial pollution can result from the production,
transportation, storage and voluntary land disposal of petroleum
products. Exposure to these products may result in adverse effects
on public health and on the environment. This paper examines
alternatives for developing acceptable soil cleanup levels (ASCLs)
for soil contaminated with petroleum products. The focus is  on
exposure to contaminated soils through ingestion (pica) and in-
halation and  on the migration of  petroleum compounds into
groundwater. The literature was reviewed to identify constituents
in  petroleum products that are toxic and/or harmful in the en-
vironment. ASCLs were determined for carcinogenic constituents
using  risk assessment  techniques.  Soil cleanup levels for car-
cinogenic constituents were compared to a total petroleum hydro-
carbons (PHs) cleanup objective of 100 ppm.

DETERMINATION OF ACCEPTABLE
SOIL  CONTAMINANT LEVELS
  Risk assessment and soil cleanup objectives are presented only
for individual chemical constituents of petroleum products which
have the highest toxicity, the ability to migrate and/or are present
in significant amounts. Emphasis was placed on chronic toxic ef-
fects (particularly carcinogenic) due to long-term exposure to con-
taminated soils. In  order to accomplish this task, the following
was performed: (1) review of physical/chemical properties, health
effects and environmental fate of selected petroleum products; (2)
identification of carcinogenic constituents and estimation of con-
centrations present in each particular petroleum product; (3) esti-
mates of residual soil concentrations of carcinogenic constituents
when  100 ppm total petroleum hydrocarbons is used as a cleanup
objective; and  (4) comparison of estimates with concentrations
resulting in a cancer risk of one in a million (1  x 10-6) and with
"normal" background soil levels to determine if any of the car-
cinogenic constituents  remain in the soil in concentrations that
may pose a threat to public health.  Comparison to background
also was made since many regulatory agencies consider cleanup to
background an appropriate remedial objective.
  Due to space limitations, all areas reviewed are not covered in
detail. This paper discusses briefly the toxicity and environmental
fate of selected petroleum  products.

IDENTIFICATION AND TOXICITY OF
CONSTITUENTS IN PETROLEUM  PRODUCTS
  The following petroleum products are reviewed in this  paper
for identification of toxic constituents:  crude oil, petroleum
naphtha, gasolines,  fuel oil No.  1 (jet fuel, kerosene), fuel oil No.
2 (heating oil, diesel oil), fuel oil No. 6 (Bunker "C"), lubricating
oil, fresh motor oil, and used motor oils (over 5,000 and 10,000
km usage).
  A review  of the literature indicates that the most  toxic consti-
tuents of petroleum  products  are  the  aromatics.  After the
aromatic fractions, toxicity  decreases  from  olefins through
naphthanes to paraffins.1  Within each group, the hydrocarbons
of lower molecular weight tend to be more toxic.1  Exposure to
high   and  low  boiling fractions of  petroleum  (crude  oil)  is
associated with carcinogenic effects. Moreover, exposure to poly-
cyclic  aromatic  hydrocarbons  (PAHs) present  in certain  heavy
342   RISK ASSESSMENT

-------
residual oils2 and to benzene present in unleaded gasoline3 cur-
rently is being investigated for carcinogenic effects of these prod-
ucts. Crude oil is mutagenic4 and is at the present time under toxic-
ity review.5 Kerosene currently is being tested for carcinogenicity.6
  This paper primarily is concerned with chronic effects due to
long-term exposure to soils contaminated with petroleum prod-
ucts. Of all the constituents identified, the carcinogenic polycyclic
aromatic hydrocarbons (CaPAHs) and benzene (a known human
carcinogen) are of most concern,2-6 and cleanup objectives for
soils contaminated with petroleum products are based on these
compounds.

ENVIRONMENTAL FATE
  Little is known about the environmental chemistry and ultimate
fate of petroleum  hydrocarbons in the soil environment. Their
fate in land is affected primarily by their distribution, volatiliza-
tion and  leaching potential.7 Low molecular weight aromatic
hydrocarbons such as benzene, toluene, xylene, etc., which have a
high Henry's Law constant,  tend to partly evaporate.  The re-
mainder will migrate to different depths of the soil column where
little or no volatilization to the atmosphere occurs.
  The proportion of the petroleum hydrocarbons that will bind to
soil versus those which will continue to migrate toward ground-
water  depends  primarily  on  the  type  of soil, the particular
petroleum product, the size of the spill and the amount of rain-
fall.7 In general, leaching to groundwater is favored by high rain-
fall and  permeable soils and increases for chemicals with high
solubility, low diffusion coefficients and low absorption coeffi-
cients' such as the aromatics benzene, toluene and xylene.
  Biodegradation is an important factor for removal of PAHs in
general, but high molecular weight multiring compounds such as
polycyclic aromatic hydrocarbons (PAHs) tend to remain in soil
for long periods of time. Biodegradation may be enhanced in soils
previously contaminated with PAHs. Naphthalene, a PAH with
two rings, behaves differently from other PAHs because of its
lower molecular weight and relatively high water solubility.8 The
rate of microbial degradation of PAHs in subsurface soil and in
groundwater is favored under aerobic conditions.
  Although migration of contaminants to groundwater is of con-
cern, this paper is concerned primarily with chronic effects due to
long-term exposure to contaminated soils.

ESTIMATED CONCENTRATIONS OF
CaPAHs AND BENZENE
  Concentrations of CaPAHs and of benzene in petroleum prod-
ucts vary depending on the type of crude oil and on the fractiona-
tion process used to derive the petroleum products. Table 1 sum-
marizes  information  on  the  concentration  of CaPAHs and
benzene in the various petroleum products studied. The CaPAHs
(benzo[a]pyrene  (BaP),  benzo[b]fluoranthene,  chrysene,
dibenz[a,h]anthracene, fluoranthene and  indeno[l,2,3,c-d]py-
rene) often are  present in crude oils in very small quantities
relative to the other PAHs.10  However, as indicated in Table 1,
used or spent petroleum products are enriched in content of BaP
and other CaPAHs by as much as 200-fold in some cases.2
  Information on  the concentration of benzene in the various
petroleum products studied is very limited. Of all the products
listed, gasoline has the highest concentration of benzene ranging
from 0.6 ppm to 2.9 mg/1 for all grades and octanes. Of the
gasolines, the concentration of benzene is highest in leaded gaso-
line (regular) during winter time.  The concentrations of total
benzenes  (benzene, toluene,  xylene  and other  substituted
benzenes) of 10.2% in a highly aromatic fuel oil No. 2 and of
1.0% in fuel oil No. 6'° indicate, as expected, that higher concen-
trations of benzene are found in low boiling  fuels than in high
boiling fuels. The concentration of benzene alone in these two
products was not reported.
                           Table 1
      Estimated Concentrations (ppm) of CaPAHs and Benzene
                 in Selected Petroleum Products
Petroleum Product
                                Total
                                CaPAHs*, ppm
                   Benzene, % Vol.
Crude Oil
Gasoline (regular)
Gasoline (high octane)
Gasoline (low octane)
Gasoline (unleaded)
Petroleum naphtha
    12;26.3(a)
Below  100(b)
      22.6(c)
       9.3(c)
       7.4(c)
                                                 0.2(e)
                     0.6-2.3(g)
                     0.2(h)
Fuel Oil No. 1 (jet fuel)
Fuel Oil No. 1 (kerosene)
Fuel Oil No. 2 (heating oil)
Fuel Oil No. 2 (diesel oil)


Fuel Oil No. 6 (Bunker "C")
Lubricating Oil
Fresh Motor Oil
Used Motor Oil
Used Motor Oil (5,000+ km)
Used Motor Oil (10,000+ km)
0.15(d)
0.05(b)
4.0(a)
0.7(b)
0.03(c)
37.3(d)
329(a)
0.3(c)
0.2(c)
5.8(c)
303. 6(c)
466. 6(c)
 'Ranging from one to four CaPAHs from a total of six depending on the study and on the par-
 ticular  petroleum product:  Benzo[a]anthracene,  chrysene,  benzo[a]pyrene  fluoranthene,
 dibenz[a]anthracene and indeno[ 1,2,3,c-d)pyrene.
 -Insufficient information available
 (a) National Academy Press10
 (b) U.S. EPA"
 (c) Verschueren11
 (d) U.S. EPA"
 (e) Sachanen"
 (0 Shelton"
 (g) Shelton"
 (h) Mellanl!
ESTIMATED RESIDUAL SOIL
CONCENTRATIONS OF CaPAHs AND BENZENE
  Assuming a soil cleanup level of 100 ppm of total PHs, residual
levels of CaPAHs  and of benzene  in  soil contaminated with
petroleum  products were estimated.  Residual  soil levels were
estimated using the highest concentrations of CaPAHs and of
benzene reported in the various studies reviewed. It was assumed
that no losses of these compounds occurred due to evaporation,
biodegradation or migration to groundwater. Residual soil levels
of CaPAHs and benzene were then compared to concentrations
of CaPAHs and benzene which would result in a one in a million
cancer risk if exposure were to occur. The residual soil concentra-
tions also were compared to "normal" soil background levels of
CaPAHs and benzene.

Calculation of Acceptable Soil Contaminant Levels
  The acceptable soil contaminant level (ASCL) that would pro-
tect  human health is  based on direct contact exposure with the
contaminated soil. A worst case exposure model  for carcinogens18
modified from Ford and Gurba" is shown below:
 / Acceptable Soil
  Cleanup Level
                                            1000 g/kg
                             (1)
                                                                                                    RISK ASSESSMENT    343

-------
where:
  A = Acceptable cancer risk =  1  x 10 ~6 (one in a million)
  C = Carcinogenic potency factor (U.S. EPA CAG)
         =  11.53 (mg/kg/day) - i for BaP-and 0.0052 (mg/kg/
        day)~ '  for benzene
     •  1000 g/kg = conversion factor
   L  = Lifetime average daily soil intake  = 0.0028 g/kg/day
  The model takes into account exposure  to contaminated soils
over a lifetime. Soil exposure is based on pica tendencies (in chil-
dren),  hand to mouthing and inhalation of dust.  It is assumed
that a  1  x  10-6 cancer risk is acceptable. Potency factors were
obtained from the U.S. EPA, Cancer Assessment Group (CAG).
The modified model does not take into consideration the half-life
of the contaminant in soil. Intake of soil during peak pica years is
assumed to be 2.5 g/day and not the 10 g/day contained in  the
original model.
  Unfortunately, carcinogenic  potency factors do  not exist  for
specific CaPAHs except for BaP. It is assumed that all CaPAHs
are as  carcinogenic as  BaP, and the potency factor for BaP was
used in the calculation. This  assumption  is consistent with  the
U.S. EPA's approach to estimating cancer risks from exposure to
mixtures of PAHs.  Acceptable soil contaminant levels of 0.03
ppm and 6.9  ppm were calculated for CaPAHs  and benzene,
respectively.

Soil Background Levels of CaPAHs and Benzene
  Information on soil background levels of CaPAHs and benzene
was obtained  from the literature. Limited  information exists on
soil background levels of PAHs in general.
                           Table 2
     Estimated Residual Soil Concentrations (ppm) of CaPAHs
     Exceeding Concentration of CaPAHs at 10 - ' Cancer Risk
         and "Normal" Soil Background Level of CaPAHs



Petroleum Product

Crude oil
Gasoline (regular)
Gasoline (high octane)
Gasoline (low octane)
Gasoline (unleaded)
Petroleum naphtha
Fuel Oil No 1 (jet fuel)
Fuel Oil No 1 (kerosene)
Fuel Oil No 2 (heating oil)
Fuel Oil No 2 (dlesel oil)
Fuel Oil No 6 (Bunker "C")
Lubricating Oil
Fresh Motor Oil
Used Motor Oil
Used Motor Oil (5,000+ Km)
Used Motor Oil (10,000+ Km)


Total
CaPAHs*,
ppra
3xlQ-3(a)
9xlO~£(b)
7x10" (b)
2x10 (b)

j
2xlO"'(c)
5xlO~°(d)
4x10". (c)
4x10 j(c)
3x10 ,(b)
3x10 '(a)
2xlO"'(b)
6xlO"*(b)
3x10 , (b)
5x10 (b)
Total
CaPAHs
Exceeding
Cancer
Risk**
N
N
N
N


N
N
N
N
N
N
N
N
N
Y
Total
CaPAHs
Exceeding
"Normal"
Soil
Background* *
N
N
N
N


N
N
N
N
N
N
N
N
N
N
*One 10 four CaPAHs from a tola! of six depending on study and on particular petroleum prod-
uct: Benzofajpyrene, benzo[a]anthraccnc, benzo(b)ftuoranthene, chryscne, dibcnz[a,b]amhra-
cenc, and indeno[l,2,3,c-d]pyrenc.
"CaPAHs at 10 ~6 cancer risk (based on BaP): 003 ppm
•""Normal" soil background of PAHs: 0.05 ppm
- Insufficient information available
Y Exceeding
N Not exceeding
(a) National Academy Press,"'
(b) Verschuercn"
(c) U.S. EPA"
(d) U.S. EPA"
  Natural  PAHs are due to plant synthesis, forest  and prairie
fires, volcanoes, etc.  Anthropogenic sources are primarily from
the extraction, processing and burning of fossil fuels.  Most of the
research on soil background levels of PAHs has been performed
on BaP. Therefore, selection of "normal" soil background levels
of PAHs was based on BaP.
  The concentration of total PAHs (the sum 5 to 20 PAHs) usual-
ly exceeds 10 times the concentration of BaP.'° BaP levels ranging
from less than 5 to  10 ppb are  generally found  in  agricultural
soils.1 Typical concentrations of  BaP in soils of the world range
from about 100 to 1,000 ppb, and values exceeding 100,000 ppb
have been  reported near known sources." It has been suggested
that endogenous BaP concentrations in soil are 1-3 ppb and never
exceed 10 ppb.10 A level of 50 ppb was selected as the "normal"
background of total PAHs in soil  since this level is typically found
as background.'
  Limited  data are available on  benzene levels in soils. Levels
ranging  from 13  to   115 ppb of  benzene were  reported in soil
samples  taken in the  vicinity of chemical  plants that  use or pro-
duce benzene." No information exists on the level of benzene in
"clean" soils. A soil background level of less than  10 ppb was
selected  since this level Is  within the range  of "normal"  soil
background levels  of most organics.'
  The ASCL of  30  ppb  for CaPAHs (assuming a 1  x \Q-6
cancer risk) is within  the range of typical soil background levels of
PAHs.  However,  the ASCL of  6.9 ppm  for benzene exceeds
typical soil background levels of benzene.
  Examination of Table 2 reveals that, after site remediation to
110 ppm of total PHs, only soil  contaminated with  used motor
oils over 10,000 km will have concentrations of CaPAH that ex-
ceed a I  x  10-6 cancer risk (by approximately 67%). As shown

                           Table 3
    Estimated Residual Concentrations (ppm) of Benzene Products
      Exceeding Concentration of Benzene at 10-* Cancer Risk
         and "Normal" Soil Background Level of Benzene
Petroleum Benzene. ppn
Crude oil 0.2U)
Gasoline (regular) l.l-2.9(b)
Gasoline (high octane)
Gasoline (low octane) 1.1-?. 9(b)
Gasoline (unleaded) 0.6-2.3(c)
Petroleum naphtha 0.2(d)
Fuel Oil No. 1
(Jet fuel)
Fuel Oil No. 1 (kerosene)
Fuel Oil No. ! (heating oil)-
Fuel Oil No. 2 (dlesel oil)
Fuel Oil No. 6
(Bunker "C")
Lubricating Oil
Fresh Motor Oil
Used Motor Oil
Used Motor Oil
(5,000+ Km)
Used Motor Oil
(10,000+ Km)
Benzene
Exceeding
10
Cancer
Risk**
N
N
N
N
N
N




-


_
_
_

.


Benzene
Exceeding
"Noroal"
Soil
Background**
Y
Y
Y
Y
Y
Y

.




.

_
.

_

-
"ASCL for CaPAHs at 10 ~ * cancer rijk: 6.9 ppm
""Normal" SBL of benzene: less than 0.01 ppm
- Insufficient information available
Y Exceeding
N Not exceeding
(a) Sachancn"
(b) Shelton"
(c) Shellon"
(d) Mellan"
344    RISK ASSESSMENT

-------
in Table 3, soil contaminated with crude oil, naphtha and gasoline
(all grades and octanes) has benzene levels below the ASCL for
benzene.  Comparison with  "normal"  soil  background levels
reveals that residual soil concentrations of CaPAHs are within the
range of "normal" soil background levels of PAHs. However,
residual soil concentrations of benzene from contamination with
crude oil, petroleum  naphtha  and gasoline  (all grades and oc-
tanes) are above the "normal" soil background level of benzene
as expected.  The residual soil concentrations of benzene from
crude oil and naphtha exceed the soil background level of benzene
in excess of 20 times and that from gasoline by 90 times.

CONCLUSIONS AND RECOMMENDATIONS
  Assuming a soil cleanup level of 100 ppm for total petroleum
hydrocarbons, most of the residual soil  levels of  CaPAHs and
benzene are below their corresponding concentrations resulting in
a 1 x  10~6 cancer risk. An exception is the residual soil level of
CaPAHs from contamination  with used  motor oils over 10,000
km which has a risk of 1.6 x 10~6. In this case, lowering the soil
cleanup level of 100 ppm  to 60 ppm for total PHs, will decrease
soil concentrations of total CaPAHs from contamination of the
above  product to levels resulting in a cancer risk of 1 x  10 ~6.
Alternatively, testing soil samples for CaPAHs could be per-
formed after soil is cleaned to 100 ppm of total PHs to determine
if CaPAHs levels remaining in  soil are acceptable.
   Comparison  of residual soil levels of CaPAHs and benzene
with their respective  soil background  levels,  indicates that, in
general, residual levels of CaPAHs are within the range of typical
soil background levels of PAHs. On the other hand, residual soil
levels of benzene exceed the assumed "normal" soil background
level of benzene. Further refinement of this comparison cannot be
made at this time due to  limited information  on "normal" soil
background levels of these compounds  and due to the volatiliza-
tion of benzene with time.
   A more extensive data base on petroleum products is needed to
refine the risk assessment  approach proposed.  Specifically, more
data on  the identification and  concentration of chemical consti-
tuents, behavior in the soil environment, toxicity and normal soil
background concentrations are needed.
   More toxicity testing of petroleum products and individual con-
stituents is needed to better understand the carcinogenic potency
of individual CaPAHs and  of mixtures  before a more  refined
ASCL can be estimated.
   In  summary,  with the  exception of  motor oils used over
 10,000 km, 100 ppm of total PAHs as a soil cleanup objective ap-
pears to  result in residual  soil levels of CaPAHs and benzene not
exceeding a 1 x  10~6 cancer risk.  Although this cleanup objec-
tive is protective in a direct soil contact scenario, it may not ade-
quately address ground water protection.

ACKNOWLEDGEMENTS
   We greatly appreciate the assistance of  Professor Richard Bar-
tha,  Rutgers University, Dr. Neill  K.  Weaver,  American
Petroleum Institute and Dr. Merry L. Morris, New Jersey Depart-
ment of Environmental Protection, in  the development of this
manuscript.

REFERENCES
  1. Curl, H. Jr. and Donnel K., "Chemical and Physical Properties of
    Refined Petroleum Products," Boulder, CO,  National Oceanic and
    Atmospheric Administration/Environmental Research Laboratories,
    Marine Ecosystems Analysis Program, NOAA Technical Memoran-
    dum ERL MESA-17, U.S. Department of Commerce (Oct. 1977),
    5-26.
 2. Bingham,  E., Trosset, R.P. and Warshawsky, "Carcinogenic Po-
   tential of Petroleum Hydrocarbons. A Critical Review of the Litera-
   ture," J. of Environ. Path, and Toxicol. 3, (1979, 483-563.
 3. U.S. EPA, "Estimation of the Public Health Risk From Exposure to
    Gasoline Vapor Via the Gasoline Marketing System." Staff Paper
    submitted for review to the Science Advisory Board, Office of Health
    and Environmental Assessment, Office of Air Quality and Planning
    and Standards, U.S. EPA, Washington, DC, June 1984.
 4. Sheppard, E.P., Wells,  R.A. and Georghiou, P.E., "The Mutagen-
    icity of the Prudhoe Bay Crude Oil and Its Residues from an Experi-
    mental In Situ Burn," Environ. Research 30, (1983), 427-441.
 5. National Institute for Occupational Safety and Health, "Registry of
    Toxic  Effects of Chemical Substances," 1983 Supplement to the
    1981-82 Edition, U.S.  Government  Printing Office, Washington,
    DC, [DHHS(NIOSH), 84-101], 1984.
 6. Sittig,  M., Handbook of Toxic and Hazardous Chemicals and Car-
    cinogens, 2nd. Ed., Noyes Publications, Park Ridge, NJ, 1985.
 7. Bossert, E. and Bartha, R., "The Fate of Petroleum in Soil Ecosys-
    tems." In: Petroleum Microbiology, Ronald H. Atlas, Ed., Mc-
    Millan Publishing Company, New York, NY 435-470.
 8. American Petroleum Institute, "The Land Treatability of Appendix
    VIII Constituents Present in Petroleum Industry Wastes," American
    Petroleum Institute, Washington, DC, API  Publication 4379 Docu-
    ment B-974-220, May 1984.
 9. Bonazountas, M.,   "Mathematical  Pollutant  Fate  Modeling of
    Petroleum  Products in  Soil Systems," Lectures at  Conference on
    Environmental and  Public Health Effects  of Soils  Contaminated
    with Petroleum Products, University of Massachusetts, Amherst,
    MA, Boston, MA, Epsilon International Inc., Oct. 1985.
10. National Academy Press, "Chemical  Composition  of  Petroleum
    Hydrocarbons Sources." In: Oil in  the Sea, Inputs, Effects, Na-
    tional Academy Press, Washington, DC, 1985, 17-42.
11. U.S. EPA, "An Exposure and Risk Assessment for Benzo(a)pyrene
    and Other Polycyclic Aromatic Hydrocarbons," Final Draft Re-
    port, Office of Water Regulations and Standards, U.S. EPA, Wash-
    ington, DC, 1982, 5-187.
12. Verschueren, K., Handbook  of Environmental Data  on  Organic
    Chemicals, 2nd. Ed., Van Nostrand Reinhold Company, Inc., New
    York,  NY, 1983.
13. U.S. EPA,  "Quantitative Analysis of Polynuclear Aromatic Hy-
    drocarbons  in Liquid  Fuels,"  Environmental  Sciences Research
    Laboratory, U.S. EPA, Research Triangle  Park, NC, EPA-600/2-
    80-069, Apr. 1980.
14. Sachanen, A.N.,  "Hydrocarbons in Gasolines, Kerosenes, Gas Oils,
    and Lubricant Oils." In:  The Chemistry of Petroleum Hydrocar-
    bons, Reinhold Publishing Corporation, New York, NY, 1,  1954.
15. Mellan, I., "Pure Hydrocarbons," In: Handbook of Solvents, Rein-
    hold Publishing Corporation, New York, NY, /, 1957.
16. Shelton, E.M. and Dickson, C.L., "Motor Gasolines, Winter 1984-
    85," National Institute for Petroleum and  Energy  Research, Bar-
    tlesville, OK. Work performed  for the American Petroleum Insti-
    tute, Washington, DC, NIPER-140PPS 85/3, June 1985.
17. Shelton,  E.M. and Dickson,  C.L.,  "Motor Gasolines, Summer
    1984," National Institute for Petroleum and Energy Research, Bar-
    tlesville, OK. Work performed  for the American Petroleum Insti-
    tute, Washington, DC, NIPER-138PPS 85/1, Feb. 1985.
18. Dime,  R. and Greim, B., "Calculation of Cleanup Levels for Con-
    taminated  Soils," Unpublished Manuscript, N.J. Department of
    Environmental Protection, Hazardous Site  Mitigation Administra-
    tion, Trenton, NJ, 1985.
19. Ford, K.L. and Gurba,  P., "Health  Risk Assessment For Contam-
    inated  Soils," Proc. of the 5th National Conference on Management
    of Uncontrolled  Hazardous Waste Sites, Washington,  DC, Nov.
    1984.
20. Edwards, N.T., "Polycyclic Aromatic Hydrocarbons (PAHs) in the
    Terrestrial Environment—A Review," /. Environ. Qual., 4, (1983),
    427-441.
21. U.S. EPA, "An Exposure and Risk Assessment for Benzene," Final
    Draft Report, Office of Water Regulations and Standards,  U.S.
    EPA, Washington,  DC, 1980.
                                                                                                       RISK ASSESSMENT    345

-------
                                 Risk Management:  Personnel,
                                     Equipment and Indemnity
                                                  Denny M.  Dobbs
                                                  Arlene B.  Selber
                                                  HAZTECH,  Inc.
                                                  Atlanta,  Georgia
ABSTRACT
  Risk management always has been a key ingredient in the over-
all management scheme of the hazardous materials cleanup and
site restoration industry. Since the industry's beginnings in the
early 1970s, companies have been  practicing risk management,
whether or not identified b.y that particular name. As the industry
has evolved and more has been learned about potential risks, the
need for a well-defined risk management program has increased.
Current health and safety regulations, increased litigatory actions
and the ominous problems with insurance  coverage have pro-
duced a realization of the critical nature and overall importance
of risk management. Three areas where risk management  can
have a significant impact on company operations are personnel,
equipment and general indemnity.
INTRODUCTION
  Risk management can be defined as the management or con-
trol of the possibility  of loss  or injury. This management has
both a social or moral aspect as well as practical ramifications.
A waste of manpower, time, resources  and efficiency are just a
few of the effects  of accidents  that could have been managed or
prevented. In order to manage risk, one  must identify the follow-
ing for all potential risks: degree of hazard or danger; probability
of occurrence; methods of prevention; and preferred health and
safety plan.
  Identifying and  managing risk has become so important to the
hazardous waste industry and  to society that  many  universities
have recently added risk management courses to their core insur-
ance curriculums.  Businesses are beginning to employ people that
have risk management as their prime job function. Personnel,
equipment and general indemnity  are the three areas that this
paper will address.
PERSONNEL
  The one area most often overlooked in an active growing com-
pany  is  the area of personnel  management. Personnel manage-
ment  encompasses a wide range of functions, but for the pur-
poses of this paper, only the area of risk management will be dis-
cussed.  Risk management as it relates to personnel involves min-
imizing  exposure to hazardous conditions, providing and requir-
ing proper training, publicizing and maintaining employee  inter-
est in safety and instituting an efficient record-keeping system.
  Hazardous materials cleanup work involves, by its very nature,
a certain degree of exposure to risk. The job of risk management
is to minimize the degree and quantity of that exposure. The de-
gree of  exposure relates to the  individual's physical exposure,
while quantity relates to the number of individuals actually ex-
posed. Both of these concerns usually arise in operational settings
where prior planning is the best solution to potential problems.
Exposure
  Physical exposure is minimized by a thorough knowledge of
the work site and conditions present. What  chemicals are in-
volved?  What  properties, concentrations, etc., did the analysis
reveal?  Is there potential exposure to irritants, poisons,  skin
absorption, etc?
  Once  the hazardous conditions are defined, the proper selec-
tion of protective equipment  and safety measures can be made.
The equipment selected may include what types of boots, gloves,
respirators or supplied air system, protective suit or coveralls, are
required. Once the proper type of equipment has been selected,
management must make  sure that the employees use and main-
tain the equipment correctly.
  The investment in time and effort required to correctly  pro-
tect employees and minimize physical exposure pays significant
dividends in workplace efficiency and decreases the risk of litiga-
tory actions.
  Quantity exposure as described earlier involves numbers. Gen-
erally, in the hazardous materials business, numbers relate most
directly  to  the number of  employees  specifically  involved.
Thorough planning before a  job begins is the best method of
controlling quantity exposure. A careful study of the scope of
work and the proposed work  plan will allow the project manager
to properly select his  project  team. The project manager's work
plan should  be designed to expose the  fewest number of em-
ployees. Fire departments, police departments and military units
have successfully  followed  this principle  for many  years. In
reality,  basic common sense plus careful  analysis and safety
planning will minimize the number of persons exposed.

Training
  How  important is training? Would you fly with an untrained
pilot? Would you hire an untrained lawyer? It is obvious, given
the current regulatory  attitude toward the work place, that proper
employee training is critical.  In  the early days of this industry,
many things  were done without adequate training because the in-
dustry was so new and procedures and standards were ill-defined.
As the industry has grown and aged, it has defined many of the
skills required  to work safely and can now train people in these
skills.
  Training in the hazardous materials industry is a cross between
EMT, fire department and military training programs. A train-
ing program should include a discussion of standard work prac-
tices,  personal hygiene, first aid, CPR, etc. Detailed training in
the selection and proper wearing of protective equipment is essen-
tial. Training with the proper content and attitude  is like preven-
tive maintenance for your work force.
346    RISK ASSESSMENT

-------
Record Keeping
  The last area that will be discussed under personnel is record
keeping. So far, the discussion of risk management has had a pure
health and safety theme. Record keeping, however, is necessary
to facilitate the treatment of an employee and to serve as proof
that adequate training was offered and required. The question of
risk management versus liability will be  discussed in the last sec-
tion of this paper.
  Keeping accurate records is proof that a particular health con-
dition was or was not job related, or that someone  was or was
not injured by exposure to a hazardous situation. Given the occu-
pational health and safety  and workmen's compensation laws
currently in existence, detailed personnel records are  an absolute
must.

EQUIPMENT
  Risk management with regard to equipment, particularly over-
the-road equipment, in view of the current insurance crisis and
large injury awards for equipment-related injuries must be an in-
tegral part of business strategy. The following example illustrates
this.
  An employee driving a company truck is involved in an "at
fault" accident. It is determined that the employee was DUI in a
truck with faulty brakes, and the other  driver sustained injuries.
Is this scenario not an example of the importance of driving safely
and adhering to company policy adequately communicated to the
driver? No business can afford to fail to communicate the impor-
tance of proper equipment maintenance to its operations depart-
ment. The difficulty of obtaining insurance continues  to reinforce
the necessity of the proper maintenance and use of equipment.
  The hazardous waste industry requires a variety of equipment
including trucks, cars, heavy equipment, protective suits, respira-
tors,  masks, hand  tools, eye  wash bottles, decontamination
equipment, etc. Often, the major emphasis or interest is focused
on equipment which can be driven on public highways. However,
no piece of equipment should be placed in operation until it has
been inspected by a qualified person and found to be in safe oper-
ating condition. Routine inspection after each project should be
made, and inspection reports and  service records should be kept
on file.
  Obviously, the most logical way to minimize equipment related
injuries and  losses is through the proper maintenance and use of
that equipment. A U.S. Army Corps of Engineer study revealed
that a "great majority of accidents were  caused by inadequate
maintenance of equipment,  insufficient instruction in safe prac-
tices or lack of insistence that safe practices be observed." Un-
safe practices contributed to 80% of all injuries. Remember the
example given earlier in this section: equipment related losses can
become extremely complicated and costly if they are not posi-
tively affected by good risk management practices.

INDEMNITY
  According to Webster, indemnification is "the action of secur-
ing against hurt, loss or damage." Nothing causes more concern,
argument, loss of coverage or law suits than the subject of indem-
nity. The most difficult  area to negotiate for any contract is the
indemnity section. Ideally, you would want to be indemnified
for everything and yet give indemnification for nothing. Prac-
tically, you must negotiate the best  position you can with the
other party. One of the important things to remember is that most
insurance  coverages are  tied very  closely to indemnification and
how it affects the risk that you are being insured against. The first
rule of good business practice is to be absolutely sure that you can
live with the indemnification you  agree to  because you just may
have to live with it. Never agree to accept a risk expressly prohib-
ited by your insurance coverage unless you intend to self-insure
that risk.
  This paper  will not discuss all the legal ramifications of indem-
nities. However, one major point  should be recognized. The law
does not make an engineer completely responsible for a project's
success. Engineering is not a perfect science, and "acts of God"
can occur  even when an  engineer exercises professional judgment
and utilizes adequate skills. Indemnification clauses, written to
the advantage of your client, can make the engineer (and your
company) a guarantor, an insuree, of a project's success.  Do not
agree to this.
  Indemnification is a major vehicle to either increase or decrease
risks, so be aware of what you agree to indemnify. It is the recom-
mendation of the authors  that you only be liable for your  own
negligence.

CONCLUSIONS
  Risk Management is a function of management that requires
thought, planning, implementation and considerable expertise.  It
becomes more important as the  hazardous waste industry con-
tinues to be plagued by loss of insurance  coverage, workmen's
compensation claims, litigation and huge jury awards. The im-
portance of managing risk is being recognized by industry and
academia and is fast becoming an integral part of successful com-
panies' management plans. Good risk management in the areas of
personnel, equipment and indemnification  is good business, and
good business is what makes a good profit.
                                                                                                    RISK ASSESSMENT    347

-------
                      Innovative Approach  to  Site Remediation
                            Involving an  RCRA  Part B  Permit

                                            Michael J. Conzett, P.E.
                                             Michael E. Harris,  P.E.
                                    HDR  Environmental  Technologies,  Inc.
                                                Omaha, Nebraska
ABSTRACT
  A preliminary investigation of an abandoned industrial waste
disposal site indicated the presence of materials containing haz-
ardous substances and groundwater contamination. The parties
involved are the owner, the Iowa Department of Water, Air and
Waste Management (IDWAWM) and the U.S. EPA Region 7. As
a result of the preliminary investigation, a stipulation between
the owner and IDWAWM was enacted. A detailed site investiga-
tion confirmed the following:
• A complex geophysical structure
• The presence of material containing hazardous substances
• The presence of a layer of floating volatile organic compounds
  (VOCs)
• The presence  of contaminated groundwater in the vicinity of
  the site
  To date, the owner has undertaken the remedial investigation
activities and proposes to voluntarily undertake  the  feasibility
study and cleanup. To meet the needs of the owner and satisfy the
requirements of the regulatory agencies, it was determined that a
Part B permit for Waste Piles would be required as part of an
innovative approach to site remediation.

INTRODUCTION
  Industrial wastes resulting  from the manufacture of paint and
varnish products were placed into a land disposal facility.  The
aquifer directly under the site is utilized as a source of water for
domestic and agricultural uses throughout the area. This prox-
imity of waste and groundwater was sufficient  cause to place the
site on the proposed  National Priority List. In this paper, the
authors describe the following: site history; remedial  investiga-
tion activities; summary of releases/contamination; proposed re-
medial action plan; owner/regulatory agencies  relationship; Part
B permit; and the innovative approach to site remediation.

SITE DESCRIPTION AND HISTORY
  Prior to the early 1970s, the owner disposed of his  wastes by
open burning. Enactment of the Clean Air Act, however, pro-
hibited this practice, and an  alternate method of disposal  was
necessary. The owner sought the aid of the Stale of Iowa, and an
agreement was reached whereby the owner's wastes were to be
disposed at an abandoned gravel pit in rural Iowa.
  Between 1971 and 1979, solid and liquid wastes generated from
the owner's paint and varnish  manufacturing operations were dis-
posed in a series of 12 trenches at the site (Fig. 1).  The trenches
were slot-dozed  to a depth  of between 8 and 12  ft.  After the
trenches were opened, they v»ere filled with waste fluids poured
from 55-gal drums. Each trench remained open for an extended
period of time to allow for evaporation of volatile hydrocarbons.
Before a trench was closed, it was topped off by placing in it a
relatively large quantity of drums and small buckets containing
waste paint sludges, resins, solvents and general trash. After the
level of  waste material  in each trench approached the natural
ground surface, the trench was covered with 1 to 2 ft of the clayey
silt loess soils initially excavated from the trenches.
                         Figure 1
             Plan View of Waste Disposal Trenches
  Although accurate records were not kept by the owner, an esti-
mated  150 to 300 drums (55-gal) were buried at the site and as
many as 43,000 gal  of liquids were dumped in the trenches. The
wastes  contained approximately 6,000 Ib of heavy metals (e.g.,
lead, cadmium, chromium, zinc and mercury) and an unknown
quantity of volatile organic compounds (e.g.,  xylene,  toluene,
ethylbenzene, methylethyl ketone and mineral spirits).
  The  waste disposal site was closed  in 1979. Wastes generated
after 1979 were disposed in properly permitted facilities.

REMEDIAL INVESTIGATION
  Remedial investigation activities began in 1979 at the request of
the State of Iowa.  The ultimate goal of these  activities was to
obtain  sufficient information regarding the soil and groundwater
contamination at the site so that a closure plan could be devel-
oped and implemented. Because of the complex geophysical na-
348
       RCRA AMENDMENT l-.XPCRH-.NC'I.

-------
ture of the site, remedial investigation activities were conducted
in several phases between 1979 and the end of 1984. A brief sum-
mary of the activities performed during this time is presented be-
low.

  1979—Activities were limited to a subsurface investigation to
determine potential off-site migration of wastes. Thirty of thirty-
five borings were drilled on and immediately adjacent to the site.
Only water level measurements were taken, and the borings were
subsequently backfilled.
  1982—Five monitoring wells were installed. Floating hydrocar-
bons were detected in one of the wells and samples were taken
during subsequent months for chemical analyses.
  1983-1984—The presence of a substantial thickness of floating
hydrocarbons  in  two  of  the monitoring wells  prompted
IDWAWM to issue an administrative order requiring contain-
ment  and recovery of the contaminant plume, proper disposal
of recovered waste and proper closure of the site.
  May 1984—Observation pits were dug at the site to determine
the approximate  quantities  of waste materials  buried  in  the
trenches. Samples  also were  taken  from the trenches and from
other soil borings to determine the concentrations of hazardous
constituents in the trenches and soil. The geological stratigraphy
of the site was evaluated. Seventeen monitoring wells were in-
stalled and developed.  Groundwater samples were subsequently
obtained and analyzed for contamination.
  August 1984—IDWAWM  ordered the owner to remove float-
ing hydrocarbons  periodically  and to take quarterly  samples
from the monitoring wells.
  September  1984—Nine additional  monitoring  wells were in-
stalled and developed.  Testing was conducted to determine the
aquifer characteristics.
  Present—The owner continues to remove floating  hydrocar-
bons, make biweekly water level measurements in the monitor-
ing wells and take quarterly groundwater samples.

SUMMARY OF RELEASES/CONTAMINATION
  As  a  result of the disposal of hazardous waste liquids in the
trenches beginning in 1971, the most immediate impact has been
the release of hazardous materials to the underlying shallow
aquifer system. A secondary impact due to the continuing deteri-
oration of buried drums probably is resulting in more recent and
continual releases of hazardous materials to the aquifer.
  Due to  the unique characteristics of the hydrogeology in the
area of the site, the extent of groundwater contamination appears
to be limited  to the site and to a distance only several hundred
feet away from the site. The degree of groundwater contamina-
tion depends on the compound. The maximum concentration of
the volatile organic compounds ranges from 28 mg/1 for toluene
to 260 mg/1 for xylene. The maximum concentration  of heavy
metals ranges  from 50 ng/1 for lead to 80 ug/1 for chromium.
  The extent  of soil contamination is limited primarily to  the
waste disposal trenches and the vadose zone directly below the
site. The maximum degree of soil contamination ranges from 15
ppm for chromium to 1,700 ppm for xylene.
  As mentioned in the remedial investigation section, there is a
floating layer  of relatively pure hydrocarbons that is unique to a
small  portion of  the groundwater adjacent to the site. These
hydrocarbons are removed by  pumping biweekly as part  of an
IDWAWM stipulation.

PROPOSED REMEDIAL ACTION PLAN
  A feasibility study currently is being written to compare vari-
ous remedial action alternatives for cleanup of the disposal site.
The objective of the feasibility study is to recommend a remedial
action plan for cleanup of soil and groundwater that effectively
provides protection of human health and the environment at the
least cost.
  The proposed remedial action plan will likely include the clean-
up activities discussed below.

Source Removal and Ultimate Treatment
  Hazardous wastes will be exhumed from the trenches and will
be disposed in properly permitted facilities or will be treated in a
manner to render the wastes non-hazardous.
  Free-standing liquid wastes in the trenches and liquids taken
from drums will be stored temporarily in a tanker truck. Then
the liquids either will be recovered in the owner's distillation
equipment or will be sent off-site and ultimately destroyed by in-
cineration.
  Solid wastes (e.g., contaminated soil, drums, buckets,  trash,
etc.) will be stored temporarily on-site  in above-ground concrete
bunker silos. Treatment technologies for the solid wastes will be
investigated and/or tested while the wastes are safely stored in
the silos. Treatment methods may include incineration, air strip-
ping, encapsulation, biodegradation and physical/chemical treat-
ment.

Recovery and Treatment of Groundwater
  Groundwater will be pumped, treated and reinjected into the
aquifer. The  treatment/flushing action will  continue until the
groundwater is rendered non-hazardous. Treatment methods like-
ly will include air stripping and carbon adsorption.

Owner's Involvement
  The owner has  expressed a high degree of interest in the pro-
posed remedial action plan presented above. His basic philosophy
is to have as much control of his present and future liability as
possible. He believes that the proposed remedial action plan will
provide effective remediation of the site at the least cost and with
the least liability. In addition, the least  cost treatment technology
which will render  the wastes non-hazardous may be found while
the waste is safely stored on the owner's property.  Not surpris-
ingly, this general type of remedial action (i.e., on-site storage/
treatment) is favored by the U.S. EPA.

OWNER/REGULATORY AGENCIES
RELATIONSHIP
  Three parties (the owner, IDWAWM and the U.S.  EPA Region
7) are involved with the cleanup of the disposal site. Due to the
nature of the relationships among the  three parties, there could
have been some potential problems. The owner is a responsible
private party willing to finance  the remedial investigation/feasi-
bility study and site cleanup. IDWAWM is responsible for the
cleanup and monitoring of uncontrolled hazardous waste disposal
sites in the State of Iowa. The U.S. EPA Region 7 is responsible
for all RCRA programs.
  The State of Iowa returned RCRA  responsibility to the U.S.
EPA in mid-1985. Included in the U.S. EPA's RCRA responsi-
bility is  the issuing of Part B permits for treatment, storage and
disposal facilities  and the monitoring of these  same facilities.
Cooperation among the three parties in the past and at the pres-
ent time has allowed effective progress toward site cleanup.
  When the owner expressed  interest in storing the  exhumed
wastes on-site, a meeting was convened with the regulatory agen-
cies to discuss the ramifications of this proposed action. Although
the owner's disposal  site is on the proposed National Priorities
List (NPL), the owner learned that an RCRA permit would be re-
quired prior to the construction of the storage facility.  Because
the site was not under an emergency cleanup order, the require-
ments of other environmental laws (e.g., RCRA) would have to
be followed. The  regulatory review of a Part B permit applica-
                                                                                    RCRA AMENDMENT EXPERIENCE    349

-------
tion can take a great deal of time. The owner wished to Icean up
the site as quickly as possible and  was naturally disappointed
when he learned a permit was required.  It would have  been easy
for the owner to criticize the U.S. EPA for "dragging its feet"
regarding the cleanup of the site; however, the requirements of
the RCRA permit program were seen by the owner as  necessary
to protect human  health and the environment from  potential
damage as a result of the cleanup.
   It is especially important to develop good relationships between
the owner and the regulatory agencies involved in a site cleanup.
Issues can be addressed and problems can be solved only if each
party  keeps  an  open  mind and  communicates  and cooperates
with the others.

PART B PERMIT
   To  fulfill the owner's desire to store the exhumed hazardous
wastes in  an above-ground facility, a Part B permit was required.
As mentioned previously, the cleanup  of  this site required  the
adherence to other environmental laws, most notably RCRA. The
filing of a Part B permit application  for this innovative approach
to site remediation is unique.  Historically,  cleanup of an uncon-
trolled waste site involves the transportation of the wastes to an
off-site RCRA-permitted disposal facility. This  action does not
require the owner of the uncontrolled waste site to wait while a
permit is  reviewed  and approved; however, it  also does not give
the owner much control over his liability. In the  case study re-
ported in this paper, the owner is willing to wait for a Part B per-
mit in order to implement an innovative  approach that allows him
to better control his liability.
LOCATION CT CONCRETE -*•   •  C.HION
1TOHAOI IlLOS U«0 TO   /
 1TOAI ElMtMEO WASTE
 NO CONTAMINATED  SOIL   /

         V         I
                            Figure 2
           Location of Bunker Silos Relative to Disposal Site
 Facility Description and Use
   The Part B permit application was made for a new on-site stor-
 age facility comprised of four above-ground bunker silos located
 adjacent to the waste disposal site (Figs. 2 and 3). Each silo will be
 rectangular with its dimensions being 56 ft wide, 104 ft long and
 a wall height  of  12 ft. Each silo is expected to store,  in piles,
 approximately 1,750 yd' of waste.
                                            DNAINAOC O'TCN
                                            torn HUM OM
                                            COMTMOL V"^
                                                                 Figure 3
                                                          Site Plan of Bunker Silos
                                                                     The sole purpose of the facility is to temporarily store contam-
                                                                  inated soils and other hazardous solid wastes that will be removed
                                                                  from the waste disposal site. Liquid wastes or soils containing free
                                                                  liquids will not be stored. These materials are to be either solidi-
                                                                  fied on-site prior to storage or manifested  and transported to a
                                                                  licensed facility for incineration.
                                                                     The waste materials are to remain in storage while various tech-
                                                                  nologies are investigated and/or tested to  treat them. The ulti-
                                                                  mate goal is to treat the materials in the storage facility to perma-
                                                                  nently render them nonhazardous. Then the waste materials can
                                                                  be delisted. The treated soils can be returned to the disposal site,
                                                                  and the other solid materials can be disposed as non-hazardous
                                                                  wastes.
                                                                                  ACC1S1 PANfl
                                                                                  '01 WASTI
                                                                                  PLACtUfNT
                                                                                                          WlM* VCNT1LATIOM
                                                   *—ACCCIS DOOR
                                                                              SUMP NOOM-
                                                                  Figure 4
                                                     Elevation View of Typical Bunker Silo
 350    RCRA AMENDMENT EXPERIENCE

-------
  Bunker Silo Design Elements
    The bunker silos  are designed to provide protection of the
  waste piles from precipitation so that neither runoff nor leach-
  ate is generated. Each silo will have a reinforced concrete base,
  concrete walls and a steel roof (Figs. 4 and 5). A 6-in. concrete
  curb will surround the silo, thus providing protection from sur-
  face water run-on. Because the waste pile will be enclosed by walls
  and a roof, the wastes will not be dispersed by the wind. Because
  liquids will not be placed in the piles, leachate is  not expected to
  be generated  by biological or chemical degradation during the
  time the piles are in existence. Each silo also will be equipped with
  a ventilation fan to  minimize the buildup of organic vapors to
  explosive levels.
    The above elements of the structure are considered sufficient to
  allow the owner to request an exemption from liner and leachate
  standards and from  groundwater monitoring requirements under
  the 1984 RCRA Amendments. Although a request was  made for
  the above exemptions, it is the owner's stated intent to provide
  a state-of-the-art facility.  Therefore, for added protection  of
  human health and the environment resulting from the potential
  release of hazardous wastes, a double synthetic liner system will
  be installed accompanied by leak detection and  leachate collec-
  tion systems (Figs. 5  and 6).
6" CONCRETE CUHI
EXTENDING 6"
ABOVE GRADE
             -CONCRETE FOOTING
                          a- REINFORCED'
                          CONCRETE SLAB
LEAK DETECTION
SYSTEM - r-0- GRANULAR
BED WITH 4- PVC
PERFORATED PIPE (TYPICAL)
                           Figure 5
        Typical Cross Section of Silo Showing Design Elements
    The owner chose to include synthetic liners and leak detection/
  leachate collection systems in the design of  the storage silos to
  serve as an additional safety factor. The owner had enough fore-
  sight to  realize that,  for a  relatively small increase in cost,  he
  could decrease his liability considerably by minimizing the poten-
  tial for a non-sudden release of hazardous waste.

  Placement of Wastes
    An auger will be used to transfer the waste materials into each
  silo via open panels in the roof. Inside the silo, a small end loader
  will move the materials into  a pile. Plastic sheeting will be placed
  on the ground around the auger to collect spillage. Portable wind
  screens near the auger will prevent wind dispersal of dry waste
  materials.

  Closure of Facility
    As mentioned previously, the goal of the remediation program
  is to ultimately treat the soils and wastes so that they can be dis-
  posed as non-hazardous materials. The duration of storage in the
  bunker silos is unknown and will be dependent upon the develop-
  ment of appropriate technologies and subsequent testing of these
                                                                    ACCESS RAMP OPENING-
                                        ' LEAK DETECTION SUMP—
                                         <2'-0-i2'-0-*3'-0- DEEP)
                                                                   'PRECAST CONCRETE
                                                                    WALL
                                                                                                 SUMP ROOM -
                                                                  ^SYNTHETIC
                                                                  '  LINERS
                                                                   -TYPICAL PERFORATED
                                                                   PVC PIPE (FOR LEAK
                                                                   DETECTION)
                                                                                                               - LEACHATE
                                                                                                                COLLECTION  SUMP
                                                                                                                (2''0~x2''Q~x3''0~ DEEP)
                                                                                              Figure 6
                                                                              Plan View Cross Section of Typical Bunker Silo
technologies on the stored wastes to determine what constitutes
effective treatment. It is possible that future technologies will not
prove completely effective and that one or more of the silos may
need to be repermitted as disposal facilities.
  If treatment technologies are proven effective and the wastes in
the piles are removed and rendered non-hazardous, the storage
silos will be closed. The bunker silos and equipment  stored with-
in them will be tested  to determine if they are contaminated.
Appropriate procedures will be used to decontaminate any con-
taminated structure or equipment.

INNOVATIVE APPROACH TO SITE
REMEDIATION
  Because the owner wanted to clean up his site as soon as pos-
sible,  a two-stage  approach to remediation originally was  con-
ceived. The scenario was  as follows.  The first stage of this ap-
proach would involve removal of the source of contamination. A
feasibility study addressing source removal  would  be written,
followed by review and approval by ISWAWM. Following ap-
proval, the cleanup of the source would begin and take approx-
imately 6 to 7 months using the services of local contractors.
  While the source is  being removed,  the second stage  of the
approach, remediation of groundwater, would be addressed.  A
feasibility study for this stage  would  be prepared. Following
IDWAWM review and approval,  tne groundwater cleanup would
begin in the subsequent calendar year.
  When it was learned that a Part B permit would be required
for on-site storage, it was determined that a two-stage approach
would  not  work. Instead, the owner was faced with making a
choice between two options concerning the logistics of remedia-
tion. The first, and more traditional, option involved writing a
feasibility  study  for  total  site remediation,  followed  by
                                                                                        RCRA AMENDMENT EXPERIENCE    351

-------
 IDWAWM review and approval of the recommended alternative.
 If a Part  B permit was required before site cleanup  could be-
 gin, the permit application would then be prepared, followed by
 U.S. EPA review and approval. The problem with  this option is
 that it would take a minimum of 1 year to receive a Part B permit
 after the U.S.  EPA received the application. All parties involved
 (owner, ISWAWM and U.S.  EPA)  thought that this option
 would unnecessarily delay site cleanup and  a second option was
 considered.
   The second  option involved a more innovative approach to re-
 mediation. This  option involved first preparing a Part B permit
 application under the assumption that the on-site storage  alterna-
 tive ultimately would  be the approved alternative for source
 cleanup. In conjunction with the U.S.  EPA review  of the permit
 application, a  feasibility study for total site remediation would be
 prepared,  followed by IDAWAWM  review and approval. Copies
 of the Part  B  permit application and feasibility study  would be
 shared  by  both regulatory agencies  so  that a  coordinated
 approach  to site cleanup would result.  The major  risk involved
 with this option  would be that if on-site storage was not  the ap-
 proved  alternative, the  expense of  preparing a Part  B permit
 application would be wasted.
   The owner chose to pursue the second, more innovative, op-
 tion to  site  remediation. At  the present time, a Part  B permit
 application for the on-site bunker silos is being reviewed by the
 U.S. EPA. A  feasibility study is being written concurrently and
 will be reviewed  by IDWAWM. Because the owner and  regula-
 tory agencies discussed the options at an early stage, the  risk in-
 volved in  pursuing the innovative approach is considered mini-
 mal. The result of this action likely will be a faster cleanup of an
 NPL site (savings of time of at least 1 year) than if the traditional
 approach was pursued.
CONCLUSIONS
  The proposed remediation of an uncontrolled waste disposal
site in rural Iowa is being implemented using an innovative altern-
ative for storage of wastes and an innovative approach to  the
regulatory review process. Conclusions made from the work per-
formed to date are:
• Off-site storage  of  exhumed solid wastes in  above-ground
  bunker silos is an attractive and innovative alternative com-
  pared to simply transporting and disposing the wastes in a dis-
  tant  landfill. The owner of the wastes is  better able to control
  his liability with an on-site facility, and the wastes can be stored
  safely while subsequent testing is conducted to  determine the
  most effective and least costly ultimate treatment  alternative.
• Although  the above-ground silos  are  designed  to effectively
  contain the waste, the owner likely will implement  a double
  synthetic liner system to give added protection.
• An innovative approach to the regulatory process is being im-
  plemented whereby a Part B permit application for the storage
  facility is being submitted prior to the submission of a feasibil-
  ity study addressing site remediation. This approach likely will
  result in a  more expedienl cleanup  of the site relative to a tra-
  ditional approach.
• Innovative approaches to the regulatory process are more like-
  ly to be successful if all  parties  involved in the  process (e.g.,
  owner, regulatory agency) can effectively communicate, coop-
  erate and keep an open mind. This is especially important when
  more than one regulatory agency  is involved.
352    RCRA AMENDMENT EXPERIENCE

-------
                                    Overview  of  the  Proposed
              Natural Resource Damage  Assessment Regulations
                                                Richard J.  Aiken*
                                            Willie R. Taylor, Ph.D.*
                                        U.S. Department of the Interior
                                          CERCLA 301  Project Team
                                                Washington, D.C.
 ABSTRACT
  This paper describes and explains the natural resource dam-
 age assessment regulations issued in proposed form in December
 1985. These regulations were developed by the U.S. Department
 of the Interior under the authority of section 301 (c) of CERCLA.
 The regulations are to be used for assessing  damages to natural
 resources which result from a release of a hazardous substance or
 oil. Claims for such damages can be presented to a responsible
 party under the liability provisions of CERCLA, section 311(f)(4)
 and (5) of the Clean Water Act or the Superfund. Some examples
 of the resources involved include groundwater,  surface water,
 park and recreation areas, wetlands,  wildlife habitat, fish and
 wildlife, air and soil. They are applicable to all Federal and State
 agencies that manage natural resources including, among others,
 the U.S. Department of Defense, State  fish  and game agencies,
 the U.S. Department of Energy and State park authorities.

 INTRODUCTION
  CERCLA and the Clean Water Act are the major federal
 authorities for responding to releases of hazardous substances
 and oil. Although the response authorities and the funding mech-
 anism for these responses constitute the focus for most discus-
 sions of these laws, this paper focuses on the issues embodied in
 the second half of CERCLA's title: compensation and liability.
 Specifically, the paper deals with the liability of the parties re-
 sponsible for  the release for resulting  damages to  natural re-
 sources and the appropriate measurement of the compensation.
 These issues form the nucleus of the proposed natural resource
 damage assessment regulations required by section 301(c) of
 CERCLA and promulgated by the  U.S. Department of the In-
 terior in December 1985.
  Although promulgated by the Department  of Interior, the sec-
 tion 301(c) natural  resource  damage assessment regulations are
 written for the use of all Federal and State agencies which claim
 trusteeship for natural resources  under CERCLA. CERCLA
 broadly defines natural resources to include  all elements of the
 environment, including such living resources as fish, wildlife and
 other  biota and such physical resources as  land, air, drinking
 water and groundwater. Examples of agencies that may make use
 of the regulations include park authorities, fish and wildlife agen-
 cies, and land and water management  agencies.  A rebuttable
 presumption is a legal mechanism which effectively gives  the
 claimant an advantage in any administrative or litigative actions.
 The use of these rules in some instances will provide a rebuttable
 presumption for the assessment.

'This paper represents the views of the authors and  not those of the U.S. Depart-
ment of the Interior.
  Section 301 (c) requires that two types of regulations be devel-
oped; the so-called type A and type B regulations. The type A
regulations are to  be standard procedures for simplified assess-
ments involving minimal field work. The type B procedures are
to include alternative protocols for the assessment of damages On
a case-by-case basis. This paper provides an overview of the en-
tire damage assessment process and focuses on the type B assess-
ment procedures. It does not address the type A procedures.

DAMAGES ARE THE RESIDUAL OF
RESPONSE ACTIONS
  Section 107(a) of CERCLA outlines  the liability of respon-
sible parties for CERCLA-covered releases to include:
• The costs  of removal and remedial  action  incurred by the
  United States Government or a State that are  not inconsis-
  tent with the National Contingency Plan
• Any other necessary costs of response incurred  by any other
  person consistent with the National Contingency Plan
• Damages  for injury to, destruction of, or loss of natural re-
  sources, including  the reasonable costs  of assessing such in-
  jury, destruction or loss resulting from such a release
  Liability under CERCLA can be  summarized to removal, re-
medial actions (jointly referred to as response actions) and dam-
ages (hereafter referred to as "damages"). Liability for the  first
two components  has  long  been  an established  part  of the
CERCLA program. However, liability for damages only recently
has been raised as a concern. A major reason for this situation is
the lack of a consistent understanding throughout all levels of
government, industry and the general public as to what consti-
tutes damages. CERCLA's definition of damages provides little
guidance. Section 101(16) of CERCLA simply states that damages
"means  damages for injury or loss of natural resources..."
  The first boundary on the  term "damages" is provided by the
method in which CERCLA establishes the liability. Section 107(a)
discusses the three elements  of liability (costs by the Federal or
State government,  costs by any other party and damages) as sep-
arate and distinguishable components. Therefore, damages are
those injuries or losses of natural resources that are not covered
by the first two components. In other words, damages are those
residual  impacts that remain after factoring in the effects of the
removal and remedial actions.
  The remainder of this paper looks more closely at the issue of
what constitutes damages. In so doing, the paper discusses what
constitutes "injury to, destruction  of, or loss of natural re-
sources," how these injuries should be measured and finally, the
process  of natural resource damage assessment proposed in the
Department of the Interior's regulations.
                                                                                RCRA AMENDMENT EXPERIENCE   353

-------
WHAT ARE DAMAGES?
  CERCLA requires that damages be determined for the injury,
destruction or loss of a natural resource due to the release of a
hazardous substance or oil. For simplicity, the terms  "injury,"
"destruction," or "loss" all can be referred to generically  as in-
jury since they refer to degrees of injury. In common law, at least
two major categories of damages are recognized: punitive  dam-
ages and compensatory damages.  To determine which  type  of
damage applies to CERCLA cases, both the statute and its legis-
lative history need to be examined.
  That CERCLA requires compensatory, rather than punitive,
damages is clear  from its  language. Section 301(c)  of  CERCLA
requires that an assessment of damages be performed using the
"best  available procedures." Use of best available procedures
would not  be required if damages were to be punitive.  If Con-
gress had meant damages under CERCLA  to be punitive, all that
would be needed is a penalty  table. This penalty table  could
have been but was not included in the legislation.
  The legislative history  of  CERCLA  also  demonstrates that
punitive damages were  not intended. For example, the Senate
Committee on Environment  and Public Works stated that the
type B regulations should  contain "methodologies for  determin-
ing value [for injury to natural resources due to a release of a
hazardous substance or oil]..." This determination of value,  as
discussed below,  is a critical part of determining compensatory,
not punitive, damages.
  Damages are thus  the compensation for an injury to a natural
resource. Liability for this compensation is to the agency acting
as trustee of the  injured natural resource. As compensation for
the injury resulting from the release, the agency is entitled  to be
made  "whole," at least in monetary terms. That is, the agency
should be made approximately as well off as it was prior to the
release.

HOW ARE DAMAGES MEASURED?
  Economics, common law and CERCLA  agree that when meas-
uring damages, making  the agency "whole" should be achieved
in a cost-effective manner. This requirement gives the agency the
lessor  of: (1) the costs of restoration, rehabilitation,  replace-
ment or acquisition  of an equivalent to the injured natural re-
source (restoration for short); or (2) the amount of lost use values
resulting from the release.
  That is, if use values are higher than the  costs of restoration, it
would be rational to restore  the injured natural resource  to its
former condition. In this case, society regains its lost uses  for a
cost that is less than the value of those  lost uses. Conversely, if
restoration costs  are higher  than  the  value  of lost  uses,  it is
rational to compensate society by giving  it  the value of those lost
uses. In the latter case, society receives the value of  the lost uses,
which  is less  than the costs of restoration  of the resource which
would regain those lost uses.
  Federal and State agencies should be required to decide whether
restoration or use values will form the basis of further work  to
determine damages.  This decision  should  be made early in the
assessment process. "Off-the-shelf" information should  be used
to make this decision. Past studies on restoration techniques and
valuation estimates should be used to obtain an "order  of magni-
tude"  estimate of restoration costs and  use values.  These rough
estimates are a surrogate  for the site-specific determination  of
either restoration  costs or the loss of use  values that will be  made
later in the damage assessment process. The sole purpose of this
early estimate is  to determine whether  restoration  costs or use
values will form the basis for further analyses.
  One major reason  for requiring  the decision between  restora-
tion and use values relatively early in the assessment  process is
that it helps focus the assessment process on the determination of
damages rather than on general research. This agreement can be
achieved by an early meeting of the entire group of specialists that
will perform the assessment. In  most instances, these personnel
will have different areas of specialization. These individuals must
work together if reasonable costs, as discussed later in this paper,
are to be achieved. If these individuals do not get together early
in the assessment process, there is a possibility that either the
information collected in the measurement of injury will not be
useful in determining damages, or that the demands for informa-
tion to  determine damages  will  be unrealistic. These problems
have been acute in several previous damage assessments.
  Congress and State  legislatures have determined that certain
natural resources are worthy of protection, even if their use values
are relatively low. In effect, these legislative bodies have removed
these special resources  from the ordinary  workings of economics
and common law. If agencies were held to a strict rule of the
lessor of a loss  of use values or restoration costs, some  of  these
special resources might be left unrestored which would be con-
trary to Congress' or the State legislature's intent. Therefore, it
would be rational to assume that when a special resource  is in-
jured due to a release,  restoration costs would be the measure of
damages.  However, even in these cases, the costs of restoration
should not be grossly disproportionate to the benefits of restora-
tion.
  To determine restoration costs, the Federal or State agency
must determine the services lost due to the release. Natural re-
sources  provide services to other resources and also to people.
When people are the ultimate recipient of the service, the  serv-
ice  is called a use. Restoration costs are the costs of restoring
the lost services previously provided by  the injured natural re-
source. For example, a lake can serve as a  habitat for a fish popu-
lation.  If a hazardous substance is  released into  the lake, the
water may be contaminated to a point where the lake is no longer
habitable  for the fish.  In this  instance, the services provided by
the lake have been lost to the fish population. Restoration would
entail  the restoration of the lake as a suitable habitat for a fish
population. In this example, replacement of the fish population
should not be attempted if the habitat would not be restored to
support the fish population. However, damages based on restora-
tion are limited to the amounts required to return the resource to
its baseline condition. The costs  of enhancement of the resource
beyond  its baseline condition  should not be included in deter-
mining compensation for an injury.  In the above example, this
would imply that the lake be restored to a suitable habitat for the
species of fish that existed in the body of water prior to the re-
lease, not to a species that an agency would like to introduce.
  The  restoration  of  services, not simply  the  replacement of
organisms, is paramount in determining the costs of restoration.
This concept was affirmed in  the decision reached in the Com-
mon wealth of Puerto  Rico v. The SS Zoe Colocotroni, 628 F.
2d 652, 673 (1st Cir. 1980) where  the cost of replacement was held
not to be a measure of damage if the replacement is not  contem-
plated.
  In determining damages based on restoration costs, the loss of
use values, over the time required for restoration to be completed,
would also be compensable  to the agency. This compensation is
owed because society suffers a real loss by not being able to utilize
the resource at its fullest potential during the period of time re-
quired for its restoration.
  In addition to defining restoration in terms of services, agencies
must consider the ability of the resource to recover naturally. Sec-
tion 301(c)(2) of CERCLA requires that  the "ability of the eco-
system or resource to  recover" be factored into the determina-
tion of  damages. The  time required for  the resource to recover
354    RCRA AMENDMENT EXPERIENCE

-------
naturally will provide the period during which a loss of use will be
experienced  if no restoration actions are performed. Thus, the
natural ability of the aforementioned lake to recover must be in-
cluded in determining restoration costs.
  Uses, as stated above, refer to the human use of services pro-
vided by natural resources. In the example, fishing may have been
a use of the lake. If so, the loss of fishing  could be an alterna-
tive to restoration  costs in determining damages.  Common law
and economics dictate that an ordering of methodologies be used
to value injuries to natural resources under CERCLA. This order-
ing would start with the change in market prices when applic-
able.  Second, if market prices are not applicable, appraisals
should be used. Third, if neither prices nor appraisals are applic-
able, certain types of non-market methodologies should be used.
The  logic for this ordering is that  if a market or a  comparable
market  exists for the  injured natural resource, the information
from these markets always should be used to  value the resource.
  If neither the market price nor appraisal methodology is appro-
priate, a non-market methodology that measures  either willing-
ness to pay or willingness  to  accept compensation would be
appropriate. Willingness to pay and willingness to accept com-
pensation are the fundamental concepts which underlie the com-
pensation principle in economics. As such,  any methodology to
determine compensation must be  consistent with one  of these
concepts.
  A paramount concern in determining damages  is  the type of
values for which the agency may be compensated. CERCLA pro-
vides that a Federal or State agency is acting as a trustee when
seeking to recover  for an  injury to a natural resource  resulting
from a release. As such, the agency can collect only losses to the
public at large because of the release. These include:  (1) losses in
recreation or other public uses;  (2) losses in fees and other pay-
ments made to the public for the private use  of the public's re-
source; and (3) economic rent lost due to the  release.
  The third type of loss may need additional explanation.  Eco-
nomic rent is the excess  of the total earnings of a producer  of a
good or service over the payment required to induce that producer
to supply the same quantity currently being supplied.  This  eco-
nomic rent usually exists because the agency  does not  charge a
price or fee for the private use of the  public's resource. These
amounts are compensable to the agency as  the use value of the
resource in place, i.e., the value of the natural resource in the
production of marketed goods  when the government does not
charge a fee or price for the use of its resource.
  Two categories of revenues associated with the  private use of
the public's resource  should be excluded in  the calculation of
damages. First, losses in tax revenue should not be  included in
damages. Taxes are transfers of income from individuals to the
government. As such, taxes are  not considered "real" damages.
Second, wages and other income lost by private individuals, with
the exception of economic rent, should not be compensable to the
agency. This income does not accrue to the  agency or the public
at large and may be subject to privately brought non-CERCLA
law suits.
  To determine damages,  the  existing and planned uses  (and
services) of  the injured natural resource must be determined.
The  determination of damages  is dependent  upon the uses (or
services) forgone because of the release. The basis for determin-
ing damages should be the committed use of the resource prior
to the release. Committed use would mean that a current or plan-
ned use has  been documented through a financial, legal, budge-
tary or other similar type of commitment made by the agency or
some other individual or body with the authority to  commit the
use of the resource. That agencies must be able to document cur-
rent  and prospective uses follows  from the need to determine
compensatory damages. To determine compensatory damages,
purely speculative uses of the injured natural resource cannot be
allowed.

WHAT CONSTITUTES AN INJURY
TO A NATURAL RESOURCE?
  To this point, the discussion has focused on methods for cal-
culating compensation for damages to natural resources. How-
ever, damages are for injury to a natural resource. Therefore, to
calculate a damage, an injury first must be established.
  The first step in identifying what constitutes  an injury is to
provide a general definition  of the term. This definition should
develop from  the purposes to which the definition will be put.
The purpose to which this concept of injury will be put is to pro-
vide justification for a compensation. Since the level of compen-
sation will be tied to the level of injury, the first element that must
be included in the concept of injury is measurability. Compen-
sation obviously implies a loss or some reduction. Therefore, the
second element of this definition, this reduction, can be viewed
as a decrease  in those basic characteristics that constitute the
quality of that resource. Finally, in the discussion of the liability,
CERCLA states that the injury must have resulted from the re-
lease. Therefore, injury can  be viewed as a measurable adverse
change in the  physical, chemical or biological characteristics of
the resource which resulted from the release.
  To this definition should be added CERCLA's guidance that
requires the inclusion of long-term and short-term as well as direct
and indirect injuries.  In addition, the term "resulting from the
release"  would have to include products of the substances  re-
leased and reactions caused by such products or substances.
  This general definition needs to be translated into specific re-
source definitions for applicability. As stated earlier, CERCLA
covers a wide  range of natural resources. These  resources need
to be grouped into some manageable number that  still provides
meaningful guidance.
  The emphasis on physical, chemical and biological characteris-
tics in the general definition suggests distinguishing between
physical and biological resources.  Although biological resources
can be considered as one group of resources,  physical resources
are more easily discussed as  either air, groundwater,  surface
water or geologic resources. All of these resource groups include
a wide range of resources. However, the type of guidance that
can be provided in such a general set of regulations suggests that
these five groups of resources are adequate.
  The physical, chemical or biological characteristics of concern
in an injury  definition would be those characteristics that relate
to the quality of the resource relative to the services that it is ex-
pected to provide. The concept of services as explained earlier
includes such services  to people and the environment as drinking
water supplies, habitat and growth medium for vegetation. The
most obvious services that can be considered include those link-
ages that exist between resources that may cause exposure  and
injury to other resources. In the earlier example, a body of water
may have been contaminated  such that the fish within that body
of water die. Obviously, the fish are injured, but the  water can
be considered injured as well since it no longer serves as a healthy
environment for the  resources that depended upon it. Similar
linkages  exist  among  all of the resource groups.  These linkages
are easier to identify in some resources than in others.  However,
in no cases should the linkages between the  resource groups con-
stitute the only method of identifying an injury.
  Several of the resource groups include Federal  and State stan-
dards which can be viewed as defining a characteristic of quality
for the resource. The clearest sample of such standards are the
water quality  criteria developed  under the Clean Water Act.
                                                                                     RCRA AMENDMENT EXPERIENCE    355

-------
These standards were developed with a view toward the services
that the water provided. Frequently, these standards were focused
on  the  potential affects on aquatic life.  Similar  quality stan-
dards or threshold  levels exist  for drinking water  and air. For
geologic and biological resources, equivalent standards have had
to be developed.
  Geologic resources include such  elements of the  Earth's crust
as soils, rocks and minerals. The major services provided by geo-
logic resources  include providing a medium for growth of vege-
tation, a habitat for biota (soils), a source of minerals for devel-
opment purposes and the unsaturated zone for aquifers.  After
a release, geologic resources also may qualify as hazardous sub-
stances under the provisions of RCRA. The physical,  chemical
and biological characteristics of geologic resources that affect the
quality of the  resource to provide such services (i.e.,  respira-
tion, pH or microbial populations) can be defined and used as
an easily identified standard to judge if the resource has been in-
jured.
  The primary  service of concern in biological resources is con-
tinuation of the species or  population viability. However, for a
small but important  subset of biological  resources (those that
are consumed)  a resource should be considered injured if con-
centrations of the substance released within the edible portions of
the resource  exceed applicable  Food and  Drug Administration
(FDA) action or tolerance levels  or equivalent State standards.
  Examples of  injuries that affect population viability are death,
disease,  behavioral  abnormalities,  cancer,  genetic mutations,
physiological malfunctions and physical deformations. Unfortun-
ately, determining whether or not one of these injuries has occur-
red as a result  of the released substance is difficult. Studies on
the toxicity of  chemicals on  various non-human biological re-
sources are not widespread and frequently are not universally
accepted. Examples abound of results appearing in  laboratory
situations that  cannot be  replicated  in field studies and vice
versa. However, if the use of the procedure is to provide a rebut-
table presumption to the assessment, it is critical to include only
those responses that are caused by the substance released. If a
procedure cannot prove that a particular chemical  caused an
effect,  the procedure  should not  be used in any litigative  or
administrative action. Therefore, the specific definition of injury
to a biological  resource is necessarily tied  heavily to the testing
procedure for the biological response.
  This  last point highlights a dilemma inherent in CERCLA's
damage assessment requirement. CERCLA requires that the regu-
lations include  the  best available procedures for assessment  of
damages.  Clearly, it  is not advisable to  hinder research  into
natural resource injuries caused  by chemical releases by requiring
the use of the few techniques currently available, given the  state-
of-the-art of assessment techniques. In addition, it is not feasible
to include every testing procedure that has ever been attempted
in a set of regulations that  is general in nature. The solution to
this problem is to provide criteria by which the techniques can be
judged rather than specific techniques. Such criteria would neces-
sarily center on  the general acceptability of the procedure and the
replicability of the results.
  Identifying that a resource has suffered an injury is insufficient
for the  purposes of establishing a  damage  claim. CERCLA re-
quires that some linkage between the source of the release and the
injured resource be established. This process requires the identifi-
cation of a potential pathway through which the substance  could
have traveled. Each of the resource groups has  the capability of
acting as a component in this pathway.
Substances may  be transported through  direct exposure,  the
food chain or via one of the air or water mechanisms. Model-
ing  or sampling would be appropriate means of determining the
pathway.
   In  summary,  injury  for the  purposes  of  liability  under
CERCLA  can be viewed  generally as a measureable, adverse
change in the physical, chemical or biological characteristics of
a  resource that resulted  from a release. This general definition
can be applied to specific groupings of resources so that clearly
recognizable affects can  be attributed to the substance released.
Once identified as injured, these resources can be measured, and
the extent of the impacts can be calculated.

HOW IS INJURY MEASURED?
   The measurement of injury is the link between the physical and
biological demonstration of injury and  the monetary determina-
tion  of damages. Injury determination demonstrates that either:
(1) a  physical or  chemical  based standard has been  violated;  or
(2) an adverse biological response is the result of a release. This
injury determination  must then be measured  to calculate how
much of the resource has been injured, how badly the resource
has been injured and over what time period the injury has taken
place. This process is the quantification of the injury.
   One major task in the quantification of the injury is to deter-
mine  the changes in services previously provided by the injured
resource. To quantify the effects of the release, the agency must
determine baseline services; the amount, type and quality of serv-
ices that would have been supplied by the injured natural  re-
source in  the absence of the release. This amount is then com-
pared to the amount, type and quality of services supplied by the
injured natural resource.  The difference between these two quan-
tities  is the reduction in  services  that an agency can use as the
basis  for either restoration  costs or use values when determining
damages.  Since damages are the residual after any response ac-
tions, the quantification of the injury must incorporate the effects
of any actual or anticipated response actions on the current and
future services of the resource.
   Quantification  extends to baseline,  rather than standards or
conditions causing the biological responses,  and is a direct con-
sequence  of  providing  compensation  to   make  the agency
"whole." Because baseline is  the  point from which compensa-
tion  is determined, the agency is being compensated  for what
would have occurred had the release not occurred.
   As mentioned  earlier, any quantification of injury must in-
corporate the ability of the resource to recover naturally. In addi-
tion, recovery rates of the resource with alternative management
actions  need to be specified. This calculation allows specifica-
tion  of alternative costs of restoration which, in turn, allows
selection of the most  cost-effective alternative to restore the re-
source.

THE NATURAL RESOURCE DAMAGE
ASSESSMENT PROCESS
   The concepts of damage and  injury are important for the
understanding of the  proposed natural  resource damage assess-
ment  process. However, there are three additional important
concepts that surface  throughout the proposed natural resource
damage assessment process. These concepts,  discussed briefly be-
low, include:

•  Cost-effectiveness and reasonable costs as  they apply to the
   assessment process
•  The involvement of the public and the potentially responsible
   party in the damage assessment process
•  The application of any awards for restoration purposes

Cost-Effecllveness/Reasonable Costs
   CERCLA requires  that natural  resource damage assessments
achieve reasonable costs.  CERCLA's legislative history also indi-
cates that these assessments be cost-effective.  Cost-effectiveness
is  the achievement of an objective with the  least expenditure of
356    RCRA AMENDMENT EXPERIENCE

-------
financial or other resources. Agencies should use cost-effective-
ness  in  determining  how  they  will perform  all aspects of the
assessment process. To achieve this requirement, agencies must
have well-defined goals. For example, testing and sampling for
injury to surface water should be limited to those tests which can
determine the injury  as opposed to any general research studies.
Given local conditions, the agency should select, from the set of
tests that could show injury, the one which is the least costly.
  Section 107(a)(4)(C) of CERCLA stated that a responsible
party is liable for the "reasonable costs of assessing..."  injury
as well  as the amount of damages determined in the damage
assessment. The concept of reasonable costs incorporates the con-
cept of cost-effectiveness, but is broader in scope.  Reasonable
cost  criteria require  that the benefits of alternative testing or
estimation procedures outweigh their costs.
  In addition, reasonable cost criteria require that the parts of an
assessment plan be coordinated  and supply only the information
necessary to move from  one section to another. This require-
ment exists because the sole purpose of a damage assessment is
to determine the damages caused by a release.

Involvement of the Public and
Potentially Responsible Party
  As indicated by the previous discussion,  a damage assess-
ment requires a significant number of choices. The performance
of a damage assessment requires the selection of methodologies
to:  (1) test and  sample for injury; (2)  determine the pathway
which exposed the resource to the release; (3) measure of the ex-
tent of injury; and (4) determine a dollar amount due the agency
as compensation for the injury.  Given the universe of methodol-
ogies that exists, the selection of appropriate methodologies to
perform all of these tasks is difficult.
  The  amount of discretion available argues for the use of an
administrative process in the damage assessment. Administrative
processes are characterized by their openness and ability to gain
the views of all interested individuals. The outcome of such a pro-
cess generally is perceived to be fair because all parties have an
opportunity  to review and comment on the process. This per-
ception of fairness is an important issue in the performance of a
damage assessment. If the courts do not perceive the assessment
process as being an impartial assessment of damages, the likeli-
hood that the results of the assessment will  be successfully chal-
lenged in court increases. This increases costs for all involved.
  Any  administrative process to determine damages must give
final authority to the agency performing the assessment for the
selection of methodologies. Input from the public and potentially
responsible parties should be useful in selecting methodologies
for and actual performance of damage assessments. In addition,
there may be instances when the potentially responsible party may
wish to perform the damage assessment. However, the agency
generally will have the management authority for the resources
in question and must have the  final  authority to determine  all
aspects of the assessment  process as an extension of its manage-
ment responsibilities.

The Application of Any Awards for
Restoration Purposes
  One final concept that  needs  to be highlighted in the process
is the requirement in CERCLA that all  funds received as  com-
pensation for the natural resource injuries be made available for
restoration. This concern for the disposition of the awards was
taken very seriously in the development of the proposed rule and
is considered critical for  the general credibility of the damage
claim. In the Zoe Colocotroni case, one reason that the courts
disallowed the initial claim was the determination that the gov-
ernment had no intention to restore the resources, and the re-
placement of resources formed  the basis of the claim. There-
fore, the commitment to restoration should strengthen the accep-
tability of the damage claim.
  Obviously, restoration is not always feasible. In those instances
when restoration is not feasible, the acquisition of resources pro-
viding services similar to those lost is a viable alternative.

PROPOSED NATURAL RESOURCE
DAMAGE ASSESSMENT PROCESS
  The natural resource damage assessment process proposed by
the Department of the Interior is  designed to be performed in
phases. The more costly phases occur after several screens have
been passed to ensure that assessments  are not performed need-
lessly. The major steps in the  process are provided on Table 1.
The remainder  of this paper will briefly describe each of these
steps.
Initiation of Process
  Natural resource damage assessments should be a follow-on to
any removal or  remedial  actions taken by the U.S. EPA,  the
U.S. Coast  Guard or any other agency designated as  the lead
agency under the National Contingency Plan. That plan requires
that the lead agency notify trustees of any potential natural re-
source injuries. In instances where  the agency identifies a pos-
sible injury  to resources under  its  jurisdiction and suspects a
CERCLA or CWA covered source,  the agency is directed to  the
procedures in the National Contingency Plan for reporting  the
release. Regardless of how the agency was  notified, once alerted
to the potential injury, the agency should determine whether any
emergency restoration actions are necessary.

Emergency Restorations
  Section lll(i)  of CERCLA provides authority for emergency
restorations. Emergency restoration actions are those limited ac-
tions necessary to abate a  continuing threat to or the immediate
destruction  of the resource. The agency should notify the Na-
tional Response  Center of the need  for emergency action. If the
U.S. EPA or the U.S. Coast Guard cannot take the necessary
actions, the agency may use whatever authorities it has to pre-
vent the  transport of the substance to the  resource. Upon com-
pletion of the limited actions, the agency would start the natural
resource damage assessment process.

Pre-Assessment Screen
  The first  step  in the assessment process is the pre-assessment
screen. The purpose of this screen  is to determine whether  the
release justifies a natural resource damage assessment. This screen
is viewed as a "desk  top"  review of existing data, capable of be-
ing completed in a matter  of days. It should not duplicate or re-
peat information gathered by the agency or by other parties as
part of the response action.
  A documented determination is required upon completion of
this screen. The  determination to proceed will be based upon en-
suring that:
• The release was covered by CERCLA or the CWA
• It  might have  resulted in some injury to a resource under the
  jurisdiction of the agency
• The resource and the extent of potential injury was of sufficient
  concern to the agency to continue
  If,  as a result of the pre-assessment screen, the agency deter-
mines that a natural resource damage assessment is in order, the
agency will initiate the Assessment Plan phase.
                                                                                     RCRA AMENDMENT EXPERIENCE    357

-------
                             Table 1
             Natural Resource Damage Assessment Process
PREASSESSMENT PHASE


Initiation of process
Decision on whether
 emergency exists
Determination of
 whether to proceed
 with a damage assessment
ASSESSMENT PLAN PHASE


Planning for the assessment


Decision on A or B
                                    Injury to Resource
                                    Suspected CERCLA or
                                    CWA Source
                        Notification	|
                            Not
                          Confirmed
                          End
Confirm presence in
 resource (Type B only)
Restoration costs or use
 value (Type B only)

TYPE B ASSESSMENT
Determine whether an injury
 has occurred that is linked
 to CERCLA Of CWA release
Review of planned methodologies,
 especially Economic Methodology
 Deteonination/ in light of results
 of Injury Determination Phase
                               Assessment Plan
                                   Phase
                               Type A or Type B


                            Confirmation of Exposure
                                     Economic Methodology
                                        Determination
                                                            Type A
                                        Type
                                         B
                                        Injury Determination
                                             Phase
                                     Review
                                 Assessment Plan
                      CERCLA or CWA Injury
                      Not Confirmed	
                    End

Quantification of effects
 of release
 -Identification of  service';
 -Baseline
 -Recovery time and  ability
Estimate of loss of value
 or restoration or replacement
 costs
PO6T-ASSESSMENT PHASE

Documentation of results of the
 assessment

Establishment of trust fund and
 development of Restoration Plan
                                  Quantification
                                     Phaae
                                Damage Determination
                                    Phase
                                Report, of Assessment |
                             I     Post-Assessment    |
Assessment Plan Phase
  The agency should document its decisions on the selection of
the methodologies that will be performed.  This documentation
must  be given in an Assessment Plan. This  Assessment  Plan
should ensure that the mandate to keep assessment costs reason-
able will be fulfilled.
  There are several  efforts that  need to be undertaken  by the
agency  prior to developing the Assessment Plan. These  efforts
center on  initiating the  involvement of multiple trustees,  poten-
tially responsible parties and the public in the assessment.
  The potentially responsible party or  parties  should be  identi-
fied. A Notice of Intent to Perform an Assessment must be sent
to any identified potentially responsible party (PRP). This notice
is intended to involve the PRP early in the assessment in an at-
tempt  to  minimize  disagreements in the performance  of the
assessment. However, all decisions concerning the assessment are
ultimately the responsibility of the lead official.
  The public is encouraged to become involved  in the Assessment
Plan.  The  public participation element of the  Assessment  Plan
includes a 30-day review period prior to implementation  of the
plan or major modifications. Any comments and responses must
be maintained as part of the administrative record.

  There are  several requirements  specific  to  a  type B  assess-
ment  that  must be included in the  Assessment Plan. These re-
quirements include:
• Confirmation of exposure
• Quality Assurance Plan
• Objectives of testing and sampling for injury or pathways
• Economic Methodology Determination

  The confirmation  of exposure is the second screen in the dam-
age assessment process. It is intended to ensure prior to initiating
an expensive type B assessment that the released  substance has
actually come  into contact with the resource of concern. The
Quality Assurance Plan should  be prepared following the same
requirements that apply to other actions taken under CERCLA.
The testing and sampling objectives should clearly state the pur-
poses of all proposed testing and sampling.
  The Economic  Methodology Determination is where the agency
must  make a  choice between using the costs of either restoring
or replacing the injured resource of the value of lost uses as the
measure of damages.  The  decision will  affect the choice  of
methodologies to be selected in the second phase of the assess-
ment, referred to as  the Quantification phase, and to a lesser ex-
tent in  the first phase, referred to as the  Injury Determination
phase. Therefore, the decision must be made early, but may be
deferred or changed  after the Injury Determination phase is com-
pleted.
   The selection of use value or restoration costs only affects the
method of damage determination. It does not imply any decisions
concerning whether  the resources will be restored.  Except in the
case of defined special resources, the method of determining the
damage will be the lesser of the lost use value or restoration costs.

Type B Assessment
   Performing a damage assessment by using a type B procedure
involves three general steps:
•  Establishing that an injury has occurred and  that the injury re-
   sulted from the release
•  Quantifying the effects of the release on the injured resource
•  Determining the damage
 358
RCRA AMENDMENT EXPERIENCE

-------
Injury Determination Phase
  This phase of the type B assessment acts as the third screen of
the natural resource damage assessment. To assert a natural re-
source damage claim, the agency would have to establish that an
injury occurred or to link that injury to the release. The Injury
Determination phase is where the agency must determine whether
one of the specific definitions  of injury for the five resource
groups has been met. Acceptance criteria for determining when a
procedure meets the requirements for the rebuttable presumption
are provided to assist the agency in making this determination.
Finally, it is in this phase that the agency must determine the path-
way that the substance  could have traveled to reach the  injured
resource.

Review of the Assessment Plan
  Upon  completion  of  the Injury Determination phase,  the
agency will review  the methodologies selected in the Assessment
Plan. The purpose  of this review is to ensure that the selection of
methodologies for  the last two phases of the type B assessment
is compatible with the findings  of that first phase. If an injury
cannot be determined or  cannot be linked to the release, the re-
sults of the  Injury Determination phase should be documented
and any further assessment efforts included in the plan should be
deleted.
  If it is determined that the injury resulted from the release, the
agency will ensure that the methodologies selected for the next
two phases are consistent with the findings of the Injury Determi-
nation. At this stage, if the agency has not previously done so, a
decision must be made on whether use value or restoration or re-
placement costs will  form the basis of the damage determina-
tion.

Quantification
  Having established that a resource was injured by the  release,
the second step in the type B assessment is to quantify the effects
on the injured resource. As the purpose of the natural resource
damage assessment is compensation rather than a decision on the
cleanup level, this phase requires ascertaining the  baseline level
for the resource. The baseline level is compared  to the level pro-
jected upon  the completion of any response actions to determine
the residual change resulting from the release. The baseline level
will include consideration  of  the resource's   natural  cyclical
changes.

Damage Determination Phase
  The outputs of the second step should act as inputs to the third
general step in the type B process, estimating the damage. The
method of estimating the damage, either using the cost of restora-
tion or replacement or the loss of use, was determined earlier dur-
ing the development of the Assessment Plan.
  If restoration or replacement is  to be the measure, a plan for
the restoration or  replacement, referred  to as  the Restoration
Methodology Plan, must be developed in  this  Damage Deter-
mination phase. This plan must have sufficient  detail to ensure
that all major elements of costs are included and that these costs
represent the cost-efficient means  of restoring or replacing the
services lost  as a result of the release. This plan also will serve as
the foundation for the final restoration plan that must be devel-
oped after the award.
Report of Assessment
  In some instances, damage assessments using this natural re-
source damage assessment process will receive a rebuttable pre-
sumption. To assert this rebuttable presumption, at the conclu-
sion of a type B assessment, the agency must document the results
of the major steps of the  process in  a report referred to as the
Report of Assessment. These major steps include the Pre-Assess-
ment  Screen Determination, the Assessment Plan with all com-
ments and responses, the Injury Determination, the Quantifica-
tion Determination and the Damage Determination, including the
Restoration Methodology Plan if prepared.

Post-Assessment
  Based upon CERCLA's  mandates,  funds recovered as a result
of the assessment process provided in the proposed rules must be
available for restoration, rehabilitation, replacement or the acqui-
sition of the equivalent of the injured resource. To accomplish
this mandate, a trust fund must be established to hold all awards
except those amounts which represent reimbursement of costs in-
cluding  assessment, administrative and litigative costs.  These re-
imbursements may be returned to the general treasury of the gov-
ernmental unit that incurred the costs. The agency shall prepare
a Restoration Plan. This plan shall be based upon the decisions
made in the Restoration Methodology Plan, if one has  been pre-
pared; the  plan shall be  modified to the extent necessary to
accommodate new information before calculating the amount of
the award. The trust fund  will be used to pay for the implemen-
tation of this Restoration Plan.
  In recognition of the fact that restoration of some injured re-
sources  is technically infeasible,  replacement and acquisition of
the equivalent are defined to include acquisition of resources that
provide similar services to the injured resource. However, there is
a limitation on the use of the fund. Where acquisition of similar
resources would involve purchase of  land  for federal manage-
ment, the award must be paid to the general treasury. The appro-
priations process must be  used  where land is being acquired to
expand the total federal landholdings.

CONCLUSIONS
  CERCLA and the  Clean Water Act authorize liability for re-
sponse actions and damages to natural resources caused by re-
leases of hazardous substances and oil. This paper has  examined
the extent of the liability of parties responsible for the release and
resulting damage to natural resources and the proposed  process
through which these damages can be calculated.
  The proposed process is driven by  two  basic premises:  dam-
ages under CERCLA are  compensatory;  and a rebuttable pre-
sumption attaches  to assessments performed in accordance with
these  regulations. The fact that damages are compensatory stems
from  a reading of the law and its legislative history. The rebut-
table presumption is specified in section 11 l(h)(2) of CERCLA.
  The concepts discussed in this paper are the basis of Interior's
proposed natural resource  damage assessment regulations. These
regulations provide a proposed procedure  for agencies to deter-
mine  damages under CERCLA. It is hoped that by exploring the
issues of damages, injury  and the appropriate measure of these
terms, the proposed rule has initiated a discussion of the appro-
priate role of natural resource damages in the overall hazardous
material control strategy.
                                                                                     RCRA AMENDMENT EXPERIENCE    359

-------
                               Regulatory  Impact  Analysis  for
                                    the Toxicity Characteristic

                                                 John  L. Warren
                                         Center for Economics  Research
                                           Research Triangle  Institute
                                    Research Triangle Park, North Carolina
ABSTRACT
  The U.S. EPA is expanding the use of the toxicity character-
istic as a means of identifying hazardous waste.  Because of the
projected impact of this new regulation, a detailed Regulatory
Impact Analysis (RIA) was prepared to support the development
of the new regulations. This RIA addresses the benefits and costs
of the regulation including a benefit/cost assessment, risk assess-
ment and uncertainty analysis. This paper provides an overview
of the methodology  employed and the results obtained; it does
not represent the views of the U.S. EPA.

INTRODUCTION
  The U.S. EPA is  proposing to  expand its  regulation of haz-
ardous wastes to comply with  RCRA by regulating additional
wastes. Executive Order 12291 requires regulatory agencies to
maximize the net social benefit of regulations and  to conduct a
Regulatory Impact Analysis (RIA)  for any major rule. A major
rule is one likely to result in one or more of the following:
• An annual effect on the economy of $100 million or more.
• A major increase in costs or prices for consumers; individual
  industries; Federal, State or local government agencies; or geo-
  graphic regions.
• Significant  adverse effects  on competition, employment, in-
  vestment,  productivity, innovation or the ability of United
  States-based enterprises to compete in d< mestic or export mar-
  kets.

  This RIA provides an analysis  of tne U.S. EPA's  proposed
regulation based on the guidelines  contained in the Office of
Management and Budget's "Interim Regulatory Impact Analysis
Guidance"  (June 12, 1981) and the U.S. EPA's Guidelines for
Performing Regulatory Impact Analysis (1983).

PROPOSED ACTION
  The U.S. EPA is proposing to expand the existing Extraction
Procedure Toxicity Characteristic (EPTC) (40 CFR 261.24). This
characteristic currently involves a leaching test (known as the Ex-
traction Procedure or EP), which is used to determine how much
a toxic chemical would leach if a  waste containing it  were  dis-
posed of along with municipal wastes in a sanitary landfill. A
problem with  the EP toxicity test is that it originally was not de-
signed to  measure leaching  of organic chemicals from solid
wastes. It was designed primarily to model the leaching of metals.
  Currently, wastes containing toxic constituents are designated
as hazardous primarily through the listing mechanism. Listing is a
slow,  resource-intensive  process that  is specific  to  hazardous
wastes generated in a particular industry. At present, only a small
percentage of potentially hazardous wastes containing toxic con-
stituents are listed  because the U.S. EPA has had limited  re-
sources to devote to the  industry studies program that prepares
the listings.
  The U.S. EPA perceived a need for a more comprehensive
and self-regulating means of identifying wastes as hazardous and
believed the best way to meet this need is  to expand the existing
EPTC. The expanded EPTC (known as the Toxicity  Characteris-
tic Leaching Procedure—TCLP) applies to  all wastes, not just
metals and pesticides, and requires the waste generator to deter-
mine if its waste exceeds an acceptable concentration level.
  The U.S. EPA is proposing to initially regulate the 52 chemicals
listed in Table 1. These chemicals have been selected for regula-
tion on the basis of:

• Existence of an acceptable exposure level
• Existence of adequate fate and transport data needed to run the
  groundwater transport  model

  For a given toxic constituent, the U.S. EPA proposes to de-
termine  a toxicity threshold of that constituent in groundwater.
For noncarcinogens, the  toxicity threshold can be approximated
using a percentage of the acceptable daily intake (ADI). The per-
centage  used depends  on the extent to which simultaneous ex-
posure through air and ingestion occurs. The ADI is the amount
of a chemical (in mg/day) that can be ingested every day for a
70-year lifetime without adverse health effects. (Criteria for estab-
lishing ADIs are discussed in 45  FR 79318.) The resulting frac-
tion of the ADI is referred to as the Health Based Starting Limit
(HBSL). For  carcinogens,  the U.S. EPA's Carcinogen Assess-
ment Group (CAG) has calculated the probability of cancer aris-
ing per unit dose. This unit cancer risk (UCR) is the risk of cancer
for a dose of 1 mg/kg per body weight/day over a  70-year life-
time. The dose corresponding to a cancer risk of 1 in 10* for each
carcinogen is  derived from the UCR.  That level is  the virtually
safe dose (VSD). For  carcinogens, the HBSL can  be approxi-
mated using the VSD.
  Once the HBSL is established for a toxic waste constituent and
the degree of attenuation between the disposal site leachate and
the well water is estimated, the level of the constituent in a leach-
ate that yields constituent concentrations at or above  the HBSL at
a well near a land disposal facility is calculated.  That level is called
the regulatory level (RL). The RL determines the leachate con-
centration of a constituent (as measured by the TCLP) that would
cause the waste containing that constituent to be considered haz-
ardous under RCRA.  Any waste that has a  TCLP concentra-
tion greater than its RL  is hazardous  by  definition and subject
to regulation under  RCRA as a hazardous waste.
360   RCRA AMENDMENT EXPERIENCE

-------
                           Table 1
      Selected Chemicals Considered for Regulation Using TCLP
  Acrylonitrile
  Arsenic
  Barium
  Benzene
  Bis(2-chloroethyl)ether
  Cadmium
  Carbon bisulfide
  Carbon tetrachloride
  Chlordane, technical
  Chlorobenzene
  Chloroform
  Chromium(VI)
  o-Cresol
  m-Cresol
  p-Cresol
  2,4-D, salts and  esters
  1,2-Dichlorobenzene
  1,4-DiChlorobenzene
  1,2-Dichloroethane
  1,1-Dichloroethylene
  2,4-Dinitrotoluene
  Endrin
  Heptachlor
  Hexachlorobenzene
  Hexachl orobutadiene
  Hexachloroethane
Isobutanol
Lead
Lindane
Mercury
Methoxychlor
Methylene chloride
Methyl  ethyl ketone
Nitrobenzene
Pentachlorophenol
Phenol
Pyridine
Selenium
Silver
1,1,1,2-Tetrachloroethane
1,1,2,2-Tetrachloroethane
Tetrachloroethy1ene
2,3,4,6-Tetrachlorophenol
Toluene
Toxaphene
1,1,1-Trichloroethane
1,1,2-Trichloroethane
Trichloroethy1ene
2,4,5-Trichlorophenol
2,4,6-Trichlorophenol
2,4,5-TP
Vinyl chloride
 REGULATORY ALTERNATIVES ANALYZED
   Benefits and  costs  are  determined  for  the  following  three
 selected regulatory alternatives and the Status Quo (i.e., no regu-
 lation):

 • Alternative 1: Includes all currently unregulated wastes that
   would produce a TCLP extract containing any of the 52 con-
   taminants  at a level greater than or equal to  100 times the
   chronic health based threshold.
 • Alternative 2: Includes all currently unregulated wastes that
   would produce a TCLP extract containing any of the 52 con-
   taminants  at a level greater  than  or equal to  10  times the
   chronic health based threshold.
 • Alternative 3: Includes  all currently regulated wastes  that
   would produce a TCLP extract containing any of the 52 con-
   taminants at a level greater than or equal to 1,000 times the
   chronic health based threshold.
 • Status Quo: No regulation.
   Benefits and costs for each regulatory alternative are compared
 to those of the baseline Status Quo. Under the Status Quo, indus-
 try does  not incur additional waste management costs for  more
 stringent requirements. However, society will  incur the costs of
 not regulating these wastes. These social costs  of the Status Quo
 are assumed  to be  the benefits  that would occur if  the wastes
 were regulated; they vary with the number of affected facilities.

 AFFECTED INDUSTRIES
  Because the proposed action is chemical-specific rather than in-
 dustry-specific, it affects  a wide range of industries. Table 2
 shows the affected  industries by Standard  Industrial Classifica-
 tion (SIC) code. Most of the plants that produce and use the pro-
posed chemicals appear in the  chemical industry. Any facility
that has one of the selected chemicals in a waste stream at con-
centrations greater than the RL will be affected.  (Those wastes
currently regulated under RCRA are not included in the analysis.)
The number of affected facilities may include plants that produce
or use more than one of the chemicals.

ECONOMIC IMPACT METHODOLOGY
  A partial equilibrium multimarket (PEM) model is employed
to project the economic impacts of the regulatory alternatives.
All 11 industries that would be directly affected by the regulatory
alternatives are modeled.
  The primary objective of the PEM model is to project changes
in equilibrium  prices and output rates in both directly  and in-
directly affected industries. Related impacts on industry employ-
ment also are projected by the model.
  The projected economic impacts are of interest for two reasons.
First, they indicate the types and magnitudes of industry changes
that might be expected as a consequence of the regulatory altern-
atives.

                           Table 2
                    Directly Affected Sectors
        Industry
                                                   SIC Code No.
Plastics materials  and resins

Synthetic rubber

Medicinals and botanicals

Soap and other detergents

Surface active agents

Paints and allied products

Cyclic crudes and intermediates

Industrial organic  chemicals

Agricultural chemicals

Petroleum refining

Nonferrous wire drawing and insulating
2821

2822

2833

2841

2843

2851

2865

2869

2879

2911

3357
                               Second, the price and output rate impacts represent resource
                            reallocations  that impose  costs on affected parties.  The  pro-
                            jected equilibrium impacts are thus the key ingredient in the regu-
                            latory cost projections.
                               As modeled, industry supply curves are perfectly elastic. Regu-
                            latory price impacts thus reflect production  cost changes. The
                            model accounts for both direct cost changes attributable to con-
                            trol and indirect cost changes attributable to inter-industry trans-
                            actions.
                               Demand curves for outputs of  modeled  industries are down-
                            ward sloping.  When product prices increase, c uantities demanded
                            decline. Production rates  decline to restore  equilibrium, and
                            changes in employment and  capital expenditures  result.  The
                            model does not account for demand shifts that might occur, but
                            these are believed to be insignificant.

                            BENEFIT ESTIMATION METHODOLOGY
                               Regulation  of wastes containing any one of the selected chem-
                            icals should result in a reduced risk of groundwater contamina-
                            tion and  subsequently contamination of the drinking water  of
                            many communities. If the contaminating chemical is  a carcino-
                            gen, drinking  the water may result in an excess incidence of can-
                            cer cases in the population. Ingestion of noncarcinogenic chem-
                                                                                       RCRA AMENDMENT EXPERIENCE    361

-------
icals in drinking water at a level  above the ADI may be corre-
lated with toxic,  reproductive or genetic effects,  depending on
the particular chemical. Switching to an alternative water source
to avoid drinking contaminated groundwater imposes substantial
costs for the affected communities. Often, if a chemical has been
detected in the groundwater, the contaminated aquifer is cleaned
up and the landfill treated, resulting in additional costs to  the
community.
   For each chemical, estimates are made of  the health effects,
switching and cleanup costs (corrective costs) attributable to  the
presence of that chemical for wastes currently disposed of in  un-
regulated landfills. Regulation of the  waste is assumed to prevent
these estimated  health effects and corrective costs completely.
The estimated benefits attributable to the regulation  are the  ad-
verse health effects and corrective costs avoided by its implemen-
tation.
   Four steps are used to determine benefits:
•  Estimate quantity and concentration of chemical in landfill
•  Estimate concentration of chemical in leachate
•  Estimate chemical concentration at well
•  Estimate health effects and corrective costs attributable to con-
   tamination of the wellwater.
   The unregulated wastes are assumed to be disposed of in a land-
fill each year for 20 years (the average lifetime of a landfill). The
amount of the chemical contaminant that  leaches through  the
landfill and the leaching duration are determined  using data on
the persistence and mobility of each chemical and the form of the
waste stream. From the bottom of the landfill, the contaminant is
transported through the aquifer to the community well. The con-
centration of the contaminant at the  well varies over  time and is
tracked over a maximum of 100 years. The  attenuation factor
assumed depends both on dilution of leachate in the aquifer and
on the specific properties  of the chemical. The adverse health
effects and corrective costs attributable to the contaminated well
are then estimated. Excess cancers are estimated using a life table
model. For noncarcinogenic  health effects, the person-years of
exposure above the HBSL are estimated.
   The Base Case Method  is employed  to estimate  the  concen-
tration of the  chemical in the  leachate and at the well.  Key
assumptions of this method include:
•  The landfill receives predominantly domestic refuse, with only
   5% of the landfill holding industrial waste.
•  The character of the leaching fluid  to which wastes are exposed
   is primarily a function of the noninduslrial material  in the land-
   fill.
•  The landfill is located over an aquifer that is a source of drink-
   ing water for 2,000 persons in the region of the landfill.
•  Soil below the landfill has limited attenuative capacity.
•  The nearest drinking  water wells  are 150  m (500 ft) down-
   gradient from the landfill.
•  As constituents migrate from  the landfill through  the  un-
   saturated and saturated zones to the source of drinking water,
   they are attenuated by a factor of 100.

COST ESTIMATION METHODOLOGY
   The current disposal costs  or baseline must be  established to
estimate the increased disposal costs incurred by waste generators
resulting from the proposed regulation. Current disposal costs are
a function of the  disposal alternatives in use. Where the waste is
not a listed hazardous waste, current disposal practices are iden-
tified by examining the technical literature, by analogy to similar
wastes for which disposal practice is known, or by assumption.
   Some baseline disposal alternatives may understate the actual
treatment and disposal applied to that waste because no effort  has
been made to determine which wastes may be affected by State
and local regulations that are more stringent than Federal regula-
tions. This also may occur because firms may be voluntarily ap-
plying  more thorough  treatment and disposal than required by
regulation. The result of this potential understatement of baseline
treatment and disposal alternatives is that the estimated increase
in disposal costs to comply with the  characteristic approach will
be greater than the actual increase.
  For currently landfilled wastes not  listed as hazardous but sub-
ject to the regulation,  disposal practice after regulation will be-
come more stringent and  costs will increase. Disposal costs are
assumed to remain  the same for wastes currently incinerated or
deepwell-injected.  Solvent wastes  and  a few other wastes that
have been assumed  to be landfilled will be containerized prior to
disposal.
  Using model  plant information, estimates of the incremental
disposal and operating and maintenance costs associated with the
implementation of the alternatives are projected. These estimated
costs then are compared to the cost of contracting with com-
mercial disposal services to estimate properly the minimum costs
incurred by the affected facilities.  These costs are annualized to
reflect  an accurate measure of the  increased  production costs
associated with  the  new regulation. Estimates of percentage cost
changes are generated  for use in  the  production/consumption
model.  Under  the  assumption of full-cost pricing, these  per-
centage estimates are determined by  dividing the annualized in-
cremental costs  by the value of shipments in affected SIC indus-
tries.
  The economic impacts model is used  to derive all costs or wel-
fare  losses borne by consumers of  directly affected products.
Consumers suffer a welfare loss when regulatory compliance in-
creases  prices because they lose consumer surplus, or the value
placed on consumption in  excess of  the  amount  required  to
purchase a product. Economic theory  allows the estimation  of
total consumer costs through impacts  in the directly affected mar-
kets. Thus, input and output market data are not required.
  Consumer surplus losses represent the only recurrent or annual
costs. Changes in waste disposal methods in response to a regula-
tion are represented by an upward shift in the  supply function.
The higher production  costs that result create a new equilibrium
and a consumer surplus loss. The new equilibrium will have lower
production at a higher cost than the initial equilibrium. A real
resource cost is the value of the additional costs incurred at the
new lower level  of output. A deadweight loss is the loss in surplus
value consumers placed on those  units that no longer will  be
produced.
  Extension of the above analysis to a  multimarket situation is
straightforward. Because impacts  in input and output markets
need not be considered to project approximate welfare costs, total
welfare costs are developed  by summing welfare costs in the 11
directly affected markets.
  Consumer surplus and deadweight costs represent annual costs.
Within  this analysis, all baseline data are presented for the year
1982. However, consumer surplus  losses will be incurred for an
unknown  number of years. To develop cost estimates for future
years, costs first are estimated for  1982 and then assumed to be
constant for all subsequent years. This simplifying assumption is
used because time constraints precluded  the projection of market
trends.
  Implementation  costs, consisting of transaction costs and em-
ployment  losses, represent one-time  losses in welfare. Transac-
tion costs represent the value of resources that would be expended
to determine  if  a waste stream  is to be regulated. These costs are
based on an estimated cost of sampling and analyzing each waste
stream by  affected  facilities. Employment losses  occur because
362    RCRA AMENDMENT EXPERIENCE

-------
consumption of goods and services are foregone when individuals
are unemployed. Losses are based on the projected changes  in
production  and employment-to-output ratios for each directly
affected market.  These losses are not valued in dollar terms be-
cause it is difficult to project the length of time an employee
will be  unemployed. Also, the value placed  on this time period
depends on a variety of such  unknown parameters  as  an indi-
vidual's job skills, age, education and personal dislike of being
unemployed.

RESULTS—
Aggregate Benefits
   Continued use of current practices for managing wastes con-
taining the selected chemicals is expected to result in the deteriora-
tion of environmental quality. This  deterioration may elevate
risks to human health and reduce the quality of environmental
resources, such as  uncontaminated drinking water.  The  major
route expected to affect environmental  quality is through the
leaching of contaminated wastes into groundwater. Over 50% of
the U.S.  population uses  groundwater for  drinking water.
Further, contaminated groundwater can enter surface water, re-
ducing its quality. The capacity of both groundwater and surface
water to assimilate toxic chemicals is limited.
   If people drink contaminated  groundwater, a  wide range of
health effects may occur, from simple gastrointestinal problems
to cancer  and birth defects. We focus on the possible excess can-
cer cases if the selected chemicals are not regulated. We assume
that contaminated water would continue to be used as a drinking
water source  unless  the concentration reached  taste or  odor
thresholds of the  average person.  If that threshold is attained, we
assume  they would switch to alternative water sources and no ad-
verse health effects would occur. When a landfill is recognized
as a source of groundwater contamination,  we also  assume the
municipality would take action to prevent further leaching of the
chemicals. Switching and cleanup incur significant financial costs.
Estimates were developed for a representative community  of
2,000 persons and were aggregated to obtain national totals. The
aggregation process  assumed that each plant's waste affected a
single community. Because this aggregation process  is not very
precise, the reader is cautioned to interpret the results carefully.
The results  for each regulatory  alternative  are summarized in
Table3.

Aggregate Costs
   Benefits of the regulatory alternatives would be accompanied
by costs.  Total costs of the regulatory alternatives include real
resource costs, deadweight consumer surplus losses, deadweight
producer surplus  losses (capital value losses), employee disloca-
tion costs and government transaction costs.
  Two  of these welfare costs  have not been projected in this
analysis. Employee dislocations have been quantified, but their
social costs  have not  been evaluated. Capital value losses in-
curred by owners of affected  capital also have not  been eval-
uated.
  Projected real resource and deadweight consumer surplus costs
of the three regulatory alternatives are reported in Table 3. Pro-
jected employee dislocations also are presented.

BENEFIT-COST ANALYSIS
  Table 3  summarizes the benefits and costs  of the three regula-
tory alternatives. The differences between the monetized benefits
(i.e., avoided  corrective costs) and monetized costs (i.e., real
resource and deadweight consumer surplus costs) are compared
using the annualized method. The differences are positive for all
regulatory alternatives. Thus, each alternative would provide an
                            Table 3
                     Benefit-Cost Assessment
                                         Regulatory Alternative
          Benefits/costs
Monetized benefits:

  Avoided cost of alternative water
  source
    Present value ($106)b
    Annualized ($106/yr)
3,218
  378
  Avoided  cost of aquifer cleanup
    Present value ($106)b         11,897
    Annualized ($106/yr)           1,398

Monetized  costs:

  Real  resource cost
    Present value ($106)b          1,285
    Annualized ($106/yr)             151

  Deadweight consumer  surplus cost
    Present value ($106)b          1.15
    Annualized ($106/yr)           0.14

  Transaction cost ($106)''         1.22

Net monetized benefits:

  Present  value ($106)b           13,830
  Annualized ($106/yr)            1,625

Nonmonetized benefits:3

  Avoided  cancer  cases               54

  Avoided  person-years of exposure    4.8
  above the HBSL  (106)

Nonmonetized costs:

  Employment dislocations           407
3,317
  390
           12,316
            1,447
            1,287
              151
            1.16
            0.14

            1.26
          14,345
           1,685
              54

               4.8




             407
3,174
  373
           11,719
            1,377
            1,186
              139
            1.03
            0.12

            1.17
          13,706
           1,610
              53

               0




             372
aBenefits are based on the Base Case Method.
b20-Year cost discounted at 10%.
cOne-time cost incurred first year.
dMonetized benefits minus monetized costs excluding transaction costs.
eHBSL =  Health based starting limit.

improvement in economic welfare. The maximum improvement,
ignoring health benefits, is under Regulatory Alternative 2.
  An evaluation  of Regulatory Alternatives 1 and 2 will allow a
comparison of the results of the 100 x HBSL and the more strin-
gent 10 x HBSL regulatory levels. These two alternatives  have
virtually no differences in health benefits. However, Regulatory
Alternative  2 yields  an increase  in  net  monetized  benefits of
$60.59 million annually.
  Moving from Regulatory Alternative 1  to Regulatory Alterna-
tive 3 results in a decrease in the stringency of the regulatory level
from 100 x  HBSL to 1,000 x HBSL. As a result of this change,
avoided cancer cases will decrease slightly by a total of 0.39 cases.
However, the  4.8  million  person-years  of avoided exposures
above the HBSL in Regulatory Alternative 1 decreases to 0 under
Regulatory  Alternative 3.  Additionally, net monetized benefits
are $14.44 million annually higher in Regulatory Alternative 1
than in Regulatory Alternative 3.

CONCLUSIONS
  The proposed expansion  of the toxicity characteristic will pro-
vide an additional, cost-effective method for the U.S. EPA to use
in regulating hazardous wastes. For those regulatory alternatives
analyzed,  the benefits are significantly greater than the costs of
regulating additional wastes containing any of the 52 chemicals.
                                                                                        RCRA AMENDMENT EXPERIENCE    363

-------
                 Preliminary Assessments  and Site  Investigations
              Under  the Corrective Action Authorities  of RCRA;
                                Analysis of Early  Experiences
                                                 John W. Butler
                                               A.T.  Kearney,  Inc.
                                              Alexandria,  Virginia
                                               Robert D. Volkmar
                                            Michael Baker,  Jr.,  Inc.
                                              Beaver,  Pennsylvania
ABSTRACT
  Implementation of Section 3004(u) of RCRA requires the eval-
uation of solid waste management units not regulated under
other RCRA authorities to determine if these units have contin-
uous releases of hazardous wastes or constituents. The first step
of this evaluation, a preliminary assessment and site investiga-
tion (PA/SI), often is conducted in conjunction with the RCRA
Part  B permit  application review  process.  The authors have
assisted the U.S. EPA in conducting several of the earliest RCRA
PA/SIs. In this paper, the authors  present a composite descrip-
tion of the PA/SI methodologies employed in those reviews, dis-
cuss critical  obstacles encountered and describe the process used
to determine the need for RCRA permit conditions to address re-
leases of hazardous wastes or constituents to the environment.

INTRODUCTION
  One of the most important new authorities provided  to the
U.S. EPA under the Hazardous and Solid Waste Amendments of
1984 (HSWA) is Section 3004(u) which requires that  all RCRA
permits issued to treatment, storage and  disposal facilities after
November 1984 shall require corrective action for all releases of
hazardous wastes or constituents from solid  waste management
units at these facilities. In addition, under Section 3008(h), the
U.S. EPA is given authority  to issue corrective action enforce-
ment orders to facilities managing hazardous  waste under RCRA
interim status. Another extremely  important related  provision.
Section 3004(v),  extends RCRA corrective action authority be-
yond the boundaries of the facility.
  The authorities  provided under  these provisions  extend the
scope of RCRA well beyond  that to which treaters, storers and
disposers of hazardous waste previously had been covered. Prior
to the amendments, facilities and operators had to be concerned
with obtaining permits for regulated units used for treating, stor-
ing and disposing  of hazardous wastes as defined by the U.S.
EPA under 40 CFR Part 261. Now, a new process has been put
into place which extends the permitting process to include a re-
view of essentially every unit within a facility's boundaries which
is used or has been used to manage solid wastes.
  The U.S.  EPA is approaching implementation  of these new
authorities under a process involving three general steps which
roughly parallel the process used under CERCLA:'

  Stage 1: Assessment of Need for Corrective Measures. This
stage usually involves a preliminary  assessment and site investiga-
tion (PA/SI), which includes an evaluation  of readily available
information to determine if releases posing  a threat  to human
health and the environment have occurred or if there is a likeli-
hood that such releases may have occurred. The U.S. EPA issued
a draft guidance document on August 5,  1985, providing pro-
cedures for conducting RCRA PA/SIs.'
  Stage 2: Remedial Investigation and Development of a Pro-
posed Program  of Corrective Measures. If the  results of the
PA/SI indicate that releases have occurred or may be occurring
at one or more units, owners and operators will be required to
conduct a remedial investigation  to make a more definitive de-
termination regarding the potential for releases and to identify
the nature and extent of releases that are found.
  The second step in this phase of the corrective action process
is the development of a program of corrective measures to clean
up the releases found in the remedial investigation. If this pro-
gram is developed as part of a RCRA permit condition, the per-
mit will need to be modified to incorporate the corrective meas-
ures program after public comment is solicited and received on
the proposed measures.
  Stage 3: Selecting and Performing Corrective Measures. This
stage is the actual establishment and implementation of the
corrective measures program, which may be carried out accord-
ing to schedules of compliance in the facility's modified RCRA
permit or under an enforcement order.
  The authors have assisted the U.S. EPA in conducting some
of the earliest PA/SIs under this process,  using the procedures
in the Agency's August 1985 guidance. The guidance currently is
being reviewed by the U.S. EPA based on the results of field test-
ing of the document. From their early experiences with the RCRA
PA/SI process, the authors present  some observations  which
they believe will be helpful to facility owners and operators in
determining how they can participate most effectively in the PA/
SI reviews of their facilities and  to regulatory officials who are
faced with the challenging task of conducting RCRA PA/SIs.

CONCEPTUAL APPROACH TO PA/SIs
  The approach used by the authors in conducting RCRA PA/
Sis roughly involves a sequential exclusionary process  for each
unit  at a  facility  in which solid waste  management  units
(SWMUs) are systematically eliminated from further considera-
tion if certain criteria are met. Figure 1 illustrates the line of in-
quiry used  by the investigator in this process. In practice, data
limitations restrict the ability of the investigator to make simple
"yes/no" determinations. Rather than being a sequential pro-
cess,  the  actual  process almost always requires a degree of pro-
tessional  judgment in weighing a number  of factors. Neverthe-
less, in an ideal situation, the investigator would have all the in-
formation needed and the line of  inquiry would take place as fol-
lows: Is the unit an SWMU?
     Units at a  facility which are considered SWMUs for pur-
poses of RCRA corrective action reviews are very broadly de-
364    RCRA AMENDMENT EXPERIENCE

-------
fined. In general, the types of units include:  landfills,  surface
impoundments, waste piles, land treatment units, incinerators, in-
jection wells, tanks (including 90-day accumulation tanks), con-
tainer storage areas, transfer stations and waste recycling op-
erations.  Units specifically  exempted from RCRA  regulation
under 40 CFR Part 264 are included in this universe, i.e., waste-
water treatment and elementary neutralization tanks.  Units may
be active, inactive or completely abandoned. Units which are not
managing or have never managed RCRA-regulated  hazardous
wastes still are considered as SWMUs  subject to review,  since
the potential for the release of hazardous constituents also must
be evaluated.

Does the Unit Manage Hazardous Waste?
  Waste  characterization is one of the most critical steps in the
PA/SI process. The scope of materials of concern is considerably
larger than wastes which are regulated under RCRA on the basis
of listing or hazardous characteristics. In a PA/SI, units which
manage or have managed wastes  containing hazardous  constit-
uents included in 40 CFR Part 261, Appendix VIII, must be re-
viewed for the potential of releases of these constituents to the
environment.  Further,  units used to manage bulk  putrescible
wastes (such as sanitary sludges) must be assessed for the poten-
tial of methane gas generation and subsequent migration of sub-
surface gases to adjoining structures.
   In an ideal case, the PA/SI investigator would have detailed in-
formation available about the types, volumes and locations of
waste within the units. If the data clearly show that no hazardous
wastes, hazardous constituents or wastes capable of methane gen-
eration have ever been managed  in  the unit,  the  unit  may be
eliminated  from further  consideration. If hazardous wastes or
constituents are present,  the chemical/physical properties, tox-
icity, persistence and mobility of the wastes can be evaluated to
determine the potential of the wastes for releases to air, ground-
water, surface water and/or soils; also the threat posed to human
health and the environment by such releases can be assessed.
(1)
DOES OR HAS
UNIT MANAGED
HAZARDOUS
WASTE
OR CONSTITUENTS*


1

(J)
DOES UNIT HAVE
A HISTORY
OF OR POTENTIAL
FOR RELEASE 7

              (41
         DO RELEASE SPOSE

          A THREAT TO
          HUMAN HEALTH
              AND
          ENVIRONMENT »
   IS THERE A

'.SIGNIFICANT RISK
 TO POPULATIONS?
                          Figure 1
                Conceptual RCRA PA/SI Process
Is There a Potential for Releases?
  If the investigator has key design and operating data, he can
make a reasonable evaluation of the potential for releases from
the unit. A common unit investigated is a tank which has an ade-
quate secondary containment system, closed roof and sufficient
flood protection; commonly available are records demonstrating
that the tank has  been inspected frequently and  maintained. In
this case, the investigator may  eliminate the unit from further
consideration on the basis of the unlikely potential for releases,
regardless of the types of wastes or constituents managed.
Is There a Threat to Human Health and
the Environment?
  At this point in the "idealized" evaluation, the investigator has
determined that a solid waste management unit has managed haz-
ardous wastes or wastes which have hazardous  constituents and
that  there is  a high likelihood of  releases. It still may be pos-
sible to eliminate this unit on the  basis of consideration of the
pathways available to the releases into each environmental med-
ium. For example, subsurface releases may not present a threat to
groundwater  if the  underlying aquifer is deep,  if soils are rela-
tively impermeable  and/or if the aquifer underlays sufficiently
impermeable  bedrock. Even if the hydrogeological setting indi-
cates that there  is a potential for  infiltration to the underlying
aquifer, the owner/operator may have monitoring data available
demonstrating that no  contamination  attributable to  releases
from the unit are occurring.

Is There a Risk to the Neighbors?
  Even if the unit has failed all of the above tests, it may be pos-
sible to determine that the risks posed by the releases do not pose
a threat to human health and the environment. For example, air
releases may not be  of concern if there are no downwind popula-
tions. However, since U.S. EPA policies regarding risk assess-
ment in RCRA PA/SIs still are evolving, the role of risk in the
process is unclear at  this time.

Key Information Used in RCRA PA/SIs
  The process just described assumes that the investigator has in-
hand all necessary information required  to make the determina-
tions just described.
  The  most important pieces  of information and their common
uses  in RCRA PA/SIs are displayed in Figure 2. The  following
discussion summarizes the  common roles of these data  in RCRA
PA/SIs.
  In many circumstances, the PA/SI for a  facility will begin
after submission of a Part B permit application  and  after the
owner/operator has responded to the regulatory  agency's request
for information  about its solid waste   management units. Al-
though these information requests have been fairly general  in
practice, there are certain specific data items which should be
included in each letter responding to such information requests.
These  include identification  of each  solid waste management
unit:3
• Description of the type of unit
• Location of each unit within the facility on a topographic map
• Dimensions and available information on release controls
• Status and history of operation
• Description of wastes managed
  All available information which could indicate the potential for
releases from  any of the units,  including:

• Available environmental monitoring data (groundwater, sur-
  face  water,  air, soil samples)
• History of releases to any environmental medium

  Information in the applicant's response to this information re-
quest,  combined with the  RCRA Part B application,  generally
provides the core of data for  initiation of a PA/SI. Some of the
most relevant data for PA/SIs which are generally found in Part
B applications include:
  Waste Characterization: Although the waste  analyses in Part
Bs are only applicable to those wastes which are managed at regu-
lated units in  the application, the information may be useful for
providing  a general picture of waste generation trends at  the
facility.
                                                                                     RCRA AMENDMENT EXPERIENCE    365

-------

\"
Inl o r mil ion ^N^
aourc • >v
RCRA
P«rt B Applies lion
a WMU

tnlormcllon L •n«r


EPA / Slat* RCRA
Inspection F H* t
(••rmii Fil«s
F torn





ol Unll





t







*

1


; h • r • c i • f 1 1 • 1 1 o n


*


g







*




a,...n8


t


i




'


*



Swb
• •i Iln0


4







"


t



Op*f Atlflfl
..i....





•







*




Oil*


t


1




*


t




IfilOf -
flt*ll0fl

















                                                           Figure 2
                                        Information Sources Frequently Used in RCRA PA/Sls
    General Facility Description: Descriptions of the facility can
 be used by the PA/SI investigator to gain some understanding of
 the overall  facility layout, surrounding land use and  general
 topography.
    Floodplain Delineation: The determination of whether or not
 individual solid waste management units are situated in flood-
 plains is a critical part of the  PA/SI.  Corrective measures  for a
 unit which may not otherwise  have a potential for releases to the
 environment may be required if flooding could threaten the struc-
 tural integrity of the unit and/or result in direct runoff or dis-
 charge to adjoining surface waters.
   Hydrogeological Data: For Part Bs for land disposal units, an
 extensive amount of data regarding the hydrogeologic setting of
 these units is required. While these data are specific to the  regu-
 lated units at the facility, they sometimes  can be used in PA/SIs
 of other units at the facility in lieu of available information for
 these units.
   Exposure Information:  Under Section 3019 of RCRA, permit
 applications  for landfills  and surface impoundments must in-
 clude exposure information reports. This information is valuable
 in determining the potential migration pathways of contaminants
 released from SWMUs and for identifying populations potential-
 ly at risk from such releases.
   Contingency Plans: Contingency plans, required of all RCRA
 permittees, may be  useful in determining if the permit applicant
 has identified all SWMUs at the facility.  For example, the  plan
 may provide for temporary storage or impoundment  of  haz-
 ardous wastes in specific areas at the facility; these areas would
 be solid waste management units which should be included in the
 PA/SI review.
  As indicated in Figure 2, there are numerous other important
 sources of information which should be available to the investi-
 gator to supplement the facility's Part B  and SWMU informa-
 tion letters. These sources are noted below.
  CERCLA  Sources: If a preliminary assessment and/or site in-
vestigation has  been performed for  the  facility  under  the
CERCLA program, the results and supporting documentation
are clearly a critical imput to the RCRA PA/SI. Hazard Ranking
 System documentation  and CERCLA Section 103(c)  notifica-
 tions also can be valuable information sources. The most im-
 portant differences between CERCLA and RCRA PA/SIs, which
 will be described in the  following paragraphs,  must be  borne in
 mind when using this information.
   U.S. EPA/State Inspection Files: Generally, the PA/SI inves-
 tigator will find it extremely beneficial to perform a file search of
 the regulatory agency's  (State or U.S.  EPA) records of inspec-
 tions performed previously at the facility. RCRA inspections, as
 well as inspections performed under the NPDES program, some-
 times can help the investigator identify environmental problems
 and releases which have occurred. Inspection records for State-
 regulated units, such as  non-hazardous solid waste landfills,
 also can be extremely valuable.  An unquantifiable, but never-
 theless important, aspect of this review is the general insight pro-
 vided about general environmental management practices at the
 facility.
  Monitoring Data:  Any  relevant type of environmental mon-
 itoring data, even if the monitoring  was not performed for
 RCRA-hazardous constituents, may be important to the PA/SI
 investigation.  Indications  of conventional  contaminants  in
 groundwater, for example, may be indicative of releases of other
 constituents of concern from a solid waste management unit.
  Other  Data Sources:  Other sources of information, such as
 aerial photography, may be very important to the PA/SI investi-
 gator. Basically, any  type of information which can help the in-
 vestigator understand the nature and operation of the unit, its his-
 tory,  wastes managed and operational practices should be con-
 sidered in the review.

 Relationship Between RCRA and CERCLA PA/SIs
  There  are obvious analogies between the PA/SI procedures
used in the CERCLA and RCRA programs. Generally, the same
types of factors are considered by the investigator, including the
 types of wastes managed at the unit, environmental setting, struc-
tural  integrity of the unit  and release controls, history of unit
operation, history of releases and surrounding land uses.  For
this reason, prior CERCLA investigations are generally extremely
useful sources of information for the RCRA PA/SI.
366   RCRA AMENDMENT EXPERIENCE

-------
  While many of the procedures for conducting PA/SIs for facil-
ity SWMUs under RCRA are similar to those used by CERCLA
field teams, there  are fundamental differences between these two
processes that may render different outcomes in terms of result-
ing requirements placed upon facility owner/operators. The most
basic difference is in the ultimate objectives  of each  of the re-
views.
  The objective of a CERCLA PA/SI is to develop information
sufficient to develop a score under the Hazard Ranking System
(HRS), which in  turn is used to establish priorities for further
action at the site. Ultimately a site may be placed on the National
Priority List with subsequent remedial investigations, feasibility
studies and selection and  implementation of remedial actions.
An HRS score is established for an entire facility and may be the
result of cumulative environmental impacts of a number of units
within the facility.
  In contrast to the CERCLA process, the objective of an RCRA
PA/SI is to  determine whether or not releases may be occurring
at individual units within a facility. The outcome of the investi-
gation is not a numeric score, but a finding for each unit regard-
ing its potential for releases to the environment. This distinction
can  mean that  facilities which  have relatively low HRS scores
may nevertheless  be required to conduct  corrective measures
under RCRA.
 OBSTACLES TO THE PA/SI REVIEW
 PROCESS
  The conceptual RCRA PA/SI process has been described in
 detail to highlight two of its  most significant features: first, an
 extensive amount of information is needed to fully complete every
 step of the process and, secondly, facility owners and operators
 need to be actively involved in the process to assure that the most
 accurate and timely information is used. The lack of data neces-
 sary to derive conclusions regarding the likelihood of releases of
 hazardous wastes and constituents to the environment likely will
 result in RCRA permit conditions requiring  the  facility owner
 and operator to develop the needed information under  man-
 dated deadlines.
  It is clearly in the facility  owner/operator's best interests to
 come forward with all available relevant information at the PA/
 SI stage of the permitting process rather than to  defer substan-
 tive input  until the RCRA  permit reaches  the formal proposal
 stage. Of course, in many cases, critical information may not be
 available to the owner/operator. Facilities simply did not have a
 need to keep records  on many of their units operated prior to
 RCRA enactment. The lack of adequate records poses one of the
 most critical obstacles to the RCRA PA/SI process.
  The most frequently occurring  data problem  is inadequate
 characterization of the wastes  placed in the unit. Without reliable
 waste management data, it  is not possible to  determine if there
 are or are  not hazardous constituents within the waste material.
 Since the entire corrective action program is keyed by releases of
 hazardous constituents, the PA/SI process is truncated by lack of
 waste characterization data. A case frequently encountered is a
 landfill unit known to have been used to dispose  of plant trash
 which may (or may not) have included off-specification produc-
 tion products with hazardous constituents or containers with haz-
 ardous residues.
  In some cases, it has been possible to ascertain the presence of
 hazardous constituents in a particular disposal  facility through
 process knowledge of the industry in question. In many cases,
 however, this is not possible—particularly where several disposal
 areas have been utilized  and no records have been kept to define
 the wastes placed in each area.
  In these cases, the PA/SI review process essentially becomes
stalled until adequate data are obtained. In some cases, this re-
view  has  been  accomplished  by  mandating data  collection
through permit conditions in the RCRA permit. In other cases,
facility owners and operators have been able to provide historical
data  on general facility waste management  practices which
demonstrated  that the facility's hazardous wastes were trans-
ported off-site during the period of operation of the unit in ques-
tion.
  Another frequent obstacle to the PA/SI review  process is in-
adequate description of the surrounding environment, primarily
those elements which serve as vehicles for contaminant trans-
port (soils, groundwater and surface waters). Typically, existing
waste management units  are not located in proximity to aban-
doned units. As a result, environmental information provided
in the Part B application or Exposure  Information Report may
not be germane to remotely located, abandoned units. If the Part
B application is for non-land disposal units, facilities are not re-
quired to provide hydrogeologic information which might be
helpful to the PA/SI review.
  Lacking critical information,  such  as  soils permeability or
depth to groundwater, it is not possible  to make reasonable judg-
ments regarding the potential for releases of hazardous constit-
uents, even when the presence of hazardous constituents in the
unit has been  firmly established. Again, in this case it is  likely
that the responsibility for providing the needed data will fall to
the owner/operator. Often,  this step requires  extensive  field
studies, including  soils testing,  groundwater  investigations or
sampling of surface waters.
  A third obstacle that frequently  impedes the PA/SI decision-
making process is the lack of evidence regarding releases  from
SWMUs.  This is especially true in the case of long-abandoned
units that predate the current era of environmental awareness. In
these and  other cases, the facility has no reason to monitor
groundwater, surface waters, air or soils. Lacking direct evidence
of releases, it sometimes  is possible to make judgments regard-
ing the  potential for or probability of a release. This decision
can be made when the properties of the waste and the local trans-
port media (groundwater, etc.) are well defined.
  More often than not,  however,  only sketchy information is
available. In those cases, there are two approaches that can re-
move the decision-making obstacles. The first  step is to directly
monitor the appropriate media to detect evidence of contamina-
tion resulting from a release from the unit in question. This pro-
cess frequently involves monitoring of surface of ground waters.
  Alternatively,  it may be more feasible to conduct the studies
necessary to assess the potential for release. This  approach in-
volves investigation of the waste properties and characterization
of the surrounding environmental media.
  In some complex situations, a hybrid of both approaches must
be used. This can occur when there are  several units in proximity
and knowledge  of environmental  conditions  is insufficient to
properly design  a monitoring program to  detect releases.  This
situation presents a more difficult  case where both background
environmental studies (e.g., geohydrological investigations) and
monitoring must be undertaken to  detect contamination.
  Another obstacle arises when environmental data indicate that
there is a problem at the facility,  but the PA/SI investigation
cannot identify the unit or units most likely to be causing the
problem. This situation can arise,  for example, when ground-
water monitoring data from samples collected at various locations
within the  facility indicate concentrations of hazardous constit-
uents above natural  background levels,  but inadequate hydro-
geologic data preclude  determining the source. In  this case, the
facility might be required to conduct a remedial investigation for
                                                                                     RCRA AMENDMENT EXPERIENCE    367

-------
 the area of the facility where the problem has been found. The
 remedial investigation may, in certain cases,  be  mandated  for
 the en tire facility.
   Finally,  difficulties in identifying and locating all solid  waste
 management units  within a facility often  are encountered. In
 some cases, facility owners and operators have not understood
 that units  which have not  previously been  considered in  the
 RCRA program, such as 90-day waste accumulation areas, non-
 hazardous waste disposal units and  elementary  neutralization
 tanks, are  now included in the PA/SI process. In other cases, the
 owner/operators were not aware of the existence  of abandoned
 units until the PA/SI investigators  questioned how wastes were
 measured at the facility prior to initiation of current management
 practices.  In answering these questions, which may require  the
 owner/operator to  interview current or past facility employees,
 the owner/operator may be surprised by the number of old  waste
 disposal units within the facility's boundaries. In the authors' ex-
 perience, it is not uncommon  for older, large chemical manu-
 facturing facilities, particularly those which  are  more than 20
 years old, to have 20 or more SWMUs which must be evaluated.
   This discussion has focused on  the critical obstacles exper-
 ienced while conducting some of the first PA/SIs  under RCRA.
 More experience by PA/SI investigators and an increased under-
 standing of the process by facility  owners  and operators  likely
 will serve to sharpen  the analytical  and implementation proced-
 ures to address these issues.

 CONCLUSIONS
   This paper has provided an  explanation of the  RCRA PA/SI
 process, likely sources of data and their uses in the process and
 some of the critical obstacles which have impeded the process.
   Several conclusions can be derived:
• The RCRA PA/SI process requires the use of as much relevant
  available data as can be reasonably assembled. Key informa-
  tion sources include the facility's Part B application, response
  to the regulatory authorities' request for SWMU information,
  RCRA and other environmental program  inspection files and
  previous CERCLA  investigation data. Facility owners and op-
  erators should be prepared to provide significantly more infor-
  mation in response to RCRA PA/SIs than to  the pre-HSWA
  permitting  process.
• The fundamental differences between RCRA and  CERCLA
  PA/SIs may  result  in different  outcomes. Facilities with rela-
  tively low MRS scores may nevertheless be  required to perform
  corrective measures under RCRA.
• Since the lack of dala during an RCRA PA/SI review can re-
  sult in mandatory requirements  to obtain the needed data, it is
  in the  best  interests of facility owners  and  operators to collect
  and provide as much relevant information as possible and to be
  involved early in the process.
REFERENCES
I.  McGraw, J.W.,  "Reaulhorization Statutory Interpretation #3: Gui-
   dance on Corrective Action for Continuing Releases," Office of Solid
   Waste and Emergency Response. U.S. EPA memorandum dated Jan.
   30, 1985.
2.  "RCRA   Preliminary  Assessment/Site   Investigation  Guidance
   (Draft)," Permits and Slate Program Division, Office of Solid Waste,
   U.S. EPA, Aug. 5,  1985.
3.  McGraw, J.W.,  "RCRA  Reauthorization Statutory  Interpretation
   #3: Immediate Implementation of New Corrective Action Require-
   ments," Office of Solid Waste and Emergency Response, U.S. EPA,
   memorandum dated Feb. 5, 1985.
368    RCRA AMENDMENT EXPERIENCE

-------
               Assessment of  the Application of RCRA Part 264
                       Standards  to  CERCLA Site Remediation

                                             Rebecca N. Fricke, P.E.
                                       EMPE, Inc., Consulting Engineers
                                               Nashville,  Tennessee
                 (Formerly: Technical Coordinator, Tennessee State Superfund  Program)
ABSTRACT
  The purpose of this investigation was to evaluate recent revis-
ions to the National Contingency Plan requiring application of
RCRA Part 264 standards to CERCLA site remediation. This
assessment was conducted in three phases. The first phase in-
volved review of the regulatory background to determine consis-
tency  with  existing  laws and regulations.  The second phase
assessed technical validity and identified technical problem areas
of the proposed policy with particular attention to requirements
of Part 264 Subparts F and G.
  The third phase investigated the cost-effectiveness of applying
RCRA Part 264 standards to CERCLA sites. Two National Prior-
ity List sites were studied. Costs were based on actual closure
costs under the original NCP policy and projected  expenditures
for closure under the revised policy  requiring application of
RCRA Part 264 standards. Effectiveness was quantified in terms
of risk reduction using the U.S. EPA Hazard Ranking System
and a computerized deterministic groundwater risk model for cal-
culation of lifetime mortality risks.

INTRODUCTION
   In the Feb. 12, 1985, Federal Register, the U.S. EPA proposed
revisions to the National Contingency Plan (NCP) which would
require that CERCLA (Superfund) remedial actions "attain or
exceed all applicable or relevant Federal public health or environ-
mental standards."  The U.S. EPA specifically explained in the
Federal Register that this includes compliance with RCRA Part
264 standards.1
  On  Oct. 10, 1985, Lee Thomas, U.S. EPA Administrator,
signed the final version of the revised National Contingency Plan.
The final version is virtually the same as the version proposed in
February with a change in the phrase "applicable or relevant"
to "applicable or relevant and appropriate." The addition of the
words "and appropriate" suggests that the U.S. EPA received
sufficient public  comment to recognize the problems associated
with strict application of Federal public health standards, includ-
ing RCRA Part 264.
  Among the comments received by the U.S. EPA was an exten-
sive study conducted by this author as an employee of the State of
Tennessee. The study specifically  assessed the  applicability of
RCRA Part 264 standards to closure of CERCLA sites. The pur-
pose of the following report is to relate the results of that study.
The study was broken into three phases; regulatory review, tech-
nical assessment and cost-effectiveness investigation.

PHASE I: REGULATORY REVIEW
  The pertinent laws and regulations reviewed in  association with
this issue were: (1) RCRA and its amendments;  (2)  RCRA Part
264 Standards for Owners and Operators of Hazardous Waste
Treatment,  Storage  and  Disposal  Facilities;  (3)  CERCLA
(authorizing the Federal Superfund program);  and (4) the NCP
used to direct Superfund activities. The review  of these laws and
regulations revealed information on the intentions of Congress
and the U.S. EPA.
  RCRA Section 3004 directed the development of regulations
for "treatment, storage or disposal of hazardous waste...as may
be necessary to protect human health and environment."3 The
same section also required that the Administrator shall, where
appropriate, distinguish in such standards between new and exist-
ing facilities. Through these sections Congress limited the regula-
tory requirements to that which is actually necessary for protec-
tion and provided recognition of the problems associated with
applying new standards to existing facilities.
  Part 264 standards were originally promulgated in the July 26,
1982, Federal Register. Section VI.A.I. of the Preamble listed
five important considerations involved in the U.S. EPA's strategy
for protection of groundwater: (1) avoid complicated uncertain
predictions;  (2) recognize  the  ability to  limit the  impact of
polluted groundwater by limiting use; (3) prevent stifling of inno-
vation; (4) provide timely  and expeditious regulation; and (5)
use limited financial resources in the most cost-effective man-
ner possible."
  Section 105  of CERCLA specified that the NCP provide a
means of assuring that remedial action measures are cost-effective
(Section 105(7)). The law also required the NCP to provide cri-
teria for determining priorities among releases  or threatened re-
leases and methods  for determining the appropriate extent of re-
medial action to be conducted, "including analyses of relative
costs" (Section 105(2,3,8))[5].
  The NCP was revised pursuant to CERCLA Section 105 to pro-
vide  for  Superfund activities.  It was published in the Federal
Register on  July 16,  1982, and became effective on Dec.  10,
1982.6'7 Section 300.68 defined the method of selection for remed-
ial action at a Superfund site.  The final selection criteria in the
NCP were dependent  on cost-effectiveness, i.e., the  lowest cost
alternative that is technologically feasible and reliable and which
effectively minimizes damage to and provides  adequate protec-
tion  of public health, welfare  or the environment."6 Thus,  the
three factors involved in the  cost-effectiveness selection were
cost, technical feasibility and adequate risk reduction. An ob-
vious question is "What is adequate?"
  The proposed revisions to the NCP published in the Feb. 12,
1985, Federal Register were signed into effect on Oct. 10, 1985.
The  final version was not available at the time of this writing.
U.S. EPA officials  state that the final version was virtually the
same as  the proposed version.  The final selection  process as re-
vised deleted the reference to the "lowest cost alternative" and
added a  requirement to attain  or exceed "all applicable or rele-
                                                                                RCRA AMENDMENT EXPERIENCE
                                                                                                                   369

-------
vant and appropriate" Federal public health  standards.2 Al-
though the term "cost-effective" remains in the selection criteria,
it is overridden by the requirement to attain or  exceed  public
health standards.
  The only situation in which cost would override the require-
ment to attain standards  is in "Fund-balancing" of Federal
Superfund dollars.  In  a situation where limited Federal Super-
fund dollars cannot take care of all sites requiring remediation
and the site represents less hazard than other sites, a lesser alterna-
tive can be accepted. Unfortunately, this waiver does not extend
to State Superfund programs with multiple sites and limited funds
or to Responsible Parties with multiple environmental problems
(possibly including more serious air, water, solid waste or radio-
logical waste problems).
  The elimination of cost-effectiveness as one of the major cri-
teria in  selecting remedial actions creates a conflict with the
apparent intentions of  Congress stated in RCRA, Part 264 (Pre-
amble), CERCLA and the original NCP

PHASE II: TECHNICAL ASSESSMENT
  In the preamble to the proposed version of the National Con-
tingency Plan, the U.S. EPA specified that RCRA  Part 264 stan-
dards would be considered relevant to response at CERCLA
sites.1 The purpose of this section is to assess the technical validity
of applying Part 264 standards to CERCLA sites. The two major
subparts of Part 264 which would be of importance to CERCLA
sites are Subpart  F—Groundwater Protection and Subpart F—
Closure and Post-Closure Standards.
  Subpart F—Groundwater  Protection  regulations  outline  a
three phase  program:  (1) detection monitoring; (2)  compliance
monitoring;  and (3) corrective action. The standards defined by
Subpart  F generally could be applied to CERCLA sites to help
establish the point(s) of compliance, location of monitoring wells,
frequency and duration of sampling, analytical parameters and
cleanup levels.  However, two areas of controversy exist  which
create severe problems in the utilization of the Subpart F stan-
dards in the CERCLA  program. These two areas are the analyti-
cal  requirements for the 375 constituents  in 40 CFR Part 261
Appendix VIII and the determination  of Alternate Concentra-
tion Limits (ACLs).
  The major problems associated with analysis of 40 CFR Appen-
dix VIII  constituents involve the high cost of analysis, technical
questions on the  validity of analytical techniques required,  in-
stability of some contaminants in water (which are therefore not
expected to appear in sample analysis) and frequency of analysis.
The U.S. EPA recognizes that these problems exist  and is making
efforts to address these concerns. The U.S. EPA has proposed a
revision to Appendix VIII to exempt requirements for 23 constit-
uents which are unstable in water for  for which no laboratory
standard for analytical calibration  has been developed.  Until
other concerns are resolved, the public  still is required to follow
the standards as written.'
  Subpart F—Groundwater Protection standards  provide three
bases for groundwater cleanup. The three cleanup levels which
are considered acceptable for groundwater under Subpart F are:
(1)  Background Levels, (2) Maximum  Concentration  Limits
(MCLs for Drinking Water Standards), and (3) Alternate Con-
centration Limits (ACLs, standards for compliance  at the site
based on acceptable levels of exposure some distance from the
site). It is anticipated that for most CERCLA sites it will not be
technically or economically practicable to achieve background or
MCL levels. Therefore, groundwater protection standards typic-
ally will be based on Alternate Concentration Limits.
  Section 264.94(b) outlines the standards used in the develop-
ment of ACLs. These standards provide a reasonable frame-
 work for information which should be collected.  However, the
 level of detail required by the U.S. EPA in the guidance docu-
 ment for ACLs' requires  time-consuming and burdensome data
 development and collection programs. For example, detailed in-
 formation is required on  exact waste quantities and types, de-
 finitive hydrogeologic data and specific lexicological effects (in-
 cluding synergistic effects of combinations).
   Although all of these pieces of information are important,  it
 is questionable whether precise  quantification  is possible and
 whether  it will appreciably improve assessment over that attain-
 able with more generalized information. It is technically incon-
 sistent to require extensive detailed  investigations to obtain high
 levels of confidence in some of the data used in an ACL determin-
 ation while recognizing the large number of speculative input fac-
 tors involved,  such  as meteorological conditions, population
 growth in the area and future uses of surface water and ground-
 water over the life of the site. It is suggested that a greater degree
 of consistency among the required  levels of  accuracy be estab-
 lished. The excessive  quantities of time and money currently re-
 quired for field and laboratory research could better be invested
 in remedial action activities.
   The RCRA Part 264 Subpart  G—Closure and  Post Closure
 standards provide general  closure performance standards. These
 standards emphasize  prevention and base closure  on a "to the
 extent necessary"  performance  criteria (Section 264.11).  This
 approach allows the most  cost-effective methods of closure to be
 used and provides for site specific interpretation of the "to the
 extent  necessary" terminology. In contradiction to this approach,
 Subpart  G references closure standards in Subpart  N—Land-
 fills where closure criteria  more specifically require use of a final
 cover which provides  long-term minimization  of liquid migration
 through the closed landfill  (Sections 264.310(a)).

  The  RCRA Guidance Document entitled  Landfill Design-
Liner Systems and  Final Cover, July, 1982,'° defines the min-
imum cover design considered acceptable. The required cover de-
sign consists of 2 ft of clay (with  a permeability of 10"7 cm/sec
or less) overlain with  1 ft of gravel (with a permeability of 10"'
cm/sec or greater) for  drainage. These two layers are covered by a
layer of soil (1.5 ft) and topsoil (0.5 ft) graded  and seeded to sup-
port vegetation.  For sites  having  an impermeable  underliner, a
synthetic membrane liner is required between the clay and gravel
layers.'"
   Strict application  of these RCRA Part 264 closure requirements
is not technically valid for  remediation of all CERCLA sites. For
many inactive hazardous substance sites, prevention or minimiza-
tion of infiltration is not a critical factor. Some situations where
sites would not greatly benefit by direct application  of these tech-
nology based standards are: (1) sites where waste materials are vir-
tually insoluble; (2) sites subject  to surface-water  flooding and
subsequent rising groundwater; and (3) sites located  in quarries or
hollows where underground springs feed into the waste. For these
sites, an impermeable  cap would not provide a cost-effective solu-
tion to the problem  and, for the second and third cases, would be
completely ineffective in preventing  liquid migration through the
site.
   It is  recommended that each CERCLA site be assessed to deter-
 mine the potential sources  of liquids which may reach the site and
 the potential impact of liquid migration through the site on pub-
lic health and the environment. Following this assessment, the
original language of  Section 264.11 concerning "to the extent
 necessary" should be considered instead of the strict infiltration
 technology based standards. Flexibility in closure design must be
 allowed in order to  provide appropriate and cost-effective remed-
 iation on a site specific basis.
370
       RCRA AMENDMENT EXPERIENCE

-------
PHASE III: COST-EFFECTIVENESS INVESTIGATION
  The purpose of this phase was to investigate the cost-effective-
ness of applying RCRA Section 264 standards and guidelines to
closure of CERCLA sites. The investigation was conducted using
two case studies: the North Hollywood Dump Site in Memphis,
Tennessee and the Hardeman County Dump  Site (also called the
Velsicol Chemical Corporation Dump) near Toone, Tennessee.
Both sites are on the CERCLA National Priority List and have
been closed under the original NCP policy. Three site conditions
were considered: (1) the original condition of the site when dis-
posal activities ceased prior to any remedial action activities, (2)
the condition of the site following closure under the original NCP
policy (as published on July 28,1982) and (3)  the condition of the
site if closed under the revised policy by following RCRA Part
264 standards and ensuing guidance.
   Costs for the cost-effectiveness determination  were  developed
 from actual closure costs under the original NCP policy including
 emergency  remedial  action  activities,  Remedial Investigation/
 Feasibility Study (RI/FS) development and covering/capping of
 the sites. Both sites under study have been closed under the orig-
 inal NCP policy, so actual costs for closure were available.11'12'13
 Estimated costs for closure under RCRA Part 264 included costs
 for emergency remedial actions, extensive RI/FS  development,
 ACL determination and a RCRA cap as required under Subpart
 G guidance. Costs for extensive RI/FS development and ACL de-
 termination for the sites were drawn from U.S. EPA Region 4
 estimates and work plans.14'15'16 Costs for RCRA capping activ-
 ities were based on  contractor estimates  for cap construction."
 The cost of the closure activities conducted when disposal ceased
 (nominal covering) was not taken into account for  this study,
 allowing the original condition to be the baseline for cost. Table 1
 provides a summary and comparison of costs for  closure under
 each policy for Hollywood and Hardeman County Dump Sites.
   The cost for closure of the Hollywood Dump using RCRA Part
 264 standards was estimated to be $9,212,300, over four times the
 cost for  the  current  closure  under  the original NCP  policy
 ($1,693,600). The largest cost increase was due to the addition of
 $5.07 million for RCRA capping activities.
   The difference  between closure costs for Hardeman County
 under the two policies was less dramatic, although  there was still
                           Table 1
  Summary and Comparison of Costs for Closure Under Each Policy for
           Hollywood and Hardeman County Dump Sites
  Site/Activity
                         Cost in Dollars
                    original       RCRA
                     Policy        Policy
                                                Increase in Cost3

                                             Dollars      Percent
HOLLYWOOD DUMP SITE
                       $  605,000   $1,963,500   $1,358,500  225%
RI/FS
Alternate Concentration
   Limit Determination       	b    1,091,800    1,091,800
                                             5,068,100
 Remedial Action

 TOTAL CLOSURE COST
                      1,080,600

                     $1,693,600
 6,157,000

$9,212,300
$7,518,700   "i<
-------
  The Hazard Ranking System was not developed to assess clos-
ure options and may not be sensitive enough to properly assess
effectiveness of the two policies. However,  it cannot be  disre-
garded. The model was developed by the U.S. EPA for the pur-
pose of assessing site hazards and is one of the nationally standard-
ized systems for prioritizing hazardous waste sites. Due to  the
many uncontrolled influences  on a site, the physical differences
between closure under the original NCP policy and closure under
the revised NCP policy may not be significant when incorporated
into the variable framework of the entire site surroundings.
   The second method used to assess effectiveness involved the use
of a computerized deterministic groundwater model to quantify
risk associated with each site in terms of annual and lifetime mor-
tality risk. The model used hydrogeological, demographic, chem-
ical, physical and lexicological data with  a one-dimensional ad-
vection-dispersion equation to project contaminant  concentra-
tions and levels of annual and lifetime mortality risk at selected
points." This groundwater risk model is  schematically depicted
in Figure 1. The acceptable level of lifetime mortality risk for this
study was based on the U.S. EPA's designated level of 10~4 to
 J0-8M-2*
   The hazards under consideration for each site were drinking
 water risk associated with contaminated groundwater, direct con-
 tact/erosion  hazards and  risk associated with  consumption of
 contaminated fish in nearby surface waters. Indicator parameters
 used for modeling were chosen based on toxicity, prevalence in
 the site  and relative mobility.  Chlordane was chosen as the indi-
 cator parameter  for Hollywood Dump Site.19-27 Carbon tetra-
 chloride, chloroform and heptachlor were chosen for Hardeman
 County Dump Site to respectively represent two highly soluble,
 prevalent compounds and one highly insoluble low level concen-
 tration compound.20-22
   Table 3 summarizes the results of this risk modeling effort for
 each site under the three site conditions being studied. A variety
 of site specific and technical information references was used to
 develop these results.19'20-22-27-"
   Drinking water risk for the Hollywood Dump Site was extreme-
 ly low under the original site conditions due to the deep layer of
 relatively impermeable clay beneath the  site. Since the original
                                                                                 INFILTRATION
                                                                                                  EXPOSED
                                                                                               POPULATION
                                                                    DEGRADATION,
                                                                    SOLUBILIZATION,
                                                                    LEACHING
                                               INGESTION
                                                          RISK
                                                                               ADVECTION,
                                                                               DISPERSION
                                                                               RETARDATION,
                                                                               DEGRADATION
                                                                             Figure I
                                                           Schematic Diagram for Groundwater Risk Model
                                                  risk was so low, mitigation would be unnecessary. Therefore, clo-
                                                  sure under either policy would not be cost-effective for mitigation
                                                  of this hazard. Alternatively, the drinking water hazard for the
                                                  Hazardous County Dump Site was so high that neither closure
                                                  method could adequately remediate the site to provide acceptable
                                                  risk levels. Therefore, the groundwater would still be unusable
                                                  and neither policy provided a cost-effective solution.
                                                    Since model calculations were based on  ingestion,  direct con-
                                                  tact hazards were not quantified.  Direct contact hazards were
                                                  considered to be completely mitigated under either  method by
                                                              Tible3
                   Summary of Risk Reduction Provided by Closure Options, North Hollywood and Hardenwn Count; Damp Sites
                    Site/Hazard Category
                                                       Maximum Lifetime Mortality Risk
                                                Original
                                                Condition
                                                                                                   Risk Reduction
                                              Original
                                               Pol icy
               (RCRA) Policy
                                                                                                 Policy
                                           R«vi««d
                                         (RCRA) Pol.o
North Hollywood Dump

   Drinking Water
   Direct Contact/Erosion
   Fiih Ingettion
       (Maximum Consumption)

Hardeman County Dump

   Drinking Water

       Carbon Tetrachloride,
         ecu
       Chlorolorm, CHCIj
       Heptachlor, C\oHjClj

   Direct Contact/Erosion

   Fish Ingellion
       (Worst Case Scenario
       Minimum Consumption)
                                               9.0 « 10'' J
                                               _ c

                                               3.9
-------
creating an earthen barrier which eliminates access to waste ma-
terials. For this hazard category, the additional cost for closure
under the more stringent and expensive revised policy is not con-
sidered cost-effective.
  The risk associated with fish ingestion at both sites under orig-
inal conditions was above the acceptable EPA  range of 10~6 to
10~8. Risk is decreased by closure under the original NCP policy
at each site but remains just above acceptable levels. The applica-
tion of RCRA Part 264 standards draws the risk value within the
U.S. EPA's acceptable range but results in a dramatic increase in
costs (see Table 1).
  At this point, the cost-effectiveness issue reverts to a more sub-
jective debate. The type of risk is a voluntary risk where the ex-
posed individual has been warned of the risk through posting of
signs and,  at Hollywood, has been limited by fencing. The risk
calculations were based on worst case conditions using high solu-
bility values and the unlikely consumption rate  for an individual
of 21  kg/yr of fish strictly from these waters every year for 70
years."  Recognizing  that the citizens have been appropriately
warned, that the risk values were purposely inflated and the risk
under the original NCP policy is close to the suggested U.S. EPA
risk range, it  is concluded that the  greatly increased cost asso-
ciated with closure under  the revised policy is not justifiable for
purposes of decreasing fish ingestion risk.
  Figure 2 provides a graphical representation of the costs  of
closure under the two policies versus the lifetime mortality risk
associated  with  fish ingestion at both sites and with drinking
water (carbon tetrachloride and chloroform) at Hardeman Coun-
ty.  The U.S.  EPA target risk range is shown for comparison.
Additional comparison is provided by the insertion of some "gen-
erally accepted" risks along the top of the figure."
CONCLUSIONS
  RCRA Part 264 standards provide a reasonable framework for
assessment and cleanup of hazardous waste sites but strict adher-
ence to standards and guidance eliminates the flexibility neces-
sary for remediation of CERCLA sites.  RCRA Part 264 Subpart
F—"Groundwater Protection"  standards  outline  a  rational
methodology for assessment  and remediation of groundwater
contamination.  However, compliance with  Appendix VIII  and
ACL determination requirements would be time consuming, cost-
ly and  technically questionable. If  problems in dealing  with
Appendix VIII analytical requirements and ACL  determinations
can be resolved, the approach outlined in Subpart  F would be de-
sirable for application to CERCLA  sites. RCRA Part 264 Sub-
part G—"Closure and Post-Closure" standards in Section 264.11
are reasonable but the standards specific to landfills in Subpart
N, Section 264.310 and  the ensuing  guidance  are too restrictive
to be technically valid for application to all CERCLA sites.
l£





10


8





6



4



2

H
0
2
z
UJ
E
o
Ul
UJ
f 5

KEY

3 f

u-
>
h-
3
, -,.8-
1


OO •-FISH ING. RISK


" , ' ' * -

s.



_. *

- HOLLYWOOD
DOB- FISH ING. RISK- HARDEMAN CO.

O^^ -DRINKING WATER CH

Ct>- HARDEMAN Q
A A A- DRINKING WUIER CCU-HARDEMAN Cc
- O DO ^~ ORIGINAL CONDITIONS
Q B  £ -CURRENT
^ B • A - PROPOSED








h




EPA

' "•
-9 -a
10 10
POLICY
POLICY

' j •"/-,





d
UJ
_l

1-
UJ
a:
i-
<
a.
, UJ



TARGET
Z
'* , •''.'•
*?

£
^
? CO
*""; , • ' a:
o
CO
i
\ ,
\
, -\
V-
, - ; \
1 ' , - V> \
\
\
',; , , "^
' - \X
" f • '"' ' \\
^ \
FISH INQESTION RISK_
HOLLYWOOD PUMP



RISK RANGE
'

-7 -6 -i
10 10 10







*


-







M

_i 3 _,
^ « O O ° o
< * z s S 5
Q_ ™~ ^
UJ rf 3 /"» 5 —
£ g i 1 i
S |u § v 5 i T
O < ? O ^ (J
1 f }t ^ r v
ACCEPTABLE MCL'S






*V \
\,
\
vi FISH INGESTION RISK \ \
VT — HARDEMAN CO. DUMP \
\ \ \ \
\ ^. DRINKING WATER RISK, CHCIj 	 .) \
J, I ^ HARDEMAN CO, DUMP ^ ^
\ X \ \


^—

\ \ DRINKING WATER RISK, CCU 	 »\ ^A
\ \ HARDEMAN CO. DUMP v \
\ \ \ \
"- \ ^
\ \ \ \
V V, \>
^ ' U ' ' t£
^^ fc=f «^
-4 -3 -Z -1
10 10 10 10 1
 o
 a
 O

 I
 o
 o
 UJ
 CE
 V)
 3
 o
                                               LIFETIME   MORTALITY  RISK
                                                          Figure 2
                                           Closure Cost Versus Lifetime Mortality Risk
                                                                                   RCRA AMENDMENT EXPERIENCE    373

-------
  The cost effectiveness of the original  and the revised policies
was investigated through estimates of cost and quantification of
risk reduction. It was determined that  the revised policy, requir-
ing application of  RCRA Part 264 standards and ensuing guid-
ance to CERCLA  sites, was not cost-effective for mitigation of
any of the hazard categories investigated at either site studied in
this report. This determination was based on the high incremen-
tal cost required to achieve little or no  incremental risk reduction
under the revised policy.
  The proposed revisions to the NCP required compliance with
"all applicable or relevant" Federal public health standards. The
final  revised version of the NCP changed these  words to read
"all  applicable or relevant  and appropriate."1  It  is  recom-
mended that RCRA Part 264 standards be considered as a refer-
ence rather than a requirement for CERCLA sites and that site
specific conditions dictate when  RCRA  Part 264 standards are
"appropriate."
REFERENCES
 1.  U.S.  EPA, "40 CFR Part 300, National Oil and Hazardous Sub-
    stances Pollution  Contingency  Plan—Proposed Rule,"  Federal
    Register, 50, No. 29. Feb. 12, 1985.
 2.  "Revised  National Contingency Plan Reflects  Proposal, Includes
    "How Clean is Clean Policy," Environmental Reporter, Oct.  18,
    1985, p. 1044.
 3.  "Resource Conservation Recovery Act of  1976," Public Law  94-
    580, 94th Congress, Oct. 21, 1976.
 4.  U.S.  EPA, "40 CFR  Part 264, Standards for  Owners and Oper-
    ators of Hazardous Waste Treatment Storage and Disposal Facilities
    —Final Rule," Federal Register, 47, No. 143, July 26, 1982.
 5.  "Comprehensive  Environmental  Response,  Compensation,  and
    Liability Act of 1980," Public Law 96-510,  96th Congress, Dec.  11,
    1980.
 6.  U.S.  EPA, "40 CFR Part 300, National Oil and Hazardous Sub-
    stance Contingency Plan—Final Rule," Federal Register 47,  No.
    137, July 16, 1982.
 7.  U.S.  EPA, "40 CFR Part 300, National Oil and Hazardous Sub-
    stances Contingency Plan—Notice of Effective Date,"  Federal
    Register 47. No. 238, Dec.  10, 1982.
 8.  Price, C.M., Assistant Administrator for Enforcement and Com-
    pliance Monitoring and  L.M. Thomas, Assistant Administrator,
    Office of  Solid  Waste  and Emergency  Response, "Enforcing
    Groundwater Monitoring  Requirements in RCRA Part  B Permit
    Applications," Memo to Regional Counsels, Regions 1-10; Air and
    Hazardous Materials Division Directors, Regions 1-10, U.S. EPA,
    Washington, D.C., Aug. 16,  1984.
 9.  Office of Solid Waste, U.S.  EPA,  Priliminary Draft—Alternate
    Concentration Limit Guidance—Based on Section  264.94(b)  Cri-
    teria, Nov. 1984.
10.  U.S.  EPA, Draft—RCRA Guidance Document, Landfill Design,
    Liner Systems and Final Cover, July 1982.
11.  Velsicol Chemical Corporation, "Hollywood Dump Site Accounting
    Records, Memphis,"TN, Dec. 21, 1984.
12.  Apple, J.M., Deputy Director, Division of Solid Waste Manage-
    ment, Nashville, TN, Interview on Mar. 29, 1985.
13.  Velsicol Chemical Corporation, "Miscellaneous  Project Status  Re-
    port,   February  1985;  Hardeman County Dump   Accounting,"
    Memphis, TN, Feb. 1985.
14.  E.C.  Jordan  Company, "North Hollywood Dump,  Supplemental
    Remedial Investigation/Feasibility Study Work Plan," prepared for
    NUS  Corporation (Contractor to  U.S.  EPA),  Portland,  ME, Feb.
    1985.
 15. E.C. Jordan Company. "Addendum to Work Plan for Supplemen-
    tal  Remedial Investigation/Feasibility Study,  North Hollywood,
    Memphis, TN," prepared for NUS Corporation (Contractor to U.S.
    EPA), Portland, ME, Jan. 1985.
 16. Jeter, C.R., U.S.  EPA Region 4 Administrator,  Letter  lo Saralee
    W.  Terry,  Director, Tennessee State Planning Office, regarding
    Hardeman County Dump Site, Jan. 4,  1985.
 17. Loyd, R.,  Loyd Excavating Company, Murfrecsboro,  TN,  Inter-
    view on Mar. 26, 1985.
 18. Tennessee Division of Solid Waste Management, "Hazard  Ranking
    System Work Sheets and Documentation Record; North Hollywood
    Dump Site," Aug.  9, 1982.
 19. E.C. Jordan Company, "North Hollywood Dump  Environmental
    Assessment and Action Plan, Task Element V-B, Corrective Action
    Alternatives  Report," prepared for NUS  Corporation (Contractor
    to U.S. EPA), Portland, ME, Jan. 1984.
 20. ERM Southeast, Inc., "Environmental Evaluation and Assessment
    of Control Measures at the Velsicol Disposal Site; Hardeman Coun-
    ty, TN," prepared for Velsicol Chemical Corporation, Brentwood,
    TN.Feb. 1985.
 21. Tennessee Division of Solid Waste Management, "Hazard Ranking
    System Work Sheets and Documentation Records, Hardeman Coun-
    ty Dump Site," Aug. 6. 1982.
 22. Rima, D.E., Brown, E.,  Litz,  D.F.G. and Law, L.M., "Potential
    Contamination of  the Hydrogeologic  Environment from the Pesti-
    cide Waste Dump in Hardeman County, TN," United States Depart-
    ment of the Interior, Geologic Survey, Aug. 1967.
 23. Broshears,  R.E., PhD Dissertation—Working Papers,  Vanderbilt
    University, Apr.  1985.
24. "Superfund to Determine 'How Clean is Clean' Using Wide Can-
    cer Risk  Range," Inside EPA  Weekly Report, 5, No. 48. Nov. 30,
    1984. p.  1 and 6.
 25. U.S. EPA,  "Effect of RCRA on Remedial Superfund Activities."
    Hazardous Site Control Division In-House Training  Outline, Jan.
    1985.
 26. "Effect  on  RCRA on Remedial Superfund Activities," Hazardous
    Waste Report. 6, Feb. 4, 1985.
 27. E.C. Jordan Company.  "North Hollywood Dump Environmental
    Assessment  and  Action Plan;  Task  Element V-A, Data Interpreta-
    tion Report," prepared for NUS Corporation (Contractor  to U.S.
    EPA), Project No.  0705-01, Portland, ME,  Jan. 1984.
 28. Lyman,  W.J., Reehl,  W.F. and Rosenblatt,  D.H., Handbook of
    Chemical Property Estimation  Methods, McGraw-Hill Book Com-
    pany, New York. NY. 1982.
 29. U.S. EPA.  "Water Quality  Criteria Documents;  Availability,"
    Federal Register, 45. No. 231, Nov. 28, 1980.
 30. Zeith, G.D.. DeFoe.  D.L.  and Bergftedt. B.J..  "Measuring and
    Estimating the Bioconcentration Factor of Chemicals in Fish," J. of
    Fisheries Research Board of Canada. 35, 1979, 1040-1048.
 31. Jacobson,  J.C. and Hastings, D.W., Population Projections for
    Tennessee Counties by Age and Sex;  1990 and 2000; University of
    Tennessee Department of Sociology, Knoxville, TN, May 1983.
 32. Wilson,  J.T. and  McNabb, J.F., "Biological Transformation of
    Organic  Pollutants in Groundwater," EOS, 
-------
                      Small  Quantity  Generators:  The  Maryland
                        Approach  to  Regulation and  Assistance

                                              Alvin L. Bowles, P.E.
                                Maryland Waste Management Administration
                                               Baltimore, Maryland
ABSTRACT
  With the enactment of the Hazardous and Solid Waste Amend-
ments of 1984 (HSWA), the universe of regulated hazardous
waste generators nationwide increased by 1,000%. The provis-
ions  for  small quantity generators are the prime reason for the
increase. While the U.S. EPA and various states rush to under-
stand the amendments and meet Congressionally mandated due
dates and requirements, the State of Maryland is actively im-
plementing its program.
  The Maryland Department of Health and Mental Hygiene re-
ceived the statutory authority for regulating small  quantity gen-
erators, that is, those generators producing 100  kg to 1,000 kg/
month of hazardous waste, in July  1984. This  authority came
from the State legislature. The law  is more stringent than the
HSWA provision,  and the  regulations supporting the  law are
more stringent than those proposed by  the U.S. EPA. These
facts, in addition to earlier effective dates for manifesting and
disposing at hazardous waste facilities, confuse the business com-
munity.  However,  the State developed an approach that helps
ease  the burden of notification and manifesting  requirements
and helps the small businesses to identify  their hazardous wastes
and means for disposal.

INTRODUCTION
  During the one-year grace period, from July 1, 1984 to July 1,
1985, provided in the legislation, the State  participated in  a multi-
media effort to inform the  newly regulated community and to
offer technical assistance and  information. The  approach in-
volved the Maryland Chamber of Commerce, the  University of
Maryland, the Controlled Hazardous Substances Advisory Coun-
cil, the Maryland Environmental Service and the Department of
Health and Mental Hygiene (DHMH).
  Through the production and distribution of  a  question  and
answer brochure,  mass mailings of regulatory requirements and
seminars with trade  associations  and county  governments, the
State reached the vast majority of affected businesses and munic-
ipal and  local governments. While the response may not  be con-
sidered favorable, it has been cooperative. The interaction with
the affected community has helped to identify problem areas in
existing regulations and has  led to regulatory amendments. Par-
ticularly  troublesome areas involved recycling  and  on-site treat-
ment at non-permitted facilities. The approach is proving  instruc-
tive to the regulatory agency and responsive to the needs of those
being regulated.

REACHING THE REGULATED
  The first response to the new legislation lowering the  exemp-
tion limit to 100 kg/month of generated hazardous waste was in
effect no response. Very few inquiries  were  received  after the
first  mass mailings. Notices of  new regulations and legislation
were probably treated as unwanted junk mail. Many were prob-
ably  never read by small business people.  When  the Adminis-
tration directly contacted the trade associations of some of these
small businesses, a frequent response was  "we do  not have a
problem; this legislation does not affect us."
  It was obvious to the Administration that the lack  of response
was not an indication of complete compliance, but  one  of not
understanding the breadth of the law. As a public agency, the
Administration could not wait for the law to become effective
and then enforce it against an unknowing regulated community.
That approach  can be very effective in getting attention and
compliance, but it creates resistance and ill will.
  During the initial phase of notifying the public, contact was
made with the Maryland Chamber of Commerce. It was felt that
the Chamber could further disseminate information to the busi-
ness  community. At the same time, the U.S. EPA was accept-
ing proposals for grants to assist in the dissemination of informa-
tion  about the small quantity generator requirements.  The
Chamber, along with the Maryland Environmental Service,  pro-
posed a multi-faceted program  involving the business, govern-
ment and university sectors with the Chamber taking  the lead.
Rather than draw on its resources only, the Administration joined
this effort. After several months of work, this group produced
a question and answer brochure that attempted to lead affected
persons through the maze of requirements of the new law. It was
distributed to nearly 7,000 businesses and organizations.
  In addition to the brochure and regulation mailouts, the Ad-
ministration made itself available to speak to various groups
and organizations to further explain the law  and requirements.
Also, a second part of the grant provided for technical assistance
by the University of Maryland.
  The total program is basically  structured as follows: the Mary-
land  Chamber  of Commerce is the lead agency since it has the
primary contacts with the business community. The Maryland
Environmental Service (MES), which is an agency  of the  De-
partment of Natural Resources, works closely with the Chamber
since it, too, has many contacts with industry; MES  also admin-
isters the grant. Also working with these two groups is the Con-
trolled Hazardous Substances Advisory Council, which is the
State advisory board for hazardous waste. It was reasoned  that,
with  the non-governmental groups  in the lead, the business com-
munity would be more receptive to the assistance.
  The role of the  Waste Management (WMA) is to explain the
requirements and  assist the business sector to come into com-
pliance. The University's role is to provide technical assistance to
business  on an industry-wide rather than individual basis: for
                                                                           STATE, REGIONAL & LOCAL PROGRAMS    375

-------
example, the University is prepared to work on problems of the
drycleaning industry, but not for individual drycleaning estab-
lishments. Hopefully they can study two or three representative
facilities  and develop solutions  that can  be  applied  industry-
wide.
  The business community is not the only regulated sector.  Fed-
eral,  State and  local government agencies  also  are  affected.
County school systems generate hazardous waste in their science
laboratories and maintenance shops.  State, county and munici-
pal maintenance shops generate hazardous waste from automo-
tive maintenance. People in these agencies usually do not apply
the requirements to themselves and certainly do not budget for
the increased costs of hazardous waste management.

PROBLEM AREAS
  The most significant problem for the small quantity generator
consists of two parts: understanding the  requirements and deal-
ing with the paperwork.  RCRA was the most massive regulatory
program ever, prior to  the HWSA of 1984.  The amendments
have expanded RCRA even more. Although only a small portion
of the Act affects  the small quantity generator, it is enough to
be overwhelming.  Small business people in  particular are  not
familiar with environmental requirements. They are not prepared
to deal with the information needs and the paperwork.  They sel-
dom have technical backgrounds or knowledgeable staff to coor-
dinate the paperwork.
  One must add that the increase in paperwork affects  the regu-
latory agency too.  All of the  manifests go to an agency for re-
view. The WMA receives 30,000 manifests per year as of FY
85. Estimates of the increase  for FY86 range from 10,000 to
80,000 manifests. These higher  numbers will  have a significant
impact on personnel resources.
  A second problem is the amount of helpful  information avail-
able.  One would think that more information on a topic would
be better. With  the small quantity generator requirements, this is
not the case. Primarily,  the problem  lies in the fact that Mary-
land implemented its own regulatory program  before RCRA was
enacted. Businesses, through their trade organizations, receive in-
formation on the Federal program.  At the same time, they re-
ceive  information on the State program. Since the two programs
are somewhat different, the business person becomes  confused
or frustrated.
  Recycling of hazardous waste is also  complicated. State  law
requires that hazardous wastes being recycled must be treated as
hazardous waste. In the case  of listed wastes and sludges,  they
must  be  manifested and transported by  certified haulers. The
statute also can be read to include characteristic wastes under the
transportation requirements. There is also a very thin-line distinc-
tion between what  is a recycling activity and what is  a treatment
activity.  Recycling  is  exempt  from permitting  activities  while
treatment must go  through  the entire RCRA  review and permit
process.  The  WMA  encourages recycling  and  makes  every
attempt to be open to new ideas.
  A good example  of the difficulty with recycling involved a new
car dealer association. The question was whether or not the body
shop  was a TSP facility. One dealer  had installed a solvent re-
covery still in  the  autobody shop. The distillation process was
very simple, consisting of a hot oil jacketed still and a condenser;
4 to 6 gal of paint solvent can be distilled in four hours. The clean
solvent is reused in paint or reused as a degreaser  in the shop.
The still bottoms are removed and shipped as  hazardous waste.
No off-site solvent is distilled.
  A fourth problem area that has serious potential is the existence
of numerous exempt sources.  When hazardous waste is shipped
from  a generator, it must go to a RCRA-permitted facility.  Ex-
empt quantities (that are non-liquid) may go to a permitted san-
itary landfill.  Exempt generators,  such as schools  and county
garages, may not generate more than 100 kg/month or  accumu-
late that quantity on an individual basis. In the aggregate, they
likely do exceed the exempt limit.
  Existing  regulations  prohibit the establishment  of a central
collection center unless it is a RCRA-permitted facility. In theory,
this center could serve as the generator. For example,  all the haz-
ardous waste from the school system would go to the central loca-
tion  where it is properly packaged and transported  to a proper
treatment, storage or disposal (TSD) facility. This approach can-
not be  used in Maryland because the central site is, in effect, an
unpermitted TSD facility.  Schools, being  individually exempt,
may then dispose of their hazardous  wastes in  municipal  land-
fills. Thus the move to more properly manage hazardous waste
from numerous small sources is abridged.
  Another problem  that is closely related to the previous one in-
volves the transporters. Maryland regulations allow for stoppage
of a vehicle hauling  hazardous waste for no more than 72 hr at a
permitted TSD facility. The waste load must be transported to a
facility or out-of-state within 5 days from initial pickup. Haul-
ing companies are having difficulty meeting these requirements
because of the dispersed location  of small generators and the
small quantities of generated waste. It takes longer than the per-
mitted  periods of time in the rural areas of the State to collect a
full truck load of waste.
  One  last area that may become a problem is transportation
and treatment/disposal capacity. Several trade organizations and
counties have raised the question: Is there sufficient capacity to
handle the additional hazardous waste? That is a difficult ques-
tion to answer.  Although the State has no hazardous waste  land-
fills and only three  TSD facilities that accept off-site generated
waste, there does not seem to be difficulty in shipping waste. The
real question may be a related question: Is there an economical
means  for  transporting/disposing of hazardous waste?  This
question is more pertinent  to the  small  generator,  and partic-
ularly so to a small generator in a rural area. The owner can ship
the waste if willing to pay the price. The capacity is there but per-
haps not within the financial means of the generator.

ANSWER TO PROBLEMS
  It is  very difficult to ensure that all the regulated community
understands the regulatory requirements.  As described previous-
ly, Maryland's  approach to multiple  mailings and seminars or
talks to specific groups is  the  most  practical approach. Keep-
ing the message simple and to the point is imperative. A  regu-
lated business owner need only hear what the State requires him
to do.  The owner does not need a discussion of what the U.S.
EPA and the other  states are doing. Information sent to a  busi-
ness  should provide the following: (1) regulatory requirements;
(2) types  of waste that the  business may generate that are haz-
ardous; (3) application packets for U.S.  EPA ID numbers; (4)
how to order and use manifests; and (5) telephone numbers for
help. Also, the State agency should  have listings of certified  haul-
ers and TSD facilities.
  The  anticipated increased volume of manifests did generate
concern in the WMA about how to cope with the increase. One
particular area where the number of manifests can be reduced is
the batch toll operation. This operation involves a company that
in effect leases  cleaning/degreasing solvents to commercial con-
cerns such as auto maintenance shops. When the material needs
to be replaced, the  company delivers new solvent and removes
the old. The company claims that  the material is not generated
by the  commercial  shop but  that it is owned by the company.
Thus, the company is the generator.
376    STATE, REGIONAL & LOCAL PROGRAMS

-------
  If the commercial shop or user is considered the generator, it
needs a fully completed manifest with each shipment of used sol-
vent. The WMA is not  anxious to receive possibly  50,000 to
60,000 additional manifests for this type of operation from one
company. The State, after review of its statutes and regulations,
reached a compromise: the company distributing the solvent is the
generator; the generator may use one manifest  per  collection
truck per day; each pickup must be recorded on the manifest, but
the name and location of the pickup is not needed; each solvent
(if more  than one at a pickup site) must be listed individually;
daily totals must be shown on the manifest; and  the agreement
only applies to generators of less than 1,000 kg/month  of haz-
ardous waste. It is estimated that this method of handling the
manifest  will reduce the  number of manifests by 80%  for one
company.
   Recycling activities have required extensive review of the Mary-
land statute and regulations along with the redefinition  of solid
waste. The WMA sought the advice of the Office of the Attorney
General in determining whether or not an  auto body shop redis-
tilling its own solvents was a treatment facility. The factors lead-
ing to the determination of whether or not it was  a TSD facility
and required a permit were: the distillation product is reused
directly in the original process;  it is generated on-site;  the still
bottoms  are treated as hazardous waste; and used solvent is not
stored before recycling.
   This decision fits very well into the redefinition of solid waste
as promulgated by the U.S. EPA on  January 4, 1985 (40 CFR,
Parts 260-266). However, there  are parts  of the new definition
that are difficult to understand and apply. It is going to be partic-
ularly difficult to apply the regulations to the small quantity gen-
erator.
   A question being asked is: "Can the waste from multiple small
sources in one system be collected at a central location and ship-
ped from that point?" The argument raised in favor of this con-
cept is that it is efficient to ship from one central location and
that it will provide for environmentally safe disposal. It is argued
further that, left to individual means, each school probably will
dispose of hazardous waste in municipal  waste  sites. This dis-
posal practice certainly is legal under the small quantity generator
regulations, but it defeats their intent.
   An answer is to allow diverse generators of a single school sys-
tem or government agency to centralize waste collection  without
requiring the central location to become a permitted TSD facility.
County and municipal governments are reluctant to pursue a haz-
ardous waste facility permit. This issue is under review.
  Another of the problems that does not have a definite solution
is that of stoppage. An attempt has been  made to allow longer
stoppage periods in Maryland (7 to 10 days). Consideration has
been given to  providing for stoppage or holding  terminals that
would be exempt from the need for a permit. These terminals
would be required to have a bond and to meet certain hazardous
waste management requirements. However,  no  transfer could
take place at the terminals.
  There are problems with this approach because enforcement of
storage times would be difficult. Also, haulers may choose to
use large trailers  to collect waste since no transfer at the terminal
would be permitted. This system could cause a safety hazard by
having large trailers  full of hazardous waste making collection
runs.  The WMA decided to leave its regulation unchanged and
review problems on a case-by-case basis.
  If there is a problem with TSD capacity, the business commun-
ity must solve it. The question is raised: "Why doesn't the State
open facilities?" The State, through MES, unsuccessfully made
an attempt to operate a hazardous  waste landfill. For  various
reasons, it could not compete with industry and was closed. The
difficulty  with bringing industry into the off-site TSD facility
business is that the cheap and easy method of landfilling is no
longer a viable alternative. A new hazardous waste landfill will be
extremely difficult if not impossible to build. The other methods,
incineration and chemical treatment or recycling, are expensive
and technology-intensive. Supply and demand will create the in-
centive and wherewithal to deal  with any capacity problems. A
market is emerging.

CONCLUSIONS
  The small quantity generator requirements appear massive on
first impression. With work, they can be made less imposing and
can be presented to  the generator in understandable language.
Every effort needs to be  made to accommodate  recycling and
disposal needs within the regulations. Waste from small quantity
generators can be minimized as evidenced in the recycling efforts
of the automobile dealers.  If an agency takes too strict a posture,
the  small quantity generator may seek out environmentally dam-
aging or illegal alternatives for disposal.
                                                                               STATE, REGIONAL & LOCAL PROGRAMS    377

-------
                           Siting  Efforts  in  Southern California

                                                  Kieran D.  Bergin
                       Los Angeles County  Hazardous  Waste Facility  Siting Project
                                                Whittier, California
INTRODUCTION
  The management of hazardous waste in Southern California is
closely linked to the history of the  development of this area.
Southern California, particularly Los Angeles County, developed
extensively following the turn of the century with  an even speed-
ier pace after the second World War. There were  several factors
that predisposed the area to rapid development at this time. The
first  was the fact that large quantities of water  were available
throughout  Los  Angeles  County,  primarily from imported
sources. In the  1910s, the  City of Los Angeles constructed an
aqueduct to bring water approximately 270 miles from the Owens
Valley (in the Eastern Sierra Nevada Mountains) to Los Angeles.
In the 1930s work began on  an aqueduct from the Colorado River
(at the California-Arizona border) to move water  approximately
300 miles across the Colorado desert.
  A second factor that was  very important in the region's growth
was the presence of large petroleum deposits. Development of oil
began around the turn of the century in Los Angeles and financed
a great deal of the public improvements that were constructed in
the form of aqueducts. Oil  also provided a great deal of the eco-
nomic vitality of the region.
  A third factor that played a very large part in the development
of the  region was  its climate. The Mediterranean-type climate
of Southern  California was  an excellent place for testing aircraft.
This  led  to  a cluster of aircraft manufacturers, including the
Douglas Aircraft Company, Northrop, Lockheed and Hughes.
This industry, which was very important during the second World
War, eventually led to a high concentration of technical talent in
the aircraft and,  later, the  aerospace  industries and fostered the
development of the aerospace, defense and transportation indus-
tries as major components of the Southern California economy.
  The last factor that contributed to  the development of South-
ern California was a general attitude  on the part  of government
that government intervention was not only acceptable, but neces-
sary, in order to develop and foster the economy of the area. This
philosophy led to government involvement in the development  of
aqueducts,  in the dredging and  filling of harbors  around the
Ports of Los Angeles and Long Beach and in the construction  of
an extensive system of  flood control that allowed the develop-
ment of many otherwise unavailable areas.

HAZARDOUS WASTE DISPOSAL HISTORY
  The  history  of hazardous waste disposal in  Southern Cali-
fornia is  not as well recorded as that of other major develop-
ments.  However, evidence  from the unearthing of old industrial
sites and from the legacy of past improper disposal gives us a rea-
sonably good idea of what  occurred. Prior to 1965 most of the
hazardous wastes generated by the region's industries were dis-
posed of at the point of generation, offshore or into the closest
municipal landfill. Around 1965 three Class I disposal sites were
opened in the Palos Verdes Peninsula, the San Jose Hills in the
City of West Covina and in the Santa Monica Mountains west of
the San Fernando Valley. These sites, which were allowed to take
virtually all wastes except  radioactives and explosives,  became
the backbone of the disposal system for hazardous waste within
the County for approximately 20 years.
  During this same time period, the nation was becoming more
conscious of the environmental consequences of its actions and
began, through a series of Federal Laws, to clean the nation's air
and water. Major federal legislation in these areas was enacted in
1970 and 1972, respectively. These laws, which were designed to
protect the air and water, did so by concentrating pollutants that
formerly were dispersed and placing them upon the land.  In the
late 1970s, it was becoming increasingly obvious that all three en-
vironmental media (air, water and land), needed protection, and
that careful attention had to  be paid to pollutants going into all
three.

RCRA IMPACT
  In 1976 Congress passed the first version of RCRA which made
much progress in regulating hazardous wastes but, unfortunately,
did little to force hazardous wastes to go to treatment facilities.
In  1984, with the  major  amendments to  RCRA,  Congress
changed the law radically and will require virtually all hazardous
waste to receive treatment prior to land disposal.
  In 1980 two of the previously mentioned sites, Palos Verdes
and Calabasas, were dosed. The first because it was full and the
second because the  geologic requirements  for hazardous waste
disposal sites had changed (whereas the geology  of the site had
not). This left Los Angeles County heavily dependent upon one
hazardous waste disposal site, the BKK Landfill in West Covina.
Pressure immediately began to build; the citizens of West Covina
wanted closure of this landfill, which they felt represented an ex-
cessively large threat to the community.
  In early 1981 the Environmental Protection Agency, the State
Health Department,  the State  Water Resources  Control Board
and the Counties  of Ventura, Los Angeles, Orange, San Diego,
Imperial, Riverside and San Bernardino banded together to form
the Southern California Hazardous Waste Management Project.
The initial charge of this study was to find replacement disposal
sites for BKK.

Search for Alternatives
  There was soon an increasing awareness at the State level that
treatment without disposal would be unacceptable both environ-
mentally and politically. The efforts of the Southern California
378    STATE, REGIONAL & LOCAL PROGRAMS

-------
project were redirected to include substantial work toward getting
waste treated  close to its generation. This included performing a
detailed waste characterization  and facility  needs  assessment,
developing goals and policies for hazardous waste management
and developing detailed siting criteria for hazardous waste treat-
ment and disposal facilities. However, the redirected project was
unable to come to terms with the hard practical question of where
to site hazardous waste treatment and disposal facilities.

BKK Closes to Hazardous Waste
  At that time it was assumed that there were  many locations
where hazardous waste treatment could be performed  and that
the private sector would  be  willing to establish treatment facil-
ities in Southern  California. As it turned out, the extremely vola-
tile nature of the business  of hazardous  waste management,
coupled with the lack of any firm guarantees of business  for
treatment facilities, has thus far prevented  the establishment of
any major hazardous waste treatment facilities in Southern Cali-
fornia. Therefore, when the BKK Landfill announced  that  it
would cease accepting hazardous waste at the end of November
1984, the Board of Supervisors of Los Angeles County asked the
Los Angeles County  Sanitation Districts and the Department of
Public Works to prepare a report listing specific locations that
would be suitable  for hazardous waste treatment and disposal
sites. The Board initially asked  for this report  in 90 days and
appropriated a total of $500,000 for this effort.

TASK FORCE ESTABLISHED
  The task force for this project began screening to eliminate
grossly unsuitable  areas  from any further consideration. This
process was carried out for both treatment and disposal facilities.
For treatment facilities, it was decided that compatible land use
was the most  important criterion and, therefore, the search cen-
tered on the industrial areas in the County. Other significant fac-
tors entering  into the siting  of a treatment facility were access
(primarily by road), isolation from homes and availability of ade-
quate sewer connections to remove treated waste water.
  After the industrial areas were screened using these criteria, an
industrial real estate  firm was hired to find available properties
between 5 and 15 acres in size that potentially could be used as
waste treatment facilities. The realtor reported that there were 56
such properties available. The project then  studied these further
and came out with a list of the best 20.
Potential Sites Identified
  Concurrent with the search for waste treatment facilities, a
search for disposal sites in the County was done.  Los Angeles
has a  population of approximately 7.5 million  people,  and a
variety of land uses preclude waste disposal sites in  all but a few
areas of the County. State law requires that waste disposal facil-
ities be at least 2,000 ft from any homes. The central part of Los
Angeles  County is occupied primarily by National Forests which
are excluded from consideration  by Federal  law.  In addition,
most of the valleys are geologically unsuited for hazardous waste
disposal.
  Using these criteria, eliminating earthquake zones, flood-prone
lands and significant ecological areas, the project quickly focused
on eight areas as potential sites for  hazardous waste disposal.  The
disposal operations that are envisioned in these areas would be for
treated residuals only  and would  not take raw,  untreated haz-
ardous waste. The sites would accept no liquids and would be
specially engineered and operated to keep rainwater infiltration to
a minimum.
  Preliminary geologic investigations of available literature were
performed on all eight perspective locations.  One piece of land,
on County-owned property, was drilled to determine if the sub-
surface met the promise that the surface provided. At the time of
this writing, attempts are being made to obtain permission to drill
three other pieces of property.
CONCLUSIONS
  In 1985 several  of the  counties around Southern California
formed a Joint Powers Agency dedicated to better management
of hazardous wastes throughout the region. At the time of its in-
ception, Los Angeles County (which  dominates the region  in
terms of waste generation)  declined to join this Joint Powers
Agency. The regional political dynamics of Southern California
make it imperative for Los Angeles to have a system in place for
managing its wastes. Unless this happens, other political jurisdic-
tions will be reluctant to site hazardous waste management facil-
ities, fearing that these will become magnets for waste from Los
Angeles.  Therefore, it is critical that Los Angeles County lead
the area in terms of hazardous waste management in  Southern
California.
                                                                               STATE, REGIONAL & LOCAL PROGRAMS    379

-------
                    The California Site Mitigation  Decision Tree

                                                   Paul W. Hadley
                                                    William Quan
                                        Toxic Substances  Control  Division
                                   California Department of Health Services
                                              Sacramento, California
ABSTRACT
  The California Department of Health Services has developed a
process that provides state decision-makers with a consistent ap-
proach to  mitigating abandoned  and uncontrolled  hazardous
waste sites.  This process, called the California Site  Mitigation
Decision  Tree Process, contains unique elements  that:  specify
how exposure criteria, known as Applied Action Levels (AALs),
will be developed; identify site characterization data that must be
collected or developed; present  preferred data collection strate-
gies  and  techniques;  and set appropriate site mitigation  criteria
for various  alternative remedial actions. Selection of the best
alternative remedial action is based upon technical and scientific
evaluations  concerning  protection of  public health and  the
environment while considering other nontechnical factors such as
demography, land use and input from the public.


INTRODUCTION
  The mitigation of  abandoned  and uncontrolled  hazardous
waste sites is a challenge faced  by federal, state and local agen-
cies across the United States. The  citizens of California demon-
strated their concern and commitment to facing this challenge
by overwhelmingly  passing Proposition 27, the $100 Million
Hazardous  Substance Cleanup Bond Act of  1984. With the pas-
sage  of Proposition 27, the Toxic Substances Control Division
(TSCD) of the California Department of Health Services (DHS)
was designated the lead state agency responsible for investigating,
characterizing and mitigating abandoned and uncontrolled haz-
ardous waste sites ranked on the California  Site  Priority Rank-
ing List (SPRL).
  To facilitate expeditious and  consistent mitigation of sites on
the SPRL as well as other  unlisted  hazardous  waste sites,  the
Alternative   Technology  and   Policy  Development  Section
(ATPDS) of TSCD developed the California  Site Mitigation De-
cision Tree Manual.  The Decision Tree Manual provides stale
decision-makers with a technical guidance document that pre-
sents a systematic and standardized  method  for evaluating the
existing and potential public  health and environmental risks asso-
ciated with  hazardous waste sites  under  various alternative re-
medial actions.
THE DECISION TREE PROCESS
  The Decision Tree Process,  as presented in the Decision Tree
Manual, consists of five basic components:

• Preliminary Site Appraisal
• Site Assessment
• Risk Appraisal
• Environmental Fate and Risk Determination
• Development of Site Mitigation  Strategies and  Selection  of
  Remedial Action
  Each of  these five basic components is made up of several
steps, procedures and decision points developed to provide a tech-
nical and scientific basis for selecting remedial actions to miti-
gate hazardous waste sites. Each of the five basic components
of the Decision Tree Process is discussed below. Also presented is
a discussion of the relationship between risk management and risk
appraisal under the Decision Tree Process.

Preliminary Site Appraisal
  The Decision Tree Process is activated through discovery and
ranking of an abandoned or uncontrolled hazardous waste site
on the SPRL. The procedures used  by the State of California to
assign a numerical score for migration potential and public health
and environmental impacts to a site are identical with those devel-
oped by the U.S.  EPA. However, once the site has been scored
according to its potential public health and environmental im-
pacts, the California procedure assigns a final rank to the site
after also developing a preliminary cost/benefit  analysis. Thus,
DHS can more efficiently manage its resources by focusing initial-
ly on those sites that can be most effectively mitigated.
  Those hazardous  waste sites with sufficiently  high numerical
scores for migration potential and public and environmental im-
pacts are submitted to the U.S. EPA to be placed  on the National
Priority List (NPL). Funds made available through passage of
Proposition 27 can be utilized lo provide the Slate's share of costs
associated with NPL sites within California. By utilizing  both
state and federal Superfund processes, DHS can direct stale re-
sources to mitigate those sites with  the greatest potential public
health benefits per unit cost and can still direct federal resources
to those sites that require the greatest costs.

Site Assessment
  Under the Decision Tree Process, hazardous waste sites are re-
garded as  consisting of source materials (including wastes and
contaminated soils) and exposure pathways that  facilitate trans-
port of hazardous chemicals from the  location of the source to a
point of exposure. Thus, there are two major objectives of the
Site Assessment component.  The first objective  is to character-
ize the nature and extent of chemical contamination in all perti-
nent media of exposure. The second objective is to characterize
the  migration or accumulation of chemical contaminants along
pertinent exposure pathways.
  The first objective of the Site Assessment component is  satis-
fied through the collection and chemical analysis of samples of
the  various exposure media to determine the extents of contam-
ination and concentrations of chemicals of concern. Guidance is
offered throughout the Decision Tree  Manual on preferred sam-
pling, handling and analytical techniques. A section on the neces-
sary quality assurance/quality control (QA/QC) procedures to
verify the accuracy and representativeness of chemical analyses
is also included.
380   STATE, REGIONAL & LOCAL PROGRAMS

-------
  With regard to the  second objective, pathway characteriza-
tion, the Decision Tree  Manual includes discussions  of those
data, parameters and interpretations necessary to characterize a
pathway. The manual also presents preferred and recommended
techniques  for  sample  collection and  analysis.  For  example,
recommended components of a drilling program may include, as
appropriate, continuous  coring, collection and analysis  of un-
disturbed samples, as well as paired borings where the strati-
graphy and observations  made in the first boring can focus and
direct the collection of  samples for chemical analysis  in the
second. With regard to groundwater investigations, discussions of
preferred data collection techniques include drilling methods, well
and piezometer design, construction, development and sampling.
  The representations of transport pathways are referred to as
modules and are developed from data collected in the Site Assess-
ment Phase. Each module may consist  of observations,  deduc-
tions, calculations, numerical models  and professional judgments
that allow the analyst to make scientifically and technically de-
fensible statements and conclusions regarding the behavior and
transport of chemical contaminants at the site.

Risk Appraisal
  In the  third component of the Decision  Tree  Process,  Risk
Appraisal, a basis is developed for evaluating risks  of adverse
toxicological effects associated with exposures to chemical con-
taminants identified at  a given site. The basis for evaluating the
risk of an adverse toxicological effect is  a  numerical  criterion
known as the Applied Action Level (AAL).
  The key to understanding the risk appraisal concept is that an
AAL is an exposure criterion. An AAL is developed in  a consis-
tent and documented manner  as  specified in the  Decision  Tree
Manual and is intended for use only to evaluate  potential  risks
of adverse effects associated with chronic, low level exposure to
toxic chemicals on hazardous waste sites. The appropriate loca-
tion for evaluating potential risk through the Decision Tree Pro-
cess is at the terminus of an exposure pathway where a sensitive
organism, referred to as a sensitive biological  receptor, may come
into contact with toxic chemicals associated with the particular
hazardous waste  site. An AAL is a media-specific, biological re-
ceptor-specific  criterion that is applicable statewide to  evaluate
the risk of an adverse toxicological effect from exposure to toxic
chemicals.
  The formal evaluation of risk associated with exposure to con-
taminants  on a hazardous waste site can be accomplished by
applying a series of three risk appraisal tests. These risk appraisal
tests evaluate single medium/single chemical exposure,  multiple
media/single chemical  exposure  and multiple media/multiple
chemical  exposure,  respectively.  Risk  appraisal  tests  are  per-
formed first to determine if existing  exposures of biological re-
ceptors present a risk of an  adverse  toxicologic effect. If so,  a
risk management process should be initiated to mitigate  that risk
by an interim remedial measure.
  The first test in the risk appraisal process compares the level of
exposure in a medium,  abbreviated as Cmedium, with the AAL
criterion for that medium, AALmedium- The first risk appraisal
test is:
 If
                   medium
(1)
then the test fails, a sensitive biological receptor is considered to
be at risk of an adverse impact and a risk management process
should be initiated.
  The second test in the risk appraisal  process considers risks
posed to a sensitive biological receptor  through exposure to a
         single contaminant via all  pertinent media of exposure. The
         second risk appraisal test is:
         If
                      medium  =  1
                             medium
                           AAlmedium
                                               >  1,
                                                                  (2)
         then the test fails, a sensitive biological receptor  is considered
         to be at risk of an adverse impact and a risk management pro-
         cess should be initiated.
           The third test  in  the risk  appraisal mechanism  determines
         whether a sensitive biological receptor may be considered at risk
         due to exposure to an aggregate of substances that produce the
         same toxic manifestation. Lacking data that are contradictive of
         the assumption of additivity, the third test in the risk appraisal
         mechanism is:
         If
                    Sub = 1  Medium  = 1
                                  medium.  sub
                                             sub
                                                           >  1,
                                                                  (3)
then the test fails,  a sensitive biological receptor  is considered
to be at risk of an adverse impact and a risk management process
should be initiated.

Environmental Fate and Risk Determination
  The fourth component of the Decision Tree Process, Environ-
mental  Fate and Risk  Determination,  is designed to evaluate
potential future risks associated with a particular  site. The en-
vironmental fate and transport modules  developed from  data
collected under the  Site Assessment component are employed in
the fourth  step to project  environmental  impacts  from current
conditions  associated with  the site to future conditions on and
distant  from the site at points  where sensitive biological recep-
tors might be exposed to contaminants.
  The three tests of the risk  appraisal  mechanism  again are
applied to  evaluate whether sensitive biological receptors could
be at risk in the future, utilizing projected concentrations in the
various exposure media. Thus, the potential risks associated with
contaminants on a  waste site are evaluated for existing current
exposures under the Risk Appraisal Component and for possible
future exposures under this step, Environmental Fate and  Risk
Determination. If a sensitive biological receptor is evaluated to be
at risk through either risk appraisal or risk determination, a risk
management process should be initiated.

DEVELOPMENT OF MITIGATION STRATEGY
AND SELECTION OF REMEDIAL ACTION
  The first four components of the Decision Tree Process focus
on evaluating risks associated with exposure to  contaminants
from hazardous waste sites. Once it has been determined that a
sensitive biological receptor currently is, or will be at risk of being
adversely affected, mitigation of that risk should be investigated.
The development, evaluation and  selection  of remedial alterna-
tives to mitigate that risk are discussed in  this section. Although
this component is presented as the final section of the Decision
Tree Manual, activities associated  with  development,  evaluation
and selection of remedial alternatives actually should  begin with
the second  component,  Site Assessment,  and run concurrently
with the subsequent three components.  In this manner, remedial
alternatives can be evaluated and screened throughout the site
assessment and characterization project as constraints and limi-
tations  of various alternatives become  evident enough to justify
exclusion from further  evaluation. As the extent of contamina-
tion in each medium of exposure and the  processes associated
with transport and accumulation of contaminants along exposure
                                                                               STATE, REGIONAL & LOCAL PROGRAMS    381

-------
pathways become better  understood,  likely  remedial  strategies
and alternatives  can be  selected for further evaluation,  and
unique data needs can be met concurrent with other site assess-
ment activities.
   The contents of the last component have  been derived  prin-
cipally from those guidance documents available on  the  U.S.
EPA's feasibility study process. The Decision Tree Process, there-
fore, is consistent with the U.S. EPA's specified process of select-
ing remedial alternatives for federal Superfund sites. The Decis-
ion Tree Process also is consistent with  the  remedial investiga-
tion process set  forth in U.S.  EPA guidance documents; sile
characterization activities are designed to develop the data base
necessary for selecting the technically appropriate remedial action
plan.

Risk Management and Risk Appraisal
   The Toxic Substances Control  Division (TSCD) of  DHS is
organized into a central headquarters and three regional offices.
In general, those issues and problems that require uniform state-
wide solutions  are referred to DHS headquarters. Those issues
and problems that require solutions consistent  with other regional
operations are addressed by DHS regional offices, with technical
support available from headquarters upon request. One example
of this division  of responsibility is in the area of hazardous waste
management. The headquarters office of  DHS is responsible for
making determinations and classifications on a statewide basis as
to which wastes are hazardous and which are non-hazardous.
The regional offices of DHS are responsible  for  overseeing the
management of those wastes, including  issuance of waste dis-
posal variances where mitigating factors such as volume and con-
centration justify disposal in other than a Class I facility.
   With the passage of Proposition 27,  the activities and respon-
sibilities  associated with  mitigating hazardous waste sites  also
have been  distributed between  DHS headquarters and regional
offices. Through the development of the  Decision Tree Manual,
the DHS headquarters office has provided a consistent technical
approach to investigating, characterizing  and  mitigating hazard-
ous waste  sites. Through the ongoing development of AALs,
DHS headquarters also provides statewide  exposure criteria for
toxic chemicals associated with  waste  sites. The  DHS regional
offices, on the other hand,  are responsible  for the  day-to-day
progress at waste sites. This includes developing working relation-
ships with  other state and local  agencies  whose jurisdictions in-
clude environmental monitoring and management as well as inter-
acting with the interested public and their elected representatives.
   On an organizational level, staff in  DHS regional offices are
responsible  for risk management; staff  in DHS headquarters
offices are  responsible for risk appraisal, as well as for providing
technical support  and guidance.  Risk management includes pro-
ject management  to ensure collection  of  accurate and represen-
tative site characterization data, identification  of sensitive biolog-
ical receptors, identification of likely remedial alternatives based
upon a working understanding  of technical  and  cost consider-
ations  pertinent to each site and identifying  non-technical  con-
siderations including demographics, land  use,  politics and public
sentiment.  Risk appraisal  includes evaluation of risk associated
with potential exposures  not completely  eliminated by remedial
action. Thus, the DHS  approach allows development of risk
management decisions based upon a  consistent  risk appraisal
process.
   Finally, it is  noted that consistency in  the  risk appraisal pro-
cess does not equate to statewide uniformity in  site mitigation
criteria. For example, a level of residual contamination could be
allowed to  remain on one  particular site and poteniial exposures
to that contamination could be managed by engineering and legal
controls such as capping and a notice on the title to the property.
On  another site where the environmental,  economic,  demo-
graphic and political factors are different, it might  be unaccep-
table to allow that same level of contamination to  remain on the
site without physical control or management beyond that required
for the first site. In both cases, statewide consistency is achieved
by having  a uniform exposure criterion to evaluate risk and by
having a risk  management process that addresses the same gen-
eral  technical  and nontechnical considerations, though the rela-
tive importance of each issue may vary from site to site.
APPLICATIONS OF THE DECISION
TREE PROCESS
  Two representative examples of the application of the Decision
Tree Process are presented below.

Arsenic Contaminated Site
  This first example site had been a pesticide formulating facility
for more than 40 years; it is located adjacent to a saltwater marsh.
Extremely high levels of arsenic were measured in soils underlying
former waste disposal impoundments and storage areas and near
former loading and handling areas. Elevated levels also were
measured across the site as a result of waterborne transport of
contaminants by seasonal flooding.
  Arsenic contamination  also was observed  in shallow, saline
near-stagnant groundwater underlying this site. Deeper ground-
water zones approximately 200 ft below the shallow zone are ex-
ploited for drinking water at  a series of wells  located hydraulic-
ally upgradient of the site. These  deeper zones are separated from
the shallow zone by approximately 100 ft of low permeability de-
posits.
  As a result of site characterization studies, the site manager was
able to identify human beings as  the sensitive biological receptors
of concern and contact with  contaminated soils as the only sig-
nificant exposure pathway. The  remedial alternatives being con-
sidered included various amounts  of excavation and capping in
conjunction with notices attached to the deed on the property to
limit  potential  usage of the land, thereby limiting future ex-
posures to those associated with on-site activities that might dis-
rupt the capping mechanism. At this site, the site manager had
identified a  likely remedial action  plan. The Decision Tree Pro-
cess was therefore activated to make an appraisal of the potential
risks associated with a particular remedial alternative and to de-
termine what level of contamination could be  allowed to remain
on-site.
  Two exposure scenarios were considered for this site.  In the
first, exposures for on-site construction workers whose  activities
might disturb the capping mechanism were evaluated. The appro-
priate  exposure criterion for  this scenario was the 8-hour time-
weighted occupational standard of 0.01 rog/m' for arsenic in the
workplace rather than a long-term exposure criterion. The second
scenario was to evaluate  potential off-site exposure to wind-
blown dust particles contaminated with arsenic.
  The data  necessary to evaluate these scenarios included char-
acterization of on-site soils and pertinent meteorological informa-
tion.  In addition to environmental data, the literature and prac-
ticing industrial hygienists were surveyed to develop a reasonable
estimate of dust levels that might result from on-site construction
activities. An upper bound 25m8/mJ of on-site dust  and particu-
lates was arrived at through review of survey results. Based upon
this upper bound of 25m8/m-1, the level of arsenic allowable in on-
site  soils stays below the occupational exposure  standard of
0.01m8/3 arsenic and was calculated to be 400 ppm of arsenic.
382    STATE, REGIONAL & LOCAL PROGRAMS

-------
  The second scenario required evaluation of the potential off-
site exposures to wind-borne arsenic contaminated  soils  should
the capping mechanism be disrupted. The particulate emissions
from the site were estimated with standard air dispersion methods
described in the Decision Tree Manual. With a residual level of
400 ppm arsenic in on-site soils, the level of arsenic in airborne
particulates was undetectable and therefore would not represent
a measurable risk to nearby residents. Thus, through the exposure
assessments just described, a basis for setting site mitigation cri-
teria was established for a site where the question "how clean is
clean?" had previously seemed unanswerable.

Leaking Underground Fuel Tanks
  In California, a statewide task force has been established to
develop a guidance document on leaking underground fuel tanks
(LUFT). Members of the task  force include representatives of
various State agencies, county health departments and water
agencies. The format of the guidance document will follow that
of the California Site Mitigation Decision Tree Manual and is ex-
pected to be available for distribution in the spring of 1986.
  The LUFT task force is organized into subcommittees parallel-
ing the five basic components  of the Decision Treel Process.
Thus, as with the Decision Tree Manual, a multi-disciplinary
approach has been adopted that  allows task force  members to
contribute in their particular areas of technical expertise.
   Specific guidance will be presented on preliminary site apprai-
sal and site assessment. Applied action levels for selected com-
ponents of fuels are being developed, and the transport and en-
vironmental  fate of fuels are being examined so that the risk
appraisal and risk determination processes can be applied to leak-
ing underground fuel tank problems.  Remedial alternatives are
being researched and evaluated  with respect to their efficiencies
and applicabilities under various site conditions. This application
of the Decision Tree Process will fill a major need in California
for developing a technical guidance document to address the in-
vestigation and mitigation of contamination from leaking under-
ground fuel tanks.
CONCLUSIONS
  The California Site Mitigation Decision Tree Process provides
decision-makers with a consistent approach in investigating, char-
acterizing and mitigating abandoned and uncontrolled hazardous
waste sites. The Decision Tree  Process provides  a basis for
appraising potential risks associated with  alternative remedial ac-
tions  at hazardous  waste sites. The  Decision Tree  Process has
been utilized by the Department of Health Services and other
agencies within California and currently  is being followed  by a
multi-agency task force to develop a technical guidance docu-
ment on leaking underground fuel tanks.
                                                                                STATE, REGIONAL & LOCAL PROGRAMS    383

-------
                               Hazardous  Waste  Management:
                      The  Role  of  the  Local  Health  Department

                                               Michael J. Pompili
                                                 Philip G. Brown
                                         Columbus Health  Department
                                       Division of Environmental  Health
                                                 Columbus,  Ohio
ABSTRACT
  The Columbus Health Department (CHD) has  defined three
functional areas for participating in the  management  of  haz-
ardous waste on a local level. These areas are: (1) coordination
of local agencies, (2) emergency response and (3) review of haz-
ardous waste facility permit applications.  A core  group of city
agencies was formed at the Health Commissioner's request to re-
view existing local responsibilities relative to  hazardous waste
management. Among the documents produced by this group for
the city were notification procedures and functional responsi-
bilities of agencies responding to a hazardous material  incident
and evacuation procedures and policies.
  The CHD has identified its primary responsibility as the in-
vestigation of health effects from a hazardous material (HAZ-
MAT) incident. A procedures manual for CHD was developed
using the Centers for Disease Control document "A System For
Prevention, Assessment, and Control of Exposures and Health
Effects from Hazardous Sites" as  a guide. The CHD  manual
later was used for the investigation of health effects from  an acci-
dental discharge of hazardous substances within  the city by a
chemical plant.
  The CHD also has reviewed applications submitted by several
industries for federal and state permits to store hazardous waste.
These reviews have resulted in written agreements between the
City of Columbus and  the permittee that go beyond the federal
and state statutory requirements.

INTRODUCTION
  Since 1980, a national regulatory framework administered by
the U.S. EPA has been  in place to protect the public's health and
safety from the improper management or illegal disposal of haz-
ardous wastes. Under this federal legislation, states may  exercise
primary responsibility for administration and enforcement  of a
hazardous waste management program under state law in lieu of
the federal U.S. EPA program provided the state program meets
certain minimum federal standards.
  On a local level, however, a combination of federal and state
rules often precludes local authorities from passing laws and/or
adopting  regulations pertaining to the treatment, storage or dis-
posal of hazardous wastes.  Although such a preclusion  exists in
Ohio's hazardous waste laws,' the Columbus Health Department
has attempted to initially define  a role for our local government
in managing hazardous  wastes and materials.

REVIEW OF PERMIT APPLICATIONS
  The CHD's  initial involvement began with the review of a
Part B Permit application for a local branch of a national inde-
pendent distributor of chemical products. The company intended
to collect spent solvents in 55 gal drums from its customers and
store them  at its Columbus location until an economic truck-
load could be arranged (this might necessitate storing the waste
for longer than  90 days) for transfer to a reclaiming facility in
an adjacent state. The Mayor of Columbus, as a statutory party
to the judication hearing before the State of Ohio Hazardous
Waste Facility Approval Board (HWFAB), requested that the
CHD review the application and represent him at the hearing.
Since the application had been reviewed previously by both the
U.S. EPA and Ohio EPA for compliance with their respective
regulatory standards, the CHD's task was to identify the poten-
tial hazards posed by the facility to the  surrounding community
and attempt to minimize any adverse environmental impact pre-
sented by these hazards. The CHD's first action was to form a
committee of various city agencies to review the application.
  The committee was comprised of representatives from the city's
Fire, Police, Sewerage and  Drainage,  Water and  Health De-
partments. Recommendations from each agency were recorded.
Based upon tl'ese recommendations, a staff member of the CHD
prepared testimony for presentation at a  public hearing scheduled
by the HWFAB.
  Following the public hearing, a meeting was arranged between
the statutory  parties and  applicant by the administrative law
judge  for this case. The goal of this meeting was to reach an
agreement on those concerns raised  at the public hearing. With
the administrative judge acting as a facilitator, the CHD nego-
tiated  an agreement  with the applicant. The negotiated agree-
ment,  or Memorandum of Understanding,  incorporated  the
committee's recommendations and was to be  included as a con-
dition of the permit, if granted. The terms of this agreement in-
cluded the following: (1) quarterly inspections of the hazardous
waste storage area by the Division of Fire, (2) an alternative route
 for the  transportation of hazardous materials and wastes enter-
 ing and leaving the facility to bypass a nearby residential area,
 (3) the  installation of a backflow device to prevent contamina-
 tion by  the hazardous waste of the public water supply and (4)
 payment of the performance of a community health assessment
 survey by the Columbus Health  Department  in the event of an
 incident at this site.
   Two  additional Columbus facilities  have  submitted applica-
 tions for a  Part B Permit to the U.S. EPA. Each application has
 undergone  the  same review process as described above. Al-
 though  Memorandum  of Understandings have been signed by
 each applicant and the CHD, each agreement addresses  those
 concerns particular to that facility. For example, one company
 which failed to make familiarization arrangements with the Fire
 and Police Divisions as required  in Section  264.37 of 40 CFR
 signed an agreement that such arrangements would be  made.
 Another company agreed to permit the Division of Water to
384    STATE, REGIONAL & LOCAL PROGRAMS

-------
conduct a cross connection survey of its facility and then cor-
rect any identified deficiencies. To date, only three of the 13 haz-
ardous waste storage facilities in the City of Columbus have
sought a Part B Permit; consequently, the Health Commissioner
has requested the formation of a standing committee to review
the remaining applications.

COORDINATION OF LOCAL AGENCIES
  The  CHD's role in the management of hazardous substances
within  the city expanded as a result of the application review
process. Within a year, the CHD had initiated the development
of three documents which became the backbone of the city's re-
sponse to hazardous material incidents.
  At the request of the Health Commissioner, a committee com-
posed of representatives from Emergency Management Services,
the Divisions of Fire  and Police,  the local chapter of the Amer-
ican Red Cross and the CHD was established to review and re-
vise evacuation procedures to be taken during a hazardous  ma-
terials  incident. After months of discussion, the committee pro-
duced  a document that contained a series of guidelines designed
to serve as an assurance that every reasonable effort would be
made to preserve life and protect property but at the same time
recognizing that these policies and procedures could not be a gov-
ernmental guarantee  of absolute  safety to area residents when it
was necessary to withdraw from an area due to a hazardous ma-
terials  incident. Highlights of this document include: (1) six basic
premises that form the basis for the policies and procedures, (2)
organizational responsibilities of  these agencies responding to an
incident, (3) key activities associated with each phase of an evac-
uation process and (4) planning considerations for facility admin-
istrators who manage large populations which  generally are diffi-
cult to evacuate. The evacuation  policies and procedures are not
site-specific, but they apply to any fixed site or transportation in-
cident  which would  require an  evacuation of the surrounding
area.
   During the evacuation planning meetings, committee members
were cognizant of the need to understand the role of each agency
when responding to an incident.  Consequently, a document was
developed which formalized alert notification procedures and ex-
plained the  functional responsibilities of  responding  agencies.
The key to the notification chain is the Division of Fire. If, in
the judgment of a responding Battalion Chief, a hazardous inci-
dent warrants the expertise or assistance of another city agency,
the notification plan is initiated.

EMERGENCY RESPONSE
   Since the  CHD had taken the responsibility for investigating
human health effects from exposures to  hazardous  materials
which  occur within the  city, an intra-departmental procedures
manual was  formulated. The procedures within this manual were
based  upon  the Centers for Disease Control  guide,  "A System
for Prevention, Assessment and Control of Exposures and Health
Effects from Hazardous Sites."2  This response procedures man-
ual outlines  CHD's involvement in any response and describes
the methodology for performing a health assessment study.
  With internal response procedure in place, the CHD sought to
improve the Environmental Health Division's response capabil-
ities. To achieve this objective,  access to  a computer database
was obtained and  personnel protective gear and sampling equip-
ment were purchased. With the  acquisition  of  "Hazardline,"
an  on-line, interactive, time sharing database,  the risk assess-
ment process was made  more timely and efficient. "Hazard-
line"  contains hundreds  of pieces  of data for approximately
4,500 chemicals. The  information includes physical and chemical
properties,  needed  personnel protective  equipment,  chemical
toxicology, symptoms of exposure, and  leaks and spills  pro-
cedures.
  A self-contained breathing unit,  chemical  splash suits, half-
face respirators and a variety of other gear were purchased to pro-
tect the CHD's response personnel. For site assessment and inves-
tigation purposes, a high/low  flow pump,  charcoal collection
tubes and a detection tube test kit were purchased.
  A test  of  CHD's  internal response  procedures occurred 5
months after its initial involvement in spill response. On May 7,
1984, a chemical manufacturer accidentally released 2,000 Ib of a
phenol-formaldehyde resin into the air in less than 5 min. A slight
northeasterly wind blew the resinous mixture over a residential
area, although a  great percentage of the material  precipitated
onto the company's ground. Sixty-eight individuals in 18 house-
holds were potentially exposed.  The CHD collected water and
soil samples within a 2-mile radius of the plant. Initial samples
were analyzed for  phenol  and  formaldehyde with subsequent
samples analyzed only for phenol. Wellwater samples collected
from the 18 households indicated no contamination from the in-
cident.
  The  CHD's role within hazardous material incidents became
more clearly defined with this incident. It was evident that OEPA
and the Columbus Division of Fire were capable of handling the
emergency response procedures during this incident. What was
missing, however, was the evaluation of health effects on the area
residents. Ohio EPA response teams have a natural tendency to
worry more about the environment than the residents, while the
CHD's training is based upon worrying about the residents first
and the environment second. In this incident,  the two agencies
complimented each other's actions to the benefit of both parties.
  Upon completion of a study of the incident, the  CHD released
a document that explained what  was released, what amount was
released, who was affected  and whether any long-term health
effects were expected.  Furthermore, the CHD recommended that
followup soil and  water  testing  for phenol be continued for a
definite period to ensure there  was no  contamination of the
ground water.  The company which had  the  accidental  release
agreed to pay for any and all required testing using CHD guide-
lines as well as the cost  of  the  complete CHD study. None of
the information about health effects to the residents would have
been available through any  government agency other than the
CHD.

CONCLUSIONS.
  Although local governments often lack the statutory authority
to regulate hazardous waste  facilities, much can be done outside
the legal framework to ensure that local concerns about such sites
are addressed. Without  benefit  of legislation,  the Columbus
Health Department secured voluntary agreements with local haz-
ardous waste facilities which surpassed federal and state legal re-
quirements.
  In addition, the CHD coordinated the formulation of spill re-
sponse procedures and evacuation policies which are implemented
through interagency cooperation. The CHD believes that local
governments have a role to play in the management of hazardous
materials and  wastes, notwithstanding the existence of formal
regulations; it is the CHD's conviction that this involvement must
complement, not duplicate, federal and state program efforts.

REFERENCES
1. Chapter 3734 of the Ohio Revised Code.
2. CDC, "A System for Prevention, assessment, and  Control of Ex-
  posures and  Health Effects  from Hazardous Sites (S.P.A.C.E. for
  Health)", Chronic Diseases Division of the Center for Environmental
  Health, Centers for Disease Control, Atlanta, GA, Jan. 1984.
                                                                               STATE, REGIONAL & LOCAL PROGRAMS    385

-------
                    U.S. EPA's  Initiatives for Expanded  Public
                 Involvement in  the  RCRA  Permitting Program

                                                Vanessa Musgrave
                                    U.S.  Environmental Protection Agency
                                                Washington, D.C.
                                                    Edwin Berk
                                                ICF Incorporated
                                                Washington, D.C.
ABSTRACT
  The purpose of this paper is to provide an overview of the U.S.
EPA's  initiatives for public involvement  in the permitting of
hazardous waste treatment, storage and disposal facilities under
RCRA. The authors first discuss the rationale for fostering early
and expanded public involvement. They then  describe the ap-
proach and  main  steps recommended by the  U.S.  EPA to
facilitate public involvement in decision-making on a permit. The
authors conclude with a brief discussion  of some  of the U.S.
EPA's supporting activities in this area.

INTRODUCTION
  In the 1970s, the  potential threat to human health and the en-
vironment posed by hazardous wastes emerged as  the nation's
most prominent environmental issue. In scores of communities
throughout the country, citizens banded together to press for the
cleanup of abandoned hazardous waste disposal sites or to pre-
vent new waste management facilities from being created. More
often than not, these  citizens  had little prior experience with
political  activism  and  were far from  being  traditional en-
vironmentalists; their concern was over the direct personal health
and economic effects of a particular hazardous waste facility. It
usually made little difference whether the facility was inoperative
and already the cause of extensive contamination or newly pro-
posed and designed according  to the strictest safety guidelines:
hazardous wastes became perceived as potential threats under any
circumstances and  were regarded as generally  intolerable in
anyone's "backyard." The concerns raised at scores of localities
led to national attention and, ultimately, major federal hazardous
waste regulatory and cleanup legislation.
  Public  concern over hazardous waste has continued and, if
anything, intensified through the 1980s. The primary  source of
the concern remains localized; that is, public concern over hazar-
dous waste arises among citizens who consider themselves and
their children to be  directly affected by a particular  disposal site
or waste management facility. The U.S. EPA has recognized that
to fulfill its legislative mandates to protect human health and the
environment from the threat of hazardous wastes, an active, ef-
fective public involvement program is essential. There is no ques-
tion of whether or  not to involve the local public  in the U.S.
EPA's decisions. The question  is how to involve the local  public
so that public participation is constructive and results in sounder
environmental decision-making.
  In the  context of the permitting program  conducted  under
RCRA, the U.S. EPA's answer to this question has  taken the
form of initiatives to foster early and expanded public involve-
ment. These efforts are being pursued in accordance with the U.S.
EPA's  National  Permits  Strategy.  They  incorporate, but go
beyond,  public participation  requirements in RCRA section
7004(b)(l) and  in  U.S.  EPA  regulations on procedures  for
decision-making in 40 CFR Part 124, Subpart A.

IMPORTANCE OF EARLY PUBLIC
INVOLVEMENT IN RCRA PERMITTING
  The U.S. EPA's  initiatives for public involvement in RCRA
permitting are aimed at enabling the local public to have meaning-
ful and constructive input to decision-making on a facility's per-
mit. For public input to be meaningful, it must occur early in the
permitting process,  before  a permit has been drafted: that is,
before all the major decisions on the permit and the actual condi-
tions it is to include  have been made. The public participation re-
quirements in the regulations cited above, however, ordinarily
would result in public involvement late in the permitting process,
so the activities they call for must be supplemented. Thus, the
U.S. EPA's  National Permits  Strategy calls  for early and ex-
panded public involvement.
  There are several  reasons why meaningful and early public in-
volvement is important to the RCRA permitting program:
• Public involvement can yield  information of value to decision-
  makers in evaluating a permit application and drafting a per-
  mit.  For  example,  it can yield information concerning  the
  facility's operational history and site characteristics.
• Public opposition often arises when citizens feel they are being
  excluded from decision-making that will have direct and long-
  lasting  effects on their community; that they are being "rail-
  roaded" by a distant government agency. This kind of opposi-
  tion  is always unnecessary.  It can  largely  be  avoided by
  ensuring that the comments  and concerns of the local public
  are considered throughout the process of debate and delibera-
  tion leading to decision-making. The result will be a permitting
  process focused on the technical merits of a facility rather than
  on the public's dissatisfaction with its lack of opportunity for
  participation.
• Public involvement from an early stage in the permitting  pro-
  cess helps government staff  identify  areas of concern to the
  community  before a permit has been drafted. When identified
  in  advance, such concerns often can  be addressed in the draft
  permit. Less time and energy will be spent developing permits
  that later provoke public opposition by failing to address com-
  munity concerns.

  Early public involvement in RCRA permitting is also important
because of the enactment of the 1984 RCRA amendments. These
amendments are a direct consequence of the increased public con-
cern over the issues  of hazardous waste management. In return,
because of certain provisions in  the 1984 RCRA amendments and
386   PUBLIC COMMUNICATIONS

-------
the debate over enactment of those amendments, public interest
in the RCRA permitting program has broadened and public in-
volvement in the program has become more complex. The follow-
ing issues, in particular, are likely to be the focus of special public
interest as the permitting program develops.

Groundwater Protection
  There has been widespread failure to comply with  RCRA's
groundwater protection requirements. The  removal of  wastes
from Superfund sites to RCRA facilities has intensified concern
over the possibility of groundwater contamination and its effects
on human health.

Protective Standards and Enforcement
for Operating Units
   Public interest  is already  strong on the question of whether
RCRA's design standards and operating specifications (e.g., land-
fill liner requirements) are strict enough and are being met.

Exposure Assessments
   RCRA section 3019 requires each final permit application for a
landfill or surface impoundment to be accompanied by informa-
tion on the potential for the public to be exposed to hazardous
constituents released from the facility. Citizens also can submit
exposure information. There often will be considerable local in-
terest in this information.

Corrective Action
   Significant public interest  can be expected in all facets of cor-
rective action requirements.  Have releases occurred from a facil-
ity? How will  releases be cleaned up? Will corrective action be
sufficient to prevent future releases? What kinds of investigations
will be conducted to determine the need for corrective action?

Permit Process Itself
   The length of time involved in issuing  a permit as well as the
adequacy of public involvement opportunities are examples of
issues related to the permit process (as opposed to the contents of
permits) that can be of public concern.

Transportation of Hazardous Wastes
   Many times  the public's strongest concern centers on the
transportation of hazardous wastes  to or from a facility. The
possibility of traffic accidents and the proximity of transportation
routes to homes and schools heighten the public's  concern over
releases during transportation. Often, though, the sheer volume
of traffic and  the associated noise and congestion are of even
greater concern.

Evacuation Plans
   The very existence of evacuation plans implies that accidents
can and will occur. Moreover, the public has shown interest in the
adequacy of evacuation  plans. Who will be evacuated? How will
they be notified? Who pays? Fire and explosion emergency plans
also acknowledge the potential for threats to the public's  safety;
despite their value, their existence may undercut assurances that
such occurrences are unlikely.

Omnibus Provision
  RCRA section 3005(c) states that "each permit .  . . shall con-
tain such terms and conditions as the Administrator (or the State)
determines necessary to protect human health and the  environ-
ment." The public may be inclined to read great flexibility into
this provision.
Consequences of Permit Denial
  Whether an operating permit is approved or denied, the conse-
quences will be of interest to the public.  The denial of a permit
will cause special concern because  of the potential impacts on
employment, property values and the local tax base. Moreover,
the public may not at first appreciate the  environmental implica-
tions of permit denial. Denying a permit for incineration, for ex-
ample, might mean that the  applicant has to continue to landfill
wastes.

U.S. EPA'S APPROACH TO PUBLIC
INVOLVEMENT IN THE PERMITTING PROCESS
Guidelines
  Early public involvement in the permitting process for a waste
management facility means doing more than simply fulfilling the
formal public  participation requirements  in federal regulations.
How much more, however, is something  that depends upon the
particular circumstances  of  each case. Thus, the U.S. EPA's
policy for public involvement in RCRA permitting does not in-
clude the addition of long lists of requirements. In fact, there are
no new requirements.  Instead, the U.S. EPA offers the following
general guidelines:
• Public involvement efforts should be tailored to the  distinc-
  tive issues and individual features of the facility and the sur-
  rounding community.
• The  applicant and other responsible  government  agencies
  should have a role in public involvement efforts, especially to
  help clarify or resolve issues that may be related to the facility
  and its  operations but cannot be addressed appropriately by
  the RCRA permit process.
• Small-scale,  low-profile, informal communications  techniques
  are  preferred. Public meetings may be held in informal set-
  tings, before small  audiences and without elaborate presenta-
  tions (e.g., in living rooms). They need not be conducted by
  high-level staff.
• In general, public involvement actions  should  extend  beyond
  providing information to the public; they should actively reach
  out to the public, encourage participation and  provide an op-
  portunity for public input on permit decisions made by EPA
  or the State.

Targeting
  Following the above guidelines,  public involvement efforts
should be targeted to individual circumstances; the same program
of activities will not be necessary, and would not be effective, in
every case. Public involvement efforts also should be targeted in a
second sense:  the  limited resources available to a U.S. EPA
regional office or State agency for public  involvement should be
concentrated on those facilities where the need is greatest.
  Targeting in this second sense means that certain facilities will
have to be selected  for special attention. The selection of facilities
for early and expanded public involvement should be based not
only on environmental significance (i.e., the potential  to pose
significant risks to  human health and the  environment), but also
on social significance (that is, the extent of public interest already
generated and the  potential  to arouse significant interest in the
future). While the development of widespread interest as the per-
mitting process unfolds cannot be predicted with certainty, the
following are good indicators of the potential to  generate public
interest:

• The owner or operator of the facility lacks credibility with, or
  the trust of, the  public or  local officials
• The public perceives that the facility poses major health risks
                                                                                           PUBLIC COMMUNICATIONS    387

-------
• Major hazardous  substance  releases  or  accidents have been
  publicized recently in the area
• The type of technology proposed in the permit (e.g., incinera-
  tion, underground injection, landfill, etc.) has a negative repu-
  tation in that particular area
• Facility non-compliance or violations have been highly publi-
  cized or will be made known by the permit process (in general,
  the more serious  the continuing non-compliance,  the  more
  public interest likely to be generated)
• The facility has been or likely will become an election issue
• The permit could allow the transportation of Superfund haz-
  ardous wastes to or  from the facility
Permit writers in the U.S. EPA regional offices and in State agen-
cies have found that  often there is little correlation between a
facility's environmental significance and its  social significance. It
is not uncommon  for  an environmentally significant  facility  lo
meet  an apathetic reaction from the local  community, while a
smaller facility creates  intense public concern.

The Field Assessment and
Public Involvement Work Plan
  Two critical elements in  a public involvement effort  for a
targetted facility are the "field assessment"  and "public involve-
ment  work plan." Both help to ensure  that public  involvement
efforts are well-planned and meet the distinctive needs of a partic-
ular community and a particular facility.
  A field assessment consists of interviews, both  in person and
over the telephone, with local citizens and officials to  obtain in-
formation on the following points:
• Major community concerns regarding a facility
• The citizens, community leaders and officials in the area who
  are especially interested in the facility and should be kept ap-
  prised of developments
• The best means  to provide information to the public and,  in
  return, to obtain public comment and input

  To this end, the community interviews conducted  for the field
assessment address the following topics:
• The local history  surrounding the  facility (e.g.,  the circum-
  stances under which it was built); any  past or present opposi-
  tion to  the  facility;  previous relations between citizens, gov-
  ernment  agencies  and the  facility's  owner or operator; and
  news stories about the facility. (For instance, it might be use-
  ful  to know that the local newspaper has opposed the facility's
  operation for ten years and has just finished publishing a series
  on  the facility's non-compliance with  interim status require-
  ments.)
• The negative as  well as the positive economic impacts of the
  facility's operation  on various segments of the  community.
  This information  is useful for  anticipating and  interpreting
  people's concerns and positions.
• The current level of information and  understanding that the
  public (including local officials)  has  about the  facility's op-
  eration. This information is needed to assess what additional
  information will be  necessary to  enable the public to partici-
  pate intelligently.
• Other significant political  or social events or  circumstances
  that might  affect  the community's attitude or  involvement
  with the facility's permitting.  It would  be imperative to know,
  for example, that the recently elected mayor campaigned on a
  promise to close the facility.
  After two or three days of interviewing to gather such informa-
tion, a public involvement work plan can be prepared. This plan
will include the following elements:
• A capsule facility description
• An identification of major community concerns and leaders
• An outline  of the  minimum activities  the  U.S. EPA or  the
  State will conduct to ensure both that the local  public is well-
  informed about permitting decisions and also has meaningful
  input into those decisions
• An indication of the timing of these activities
  The public involvement work  plan provides a strategy for en-
abling the public to have meaningful input  to the decision-making
process for a  permit. It also serves to coordinate the roles and
responsibilities of the various participants in  the public involve-
ment effort: The U.S. EPA, the State, local officials, the owner
or operator and the general public. And it  reassures local citizens
that there indeed has  been some forethought given to providing
opportunities for meaningful public involvement.

Public Involvement Activities
  Public involvement activities will vary  by facility and by the
stage in the permitting process at which public involvement ef-
forts are initiated. The following types of activities, however, will
be conducted in each  instance:
• Outreach activities, such as informal informational briefings
  and meetings, including public notice of the draft permit
• Dialogue and assimilation activities,  such  as work sessions,
  public meetings and public hearings (if appropriate)
• Response activities,  including informal responses to questions,
  concerns and requests from the public during the permit pro-
  cess as well as formal, final responses to public comments
The only activities that are required are those specified in RCRA
section 7004 and 40 CFR 124 Subpart A, the  most important  of
which is the public comment period on the draft permit. These re-
quirements hold for every determination on an  RCRA final per-
mit application, including those for facilities that are not targeted
for expanded public involvement.
  Often the most effective public involvement activities are very
different from "textbook" information techniques. For example,
a permit writer in Hawaii  responded to citizen concern over the
permitting  of a waste management  facility by serving as an in-
termediary between  citizens  and the  owner,  encouraging the
owner to modify his operating procedures in a way that could not
have been specified in the permit but that  successfully addressed
public concerns.

Roles and Responsibilities
  Public involvement  in the RCRA  permitting process is a team
effort, with contributions from the U.S. EPA and the State. And
within each permitting agency, there will  be roles and  respon-
sibilities for staff from a number of offices. For example,  each
EPA regional office, and each State authorized to issue final per-
mits, should already have designated someone on its staff as the
RCRA permitting program public involvement coordinator. The
coordinator works with permit writers, enforcement personnel
(both U.S. EPA  and  State), facility owner/operators and other
appropriate individuals or groups to implement public  involve-
ment activities. Permit writers may take the lead in discussions of
specific permit conditions  on facility operations;  they  may  be
responsible for providing information to answer public inquiries.
  In some  cases, the owner or operator may contribute to public
involvement activities  conducted by the U.S. EPA or the State in
addition to his own independent outreach  or public information
efforts.  In fact, owners and operators are encouraged to take
steps to inform and involve the local public, although in so doing,
the distinction between the owner or operator's interests and the
government's regulatory mission should be made absolutely clear
to the public.
388
       PUBLIC COMMUNICATIONS

-------
SUPPORTING INITIATIVES
  The U.S. EPA's Office of Solid Waste is pursuing a number of
supporting initiatives  to advance the policy for public involve-
ment outlined above. The goal is for local public involvement to
become a regular feature  of the U.S. EPA and State decision-
making on each RCRA permit, with the extent of the public in-
volvement effort for each facility and the scope of the activities
conducted in the locality scaled to the need.
  First, we are now providing a two-day training course in public
involvement  techniques for RCRA  staff in each  U.S. EPA
regional office. State staff also are invited to these sessions,  and
we expect to offer the training in authorized States in the future.
The training  covers the rationale  for early and expanded public
involvement in RCRA permitting, the targeting of facilities for
special attention, procedures for conducting field assessments and
developing public involvement work plans, as well as techniques
for conducting effective public meetings and handling relations
with the news media.  There is also instruction in  basic skills for
minimizing or resolving disputes and conflict between govern-
ment and  local citizens.
  Second, we are preparing a variety of educational and informa-
tional materials for the lay public on the RCRA permitting pro-
gram. Already in draft is  a "Citizens' Guide to RCRA Permit-
ting" that explains the steps in deciding whether to grant a final
permit to a hazardous waste management facility, the conditions
that may be included in a permit and the role of the local public in
decision-making  on the permit. We plan to prepare brief  fact
sheets on specific topics of widespread general interest (for exam-
ple, corrective action requirements and exposure assessments).
Fact sheets also  will be prepared  on individual facilities, when
needed, to explain to the public the technical features of the facil-
ity, the operation and design standards to which it is subject  and
progress in reviewing the facility's permit application.
  Finally, we are  continuing  to  analyze  the  best approach to
public  involvement  and  community  relations  in  special  cir-
cumstances (for example, in siting new hazardous  waste manage-
ment facilities or in conducting exposure  assessments).  As the
analyses reach completion,  we will prepare guidance materials to
assist the regional and State staff who will be on the "front line"
in public involvement efforts. As noted, the U.S. EPA already
has issued draft guidance on public involvement in  RCRA permit-
ting.
CONCLUSIONS
  The discussion in this paper should serve to dispel any impres-
sion that the U.S.  EPA's initiatives for public involvement in
RCRA permitting will just "stir the pot," creating public opposi-
tion where there would otherwise have been none and ultimately
forcing  the closing  of facilities that could treat and dispose of
hazardous wastes in an environmentally sound manner. That is
not, of course, the desired consequence of these initiatives.  In-
stead, these initiatives are aimed at preventing needless public  op-
position, the opposition rooted in public distrust of government
motives, rather than in the technical flaws of a facility or permit,
and are precipitated by neglecting the local public in government
decision-making. They are, moreover, aimed at ensuring that
RCRA permits  take local concerns fully into account and protect
the local public's health, welfare and environment. Early and  ex-
panded public  involvement in  RCRA permitting  can  thereby
facilitate the permitting process and lead to better environmental
decision-making.

BIBLIOGRAPHY
 1.  A National Permits Strategy for Implementation of the  Resource
    Conservation and Recovery Act, U.S. EPA Office of Solid Waste,
    Aug. 1984, as revised on Sept. 9, 1985.
 2.  Bleiker, Annemarie and Hans,  Citizens Participation Handbook
   for Public Officials  and Other Professionals Serving the Public,
    1981.
 3.  Community Relations in Superfund: A Handbook, prepared by 1CF
    Incorporated for the U.S. EPA, Sept. 1983.
 4.  Environmental Protection Agency Regulations on Procedures  for
    Decisionmaking, 40 CFR Part 124, Subpart A.
 5.  EPA and the Public: A Handbook on Public Participation Con-
    cepts and Skills, Barry Lawson Associates,Inc., 1981.
 6.  Guidance on Public Involvement in the RCRA Permitting Pro-
    gram, U.S. EPA Office of Solid Waste, Permits Branch, Dec. 1985.
 7.  Implementation of the Resource Conservation and Recovery Act,
    U.S. EPA, No. EPA/530-SW-84-007, 1984.
 8.  Jordan, Barry H., How  to Write a Public Notice: A Collection of
    Examples, U.S. EPA Water Program Operations, Dec. 1979.
 9.  Resource Conservation and Recovery Act,  Section 7004(b)(l),  PL
    96-482.
10.  "Responsiveness Summary and Preamble on Public  Participation
    Policy," Federal Register, 46, No. 12, 1981.
                                                                                             PUBLIC COMMUNICATIONS    389

-------
                      Cleanup  in the Sunshine:  Florida  DOT's
              Public Information  Program  at  the Fairbanks  Site

                                               Robert  C. Classen
                                              O.H. Materials Co.
                                                Atlanta, Georgia
                                                Charles C. Aller
                             Florida Department of Environmental Regulation
                                              Tallahassee,  Florida
ABSTRACT
  A routine safety inspection of the  Florida Department of
Transportation (DOT) Bureau of Materials and Research Labora-
tory was conducted in November 1982. This inspection indicated
that chemical waste generated at  the laboratory was being dis-
posed at a borrow pit located on DOT property in the rural com-
munity of Fairbanks, Florida, 7 miles northeast of Gainesville.
  In January 1983, a number of private water supply wells in
Fairbanks were tested for volatile organic compounds (VOCs).
The results indicated the presence of VOCs in some of these wells.
  In response to public health concerns, the well sampling pro-
gram was intensified, and a comprehensive site investigation was
initiated. The purpose of the investigation was to determine the
source of VOC contamination  affecting private wells to estimate
the extent of VOCs in the groundwater, to develop immediate re-
medial procedures  and to recommend a water supply alternative
for the Fairbanks community.
  The results of the site investigation indicated that the  DOT
borrow pit was the source of VOC contamination affecting pri-
vate water supply wells. Laboratory waste materials which form-
erly were disposed of at the sand pit had seeped into a multi-
aquifer  system.  This contamination had entered  the shallow
aquifer in substantial concentrations within a limited area near
the sand pit and had slowly leaked into the uppermost secondary
artesian aquifer. VOCs  that had entered the upper  limestone
aquifer spread to private wells through the natural flow system
and in response to pumping.
  Remedial actions were implemented to eliminate the source of
VOCs and to cleanse the shallow aquifer. These actions included
the removal of drums of buried waste from within the borrow
pit and the installation of a recovery well system designed to pre-
vent further movement  of contaminated groundwater in the
shallow aquifer, as well as to treat the groundwater by removing
VOCs.
  After the initial sampling, when it was determined that private
water supply wells were contaminated, bottled water was distrib-
uted to the community.  At that  time,  an understandably high
level of anxiety existed in the  community. For this reason, the
DOT decided to initiate  a public  information  program consist-
ing of a newsletter, weekly public information  meetings and lia-
ison with the citizens' group.  By the  end of  the crisis, it was
apparent that the program was extremely successful.

INTRODUCTION
  During a routine safety inspection of the Florida Department
of Transportation (DOT) Bureau of Materials and Research Lab-
oratory, the Second District Industrial Safety Manager noted that
chemical waste generated by the DOT laboratory was  being dis-
posed at Fairbanks Sand Pit (FSP), located on DOT property
approximately 7 miles northeast of Gainesville in Alachua County,
Florida (Fig. 1).
  Private water supply wells in the vicinity of the borrow pit were
sampled, and the results of these tests indicated that 9 of the 18
wells sampled were contaminated with  the  same halogenated
chemical solvents disposed at FSP or with the degradation pro-
ducts of these solvents. The DOT immediately began distributing
bottled water to the residents of Fairbanks, the rural community
which surrounds FSP.
  On Feb. 4, 1982, Dames & Moore was retained by the DOT to
work with the agency toward resolution of the Fairbanks prob-
lem. It  was agreed  that a detailed study of the Fairbanks prob-
lem was necessary and that this work should incorporate a com-
prehensive hydrogeologic investigation of the site.
                                              FAIRBANKS
              MILES
                         Figure 1
                       Site Location
Problem Definition
  The initial objective of the project was to provide a compre-
hensive analysis of the nature, extent and transport mechanisms
causing the observed groundwater contamination. Specifically,
the work included:
390    PUBLIC COMMUNICATIONS

-------
• Implementation of immediate procedures to clean up and re-
  move chemical waste contamination within the borrow pit
• Characterization of historical waste disposal at the borrow pit
  and adjacent areas
• Evaluation of adjacent off-site areas  relative to potential con-
  tamination sources
• An estimate of the  horizontal and vertical  extent of contam-
  ination in the aquifer(s)
• Development of conceptual control/abatement alternatives for
  the zone of groundwater contamination

  As the investigation progressed, the information developed led
to expansion of the scope of work outlined in the plan of study to
include: additional investigation of the extent and rate of travel of
contamination in confined aquifers; logging  of private water
supply wells;  toxicological analysis; design, installation and per-
mitting of a recovery system;  and an analysis of alternatives for
providing potable water to the residents of the community.

PROGRESS OF THE INVESTIGATION
  A major objective of the program was to  establish the areal
and vertical extent of the contamination affecting the aquifer sys-
tem in Fairbanks.

Sampling
  Sampling of private water supply wells had indicated that vola-
tile organic compounds (VOCs) were  contaminating one or more
 aquifers in the Fairbanks area.  The first step was to determine if
     I OBSERVATION WELLS

     JAREA OF CONTAMINATION
                                              N.E. S3 TEBR.
the Fairbanks Sand Pit was, in fact, a source of this contamina-
tion. In order to determine the types of waste present in the pit, as
well as historical disposal practices, DOT  laboratory personnel
(the generators) and maintenance personnel (who had placed the
chemicals in the pit) were interviewed.
  Next, four clusters  of two monitoring wells each were placed
on  the pit's  perimeter (Fig.  2), sampled and analyzed for the
chemicals on the U.S. EPA's Priority Pollutants  List. The re-
sults indicated that the chemicals which were buried in the pit or
their  degradation products  were  contaminating  the  surficial
aquifer and that among these chemicals were the same chemicals
detected in the private water supply wells.
  The next task was to construct eight clusters  of two wells
around the pit at a distance of 500 ft. Analysis of samples and
water levels from these wells showed that the direction of flow in
the surficial aquifer was to  the north, towards  Hatchet Creek,
and that  VOC contaminated groundwater was leaving the pit in
this direction.
  The next monitoring wells were placed downgradient of other
borrow pits in the vicinity of FSP to determine if they were con-
tributing to the contamination problem. They were not.
  Other  wells were placed into the semi-confining unit at the
base  of  the surficial aquifer  and north  of  Hatchet Creek.
Analysis  of samples from these wells indicated that the contam-
ination was penetrating the semi-confining unit but was not cross-
ing Hatchet Creek.
  With this information, work shifted from the shallow to the
deeper aquifers. First, borings were made to determine the posi-
tion and  continuity of water-bearing units in the Fairbanks area.
When these units had  been mapped, deep monitoring wells were
placed so that each of the three aquifers identified could  be sam-
pled and  aquifer characteristics could be measured (Fig. 3).
  While the field program was underway, additional information
on private water supply wells was being collected through a house
to house well inventory and well logging.
                          Figure 2
  Monitoring Well Location and Approximate Lateral Extend of VOC
              Contamination in the Shallow Aquifer
                                                                     • TEST BORING LOCATION ONLY

                                                                     ® ARTESIAN AQUIFER MONITOR WELL
                                                                      AND TEST BORING LOCATIONS
                           Figure 3
          Location of Artesian Wells and Deep Test Borings
                                                                                            PUBLIC COMMUNICATIONS    391

-------
Pit Closure
  Another of the project's objectives was to properly close, con-
current with the field investigation, the Fairbanks Sand Pit, elim-
inating it  as a source  of contamination. Closure consisted of a
magnetometer survey  to ensure that all buried drums had  been
located, excavation and proper disposal of the drums and installa-
tion of a  recovery system  to control the movement of contam-
inated water in the immediate vicinity of the pit.
   Residents in Fairbanks whose water supply wells were contam-
inated were using bottled drinking water. A priority objective of
the project was to restore potable tap water to the community as
soon as possible. Thus, while the field program and pit cleanup
were underway, an analysis of alternatives  to provide potable
water was conducted. Completion of this analysis was timed so
that important information from the  geohydrological program
could be considered in deciding on the recommended alternative.
   Although the residents were  drinking bottled water,  another
issue of concern was whether or not they would be able to safely
eat (1) fruit and vegetables in  which  contaminated water was
used in the process of home  canning  or freezing and (2) meat
from livestock which had consumed contaminated water before
slaughter. To answer these and other health questions, a toxicol-
ogist was  retained and samples of food were analyzed. The food
samples were found to be free from contamination.

PUBLIC INVOLVEMENT
   The need for  public involvement was, from the outset, iden-
tified as vital  to the successful resolution  problem. When De-
partment  of Transportation (DOT) and Dames & Moore project
personnel arrived on-site  in February  1983,  it was evident that
the level of community concern was extremely high.  Following
an on-site review, a press conference was held to announce plans
to initiate site investigation and cleanup activities. This press con-
ference was the beginning point for a project which, from the
start, focused local, state and national attention on public agency
response to a serious groundwater contamination incident.
   Development of the project plan of study proceeded with close
coordination between all involved governmental agencies:
•  Alachua County, Board of County Commissioners
•  Alachua County Health Department
•  Florida Department of Environmental Regulation
•  Florida Department of Health and Rehabilitative Services
•  City of Gainesville, Utilities Department
•  St. Johns River Water Management Dr'.rict
   The plan of study was publicly presjnted at an evening meet-
ing in the County Commission board room in mid-March 1983.
Public comment at that time made it clear that, although the
Department of Transportation had done an excellent job of con-
sulting with other agencies and local elected  officials, members of
the Fairbanks  community  felt excluded from the decision-mak-
ing processes.

Community Group Formed
   During this time, in response  to the crisis,  people living in the
area joined together in a voluntary group known as the Fairbanks
Community Association  (FCA).  They  held meetings,  elected
officers, began to develop  an agenda of concerns relating to the
contamination and  contacted a local  legislator, Representative
Sid Martin (D-Hawthorne), who was active  in environmental and
conservation issues. In late March, Representative Martin chaired
a volatile and well attended, legislative committee meeting which
provided a mechanism for  DOT, Dames &  Moore and local resi-
dents to begin  the process of working together. From this point,
an extensive community involvement program began which in-
cluded the following key elements.
Community Involvement
• Daily contact between the DOT project  manager, the FCA
  President, the Alachua County Administrator (acting for the
  Board of County Commissioners) and the County Health De-
  partment.
• At a minimum, weekly contact  with legislative staff and  the
  Department  of Environmental  Regulation.  Other  involved
  agencies were contacted regularly as the situation required.
• A series of evening community meetings (weekly during April
  and May 1983,  then  biweekly for the project duration) was
  conducted at the Fairbanks Baptist Church.  The purpose of
  these meetings was to brief residents on project programs and
  provide an opportunity to  air questions and  respond to con-
  cerns.  Prior to each of these meetings,  face-to-face briefings
  were provided to FCA officers and their concerns were solicited.
• Site visits were arranged for interested residents.
• A newsletter  was published and mailed to  all  community resi-
  dents, not just to FCA members.
• Community members and agency representatives were invited
  to participate in all meetings involving major project decisions,
  including:
  -Study design
  -Monitor well construction
  -Monitor well and test boring location
  -Sampling protocol
  -Borrow pit closure
  -Water supply alternatives analysis
• Complete public and media accessibility of DOT project man-
  agement personnel.
  The cornerstone of the public involvement program  was the
open attitude and effort to maintain communication on the part
of DOT  personnel; from the beginning,  all project operations
were open to public  scrutiny. Of the elements  briefly reviewed
above, the most critical to project success probably involved the
meetings with the residents and the newsletters.

Public Meetings
  At the outset, the FCA meetings were well attended with heated
discussion. Residents were understandably  scared, angry and
frustrated. As the project progressed, DOT, Dames & Moore and
community members began  to work together to  resolve issues of
mutual interest. Examples include:
• Site excavation cleanup and progress
• Bottled water deliveries
• Site security
• Monitor well construction and testing progress
• Private water supply sampling
• Safety of foods grown and irrigated or preserved with contam-
  inated water
• Provision of  independent lexicological expertise  for respond-
  ing to health concerns
• Community well construction survey
• Rumor control
• Property values
• Long-term project planning
• Remedial action water supplies alternatives
REMEDIAL PLANS
  The  last item noted probably is most representative of the
overall success of the public involvement program. A final pro-
ject step required an analysis of possible solutions to the contam-
ination problem, the election of a  preferred remedy and imple-
mentation of an equitable settlement for the residents.
392   PUBLIC COMMUNICATIONS

-------
  This remedial evaluation process, which examined drilling new
wells, home filtration systems, a community water supply and
connection to an extension of the Gainesville Municipal system,
was jointly conducted by DOT and a citizens committee of 12
volunteer  community members.  Issues of  fire protection and
financial settlements also were considered. The result of this work
was a decision to connect residents to the Gainesville water sys-
tem, provide enhanced fire protection (hydrants) and include a
water rate subsidy for a specified period of time.

COOPERATION ATTAINED
  Toward the end of the site cleanup and contamination inves-
tigation activities, the remedial action decision for water system
connection was announced at a well attended public  meeting
which received extensive media coverage. Unlike earlier meetings,
the only hostility toward the DOT was evidenced by media repre-
sentatives not familiar with the project.  Residents and officers of
the FCA voluntarily spoke in support of the different phases of
project activities and on behalf of DOT.
  The final few public meetings on this project clearly demon-
strated the accomplishment of this multifaceted public involve-
ment and educational program. A community, completely frus-
trated and angry at the state over the contamination of the water
supply, eventually became totally involved  in all phases of the
project. The initial hostility gradually led to an active participa-
tion which  resulted in a sense of satisfaction on the part of resi-
dents in having played an important part in the solution.
                                                                                           PUBLIC COMMUNICATIONS
                                                                                                                          393

-------
            Public Involvement in  the  RCRA Permitting Process-
                                         A  Facility Perspective
                                                    Gordon Kenna
                                        Chemical  Waste Management,  Inc.
                                                  Emelle, Alabama
INTRODUCTION
  Few issues in America today generate more public concern than
that of hazardous waste handling and disposal. By statute, reg-
ulation, policy and  inclination, the U.S. EPA and the states en-
courage public involvement in permit decisions. While the prin-
ciple is well established, there is little agreement on how best to
accomplish the objective.  There  also is considerable debate on
how best to evaluate, respond and act on issues raised by the pub-
lic. As a result, regulatory agencies sometimes find themselves in
the uncomfortable position of being sued on the same issue by in-
dustry for over-regulating and  by  citizen  and  public interest
groups for under-regulating. Moreover, while the agencies have
the ultimate responsibility for making and  defending their de-
cisions, they must consider issues raised by citizens even though
those issues may not be germane to the subject or within the scope
of the regulatory agency.
  With this  responsibility is also the reality that the regulatory
agency does not have the  staff, resources or ability to know the
real community consensus and composition  in depth. As  is well
known, those  who  attend public meetings and  write comments
are nearly always a very small minority of the population. The
majority of citizens who are aware of the issues but do not make
their opinions known are,  more often than not, supportive of the
applicant or the agency's professional judgment. Imperfect as the
process is, it is a fact of life in major regulatory actions.

CITIZEN PARTICIPATION ENHANCEMENT
  The primary variable available to improve the quality of citi-
zen participation is  the owner/operator working with citizens to
promote a full  understanding of the issues. Ideally, this is done at
three  levels  of involvement—the company  in the community,
the company in concert with  similar groups acting as an industry
or trade group and  the company coordinating its approach with
the regulatory agencies.
  Obviously, the industry can and should do more to generally
educate the public about siting and permit issues; but even if this
were to occur, the effects  would be  very limited on specific de-
cisions. This approach is most widely and effectively used in leg-
islative and rule-making activities.
  A coordinated  approach between  government and industry is
certainly a good idea but  is fraught with problems. Aside from
the regulatory resource limitations already mentioned, a cooper-
ative effort often leads to the allegation that the agency and appli-
cant are too close or "in bed with one another." It is, of course,
the nature of a highly regulated industry to have a  close rela-
tionship and intimate knowledge of operations and issues. While
distance must be maintained, the  agencies should do a better job
of publicizing their responsibility and the need for close commun-
ication and coordination.
ROLE OF REGULATORY AGENCY
  At a minimum, the regulatory agencies should plan and dis-
cuss the public involvement program with the applicant well in
advance of implementation. An exchange of information should
include citizen comments received to date, an exchange of any
community relations or public involvement plans, suggestions for
action by either party and  joint informal meetings,  workshops
and information booths to identify issues in advance of a formal
public hearing. The burden  of this activity should be on the own-
er/operator/applicant,  but  the regulatory  community can im-
prove the effectiveness and  "reach" of the program significantly
through their participation. In so doing, the agency also will en-
hance the quality of public comment and its own image as a
public agency.
  In reality,  however, there are severe limits to how much the
government as a regulatory  agency really can participate. Except
in the case of the largest and most visible sites or permit issues,
the agency will be able to provide little more than a cursory over-
view of the community dynamic. It  is the local  site that must
take the lead  for generating understanding that will give way to
acceptance in the  community. Guidance, participation and en--
dorsement from government regulators of this local effort simply
lend credibility and acceptance to the  process. It is most assured-
ly a process and not an overnight sensation.

PUBLIC INVOLVEMENT PROGRAM
  There are  many  considerations in  planning a public involve-
ment program. Among  the most important is timing, and there
are several key elements in that. It is not possible to  start too
early. There will be public involvement sooner or later; the earlier
the public has an opportunity to become familiar with the process
and the issues, the better. One common citizens' complaint is that
there  was no opportunity to be involved until after the major
issues had been cast in a draft permit.
  Public interest at this early stage  may not be evident, so it
usually is necessary to identify those who are potentially inter-
ested  and solicit their participation.  The ultimate objective in
this effort is an exchange of information and ideas—a two-way
dialogue. To  set the stage for this two-way communication, the
community members need  to  feel that it is  early enough in the
process to make a difference, that they are being dealt with open-
ly and honestly and that they have access to information  and to
key regulatory and company personnel.

Communications
  Some of this feeling of openness and availability can be com-
municated by the press; they, too, certainly need to have a regu-
lar and  reliable point of contact and access to key people. It is
very difficult for a complex and controversial issue to be consis-
394    PUBLIC COMMUNICATIONS

-------
tently well done by the press, but the press can convey the mess-
age that the company is open and invites honest inquiry. Discuss-
ing the  public involvement  strategy with the press  or allowing
coverage of a tour or meeting with citizens can be a good way to
communicate that feeling of openness. Failing to generate  that
feeling of being open  and responsive  seems to make virtually
everyone suspicious and undermines credibility.

Newsletter
  Another means of communicating that open feeling and giving
people information about the operation is through a newsletter
or other general printed information. Although this is a one-way
effort, it is a good precursor for a real exchange. The newslet-
ter is also a good way for the operator to focus attention on issues
of his choice. A newsletter also can  be  used to feature informa-
tion about tours,  meetings with environmental groups, question
and answer columns and other two-way exchanges.  The regular
appearance of a newsletter is also proof  that the facility has made
an effort to  communicate with and  involve citizens. The longer
the newsletter comes out, the more knowledgeable people become
about the issues and people at the facility. Moreover, the reader-
ship  seems to increase  with each issue as the product and place
become more familiar.

Documents Repository
   Another important means of making information available is
a  local  repository for documents, applications, permits and so
forth. Traditionally, these documents have been kept at a court-
house or other government  office (i.e., the local health depart-
ment). Sometimes this "official" repository is not  available to
the public during evening or weekend hours or is not seen  as a
neutral location. A more appropriate location would be a public
library or local  school. The library has the added advantage of
having other research resources available for serious inquiry. In
addition to the required documents  for review and inspection,
other material should be placed on reserve. This reserve list might
include: newsletters and other company  publications; indepen-
dent material such as state  and  federal regulations; and books
and pamphlets on chemistry, industrial hygiene and safety, re-
search reports and so forth. A summary of the important issues
and the process for resolution of these issues also should be avail-
able. A good summary will be sufficient for many people and
can offer a clear  perspective and context for understanding the
issues.
   There are many other means of disseminating information
that are, although basically one-way vehicles, good for educat-
ing citizens about certain activities and people and encouraging
participation in meetings and other "feedback" opportunities.

COMMUNITY RELATIONS
   The  most important opportunity  for generating  community
understanding and confidence  is simply a presence  in the com-
munity. The extent to which facility  management and employees
are integrated into the social fabric of the community is of critical
importance.  First, this means  that  personnel from the facility
should live in and have  roots in the area and be visible and active
in the community. Employee community involvement programs
and a recognition and  incentive system are  good tools for en-
couraging that presence.
  Corporate involvement in community  projects is also an impor-
tant means of demonstrating a commitment to the community.
Without being able to demonstrate that  presence in the commun-
ity and  commitment to citizenship,  other efforts are seriously
hampered. It is important to understand, however, that the objec-
tive is citizen understanding and acceptance  and, in the case of
controversial facilities, much more than community involvement
and a presence is required. The emphasis must be education and
an exchange of ideas.

Tours
  One of the best ways to accomplish this community presence
is to give tours of the plant or facility with ample opportunity
for questions about what is on the tour or any other issues that
might arise. These tours should be personalized as much as pos-
sible to the particular tour groups. Involvement by senior man-
agement in these tours and question and answer sessions is very
important to maximize credibility and commitment to public in-
volvement. The importance of first-hand observations cannot be
over-emphasized; often a facility will "sell itself" on such a tour,
and the strong positive image from such an experience will make
a lasting impression.

Speeches
  Talks and speeches to interested audiences are another way to
reach out to a particular group  and encourage understanding
and exchange of views.  However, it is difficult to have enough
time to cover the major issues and respond to questions as well.
The question and answer session following a talk is probably the
most important part of this kind of effort.

Meetings
  The activities that can be expected to go the furthest in clarify-
ing issues and settling disputes are the small informal meetings
and one-on-one discussions. These meetings may  take  several
forms but have several common characteristics: listening for key
points,  responding to questions and concerns and generally in-
teracting as an equal.
   While the exchange that occurs in these meetings is critical, it is
also important that there be  appropriate follow-up and eval-
uation. These meetings should be targeted for those people who
clearly want to know more and are likely to vocalize their feel-
ings.
   Groups likely to be targeted are environmental organizations,
elected officials, local community organizations, residents, ad hoc
groups  and public interest groups. While it is important to an-
ticipate issues and prepare for them with an agenda  and objec-
tive, it is equally important to make that agenda flexible if the
group demands it.
   It is also  useful to have such  meetings for anyone willing to
come rather than just targeting a particular group. Ideally, these
meetings should be fairly regular, held in a neutral place and have
a specific objective relative to the permit issue. These meetings
are designed to explain the issue and draw out opinion  on the sub-
ject. Once positions are  disclosed, follow-up work is appropriate
to influence or support the opinion.

Information Booth
   A variation of this is an "information booth" approach that
encourages local residents to ask questions or offer  their  opin-
ions on a permit issue. Regulators  have  much more credibility
in this role than operators, but it is an area where the two interests
can work effectively together. This kind of open-ended  small
meeting can be unpredictable but  may  produce unexpectedly
good results.
   An exchange of information is the most powerful and useful
communication that can be used to convince and persuade. Tar-
geting the right audience for this dialog is very important since
the process  is very labor and  time intensive. An even distribu-
tion of public  sentiment might  look like a  "bell curve"  with
communication techniques overlain:
                                                                                           PUBLIC COMMUNICATIONS    395

-------
                 One way communication:
 Community position, newsletter, press, handouts, perception
                        Neutral
 Speeches, personal contact, information booth, meetings, two-
                         way talks.
  It is the groups predisposed to be against a permit that a small
meeting may be best able to influence. If these groups are iden-
tified and are willing to meet  and discuss issues, some movement
is possible. While these dynamic two-way exchanges are at the
core of the objectives in a public involvement program,  it is the
static  or one-way  information  program  that educates, builds
support and makes the opportunity for an exchange of informa-
tion more meaningful and more difficult to resist.
   Somewhat surprising to those concerned with this activity may
 be citizens who resist such small meetings and will not verbalize
 issues of importance to them. There may be several reasons for
 this including:

 •  Uncertainty about their grasp of technical issues
 •  Unwillingness to disclose their position because of a preference
   to "surprise" the applicant
 •  Unwillingness to believe anything  that regulator or applicant
   says
 •  Unawareness of opportunity
 •  Personal embarrassment and other reasons
   At any rate, many of these reasons become unimportant at a
 public hearing or in front of a television camera. It seems that
 nearly everyone wants to make his  feelings known at that point. It
 is important to become aware of the feelings early so they can be
 identified, responded to and dealt with. Otherwise they represent
 "new" issues  and come as a surprise. If there are  no questions
 from the public, do not assume that  there are no problems.

 GENERAL TO THE SPECIFIC
   As with most problems, work on  the issue should move from
 the general (where some agreement and common ground can be
 identified)  to  the specific (where  differences are most evident).
 Even with  early identification of  issues,  good  listening and
 analytical skills and a spirit of compromises, many problems may
 remain. However, if any semblance of a working relationship has
 evolved  between the two parties, there  will  be room for some
 respect for the other's position and possibly an acknowledgement
 that there is some legitimacy to the position.  If true,  then the
 minimum outcome is  the recognition of having acted  in good
 faith—an important achievement in  and of itself.

 CONCLUSIONS
   In chronological summary, the basic objectives of a public  in-
 volvement program are to:
                                                          •  Inform
                                                          •  Educate
                                                          •  Target
                                                          •  Listen
                                                          •  Respond
                                                          •  Follow  up
                                                            This movement and trend toward negotiation and mediation of
                                                          environmental issues directly between citizen groups and corpora-
                                                          tions is clearly a positive development. Certainly it is more effi-
                                                          cient than litigation and through early identification of the issues
                                                          can go a long way to help prevent problems.
                                                            In  the RCRA permitting process, however, it is especially im-
                                                          portant to begin this process of exchange and negotiation early
                                                          enough to accomplish the major objectives before the public hear-
                                                          ing process has  run its course. The process needs to be open-
                                                          ended to the extent that the agenda may include new issues raised
                                                          during the process and  may continue indefinitely past the hearing
                                                          dates, but some priority for dealing with major differences should
                                                          be established.  Initially,  those priorities should  be outlined,
                                                          ranked  and agreed upon by both parties.  Any area of agreement
                                                          and  common ground  related to this issue should be used as a
                                                          preface and introduction to begin talks.
                                                            If agreement fails to emerge  from this process after a consis-
                                                          tent,  determined  effort,  the participants still are rewarded for
                                                          having tried. Some of these benefits are:

                                                          •  Respect from  all sides, especially from press and public, for
                                                            initiating  a good faith effort
                                                          •  Improved public understanding
                                                          •  A narrower range of critical comments
                                                          •  A lack of surprises in the hearing process

                                                            There is no substitute for an early start  to educate, inform and
                                                          involve the public in RCRA permit decisions. There is no magic
                                                          recipe for success either. Each approach should be specific to the
                                                          community  involved and to the nature of the permit under ap-
                                                          plication.  The goals of  a program generally are consistent and in-
                                                          clude such things as:
                                                          •  Being open to inquiry
                                                          •  Being responsive and flexible
                                                          •  Making a real effort to inform and educate the public
                                                          •  Anticipating issues and being prepared
                                                          •  Being cooperative with the community
                                                          •  Being innovative and creative with new techniques
                                                          •  Being persistent in your efforts
                                                            While these are elements in a  community based program, these
                                                          elements also are applicable to working with the press.
                                                            A  closing thought—not an afterthought.  There is a natural
                                                          tendency to concentrate on those who are contentious since that is
                                                          where the conflict and the interest are found, but it may be just as
                                                          important to work with those who are neutral or likely to be sup-
                                                          portive of the application. There are several reasons for this posi-
                                                          tion: (1) the neutral group frequently includes the majority of the
                                                          citizens and hence usually is neither organized nor recognized; (2)
                                                          this group can give the regulatory body a bigger "comfort zone"
                                                          in which to make a decision; and (3) people  often want to help,
                                                          they simply do not  know how to do it.
                                                            This   neutral  group   may  include   customers,   suppliers,
                                                          employees,  elected officials, industrial development groups and
                                                          others.  Many of them are your friends;  at least let them know
                                                          what you are going through and what they may be able to do to
                                                          help. In the final analysis, we are all in this effort together and we
                                                          need support and understanding wherever we can find it.
396
PUBLIC COMMUNICATIONS

-------
                                  Public Participation  in  Siting
              Hazardous  Waste Management Facilities  in  Alaska
                                                Sharon O. Hillman
                                       Sohio Alaska Petroleum Company
                                                Anchorage,  Alaska
ABSTRACT
  Hazardous waste management remains a major controversial
issue in Alaska. The RCRA amendments provided impetus for
immediate action on the issue of disposal of hazardous waste. The
State of Alaska was directed by the Alaska Legislature, specif-
ically the Committee Substitute for Senate Bill (CSSB) 503, in
June of 1984 to address the problem of hazardous waste dis-
posal.
  Alaska Statute (AS) 46.03 assigns the Alaska Department of
Environmental Conservation  (DEC) the responsibility  to study
and identify a viable hazardous waste management and disposal
facility for Alaska. This task is a large undertaking with consid-
erable opportunity for controversy among various interested par-
ties. One approach would be to address the issue actively with
public participation from various groups including  state agencies,
local government, environmental groups  and industries. The
group could assist in the development of action plans and pro-
vide recommendations for the best options for in-state hazardous
waste management facilities including disposal.

INTRODUCTION
  The  1981 Alaska State Legislature passed a law (Chapter  93
SLA 1981)  requiring the Alaska Department of Environmental
Conservation (DEC) to develop hazardous waste  regulations at
least equivalent to the requirements of the  federal RCRA.  Such
regulations would be the basis for a federally approved hazardous
waste program in Alaska. In order for Alaska to be in a position
to receive RCRA authorization, the U.S. EPA determined that
the original state legislation must be changed to provide  civil and
criminal penalties of at least $10,000 per day. This legislation was
introduced in the 1984 Alaska Legislature, and several bills affect-
ing the DEC hazardous waste program were introduced as a re-
sult. Eventually, the chair of the Senate Resources Committee
proposed merging the bills into one compromise bill (CSSB 503)
to address the diverse positions of the parties. The compromise
legislation, which was  passed into law, included  the following
three points:
• DEC shall develop and implement a program to educate busi-
  nesses and the general public about their responsibilities under
  the regulations.
• DEC shall evaluate and select in-state sites for hazardous waste
  management.
• DEC shall track the transportation of hazardous waste through
  the state.
  These points were based on input from the Hazardous Waste
Advisory Work Group (HWAWG), an advisory committee to the
state. The HWAWG group was originally formed in 1982 and met
throughout the year to address the 1981 legislation and provide
input for developing regulations. In addition to the DEC staff,
the group included representation from the Associated General
Contractors, the Alaska Oil and Gas Association, the Alaskan
Air Command, the Alaska Center for the Environment, the Sierra
Club, the League of Women Voters, the  Alaska  Federation of
Natives, the Alaska Truckers Association and the Alaska Public
Health Association. The new Alaska legislation and accompany-
ing funding renewed the process for a comprehensive hazardous
waste program. New mandates address not only the problems
of hazardous waste generated by businesses, but also those of
household waste, tracking waste shipments and hazardous waste
disposal facilities.
  The ability of the states to implement and enforce the RCRA
program adequately is the subject of controversy.  Until and un-
less a state is authorized by the U.S. EPA  to implement RCRA,
the U.S. EPA is required to implement the RCRA regulations, in-
cluding the issuance of permits for waste storage, treatment and
disposal facilities in that state.
  RCRA regulations primarily address the permitting and oper-
ating of hazardous waste  management facilities. RCRA does not
specify the process for selecting a hazardous waste management
site. However,  RCRA does state that certain regulatory require-
ments must be met before hazardous waste  management activities
may be  performed legally at the  site. Several states have estab-
lished programs for siting hazardous waste management facilities.
These programs generally specify  the process to be followed in
selecting a site, provide for public participation and determine the
role of state and local government in site selection and permitting.
  Many hazardous waste management facilities may require state
and federal permits or approvals in addition to the RCRA facility
permit(s).  These  permits  may govern: facility construction, air
pollution emissions,  water usage, wastewater discharge, under-
ground injection or disposal, highway rights-of-way and leasing
of state lands.
  The 1984 Amendments to RCRA have further forced the siting
issue to the forefront. The 1984 Amendments address the phase-
out of land disposal for hazardous  wastes, greatly increase the
number of hazardous waste generators by lowering the regulated
quantities, require the study and listing of additional materials
as hazardous waste, add a citizen's suit provision and increase
the penalties for  violations. The amendments increase the costs
and decrease the options currently available for disposal in the
lower 48 states. Currently, no disposal site  is available in Alaska;
however, many generators exist, and a site is needed.
  An evaluation of a proposed facility should include an assess-
ment of the need for hazardous waste management facilities of
some sort (what's generated, where and how much).  Additional
                                                                                       PUBLIC COMMUNICATIONS    397

-------
concerns to be addressed include whether the proposed facility
represents  the  most appropriate technology  for  handling  the
particular wastes generated and whether the location proposed
for such a facility  is suitable given environmental, social and
economic concerns in Alaska.
  There are three basic types of facilities which may provide in-
dependent or complementary management:
• Storage—holding  the  waste  prior to  treatment, disposal  or
  transportation to another state
• Treatment—turning the waste into something less hazardous
  or reducing its volume
• Disposal—final placement of waste
  The siting process should:
• Be well thought out by government agencies, the Alaskan pub-
  lic, industry and potential site developers
• Provide for early and full public information and participation
• Be understandable to tne public
• Provide  the public with accurate information concerning the
  benefits, risks and alternatives to a proposed facility
• Address the legitimate concerns of all parties
• Be a shared consultative process among site developers, facility
  operators,  communities,   government  agencies  and  other
  affected parties
• Provide  opportunities  for all concerned to negotiate areas of
  dispute
  Hazardous waste management sites are needed within the state
boundaries. Landfills, injection wells and process facilities all are
becoming more strictly regulated. Proper design, siting, permit-
ting and construction of a hazardous waste disposal facility could
take more than  10 years. Alaska must start now and not plan to
send its wastes to other states.

COMMUNITY INVOLVEMENT IN FACILITY
SITING—THE PROBLEM
  Environmental and health problems resulting from some  im-
proper hazardous waste management practices have caused peo-
ple to raise legitimate concerns about the nation's ability to man-
age  hazardous   wastes  effectively. Appropriate  technologies
should be applied at environmentally acceptable storage,  treat-
ment and disposal locations for hazardous waste. Public informa-
tion and participation programs that include the public early in
the siting process are increasingly viewed as necessary  steps for
addressing public concerns and for public acceptance of the tech-
nologies. The fact that hazardous wastes can be dangerous and
that the management of hazardous waste  is a highly  technical
area has been well documented.
  Another problem  with  the siting  of hazardous waste facilities
is the  "not in my back yard" syndrome—which has stopped the
permitting  of many  otherwise acceptable sites.  The fact that no
one wants to live near a hazardous waste facility further empha-
sizes the need for public participation in the siting process.

THE PROCESS
  Components  of  public information/participation programs
that facilitate negotiation, cooperation  and dispute resolution
can help overcome mistrust and skepticism and avoid legal con-
flict. Once a facility has been proposed, there may be only a short
time to institute dialogue before polarized viewpoints are  estab-
lished. Dialogue should be based, among other things, on cred-
ible information about the environmental integrity of a site, the
need for the facility, the performance characteristics of the facil-
ity and the financial stability and competence of the facility oper-
ator. The real challenge to industry and government is to provide
the public  early in  the  planning process with non-adversarial
points of contact that can reduce polarization and provide an
opportunity for addressing questions and concerns with candor
and clarity.
  What is being  suggested is the need for consensus  building
through an innovative process of private and public cooperation.
This is an important starting point which will:
• Foster positive involvement and dialogue among the interested
  and affected parties
• Help focus the issues and narrow the areas of real disagreement
• Provide ideas and information that may improve the quality of
  solutions and facilitate decision-making
  For hazardous waste management facility siting in Alaska, the
1984 legislation mandated  the consideration of several factors in-
cluding:  economic  feasibility,  intrinsic suitability of the site,
pollution control regulations, the risk  and effects on local resi-
dents, consistency of a facility with local land use planning and
potential adverse effects on agriculture or natural resources.
  The design of a  public information/participation  program
communicates how  much  the public's  involvement is valued or
desired. Providing information and avenues of response to pub-
lic concerns may  earn  public approval of the facility, result in
changes that make the facility  acceptable or justify disapproval of
a proposed facility. Regardless of the outcome, the actions will
help the community and the  state to  make decisions which en-
sure that new facilities will  be as safe as possible.
  RCRA has established the following procedure for public par-
ticipation after the notice to issue a draft RCRA permit is estab-
lished:
• The intention to issue  a draft RCRA  permit  must be pub-
  lished in major newspapers or  broadcast over local radio sta-
  tions
• This notice of intent  must  be sent to each unit of local gov-
  ernment that has jurisdiction over the area where the facility is
  proposed and to each state agency with authority over the con-
  struction or operation of the facility
• At least 30 days  must  be  allowed  for  comments on a draft
  permit
• Hearings are required, if requested, within 45 days of public
  notice of intent to issue a draft RCRA permit
• Public notice of  a hearing must  include, whenever  possible,
  the date, time, location and nature and purpose of the hear-
  ing. The hearing must  take  place  in  a location convenient,
  whenever possible, to the population center nearest to  the pro-
  posed facility.
• A fact sheet for every draft hazardous waste permit  must be
  prepared by the chief officer for the state's U.S. EPA-approved
  program, if the officer  finds  that the permit application is of
  widespread public interest or raises major issues.
  Since the State of Alaska has not yet promulgated siting regu-
lations, conditions for confrontation rather than cooperation are
enhanced. It  is  important for  individuals and groups to meet with
appropriate State officials to  encourage the development of a sit-
ing process and to open up the decision-making process in a way
that seeks public involvement, keeps the public informed and in-
corporates viable suggestions about  decisions that may affect
public health and the environment.
  State and  local government officials and potential site devel-
opers should recognize the planning and providing opportunities
for all concerned to participate in a proposed project will facil-
itate timely siting decisions that are supported by the community.
An important  aspect of the  process involves assuring the pub-
lic that its concerns will be considered, and, when appropriate,
will  influence decisions on whether or how to proceed  with the
siting and operation of a hazardous waste management facility.
398    PUBLIC COMMUNICATIONS

-------
THE PARTICIPANTS
  Participants in the siting process range from the general pop-
ulation to community leaders to those who may have a particular
interest in or may be affected by siting decisions. Serious efforts
must be  made  to inform, involve and respond  to these some-
times overlapping groups during different phases of a public in-
formation/participation program. Possible participants are:
• Public representatives:
  -Public officials, both elected and appointed
  -Federal, state and local government agencies
• Generators and Users:
  -Businesses and industries
  -Hazardous waste generators
  -Site developers and facility operators
  -Property owners in the vicinity of the site
  -Consulting scientists, engineers and other experts
• Public and Interested Parties:
  -General public
  -Public interest groups
  -Environmental and conservation groups
  -Local scientists and engineers
  -Ad hoc citizen groups
  -Civic associations
  -Community and religious groups
  -Consumer groups
• Other Organizations:
  -Trade, industrial, agricultural and labor organizations
  -Public health, scientific and professional societies
  -Educational institutions
  -Media, including editorial boards

THE TECHNIQUES
  The use of early and broad public information and consulta-
tion techniques will inform interested parties about the facts of a
proposed facility and may help to narrow areas of potential con-
flict and promote understanding. Many  fears about  sites  are
based on unknown factors or concern that another "Love Canal"
will result without the people having any input, therefore exten-
sive public education and input programs are essential. Whether
information and participation programs are designed by  govern-
ment regulation or by community/developer interaction, numer-
ous techniques may be used  to encourage understanding and
evaluation of a proposed siting project:
• Information Dissemination:
  -Fact sheets
  -Agency and interest group newsletters
  -Education of the media for proper use of use of print, radio
   and TV education
  -Letters to people on mailing lists
  -Information centers and public libraries
  -Spring Cleanup  (agency-sponsored  hazardous waste collec-
   tion)
  -Hotlines
• Consultations:
  -Advisory committees drawing on  major interest groups and
   representatives of the affected local community
  -Technical advisors
  -Public workshops for large and small group discussion
  -Public opinion surveys
  -Small group meetings covering specific concerns
  -Use of models
  -Public tours and walk-throughs of facilities
  The critical influence of the facility's type and location on the
potential  for adverse environmental impacts makes it important
to have comprehensive state or regional siting criteria and man-
agement plans for hazardous waste facilities. These criteria and
plans should be developed with public participation and should
identify, based on hazardous waste management needs and en-
vironmental sensitivity, the most environmentally acceptable type
of facility and location for the state.
  The plans must be realistic in  evaluating what is technically
feasible and economically supportable given the type, quantity
and  generation rate  of hazardous waste. The Alaska DEC re-
cently has initiated such a plan for Alaska. The  first part of this
plan includes  a  "Waste  Disposal Study" to be completed by
January 1986,  and a  second study, "A survey of Approaches by
Other States in Establishing Criteria for the Location of Haz-
ardous Waste Facilities," also to be completed in January. Both
of these studies are in support of the draft regulations to be issued
in early 1986.

COSTS AND BENEFITS
  A community may experience increased costs as a result of the
new facility. One major concern is the remote possibility of health
and  environmental injuries that may  arise if there  are releases
from the facility. Other increased costs may be related directly to
the site: security patrols, fire equipment and new tasks for fire
and police personnel. For example-, a small local fire department
or hospital might need new equipment and different or additional
training to handle emergencies related to the facility.
  While the developer and/or  operator of  a hazardous waste
facility is interested in  a profitable investment as well as a safe
and effective business, the community's prime concerns relate to
health, quality of life and property values. If the proposed  facil-
ity is viewed as a threat to these values, the community may want
some sort of community benefit in return for accepting the facil-
ity.
  The community may request the operator to provide funds for
services related to the facility including additional fire-fighting or
hospital equipment or improved roads in the vicinity of the site.
Alternatively, the developer or operator could contribute a fixed
sum to the local government for needs that may arise in connec-
tion with the project. Another solution could be a type of sales
tax whereby the operator would pay a fixed percentage of oper-
ating revenues  to the local government.
  These additional costs,  whether a part of the initial investment
or a part of higher operating costs, will add to  the general eco-
nomic feasibility of the project. As such, it is very important in
studying the project economic viability to address all costs at the
very beginning. In some cases, the state may  assist the local gov-
ernment in mitigating potential community impacts,  especially if
they encompass a larger area or population such as would be the
case in Alaska.

CONCLUSIONS
  The key to developing a good program which involves the pub-
lic in the siting process must:
• Provide adequate information to the community
• Provide a means to listen to and address the community's con-
  cerns
• Develop sound methods of interaction among all interested
  parties
  Communities have developed their own task  forces, advisory
committees, information systems and other educational and com-
munication methods. Volunteer leaders have been used to work
with site developers,  facility operators and the government. Po-
tential sources of funding for community participation programs
are government, civic,  industry and other groups. The goal is to
develop a program which will work effectively toward a disposal
                                                                                           PUBLIC COMMUNICATIONS    399

-------
 answer for the state while enhancing the cooperation and direct-
 ing the efforts of all interested parties toward the right answers.
   Introduction of the public  to  hazardous waste  management
 facilities and siting options could create a greater  appreciation
 and knowledge of the  complexity of properly managing haz-
 ardous waste  in  Alaska. Several objectives would be accom-
 plished including bringing  more  affected entities into the dis-
 cussions, educating the regulators to practical application and
 focusing  the hazardous waste issue on  an area  of common
 ground. This process should lead to a viable hazardous waste dis-
 posal option in Alaska.
   The Hazardous Waste Advisory Work Committee should be re-
 established to work with the DEC in developing a plan of action
 for disposal option(s) and site(s) in Alaska. The committee would
 include  representatives of state agencies, local governments, en-
 vironmental groups, industries and the state legislature. The draft
 regulations currently being  developed should be shared with this
 group to get early input prior to  the legislative deadline of July
 1986 for the adoption of regulations. The State DEC should con-
 tinue its current effort to use newsletters,  fact sheets and spring
 cleanups to promote the  local education process.
   A  balanced picture of  the hazardous  waste  threat  to the
 Alaskan environment and its people is an important perspective
 for the State to consider. The current DEC study will help answer
 this question. A strong emphasis on the conservation and recov-
 ery aspects of the Resource  Conservation and  Recovery  Act
 (RCRA) program  will provide direction for the Alaskan agencies,
 public and industry to educate each other about the hazards  in-
 volved and to minimize the  hazards (by reduction and treatment)
 to the maximum  extent  possible. And last, but  not  least,  one
should recognize that this is not  a one-time education process,
but one that is required on a continuing basis for all of us to live
in our world, together.
REFERENCES
1. Fawcett, H.H.. Hazardous and Toxic Materials, Safe Handling and
   Disposal, John Wiley & Sons, New York, NV, 1984.
2. Peirce, J.J. and Vesilind. P.A., Eds, Hazardous Waste Management,
   Ann Arbor Science Publishers, Ann Arbor, MI, 1981.
3. Quarles, J.,  Federal Regulation of Hazardous Wastes, A Guide  to
   RCRA,   Environmental  Law  Institute, Washington,  DC,  1982.
    (RCRA," 42 U.S.C. 6901-6987 1986 and Supp. Ill 1979) as amended
   by the Solid  Waste Disposal Act Amendments of 1980.
5. Rodgers,  W.H., Jr.,  "Directions  of Environmental  Law  in the
   1980's", presented at the Hazardous  Waste Law and Management
   Conference,  University  of Washington Law School, Seattle, WA,
   1984.
6. The Committee Substitute for Senate Bill 503 (Finance), in the Legis-
   lature of the Slate of Alaska. Thirteenth Legislature—Second Ses-
   sion (CSSB 503). May 25, 1984.
7. The Conservation Foundation, Siting Hazardous Waste Management
   Facilities.  White Plains, MD, 1983.
8. The Hazardous and Solid Waste Amendments of 1984, Title II, Pro-
   visions Relating Primarily 10 Subtitle C of the Solid Waste Disposal
   Act (Subtitles A, B and C)
9. Wood,  L.,  "Alaska  Department  of Environmental Conservation
   Hazardous Waste Management Program  Annual Narrative, 1983-
   1984," Proc. Hazardous Waste Law and Management Con/., Uni-
   versity of Washington Law School, Seattle. WA, 1984, 258-273.
400    PUBLIC COMMUNICATIONS

-------
              Developing  a  Comprehensive  Public  Affairs Policy
                     for  New and  Existing Industrial  Operations
                                                   Ann E.  Burke
                                           Environmental Impact, Inc.
                                             Baton Rouge, Louisiana
ABSTRACT
  The primary public affairs policy problem faced by most indus-
tries involved in the handling of hazardous wastes/materials is
simply devising specific programmatic approaches to the desired
audiences. This paper offers specific program suggestions for five
groups defined as the most important audiences for an industrial
public affairs program: the Media, the Community, Government
officials, Customers and Employees. The focus of the suggested
strategies is the constructive dissemination of unbiased, technical-
ly sound information to all these groups.

INTRODUCTION
  Your secretary buzzes your office. You detect a note  of trepida-
tion in her voice. Suddenly, the dread words waft over the inter-
com..."Mr. Johnson—there's a Morley Safer here to see you...
and Mr. Johnson, he wants permission to film while he talks to
you...Mr. Johnson?...Mr. Johnson!...."
  There are four possible outcomes at this point:
• With luck, you will wake up.
• You will refuse comment and be made to look like Attila the
  Hun on next week's Sunday night broadcast.
• You will go into the interview unprepared and watch  yourself
  be battered by the reporter who has already researched the en-
  tire story, who has made up his mind about the guilty party
  (undoubtedly you) and who is only in your office to fulfill his
  "ethical" duty to get your rebuttal.
• You will be prepared for  the interview since one of  your com-
  munity contacts  will have alerted you  days earlier that the
  "60 Minutes" crew  was in town. Your staff will have ap-
  proached the crew and offered them  access to your facilities
  and personnel. Your  interview  will be your opportunity  to
  make a coherent and well researched statement about the issue
  at hand.
  To many of you, this final outcome seems unlikely at best. Yet,
it is my  firm belief that structured, organized and  well-directed
public affairs policies can produce a positive scenario such as that
I have recounted.
  To most industries, the advantage of anticipating the inquiries
and needs of their public has been ignored or overlooked  in favor
of adopting reactive public affairs postures. We allow the story on
our new waste incineration process to be "uncovered" by the
media and thereby sensationalized without benefit of fact or our
own point of view. We meet our  community in a forum of litiga-
tion rather than negotiation, thereby immediately casting  our-
selves  as the uncommunicative, overbearing industrial warlord.
  1 would like to recommend several very specific program strat-
egies that can be implemented to begin to  change our public re-
sponse posture from reactive to anticipatory and constructive. I
have had the opportunity to design and administer many of these
programs  for  one of  the  major waste disposal  companies,
CECOS International. I have since designed similar programs for
other clients and have seen  the positive impact that a well de-
signed outreach program  can  have on industry's image in the
community.

INTERNAL STRATEGIES—THE FIRST STEPS
  The focus of your public affairs policy should be sharply and
clearly delineated prior to  structuring the actual programs.  New
projects will require different  strategies from existing projects.
Although this  may seem obvious, it  is extremely important to
understand that while you may have the opportunity to shape and
direct perceptions about a new project, you will have to acknowl-
edge and accept perceptions about existing operations.
  I suggest that a small (no more than four people) committee be
formed and given one full day to:
• Describe the project
• Describe the need for the project
• Outline any community impact (positive and negative)
• For a new project,  describe the desired perception of the pro-
  ject by the community.
• For  an existing project, describe present perceptions  of the
  project and desired perceptions.
• Select a Public Affairs Representative for the project.
  Above all, it is important to be realistic in your goal  setting
with regard to the project. For example, you will never get 100
people to turn out at a public  hearing to voice their enthusiasm
for a new hazardous  waste site. However, you may be able to
minimize the number  of opponents to the  project by disseminat-
ing clear, straightforward project information to the public.
  Once a Public Affairs Spokesperson has been chosen, resource
people should  be selected. It is important to have at least one
management and one technical contact for the  Spokesperson.
These people will be  kept continually informed by the Spokes-
person and will be prepared to handle special interview situa-
tions (press conferences, technical journal interviews, etc.). Con-
versely, these resource people will serve as cotiduits to the Spokes-
person for new project information.
  After the  foregoing has been established, the  Spokesperson
can begin to develop  a Comprehensive Public Affairs Policy. I
will outline some specific ideas for  such a  program in  the re-
mainder of my presentation.

IDENTIFYING THE AUDIENCE
  You will note that I have selected five target audiences  for my
policy  archetype:  Media,  community,  government officials
(elected and non-elected), customers and employees. I would like
to comment briefly on specific suggested contacts within these
general groups.
                                                                                       PUBLIC COMMUNICATIONS    401

-------
Media

• National Media: National magazines,  national newspapers
  (USA Today, Christian Science Monitor, Wall Street Journal),
  network news, CNN News (they will cover what others do not),
  national radio shows (Talk Net,  Larry King,  National Public
  Radio)
• Local Media: Local television (contact all local talk and inter-
  view shows—do not forget PBS), newspapers (do  not forget
  Pennysavers, small community tabloids), local magazines  (es-
  pecially Chamber of Commerce  and business  oriented tab-
  loids), local radio (especially all-talk and all-news  formats),
• Trade Media: National and local trade journals  (Chem-week,
  various industrial/regulatory newsletters)

Community
• Citizens within your immediate community or some specified
  radius of your facility. Address lists can be purchased by zip
  code.
• Professional Groups and Societies
• Fraternal  Groups:  Lions,  Rotary,  Kiwanis  Clubs,  Garden
  Clubs, PTAs
• Schools

Government Officials
• Elected and Non-Elected (the Sewer Board,  the Planning Com-
  mission, the Health Inspector, etc.)

Employees
• The most valuable and most under-utilized resource that you
  have. Certain programs also should include your employees'
  families.

SUGGESTED PROGRAMS
Programs for the Media
Facility Tour and Press Conference
  This obvious  and  tremendously valuable  strategy  is almost
never implemented in an organized fashion  by  industries. It is
most important to familiarize media personnel with your facility
and/or project. By conducting a tour, you are allowing access to
the press as well as getting an opportunity to tell your company's
story.  Without sanctioned access, the media will  portray your
company as a series of locked gates and employees hiding their
faces. By arranging constructive photo opportunities with a sub-
sequent press conference, you will be able to  direct the image of
your company/project as it is fashioned by the media.
  I suggest that a tour-related press conference be conducted and
directed by the Spokesperson. The CEO and technical resource
should be available for specific press inquiries as directed to them
by the Spokesperson.
  At  the time of the Press Conference, a carefully constructed
Press Guide should be distributed. This Press Guide should con-
tain:
• A company history
• Brief biographies of top management
• Economic impact of the project
• Description of the project
• Reproducible diagrams or charts
• Name  of contact  person (with  home and office  telephone
  numbers)
Press Releases/Story Opportunities
  Press releases should be an on-going part of any  Public Affairs
Program. All research programs, promotions, earnings  reports
and new technologies should be subject to press releases.
 Story  opportunities  should be presented to  individual media
 members as they occur. Story opportunities can present anything
 from interesting employee/public interest stories to exclusive in-
 formation on acquisitions or new technology breakthroughs. Re-
 serve your  "story opportunity"  approaches to the media for
 especially interested cases.
 Individual Relationships with the Media
   One of the  first responsibilities of any Project Spokesperson
 should be to develop an individual relationship with those mem-
 bers of the  media with which he/she most frequently interacts.
 Personal facility  tours and meetings enhance not only the re-
 porter's knowledge of the facility, but also serve  to foster a per-
 sonal acquaintance between the reporter and the Spokesperson.
 It is much more difficult to write a totally biased story about a
 company  when  you have  had personal contact with  its  em-
 ployees.
 Meetings with Editorial Boards
   Each large newspaper has a select group of personnel who com-
 prise the Editorial Board. The Board selects the topics and view-
 point for the papers' editorials. I have found meetings with  Edi-
 torial Boards to be the single most helpful tactic in receiving posi-
 tive  press coverage. Editorial  Boards should be briefed on the
 project/company by the CEO and the Spokesperson.
   In the case of existing projects and facilities, samples of un-
 fair coverage by the  paper  should be assembled and discussed.
 A succinct but professional  presentation on the project should be
 made. Full contact information for any future inquiries should
 be readily available.
 Technical Resource Center
   The media should  be made aware that all the company  per-
 sonnel are available for technical  consultation or project or re-
 lated technical issues. At CECOS, we were able to have several of
 our staff used as experts by the media in articles relating to  gen-
 eral hazardous waste issues. All company reference and research
 materials also should be made available to the media.
   Both of these programs lend technical credibility to your  per-
 sonnel and organization as  well as giving your company stature
 as an expert in the field.
No, No Comment
   It  is my policy  to advise  clients against ever adopting a  "no
 comment" stance except in cases under litigation. The public  per-
 ceives "no comment" responses as one of three things:
 •  You do not  know the answer and are uninformed about your
   company's affairs
 •  You are hiding something
 •  Both of the above
   You cannot  benefit from  a  "no comment" response.  I recom-
 mend that if anyone is faced with a question he does not feel qual-
 ified or inclined to answer at  the moment, he should get the re-
 porter's name and number and give him a response at a later  (but
 definite) time.  Do not be afraid to tell the press that you do not
 have the  facts with you, but that you will assemble them and re-
 spond later. Another alternative is to offer a full interview the
 next day rather than an answer to a single question. Be resource-
 ful—but always comment.

 Programs for the Community
 Facility Tour and Open House
   Most people who vehemently oppose a project or facility  have
 no idea what such a facility looks like or what takes plate at  one.
 Therefore, the first community program that should be imple-
 mented is the  facility tour and open house. The tour should be
402    PUBLIC COMMUNICATIONS

-------
held on a Saturday and always should be preceded by a full pres-
entation on the company, its history and facilities. The tour
should be followed by an open question and answer session with
company personnel.  The names  and addresses  of all attendees
should be recorded and kept for future activities and mailings.
  The  Open House should be advertised in all local papers, and
all local media should be invited to cover the event.
Openline
  A toll-free number of citizen inquiries should be established. I
resist the title of "hot line" because I feel it is associated with
some  type  of disaster—"openline" has a more informational
overtone. All inquiries received on the Openline  should receive a
written followup response.
Directed Mailing Program
  A program of directed mailings to the community should be
implemented either immediately after the announcement of a new
project or to address specific issues related to an existing project.
   Examples of topics for directed mailings might  include:
•  Economic impact of the facility on the community (taxes, pay-
   roll, vendors, etc.)
•  New technological  developments
•  Industries dependent on this facility—service performed for in-
   dustries
Community Newsletter
   This newsletter should be published quarterly for  distribution
to the targeted citizens and citizens' groups in  the community.
Some suggested topics include:
•  New facility development
•  Profiles of employees (spotlight personal and professional in-
   terests)
•  New technological  developments at your facility
•  Recent community events
•  Community contributions by the company
   There should be some response  mechanism in the newsletter
—a prepaid card  for comments, requests for tours or speakers,
etc.
Speakers' Bureau
   Employees with specific expertise should be identified and in-
volved in  a Speakers'  Bureau.  A brochure  listing all available
topics should then be distributed to all professional, fraternal,
civic and educational organizations in the community. Be certain
to include employees  vocational and avocational interests.
Financial Contributions to Community Causes
   In most organizations, contributions to community causes  are
made without thought to their overall impact. In many instances,
$100 and $200 contributions  are made without being part of a
larger  focussed giving program. For example, at  CECOS we were
able to identify over $10,000 per year  in contributions that were
being  distributed without an overall giving  strategy. Once this
amount was identified, we selected one major, visible community
project for funding. We provided matching funds for a project to
buy a new fire truck  for the community in which one of our dis-
posal sites was located. This single contribution had a much more
far-reaching and visible impact on the community than a  series
of $100 and $200 contributions.
School Programs
   Secondary schools, colleges and  universities  always  welcome
speakers from local industries. Contact department  chairmen as
well as placement and guidance  counsellors with Speaker's Bur-
eau information.
   For secondary schools, a company sponsored essay contest on
the environment will highlight the company's involvement  in the
community and provide an excellent media coverage opportunity.
Household Chemical Disposal Days
  For industries with disposal facilities, a day for the free dis-
posal of household chemicals will highlight the pervasive problem
of hazardous materials in our society.

Programs for Government Officials
  As discussed previously, tours and open houses are one of the
most effective mechanisms for removing the mystique surround-
ing your operation.
  In general, major elected officials should have personal tours of
the  facility while others, such as a Town Council, can  receive
group tours. Meetings  with top  management officials  always
should follow these tours. Special emphasis on the economic con-
tributions of the company to the community should be made.
  A systematic program of followup contacts  should be estab-
lished following the tours. A contact person in the organization
should be selected and identified for all governmental inquiries.
  The personnel and reference resources of the company should
be detailed for government officials and made available to them
for  legislative  research.  Expert witnesses also  should be made
available.

Programs for Customers and Vendors
  Customers and vendors can be an important source of sup-
port within  the  community. Customers, in  particular,  can be
valuable  resources at public hearings and can help focus atten-
tion on the importance of your facility/project to the community.
Many of the aforementioned programs (Open Houses, Tours and
Newsletters) would have  obvious benefit to your customers. I also
would suggest  that you distribute periodic legislative updates to
your customers. The development of structured lobbying organ-
izations in response to specific legislative proposals will augment
the credibility of your leadership within the industrial community.

Programs for Employees
Tour of Your Facility
  A survey conducted at CECOS revealed that  over 70% of our
office personnel had never visited our disposal  site—despite the
fact that it was only 20 min away. After a full  day of tours  and
presentations for our employees, we got  a universally enthusiastic
response. An extensive period of questions and answers followed
the  tour  and presentation.  Afterward,  people  commented that
they felt  much better equipped to act as advocates for the com-
pany.
Newsletter for Employees
  A newsletter can serve as a vehicle  for developing company
morale as well as a tool for  disseminating information about
new technological or financial programs  for the company.
Employee/Legislative Liaison Program
  All employees living in the immediate area of the facility should
be asked to  make contact with their various legislative represen-
tatives. When specific legislative issues arise, employees can be
asked to write or call their legislators.
CONCLUSIONS
  I hope this paper has convinced you to take the first step to-
ward  establishing a comprehensive  public affairs program for
your organization. If you do not, you will condemn yourself to
repeating the same reactive posture you always have taken with
the press and the community—with the same unsatisfactory re-
sults.

  And, who knows, this time Morley really may be outside the
door.
                                                                                           PUBLIC COMMUNICATIONS    403

-------
                        Behind the  "F" in  Public  Education  for
              Hazardous Waste  Management:  A  Case for Special
                  "Tutoring"—Plus Some Tips  for Better  Grades
                                               Howard A. Coffin
                                                 Patricia Hunt
                                        Fanfare Communications, Inc.
                                          Philadelphia,  Pennsylvania
                                               L.T. Schaper, P.E.
                                                Black & Veatch
                                             Kansas City,  Missouri
 ABSTRACT
   In recent years, engineers and waste management companies
 have been winning the battle of technology but losing the war of
 public opinion. Hundreds of waste management projects have
 been scuttled  in the face of overwhelming public opposition
 traceable to the waste management industry's inability to com-
 municate its side of the story with any real impact or credibility.
 Engineers and waste management technicians have some built-in
 limitations in this regard, but they can learn to do better—with
 some professional help and some practical tips contained in this
 article.
Where jrubl.ir acceptance in concerned, nil ounce of
is worth a i>oun
-------
 American industry doesn't care how  many wells it poisons or
 what it does to the air we breathe.
   These sizeable handicaps notwithstanding, the hazardous waste
 management industry is still the leading culprit in the failure of
 public education. Talk about "slow to learn!" Over and over
 again, companies in  the waste management field  have invested
 huge sums of money  in feasibility studies, geologic surveys, pro-
 perty acquisitions and design and engineering fees, only to have
 their projects fail the test of public acceptance. But while be-
 moaning the fortune that keeps going down the drain, and know-
 ing full well that public relations and politics, not  technology, is
 the problem, the industry still refuses to spend any real money for
 professional public relations and marketing help. Instead, it has
 tackled  the highly specialized task of public acceptance  on its
 own. Countless companies turn to  outside help  only after com-
 munity opposition already has become so impacted that the pro-
 ject is doomed to failure.
    Judging from its abysmal record, one would think the industry
 would finally admit that the engineers, scientists and other experts
 who are so  knowledgeable about the technology of hazardous
 waste treatment have a lot to learn about how to sell the public on
 their plans. Not only  do they have a lot to learn, most of them are
 constitutionally ill-equipped to do the job themselves. Here's why
 the technical experts  can't hope  to succeed without a massive in-
 fusion of expert  tutoring and  direct, ongoing  assistance  from
 specialists skilled at persuading the public.
Winning over the public can be the key to a winning project.
  WHY PROFESSIONAL PUBLIC RELATIONS
  EXPERTS ARE NEEDED
  Engineers and Scientists Naive
   Sheer  naivete is at  the  heart of the industry's  problem.
  Engineers and scientists are rationalists. They tend to believe that
facts are—or should be—as persuasive to others as they are to
themselves. They have a hard time seeing things from the perspec-
tive of the general public. If they make a logical case for a project,
then back it up with solid facts and research, they think the public
should react favorably. They forget that there are a lot  of other
ways  to  look at waste management  problems than  from  a
technical, engineering perspective.

Operations-, Not  Marketing-Oriented
  A  second,  closely  related,  issue  is  that  engineers  and
engineering-based companies, including large waste management
firms,  traditionally  have  been  very  operations-oriented, not
marketing-oriented. Even some of the biggest firms  have been
slow  to  develop  anything more than  the  most  rudimentary
marketing communications  and public relations efforts.
  This tendency is consonant  with the personality type of the
engineers and waste management specialists Fanfare Communica-
tions has come to  know (and respect) in more than seven years of
dealing with  environmental issues. They  are generally uneasy
wrestling with subjective judgments. They prefer to avoid situa-
tions calling for salesmanship and the manipulation of other peo-
ple's feelings. They are more comfortable dealing with tangible,
quantifiable things. How much? How many? What weight? How
high?
  Even the professionals who do spend a fair amount of time sell-
ing rarely have much experience dealing with the general public.
Engineers don't sell their services to the man-in-the-street; they
sell to other engineers or to technocrats, who see the world as they
do.

Technicalese
  A third factor compounding the communications problem  of
many science and engineering types is that they don't even talk the
same  language other people do. Even when they're trying their
level best to  communicate with a lay audience, they talk  in
"Technicalese" without realizing it. "Leachate collection zones"
and "low-permeability soils" are simple enough terms to a waste
management specialist, but  they don't communicate much to an
untutored citizen.
  For example, one of the  country's  largest  waste management
firms  last year published a "Fact Sheet"  aimed at clearing up
public misconceptions about  hazardous  waste management.
"PCBs: What Are They?"  said the fact sheet, implying that the
answer which followed would clear up this mystery once and for
all. That possibility was quickly dashed. What would your non-
technical friends or spouse make of this explanation?
  "The name stands  for polychlorinated biphenyls, and  contrary
to popular belief,  PCBs are not one chemical but a class of com-
pounds. These compounds are two benzene rings joined  together
at one carbon on each ring. The rings are then substituted with
one to ten chlorine atoms."

Too Expert to Communicate?
  Fourth,  as specialists in demanding technical fields where
depth, not breadth, of knowledge is the common measurement of
true expertise, engineers and scientists are prone to write  off their
inabilities as public communicators with such rationales  as, "I'm
an engineer, not an actor." As a consequence, they invest little ef-
fort in improving  their skills and may simply avoid situations that
highlight weaknesses in this area. Like most people, if given a
choice, they'll define their jobs and responsibilities to emphasize
their strengths, not their liabilities.
  Finally, in  organizations  that traditionally have rewarded per-
sonnel for their  engineering and  operational know-how, and
which recruit new professionals by those same criteria,  qualified
professionals are  often reluctant to make marketing and com-
                                                                                            PUBLIC COMMUNICATIONS    405

-------
                                                                              l — CHrC-CI*SO,-<
                                                     H^NH-CH -COOW-r; RX —CW, -t- H,O+M2SO.,1
7o<> o/h
                                                  i talk "t('<'lininilfs('"lh(itri>iifii.
-------
Some Practical Suggestions
  These, then, are the key participants for a successful public ac-
ceptance effort. But how should they proceed, and what mistakes
should they be careful to avoid?
  Whole books and lengthy manuals have been written on the
subject of public  participation programs  for waste management
projects, but they pay far more attention to the rules and regula-
tions for formal hearings and proceedings than to the more fun-
damental issue of winning public acceptance.  Space doesn't per-
mit a full exploration of all key components of a sound public ac-
ceptance program, but here are some practical suggestions gleaned
from the experience of Fanfare Communications and its clients in
the waste management industry.

In General
• Think carefully before trying to go around the public. It rarely
  works. Have a  good public acceptance program ready to put
  into place even though you think it won't be  necessary.
• It's usually easier to  control public reactions if  you take the
  initiative instead of being forced to react.
• Communicate a sense of full disclosure. Better that YOU tell
  people the bad news than for the public to hear it  from another
  source and think you've been deceiving people.
• The bottom line is not to be declared "right"; it is to get enough
  people in the community to believe enough  of your story to
  allow your company to do what it hopes to do.
 • "Walk in the public's shoes" if you want to understand a com-
  munity's concerns. If you do not understand its concerns, you
  will never gain  its acceptance.
 • Never lie to the public or "shoot from the hip" with a  state-
  ment you might later have to retract.
 • Never make up an answer and then try to support it later.
 • Communicate in  such a way that people who know absolutely
  nothing about  the environmental sciences, chemistry or en-
  gineering can understand what you say. It makes people nerv-
  ous, worried, embarrassed and  angry to be confused.
 • Even if you can't win  outright public acceptance, try to es-
  tablish an atmosphere calm enough to permit regulatory  agen-
  cies to do their jobs free  of  public  hysteria and excessive
  political pressure.
 • Public acceptance programs do not succeed  because of what
  YOU know, but because of what the public FEELS.
 • Internal communication is critical. Develop a consistent pro-
  ject narrative and description, and make sure all parties stick
  to it.
 • Do not think of "the public" as a single entity. Think of com-
  munity residents  in terms of "active, intractable opponents,"
   "passive opposition," "disinterested neutrals,"  "persuadable
  neutrals," "allies" and "latent allies." There are also opinion
  leaders,  special  interest groups (labor, churches, business,
  environmental  groups, the PTA, ethnic concentrations,  etc.),
  the news media, elected officials, regulatory agencies and civic
  groups. "Target" messages appropriate to  each  sub-group,
  but never make contradictory statements.
 • Have  specific,  detailed objectives for  all the communication
  tools you employ, including: the official project  spokesperson
  (The Communicator), speeches  to specific groups of people,
  public information  meetings, printed brochures and  book-
  lets, audiovisual presentations,  individual meetings  with of-
  ficials and opinion leaders, radio and TV commercials, posters,
  targeted mailings, publicity releases, news conferences and an-
  nouncements, official hearings, public debates and news media
  interviews.
 • Tap the resources and  trade associations of  waste generators
  in a  community to participate in general education efforts
  about the importance of sound waste management.
 •  Line up friends.  Harness their self-interest and provide them
   with the materials and support to  make it easy for them to
   help.

 Spokespersons
 •  Insist  on having  one  official project  spokesperson.  Many
   mouths make hard work, lots of headaches and embarrassing
   contradictions.
 •  It is essential to level with the media as much as possible, es-
   tablish trust and not be  seen as pro-company zealots. Give
   them the freedom to succeed for you.

 At Public Meetings
 •  Urge all company speakers and respondents to questions to be
   "real people"  and let (the good sides of)  their personalities
   show. Humor, at the proper time, can be a real trust-builder.
 •  Your personal manner and appearance communicate at least as
   much information about your project as what you say about it.
   Look confident and  relaxed.  If you cannot, do not appear in
   public meetings.
 •  Use standard persuasion techniques, including always starting
   where your audience is, not where you are or where you want
   them to be.
 •  Insist on getting professional training in public speaking and
   responding to  hostile questions for all  company representa-
   tives likely to make public appearances.
 •  Use videotape cameras to  record the way company representa-
   tives look and sound  at public meetings and  to help them prac-
   tice their presentations beforehand.  There is no better way for
   people to learn than to see  themselves  as others see them—
   especially if they are  going to be seen on television.
 •  Address the intent of people's questions, not just the specific
   wording, which is often inarticulate. It  makes you appear to
   be more forthcoming.
 •  Always schedule ample time for give-and-take with the  public
   and for  clarifying questions and answers. Do not put  this
   crucial component of public meetings on an "if there's  time"
   basis.
 •  Be clear  about what you want to convey in  all public  utter-
   ances.  Have simple,  modest  goals for your statements.  Com-
   plexity bewilders. Bewilderment breeds anxiety.
 •  Avoid  jargon and "technicalese."  It separates you from the
   audience just when you want them to be with you.
 •  The purpose of public meetings is  to generate credibility and
   better feelings, not to flood people with a vast quantity of in-
   formation. It is not  your message that matters; it is how the
   public  feels about you and your project.
 •  Do not disengage from the audience by constantly staring down
   at a prepared speech and reading in a monotone. Master the
   material  so  that  you  can look up frequently.  Engage the
   audience.  Be expressive.
 •  Learn to tolerate people directing their  angry comments and
   concerns at  you.  It  is seldom truly personal, and it usually
   helps to defuse public hostilities when people are allowed to
   spout off at someone—unpleasant as that may be.

Audiovisual Programs
 •  Unless they are accompanied by a  skilled speaker  or are  de-
   signed  to  allow interaction with the audience,  one-projector
   slide shows without a soundtrack can be pretty static and bor-
   ing. The best format  is a technically  uncomplicated, not  overly
   slick,  two-projector  program with a pre-recorded, synchron-
   ized soundtrack.  Numerous studies have shown that  it sus-
   tains interest  better  and  produces  much fuller information
   recall.
                                                                                           PUBLIC COMMUNICATIONS   407

-------
 •  Try to put  only one  idea  on each slide.  Detailed charts,
   lengthy lists  and wordy graphics are a staple of engineering
   presentations, and  they are almost worthless as communica-
   tion devices.
 •  If the eye cannot quickly grasp and easily "hold on to" the
   visual information in a  slide, it is a bad slide.
 •  The use of all capital letters tends to be less readable than capi-
   tal and lower case letters combined—the way they are in 99%
   of the things we read every day.
 •  The typography in many slides resembles "The Lord's Prayer"
   written on the head of a pin.  Keep type clear and large enough
   to be quickly legible.
 •  Audiovisual  programs that present  information only and  fail
   to address and affect people's feelings  are failing to use the
   full potential of the medium.

 News Media Relations
 •  Educate the  news media patiently. Keep giving them  clear, in-
   formative information and their coverage  will improve.
 •  Establish your company as a dependable source of  informa-
   tion about waste management in general,  not just as  it relates
   to your  company.  It makes reporters more knowledgeable
   and builds useful bridges with the news media.
 •  See every potential story about your project as the press might
   see it. Your perspective is not a reliable index of how the media
   will, or should, deal with news about your company and its
   projects.
 •  Keep your cool when "correcting" the news media for erron-
   eous stories.  Mistakes rarely are intentional, and you will bene-
   fit in the future from not insulting a reporter who  gets things
   wrong. Never "demand"  retractions unless you  have  firm
   legal grounds for doing so. A firm request often  gets better
   results.
Booklets and Brochures
• Good printed materials should adhere to the same communica-
  tions guidelines stated for other media.  They  should  avoid
  "technicalese,"  be clear and  understandable to non-experts
  with only a high  school education and address the readers'
  interests and concerns, not simply hammer home the com-
  pany's message.
• Printed materials should communicate well on several levels:
  they should use boldface or color-emphasized headlines, sub-
  heads that add  details and  interest to the headlines, photo-
  graphs and  illustrations that convey visual information, and
  caption material that expands on them. All of these elements
  help "pull the reader in," and at a time when people are being
  bombarded with a vast amount of printed matter, every read-
  ership encouragement helps.
• Generally, waste management companies  should avoid  using
  excessively slick and sophisticated publications that may convey
  too strong a "big business'V'big money" impression to people
  in a particular community. Sometimes,  however, it is impor-
  tant to  convey a  very substantial image, in which case the
  graphic  presentation and paper quality can signal your  com-
  pany's ample resources.
• Consider using targeted mailings to direct specific messages to
  specific  audiences.  It is expensive, but effective.

Radio, TV and Print Advertising
• If you are frustrated by repeated failures  in getting the  news
  media to pay  sufficient attention to your side of the story,
  consider using advertising to communicate a message that you
  and you alone control. But get professionals to create the copy
  and graphics. Advertising that  misses the boat from a message
  standpoint, makes  no strong impression or  fails to get people's
  attention in the first place is a terrible waste of precious  com-
  munications resources.
  Now get out there  and start educating!
408    PUBLIC COMMUNICATIONS

-------
                 Recycling  of  Dust  from  Electric  Arc  Furnaces-
                                   An  Experimental  Evaluation
                                             E. Radha Krishnan, P.E.
                                                 William F.  Kemner
                                              Gopal Annamraju,  P.E.
                                                PEI Associates, Inc.
                                                  Cincinnati, Ohio
ABSTRACT
  Dust generated from electric arc furnaces (EAFs) employed
in steelmaking plants currently is listed as a hazardous waste.
Disposal of the waste at a controlled landfill is becoming expen-
sive as disposal sites become more scarce and remote from the
point of origin. Even with current standards of environmentally
acceptable disposal, long-term liability is a concern. This has led
the steelmaking industry to look to other viable options for han-
dling the dust.
  The dust from carbon and low-alloy steel plants represents a
potential source of heavy metals such as iron and zinc, the recov-
ery of which appears to  be  a logical alternative to disposal. In
addition to providing a dust management option,  material con-
servation would result from  this approach. Centralized recovery
facilities do not appear to be economically attractive for EAF
dust because it is generated in small quantities at a large number
of locations distant from industrial centers. Recycling to the
furnaces to achieve additional iron recovery and produce a zinc-
enriched blowdown dust stream for sale appears to represent the
most sensible route for managing the dust. This paper discusses
the approach  and presents results  of  an experimental program
for investigation of key issues pertaining to EAF dust pellet re-
cycling. These issues are the fate of zinc and other volatile ele-
ments in the dust, electrical energy consumption in the furnace
during recycling, effect on steel quality and the economics of re-
cycling. The research was conducted under  a grant from the U.S.
EPA's Small Business Innovation Research Program.

INTRODUCTION
  Each year  the  electric arc  steelmaking industry generates
approximately 650,000 tons  of dusts containing valuable metals
such as iron, zinc, lead and  chromium. These electric arc furn-
aces (EAFs) are becoming increasingly popular for the  produc-
tion of carbon and low-alloy steels and currently account for
about 35% of total steel production. Although only a relatively
small amount of dust is generated from the furnaces at  any one
steel plant, the cumulative quantity from all plants is large and
represents  a significant loss  of heavy metals and scarce  alloying
elements.
  More importantly, EAF dust currently is listed as a hazardous
waste  because  of the leachability  of  its toxic constituents; its
U.S. EPA-assigned hazardous waste number is KO61. Disposal
is becoming costly as disposal sites become more scarce and re-
mote from the point of  origin. Disposal costs  (including Jrans-
portation)  of $100 per ton of dust are quite common. Further-
more,  even with current standards of environmentally accep-
table disposal,  long-term liability is a  concern. This has led the
steelmaking industry to look to other viable options for handling
the dust.
  Alternatives to the landfilling option for EAF dust manage-
ment include chemical fixation, regional recovery processes and
on-site recycling. The major shortcomings of the various pro-
cesses suggested for  recycling metallic values from EAF dust
(e.g., SKF Plasmadust process, AMAX Caustic Leach—Elec-
trowin Zinc process) have been the high capital cost of the equip-
ment and the need  for large quantities of dust  for the processes
to be economical. Furthermore, the processes require dusts with
zinc contents greater than 20%. EAF dust generally is not amen-
able to these processes because it is generated in small quantities
at a large number  of locations distant from  industrial centers
where regional recycling might occur.
  Recycling to the furnaces is a logical alternative for managing
EAF dust. The recycling operation should result in the genera-
tion of a dust enriched in volatile elements such as zinc and lead;
this blowdown  dust stream  could be removed for  sale to zinc
smelters. The main advantages of the  recycling process are its
low capital cost, relative insensitivity to EAF  dust composition
and applicability to both minimills and large mills. This process
can be applied to carbon and low-alloy steel plants in both the
basic steel and steel foundry industries to recover contained min-
eral values in the dust and thus avoid costly disposal of this waste
in hazardous waste landfills.

Plant Experience in Recycling
  Several steel  plants including Bethlehem Steel,  Babcock  &
Wilcox,  Empire-Detroit Steel, LASCO  in Whitby, Ontario and
Armco at their  Kansas City  Works have conducted trials of re-
cycling with  either pellets (of dust and water)  or briquettes. On
balance, there appears to be  no clear consensus on the effects of
EAF dust recycling  in the steel industry. Although some dissatis-
faction with recycling exists,  it appears to offer  the most immed-
iate, lowest-cost alternative to disposal at this time.

Experimental Research Program
  Based on the results of a feasibility study by PEI and the exper-
ience of previous researchers, PEI designed an experimental pro-
gram to obtain data  under controlled conditions for investiga-
tion of several critical issues for the commercial application of re-
cycling: (1) the partitioning (upon recycling) of heavy metals such
as zinc,  lead and cadmium between the slag and the dust gener-
ated; (2) the effect  of recycling on energy  consumption; (3) the
effect on steel quality; and (4) the economics  of recycling. The
tests were conducted  at the Green  River Steel EAF shop  in
Owensboro, Kentucky, during the summer of 1985.
  Green River's Owensboro  site has two 65-ton electric arc furn-
aces rated at 24,000 kVA each (370 kVA/ton). The furnaces pro-
duce low-alloy and specialty carbon steels that are sold to  the
forging industry for use in the aerospace, bearings,  defense, en-
                                                                                               REUSE & RECOVERY    409

-------
 ergy exploration  and production, machine tool,  power trans-
 mission and  generation, mining  and transportation industries.
 Oxygen is injected with hand-held pipes;  lime is injected pneu-
 matically. Three to four charges of scrap are used per heat. The
 furnaces are equipped with  side-draft  hoods,  which semi-sur-
 round the electrodes and canopy hoods in the roof. Both furn-
 aces are vented to a  14-compartment baghouse that discharges
 into screw conveyors. These  conveyors  ultimately join into one
 cross conveyor that discharges into a  flexible tube. The dust then
 falls about 6 ft through the tube into a large sealed rolloff box
 situated on the ground. After being filled,  this box is hauled to a
 hazardous waste facility.

 EVALUATION OF PELLETIZING EQUIPMENT
   Prior to plant selection, fresh dust was  obtained from two oper-
 ating EAF shops representing low-alloy medium carbon steels
 (Plant A) and plain  carbon steels (Plant  B). Manufacturers  of
 pelletizing equipment were contacted for conducting pelletizing
 tests. (The agglomerating of EAF dust with water is often called
 "greenballing." In this paper, however, we refer to the product as
 pellets and the process as pelletizing.) Teledyne-Readco of York,
 Pennsylvania, was selected to run laboratory-scale and bench-
 scale tests on these dust samples.
   Based on the bench-scale test results, the following equipment
 was selected for conducting the pilot-plant studies:
 1.  3-ft-diameter disc pelletizer and hopper
 2.  8-in-diameter pin mixer
 3.  3-ft5 live bin feeder
 4.  Heavy-duty base supports for pelletizer, pin mixer and feeder

   Two feeder tubes were also used to provide the  flexibility  to
 bypass the  pin mixer. Even though  the bench-scale pelletizing
 tests did not warrant the use of a pin  mixer, one was included  to
 provide preslaking of lime in the event that this was required.
                                                                                      Table 1
                                                                   Elemental Concentration! In Electric Arc Furnace Dusts
                                                                                   (% by Weight)
          100
           90
           so
           70
           60
           50
           40

           30
           20
        5  n
            2 -
           I   I   I   I   I  I    I
          I STEEL PLANT A
          I STEEL PLANT B
                       _l	1	L   1
                                  _L
                                       _L
            1
            10   20  30 40 50 60 70  80   90  95  98 99
            CUMULATIVE PERCENT  LESS THAN  STATED DIAMETER

                          Figure 1
     Particle Size Distribution of Electric Arc Furnace Dusls from
                       Two Steel Plants
Element
Caa
ff"
Zn"
Pbb
Steel plant
A
9.17
16.7
?6.9
2.24
B
3.18
32.1
20.9
3.85
C
2.81
32.9
20.6
5.08
D
14.71
28.2
12.9
0.83
                                                             'Analyzed by ASTM Reference Method J682
                                                              Analyzed by ASTVt Reference Meihod J683
CHARACTERIZATION OF EAF DUSTS
   As part of the preliminary activities  for the experimental re-
search program, PEI conducted physical and chemical character-
ization tests for EAF dusts.
   PEI performed Bahco analyses for analyzing particle size dis-
tribution  of two EAF dusts (Fig. 1); the Bahco analyses were
performed  on  minus- 100-mesh  material,  which  represented
98.8% (by weight) of the sample for dust from Steel Plant A and
65.5% of  the sample for dust from  Steel Plant  B. The results
show that the EAF dusts were composed of  fine particles and
that approximately 50% of the dust was below 5 microns.
   Table 1 illustrates the variability of important metallic constit-
uents in the dust. The EAF dust analysis varies from heat to heat
and during a heat, depending on the materials charged, stage of
the heat and  other factors. Consequently,  the recycling  tech-
nique must be relatively insensitive to changes  in both the chem-
ical and physical characteristics of the dust.
   Some of the pellets produced during the bench-scale tests were
subjected  to weatherability tests. These pellets were kept in the
open for a period of 5 months. Ninety-rive percent of the pellets
retained their shape without  any  breakage. All  the  pellets re-
tained a strength of over 50 Ib. Although the weatherability tests
indicated reasonably good weathering characteristics without any
impact on pellet  strength or breakage of pellets, in an ongoing re-
cycling operation, pellets would not be stored for more than a few
days.
   Tests were conducted  to measure the temperature change of
EAF dust/water mixtures over time. The  dust contained 20.6%
lime.  For  one test,  the dust/water mix  was  formed into pellets
3/8 in. to 1/2 in. in diameter; for the other, the mix was left as a
solid  cake. The mixtures were held  in insulated cups and the
temperature was measured by imbedded thermocouples.
   Figure 2 shows the time-temperature curves for the pellets and
the solid cake. Maximum temperature reached in the  solid cake
was higher than in  the pellet test; however, the more  important
observation is that in both tests, temperatures increased quickly
at first and then  gradually rose to a maximum sometime after the
dust and water were mixed. The time required  to reach the max-
imum temperature was 80 and 105 min, respectively, for the solid
cake and pellets. The thermal mechanism  demonstrated in these
tests indicates the  lime  slaking and curing phenomenon taking
place.

PELLETIZING OPERATION
   Figure 3 illustrates the  pelletizing process flow diagram for the
Green River installation. The temporary storage hopper, which is
completely enclosed, receives the EAF dust from the baghouse.
The pelletizing operations were enclosed in a temporary shelter.
410
REUSE & RECOVERY

-------
                   120
                   no
                   100

             K-
             UJ

             I      90

             I


                    80
                                                                             SOLID CAKE
                                                                _L
                                                                80      100

                                                                TIME, minutes
                                                                                 120
                                                           Figure 2
                                        Results of Time-Temperature Tests on EAF Dust Pellets
The pellets produced were stored either in cardboard drums or
wooden boxes lined with polyethylene. All the pellets produced
were transferred and stored in  a covered  warehouse  before
charging into the furnaces. These arrangements met the neces-
sary regulatory requirements. Figure 4 shows the temporary en-
closure and the auxiliary facilities for the pelletizing system.
  The test pelletizer unit was a pilot model and, as such, was open
to allow adjustment of water spray or dust feed  locations. Dur-
ing pilot model operation, some dusting occurred around the ma-
chine; this would be circumvented on a production model as it
would be covered and vented to the baghouse.
  Dust composition appears to be the most important parameter
affecting pellet quality. Fortunately, variations in size (  < 1/4 in
to > 3/4 in.) did not affect the efficacy of recycling.
  The initial arrangement of the system included the pin mixer,
which was believed to offer the advantage of premixing water
with the dust to start the slaking of the lime before the dust en-
tered the pelletizer. After several trials, however, the pin mixer
became clogged with wetted dust,  which hindered the process;
therefore the mixer was removed. Even though the Green River
EAF dust varied from 10 to 20% lime, the pellet quality without
the pin mixer was not noticeably different from the pellet quality
with the pin mixer.
                        FLEXIBLE
                          TUBE
                                                           Figure 3
                                            Dust Pelletizing Process Flow Schematic for
                                                  Green River Steel Installation
                                                                                                   REUSE & RECOVERY    411

-------
                                        Temporary Weather Enclosure Over Pelletmng Equipment
                                            at Green River Steel EAF Shop. Owensboro, Ky.
                  Day Bin Constructed by Green River Steel
                           for fcAl Dust Storage
Auger Supplying Pelleti?er Feeder
                                                              Figure 4
                                               Pcllcimng Facility at Green River Steel Co.
412    REUSfc&KECOVhRY

-------
Pellet Quality
  Pellet strength tests were mostly qualitative, i.e., the pellets
were squeezed by hand. In some cases, however, a flat strip was
used to press the pellets on a scale, and the pellets showed break-
ing strength of 9 to 10 Ib (green strength). Pellets from the pelle-
tizer as made were strong enough to withstand shoveling and/or
dropping into the container.
  After aging for 24 hr, the pellets increased in strength. Tests on
three different batches (of different sizes) showed a crush strength
of 18 to 25 Ib. The  same batches showed a crush strength of
35 to > 80 Ib after 48 hr of aging,

RECYCLING OPERATION
  Transportation of pellets to the furnace and the timing of their
charging are important for a successful recycling program. Pref-
erably, the pellets should be charged into the furnace through the
use of a separate box or skip pan. They should be charged while
a slag layer is present on the molten metal from the meltdown of
scrap charged earlier. This facilitates quick assimilation of the
pellets into the bath and reduces flash off as dust. The scrap back-
charge can  take place immediately after the pellets are charged.
Another variation in charging consists of "sandwiching" the
pellets between the scrap in the charging bucket.
  At Green River, the recycling  operation involved moving the
drummed or crated pellets by a forklift to a warehouse area and
then to the melt shop floor, where they were dumped into a skip
pan. The skip pan was weighed and emptied into the furnace by
crane. The  pellets  were charged prior to the second  charge of
scrap (first  backcharge).  Figure 5 schematically illustrates both
methods of charging.

RESULTS
  Results  of pellet recycling  were analyzed  by comparison with
collected baseline data to determine:
   Effect on heat time and productivity
   Effect on power consumption
   Effect on yield
   Effect on quality of the steel
   Effect on carbon and electrode consumption
   Effect on lime and ferroalloy consumption

   The tests were conducted in two blocks. The main objective of
 the Block I tests was to determine the effect of recycling fresh
 pellets on power consumption and other furnace variables, where-
 as the primary objective of the Block II tests was to determine the
 fate of heavy metals under continuous recycling conditions. Table
 2 summarizes the data for the heats with pellet charging in Blocks
 I and II. Because the shop operation was not continuous during
 the experiments, some heats were made after  the furnace had sat
 idle for an entire turn, whereas others were made immediately
 after the previous heat. This situation affects apparent power use
 in kWh/ton and heat time regardless of pellet use. The heats are
 therefore divided into these  two categories  ("first heat,"  i.e.,
 cold furnace, versus "not first heat," i.e., continuous operation)
 for more  accurate analysis. Table 3 summarizes similar data for
 baseline heats, i.e., heats without pellet charging. The data for
 these heats were obtained largely from the plant's heat log sheets.
  During the Block  I test period, both furnaces operated 16 hr per
 day. Later, one of  the two furnaces was shut down for repairs
 and only one furnace remained in operation. During the night
 turn,  the  furnace roof  was  kept  open and  the  furnaces were
 allowed to cook down. The data collected during the Block I  and
 Block II tests include  a variety  of heats representing the normal
product mix. The operators paid no special  attention to, nor im-
plemented  any changes  in, the  steelmaking practice (other than
 adding extra carbon) because of the testing program.
  The Block II tests  were limited to 4 days of continuous re-
cycling because of  the  plant operating schedule.  During  this
period, pellets were made directly from the baghouse catch and
                                             PELLETS SANDWICHED
                                                 IN SCRAP
            PELLETIZER
                                                  PELLETS CONVEYED BY FORK LIFT
                         PELLETS HELD IN BASKETS
                              OR BOXES
                                                          Figure 5
                                      Schematic of EAF Dust Pelletizing/Recycling Operations
                                                                                                  REUSE & RECOVERY    413

-------
      TOTAL OBSBRVATIONS:
                                                                        Table 2
                                         Comparison of Data for Heals with Pellet Charging In Blocks I and II

                                  •ATS win PELLET  CHARGING  (nusT BAT ON FCTWACI Arm IMMMUM)

                                 14


N OF CASES
MINIMUM
MAXIMUM
MEAN
STANDARD DBV
TAPTAP
HRS
14
4.830
9.670
7.319
1.217
PWRUSBD
KWH/TON
14
536.960
670.740
616.919
40.442
TOTYIBLD
*
14
86.550
93.800
89.909
2.752
NETYIBLD
*
14
79.600
93.600
86.693
3.734
OPBNCRBN
*
14
0.070
0.471
0.270
0.120
FNALCRBN
*
14
0.150
0.480
0.362
0.113
COKBCHRQ
LBS
14
0.
3200.000
1132.143
799.425
PLLTCHHG
LBS
14
910.000
2180.000
1129.286
322.090
      TOTAL OBSBRVATIONS:    24
                                                 BEATS WITH PILOT CHARGING  (NOT FIBST BUT)


N OF CASES
MINIMUM
MAXIMUM
MB AN
STANDARD DBV
TAPTAP
HRS
24
4.500
8.000
5.872
0.953
PWRUSED
KWH/TON
24
493.830
672.500
560.903
37.725
TOTYIKLD
*
24
82.500
98.900
91.231
3.656
NBTYIELD
*
24
65.500
93.900
85.788
5.494
OPBNCRBN
*
24
0.050
0.860
0.349
0.203
FNALCRBN
*
24
0.160
0.470
0.338
0.112
COESCRRO
LBS
24
100.000
2000.000
1060.000
521.703
PLLTCHHO
LBS
24
100.000
2155.000
1083.542
328.019
Variable names are as follows:
TAPTAP       Tap-to-lap heal lime (in hours)
PWRUSED     Power used (in kilowall-hours, Ion).
TOTYIELD     Yield of sleel as tapped as percentage of lolal metallic charge including scrap.
                ferroalloys, and pellets.
NKTYIEl D     Yield of prime sieel as poured (in percentage)
OPENCRBN    Opening carbon m both (in percent).
I INAI l~RBN   Final tap carbon in both (in percent)
C'OKEC'HRCi    Coke charged (in pounds)
PI I UHRC.    Pellets charged (in poundi)
                                                                        Table 3
                                                        Comparisons of Data for Baseline Heats

                                             BASELINE BEATS  (FIRST BEAT ON FOHNACI  AFTO
       TOTAL OBSBRVATIONS:    28


N or CASES
MINIMUM
MAXIMUM
MB AN
STANDARD DBV
TAPTAP
HRS
28
4.670
9.083
6.777
1.211
PWRUSED
KWH/TON
28
483.220
766.090
592.898
74.942
TOTYIBLD
*
28
83.400
97.100
90.118
3.438
NBTYIBLD
X
28
73.000
96.600
87.172
4.799
OPBNCRBN
*
27
0.060
1.170
0.348
0.295
FNALCRBN
k
27
0.110
1.070
0.381
0.271
COOCHRG
LBS
28
0.
2500.000
1078.571
654.532
PLLTCHHG
LBS
28
0.
0.
0.
0.
                                                                BASILINS BEATS (NOT FIRST  BUT)
     TOTAL  OBSBRVATIONS:    35


N OF CASES
MINIMUM
MAXIMUM
MEAN
STANDARD DBV
TAPTAP
HRS
36
4.083
9.250
6.093
1.204
PWRUSBD
KWH/TON
35
466.440
634.440
540.882
37.959
TOTYIBLD
*
35
86.000
96.500
91.180
2.712
NBTYIBLD
«
35
71.000
96.600
87.397
4.290
OPBNCRBN
*
35
0.090
1.040
0.336
0.257
FNALCRBN
k
35
0.100
1.060
0.350
0.203
COKBCHRG
LBS
35
0.
2300.000
788.571
546.447
PLLTCHRG
LBS
35
0.
0.
0.
0.
 Variable names arc as follow1.:
 TAPTAP       Tap-lo-lap tieal lime (in hour1.).
 PWRUSfcD      Power used (in kilowati-hours/ion)
 TOTYIELD     Yield of sicel as lapped as percentage of lotal metallic charge including snap,
              ferroalloys, and pellets
NI TV 11-1 n     Yield of prime steel as poured (m percentage).
OPL-NCRBN    Opening carbon in Haiti (in percent).
KINAl CRBN   I'liml tap carbon in bath (in percent).
COKI-CHRG    tokc charged 
-------
                                                            Table 4
                                        EAF Dust Analysis in Block I and Block II Experiments
                                                         (% by weight)
Constituents
Phosphorus
Sulfur
Cadmium
Calcium oxide
Chromium (+3)
Chromium (+6)
Copper
Iron
Lead
Magnesium
Manganese
Molybdenum
Nickel
Potassium
Silica (SIO?)
Sodium
Vanadium
Zinc
Sample No., date (1985), and type"
!
06/12

N11
0.50
0.02
15.3
0.3!
Nil
0.1
18.4
0.76
1.87
3.0
0.14
0.17
0.8
1.8
0.6
<0.01
24.4
?
06/13

Nil
0.47
0.02
15.5
0.26
Nil
0.09
18.7
0.67
2.02
3.3
0.1
0.13
0.8
1.9
0.6
<0.01
23.0
3
06/14

Nil
0.44
0.02
15.5
0.23
Nil
0.08
18.2
0.70
1.96
3.3
0.8
0.12
0.9
1.6
0.6
<0.01
27.9
4
06/14
G.S

Nil
0.47
0.02
14.4
0.25
Nil
0.08
17.3
0.68
1.91
3.0
0.1
0.13
0.9
1.6
0.6
<0.01
26.9
5
06/14

Nil
0.47
0.02
13.9
0.22
N11
0.07
15.5
0.63
1.79
3.0
0.07
0.11
1.0
1.6
0.6
<0.01
25.0
6
06/17
Block I
Nil
0.43
0.0?
13.4
0.25
N11
0.07
16.7
0.56
1.88
3.0
0.09
0.12
0.8
1.5
0.5
<0.01
26.6
7
06/17
G.S

N11
0.47
0.02
13.8
0.26
Nil
0.08
16.1
0.52
1.87
3.1
0.10
0.11
0.9
1.6
0.5
<0.01
26.1
8
06/18

N11
0.55
0.02
14.3
0.32
Nil
0.09
16.9
0.47
1.79
3.2
0.12
0.13
1.0
1.8
0.6
<0.01
20.1
9
06/19
C

N11
1.0
0.01
17.5
1.8
0.01
0.08
21.1
0.59
1.76
2.6
0.05
0.14
0.7
3 2
0.5
<0.01
24.6
10
07/02
G

Nil
0.41
0.01
21.1
3.1
0.004
0.1
25.6
0.65
1.03
2.0
<0.01
0.16
0.3
4.1
0.4
<0.01
12.0
11
07/02
G

Nil
0.40
0.02
12.0
0.42
0.004
0.05
14.8
0.45
2.36
3.0
0.09
0.12
1.0
1 9
0.6
<0.01
32.7
12
08/26
C
13
08/27
C
14
08/28
C
15
08/29
C
Block II
0.03
0.50
0.01
19.0
0.58
0.002
0.09
24.4
0.66
3.63
5.0
0.14
0.29
1.0
3 1
0.5
<0.01
9.0
0.03
0.47
0.02
15.1
0.44
0.006
0.08
21.2
0.58
2.29
3.5
0.16
0.24
1.0
2 8
0.5
<0.0!
12.4
0.04
0.63
0.02
18.1
0.57
0.003
0.13
26.7
0.65
2.64
4.5
0.37
0.27
1.2
3 0
0.7
<0.01
12.3
0.02
0.58
0.02
16.1
0.62
0.002
0.1
28.5
0.65
2.77
4,4
0.33
0.27
1.1
3 l
0.7
<0.01
14.7
            aC = composite sample, G   grab sample, S = pellet size change.
recycled to the furnace. Two furnaces operated 16 hr per day
during this period. Slag and dust samples were collected to eval-
uate the fate of zinc.  Table 4 presents the analytical results for
dust/pellets, and Table 5 presents the results for slag analysis.
  A long shutdown occurred at the plant prior to the Block II ex-
periments. Because of uncertain  business conditions,  inventory
was reduced by consuming all available scrap in the plant during
the Block  II tests  without  regard  for a balanced scrap mix.  The
zinc content of this cleanup  scrap  was significantly lower than
that of the scrap usually used. No high-zinc scrap was observed
during the Block II testing; consequently, the zinc content of the
recycled dust was lower than in  Block I despite  the  recycling.
An increase in the zinc content of the  dust from 9% to approx-
imately 15%, however, was observed during the continuous re-
cycling period.
  Several area samplers with AA millipore membrane filters were
located in the shop to measure heavy metal concentrations in the
shop atmosphere—two at  5 ft above the ground level and two
above  the furnace level and just below the crane runway. Table
6 presents the analytical results for these samples. Although zinc,
lead and cadmium were all detected on the filters, the concentra-
tions were very low. These results indicate that general loss of zinc
and other heavy metals through fugitive concentrations is not a
significant factor in the zinc balance during recycling operations.
  During Block I testing, the  zinc content of the dust  was fairly
constant, varying between 20 and 25% based on composite sam-
ples. On the other hand, two grab samples taken within 30 min of
each other showed a marked variation in zinc content (12 and
32.7%).  The zinc content of the slag was less than 0.01% in al-
most all samples.
                                                            Table 5
                                         EAF Slag Analysis in Block I and Block II Experiments
                                                          (% by weight)
                                                   Sample No., date (1985), heat No., and type of sample
Constituents
A1203
CaO
FeO
MgO
MnO
SiO
Cr
Cu
Pb
Zn
1
06/26
69568
FS
2
06/26
69569
FS
3
06/27
69571
FS
4
06/27
69570
IF

7.21
41.2
6.87
10.7
9.36
16.8
1.22
0.01
<0.01
<0.01
7.79
41.0
3.96
15.3
4.49
17.1
0.29
0.04
<0.01
<0.01
7.48
42.4
3.25
11.6
4.98
20.3
0.48
0.02
<0.01
<0.01
4.96
41.7
18.1
10.6
6.15
7.32
0.67
0.02
<0.01
<0.01
5
07/01
69575
IF
6
07/02
69576
IS
7
07/02
69577
IS
Block I
4.99
40.3
15.4
6.86
6.69
13.9
0.87
O..01
<0.01
<0.01
5.47
39.8
11.0
9.31
7.60
17.1
1.13
<0.01
<0.01
<0.01
5.59
42.3
17.9
6.40
8.09
13.1
0.9
0.05
<0.01
<0.01
8
07/02
69578
IS

5.89
42.0
15.5
9.39
4.55
13.0
0.6
0.02
<0.01
<0.01
9
07/03
69579
IS

5.78
42.4
22.4
7.49
4.52
8.80
0.64
0.01
<0.01
<0.01
10
08/28
50028
IS
11
08/28
50028
IF
12
08/29
50030
IS
13
08/29
50031
IF
14
08/29
69616
IF
Block II
2.66
36.1
27.5
2.61
7.40
7.42
2.72
0.02
<0.01
<0.01
3.30
39.5
18.1
5.31
10.6
9.13
2.22
<0.01
<0.01
<0.01
3.93
31.1
39.5
5.89
5.49
5.86
1.07
0.02
<0.01
<0.01
4.22
33.0
26.7
5.81
8.86
10.3
1.67
0.02
<0.01
<0.01
5.54
36.2
18.4
5.99
7.68
15.1
1.25
0.01
<0.01
0.06
            = fjna) slag, IF = inside the furnace (in-site sample, IS = initial slag.
                                                                                                    REUSE & RECOVERY    415

-------
                                                             Table 6
                       Analysts of Contents of Area Samplers Located Inside the EAF Shop During Block II Experiments
Constituents
lotjl pirtk-
uljtfS
Clfolw
Ifld
Zinc
S«aplf ""»• Kite)
1* (8/26/85)
ug (gr)'
•2.0 < 10'
(0.1 « 1C'4)
<0.2
(«3.1 . IO'6)
<4.0
(•6.1 i IO"5)
7.7
(I.? . 10"4|
no/liter (gr/scf)f
•0.75
(O.J « 10"5)
.0.00075
(<3.3 • 10"')
-0.015
(.4.6 « IO'6)
0.026
11.3 . 10"5)
2b (S/28/85)
vt (gr)
1.71 i 10J
(4.2 . 10"')
.0.2
(O.I • 10"')
10.0
(1.55 • Ifl-'l
211
(1.26 • 10°)
vg/lllrr (gr/tcr)
3.49
(1.5 . I0'3|
•0.000)
(-1.14 • 10"')
0.013
(5.7 • 10"')
O.J6
(4.6 • IO"S)
Jc (8/2»/«S)
»g (gO
1.57 • IO3
(2.4 . 10'?)
•0.2
(0.1 i 10"')
• 4.0
(6.2 . IO"S)
2>l
(1.26 • IO"3)
i>9/llttr (gr/icf)
5.23
(2.J • 10"J)
.0.0007
(«2.» « 10'')
0.01}
(5.1 « 10"')
O.M
(I.I « 10"')
4* (t/n/ss)
i »g (gr)
r 3.4 > 10*
IS. 3 . 10'1)
•0.2
(O.I • I0"*|
.4.0
(.6.2 i 10'*)
60.7
(1.4 • 10"')
i)(/llt>r (gr/tcM
O.t
(3.S i 10'*)
.0.0005
(.2.1 . If'}
•a on
M.I a I0'*l
0.14
(6.) • IO"5)
         aSample I—Total volume. 267 I; duration, 133 min; location, 3 ft. above ground
         bSample 2—Total volume, 7701; duration. 390 min; location, above furnace and below crane girder
         'Sample 3—Total volume, 3001; duration, I SO min; location, above furnace and below crane firder.
         "^Sample 4—Total volume, 4Z5 1; duration, 210 min; location, 5 ft above ground.
         'Two types of uniu are indicated: micrograms (jig) and grains (gr).
          Two types of uniu are indicated: micrograms per liter (ug/liier) and grains per standard cubic foot (gr/tcf).
   Results of pellet charging on heat time are inconclusive. Heats
with pellets on a cold furnace took longer, while heats with pellets
during continuous operation were shorter. In neither case,  how-
ever, are these differences statistically significant.
   Power consumption for heats on a cold furnace was about 10%
higher than during continuous operation where pellets were used
or not. The use of pellets caused an increase in power use of about
4°7o; this was consistent for both cold furnace heats and contin-
uous operation, but the test of statistical significance is  conclu-
sive only for continuous operation. This increase in power con-
sumption is consistent with results reporttxj by some other inves-
tigators.
   Average total yield for the heats to which pellets were added did
not differ significantly from the base line heats (no pellets added).
It should be noted that several factors other than  pellet addition
affect yield. If all the iron contained in the pellets were recovered,
an improvement in yield of 0.2 to 0.3% could  be expected. Net
yield is not a significant  parameter in this analysis because it is
affected by physical  factors such as spillage, short ingots and off-
product unrelated to metallurgical yield.
   Green River produces quality steels in medium-carbon and low-
alloy grades suitable for forgings in several industries,  including
the aircraft industry. The quality and internal cleanliness of steel
are extremely important in meeting the stringent specifications for
this use. To assess the effect of recycling on steel quality, the plant
metallurgical department kept a special vigil and followed all the
heats charged with pellets from the ingot stage through  the dis-
patch of semifinished  products (bars, rods, etc.) to various cus-
tomers. Recycling had no deleterious effects on the steel quality.
This assessment is based on  38 heats (with pellets) in different
specifications.
   Although not apparent in Tables 2 and 3, comparison of aim
carbon and opening carbon data for  heats with pellets  and  base-
line data showed a higher percentage of heats using pellets were
closer to aim carbon; 28°/o of the baseline heats opened with car-
bon higher than aim carbon compared with 44% of the heats with
pellets. Because operators were concerned about carbon loss with
pellets, they compensated for this by making  carbon  additions
exceeding  the  theoretical requirements  for the pellet recycling
trials. Coke consumption during continuous operation was  about
33% (260 Ib) higher for heats with pellet charging although the
                                                            difference does not meet the test of statistical significance (at the
                                                            5% level). A large portion of this additional carbon was recovered
                                                            in the bath. Electrode consumption  data were collected for the
                                                            period when pellets were used in the Block I experiments and were
                                                            compared with the electrode consumption of the previous month.
                                                            Electrode consumption increased slightly during pellet recycling
                                                            (by about 0.26 Ib/net ton of steel produced).
                                                              In the Green River operations, pea size lime is blown into the
                                                            furnace pneumatically near the top of the furnace, above the bath
                                                            line. Lime is required to produce slag that assimilates the gangue
                                                            materials charged in the  furnace, thereby  maintaining the re-
                                                            quired basicity. Although not presented in Tables 2 and 3, com-
                                                            parison of the baseline data with the heats charged with pellets
                                                            showed a slight increase in the lime consumption when pellets  are
                                                            used. An average of 1.2 Ib of extra lime was consumed  per ton of
                                                            steel produced.
                                                              Comparison of ferroalloy consumption data did not indicate
                                                            any significant differences between heats with and without pellets
                                                            after adjusting for differences in aim specifications.
                                                              During the tests, the Green  River  baghouse operated normal-
                                                            ly.  Dust was discharged from it periodically; i.e., when sufficient
                                                            dust had accumulated in the hoppers, it was conveyed  to the day
                                                            bin for pelletizing. No abnormal baghouse operation was noticed.
                                                              Observations did  not reveal any noticeable change  in dust or
                                                            slag generation rates due to the addition of pellets.
                                                              In  summary,  the  observations and the results of the experi-
                                                            ments indicate that an acceptable quality of pellets can be made
                                                            by  pelletizing, and that recycling to the furnace can be  conducted
                                                            without any operational or air pollution problems. Recycling at
                                                            the rate of dust  generation had  no deleterious effects on steel
                                                            quality at Green River Steel.
                                                            ECONOMICS
                                                              Table 7 summarizes the design and operating parameters for
                                                            three EAF units of different sizes. Table 8 presents the capital and
                                                            annualized costs developed for each of these three units.  These
                                                            economic analyses are preliminary in nature and must be refined
                                                            for each specific plant site. When compared to landfilling costs of
                                                            SlOO/ton dust,  the  economic analysis indicates  recycling to be
                                                            attractive, especially for larger plants.
416
REUSE & RECOVERY

-------
                           Table 7
         Base Units for Economic Evaluation of Recycling
                           Table 8
   Capital and Annualized Costs for Selected EAF Recycling Units

EAF plant size (tons steel/yr)
Dust quantity processed (tons/yr)
Recycle plant capacity (tons/h)
Recycle plant operating rate (tons/h)
Operating hours per year
Onsite operators/supervisors
Units
Small
70,000
1,000
0.5
0.5
2,000
1
Medium
200,000
3,000
2.75
1.50
2,000
1
Large
600,000
9,000
6.5
4.5
2,000
2
CONCLUSIONS
  This research has proved that EAF dust can be pelletized with
water only and recycled to  the furnace. The following original
objectives underlying PEI's philosophy for EAF dust manage-
ment have largely been met:
• Use of a simple process design
• Use of existing, proven, low-cost equipment
• Use of EAF dust only, with no additives other than water
• Use of a process with no residual pollution streams
• Use of a process with a high turndown ratio and capable of
  processing widely varying dust compositions
  Thus far, PEI's experimental research has included an evalua-
tion of pelletizer performance  parameters and the effects of re-
cycling on electrical energy consumption, raw materials consump-
tion and steel quality. Future research programs should focus on
the longer-term verification of EAF operation with dust recycling
and the fine-tuning of the process economics. In the research con-
ducted, it was not possible to include sufficient heats to obtain the
desired amount of process data  to close the material balance.  The
objectives of longer-term testing are:
• To establish the material balance for continuous EAF dust re-
  cycling, with particular emphasis on  the partition of zinc be-
  tween various phases
• To determine the optimum rate and composition of the blow-
  down stream from recycling
• To finalize the economic  model  for the process to determine
  the net cost per ton of dust recycled

BIBLIOGRAPHY
 1. Babcock and Wilcox,  "Recycle of Electric Arc Furnace  Emission
   Control Dusts, Evaluation of  Two-Year Program," Tubular  Pro-
   ducts Division, Beaver Falls, PA, 1985.
 2. Barnard, P.O., et al., "Recycling of Steelmaking Dusts," U.S. De-
   partment of the Interior, Bureau of Mines, TPR 52,1972.
 3. Center for Metals Production, "Electric Arc Furnace Dust,  Dis-
   posal, Recycle and Recovery," Report No. 85-2, Project No.  RP-
   2570-1-2, 1985.
 4. Fosnacht, D.R., "Recycle of  Ferrous Steel Plant Fines, State-of-
   the-Art," I & SM, 1981, 22-26.
 5. Harris, M.M., "The Use of Steel Mill  Waste Solids in  Iron and
   Steelmaking," Presented at American Iron and Steel Institute  86th
   General Meeting, New York, NY, May 1978.
 6. Higley, L.W., Jr. and Fine, M.M., "Electric Furnace Steelmaking
   Dusts—A Zinc Raw Material," Bureau of Mines Report of Investi-
   gation 8209,1977.

Capital (Installed) Costs
Fixed capital costs
Working capital (151 FC)
Total capital
Annualized Costs
Manufacturing costs
Raw material
OiK labor (includes
onsite supervision I
clerical)
(20BO h/y/person » 16/h
burdened)
Utilities
Electricity
Water
OiH supplies (JO! of FC1)
Lab charges ($l/ton)
Total Manufacturing Costs
Fixed charges
Equipment payments
(5 yr loan t 141)
includes working capital
Local taxes & insurance
(3J of FCI)
PI art Overhead
(151 of mfg. costs)
Home Office Management
and Administration
(2S'« of mfg. costs)
Other
Credit for sale of high
zinc material (SlO/ton
dust)
Freight
Total Annualized Cost
Unit Cost for Dust
processed, i/tor

Costs per Operating Unit, $
Small
70.000
10,500
80.500


0
33.280
2,000
250
7,000
1,000
43,500

22,500
2,100
6,500
10,900

(10,000)
10,000
85, SOU
85.50
Medium
130,000
19.500
149.500


0
33.280
6,000
750
13,000
3,000
56,000

41,700
3.900
8.400
14.000

(30,000)
30,000
124,000
41.33
Large
200,000
30,000
230. 00&


0
66,560
18,000
2,250
20,000
9,000
115,800

64,220
6,000
17.370
28.950

(90,000)
90,000
232.500
25.83
 7. Higley, L.W. and Fukubayashi, H.H.,  "Method for Recovery of
   Zinc and Lead from Electric Furnace Steelmaking Dusts," Proc. of
   the Fourth Mineral  Waste Utilization Symp., Chicago,  IL,  May
   1974.
 8. Hoffman, A.O., et al., "Environmental Appraisal of Reclamation
   Processes for Steel Industry, Iron-Bearing Solid Waste," Presented
   at the Symposium on Iron and Steel Pollution Abatement Technol-
   ogy for 1980, Philadelphia, PA, Nov. 1980.
 9. Holley,  C.A. and Weidner, T.H., "New Process for Converting
   Steelmaking Fumes into Low-Zinc Pellets," Presented at  Chicago
   Regional Technical Meeting of American Iron and Steel Institute,
   Oct. 1969.
10. Holowaty, M.O., "A Process for  Recycling of Zinc-Bearing Steel-
   making Dusts," Presented at Chicago Regional Technical Meeting of
   American Iron and Steel Institute, Oct. 1971.
11. Keyser, N.H., et al., "Characterization, Recovery and Recycling of
   Electric Arc Furnace Dust," Presented at the Symposium of Iron and
   Steel Pollution Abatement Technology for 1981, Chicago,  IL, Oct.
   1981.
12. Krishnan, E.R.,  "Recovery of Heavy Metals from Electric Arc
   Furnace Steelmaking Dusts," Environ. Prog., 2, 1983, 184-187.
                                                                                                      REUSE & RECOVERY    417

-------
13.  Lehigh University, "Characterization, Recovery and Recycling of      15. PEI Associates, Inc.,  "Technical Assistance  to the Missouri De-
    Electric Arc Furnace Dust," Final Report, Bethlehem, PA, Prepared          partment of Natural Resources—Heavy Metals Recycling," Prepared
    for U.S. Department of Commerce, 1982.                                  for U.S. Environmental Protection Agency, Office of Solid Waste,
14.  Lynn, J.D., "Electric Furnace Fume Greenball Recyding-A Tech-          Washington, DC, 1982.
    nical and Economic Evaluation," Presented at  the Symposium on      16. PEI Associates, Inc., "Recycling of Dust from Electric Arc Furn-
    Iron & Steel Pollution Abatement Technology  for 1983, Chicago,          aces," Phase I  Final Report, Prepared for U.S. EPA, Office of Re-
    IL, Oct. 1983.                                                         search and Development, Washington, DC, 1984.
418    REUSE & RECOVERY

-------
               Trends  in Used  Oil  Composition  and  Management
                                                  Jacob  E. Beachey
                                             Franklin Associates, Ltd.
                                               Prairie Village,  Kansas
                                                  William L. Bider
                                                Trans World Airlines
                                               Kansas City, Missouri
ABSTRACT
  In 1983 Franklin Associates, Ltd. prepared a major technical
report  for the U.S. EPA's Office of Solid Waste. This report
developed baseline data on the composition and management of
used oil generated in the United States.
  Traditionally, a significant fraction of the used lubricating oils
generated in the  United  States  have  been environmentally
mismanaged.  Contaminated used oils with little or no processing
commonly have  been  used for  dust suppression  or burned in
boilers without pollution controls. As a result, the hazardous con-
taminants present in the oil have been dispersed into the environ-
ment, and health impacts may have occurred.
  The  U.S.  EPA has attempted to minimize the health  risks
associated with improper  used oil management by promulgating
regulations under the authority of RCRA and the Used Oil Re-
cycling Act. While the U.S. EPA has been developing and pro-
posing regulations for used oil, the quality of oil has been improv-
ing due to the decrease in the lead content of gasoline. This paper
examines the changes occurring and expected to occur in used oil
composition and management practices resulting from U.S. EPA
regulations and changes in lead levels in gasoline.

INTRODUCTION
  Over the past decade, there  has been widespread and growing
interest in used oil issues. Used oil management practices of the
late 1960s and early 1970s were criticized as being wasteful of a
valuable resource and harmful to the environment. In general, the
concerned parties, which included industry, government agencies
and environmentalists, contended that used oils could be handled
wisely or unwisely. Wise handling included reuse practices such as
re-refining and carefully controlled burning with energy  recovery.
Unwise handling included such wasteful practices as dumping and
even land disposal. Oiling roads with used oil to suppress dust was
considered more desirable than dumping or land disposal, but less
desirable than re-refining or burning as a fuel.
  Studies completed in the early 1970s showed  that more  than
50% of the generated used oil in  the United States was being
managed in what were generally considered  to be undesirable
ways.  Less than  10%  was  refined into new  lubricating  oil. A
larger  fraction  was burned as fuel,  but often without  the
necessary monitoring to assure that the public was not  being ex-
posed to any potentially  hazardous substances. Over  the  next
decade, additional studies indicated that most unwise manage-
ment practices continued to occur at significant levels.
  The  U.S. EPA has attempted to minimize the health  risks
associated with improper used oil management by promulgating
regulations under  the authority of  RCRA and the Used Oil Re-
cycling  Act. Final rules governing burning were issued on Nov. 29,
1985. On the same day, the U.S. EPA  proposed an overall
regulatory program that covers all aspects of used oil manage-
ment.   The  regulations   provide  generators,  collectors,
transporters, processors and end-users with requirements related
to storage, testing with a specification, recordkeeping and other
administrative requirements.
  While the U.S. EPA was developing and proposing legislation,
used oil quality was improving; it will continue  to improve in
1986. This change in waste oil quality is due not only to a decrease
in the lead content of gasoline, but also to behavioral changes in
anticipation of the new regulations. Since lead is one of the major
contaminants in used  oil  and  since most  lead comes from
gasoline, this change will influence used oil quality  significantly in
the future.
  This paper presents data on  baseline  used oil  contamination
levels and used  oil  flows  as developed for a  1983  Franklin
Associates, Ltd. (FAL) study for the U.S.  EPA Office of Solid
Waste.1 The 1985 regulations are then briefly  summarized  fol-
lowed by a discussion of the changes in oil composition and flows
that are expected to result from the new  regulations.

BASELINE CONDITIONS
  Two data bases were developed as part of the 1983 FAL study
for the U.S. EPA: one data base characterized  used oil flows
through the management system and  the other data base char-
acterized contaminant  levels in used oil.  The flow characteriza-
tion was accomplished  by a  literature search,  hundreds of
telephone interviews and 25 site visits to facilities involved in com-
mercial used oil management.
  Much of the information regarding the flow of used oil is un-
documented because of the unstructured nature of the used oil
management system. Federal regulations prior to 1985 did not re-
quire participants in the used oil industry to report their collection
procedures or reuse practices.
  The used oil composition characterization is based on a series
of fairly simple statistical parameters that characterize over 1,000
used oil samples. The concentrations of 19 potentially hazardous
constituents  were determined, including several  heavy metals,
chlorinated solvents, aromatic solvents,  polynuclear aromatics,
PCBs and total chlorine.

Used Oil Generation
  In 1983 approximately 1.2  billion  gallons of  used oil were
generated in  the United States.  The 1.2  billion gallons were
generated from the sale of over 2.3 billion gallons of new oil.  A
summary of the automotive and industrial used oil sales and  used
oil  generation is shown in Figure 1. Approximately 54% of
automotive oils enter the used oil market compared to about 48%
                                                                                              REUSE & RECOVERY   419

-------
                         NEW OIL SALES
Automotive -
Industrial -
Total
1,251
1.061
1731?
                                                                  USED OIL GENERATION
Automotive x
Industrial x
0.559 • 699
0.478 • 507
                                                                                            TOTAL GENERATION
                     Note: All values in million gallons
                                                           Figure 1
                                          Used Oil Generation in the United States in  1983
of industrial oils. Generation rates for various used oils take into
account losses due to leakage, spillage, combustion, disposal with
equipment  and  incorporation  into  finished products such  as
paint, putties, etc.

Used Oil Management System
  The used oil management system (UOMS) is comprised of com-
panies that collect, process and  sell used oil into several markets.
There are three basic types of companies involved in the industry:
(1) independent  collectors, who only collect and sell the oil; (2)
minor and major processors, who collect, process and sell an im-
proved product oil; and (3)  re-refiners, who collect, process and
sell a refined lube base stock. Many variations exist for each basic
type of company.
  Figure 2 shows the flow of used oil from the point of generation
through the management  system to reuse or disposal  for 1983.
Also included is a description of the reuse practices for generated
oil that does not flow through the system. This oil is  reused or
disposed of directly by the generators. Only 55^0 (669.1 million
gallons) of all generated used oil entered the used oil management
system. The remainder was reused, dumped or disposed of by the
generators (e.g., those who change their own automobile oil).
  The total number of  companies  involved  in  the used oil
management system in 1983 (excluding 500 to 1,000 independent
collectors) was 253. There  were only 13 companies involved in re-
refining used oil;  however, approximately 15% of the oil that
entered the  management system was re-refined. There were ap-
proximately 115 minor processors and  125 major processors.
Dispose! 40.8


Dinplng. (18.0 ^_ u.l
Burnt
166.8
Direr
193.9
27.1
AUTOHO
GENER/
532


TueT"5ales K.B
V.O.F.O. 15.8*

INDEPENDENT .., .

605.2
TIVE 317.4 	 "
TORS 	 \ 167.3
2 \
Fuel Use 55.2 V
Au'ttott. ft"1* >»•*: AT5V,»'-«
Flftt iQM» „
,7 	
^" IMDUST
GENERA
— *• 513.
fn-House
POS.I t.f, \rn.v i

*m "'•' /
9 FROM MINOR ; j

Fuel Use 37.1
Disposal WTr1
Roa"
	 	 -^^
V.O. EO. 41.6 ( "»"'
Hon FM! Inri 44 N. 590.1* _S
Roid Olllno *•• ,
Mej. Proe. /Re-re 4.4
On-Slte Fuel *•'
DIsposel 4.4
Fuel Seles '9.8
V.O.F.O. '".I
loed Oiling J.a u
In-SHe Fuel 9.3 (
ISPOiil *'J
Fuel Sllei 125.9
Ion-Fuel Ind. 6.8 *
In-Slte Fuel 11.0
Lulx Oil 6.5
Dlstlllete Fuel 0.7 '
IDISPOS.I 5.B '
Lube 011 56.2
Xitllleti Fuel 5.3 '
Msposi) 4.5

-^M&2.0
"^Cj_-
-»/"l5
*/^34.2
^^
-t^TTi
^/^68.S
QM.
                                                                                              VIRGIN OIL |  Fuel Seles 204.t
                                                                                              FUEL OEAURr
                                                            Figure 2
                                        Used Oil Flow Description in (he United States in 1983
 420    REUSE & RECOVERY

-------
  The fuel oil market clearly dominated with respect to end-use
applications. About 590 million gallons of used oil (nearly 50% of
all used oil entering the used oil management system) was burned
as fuel in  1983. Approximately one-half of the burned oil was
blended with  virgin fuel oil before being burned.  The industrial
market was and is the largest for burning;  however, significant
quantities also were burned in commercial, residential and institu-
tional boilers.
  Comparable quantities of used oil were re-refined and applied
to roads  for  dust control  in 1983. Approximately  83 million
gallons were received by re-refiners and nearly 74 million gallons
were used  for roal oiling. Approximately 70% of the oil received
by re-refiners  became new  lube stock, and the  remainder was
recovered as lighter distillates or was lost during processing.
  Approximately 35 million gallons of used oil were marketed for
industrial purposes other than fuel. The major application in this
category included flotation oils, asphalt extenders  and form oils.
  In 1983 over 400 million gallons of used oil, out of the total of
1.2 billion gallons generated, were disposed  of by  landfilling, in-
cineration  and dumping. Nearly 50% more oil was dumped than
was disposed  by landfilling  and incineration combined. Most of
the dumped oil is automotive oil generated by do-it-yourselfers
and large off-road equipment operators such as farmers, moving
companies, construction companies and the military.

Used Oil Composition
  Analytical data that quantify waste oil contamination in  1,071
samples were obtained  from  50 sources  for the baseline study.
The majority  of the results were obtained from state and federal
government agencies in unpublished form. Because of the variety
of data sources, there is considerable variation in: (1) the constitu-
ents that were measured; (2) the precision of the tests; and (3) the
detection limits of the equipment. These differences in technique
complicate any procedure used to combine the data  into  sum-
marized statistics.
  Table 1  summarizes used oil contamination with respect  to 19
specific constituents, 17 of which are included on the U.S. EPA's
               published list of hazardous constituents. The results are summar-
               ized according to percent detection (mean, median, 75th and 90th
               percentile concentrations), and range of measured concentrations.
                 The data in Table 1  indicate high levels of contamination for
               many constituents including lead, chromium, cadmium and sev-
               eral chlorinated  solvents.  The mean concentrations  are  badly
               distorted by a few very high values. The mean concentration for
               each contaminant is much higher than the median. PCBs were de-
               tected in approximately 18% of the samples analyzed.

                 The data in Table 1 have been categorized according to source
               and end use with the following generalizations:
               • Automotive used oils  tend to  have  higher concentrations  of
                 potentially hazardous  heavy  metals; industrial  oils tend  to
                 have higher levels of chlorinated solvents and PCBs. No signif-
                 icant differences were  noted in the concentration of aromatic
                 solvents or polynuclear aromatics.
               • Used oils from gasoline engines have much higher  lead con-
                 centrations than those  from diesel engines due to the presence
                 of leads in some gasoline additives.
               • Metalworking industrial oils have higher levels  of heavy metals
                 and chlorinated  solvents than  other industrial oils  including
                 hydraulic, compressor, turbine, electrical and others.
               • Approximately 30%  of the oil samples had a  measured flash
                 point below 140°F,* which is a criterion for classifying a  waste
                 as hazardous. Since the flash point  for virgin fuel oils is above
                 350 °F, contamination with low flash points such as gasoline or
                 chlorinated and organic solvents is  indicated.
               • Used oils do not appear to differ significantly in contamination
                 levels according to end-use markets. Similar levels of contam-
                 ination  were measured in  used oils  that  were burned,  road
                 oiled and re-refined.
               * The new used oil rules are somewhat more lenient. Used oil is subject to the full
               hazardous waste rules due to the ignitability characteristic only if the flash point is
               less than 100 °F (see Table 2).
                                                               Table 1
                                     Concentration of Potentially Hazardous Constituents in Waste Oil'
                                      Samples with
                                       Detected
                                      Contaminants
                                   Number
   Mean           Median
Concentration 21 Concentration 31
   (ppm)            (ppm)
Metal a
  Arsenic                   537        135        25        17.26
  Barium                    752        675        89       131.92
  Cadmium                   744        271        36         3.11
  Chromium                  756        592        78        27.97
  Lead 4/                   »35        760        91       664.5
  Zinc"                    810        799        98       580.28

Chlorinated Solvents
  Dichlorodlfluoromethane      87         51        58       373.27
  Trlchlorotrlfluoroethane      28         17        60     69,935.88
  1,1.1-Trlchloroethane       616        388        62      2,800.41
  Trlchloroethylene           608        259        42      1,387.63
  Tetrachloroethylene         599        352        58      1,420.89

Total Chlorine               590        568        96      4,995

Other Organics
  Benzene                   236        118        50       961.2
  Toluene                   242        198        81      2,200.48
  Xylenea                   235        194        82      3,385.54
  Benzo(a)anthracene           27         20        74        71.3
  Benzo(a)pyrene              65         38        58        24.55
  Naphthalene                 25         25       100       475.2
  PCBs                      753        142        18       108.51
                      5
                     48
                      3
                      6.5
                     240
                     480
                     20
                     160
                     200
                     100
                     106

                   1,600
                     20
                    380
                    550
                     12
                     10
                    330
                      5
Concentration
  at 75th
 Percentile 3_/
   (PP°0
       5
     120
       8
      12
     740
     872
     160
   1,300
   1,300
     200
     600

   4.000
     110
   1.400
   1,400
      30
      12
     560
      15
Concentration
  at 90th
 Percentile 31
   (ppm)
       18
      251
       10
       35
    1,200
    1,130
      640
   100,000
    3,500
      800
    1.600

    9,500
      300
    4,500
    3,200
       40
       16
      800
       50
   Concentration
       Range
  	(ppm)
 <0.01
  0
  0
  0
  0
 <0.5
<20
110
  0
                                          High
   100
  3,906
    57
   690
 21,700
  8,610
  2,200
550,000
110,000
 40,000
 32,000

 86,700
 55,000
 55,000
139,000
    660
    405
  1,400
  3.800
 1. Results determined for the analyses of 1,071 used oil samples.
 2. Calculated for detected concentrations only.
 3. For purposes of determining median  and percentile concentrations, undetected levels were
  assumed to be equal to the detection limit.
 4. Used samples collected after 1979 only.
                                                                Source: Reference 1.
                                                                                                         REUSE & RECOVERY     421

-------
U.S. EPA REGULATIONS FOR USED OIL
  In 1985 the U.S. EPA issued three proposals and one final rule
for the regulation of used oil. On Jan.  11, 1985, U.S.  EPA pro-
posed under Subtitle C of the RCRA regulations' that prohibit
the burning of non-industrial  boilers of used oil that does not
meet specification levels for certain hazardous contaminants and
flash point. Final rules were issued on Nov. 29, 1985.' The regula-
tions also provide  administrative controls to monitor  marketing
and burning activities.  These controls include recordkeeping re-
quirements and U.S. EPA notification of used  oil activities, in-
cluding an invoice system for  shipments.  The specifications  for
used oil that may be  burned  in  non-industrial  boilers without
regulation are listed in  Table 2.

                          Table 2
          Used Oil Furl Specification for Oil lhat May Be
                Burned in Non-Industrial Boilers
Constituent/Properly

Arsenic
Cadmium
Chromium
Lead
Total Halogens
Flash Point

Source: Rcfcrenct 3. r 49181
Allowable Level for
Burning without
Regulation

5 ppm maximum
2 ppm maximum
10 ppm maximum
100 ppm maximum
4,000 ppm maximum
100 °F minimum
   Also in  the Nov.  19,  1985  Federal Register,' the U.S. EPA
issued a proposed rule to establish standards for used oil that is
recycled, or "recycled oil."  These standards apply to generators
and transporters of recycled oil and owners and operators of used
oil recycling facilities. The standards would include tracking re-
quirements when  used oil is shipped off-site  for recycling and
facility management requirements when used oil is stored prior to
recycling. Recycled oil used as fuel would  be  subject to certain
regulations except that  fuel meeting the specifications listed in
Table 2 for toxic contaminants and flash points would be exempt
from regulations.  Uses of recycled oil  that constitute disposal
would be regulated as land  disposal,  and road oiling would  be
prohibited.
   The third  proposed rule,  also issued in the Nov.  29, 1985
Federal Register,'  would list used oil as a hazardous waste. The
effect of the listing, if promulgated, would be to control the treat-
ment and disposal of used oil (as well as its transportation, ac-
cumulation or storage prior to treatment or disposal) by subject-
ing it to full hazardous regulation under Subtitle C of RCRA. At
the same time, most used oil that is recycled would be subject to
the special management  standards of the previously mentioned
proposed standards for recycled oil.
   In the following sections, expected trends in  used oil  composi-
tion and management practices resulting from  the proposed and
final rules  are summarized.

CHANGES IN USED OIL COMPOSITION
   The burning regulations include specifications for metals, total
halogens and flash point for oil burned as fuel. Because of these
regulations, the concentrations of hazardous materials in used oil
will be reduced,  thus reducing the risk  to human health and the
environment.
   Most of the contaminants in used oil are the result of normal oil
usage. However,  significant quantities of  some contaminants
sometimes are introduced as a result of carelessness or intentional
mixing.
  Chlorinated solvents make up a major group of contaminants
in used oil. They are not a normal component of automotive oil
but often are introduced by careless practices. For  example,
automobile mechanics may pour solvents into tanks used primar-
ily for storing used automobile oils.
  Measuring total  chlorine  is  one  method  for assessing the
presence of chlorinated solvents. The 1983 baseline data indicate
typical chlorine concentrations range from 1,000 to 5,000 ppm.
However,  the total  chlorine content of some samples that were
identified  as  oil  were as  high as 40%. It is probable that these
materials were not oils at all, but rather solvents. It is expected
that  many industries,  when  faced with higher  disposal costs
because of mixing solvents with used oils, will modify these prac-
tices to keep oils and solvents segregated to the best of their ability
and therefore continue to manage the oil under the special used
oil rules rather than the full  hazardous waste rules.
  Most metals in used oil are the result of the oil's original use.
Metals concentrations in oil  can be  reduced only by blending.
Even  with blending at a 9:1  ratio,  many oils  will not meet
specifications.
  One of the major heavy metal contaminants in used oil is lead.
In the baseline study, the lead concentration in used oil ranged
from  0 to 21,700 ppm. The  high levels were most common in
crankcase  oils in samples taken in the early  1970s or from oils
taken from  leaded  gas-burning engines. Samples taken in the
1980s generally had  levels between 100 and 1,200 ppm. As stated
earlier, the major source of lead in automobile oil is the lead in
gasoline. As the  lead in gasoline continues to decrease (a tenfold
reduction  from 1983 is expected in early 1986), the  lead level in
used oil is expected  to decrease almost  proportionately.


CHANGES IN USED OIL FLOW
THROUGH THE WASTE MANAGEMENT SYSTEM
  Several  changes  are expected in used oil  flows through the
management  system. The U.S. EPA  has prepared a Regulatory
Impact Analysis (RIA) of its proposed standards for the manage-
ment of used oil.' The estimated total national annualized cost of
regulation is $168  million. The U.S.  EPA  predictions of how
these costs (and regulatory constraints) affect supply and demand
for used oil in different markets are shown in Table 3. A total in-
crease in  recycling  of  100  million  gallons is expected.  The
establishment of fuel standards is expected to shift recycled oil to
controlled  burners.  Although  the U.S. EPA study does  not
predict any burning  in hazardous waste incinerators, information
obtained  from industrial sources indicates that significant quan-
tities also may be managed in  this way, particularly oils recovered
from  treatment  facilities  and  various  types  of  oil/water
separators
  The use of used oil as a dust suppressant is banned by the pro-
posed rules. This displaced oil (currently 69 million gallons per
year)  is expected to  be used  as both re-refining feedstocks  and
fuel. Re-refining, which generally is considered a wise reuse prac-
tice by most segments of society, is expected to increase from 85
to 135 million gallons per year. There is, however, some question
about whether re-refiners can economically survive in a tight lube
oil market and with higher operating costs than the fuel oil pro-
cessors who compete for the  same oil.
  The RIA has estimated impacts on used oil generation, collec-
tion and  processor  facilities.  For industrial generators, used oil
management costs are usually very minor when compared to the
industry's production processes. It is expected that  used oil still
will be sold to collectors and processors but for a  lower price
because of the collector and  processor costs of regulatory com-
pliance. The generation from  industrial generators is not expected
to change significantly as a result of the regulations.
422    REUSE & RECOVERY

-------
  The impacts  of regulations  on  non-industrial (automotive)
generators are expected to be greater. The U.S. EPA has assumed
that since automotive oil  changes will cost more, more home-
owners will change their own oil. This impact could result in more
oil being dumped. The U.S. EPA estimates that full implementa-
tion of the rules will increase these homeowner oil changes by 12
million gallons per year.

                           Table 3
         Effect of Regulation on Market  Flows of Used Oil
                     (Million gallons per year)
                               Baseline

  Burning
       Industrial Boilers            249
       Asphalt and Cement Kilns      94
       Non-industrial Boilers        121
       Diesel Engines                15
       Space Heaters                 34
       On-site Boilers               _7J3
           Total Burned             586

  Re-refining
       Lube Oil                     59
       (total re-refined)            (85)

  Non-fuel Industrial                36
  Road  Oiling                       69
  Disposal                          405

       Total                     1,155


 Source: Reference 4, p. 49247.
Regulatory
  Impact
     185
     309
     117
      15
      34
      4J3
     708
     101
    (135)

     40
      0
     305

   1,155
  Collectors (particularly small collectors) are expected to ex-
perience the most significant changes in the industry structure
resulting from the U.S. EPA regulations. It is estimated that ap-
proximately 473 small and  medium-sized collectors will  find  it
uneconomical to continue operating as small independent busi-
nesses.  Although  these 473 collectors  represent  approximately
50% of the used oil recycling industry, they currently handle only
approximately 10%  of the  volume of oil entering the recycling
system. Initially, some oil may not be collected as facilities adjust
to the changes imposed by the regulations. In the long run oil will
be collected, but the nature  of the collection function is likely to
be very different after regulation.
  Some of the small collectors (an estimated 155) are expected to
grow or become part of larger businesses  and thus be able to ab-
sorb the costs of the U.S. EPA regulations.
  The  U.S. EPA  has optimistically estimated that the flow  of
used oil to re-refiners will grow by 59% as a result of the new
regulations.  This means the equivalent of six new facilities will be
opened (there are now 13).
  Minor and major processors are not likely to change drastically
as a result of the used oil regulations, although it is estimated that
12 out of the current 115 minor processors and two of the current
25 major processors may be involved in closures.
CONCLUSIONS
  The expected trends in used oil composition and management
following the promulgation of the 1985 proposed regulations are
summarized as follows:

• Reduction of lead in automotive oils as a result of the removal
  of lead from gasoline
• Reduction of chlorinated solvents in waste oil because of im-
  proved handling practices
• Reduction in the quantity of used oil burned in industrial and
  non-industrial boilers
• Increase of burning in asphalt plants and cement kilns
• Elimination of road oiling with used oil
• Large potential increase in re-refining of used oil
• Increase in  total used  oil burned  because of constraints on
  placing liquids in secure landfills
• Decrease in total disposal of used oil,  but with probable in-
  crease in dumping by do-it-yourself automobile oil changers

  The overall used  oil regulations and allowable lead level in
gasolines will result in a significantly better used oil management
system from an environmental perspective. Burning will be more
closely controlled, and the oils burned will  be of much  higher
quality.
  In the past,  generators were very careless and often purposely
mixed hazardous wastes into  oil. The new rules should modify
this behavior substantially, resulting in more oils that  are  con-
taminated only through normal use. With the planned lead reduc-
tion,  these levels of contamination are quite low compared to
earlier levels of contamination resulting from purposeful mixing.

REFERENCES
1. Franklin Associates,  Ltd., "Composition and Management of  Used
  Oil Generated in the  United States," prepared for the U.S. EPA Of-
  fice of Solid Waste, under Contracts No. 68-02-3173 and 68-01-6467,
  Nov. 1985.
2. U.S. EPA, "Hazardous Waste Management System; Standards for
  the Management of Specific Wastes and Specific Types of Facilities,"
  Proposed Rule  and  Request  for Comments,  Federal Register, 40
  CFR Part 266, 50, Jan. 11, 1985.
3. U.S.  EPA, "Hazardous Waste Management System;  Burning of
  Waste Fuel Used Oil Fuel in Boilers and Industrial Furnaces,"  Final
  Rule, Federal Register, 40 CFR Parts 261, 264, 265, 266 and 271,
  50, Nov. 29,  1985.
4. U.S. EPA, "Hazardous Waste Management System; Recycled  Used
  Oil Standards," Proposed Rule, Federal Register, 40 CFR Parts 260,
  261, 266, 270 and 271, 50, Nov. 29, 1985.
5. U.S. EPA, "Hazardous Waste Management System; General;  Iden-
  tification and Listing of Hazardous Waste; Used Oil,"  Proposed
  Rule, Federal Register, 40 CFR Parts 260, 261, 271 and 302, 50,  Nov.
  29, 1985.
6. Temple,  Barker &  Sloane, Inc., "Regulatory  Impact  Analysis  of
  Proposed Standards for the Management of Used Oil," prepared for
  the U.S.  EPA Office of Solid Waste, Washington, DC, Nov. 1985.
                                                                                                    REUSE & RECOVERY    423

-------
                             Used  Solvent Elimination Program

                                              Renato G. Decal, P.E.
                              Naval Energy and Environmental Support Activity
                                            Port  Hueneme,  California
ABSTRACT
  This paper will discuss the manner in which the Naval Energy
and Environmental Support Activity (NEESA) is supporting the
Navy in the implementation  of the Used Solvent  Elimination
(USE) Program.
  The Hazardous and Solid Waste Amendments of 1984 man-
date that industrial dischargers minimize hazardous waste gener-
ation at their facility. These amendments to RCRA provided an
impetus for launching the  Department of Defense (DoD)  USE
Program. The purpose of this program is to reduce the present
and  future risks  and cost associated  with the disposal of used
organic solvents by eliminating disposal as far as possible. This
program demonstrates DoD's conscientious efforts  to conserve
energy and protect the environment.

INTRODUCTION
  The Navy generates a  large amount of used oil and  solvents
(UO&S) in accomplishing its mission. As  an occupant of prime
coastal properties, the Navy is in a very precarious position of
maintaining a delicate balance between fleet readiness and pres-
ervation of environmentally sensitive  areas.  In  support of the
Department  of Defense's  (DoD) Solvent Elimination  System
(USE), the Navy  set a goal to  eliminate as far as possible the dis-
posal of used organic solvents. Means to achieve the goal include,
but are not limited to:

• Process changes
• Material substitution
• Recycling/reclamation

  The initial thrust of the Navy's USE Program is the recycling/
reclamation of solvents.  Process changes  and material substitu-
tion will be investigated further  under  the  Navy's  Hazardous
Waste Minimization Program. Recycle/reuse of these  by-pro-
ducts can reduce fuel consumption, save energy and provide mon-
etary savings to the Navy. Naval installations classified as major
generators of solvents (shipyards, ship repair facilities, air rework
facilities, etc.) have been identified and instructed to implement a
USE Program no later than Oct. 1, 1986.
  All other activities that generate over 400 gal of used solvents
per year must implement a USE Program by Oct. 1,  1987. Activ-
ities can receive technical assistance from their Engineering Field
Division (EFD) and Naval Energy and Environmental Support
Agency (NEESA) as they start their USE Program. This assis-
tance includes feasibility studies, equipment and training.
  Studies were initiated at six large industrial naval bases in Fis-
cal  Year (FY) 1985. In FY86, USE studies are scheduled to be
started at approximately 50 naval installations.
  NEESA studies include used oils since the U.S. EPA soon may
classify used oils as hazardous wastes and used oils often are
mixed with used solvents.
  The study focuses on developing a recycling program that is
environmentally sound, cost-effective, simple to implement, re-
quires little training, has low labor requirements and uses solvent
distillation equipment that is commercially available.

PLANNING
  Prior to initiating the UO&S study at an activity, an ad hoc
committee is organized to ensure maximum participation of per-
sonnel who will be directly affected by  this program. The  com-
mittee members will provide input on  which recycling alterna-
tives they prefer for implementation at their shops.  In addition,
the committee will be involved in the review of the interim reports
and decision-making process during the course of the study.

UO&S STUDY
  The UO&S recycling study consists of three steps:
• UO&S Inventory
• Evaluation of Alternatives
• UO&S Recycling Program Development

Step 1—UO&S Inventory
  A site visit is conducted to obtain the following information as
a minimum:

• Annual generation rates of UO&S in  gal/year
• Location (building number, ship type and hull number, etc.)
• Common name, chemical compound  name and MILSPEC of
  product
• Usage information (brief description of how UO&S are gen-
  erated)
• Probable contaminants and the corresponding chemical and
  physical properties
• Current disposal procedures
• Anticipated variations in above findings
• Purchasing and Disposal Records
• All UO&S in accordance with U.S. EPA, State and local regu-
  lations and requirements

  Previous studies related  to UO&S are reviewed and the rele-
vance of their findings to this effort indicated.
  In determining annual generation rates, maximum and  mini-
mum values as well as  average values are obtained.  Where aver-
age values are not available, as may be the case for solvent dis-
posal, estimates based on process  knowledge and characteristic
losses (i.e., solvent evaporative losses) are estimated.
  The information obtained is tabulated and reviewed by the ad
hoc committee for accuracy.
  It is preferable to conduct the inventory by visiting the  shops
and interviewing shop  personnel. This process  takes only  a few
minutes of shop personnel time and gives the investigator a first-
424   REUSE & RECOVERY

-------
hand knowledge of the operations. Questionnaires should not be
sent to shop personnel. More than likely, they will not be able to
provide all the  information needed.  Shop personnel should be
notified at least  1 month in advance of the site visit.


Step 2—Evaluation of Alternatives
  The key to a  successful recycling program is to keep it simple
and to maintain  the degree of segregation required.
  The most practical UO&S recycling program options are:
• On-site/Off-site reclamation of solvents
• Closed loop recycling of  high value hydraulics and lubricants
  with manufacturers
• Burning of appropriately segregated UO&S in Navy boilers
• Sale of UO&S through Defense Reutilization and Marketing
  Office (DRMO)

On-Site Recycling
  On-site recycling of used solvents located as close to the source
of generation as  possible usually will yield the most cost-effective
results. Accordingly, on-site recycling is emphasized as the pri-
mary USE program option.  On-site recycling of used solvents in-
volves use of distillation equipment (stills). There are many types
of stills commercially available in sizes ranging from 5 gal to over
200 gal. Stills are available which can be run on a batch or auto-
matic mode. In addition, some are explosion-proof and have op-
tional features for easy operation and maintenance.
  On-site recycling currently is being practiced at the  Norfolk
Naval Shipyard Paint Shop utilizing a small still for recycling used
paint thinners with impressive results. A savings of approximate-
ly $17,000/year  is realized due to reduced consumption of virgin
paint thinners and less volume of paint sludge for disposal. Pres-
ently,  10 stills are being ordered for four naval installations and
the Navy expects to procure more than 15 stills in FY87.

Off-Site Recycling
  Where on-site  recycling may not be cost-effective or may be dif-
ficult to implement, off-site recycling  by commercial contractors
offers a viable solution for recycling UO&S. As in the case of on-
site recycling, proper segregation of used solvents must be main-
tained. Costs for recycling by commercial contractors vary as a
function of virgin  material  costs, volumes, handling and trans-
portation costs and ease of recycling. Overall savings for naval in-
stallations of about 20-30%  can be achieved using reclaimed sol-
vents instead of virgin material for parts cleaning.
  There are some companies which provide equipment and sol-
vents for parts cleaning under a lease agreement. When the sol-
vent gets too dirty, it is replaced by the supplier. Because  the sol-
vent essentially is leased out to the customer, this type of opera-
tion becomes very attractive for smaller Navy  facilities because
the supplier assumes all responsibility for the solvent, including
ultimate disposal.  Presently, the Construction Equipment De-
partment at Naval Construction  Battalion  Center (CBC) Port
Hueneme is utilizing off-site recycling for Dry Cleaning Solvent.
Shop personnel are very satisfied with the operations.

Closed Loop Recycling
  Closed loop recycling of high value hydraulics and lubricants
with manufacturers is another recycling alternative worth inves-
tigating. This option may be employed when large volumes of
contaminated high value hydraulics and lubricants are being gen-
erated by an activity located near major industrial centers where
the manufacturers' processing facilities are available for reclama-
tion. Under this  option, the manufacturer contracts with the ac-
tivity to perform either of the following tasks:
 • Reclaim the used petroleum product and return it directly to
   the activity. A per gallon fee is charged which generally de-
   pends on the handling, transportation and processing costs.
 • Accept the  contaminated petroleum products and give the
   activity credits on its purchase of virgin materials. This option
   is not likely to be cost-effective for lubricants which have com-
   plex and diverse additive packages  and which  would require
   costly requalification testing.

Burning
   Burning of appropriately segregated UO&S in Navy boilers has
been practiced  at some  Naval  installations. CBC Port Hueneme
has burned contaminated oil at 5% blend with fuel oil #5 at the
Thompson Boiler Plant without any boiler modifications.
   Among the waste solvents being generated at Navy shops, only
PD-680 Type II is recommended for blending with fuel oils. Hal-
ogenated solvents must be kept segregated and must not be mixed
with used oils intended for use as boiler fuel. Halogenated  sol-
vents combine with  water to form a corrosive mixture that would
attack the boiler tubes and result in untimely breakdowns. Like-
wise, to avoid explosion, high flash solvents (having a flash point
greater than 140° F) must not be mixed with low flash solvents.
Sale
   Sale through DRMO is the least favorable option because, be-
fore this sale can be completed, it may go through a reuse, trans-
fer, donation and sale cycle which makes the UO&S available to
any federal, state or local government agency upon request and at
no economic benefit to the generating activity. Furthermore,
approval by knowledgeable personnel from a higher command is
required before this  option could be implemented.
   Recently, the DoD added an impetus to recycling used solvents.
The DoD directed naval activities to implement a  directive that
immediately prohibits landfill disposal  of certain  used  solvents.
This ban made  recycling of used solvents even  more  desirable
since  the amount of certain used solvents  to be ultimately  dis-
posed  of (by incineration) after recycling will be reduced by 70-
80%, thereby drastically reducing the disposal cost.
   The ad hoc committee is consulted to assist in identifying which
of the four alternatives provides a combination  of practicable
UO&S recycling programs that is in full compliance with applic-
able environmental regulations. For each recycling  alternative,
the following information is provided:
•  Present worth cost
•  Expected annual cost
•  Savings
•  Payback period
   Cost includes one-time  and recurring  costs associated  with
collection, transportation, segregation,  recycling and disposal of
non-recyclable UO&S. An interim Step 2 report incorporating the
above information is presented to the ad hoc committee for re-
view and concurrence.

Step 3—UO&S Recycling Program Development
   Once the ad hoc  committee decides on the preferred recycling
alternatives  for implementation, a recycling program is devel-
oped.  This program contains  instructions, detailed  operational
procedures and identification of key personnel to implement the
program. The UO&S recycling program consists of two parts.

Part I—Recycling
  The Part I program includes:
•  Identification of the UO&S sources with annual volumes by oil
  type or group.
                                                                                                   REUSE & RECOVERY    425

-------
• Identification  of the selected recycling  alternative  for  each
  group/source and the corresponding manpower requirements.
• Schedule for implementing the recycling program and delinea-
  tion of respective command responsibilities.
• Identification of interim practices/improved management pro-
  cedures that can be implemented immediately until the recom-
  mended recycling program can be fully implemented and re-
  quired facilities can be constructed.
• Identification of shop practices which will result in the reduc-
  tion of UO&S generation.

Part II—Recycling
  Part II of the recycling program delineates specific procedures
and  provides design parameters and specifications for necessary
facilities/equipment for the selected recycling alternative. In addi-
tion, a  description of the selected alternative for each group/
source is presented. This description  includes, but it not limited
to:

• A description of the source, recycling group and correspond-
  ing disposal program
• A description   of the responsibilities and identification of per-
  sonnel required to implement the selected recycling program; a
  description of the training program that will be required of the
  personnel who will implement the selected recycling program
• Procedures  for segregating UO&S  in accordance with respec-
  tive recycling alternative
• Procedures  for establishing  collection  points,  frequency  of
  collections and transfer centers where required
• Development of design parameters for facilities and/or equip-
  ment specifications for collection/transfer and recycling and its
  respective location
• Development of design criteria for special equipment required
  for  blending,  grouping  or  mixing chemically compatible
  UO&S. Establishment of procedures for blending,  handling
  and disposal of UO&S in accordance with the HSWA require-
  ments.
• Identification of existing collection, transfer,  treatment, dis-
  posal/reuse procedures that are  implementable within  the
  framework of the proposed program.
• Identification of disposal  procedures for UO&S sources that
  are not suitable for recycling
• Development of procedures or operational  improvements and
  providing design parameters and specifications  for  required
  support facilities or equipment for sources where reduction of
  UO&S generation is viable
  The completed UO&S Recycling Management  Program is re-
viewed and approved by the ad hoc committee.  Once  the pro-
gram is finalized, it is included in the base hazardous waste man-
agement  plan.  Instructions  for implementing this  program  are
outlined in the base instruction.

CONCLUSIONS
  NEESA UO&S studies showed that:
• Maintaining the degree of segregation required in UO&S is very
  important to ensure a successful recycling program.
• On-site recycling using  small capacity stills at the source of
  generation is the most cost-effective and simple way to recycle
  used solvents.
• For used oil, blending with virgin fuel oil and  burning in the
  boiler is practical.
• The estimated annual savings from recycling of UO&S at Navy
  and Marine Corps bases is $30,000-5200,000 depending upon
  the type of industrial operations.
• Recycling of UO&S has proven to be an effective program for
  conserving valuable resources and  preventing  environmental
  degradation without sacrificing the Navy's mission.

REFERENCES
I. NAVFAC UO&S Recycling Guide.
2. OPNAVINST5090.I.
3. "Utilization of Navy Generated Waste Oil as Boiler Fuel—Handbook
  of  Guidelines and Field Survey Results," NCEL Technical Report
  No. N-1674, Aug. 1983.
4. "Defense Fuel Supply Center Reference List of Commodities, Spec-
  ifications and Standards for  Petroleum and Related  Products," Oct.
  1982.
5. NAVFAC P-442, "Economic Analysis Handbook," June 1975.
6. "In-House Solvent Reclamation," NEESA 20.2-012, Oct. 1984.
7. "Guide for  Developing A  Recyclable  Materials Sales  Program,"
  NEESA 5-010.
8. "Assessment of Solvent Distillation Equipment," NEESA  20.3-012,
  Dec. 1985.
426    REUSE & RECPVERY

-------
                           Waste Reduction Audit Procedure—
                    A Methodology for Identification, Assessment
                   and  Screening of Waste Minimization  Options

                                              Carl H. Fromm, P.E.
                                            Michael S. Callahan, P.E.
                                     Hazardous & Toxic Materials Division
                                    Jacobs Engineering Group Incorporated
                                               Pasadena,  California
ABSTRACT
  This paper describes a waste reduction audit procedure, a man-
agement tool designed to generate a comprehensive set of waste
minimization options for a specific process and rank the options
to evolve those deserving either immediate implementation or an
in-depth evaluation. The procedure relies on a programmed sys-
tematic approach and is described in detail along with the actual
case example.
  Additionally, waste  minimization definitions are provided to-
gether with: a discussion on why waste occurs,  a characteriza-
tion of  factors inhibiting  waste minimization  activities and a
checklist to help identify waste reducing measures.

INTRODUCTION
  The Waste Reduction Audit Procedure (WRAP) described in
this paper  is a management tool intended to help  achieve  the
following objectives:
• To generate a comprehensive list of waste minimization meas-
  ures or options applicable to a specific industrial process
• To rank all identified waste reduction options and use these
  rankings to allow management to focus on options deserving
  further in-depth consideration
  The ultimate outcome of a WRAP is a report to management
that  contains  concrete  recommendations concerning which
options to pursue and/or implement.

BACKGROUND
  In 1981,  it was estimated that nearly 80% of all the hazardous
waste generated in the  United States ended up in landfills, pits or
ponds.1 Landfilling has been the most popular form of disposal
chiefly because of its low cost and simplicity. However, this situa-
tion has been changing  since the Federal government  adapted
The Hazardous and Solid Waste Amendments of 1984 to RCRA.
  The major goal of  the 1984 RCRA  Amendments is to min-
imize both the generation of hazardous waste and the practice of
land disposal. Achievement of this goal currently is being pro-
moted by at least two factors:
• Increased land disposal costs, which are the result of new regu-
  latory requirements for  mandatory groundwater monitoring,
  leachate  collection and treatment, installation of double liners
  and solidification of liquid wastes
• Waste minimization certification requirements demanding that
  generators of hazardous wastes certify that the volume or quan-
  tity and toxicity of the hazardous wastes they generate has been
  reduced to the degree economically practicable
  Land disposal costs have been increasing significantly and will
continue to do so. According to estimates over the last year, the
cost of landfilling drummed waste has increased more than 50%,
bulk solids 60 to 70% and bulk liquids 100%. The cost impacts of
waste  minimization certification requirements cannot be eval-
uated  easily since the requirements are recent (Sept.  1, 1985).
However, hazardous waste minimization already has been recog-
nized by many companies as a desirable goal, and there is strong
evidence indicating considerable interest, activity and progress in
this area.2

WORKING DEFINITIONS
  What is "waste minimization"? At the time of this writing,
the U.S. EPA had not issued a formal definition. In  lieu of a
formal definition, the following working definitions are used:
• Waste minimization—any source reduction, recycling or treat-
  ment activity  that reduces the volume or toxicity  (through
  means other than dilution) of any hazardous waste that is land
  disposed;
• Treatment  (as part of waste minimization)—any activity or
  series of activities that results  in reduction of the volume and/
  or toxicity  of the hazardous waste without attendant genera-
  tion of a valuable material that is subsequently employed in the
  manufacture of a commercial product (example—an incinera-
  tor for disposal  of spent chlorinated solvent with scrubbing
  and neutralization of hydrogen chloride from the flue gas);
• Recycling (as part of waste minimization)—any activity that re-
  duces the volume and/or toxicity of the hazardous waste with
  attendant generation of a valuable material subsequently em-
  ployed in the manufacture of commercial product (example—
  distillation  of spent cleaning solvent with subsequent reuse of
  the recovered clean solvent;
• Source reduction—any activity that reduces or eliminates the
  generation  of a hazardous waste within the industrial process
  (example—replacement of chlorinated solvent by a biodegrad-
  able, water-based  cleaning detergent solution).

  Of the three components of waste minimization, recycling and
source reduction can be subdivided further (Fig. 1). It is useful
to distinguish between reclamation and use/reuse as two separate
recycling activities.
  Source reduction,  the most prominent component  of waste
minimization, consists of product  substitution and source con-
trol. Product substitution means the replacement of an original
product and use of another product suitable for the same end use,
or the alteration of an original product use which results in a de-
crease or elimination of hazardous waste generation associated
                                                                               WASTE MINIMIZATION PROGRAMS   427

-------
                                                       MINIMIZATION OF HAZARDOUS WASTE  |




1
| RECYCLING | [SOLRCE RHXJCT1CN|


1 1 1
RECLAMATION
| |-»OR
•T |
ti pt
USE/REUSE
T?"'
TT>,
1 — OR
T
LEGEND
1 - PROCESS
1 — WASTE TREATMENT INIT
1 	 ' lEfOOFPIPE)
R — RAW MATERIAL
W — WASTE
U -PRIMARY. 2-SECOfOARY 1
| 	 | - WASTE TO DISPOSAL
»OTE5
1. «, < W. FOR BOTH VOLUME AMI/OR TOX1CITY


TOXCT SUBSTITUTION
l.USf OF COCRETE
PlLlNCS INSTEAD
OF TREATED TOO
2 HERACEitNT OF
SrNT
-------
                                                                    Table 1
                                      Summary °f Waste Generation Origins, Causes and Controlling Factors
               Waste  Origin
                       Typical Causes
                                                            Operational Factors
                                Design Factors
               Chemical Reaction
                     • Incomplete reactant conversion
                     • Byproduct formation
                     • Spent catalyst - deactivation
                       due to poisoning, sintering, etc.
• Inadequate temp, control
• Inadequate mixing
• Poor feed flow control
• Poor feed purity control
• Inadequate reactor design.
• Catalyst design or selection
• Choice of  process path
• Choice of  reaction  conditions
• Fast quench
• Inadequate instrumentation or
  controls design
• Poor heat  transfer
               Contact Between
               Aqueous and
               Organic Phases
                       Vacuum production via steam jets    • Indiscriminate use of water    • Choice of process route
               Disposal of Unusable
               Raw Materials and
               Offspec Products
                     • Presence of water as a reaction
                       byproduct
                     • Use of water for product rinse
                     • Equipment cleaning
                     • Cleaning of spills

                     • Offspec product generation
                       caused by contamination,
                       temperature/pressure excursions,
                       improper reactants proportioning,
                       inadequate pre-cleaning of
                       equipment or workpiece
                     • Obsolete raw material inventories
  for cleaning or washing
• Poor operator training
  and supervision
• Inadequate quality control
• Inadequate production
  planning and inventorizing
  of feedstocks
• Vacuum production via vacuum
  pumps
• Use of  reboilers instead of
  steam stripping
• Clingage reduction

• Extensive use of automation
• Maximizing dedication of
  equipment to a single process
  function
               Process Equipment    • presence of clingage (process side)   • insufficient drainage prior     • design for lower film temp in
               Cleaning
Metal Parts
Cleaning
                       and scale (cooling water side)
                     • deposit formation •
                     • use of filter aids
                     • use of chemical cleaning agents
                                    • disposal of spent solvent,
                                      cleaning sludge or spent
                                      cleaning solution
  to cleaning
• inadequate cooling water
  treatment
• excessive cooling water
  temperatures

• indiscriminate use of
  solvents and water
  heat exchangers
• design of reactors or mixing
  tanks with wall wiper blades
• controls to prevent cleaning
  water from overheating

• choice between cold dip tank
  or vapor degreasing
• choice between solvent vs.
  aqueous cleaning solution
               Metal Surface
               Treatment
                     • dragout
                     • disposal of spent treatment
                       solutions
• poor rack maintenance        • counter-current rinsing
• indiscriminate rinsing with    • fog rinsing
  water                       • dragout collection tanks
• fast withdrawal of workpiece
               Spills and Leaks
               Cleanup
                     • manual material transfer
                       and handling operations
                     • leaking pump seals
                     • leaking flange gaskets
• inadequate maintenance
• poor operator training
• lack of operator attention
• indiscriminate use of
  water in cleaning
• choice of gasketing material
• choice of seals
• use of welded or seat-welded
  construction.
• Lack of awareness of the benefits of waste minimization
• Lack of technical staff
• A "hands off the process" attitude caused by fear of upsetting
  a product's quality
• Organizational inertia (for example,  an "if it  isn't broke—
  don't fit it" attitude)
                                                           • Internal politics  of the organization  (for  example, an inno-
                                                             vator may feel inhibited by a fear of the lack of management's
                                                             support)
                                                           • An "it  can't be done" attitude; people may reject an innova-
                                                             tive approach merely because it  is outside their  range of ex-
                                                             perience
                                                                                                WASTE MINIMIZATION PROGRAMS    429

-------
   In implementing a WRAP, the above inhibiting factors  al-
 ways must be recognized, since  their identification and recog-
 nition allows formulation of constructive approaches to over-
 come them.

HOW TO IMPLEMENT A WASTE REDUCTION
AUDIT PROCEDURE
  The WRAP developed by the authors of this paper consists of
eight sequential steps:
• Selection of the audit team
• Compilation by the audit team of a waste stream list for the
  facility with the associated flowrates
• Generation by the audit  team of waste reduction  options for
  each waste stream
• Ranking by the audit team of each compiled  option in three
  categories: effectiveness, extent of current use and  applica-
  tion potential
• Preparation by the audit team of documentation in support
  of selected options
• Presentation, discussion and joint review with  plant personnel
  of options and their rankings
• Analysis by the audit team of revised rankings
• Final report preparation
  The above procedure is  applicable to all three categories of
waste minimization (treatment, recycling and source reduction).
However, it originally was developed and tested for source reduc-
tion options only. Source reduction measures should be consid-
ered even  when  recycling  or treatment  options are given prior-
ity. This is because reducing the quantities of waste  that are re-
cycled or treated often means an increase in revenues (e.g., due to
an increase of product yield and lower cost of treatment).

Selection of the Audit Team
  To effectively conduct a WRAP, it is necessary first to desig-
nate a leader for the audit. This person should have solid  tech-
nical credentials and demonstrated problem-solving ability and
can be either from within the company being audited  or from the
outside consulting firm conducting the audit. The leader and his
team (audit team) preferably should not have any previous exten-
sive association with the plant to be audited, primarily  to avoid
biases and to  bring an independent  perspective into analysis.
Management must  give the audit team  enough authority to gain
access to all required technical documentation (process flow dia-
grams,  piping and instrumentation  diagrams,  inventory  logs,
operations logs, etc.) and technical personnel within the plant.

Waste Stream List
  The second step consists of compiling a list of all waste streams
leaving the plant. It may prove beneficial to include not only haz-
ardous wastes destined for landfilling, but also  other emissions
such as  wastewater, air emissions and  non-hazardous  wastes.
Waste stream identification and quantification may require exten-
sive site inspection,  review  of process documentation, interviews
with plant operations personnel and, in some cases, material bal-
ance calculations.
  Reviews of process flow  diagrams, heat and material  balances
and piping and instrumentation diagrams are particularly useful.
Each diagram should be analyzed for all points where waste gen-
eration can occur. When producing a chemical commodity, typi-
cal sources include reaction byproducts leaving the process as dis-
tillation column bottoms. Also, equipment cleaning wastes (e.g.,
wastes from storage tanks and heat exchangers) must be con-
sidered,  although they are usually minor in a continuous pro-
cess.  In a  batch  process, large  quantities of waste can  be  asso-
                                                             ciated with cleaning operations, such as cleaning out a mix tank
                                                             or reactor. Since these incidental wastes rarely appear on process
                                                             flow diagrams or are mentioned  in  process descriptions,  dis-
                                                             cussions with key operations personnel are important.
                                                               Waste streams must be quantified  on a uniform basis.  If a
                                                             stream is intermittent, it should be represented as a pseudo-con-
                                                             tinuous stream.  After all waste streams are quantified, it is  use-
                                                             ful to express each of them as a percentage of the total quantity
                                                             of waste leaving the process.

                                                             Waste Reduction Options
                                                               The third step in WRAP involves the generation of waste re-
                                                             duction options  for each stream previously identified. Good re-
                                                             sults can be  obtained in  the course of a brainstorming  session
                                                             preceded by  independent preparation  by each participant. This
                                                             session must  involve only the outside audit team — no plant per-
                                                             sonnel  must be present. The reasons for exclusion of plant per-
                                                             sonnel at this point include:
                                                             • Working through the options generation process without plant
                                                               personnel forces the audit team to work harder to gain a good
                                                               understanding of the process and analyze it with a long-vision
                                                               perspective — which is often absent among the plant operators.
                                                             • It is necessary, at this  stage,  to postulate a  maximum  number
                                                               of options.  The  presence of  plant personnel may inhibit peo-
                                                               ple from submitting an option to the list.
                                                               Tables 1 to 4, developed by the authors of this paper, are pro-
                                                             vided to stimulate the discussion process to identify possible use-
                                                             ful waste reduction measures. Tables 2 to 4 contain checklists of
                                                             waste reduction measures compiled based on many process analy-
                                                             ses and on experience with actual industrial WRAP applications.

                                                                                        Table 2
                                                                    Waste Reduction Methodology Checklist: All Processes
                                                             All Waste Streams
 6.
 7.
 8.
 9.
10.
11.
12.
                                                                                     Use higher purity materials
                                                                                     Use less toxic raw materials
                                                                                     Use non-corrosive materials
                                                                                     Convert from batch to continuous process
                                                                                     Tighter equipment inspection and maintenance
                                                                                     Better operator training
                                                                                     Closer supervision
                                                                                     Practice good housekeeping
                                                                                     Eliminate or reduce water use for spill cleanup
                                                                                     Implement proper equipment cleaning techniques
                                                                                     Use improved monitoring s> stems
                                                                                     Use pumps with double mechanical seals
                                                             Commodities Produced Continuously

                                                             Examples:            Aery lonitrile, Epichlorohydrin, Petroleum Refining,
                                                                                 1,1,1 Trichloroethane, Trichloroelhylene/Perchloro-
                                                                                 ethylene, Vinyl Chloride Monomer
                                                             Heav> and Light Ends
                                                             Spent and Lost
                                                             Catalyst


                                                             Equipment Cleaning
                                                             Wastes
                                                             Leaks and Spills
 1.   Develop more selective catalyst
 2.   Optimize the reaction variables/reactor design
 3.   Use alternate process routes
 4.   Combust with heat (and HCD recovery

 1.   Develop tougher catalyst support
 2.   Use filter inside reactor freeboard
 3.   Regenerate and recycle spent catalyst

 1.   Increase equipment drainage time
 2.   Use corrosion resistant materials
 3.   Agitate and/or  insulate storage tanks
 4.   Re-examine need for chemical cleaning
 5.   Use nitrogen blanket to reduce oxidation
 6.   Use in-process HX cleaning devices

 1.   Use bellow-sealed valves
 2.   Use canned (seal-less) pumps
 3.   Maximize use of welded vs. flanged pipe joints
430
WASTE MINIMIZATION PROGRAMS

-------
                             Table 3
    Waste Reduction Methodology Checklist: Commodities Produced
                            in Batches
 Examples:
                 Dyes, Inorganic Pigments, Paint, Agricultural Chemicals
                 Formulation, Phenolic Resins, Wood Preserving
Material Handling     1.
                    2.
                    3.
                    4.
                    5.

Reaction/Processing   1.
Step                 2.
                    3.

Filtration and        1.
Washing             2.
                    3.
                    4.
                    5.

                    1.
                    2.
                    3.
Baghouse Fines
Off-Spec Product     1.
                    2.

Equipment Cleaning   1.
                    2.
                    3.
                    4.
                    5.
                    6.
                    7.
                    8.
                    9.
Leaks and Spills
                    1.
                    2.
                    3.
                    4.
Segregate containers by prior contents
Use rinseable/recyclable drums
Purchase materials in bulk or in larger containers
Purchase materials in preweighed packages
Use pipeline for intermediate transfer

Optimize the reaction variables/reactor design
Optimize the reactant addition method
Eliminate the use of toxic catalysts

Employ efficient washing/rinsing methods
Eliminate the use of filter aids
Use countercurrent washing
Recycle spent washwater
Maximize sludge dewatering

Increase use of dust suppression methods
Use wet instead of dry grinding
Schedule baghouse emptying

Tighter control of reaction temperature
Reformulation of off-spec product

Install high pressure spray wash system
Alter production schedule
Use mechanical wipers on mix tanks
Clean mix tanks immediately after use
Use a countercurrent rinse sequence
Recycle spent rinse water
Increase spent rinse settling time
Re-examine need for chemical cleaning
Dewater spent rinse sludge

Use bellow-sealed valves
Install spill basins
Use canned (seal-less) pumps
Maximize use of  welded vs. flanged pipe joints
 Option Rating
   The fourth step involves rating, by the audit team, of each op-
 tion generated in term of three categories (effectiveness,  extent of
 current use and future application potential).  While the meaning
 of "effectiveness" is fairly obvious, "extent of current use" and
 "future application potential" need more discussion. The extent
 of current use can be based on either the number of waste streams
 that are controlled using a given method or the frequency of use
 of a given method on a particular waste stream. The future appli-
 cation potential is  a measure  of the probability that  a  given
 method will be implemented in the future. This variable is highly
 dependent on capital and operating cost, level of difficulty,  im-
 plementation period,  product  yield  credits,  technological risk
 and the risk of detrimental effects on product quality.
   In the course of developing and using this  audit  approach, it
 was convenient to ask people to  rank each item on a scale of zero
 to four or zero to ten. Usually, the scale chosen will depend on
 the people conducting the WRAP and the extent of their knowl-
 edge.  When people are asked to what  extent a given method cur-
 rently is implemented, it may be difficult for them to come up
 with an  answer.  However, when  asked  to rank  these  same
 methods  on a scale of zero to four where the inference is from
 low to high,  answers often are more forthcoming. Once each
 method is ranked, the amount of waste currently being minimized
 and the  potential  future  reduction   index  can  be  calculated
 according to  the equations presented in the Appendix of this
 paper. In addition, one or more waste reduction methods should
 be singled  out for further study and implementation for each
 waste stream.
                                                                                                    Table 4
                                                                         Waste Reduction Methodology Checklist: Manufacturing Operations

                                                                       Examples:          Electroplating, Lithographic Printing, Metal Parts
                                                                                          Cleaning, Metal Surface Treatment, Paint Application
                                                                                          Printed Circuit Boards
Material Handling    1.   Segregate containers by prior contents
                   2.   Use rinsable/recyclable drums
                   3.   Purchase materials in bulk or in larger containers
                   4.   Purchase materials in preweighed packages

Solvent Cleaners     1.   Install/operate cleaning tanks properly
                   2.   Avoid cross-contamination of solvent
                   3.   Avoid water contamination of solvent
                   4,   Remove sludge continuously
                   5.   Monitor solvent composition
                   6.   Consolidate cold cleaning operations
                   7.   Recycle spent solvent
                   8.   Use plastic bead blasting for  paint stripping

Alkaline/Acid       1.   Install/operate cleaning tanks properly
Cleaners            2.   Avoid cross-contamination of solvent
                   3.   Remove sludge frequently

Plating/Etching/     1.   Increase plating solution bath life
Surface Finishing    2.   Use lower concentration plating bath
Solutions           3.   Use trivalent Cr in place of hexavalent Cr
                   4.   Use non-cyanide plating solutions
                   5.   Use in-line recovery techniques
                   6.   Regenerate spent bath solutions
                   7.   Segregate all waste streams
                   8.   Inspect all parts for proper cleanliness

Rinse Water         1.   Install/operate all rinse tanks properly
                   2.   Use multiple rinse tanks
                   3.   Install drain boards and drip tanks
                   4.   Use fog nozzles and spray units
                   5.   Agitate rinse bath
                   6.   Use deionized water for rinsing
                   7.   Recycle and reuse rinse water
                   8.   Segregate all waste streams
                   9.   Reclaim metal from  rinse water

Paint Application    1.   Use equipment with low  overspray
                   2.   Inspect all parts before painting

Leaks and Spills      1.   Install splash guards and drip  boards
                   2.   Prevent tank overflow
                                                                       Audit Documentation
                                                                         The fifth step involves preparation by the audit team of any
                                                                       documents outlining the concept, rationale and intended imple-
                                                                       mentation of options which are  not obvious and hence require
                                                                       explanation. This may involve a block flow diagram, a proposed
                                                                       revision of a piping and instrument diagram, an outline of pro-
                                                                       posed new equipment, a narrative description, etc.

                                                                       Presentation to Plant Personnel
                                                                         The sixth  step  is most crucial, as it involves presentation  of
                                                                       the options evolved to the plant personnel. The chief objective of
                                                                       this session is to stimulate the innovative process within the entire
                                                                       group  and not simply to have the plant people select from the
                                                                       menu prepared by the  audit team.  For that reason, the  audit
                                                                       leader must be firmly in control of the meeting; after all, innova-
                                                                       tion is intended to occur through an interactive process that may
                                                                       include controlled conflict. An atmosphere of conflict, or at least
                                                                       polarization, is likely to exist between the members of the audit
                                                                       team and plant personnel as each proposed option is subjected to
                                                                       review. New rankings very often will be evolved, and alternative
                                                                       options to those previously compiled often will emerge. The func-
                                                                       tion of the audit leader  is to moderate the discussion, to provide
                                                                       positive  resolution of  conflicting  opinions (e.g.,  subject  the
                                                                       option to ranking by vote) and  to  keep the discussion focused
                                                                       on the subject. Prior to  the meeting,  plant personnel  must re-
                                                                                           WASTE MINIMIZATION PROGRAMS     431

-------
view and analyze the postulated options. However, the meeting is
the culmination of the entire process. At the end of the meeting,
the leader should undertake a review  of the new rankings ob-
tained for each option now appearing on the list. Any residual
disagreements should be resolved at that time.

Analysis Re-Ranking
  The seventh step involves the re-analysis of the  rankings to
evolve new waste reduction indices for each option as described
in the Appendix.

Future Reduction Potential
  The eighth and the final step involves review of "future reduc-
tion potential" indices, options screening and preparation of the
final report in which  the WRAP results are presented to com-
pany management. The report and any recommendations must be
reviewed and concurred with by plant personnel.

A WRAP CASE STUDY
  WRAP was applied to the batch process partially depicted in
Figure 2 with the purpose of developing a process or equipment
configuration minimizing emissions  of toxic powders  into the
workspace environment. The release  of these powders,  apart
from presenting an exposure problem to the operators, ultimately
results in hazardous waste generation. The waste is generated
when the work area is hosed down and the  wastewater is sent to
a clarifier in which a toxic sludge settles  out.
  The part  of the process described involves weighing, mixing
and dispersion of two toxic powders (Powders A and B) in the wet
grinding unit. At the end of the wet grind operation, the slurry is
poured into drying trays, which are placed into a drying oven.
After the mixture has dried, the trays are manually removed and
emptied into a dry grinding unit. As the powder is ground, it
flows into a lined drum which then is sent to the existing equip-
ment for additional processing. The waste producing operations
identified were weighing operation, wet grind loading and un-
loading and dry grind loading and unloading.
  For each operation, all of the control methods originally postu-
lated by the consultant were listed (Table 5). The list was reviewed
and modified during the meeting between the client's design and
operating personnel  and the  consultant. It was during this time
that  each method was ranked for its effectiveness, extent of cur-
rent  use, future application potential and the  fraction of waste
generated by each process operation. In addition, the group was
asked to add any methods that may have been overlooked.  The
paragraphs below describe each control method for every  unit
operation shown.
  Starting with the weighing  operation, the consultant suggested
that  the company return the empty containers to  its supplier
(some suppliers of toxic materials have facilities for recovering or
disposing of material left in empty containers), purchase powder
A (the more toxic of the two  powders) in preweighed plastic bags
and use drum covers when transferring  the material. While few
thought that the first method was practical, and the third method
was currently being employed, the second method met with much
enthusiasm.  By  using preweighed  containers, the likelihood of
spills or dusting of powder A would be greatly reduced. In addi-
tion, unloading of the preweighed powder at the wet grind  sta-
tion  would allow rinsing out the bag so that the bag would no
longer be considered hazardous.
                                                                         !§§!
                                                         Figure 2
                                      WRAP Case Study—Illustration of Process Modifications
432   WASTE MINIMISATION PROGRAMS

-------
                                                           Table 5
                                     Summary of Source Control Methodology for the A/B Powder
                                                      Formulation Process

Waste Source
Weighing
Operation


1.
2.
3.

Control Methodology
Return empty containers
Use preweighed containers
Use drum covers
Waste
Reduction
Effectiveness
2
2
2

Extent of
Current Use
0
0
2
Future
Application
Potential
2
4
1
Current
Fraction of Reduction
Total Waste Index
0.00
0.00
0.25
Future
Reduction
Index
0.25
0.50
0.06
                         Overall
                                                    2.00
                                                                 0.67
                                                                              2.33
                          Overall
                                                    1.B9
                                                                 0.00
                                                                              2.33
                                                                                          0.10
                                                                                          0.45
                                                                                                      0.25
                                                                                                      0.00
                                                                                                                  0.50
Wet Grind Loading
and Unloading






1.
2.
3.
It.
5.
6.
7t
8.
9.
Use plastic funnel/collar on unit
Use smaller trays? manual operation
Place trays on rack, walk-in oven
Use elevator table to carry trays
Install roller conveyor under valve
Install fail-close valve on discharge
Reduce cleaning frequency
Bypass dry grinding unit
2
2
2
1
1
2
2
3
2
0 4
0 2
0 2
0 1
0 1
0 3
01
1
0 3
0 4
0.00
0.00
0.00
0.00
0.00
0.00
0 00
0.00
0.00
0.50
0.25
0.25
0.06
0.06
0.38
Ql T
. 1 J
0.56
0.50
                                                                                                                  0.56
      Dry Grind Loading 1. Use plastic funnel/collar on unit        2
         and Unloading  2. Do not load while unit is operating      3
                     3. Inspect all seals regularly             2
                     4. Use drum covers                   2
                     5. Bypass dry grinding unit             4
                                     0.00
                                     0.75
                                     0.38
                                     0.25
                                     0.00
0.50
0.00
0.06
0.19
0.90
                          Overall
                                                    2.60
                                                                 l.BO
                                                                              2.60
                                                                                          0.45
                                                                                                      0.75
                                                                                                                  0.90
       All Sources
                        All Methods
                                                                                          1.00
                                                                                                      0.58
                                                                                                                  0.71
  For the second operation, wet grind loading and unloading,
eight waste reduction methods were suggested. For the loading
operation, it was recommended that a plastic funnel or collar be
used when pouring powder into the unit. The wet grinder had a
small opening port and therefore the  likelihood of a spill was
great. For the  discharge operation, the original procedure re-
quired the operator to manually slide-out a filled tray from under-
neath the discharge  valve. Since the filled  tray would be very
heavy, movement of the trays on the floor would cause much
spillage to occur. Once the trays were full, they  were placed on a
rack, taken to the oven and placed inside. Therefore, many of the
methods focused on making the job easier for the operator. Sug-
gested methods were to use smaller trays (easier to move), place
the trays on a rack that could fit inside the walk-in oven, use an
elevator table to carry the trays, install a roller conveyor under
the valve and install  a pump on the unit and pump the slurry to
the oven area for tray filling. Of these methods,  installing a roller
conveyor and using a rack/walk-in oven were considered the most
practical  by both  the consultant  and plant operations group.
Since drips from the  valve onto the conveyor could occur, it was
suggested that a fail-close valve also be installed so that the opera-
tor would be discouraged from the practice of opening the valve
and sliding the trays underneath while the slurry continued to
flow. Finally, it was agreed that the cleaning frequency of the unit
would be reduced  as compared to existing units since this  unit
would be dedicated tothis-operation.
  Por the third operation {which also involved a loading and un-
loading operation), the use of a plastic funnel or collar and the
use of drum covers  were again suggested.  In  addition, it was
recommended  that the unit not be loaded while operating and
that the unit be inspected regularly. Finally, the client was asked
why the dry grind operation was necessary—subsequent opera-
tions provided  additional mixing  and million. The  discussions
that followed proved to be the most productive of all.
  At first, the facility process staff indicated that the design of
the process had been taken from the procedure developed in the
laboratory. Technical reasons for performing the dry grind opera-
tion were that segregation of the powders might occur during the
long slurry drying operation and  that powder dried in the oven
had a tendency to form lumps or a crust. This then led  the con-
sultants' attention back to the drying oven. Could something be
done to the oven to increase the drying rate (so that segregation
would not occur) and yet produce a material with no lumps or
crust? The use of a continuous drying, as opposed to batch dry-
ing, oven was suggested. The client had a Turbodryer (a contin-
uous rotary tray dryer) that was seldom used. Since the material is
dried rapidly and on a continuous basis, no segregation or lump-
ing should occur. Also, since the material is not dried to the same
degree as in the batch oven (the slight increase in moisture content
would not affect the subsequent processing step), the degree of
dusting should be reduced. Finally, the manual operation of load-
ing the drying trays back at the wet grind operation is eliminated,
since the slurry now will be pumped into the Turbodryer.  A
process flow diagram comparing the original and revised opera-
tions is shown in Figure 1.
  Implementation of the modified process is currently underway.
Emissions of materials to the workplace environment and the re-
sulting generation of hazardous  waste will be significantly re-
duced when compared to the original processing scheme.

CONCLUSIONS
  It is believed  that the systematic approach described in this
paper for identifying and evaluating waste minimization options
will prove valuable in conducting effective waste reduction audits.
  As shown  in the case study description,  the systematic pro-
cess has been applied and has led to the identification and imple-
mentation of measures  to reduce hazardous waste  generation.
This will  have the obvious effect of reducing the  company's
                                                                                     WASTE MINIMIZATION PROGRAMS    433

-------
waste  disposal costs and of deriving higher profits  realized
through better utilization  of raw materials. At the same time, in
this particular case, there is the additional benefit to the company
of reduced worker exposure to toxic emissions.
  The reader is reminded, however, that the systematic Waste
Reduction Audit Procedure by itself is no panacea.  Successful
use of this approach requires cooperation among  the audit team
and facility operations personnel; it requires a good understand-
ing of the industrial process and its underlying principles, and
finally it requires creativity and hard work.

REFERENCES
1.  Josephson,  J.,  "Hazardous  Waste Landfills," Environ.  Sci. and
   Technology, 15,  1981,250-53.
2.  Conference materials, "Waste Reduction—The Untold Story," spon-
   sored by League of Women Voters of Massachusetts, Tufts Univer-
   sity and U.S. EPA, Woods Hole, Massachusetts, June 19-21, 1985.
3.  Briner,  R.P., "Making Pollution Prevention Pay," EPA  J.,  Dec.
   1984,28-29.
APPENDIX: COMPUTATIONAL
PROCEDURES
  The text below explains how individual rankings of each option
in three separate categories  (effectiveness, extent of current use
and future applications potential) are combined to obtain qualita-
tive measures of how much  waste has already been  reduced and
how much waste can be reduced in the future for each option, for
each waste stream and for the entire process.

Current Waste Reduction Index
  The current waste reduction index  C is the fractional reduc-
tion of waste resulting from  the employment of current waste re-
duction measures.  In other words, it represents the  reduction in
waste that occurred  because of measures currently  being imple-
mented. The derivation of equations allowing computation of C
from the  ratings given to each waste reduction method is pre-
sented below.
  Consider  a process which, in the past, generated waste at  a
specific rate, Wp. Assume that, due  to implementation of certain
source control techniques, the same process now generates waste
at a specific rate, W0. Hence, the currect overall fractional reduc-
tion of waste generated is:
 C = (Wp-W0)/Wp
where:
(1)
   C  -current fractional reduction of waste  produced  by the
      process
   Wp-past specific waste generation rate (Ib waste/lb product)
   W0-current specific waste generation rate (Ib wasle/lb product)
        given source control method will be implemented to control the
        j-th waste stream. Such a variable depends on capital and operat-
        ing cost, level of difficulty, implementation period, product yield
        credit technological risk and the risk of detrimental effect on pro-
        duct quality.
           In mathematical terms, the current fractional reduction of the
        j-th waste stream by the implementation of the i-th source control
        method is given by:

        cii = eijxuij/52                                             (2)
          The normalizing factor, S, a maximum score, equals 4 if u and
        e were ranked on a scale of zero to four or equals ten if u and e
        were ranked  on a scale of zero to  ten. If the cumulative effect
        and the mutual exclusivity of all methods can be ignored, the cur-
        rent  reduction  index for the entire  j-th waste stream  can be ex-
        pressed as:
                                                                 Cj = max (cjj, i = l,2,...Mj)
                                                                  (3)
          By ignoring the mutual exclusivity of the methods, the deter-
        mination of Cj is vastly simplified. While the mathematical equa-
        tion was formulated to account for any cumulative effects, it is
        doubtful if the accuracy of the estimation would actually have im-
        proved. After all, the current reduction index is based on a theo-
        retical waste generation rate from the past, and its intent is to pro-
        vide a qualitative estimate of what already has been done to re-
        duce waste.
          Now that Cj is determined, it is useful to express it in terms of
        WQ) and Cj. Rearranging equation  (1) and writing it in terms of a
        single waste stream gives:
         w
          PJ
woj/U -C,)
                  N
                           N
        w.
£ w   = I  wo/ (l-C)
(4)
                                                                 (5)
          To obtain the current reduction index C for the whole process
        (all waste streams), equations (5) and (1) are combined to yield:
                                                                  c - 1-
    N
 w0/  z woj/(i -
(6)
          Since it is usually easier to estimate the fraction of waste, Zj,
        that a waste stream represents as part of the total waste generated
        (as opposed to estimating the actual waste generation rate), the
        following equations are used to calculate C:
  Now consider that the process has N different waste streams
and that  for  each  j-th stream (j = 1.2,.-N) there are different
source control methods available. Each method can be charac-
terized by its effectiveness, eij, its extent of current use, uij, and
its extent of further application or application potential, pij. The
first parameter, effectiveness, eij,  is defined as the measure of vol-
ume or toxicity reduction for the j-th waste stream due to full im-
plementation of the i-th (»= 1,2,..Mj) source control method. The
second parameter, extent of current use, ij, is defined as the meas-
ure of current level  of use of the particular method to control the
j-th waste stream. Finally, the future application potential, pj, is a
stochastic variable  defined as the measure of probability  that a
                                                                  w,
          oj  -
                  N
         C = 1 - I/ £ Zj/(l -C;)
                                                                  (7)
                                               (8)
           The  above equation relates the current fractional extent of
         waste reduction for the entire process to the current fraction Zj of
         each process waste stream and to the current extent of waste re-
         duction Cj for each waste stream.
434
       WASTE MINIMIZATION PROGRAMS

-------
Future Waste Reduction Index
  In addition to the  current reduction index, C for the  entire
process, an estimation of the potential future reduction index F is
required. Future reduction indices for each control method can be
calculated by using the equation:
            ij (s-Uij) Pij/53
(10)
  Again, ignoring the effect of mutual exclusivity or assuming
that a given facility will implement the most effective control
method for each waste stream:
           max (Fjj, i - l,2,...Mp
           N
           I
                 FJZJ
(11)
(12)
  Future reduction index F, expressed by equation (12) is a semi-
qualitative measure of expected fractional waste reduction from
the process due to implementation of the single most effective
control option for each waste stream.
  In summary, the method described above is based on an heuris-
tic approach where the estimates of waste reduction were derived
from the ratings given to the variables eij, uij and pij. In actual
applications, it usually is not possible to obtain precise fractional
values of e, u and p. Instead variables should be ranked on some
integer scale such as zero to four or zero to ten with the degree of
resolution based on the expertise of the individuals involved and
availability of information. Since the use of an integer scale can
result in Cj or Fj equal to 1.0 (100%  reduction of waste), it is
recommended that the maximum fractional values allowed be set
equal to 0.9 to avoid division by zero.
                                                                                    WASTE MINIMIZATION PROGRAMS    435

-------
                      Minnesota  Technical  Assistance Program:
                               Waste Reduction Assistance  for
                                    Small  Quantity Generators

                                               Cindy A. McComas
                                                  Donna Peterson
                                   Minnesota Technical Assistance Program
                                             University of Minnesota
                                             Minneapolis, Minnesota
ABSTRACT
  Waste minimization is a feasible approach to hazardous waste
management, reducing the volume of waste generated at the
source and reducing industry's reliance on landfilling.
  Unlike the federal hazardous waste regulations governing com-
pliance of small quantity generators between  100 and 1000 kg/
month,  Minnesota regulations allow no exemptions for small
quantity generators. This means that more small quantity genera-
tors in Minnesota (0 to 1000 kg/month) must abide by the haz-
ardous waste rules. In addition, because Minnesota has no waste
disposal facility, the costs of disposal can be high.
  In response to the high costs of disposal and the additional
number of small quantity generators, Minnesota has adopted a
stringent policy of waste minimization or waste reduction. The
Minnesota Waste Management Board has created the  Minne-
sota Technical Assistance  Program (MnTAP) to foster and im-
plement waste reduction in Minnesota.
  Many states across the  country are initiating technical assis-
tance programs to assist industry with the changing regulations.
Technical assistance programs have the benefits of  being non-
regulatory and confidential and thus are a good vehicle for en-
couraging compliance and waste reduction. Compliance with haz-
ardous waste regulations, too often viewed by industry as difficult
and expensive, can instead present methods such as product sub-
stitution, waste segregation, more efficient equipment and quality
improvement  in addition to recycling and treatment opportun-
ities.
  MnTAP offers several services to foster and implement waste
reduction in industry  including  on-site consultation, research
grants, the engineer intern program, educational seminars,  in-
formation gathering and dissemination and a  telephone hotline.
Waste reduction successes achieved through research, the engi-
neer intern program and  on-site  consultation are documented
and provided to other industries through reports and presenta-
tions. Significant cost saving can be realized by  seemingly insig-
nificant changes in a manufacturing process, including savings
in raw materials, savings in disposal costs and savings in ultimate
liability.

INTRODUCTION
  The trend toward hazardous waste reduction is largely in re-
sponse to changing public policy at the Federal and State levels.
One of the goals of the RCRA Amendments of  1984 is the min-
imization or reduction of hazardous waste generation and its dis-
posal  on land. The obvious  consequences of land disposal as
historically acceptable waste management have been poor man-
agement  of hazardous waste, groundwater contamination  and
other significant hazards to human health and the environment.
  The U.S. EPA has set forth its "waste minimization" policy
by requiring generator certification on the manifest that the gen-
erator has reduced waste generation to the maximum extent which
is economically  feasible.  The U.S. EPA's definition of waste
minimization may include both source reduction and recycling/
reuse. Those reduction methods which  are economically feas-
ible for a  particular generator are left  to  the generator's dis-
cretion, rather than being specified by the U.S. EPA. Waste re-
duction measures taken must b« documented and reported every
2 years.
  A programmatic trend toward  hazardous waste reduction al-
ready is taking place among the states. Currently, approximately
20 states have programs in some stage of development to respond
to the U.S. EPA waste minimization policy and generator needs.
Both the U.S. EPA and the states believe such programs are best
handled at the state level to better respond to generator needs
and specific state regulatory requirements. These state programs
typically take the form of technical assistance, regulatory assis-
tance or both.
  The  Minnesota Technical Assistance Program (MnTAP),  for
example, was established by the Waste  Management Board  for
several reasons, largely driven by state needs: (1) the adoption of
waste reduction as the most favored state policy for waste man-
agement by  the  Waste Management Board  and  the State of
Minnesota due to the  absence of hazardous waste treatment/
disposal facilities in Minnesota; and (2) the regulation of all small
quantity generators in Minnesota with no exemption for any vol-
ume of waste generated.
  Two main areas of emphasis exist for MnTAP: waste reduc-
tion and small quantity generator assistance. A five-step approach
is taken in educating generators on the options available for waste
reduction and implementing waste reduction measures.

THE APPROACH TO WASTE REDUCTION
  In the past, the availability of cheap landfill sites has acted as a
disincentive to capital expenditures  or changes in operation
needed to implement waste reduction and waste recovery.' Land
disposal now is recognized by most as the waste handling method
of last resort. Long-term liability costs of disposal have become a
major incentive to some generators to reduce their waste genera-
tion and not rely upon landfilling. When these long-term costs
are factored in, landfilling no longer appears attractive.
  In place of land disposal, the hierarchy shown in Fig. 1 is gain-
ing greater acceptance as the optimal management strategy  for
hazardous  wastes. The goal of a waste management system is to
utilize the  most  cost-effective combination  of management  op-
tions to handle wastestreams  within the confines of the regula-
tions.1 Utilizing one or more of the elements in the hierarchy may
436   WASTE MINIMIZATION PROGRAMS

-------
                WASTE  REDUCTION
       In-plant waste reduction/process modification

            On-site recycling/waste segregation

     *§?,  \.     Off-site recycling/reuse
            ^^
                        Treatment

                      Land Disposal
                          Figure 1
        Optimal Management Strategy for Hazardous Wastes
result in a  three-fold benefit:  (1) savings on raw materials; (2)
avoiding disposal costs; and (3) minimizing future liability.

Iii-Plant Changes
  Source reduction, or in-plant  reduction,  is  the  preferred
method and first step in working toward the overall scheme of
waste reduction.  Source reduction not only has the greatest
potential to reduce waste generation over other options, but also
reduces occupational exposure to toxic substances in the work-
place. Source reduction is also most likely to result in the greatest
savings on raw materials and avoided disposal costs.
  A step in the right direction toward source reduction is to con-
duct an in-plant survey, or waste audit. A survey involves collect-
ing information on the volume and  characteristics  of waste
streams generated  in a  plant. With this information in hand,
opportunities for  source reduction  can be  identified. Some of
these opportunities are reviewed below.

Improved Housekeeping
  Low cost measures such as good housekeeping practices, rou-
tine checks for faulty and leaking valves or pipes, balancing ma-
terial inputs and outputs to identify losses, better inventory con-
trol and shipping and minor process changes all help to minimize
pollution in a cost-effective manner. Cutting back water usage in
electroplating rinsewaters or eliminating vented solvent emissions
in a dry cleaner's plant are examples.

Product Substitution
  Product  substitution  involves  replacing a hazardous product
used in a production process with a less hazardous or a non-haz-
ardous product. Examples  of  product  substitution may include
eliminating the use of metal-containing paints or solvent-based
paints and substituting water-based paints containing no metals.
Many companies have eliminated the use of chlorinated solvents
in degreasing operations and have replaced them  with non-
chlorinated solvents or caustic washes, as appropriate.

Product Reformulation
  Product reformulation is substituting an end-product which re-
quires a less  waste-intensive manufacturing  process.3  As an ex-
ample, during the manufacture of polyester  resins, toxic styrene
monomer was released  into the  air. This process was changed
when chemical producers began manufacturing a line of styrene-
suppressed  resins that reduced styrene  loss  into the air by 70%
compared with nonsuppressed systems.
Production Process Modernization
  Replacing existing equipment with more efficient equipment
based on the same production methods can significantly reduce
waste generation. A classic example is  replacement of conven-
tional air-atomized spray equipment having a transfer efficiency
of 30 to 60% with electrostatic equipment (transfer efficiency
of 65 to 80%) or with powder coating equipment at 90 to 99%
transfer efficiency.'

Production Process Redesign
  Developing and using production processes of a fundamental-
ly different design than those currently used, such as waste stream
segregation, can greatly enhance opportunities for waste recycling
or reuse. Wastewater  from one part of the plant may be clean
enough for reuse in another part  of the plant. Segregation  of
metals streams may allow metals recovery with installation of a
metals recovery system.

On-Site Reuse/Recycling
  Reuse or recycling a waste stream refers to returning the waste
stream as a raw material to the original process that produced  it,
with or without treatment. A large selection of equipment is avail-
able for solvent distillation on-site. It makes economic sense  to
reclaim waste paint or overspray (through reformulation if neces-
sary) for its original purpose.

Off-Site Reuse/Recycling
  The low volumes of waste produced by some small companies
may not justify the purchase of needed equipment for on-site re-
cycling. In this case, wastes may be shipped for off-site recycling
at a centralized recycling facility. "Milk runs" by solvent trans-
porters have been established to help reduce transportation costs
and facilitate recycling.

Treatment
  Treatment of the waste to render it less-hazardous or  non-haz-
ardous can be done by applying physical, chemical, biological
or other methods. These methods may include concentration,
neutralization, detoxification, stabilization or incineration of the
waste.

Disposal
  Once all reduction, reuse, recycling and treatment options have
been exhausted, remaining wastes or residuals must be disposed
of in some environmentally acceptable manner.
                           Table 1
         Hazardous Waste Generation in Minnesota in 1984"
Waste  Category
                              Amount ( t u n s ;
Oi ly was les
Waste oil
Paint
Pesticides/Herbicides
PCBs
Organic chemical production
Halogenated solvents
Non -ha logena t ed solvents
Other organ! cs
Heavy metal solutions
and sludges
Corros ive
React ive
Other organic
Misc chemicals
TOTAL
466
8720
1606
30
1753
26
2260
13461
19025
15386

42063
982
4483
13202
123,463
• 1%
7
1
<1
1
<1
2
11
15
12

34
1
4
11
100%
                                                                                    WASTE MINIMIZATION PROGRAMS    437

-------
 MINNESOTA HAZARDOUS WASTE GENERATION
 AND MANAGEMENT
   Approximately 123,000 tons of hazardous waste were reported
 to be generated in Minnesota in 1984 in 14 waste categories (Table
 1).  Some waste streams were not reported  by disclosure but
 should be included in the overall description of Minnesota waste
 generation. All waste streams  including  disclosed  wastes, SQG
 wastes, spill and site cleanup waste and waste oil result in an over-
 all waste generation amount of 221,000 tons (Table 2).

                           Table 2
      Overall Estimate of Waste Generation In Minnesota in 1984'
Source
Disclosed
Addit lonal
SOG waste
Spill and
site c 1 c (i n
Add it i ona 1
waste oil
TOTAL
Amount
123

I,

up .11

51
221
(tons:
, 000

,000

. 000

.000
.000
( ' (• r ' e n t
<•><•,"„

3

in

23
100X
   Disclosure data for 1984 suggest 51% of Minnesota waste  is
 being sewered  with or without pretreatment,  13% is being re-
 cycled, 10% is being incinerated, 6% is being land disposed and
 14% is unknown or being treated/disposed in  other ways. Mani-
 fest data for 1984 indicated that 58,000 tons (including site clean-
 up) of waste (47% of total reported waste) were shipped out-of-
 state for treatment, recycling or disposal. Approximately 31,000
 tons of waste were managed through on-site recycling,  fuel blend-
 ing, stabilization, waste  oil rerefining, land treatment or other
 methods in-state.' The remainder is being sewered.
   This information suggests  Minnesota's hazardous  waste gen-
 eration and management concerns may not be as severe as in some
 other states. However, the absence  of permitted hazardous waste
 treatment facilities in Minnesota causes  concern when one con-
 siders that roughly one-half of reported Minnesota waste must be
 sent out-of-state.
   In order  to reduce  reliance on out-of-state  facilities and take
 responsibility  for its  own waste generation,  Minnesota has
 adopted a policy of  waste reduction. MnTAP was  created to
 educate generators on waste reduction opportunities and to offer
 direct technical  advice and assistance to small quantity genera-
 tors.

 SERVICES OF THE MINNESOTA TECHNICAL
 ASSISTANCE PROGRAM
   MnTAP was established by the Minnesota Waste Management
 Board (WMB) in November 1984 as one of the recommendations
 to the Minnesota legislature in the Waste Management Board's
 Hazardous  Waste Management Plan.  The WMB is a solid and
 hazardous waste planning  and  siting  agency  and  provided the
 MnTAP grant  to the University of Minnesota in order to call
 upon the available technical resources and  faculty. In response
 to generator needs in Minnesota,  several services are available
 free-of-charge through MnTAP

 Information Dissemination
   Three  types of information  are available to generators through
 MnTAP: (1) technical literature on waste  reduction; (2)  pro-
 fessional assistance contacts; and (3) equipment information. All
 of this information is stored on the computer and can be accessed
 by industry or waste stream type.
Telephone Hotline
   Approximately 40-50 calls per month are received from a broad
base of industries  (service and  manufacturing) covering a wide
range of topics. The majority of calls have been received from
painters/coalers and metal finishers.

On-Site Consultation
   Assistance by telephone may  not be adequate to meet genera-
tor needs. An on-site visit may be made to provide the generator
with more detailed information, clarify regulations and  look for
waste reduction opportunities. As of November 1985 30 site visits
had been made by MnTAP staff.

Engineer Intern Program
   A site visit may  reveal the  need  for longer-term waste reduc-
tion assistance.  In  this case, a student intern (paid by MnTAP)
can be provided to  the company  to work on a dedicated waste re-
duction project for one school term. The project  typically in-
volves conducting an in-plant survey, evaluating equipment, talk-
ing to vendors and assisting with design and selection of equip-
ment. An intern is a valuable resource for a small company which
may otherwise be unable to hire  a consultant. Seven interns were
employed by MnTAP during the  summer of 1985.

Education
   Education of generators  is conducted primarily by presenta-
tions and seminars  for trade associations or generators. MnTAP
has just completed a series of industry-specific seminars for dry
cleaners, metal  finishers  and others.  Fact  sheets prepared by
MnTAP  staff provide information on waste  management dis-
posal and reduction options for specific waste streams.
SUCCESS IN WASTE REDUCTION
  MnTAP  assistance to Minnesota generators covers a broad
base of industries from service related firms, such as dry cleaners
and service stations, to manufacturing including metal finishing
and plastics manufacturing.  Informational  needs of generators
may be as straightforward as providing a transporter list or may
involve a long project such as evaluating and installing equip-
ment for metals recovery from plating operations. Two examples
of actual waste reduction practices implemented as a result of
MnTAP assistance follow.

Metal Plater Assisted
  Company A is a precious metal plater of gold and silver. The
spent gold and silver rinsewaters are sent off-site for recovery, but
copper and  other metal rinsewaters historically have been pre-
treated on-site, prior to discharge to the sewer system. The com-
pany uses an existing system for  neutralization, precipitation of
metals and sludge dewatering as  wastewater pretreatment prior
to sewer discharge.
  In order to expand operation and meet sewer discharge require-
ments, this company needed to install a metals recovery system to
remove metals from its wastewaters prior to the current pretreat-
ment  stage.  An  in-plant survey  was  conducted to characterize
the  various  waste streams. This  information  enabled  the com-
pany to select and size the needed equipment.
  After reviewing technical literature and talking with vendors,
Company A selected a  scheme proposed by an equipment ven-
dor using  electrolytic metal  recovery operating in conjunction
with the existing pretreatment system.  It is estimated that installa-
tion  of the new system will result in a reduction  of as much as
90% of the hazardous metal components of the sludge.
438    WASTE MINIMIZATION PROGRAMS

-------
Electronics Firm Assisted
  Company B, a large electronics and computer manufacturing
firm,  generates two drums per day of dewatered metals sludge
from its copper plating operations. The total cost of disposal in-
cluding transportation and cost of the drum is approximately
$200 per drum of waste  for  an annual total disposal cost of
$100,000. This cost is in addition to chemicals for the wastewater
treatment process, operating personnel and utilities.
  MnTAP provided  the firm  with an engineer intern to aid in
the evaluation and installation of a metals recovery unit. It is esti-
mated that metals recovery will reduce sludge generation from
two drums per day to two drums per month. This waste reduc-
tion will  result in an  annual cost  savings of approximately
$100,000 in avoided disposal costs and transportation costs.

CONCLUSIONS
  Waste reduction, or waste minimization, as the new approach
toward waste management is clearly an improvement over histor-
ical end-of-pipe treatment for  pollution control. The U.S. EPA
is encouraging waste reduction through  its waste minimization
program. Many states have adopted waste reduction as a way to
effectively manage large and small quantity waste generation. A
number of state  technical assistance  programs are in place to
educate generators on waste reduction opportunities and  help
them implement process changes.
  In Minnesota,  the Minnesota Technical Assistance  Program
plans to maximize waste reduction efforts. Much of the empha-
sis on waste reduction is due to the absence of hazardous waste
treatment/disposal facilities in Minnesota. MnTAP believes this
approach is working and that assistance at the state level respond-
ing directly to generator needs is the best way to reduce the na-
tion's dependence on landfilling.
REFERENCES
1. Campbell, M.E. and Glenn, W.M., Profit from Pollution Preven-
  tion, Pollution Probe Foundation, Toronto, Ont., 1982.
2. Maugh, T.,  "Burial is the Last Resort for  Hazardous  Wastes,"
  Science, 204, 1979.
3. Caldart, C.C. and Ryan,  C.W., "Waste Generation Reduction: A
  First Step Toward Developing a Regulatory Policy to Encourage Haz-
  ardous  Substance  Management  Through  Production  Process
  Changes," Hazardous  Waste and Hazardous Materials,  2,  1985,
  309-331.
4. Minnesota Waste Management Board, "Draft Estimate of Need,"
  Sept. 26, 1985.
                                                                                   WASTE MINIMIZATION PROGRAMS    439

-------
              Waste  Minimization  at Air  Force GOCO  Facilities

                                              Douglas L. Hazelwood
                                              Brian J. Burgher, P.E.
                                      The Earth Technology Corporation
                                               Alexandria, Virginia
                                                   Charles Alford
                                                   U.S. Air  Force
                                    Aeronautical  Systems Division/PMDA
                                          Wright-Patterson  AFB, Ohio
ABSTRACT
  Continuing waste minimization efforts at eight aerospace man-
ufacturing facilities owned by the U.S. Air Force have identified
significant opportunities for reducing the quantities of hazardous
wastes requiring off-site treatment or land disposal. In most in-
stances, these alternatives would allow operating cost reductions
from  current practices, provide payback  of capital  investments
within 3 years  and result in  significant liability exposure  reduc-
tions.
  Hazardous waste streams  which exhibit the greatest potential
for reduction include machine coolants, hydraulic oils, solvents,
cyanide wastes, paint stripping wastes, chemical milling  wastes
and wastewater treatment sludges. The Air Force is now  imple-
menting waste  minimization measures at  its facilities to achieve
these benefits and to comply with RCRA requirements. As a  part
of this program, the Aeronautical Systems Division/PMDA  is
funding a program to demonstrate zero waste generation  at Air
Force Plant 44 in Tucson, Arizona.

INTRODUCTION
  Interest in waste minization has long been promoted by Fed-
eral legislation, including the Federal Water Pollution Control
Act Amendments of 1972, the Energy Policy and Conservation
Act of 1975, the Used Oil Recycling Act and  Department  of De-
fense  (DOD) directives such as APR 78-22  and DODD  19-14.
More recently, the impetus for waste minimization  has become
even stronger. The 1984 reauthorization of RCRA includes bans
on  landfilling and underground injection  of  certain waste types
and a requirement for certification that  waste minimization  is
being conducted by hazardous waste generators. Similarly, DOD
has issued directives requiring zero land disposal of solvents by
1986.
  The Aeronautical Systems Division/Facilities Management Di-
vision (ASD/PMDA)  of the U.S. Air Force anticipated these
developments and  initiated  programs in  1983  to address these
issues.  Studies   at  government-owned  contractor-operated
(GOCO) facilities under ASD/PMDA's  cognizance were  con-
ducted in 1983  and 1984 to identify resource conservation and
recovery activities and opportunities.
  The 1984 study demonstrated that, in many cases, plant oper-
ators  were implementing methods that could substantially  reduce
waste generation volumes and raw material requirements. How-
ever,  significant opportunities for waste minimization which ap-
peared both technically and economically feasible but were not
being pursued were also identified. In light of these  findings and
the new  certification requirements  of  RCRA, ASD/PMDA
adopted a Waste Minimization Program which includes the study
reported in this paper.
STUDY APPROACH
  The Earth Technology Corporation was given a contract by
ASD/PMDA to assist in implementing the Waste Minimization
Program through detailed studies at each  GOCO facility where
significant opportunities had been identified in the preliminary
studies.  Each of the eight selected facilities was visited by a two-
person study team  for periods ranging from 2 to 5 days to re-
view in-plant records, observe current operations and  identify
waste minimization opportunities. Based on these data, exten-
sive vendor contacts  and engineering analyses of planned and
potential waste minimization measures, detailed reports were pre-
pared for each plant describing the opportunities for  minimiza-
tion of every hazardous waste stream generated, associated costs
and benefits.
  The following sections describe the most common waste  min-
imization opportunities identified during the study. These sec-
tions, arranged by  waste type, include descriptions of the tech-
nologies available to minimize each waste  type and case studies
illustrating the benefits available from implementation.
MACHINE COOLANTS
  Machining operations (e.g., milling, turning, grinding) typical-
ly utilize a water-soluble coolant diluted to a 2 to 5% concentra-
tion with water. The coolant mixture is sprayed  directly on the
tool-part interface to provide  lubrication and  dissipate heat,
thereby preventing part damage and excessive tool wear. After
prolonged use (usually 1  to 8 weeks) the coolant  is degraded, as
evidenced  by ineffective  lubrication, rancidity and free-floating
tramp  oils. Common disposal  methods for waste coolants in-
clude underground injection and treatment through oil-water sep-
aration in  industrial waste plants. Waste coolant treatment and
disposal costs range from  $0.10/gal to $1.00/gal and  average
$0.30/gal at the eight plants studied, while coolant replacement
costs are approximately $0.35/gal (diluted).
  Advances in coolant recovery technology have  allowed indus-
trial facilities to greatly extend the life of machine coolants. Sev-
eral systems are commercially available which effectively remove
tramp oil and other impurities from waste coolants. Fresh fluids
are  added as needed to replace lost oils and biocides, and the re-
cycled coolant is returned to the machine sumps for reuse.
  Centrifugation is the most common method now used for cool-
ant  recycling. Such systems provide high efficiency separation of
swarf and tramp orls from coolant. Biocide often  is added to the
recovered coolant to curtail bacterial growth before it is returned
to the machines for reuse. Alternately, systems may be purchased
with "flash pasturization" capabilities which negate the need for
biocide addition.
440    WASTE MINIMIZATION PROGRAMS

-------
  The second type of coolant recovery now finding wide appli-
cation is coalescing plate filtration. These systems utilize series of
corrugated plates  to enhance gravity separation of swarf and
tramp oil from the coolant. Although centrifuges can achieve
greater efficiencies than coalescing plate filtration systems, their
cost is typically five to ten times that of a plate system.
  Three approaches to  coolant recycling  are  commonly em-
ployed: (1) connection of all machines to central sump systems
with continuous recycling of sump coolants; (2) manual collection
of coolants with batch processing in a central location; and (3)
use of mobile recycling systems to treat coolants at each machine
separately. Although all three approaches will achieve approx-
imately equivalent recovery efficiencies,  their economics vary
significantly. Mobile systems have  the lowest  implementation
cost but have the highest labor requirements. However, even the
comparatively high labor costs  of mobile systems are still gen-
erally less than conventional coolant replacement labor costs.
  The advantages  offered by coolant recycling can be seen by
examining operations at Air Force Plant (AFP) 4 in Forth Worth,
Texas. Coolant use at AFP 4,  operated by General Dynamics,
is approximately 20,600 gal/yr (undiluted) with replacement costs
of  $158,500/yr. Waste coolants are disposed off-site by con-
tracted underground injection at a cost of $0.53/gal or $121,2007
yr.
  Coolant use  and waste generation already have been reduced
significantly  at AFP 4 by using a centrifuge.  Six of the plant's
largest machines are connected to a centralized drainage system, a
high  speed centrifuge and  a  12,000-gal coolant holding  tank.
Approximately 500 gal/yr of biocide are added  to the recycled
coolant to  control bacterial growth at a cost of $25,900/yr. This
system extends coolant life  to over 6 months, or approximately
four  times that of untreated coolant. A net reduction in plant-
wide waste coolant generation  rates of approximately 25%  is
attributable to this system.
  Additional measures now under consideration at AFP 4  could
result in a reduction in  current coolant waste generation rates
of 90%. These include:
• Installation of a second high speed centrifuge to reduce down-
  time and improve recovery efficiency of the  existing system
  (cost $126,000; payback 3.1 years)
• Connection of six additional large machines to the central re-
  covery system (cost $100,000; payback 1.8 years)

• Installation of a chip wringer  on the central system to recover
  excess coolant  now  carried  out  on  aluminum  chips  (cost
  $233,500; payback 0.9 years)
• Installation of a coolant recovery system and collection vehicle
  for machines not on the central sump (cost $104,000; payback
  1.9 years)
• Purchase of a coolant analyzer to allow better control of  cool-
  ant quality  (cost $5,000; payback 0.7 years)
  The total estimated cost of these improvements,  $570,000, is
expected to result in savings of $410,000/yr with  projected pay-
back in 1.4 years.

HYDRAULIC OILS
  Hydraulic oils are used as internal lubricants  in mills, grind-
ers, lathes, stretchers and other metal- and wood-working ma-
chines. After extended use,  hydraulic oils may become  contam-
inated with water,  acids and particulates,  leading to a loss of
antioxidants and critical lubricating properties. To avoid internal
damage to  machines, hydraulic  oils  typically are  replaced every
6 to 12 months.
  Used oils generally are sold to off-site recyclers for eventual re-
use. However, the geographical location  of the generator or the
presence of toxic constituents in the oil  occasionally result in a
disposal fee being levied for its removal.  During the study,, man-
agement costs were found to range from a $0.50/gal fee for dis-
posal to revenues of $0.40/gal for recyclable oil.  The  average
waste oil revenues at the eight facilities studied is $0.15/gal.  Oil
replacement costs average $2.00/gal.
  On-site purification of hydraulic oils to prolong their useful life
is a viable option for  implementation at many facilities. Com-
mercial "off-the-shelf" systems are available which effectively re-
move contaminants and allow continued oil use. Most of these
systems employ cartridge filters  to remove  particulate  matter.
Filter sizes ranging from 2 to  10 microns generally are  sufficient
for removal of all particulate matter of concern.  Advanced  oil
purification systems are equipped with systems for removal of
water, acids formed during  use and other volatile liquids. Such
systems usually function by exposing a thin layer of heated oil  to
a vacuum, effectively stripping these contaminants from the oil.
  Analyses  of  oils recycled with such  systems have consistently
demonstrated higher purity  levels than those found in  new oils.
Through the routine purification of hydraulic oils (purification
intervals of approximately  1  to  2 months are  recommended),
useful life can often be extended to 10 years or more.
  Labor requirements for on-site purification generally are equiv-
alent  to  or  less than those  for routine oil replacement.  Mobile
purification systems can be attached directly to the machine's oil
return line and left unattended. This compares quite favorably
with the labor required to periodically drain, flush and refill  oil
reservoirs and has the  added advantage of keeping the machine
in use during servicing.
  Rockwell International's operations at AFP 3 in Tulsa, Okla-
homa, illustrate the potential  benefits of an on-site oil purifica-
tion program. Approximately  12,300 gal/yr  of waste  hydraulic
oil are drained from metal-working machines and sold to a local
oil  recycler for $0.15/gal. It has been estimated  that two ad-
vanced, portable purification systems  could reduce Rockwell's
hydraulic oil use by 90%.  Based on manufacturer's estimated
operating costs of $0.20/gal, avoided new oil purchase costs of
$2.00/gal and lost resale revenues of $0.15/gal,  net annual sav-
ings of approximately $17,800 are projected. Based on system
acquisition costs of $28,000, payback could be expected in under
2 years.  Additional unquantified savings would result from  re-
duced maintenance labor requirements.

SOLVENTS
  A wide variety of waste solvents were  encountered during the
course of the study. Table 1 lists these wastes and their cumula-
tive generation rates and disposal costs at the eight facilities stud-
ied. The bulk of the solvent  wastes result from cleaning/degreas-
ing operations. Solvent wastes typically are disposed off-site by
underground injection, landfilling, combustion in industrial furn-
aces or boilers and high temperature incineration. Solvents with
wide commercial  applicability such as  1,1,1-trichloroethane and
Freon usually are sold  to off-site  recyclers for reclamation and
reuse. The eight facilities studied generate 302,650 gal/yr of sol-
vent wastes. Management costs for these wastes total $254,740/yr
and average $0.84/gal.
  Several viable approaches exist for reducing solvent waste gen-
eration, including on-site distillative recovery and reuse as fuels.
In most situations, improved segregation  of solvents^ the point
of generation and/or  accumulation is required to achieve the
maximum possible benefits from these  methods. In  addition,
many facilities could markedly increase the quantities of solvents
resaleable to off-site recyclers through improved segregation.
                                                                                     WASTE MINIMIZATION PROGRAMS    441

-------
Waste Solvent Segregation
  As shown in Table 1, over 75% of the waste solvents generated
at the COCO facilities were mixed with  other solvents, water,
paints and other wastes, usually during their accumulation. These
mixed solvents generally are not suitable  for recovery or use as
fuels because of this contamination.  It is estimated that 50 to
75% of these solvents could be recycled or used as fuel supple-
ments if properly segregated. In every situation encountered dur-
ing the study, proper  segregation appeared to be achievable by
increasing the number of containers used  for accumulation  and
training workers in the importance of waste segregation.

On-Site Solvent Recycling
  Many solvents used at the COCO facilities studied are excellent
candidates for on-site distillative recovery and reuse. Although
distillation has been used widely for  many years in commercial
facilities to extend solvent life,  concern exists over the use of re-
covered solvents in military applications. Solvents purchased for
use in cleaning, degreasing and painting operations usually must
meet rigid military specifications (mil specs). Two approaches
have been suggested for resolving this  problem: (1) recovered  sol-
vents can be tested to assure that they meet mil specs and are suit-
able for reuse without restriction and (2)  recovered  solvents  can
be assumed not to meet mil specs and their use restricted to non-
critical  operations. The former  approach  is being used success-
fully at one of the facilities studied.
  AFP 59 in Binghamton, New  York, operated by General Elec-
tric  (GE), has been using a simple batch distillation system for 7
years to extend the useful life of 1,1,1-trichloroethane (TCA) de-
greaser solvents. Simple pH and specific gravity analyses are used
to determine when TCA in the degreasers must be replaced  and
to  verify that the distilled TCA  is suitable  for  reuse. This
approach has resulted in a reduction of approximately 70% in
waste TCA generation at AFP 59.
  In addition, GE now is evaluating the use  of  additives to
further extend solvent  life. Acid acceptors  and white metal stabil-
izers present  in commercial blends of  TCA will become depleted
through continued use and eventually will  render the solvent un-
suitable for reuse. Replacement additives,  available from several
solvent manufacturers and  suppliers, can be used  to replenish
lost additives, thus allowing continued reuse of the solvent.
  The second  alternative,  restricting  the  use of recovered  sol-
vents, is particularly well suited to  painting cleanup operations.
The largest single type of waste solvent at the facilities studied,
painting cleanup waste, typically contains approximately 50%
methyl ethyl  ketone (MEK) from the  cleanup of spray guns  and
other equipment.  (Some facilities recently have switched from
MEK to  nonregulated cleanup solvents  to  comply with VOC
emission requirements.) As these cleanup solvents need not meet
mil  specs, they are excellent candidates  for on-site distillative
recovery and reuse. Segregation of  concentrated paint gun resi-
dues from the washing solvents can  further  increase the  effic-
iency of recovery measures.
  The efficiency of distillative recovery operations is highly  de-
pendent on the type of equipment utilized. The continuous type
side  stills used on many vapors degreasers  do extend solvent  life
considerably, but they also produce a  solvent-rich waste stream.
Analyses have shown that many side stills  produce residues con-
taining 95% or more solvent. Several manufacturers now offer
batch systems which generate a  waste stream that is virtually  sol-
vent-free. Recovery efficiencies of  95% and recovered solvent
purity levels in excess of 99% are commonly  reported with such
systems.
  The economics  of  on-site solvent  recovery usually  are very
attractive. As an example, the system  used by GE at AFP 59 re-
                           Table 1
       Solvent Waste Generation at Air Force COCO Facilities
          WASTE
         STREAM
                       RATE
                    (GAL/YEAR)
                                               DISPOSAL  COST
 1 .


 2 .

 3.


 4 .

 5.


 6.

 7 .
Methyl  Ethyl Ketone  133,800
Paints,  Thinners
Mixed  Solvents

1,1,1-Tr ichloro-
et hane

Trichloroethylene

Freons


Naptha

Isopropyl  Alcohol

Other  Segregated
Sol vents
102,500

 40,450


 18,500

  1, 100


  2,500

     600

  3,200



302,650
$144,300


$113,400

$ 20,600
(revenue)

$ 13,000

$     500
< revenue I

$  1,320

$     320

$  3,500
                                                    $254,740
duces waste generation and solvent purchases by approximately
7,500 gal/yr, resulting in an estimated net savings of $50,400/yr.
The distillation system was purchased in 1978 at a cost of $3,500.

Waste Solvents as Fuels
  Many waste solvents contain considerable energy and are ex-
cellent candidates  for use as supplemental  fuels  in boilers and
furnaces. Six of the eight facilities surveyed currently  sell a por-
tion  of their waste streams to off-site fuel blenders.  However,
none of the facilities studied reuse waste materials on-site as fuels,
even though most do operate  industrial boilers  to meet plant
steam needs. Although nonhalogenated waste solvents such as
MEK,  Stoddard solvent and acetone could  be used to provide a
small portion of these energy needs, uncertainties regarding the
eventual regulatory requirements which will apply to hazardous
waste fuel use have deferred  such undertakings at the facilities
studied.
  In its Nov. 29, 1985 promulgation of waste oil burning regula-
tions (Federal Register, 49164), the U.S. EPA has indicated that
regulations addressing the use of hazardous wastes as fuels will be
promulgated in 1986. Current regulations, contained in 40 CFR
Part 266,  do not address  issues critical to the decision  process
for facility  operators:  equipment  permitting and waste testing
requirements. It appears that the planned  regulations will have
significant impacts on the economics of waste fuel use, particu-
larly for facilities which generate small  quantities of  hazardous
waste suitable for use as fuels.
  The average GOCO facility surveyed generates approximately
23,000 gal/yr of nonhalogenated waste solvents suitable for use as
a supplemental fuel. Assuming an average disposal cost of $0.857
gal,  an average waste heat content of 110,000 Btu/gal and  fuel
costs of $6.00/mmBtu, it is  projected that the average GOCO
facility could save $34,000/yr in fuel and disposal costs through
the use of waste solvents as fuels. However, as the costs of regu-
latory compliance and boiler system engineering modifications
could  overshadow these potential savings,  it  now appears pru-
dent to delay  implementation  decisions until the U.S. EPA's
final regulations are promulgated.
442    WASTE MINIMIZATION PROGRAMS

-------
CYANIDE WASTES
  Cyanide is used widely at Air Force GOCO facilities in elec-
troplating processes. Cyanide, present in cadmium,  silver, zinc
and other electroplating solutions, typically renders the spent
solutions, tank sludges and  rinsewater treatment sludges haz-
ardous. These wastes typically are treated on-site through cyanide
destruction but also may be sent off-site for treatment  or dis-
posal. In addition to conventional cyanide destruction, two al-
ternative cyanide-free plating technologies are being considered
for application at GOCO facilities; Ion Vapor Deposition (IVD)
and acid plating.

Ion Vapor Deposition
  IVD  technology is now being considered  at AFP 4 in Fort
Worth, Texas, as an alternative to cyanide-cadmium plating of
aluminum aircraft components. The  system, which relies on the
deposition of ionized aluminum in a vapor form, would virtually
eliminate waste generation from the cadmium  plating line and
would reduce plant-wide cyanide waste generation by 80%. Cad-
mium consumption would decrease  significantly, owing to the
high efficiency of IVD, and the need for continuous rinsing would
be eliminated, thus reducing wastewater generation.

Acid Cadmium Plating
  A promising alternative available for cadmium cyanide  plating
replacement appears to be acid cadmium plating. One such plat-
ing solution  is manufactured by LeaRonel, Inc. of Freeport,
New  York, under the trade  name "Kadizid" plating solution.
This proprietary batch solution consists of cadmium oxide, sul-
furic acid, brightener, starter and stabilizer compounds.  Lock-
heed-Georgia Company at AFP 6 incorporated this acid cadmium
plating system in August 1983. Lockheed has found no reduc-
tion in product quality following changeover, but has realized  a
slight reduction in operating  costs and total elimination of cya-
nide operations.

PAINT STRIPPING WASTES
  Two  of the eight GOCO facilities  studied conduct large-scale
paint stripping operations. Paint stripper, typically a methylene
chloride/phenolic compound, is  rinsed with water from the part
being stripped and collected in floor drains for bulk off-site dis-
posal or on-site treatment in  wastewater systems. Waste genera-
tion at these  two facilities  totaled approximately 500,000 gal/yr
in 1984 with off-site disposal costs  of $117,000/yr.  Three ap-
proaches  for reducing the generation of paint  stripping wastes
are being considered by the GOCO  facilities:  (1) segregation,
(2) on-site treatment and (3) alternative stripping techniques.

Segregation
  In both facilities involved in large-scale stripping, significant
quantities of other nonhazardous wastes also  could enter the
stripping waste collection systems by virtue of their designs. Im-
proved  segregation of paint stripping wastes from floor wash-
ings, noncontact rinsewater and surface runoff could  significant-
ly reduce waste  generation  at  these facilities.  For example,
segregation measures recommended at Lockheed's AFP 42 opera-
tions  in Palmdale, California, are projected to  cost $2,000 and
would result in a 70% reduction in waste generation with disposal
cost savings of $46,000/yr.

On-Site Treatment
  In conjunction with  waste generation reduction techniques or
alone, waste treatment represents a viable approach to reducing
off-site waste disposal. AFP 3 in Tulsa, Oklahoma, is now pro-
ceeding with the design of an on-site paint stripper waste treat-
ment  system as part of a long range wastewater treatment plan.
The planned system involves batch treatment of liquid stripper
wastes with hydrogen peroxide in the presence of an iron cata-
lyst to oxidize organics. Treated paint stripping waste then would
be discharged to the plant's general industrial waste treatment
system for further treatment.
  Through  this treatment scheme,  all stripper waste,  except
sludges currently generated in collection sumps, would be elim-
inated from off-site disposal. Some increase in sludge volumes
from  waste treatment would occur, but this would  be  insignifi-
cant. Although the economics of the planned system are not avail-
able, significant savings are expected to result as AFP 3 stripper
waste generation approaches an anticipated rate of 1.2 x  106
gal/yr.

Dry Media Stripping
  Several alternative paint stripping techniques currently are  be-
ing studied by DoD. One of the most promising methods is  dry
media stripping with plastic beads. Hill Air Force Base in Ogden,
Utah, has successfully demonstrated plastic media stripping  for
aircraft renovation. In the Hill AFB process, conventional sand
blasting equipment is being employed with recoverable plastic
beads for paint removal. A dry waste  of paint and plastic beads
is generated which can be separated to produce a relatively small
volume of paint waste. Waste volumes and labor requirements
have been shown to be significantly lower than conventional wet
stripping methods.  A full size dry media stripping  demonstra-
tion operation currently is being implemented at Hill AFB.
  In addition to Hill AFB, several other DoD facilities are inter-
ested  in  plastic media stripping  methods including Alameda
NARF and Pensacola NARF. At Pensacola, dry stripping cur-
rently is being tested for fiberglass helicopter components. Prob-
lems with dry media stripping appear  to be control of the strip-
ping operation  to prevent aircraft damage and dust generation
and collection. However, current  development efforts are  di-
rected at alleviating these problems through careful design and
operation of stripping systems.

WASTEWATER TREATMENT SLUDGE
  Five of the GOCO facilities studied operate on-site industrial
wastewater treatment plants. Approximately 2.9 x 106  Ib/yr of
sludge from three of these plants are disposed off-site at a cost of
$112,000/yr. The other two facilities produce 16.3 x  106 Ib/yr of
sludge for on-site disposal. Three of the GOCO facilities studied
are implementing renovations expected to  reduce sludge genera-
tion by 30 to 90%  at each facility.  Additional  reduction meas-
ures are under consideration at all facilities.
  The most significant sludge reductions planned will result from
the installation of more efficient sludge dewatering systems. Two
of the GOCO  facilities studied are planning to replace older
vacuum filtration systems, now producing sludge with approx-
imately 20% solids, with presses expected to increase solids levels
to 35% or greater. At AFP 3 in Tulsa, Oklahoma, a sludge  de-
watering system will be utilized to dewater a 2% solids slurry now
being disposed  on-site in lagoons. The planned AFP 3 renova-
tions also will include improved treatment measures which should
allow delisting of the dewatered sludge.
  Increasing chemical waste storage capacity often can allow  sig-
nificant decreases in sludge  production. During the study, sev-
eral waste  treatment plant operators  noted that a lack of ade-
quate holding capacity for waste acids and bases often required
the use of purchased chemicals to achieve neutralization. If addi-
tional storage capacity  were available, acidic and caustic wastes
could be mixed to  achieve neutralization, thus reducing chem-
ical usage and sludge production.
  Another waste minimization option  being considered is the use
of high-performance neutralizing and coagulating agents. Several
                                                                                    WASTE MINIMIZATION PROGRAMS    443

-------
chemical manufacturers now are offering substitutes for con-
ventional treatment compounds such as lime, caustic soda, ferric
chloride, ferric sulfate and aluminum chloride.  These high per-
formance substitutes reportedly decrease sludge bulk and improve
dewatering characteristics. In many  situations, plant throughput
can  be increased significantly  through  the  use of these com-
pounds, thus reducing or eliminating expansion needs.
FUTURE DIRECTIONS
   It is ASD's intent,  as demonstrated by this  program, to vig-
orously pursue all viable approaches for waste minimization at
the GOCO  facilities for which it has responsibility.  As appro-
priate, funding for implementing waste minimization measures
is being incorporated into current and future plant budgets.
  AFP 44 in Tucson, Arizona, is being used to demonstrate the
feasibility and benefits of waste minimization. In conjunction
with the  plant operator, Hughes Aircraft, ASD already has im-
plemented measures which have reduced waste generation by 99%
from 1983 levels. These measures include a state-of-the-art waste-
water treatment system which provides plant makeup water that is
cleaner than available city water, a coolant recycling system and
computerized plating lines.  Future plans at  AFP 44 call for in-
stallation of a  high-temperature hazardous waste incineration
system for organic wastes, replacement of cyanide plating lines
with IVD systems, and on-site solvent recycling and reuse.  It  is
ASD's goal  to demonstrate  the complete elimination of hazard-
ous waste generation at AFP 44 and eventually transfer these min-
imization technologies and management approaches to the other
facilities for  which it is responsible.
444    WASTE MINIMIZATION PROGRAMS

-------
                      Defense  Environmental Leadership  Project
                               Study  of  Industrial Processes to
                                     Reduce Hazardous Waste

                                        Thomas E. Higgins, Ph.D., P.E.
                                                  Drew P. Desher
                                                    CH2M HILL
                                            Industrial Services Division
                                                  Reston, Virginia
ABSTRACT
  The U.S. Department of Defense (DOD) operates industrial
facilities that  repair and recondition planes,  helicopters, ships,
tanks and other vehicles and equipment. At these facilities, paint
stripping, solvent cleaning  and metal plating are the industrial
processes that produce most of DOD's hazardous wastes.  In
policy directives issued in 1980, DOD directed the armed services
to significantly reduce the generation of hazardous wastes.
  CH2M HILL was contracted by the Defense Environmental
Leadership  Project and the U.S.  Army Corps of Engineers to
evaluate 42 cases of industrial process modifications attempted by
the three armed services to  reduce their generation of hazardous
wastes.  Process  modifications  evaluated  included  resource
recovery and recycling, source separation, raw material conserva-
tion, material substitution, improved housekeeping practices and
the development of innovative manufacturing techniques.

INTRODUCTION
  Traditionally, waste producers have not paid the true costs of
disposal. Prior to the twentieth century, these costs were borne by
water consumers and paid for in disease and premature death. At
the turn of the century, an understanding of the link between
polluted water and  disease  and the adoption of water treatment
led to a dramatic decline in the incidence of water-borne
epidemics. Still, the costs of waste disposal were borne by water
consumers,  this time in the cost of water treatment.
  Gradually, pressure was exerted to shift the burden from water
consumers to waste dischargers. Industries and communities were
required to  treat their wastes sufficiently to protect downstream
users.
  As our understanding of the long-term effects of trace quan-
tities of contaminants improved and our analytical capabilities for
detecting minute concentrations developed, increasingly stringent
treatment requirements were  placed on water and wastewater
treatment plants. The logical result has been to shift the burden of
compliance  closer  to  the  individual  producers  of  hazardous
wastes, since segregated concentrated wastes can be treated more
efficiently before they are mixed or diluted.
  Now that producers are faced with the true costs of disposing
of their hazardous wastes, it has become increasingly evident that
waste minimization is rapidly becoming an economic necessity.
Waste minimization can be accomplished by the  recovery and
recycling of waste  materials  or,  preferably, by modifying  in-
dustrial processes to reduce or eliminate the production of the of-
fending waste products.
  Since 1980 it has been DOD policy to limit the generation of
hazardous waste through alternative procurement policies and
operational procedures that are both environmentally attractive
and fiscally competitive. The DOD directed the Army, Navy and
Air Force to reduce quantities of hazardous waste, when feasible,
through resource  recovery and reclamation, recycling,  source
separation and raw material conservation.
  In carrying out the intent of these policies, many studies have
been performed that recommended industrial process modifica-
tions which, if successfully implemented, have  the potential to
significantly reduce the generation of hazardous  wastes at the
source, rather than treating these wastes at end-of-pipe facilities.
Several  modifications  have  been  successfully  implemented;
however, many other suggested changes either have not been im-
plemented or have failed to meet their goals.
  While there are specific circumstances and reasons behind the
success, or lack of success, of each modification attempted, two
factors have been an integral part of each of the successful pro-
cess modifications; one or both of these elements has  been miss-
ing from those modifications that have been less than  successful.
Very simply stated, in process modifications that were  successful-
ly implemented,  the  production   people  were  sufficiently
motivated to make the  change,  and the technologies  were
"elegant in their simplicity." Factors that motivated personnel in-
cluded an improvement in production rate and quality, a reduc-
tion in overall costs,  a decrease in manpower requirements and a
decrease in the quantity of hazardous wastes to be disposed of.
Technologies that were elegant in their simplicity were easy to
operate and maintain, reliable and cost-effective. Successfully im-
plemented process modifications combined effective technology
and motivated personnel to significantly reduce hazardous waste
production by substantially  changing the  process, substituting
raw materials, or recovering and reutilizing waste by-products.
  This paper reviews a few of the  more useful examples of in-
dustrial modifications attempted by the armed services. These ex-
amples include paint stripping modification, solvent recovery,
metal plating modification and the  elimination of solvent clean-
ing. These modifications affect the major sources of hazardous
waste generated by the DOD. The paper also summarizes the con-
clusions of the project on factors contributing to the successful
implementation of process changes. Detailed  analyses  can be
found in project reports1'2 and in a previous paper.3

PAINT STRIPPING
  Paint  stripping,  in  preparation  for  reconditioning  and
recoating, is performed at virtually every DOD industrial facility.
In a typical paint stripping operation, sprays or  baths containing
                                                                                WASTE MINIMIZATION PROGRAMS   445

-------
acidic methylene chloride, phenolic compounds or hot alkaline
sodium hydroxide solutions are employed to loosen and dissolve
old paint. After the paint softens, the resulting solvent-paint mix-
ture is scraped off. In addition, hard to remove paint is machine
sanded, often resulting in damage to the metal.
  The solvent-paint mixture falls to the floor, where it attacks the
concrete and makes the floor slippery. The floor  is frequently
washed with water to reduce the slipping hazard. For a typical
stripping operation on a fighter aircraft, thousands  of gallons of
solvent-laden wastewater  are generated. Wet  paint stripping  is
labor intensive, dirty and imposes a significant burden on waste
treatment facilities.
  Several alternative paint stripping processes have been studied
by private industry and the military. Among these are dry media
blasting, laser stripping, flash lamp stripping, water jet stripping,
CC>2 pellet blasting and cryogenics. The most promising of these
techniques is dry media blasting using recoverable plastic media.
  In plastic media stripping, small, angular plastic particles are
air blasted at the painted surface,  causing the coating to dislodge.
The key parameter for the successful use of plastic media blasting
is hardness.  The  paint must be  softer than the plastic media,
which, in turn, must  be softer  than  the underlying substrate.
Through careful control of the size of the particles and the condi-
tions of the process, the used plastic media can be separated from
the  loosened paint  particles and recycled. Generation  of  wet
hazardous waste (solvents and paint sludge in water) is completely
eliminated. A small  volume of dry hazardous waste is produced.
   While  there  are significant advantages to  the  plastic media
stripping  process, care must be taken in adapting it to specific ap-
plications. Stripping of  thin-skinned  aluminum,  magnesium,
fiberglas and other composite surfaces  requires skilled operators.
These operators must carefully select and control several variables
(e.g., media hardness, roughness and size; blast pressure; stand-
off distance; application angle; nozzle size; and feed rate) so that
surfaces are not damaged during  stripping.
   Following  extensive  testing  on  aircraft  components  to
demonstrate the effectiveness and safety of the process, personnel
at Hill AFB stripped a  complete  F-4 fighter plane in July 1984.
The aircraft was completely stripped with one-tenth of the man-
power  required for conventional wet  paint stripping.  This  test
demonstrated that the process is much less  labor-intensive than
solvent stripping. In addition, greater  control in  stripping was
achieved compared to that achieved with wet paint stripping and
sanding.  This degree of control  resulted in reduced damage to
underlying surfaces.
   A full-sized plastic media blasting booth has been constructed
at Hill AFB based  on  the prospects of reduced manpower re-
quirements  and  favorable environmental impact. The booth
incorporates five blast machines, a live floor vacuum system to
provide ventilation and particle and dust removal, and a separa-
tion system for media recovery and reuse. This booth was used to
blast strip an F-4 aircraft  in an elapsed time  of 5.4 hours.
  The new booth cost $650,000 to purchase and install. Yearly
savings are  anticipated  to be $5,600,000, resulting in  a  6-week
payback  period.  A significant  portion of this savings  is at-
tributable to a  reduction  in hazardous waste from the currently
estimated 10,000 Ib of wet hazardous sludge per aircraft to 320 Ib
of dry paint chips and decomposed plastic media per aircraft.
  The Naval Air Rework  Facility (NARF) in Pensacola, Florida,
has successfully stripped paint from aircraft and helicopter parts
using plastic media. Paint stripping of parts is currently being
done  in  enclosed  glove  boxes and  walk-in blast rooms.
Pensacola's long-range  plans involve converting two helicopter
hangars to accommodate  dry media stripping. These plans have
been postponed pending  resolution  of OSHA  regulations that
prohibit people from working in  rooms in which blasting is car-
ried out with  "organic"  media.  Concern was  expressed  that
generation of dust might post a possible explosion hazard.
  The Army is utilizing plastic media blasting to strip paint from
composite helicopter components at  Corpus Christi Army Depot
in Texas. Following a demonstration of the technology at Hill
AFB, Republic Airlines installed a plastic media  paint stripping
system  in  its repair  facilities in Atlanta, Georgia. Using this
system, they have successfully stripped the paint from more than
30 DC-9s.  Plastic media paint stripping is rapidly becoming the
state-of-the-art for paint stripping.
  The development of the plastic media blasting technology  at
Hill AFB in Ogden, Utah,  is a clear  example of the key elements
that contribute to the successful implementation  of a  modifica-
tion. The process is simple. Conventional sand  blasting equip-
ment was adapted to include media recovery and separation of the
media from the waste paint chips and dust. The modification was
championed  by R.  Roberts, a staff member at Hill AFB,  who
recognized  the  environmental  disadvantages of the existing
methods used  for  stripping planes. He  tried many  processes
before discovering, developing, implementing and promoting dry
plastic media stripping.
  The DOD has estimated  that millions of dollars could be saved
annually and that the generation of millions of gallons of hazar-
dous wastewaters could be  avoided by switching to plastic media
paint stripping at all facilities.

RECYCLE OF SOLVENTS
  Organic  solvents are used at  virtually every military facility.
Trichloroethylene (TCE),  1,1,1-trichloroethane and perchloro-
ethylene are used in vapor degreasers, and mineral spirits, such as
Stoddard solvent and Varsol, are used in cold cleaning baths.
Alcohols and Freon commonly  are  used for metal preparation
and  precision cleaning of electronic equipment. Solvents also are
used  in painting operations. Methylene chloride commonly  is
used to strip paint and carbon from  metal surfaces, and toluene
commonly  is used to thin solvent-based paints.  Volatile solvents,
such as methyl ethyl ketone (MEK) and xylene are used to clean
painting equipment.
  The safe disposal of waste solvents  can cost more than $100 per
drum. This cost is expected to increase substantially as a result of
RCRA  regulations  banning  the disposal  of  liquid hazardous
wastes on land. In addition, when waste solvents are disposed of,
fresh solvents must be purchased. A  recent study estimated  that
DOD purchases and disposes of  approximately 50,000 drums of
cleaning solvents per  year.' The  recycling of waste solvents can
result in a  savings of hundreds  of dollars per  drum due to the
elimination of the costs of waste disposal and initial purchase of
virgin material.
  Solvents  most frequently are  reclaimed  by  batch distillation
which typically consists of  a still pot, a heat source and a con-
denser. The waste organic mixture is loaded into the still pot, and
heat is  applied  to the contents, causing the  mixture  to  boil.
Organic  vapors separate  from  the waste mixture  and  pass
overhead to the condenser. Cleaned organic fluid then is collected
for  reuse, and the still bottoms are disposed  of as hazardous
waste.
  An atmospheric still can reclaim  organic solvents that  have
boiling points of less than 325 °F.  By adding vacuum, a distillation
unit  can be used to recover organic fluids that  have atmospheric
boiling points of up to SOOT while maintaining a 300°F limit in
the still's pot to limit degradation. A still is usually  heated directly
by electricity or indirectly by steam  generated by an electrically
heated boiler.
  Waste solvents can be collected and transported to a centralized
distillation  facility for recovery,  or they can be recycled at the
point of use. DOD facilities have  successfully used  both ap-
446    WASTE MINIMISATION PROGRAMS

-------
preaches. Regardless of where the distillation occurs, it is critical
that waste solvents  be properly  segregated and  stored so that
various solvents and impurities do not mix.
  The main advantage of operating a large centralized facility is
that capital costs can be recovered quickly due to economies of
scale. A centralized facility can redistill large quantities of various
types of solvents. The disadvantages of a centralized facility are
that, since many different types  of solvents  are  recycled, great
care must be taken with waste segregation and sample analysis,
and solvents must be transported to and from the point of use.
  Decentralized facilities are sometimes preferable because the
waste  generator has total control over the recycling  operation.
Since only a few types of solvents  are redistilled at small facilities,
laboratory analysis of waste solvents often is not required. Labor-
intensive  transportation  and  segregation  activities  also  are
eliminated.
  A centralized facility is dependent on a dedicated individual to
initiate and supervise the operation of the system and on an en-
thusiastic staff dedicated solely  to solvent collection, analysis,
recycling  and  distribution.  In a decentralized  facility, more
personnel (foreman  and operators) must be persuaded to adopt
solvent recovery as  part of their routine. Both centralized and
decentralized  solvent  recycle facilities have  been successfully
implemented by DOD.
  Warner Robins AFB, Macon, Georgia, has had a centrally
operated atmospheric still since August 1982.  Their organic fluid
recovery system consists of a single-stage batch still,  a water
separator and an electricallly powered steam generator. The still
can operate with a pot temperature up to 300 °F and can reclaim
organic fluids at a rate of 55 gal/hr.  The still has been used to
reclaim trichloroethane,  Freon-113 and isopropanol at recovery
rates of 70 to 99%. From the  initial  startup in August 1982 to
Dec. 31,1984, it was estimated that more than $230,000 was saved
due to a reduction in the need for virgin material and in hazar-
dous waste disposal costs. Reclamation of the used chemicals cost
only $13 per drum, whereas disposal of the chemicals and repur-
chase  of new materials would have cost from $250 to $500 per
drum.
  Robins AFB has been  able to successfully recycle solvents in a
large-scale operation  because of  careful  waste  segregation,
storage and transportation.  Site managers are responsible for
segregating and labeling waste drums at 30 different collection
areas,  Before solvents are reclaimed, samples are analyzed to con-
firm the labeling. Samples are also analyzed  after distillation to
ensure that they meet appropriate military specifications.
  Solvent recycling has been successful at Robins AFB because of
a strong commitment from management to reduce the quantities
of waste solvents that must be disposed of.  More importantly,
production personnel have cooperated with the recycling team so
that  waste  solvents can  be  segregated,  labeled,  analyzed,
transported and redistilled in an orderly and systematic fashion.
  One major factor in the success of the Robins solvent recovery
program was the commitment of  O.K. Carstarphen, the solvent
reclamation manager. Through his perseverance,  colleagues and
management were convinced to supply equipment and manpower
to set up and run the program. In  addition, Mr. Carstarphen had
extensive experience in distillation technology, and he selected
proven equipment that was reliable, simple to operate and easily
maintained.
  An  example of  a decentralized  facility  is Norfolk Naval
Shipyard in Norfolk, Virginia, which installed a $10,000 nonfrac-
tionating, batch still to recover waste  solvents generated in the
paint shop during cleaning operations. This small still, which has
a capacity of 2 gal/hr, is used to recover methyl isobutyl ketone,
MEK,  epoxy thinners and mineral spirits. Operators can operate
the  still with or without a vacuum system depending upon the
distillation temperature  required.  After a  15-minute startup
period, the still runs without operator attention. More than 80%
of the waste solvent is recovered at a cost of about $0.15/gal.
  The recovery process was successful at Norfolk because of the
personal dedication of J. Coulter, the paint shop foreman, and
the straightforward, uncomplicated operation  of a technically
innovative system.

METAL PLATING
  Plating is defined as the  deposition of a layer of metal on the
surface of a base metal for the purpose of changing the properties
of the base metal. Plating may be used to improve the appearance
of the metal (decorative plating), to increase its resistance to cor-
rosion or to  improve  its engineering  properties  (hardness,
durability, solderability or  frictional characteristics).  The prin-
cipal metals used for plating at military facilities are cadmium and
chromium.
  The major discharges of hazardous waste from typical metal
plating facilities are rinsewater contaminated by drag-out from
various cleaning and plating baths, cleanup of spills, disposal or
acid and alkaline cleaners and  occasional plating bath dumps.
  Sacrificial cadmium coatings normally are applied to protect
the base metal, which is typically iron or steel.  A thin surface
coating normally is applied  for decorative purposes or to provide
corrosion protection, improve wear  or  erosion  resistance, or
reduce friction. Many parts can  be efficiently plated in a few
tanks, but a significant amount of waste rinsewater is produced
due to the drag-out of plating  solutions on the parts.
  Cadmium almost universally has been plated  from  alkaline
cyanide baths, due to the improved plate resulting from the stable
cadmium cyanide complexes.  Unfortunately, cadmium  cyanide
baths are costly, dangerous to  operate and expensive to treat.
  Lockheed's  plating shop located in Air Force Plant No. 6,
Marietta, Georgia, switched from an alkaline cyanide cadmium
bath to a proprietary acidic noncyanide cadmium bath. Lockheed
found that the product quality improved as a result.  However,
more  careful process control  was required. Although the  new
plating  solution,  which  costs  approximately $3/gal, is more
expensive than the old cyanide formulation, reduced waste treat-
ment  costs have resulted in a net  cost savings for this modifica-
tion. As a result of this modification, alkaline chlorination for the
treatment of cyanide no longer is  needed.
  The new process was  implemented primarily  to reduce the
hazards associated with the operation and disposal  of  cyanide
baths. An improvement in quality and a reduction in total costs
led to the permanent adoption of the modification. In addition,
the process required a minimal amount  of change from the old
method,  and existing personnel had no trouble adapting to it.
  Getting rid of both cadmium and cyanide was the goal  of  a
modification  attempted at  North Island NARF in San Diego,
California. Wet cadmium plating was replaced by a method of
coating parts with aluminum using ion vapor deposition (IVD) for
corrosion resistance.  This  substitution  eliminated  the  en-
vironmental problems associated with cadmium and cyanide. For
the past 7 years, metal parts such  as landing gears, bolts and tail
hooks have been plated with the dry IVD process.
  Advantages of the process include safer working conditions, an
end product that can withstand higher operating temperatures,
improved throwing  power and better adhesion of the aluminum
coating compared to that achieved with  cadmium. Despite these
advantages, this process has not realized  its potential.
  The process is  considerably  more complex and requires more
labor and skill than cadmium plating. For this reason, the process
has been used for only a few parts, and new parts are evaluated on
an individual basis to determine the preferable plating technique.
The process change has not significantly reduced  the amount of
                                                                                    WASTE MINIMIZATION PROGRAMS    447

-------
cadmium plating performed at the facility. In addition, there was
no champion who assumed responsibility for implementing the
modification and ensuring its success.
  Chromium is used principally in the remanufacturing of worn
parts whose replacement with new parts would not be feasible or
economic because of their unique design. Remanufacturing con-
sists of machining the worn part or stripping a portion of the old
plate, overplating it  with a thick  layer  of chromium  (hard
chromium plating) and machining it back  to  original specifica-
tions. To achieve the required  thickness  of chromium, parts
typically are plated for more than 24  hours as opposed to minutes
required for cadmium plating. In contrast to  cadmium  plating,
many tanks are needed for a few parts, and drag-out losses are
slight. Lack of drag-out creates a buildup of contaminants in the
plating bath, and the resulting bath  dumps are the major source
of waste from hard chromium plating.
  Pensacola NARF and Charleston  NSY tested a vapor recom-
pression evaporator system for the recovery of chromium from
hard chromium plating rinsewater.  The system  consisted  of a
cation exchange unit used for the removal  of contaminating ca-
tions and a vapor recompression evaporator used to concentrate
the cleaned rinsewater to plating bath strength.
  There were several reasons  for the lack of success of this
modification.  The initial  feasibility evaluation was based on the
amount of drag-out experienced from decorative chromium baths
rather than the significantly lower quantity of chromium available
from  hard chromium plating.  The equipment was complicated
and required excessive operation and maintenance attention. The
system was plagued by the failure of seals and by severe corrosion
of the carbon  steel compressor.
  Despite the  fact that the testing programs were championed by
Ms. D. Preble (Pensacola) and  B.  Mordecai (Charleston), the
complexity of the chosen equipment and the faulty assumptions
used for the initial feasibility evaluation doomed this project to
failure.
  As a result of the lessons learned from the failure of the vapor
recompression evaporator system, the  Naval  Civil Engineering
Laboratory (NCEL),  Port Hueneme,  California,  chose  other
methods to modify the hard chromium plating  process at  Pen-
sacola  NARF to  reduce rinsewater  flows  and  recover  the
chromium. They retrofitted an existing countercurrent rinse tank
with a recirculating spray rinse  system. This new rinse system
reduces rinsewater  usage sufficiently enough to  allow all the
rinsewater  to  be used for plating bath makeup.  A pump recir-
culates  rinsewater  through eight high-velocity  spray  nozzles
located  around  the perimeter  of the  rinse tank.  The pump is
activated by a  foot pedal as parts are  lowered into the empty tank.
Clean  rinsewater is available  via a hand-held sprayer.  After
repeated use,  a portion  of the rinsewater is pumped through a
cloth filter into the plating tank and  added  to the plating bath to
replace  water  lost through evaporation. Plating baths  are
operated at elevated temperatures to increase the rates  of both
evaporation and plating.
  These modifications  reduced   the use of  fresh water  from
350,000  gal/mon  for countercurrent  rinsing to about  1,200
gal/mon for spray rinsing. Since this amount was less than the
evaporation rate, all of the spray rinse was returned to the plating
bath, resulting in a zero  discharge condition.  A total savings of
approximately $25,000 per year per bath was projected, principal-
ly due to reduced industrial wastewater treatment costs.
  Without drag-out to aid in removal of contaminants from the
bath, a  cleanup  process  was required to reduce the need  for
plating bath dumps. An electrolytic  bath purification system was
installed to continuously remove cations  from the chromium
plating solution. The system uses cathodes contained within mem-
brane modules to selectively precipitate cation impurities from the
plating solution and anodes to oxidize trivalent chromium to hex-
avalent  chromium.  Hexavalent chromium  ions  remain on the
anode side of the membrane and are returned to the plating bath.
  The purification system did not effectively remove contamina-
tion from the chromium plating bath during a trial run. The
system experienced a failure of the membrane modules caused by
a supplier's change  of material.  Replacement of the  membrane
modules is expected to rectify the problem, but further testing is
required before this technology  can  be recommended  at  other
DOD facilities.
  NCEL implemented these process modifications at Pensacola
as a prototype system. Although NCEL has proven that the spray
rinse system is effective in removing drag-out, plating personnel
remain skeptical about the effectiveness of the  system. Despite
resistance  from platers, there are plans to construct a  permanent
spray rinse and bath purification system because of the prospects
of reduced wastewater flows and treatment costs.
  At other military  installations where the spray system was im-
plemented, the rinsewater modifications were  received  more
favorably  because an extensive training program was provided
prior to startup. In  addition, engineering and management per-
sonnel have been  very supportive of  the process modifications.
That these plating modifications have succeeded is due in large
part to the dedication of C. Carpenter of NCEL, who originated
the  new system,  diligently supervised  its implementation and
remained available for ongoing consultation.

ELIMINATION OF SOLVENT CLEANING
  Prior  to servicing, tactical vehicles  and  equipment  used  at
Army bases typically are washed and  cleaned  at a common
washrack located at  individual motor pools. Each base as 2 to 45
washing locations with a total of 30  to 80 washracks.  Exterior
washing involves removing road dirt and sediment from tracked
and wheeled vehicles. Detergents and solvents sometimes are used
to assist in exterior cleaning.
  Scheduled maintenance of tracked vehicles usually is preceded
by removing the engine from the vehicle and cleaning both the
engine and its compartment. Cleaning prior to servicing often
removes large quantities  of petroleum, dirt  and vegetation.
Solvents regularly are used to assist in the cleaning operation.
  Vehicle washing produces a large volume of wastewater, which
is principally contaminated with soils and minimal concentrations
of oils and organic material. Maintenance produces low  flows of
wastewater that are contaminated heavily  with  oils,  greases,
solvents and  other  organic contaminants. Conventional Army
practice has been  to perform both washing and maintenance on
open wash stands, with the resulting large flow of contaminated
wastewater discharging to  the base  stormwater system.  Many
facilities were having difficulty meeting wastewater discharge per-
mit requirements. In addition to a lack of wastewater  treatment,
other deficiencies of the combined washing facilities include in-
adequate water pressure, ineffective  solvent and oil collection
facilities and undependable steam cleaners.
  Planning for process modifications to reduce water and solvent
use began in 1974  at the Corps of Engineers Construction
Engineering Research Laboratory (CERL), Champaign, Illinois.
J. Matherly of CERL developed the concept of segregating exter-
nal vehicle cleaning  from maintenance servicing so  that the
resulting two  waste streams can be treated separately.  It was Mr.
Matherly's influence that promoted the planning and installation
of the segregated cleaning and maintenance facilities at Fort Polk
and Fort Lewis.
  Fort Polk,  Louisiana, has two central vehicle wash facilities,
each consisting of a large washing basin or "bird bath." Two
staggered rows of 24-in. steel pipes were installed at the bottom of
each lane. This configuration causes a teetering action when a
448
       WASTE MINIMIZATION PROGRAMS

-------
tank or other tracked vehicle drives through. The tracks extend
through their entire range of motion as they move over the cor-
rugations on the bottom of  the bird bath.  This movement
dislodges caked-on soil. Water monitors (spray cannons) provide
secondary cleaning. Facilities are provided with hoses to clean the
insides of vehicles after they exit the bird bath. Waste washwater
is treated in a sedimentation lagoon and reused.
  J. Kelley of Fort Polk developed this simple concept for clean-
ing heavily soiled vehicles and implemented its use at his facility.
Aiding in the adoption of the process was a significant reduction
in manpower.  At the old washracks,  it took  an estimated 7
manhours to wash a tracked vehicle. At the central wash facilities,
only 0.5  manhours were needed to wash a tracked vehicle.
  At Fort Lewis in Washington, D. Hanke, Chief of the Sanita-
tion Branch, implemented the program to segregate cleaning and
maintenance facilities in response to notices of violations caused
by the discharge of polluted stormwater from the facility.
  In contrast to  Fort  Polk,  which uses predominantly tracked
vehicles and has heavy soils,  Fort Lewis has light soils and uses
mainly wheeled vehicles. For these local conditions, central ve-
hicle wash facilities were developed that consisted of individual
wash stations with high-pressure, high-volume (30 gal/min of 90
lb/in.2 water) hoses. These facilities met local needs better than
the bird bath installation used at Fort Polk.
  Waste washwater is reused after treatment  in a simple gravity
(API) oil/water separator and intermittent sand filters. At one of
the vehicle wash facilities, water is not recycled but is discharged
to the sanitary sewer instead.
  With the lower water pressure in  the old  system, cleaning a
vehicle required approximately 2 hrs. With the new high-pressure
system, a tracked vehicle can be washed in approximately 20 to 30
min, and a wheeled vehicle can be washed in 15 to 20 min.
  At Fort  Lewis, facilities were designed to  provide a covered
location  for scheduled maintenance  to exclude  rainwater and
thereby limit the amount of solvent- and oil-laden wastewater.
High-pressure,  hot water cleaners successfully remove oil and
grease  from engine compartments,  eliminating the  need for
solvents  and detergents  which promote oil  emulsification and
complicate  treatment. The resulting low volume  of  oil-laden
wastewater  then  is  treated in gravity oil/water separators and
discharged to the sanitary sewer system.
  The combination of reduced solvent use, separation of exterior
cleaning  from  vehicle  maintenance, installation of  oil/water
separators and recirculation of washwater has led to a 90 to 95%
reduction in the amount of  contaminants  being discharged
through the storm sewers to surface water at Fort Lewis.

CONCLUSIONS
  As demonstrated in the discussions of plastic media paint strip-
ping at  Hill AFB,  hard chromium  plating  at  the Pensacola
NARF, solvent recovery at Robins AFB and at Norfolk Naval
Shipyard and solvent  cleaning  elimination at Forts Polk and
Lewis, a number of features were common to the industrial pro-
cess modifications that were successfully implemented.
  Production people were enthusiastically and actively involved.
Attaining positive involvement usually required that the modifica-
tion result in some production benefit, such as reduced manpower
requirements or simplification of the process; the change did not
have  an adverse effect  on  product quality;  and  the change
preferably had a beneficial effect on the end product. Care was
taken to tailor the modification to the individual facility. During
design and  installation,  operations and  production personnel
were asked to provide input in order to inspire  them to adopt the
process change.
  A "champion" guided the project, overcoming developmental
problems and the inertia that protects existing processes, especial-
ly those that function, even though they may product undesirable
wastes.
  Support was provided at a sufficiently high level in the chain of
command to influence  production and  environmental policy
decisions.
  The technologies tended to require  evolutionary rather  than
revolutionary changes. That  is,  off-the-shelf equipment  was
adapted to a new application, and special or complex equipment
was avoided. Successful modifications were straightforward  and
simple to  operate, requiring minimal training  for personnel un-
familiar with the technology involved. Process  reliability  was
high, and  maintenance requirements were minimal.

ACKNOWLEDGEMENTS
  This  paper  was  supported by the Defense Environmental
Leadership Project and the U.S. Army Corps of Engineers under
contract number DAC  A87-84-C-0076.  The  views expressed
herein are the private views of the authors and  are not to be con-
strued as official or as reflecting the views of the Department of
Defense.
REFERENCES
1. Higgins, T.E., "Industrial Processes to Reduce Generation of Haz-
   ardous  Waste at DOD Facilities, Phase 1  Report,"  prepared by
   CH2M  HILL for the  Defense Environmental Leadership Project,
   Washington, D.C. DTIC No. AD-A159-239, July 1985.
2. Higgins, T.E., "Industrial Processes to Reduce Generation of Haz-
   ardous  Waste at DOD Facilities, Phase 2  Report,"  prepared by
   CH2m HILL for the  Defense Environmental Leadership Project,
   Washington, D.C. DTIC No. AD-A159-239,  July 1985.
3. Higgins, T.E., Fergus, R.B. and  Desher, D.P., "Evaluation of In-
   dustrial Process Modification  To Reduce Hazardous Wastes in the
   Armed  Services," paper presented at  the Purdue Industrial Waste
   Conference, Lafayette, IN, May 1985.
4. Bee, R.W., et al., "Evaluation of Disposal Concepts for Used Solvent
   at  DOD Bases," The Aerospace Corporation Report No. TOR-
   0083(3786)-01, Feb. 1983.
                                                                                    WASTE MINIMIZATION PROGRAMS    449

-------
                                                1986  EXHIBITORS'   LIST
 Allied Healthcare Products, Inc.                  227
 1720 Sublette Ave.
 St. Louis, MO 63110                   314/771-2400
 Allied Healthcare Products is a manufacturer of patient
 care products and medical gas delivery system products.
 We are now offering a hazardous waste compactor, the
 Red Bag Station, to improve efficiency in handling costs
 of hazardous/infectious medical waste, as well as reduc-
 ing  the potential of nosocomial infection.

 Analytical Instrument Development, Inc.           134
 Rt.  41 & Newark Rd., Box 689
 Avondale, PA 19311                   215/268-3181
 Manufacturer of portable instrumentation for the deter-
 mination of trace organic materials in the environment.
 Portable Gas Chromatograph with the Electron Capture
 Detector for the analysis of PCBs in  soil and  water.
 Model 590 OVM/GC for total organic vapors (OVM) or
 specific materials (GC) in air will also be shown. Other
 AID instrumentation for on-site organic measurements
 at  hazardous  waste sites  and  leaking underground
 storage tanks (LUST) will also be on display.
 BCM Engineers, Inc.
 P.O. Box  1784
 Mobile, AL 36633
                                               116
                                      205/433-3981
 BCM provides environmental guidance to plan for and/
 or demonstrate compliance with environmental laws and
 regulations.  BCM also provides  environmentally and
 economically sensible solutions to groundwatcr or soil
 contamination problems.  BCM provides management,
 engineering,  hydrogeological, air  quality, remedial ac-
 tion and EPA- and State-certified laboratory services for
 hazardous waste projects.
 BONDICO, Inc.
 2410 Silver St.
 Jacksonville, FL 32206
                                               325
                                      904/358-2602
 BONDICO,  Inc. has introduced a 90-gallon container
 system designed for transportation, storage, treatment
 and disposal of hazardous and toxic materials. A dual
 laminate composite of polyethylene and fiberglas, the
 container provides superior safety and extraordinary
 cost-effectiveness with multiple reuse. Manufactured in
 Jacksonville, FL, the unit may be  utilized as a salvage
 container, on-site storage  container, or encapsulate via
 its innovative on-site sealing system  (patent pending).
 BONDICO's container is  rustproof,  leakproof, corro-
 sion resistant, light-weight, nestable and reusable.

 Berghof/America, Inc.                          244
 27 Main St.
 Raymond, NH 03077                   603/895-4757
 Exhibiting the widest range of TEFLON products now
 available for research and induslry including:  Teflon-
 lined High Pressure Autoclaves and  Digesters, Teflon
 Acid Purifying Apparatus, Teflon Labwarc,  Fluid Flow
 Systems with Valves,  Fittings,  an all Teflon Pump and
 much more. Custom Fabrication a  Specialty.
Black SVealch                                 126
1500 Meadow Lake Pkwy.
Kansas City, MO 64114                 913/339-2000
Complete Hazardous  Wasle Management Services in-
clude:  RCRA Part B permit applications;  Irealmcnt,
 storage and disposal facility design; regulatory analysis;
 facility inspections;  underground storage  tank  design
 and evaluation; and  facility remediation services. These
 services  have  been  provided  for municipal,  state and
 Federal government  and private industries.

 Bryson Industrial Services, Inc.                   302
 411  Burton Rd.
 Lexington, SC 29072                   803/359-7027
 Bryson Industrial Services, Inc  is a hazardous waste
 management company.  We provide  consultation  and
 management services 10 customers on methods of reduc-
 ing, handling and disposing of Ihcir hazardous waste. In
 addition we provide secure permitted transportation and
 fully trained and experienced project  I cams for on-site
 service needs.

 Bulk Lift International, Inc.                      330
 455  Maple Ave.
 Carpentersville. IL 60110               312/428-6059
 The Bulk Life bag  is an innovative approach to cost-
 efficient material handling. Our bag  is a woven poly-
 propylene  bag  designed  for  shipping  dry  fiowable
 cargos. It is an efficient alternative to multi-wall bags,
 corrugated, rigids and bulk. The Bulk  Lift bag is de-
 signed with patented features  for product protection,
 double drawstring  closures  reduce product  pollution
 risk, and safety bands  with reinforcement. Our bags
 range from 1000 to 8000 pounds capacity for con-
 taminated soil and other hazardous wastes.

 CH2M HILL, Inc.                               103
 229 Peachtree St.. N.E., Suite 300
 Atlanta. GA 30303                      404/523-0300
 CH2M HILL is a consulting engineering firm with over
 40 offices throughout the world. With extensive experi-
 ence in hazardous waste  site investigation and cleanup,
 CH2M HILL provides services to both public and pri-
 vate sector clients. CH2M HILL is the primary contrac-
 tor lo EPA for Superfund in the western United States.

 California Analytical Laboratories, Inc.           204
 2240 Dabney Rd.
 Richmond, VA 23230                  804359-1900
 California Analytical Laboratories, Inc.  and Cal Lab
 East, Inc. provide quality environmental analyses from
 laboratories in West Sacramento and Richmond, Vir-
 ginia. Governmental, industrial and private clients pro-
 fil from Cal Lab's participation in EPA's Contract Lab
 Program. The Cal Lab family is dedicated (o providing
 prompt, high quality  pharmaceutical and environmental
 services,

Camp Dresser & McKcc Inc.                   127-129
 1945  The Exchange,  NW, Suite 290
Atlanta, GA 30339                     404/952-8643
Camp Dresser  & McKec Inc.  provides comprehensive
engineering and consulting services for hazardous waste
managcmcnl, including remedial investigations, feasibil-
ity studies, remedial design, groundwatcr modeling and
remediation, and RCRA  permitting.

Carbon Air Services, Inc.                         420
P.O. Drawer 5117
Hopkins, MN 55343                    612/935-1844
CarbonAir Services is a  groundwatcr decontamination
contractor providing treatment design and turnkey in-
stallation for removal of  organic and inorganic con-
taminants in surface, groundwater and process streams.
Treatment alternatives include airslrippmg. carbon ad-
sorption, metals precipitation and/or oil-water separa-
tion. Emphasis is on short-term leasing and  total  ef-
fluent responsibility.  Equipment can be purchased.

Chem-Mel Services                              321
18550 Allen Rd.
Wyandotte, MI 48192                  313/282-9250
Chem-Met Services provides environmentally safe treat-
ment for liquid  and  solid  wastes. Our process is eco-
logically  sound and  economically efficient.  We  have
been processing waste streams for industry since 1966.
Client-oriented,  we know that  dependable service  is
essential. Our waste disposal  treatment is proven sound
for our environment. We slay current wilh changing
government regulations. Analysis of all hazardous waste
is required and  maintained.
Dixie Poly-Dram Corp.
P.O. Box 597
Yemassee. SC 29945
                                           214-216
                                                                                                                                                   803/589-6660
Dixie Poly-Drum Corp. offers a complete line of DOT-
approved closed head and open top polyethylene ship-
ping drums. The "Super  Shipper"  line of closed head
polydrums is available in a 15. 20, 30. 35 and 55 gallon
size  The Plasti-Drum line of  open top polydrums is
available in a 14, 16, 30, 45 and 55  gallon size.
Duke Univenil) Press
6697 College Station
Durham, NC 27708
                                               227
                                      9I9'684-2173
Duke University Press was founded in 1921 as the first
scholarly publishing house in the South and became a
charter member of the American Association of Univer-
sity Presses in 1937. Duke currently publishes more than
40 books and  11 journals  annually and has some 300
titles in print.

E.NSCO                                        401
1015 Louisiana
Little Rock, AR 72202                  SOI/37S-8444
Environmental Services  •  Hazardous & Toxic Waste
Incineration •  On-Sile Hazardous Waste Incineration •
Environmental Laboratory Services • Geotechnical &
Hydrological Assessment • PCB Transformer Decom-
missioning and Rctrofill
 ERT                                          203
 696  Virginia Rd.
 Concord. MA 01742                     617/369-8910
 Full-service environmental  and engineering consulting
 firm with offices nationwide. Inactive site management,
 UST consulting,  closure and  plant decommissioning,
 land transfer, RCRA/CERCLA compliance, air toxics
 consulting and other waste-management related serv-
 ices.

 Earth Technology Corporation, The              438
 3777 Long Beach Blvd.
 Long Beach, CA 90807                  213/595-6611
 The Earth Technology Corporation provides environ-
 mental, geotechnical, gcosciemific and engineering con-
450     EXHIBITORS' LIST

-------
suiting services to industrial and governmental clients.
Our services include: remedial investigations; remedial/
corrective action; design and engineering; geotechnical
investigations; waste  stream reduction  and recovery;
facility closure; environmental  auditing/compliance as-
sessment;  hazardous waste permitting; laboratory and
special technical services.

Ecology and Environment, Inc.            121-123-125
P.O. Box D
Buffalo, NY 14225                     716/688-7876
E & E offers the complete range of services required for
generators, transporters and disposers  of hazardous
materials. Our staff of 500 scientists and engineers pro-
vides   hydrogeological  studies;  site  investigations;
remedial plans and specifications;  engineering designs
and project management;  hazards and risks analyses;
RCRA and LUST compliance programs; spill emergency
response; and extensive analytical laboratory services.

Engineering-Science                             310
57 Executive Park South, Suite 590
Atlanta, GA 30329                     404/325-0770
Engineering-Science (ES) is a major national and inter-
national  engineering  firm  with a  major specialty in
hazardous waste engineering and management. ES pro-
vides the complete range of services including health and
safety,  site  investigations,  remedial  investigation/
feasibility studies, hydrogeological, geotechnical  and
geophysics,  site cleanup,  analytical  services  and  air
pollution monitoring.
 Enviresponse, Inc.                           115-117
 110 South Orange Ave.
 Livingston, NJ 07039                   201/533-7289
 Enviresponse, Inc. offers a wide range of technologies
 and expertise relevant to the complex problems sur-
 rounding the treatment of hazardous  and toxic waste.
 Services range from consulting to design and construc-
 tion.  Enviresponse, Inc.  is a subsidiary of the Foster
 Wheeler Corporation.
 Environmental Protection Systems                206
 7215 Pine Forest Rd.
 Pensacola, FL 32506                    904/944-0301
 As a professional engineering and analytical firm, EPS
 provides the entire spectrum of services which are need-
 ed to support the hazardous materials  and hazardous
 waste  industry.  Our  master program  of  permitting
 negotiation and regulatory liaison is designed to support
 our clients needs with turnkey in-house services and con-
 tractors management. We can also provide sample pick-
 up and delivery service from sample sites with rapid
 turn-around of analytical results.  From operating per-
 mits to material characterization  to  feasibility studies
 and remediation, EPS is ready to  become part of your
 team.

 Environmental Quality Labs, Inc.                 422
 6107 E. Ten Mile Rd.
 Warren, MI 48091                      313/757-7970
 Independent  testing  lab—consultants  specializing in
 hazardous waste  sampling  and  analysis, water  and
 wastewater monitoring, along with coal and petroleum
 testing.
Environmental Science & Engineering, Inc.         406
P.O. Box ESE
Gainesville, FL 32602                   904/332-3318
ESE,  a  full-service multidisciplinary environmental
engineering firm, has performed work at more than 120
hazardous waste sites, including 15 CERCLA NCP sites.
Capabilities include:  remedial investigations,  feasibility
studies, QA/QC plans, safety and health planning and
monitoring,  community relations,  analytical services,
and expert witness testimony.
Fred C. Hart & Associates
530 Fifth Ave.
New York, NY 10036
                                                224
212/840-3990
Fred C.  Hart Associates is an environmental manage-
ment consulting firm providing technical, economic and
management service to industry. Specific services in-
clude the following: Remedial Investigations and Feasi-
bility Studies • Construction  Management Services  •
Remedial Engineering  and  Design  Services •  En-
vironmental  Audits  and   Risk  Assessments   •
Underground Storage Tank Management and Safety.

GA Technologies, Inc.                           139
P.O. Box 85608
San Diego, CA 92138                   619/455-3045
Circulating  Bed Combustor systems. Circulating bed
technology has demonstrated  safe, economical,  com-
plete elimination of hazardous wastes.  EPA require-
ments are met without the use of afterburners and scrub-
bers. In  addition to permanently  installed units, GA
Technologies has developed transportable CBC units for
on-site treatment of corporate/Superfund sites.
GSX Chemical Services, Inc.
60 State St.
Boston, MA 02109
     333-335
     432-434
617/367-8300
Complete chemical  and hazardous waste disposal and
complete solid waste collection and disposal services.

Geo-Con, Inc.                                  210
P.O. Box 17280
Pittsburgh, PA 15235                   412/244-8200
Geo-Con  is a full-service hazardous waste contractor
with capabilities of  total site cleanup including special-
ized types of remedial activities  such as:  soil/sludge
stabilization, geomembrane systems, earthwork,  drum
and tank sampling, assistance  in transportation and
disposal of hazardous wastes and total health and safety
capabilities.

Geonics Limited                                205
1745 Meyerside Dr., #8
Mississauga, Ontario, Canada
L5T1C5                               416/676-9580
Geonics Limited, a manufacturer of electromagnetic
geophysical equipment  used worldwide by  government
and industry for mapping groundwater contaminant
plumes, is proud to  introduce two innovative products.
The first, a data logger and contour plotting system, will
greatly enhance survey data presentation. The second, a
high resolution borehole induction conductivity logger,
will accurately delineate contaminant plume stratifica-
tion through PVC cased,  monitoring wells.

Greenhorne & O'Mara, Inc.                      235
9001 Edmonston Rd.
Greenbelt, MD 20770                   301/982-2800
One of the nation's largest consulting engineering firms,
Greenhorne & O'Mara  combines a wide variety of serv-
ices in the hazardous waste area with special emphasis
on remote sensing and the natural sciences. Founded in
1950, the  firm has 13 offices nationwide.

Groundwater Decontamination Systems, Inc.      326
Suite 210, 140 Rt. 17 North
Paramus,  NJ 07652                     201/265-6727
Introducing Groundwater  Decontamination Systems
(CDS), the unique new system for soil and groundwater
decontamination. The  GDS process eliminates hydro-
carbon  and  halogenated hydrocarbon contaminants
from the groundwater and from the soil through a pro-
cess of accelerated biodegradation by micro-organisms
existing in the contaminated soil. Patent awarded.

Gundle Lining Systems  Inc.                      209
1340 E. Richey Rd.
Houston, TX 77073                     713/443-8564
Gundle Lining Systems Inc. is recognized as the World
Leader in lining systems, manufacturing and installing
High  Density  Polyethylene from  20 to  100 mil thick.
Also, Driline,  Hyperlastic and Gundnet Draining Net-
ting for leachate systems. For the past 22 years, Gundle
has installed over 3 million square feet of lining systems
throughout  the world.  For more  information,  call
800/435-2008 or 713/443-8564.

HAZCO, Inc.                                500-502
1347 E. Fourth St.
Dayton, OH 45402                     513/222-1277
Turn key supplier of health and safety equipment for all
phases of remedial and emergency response. Specializing
in complete rental and  training packages.
                  HAZTECH
                  5280 Panola Industrial Blvd.
                  Decatur, GA 30035-4013
                                                                 220
                                      404/981-9332
HAZTECH, a hazardous waste contractor, provides the
following  services: contaminated soil  removal;  on-site
water treatment; drum, soil and tank excavation; lagoon
closures and  cleanings;  sludge  solidification;  24-hour
emergency spill response;  surface and  groundwater
treatment; sampling; site  assessment,  including  the
handling  of  shock-sensitive   materials  (Lab  pack
removal).

Hazardous Materials Control              223-225-227
Research Institute
9300 Columbia Blvd.
Silver Spring, MD 20910                 301/587-9390
HMCRI  is a unique public,  nonprofit,  membership
organization  which  promotes  the  establishment and
maintenance of a reasonable balance between expanding
industrial  productivity and an acceptable  environment.
Our goals are met  through  a variety of publications,
conferences, workshops, newsletter, equipment exhibi-
tions and other information dissemination programs.
We provide members and all other interested persons
with a distinctive forum in which they can exchange in-
formation  and  experiences  dealing  with hazardous
materials.  JOIN HMCRI TODAY!!!
                  Hoyt Corporation
                  251 Forge Rd.
                  Westport, MA 02790
                                                                                                                                                               317
                                       617/636-8811
                  Solvent  Vapor Recovery/Air Pollution Control Equip-
                  ment, Distillation Equipment

                  Hydro Group, Inc.                              304
                  P.O. Box 266
                  Linden, NJ 07036                      201/862-7800
                  Hydro Group, Inc. provides comprehensive solutions to
                  groundwater pollution problems—hydrogeological in-
                  vestigations, well drilling, treatment  process and equip-
                  ment design and manufacture, and  system  installation
                  and maintenance. Its wholly owned subsidiary corpora-
                  tion, Ground Water Associates, and  its six operating
                  divisions   provide  Hydro  Group,  Inc.  turnkey
                  capabilities in groundwater pollution.

                  In-Situ, Inc.                                    403
                  P.O. Box I
                  Laramie, WY 82070                    307/742-8213
                  In-Situ, Inc. manufactures computer-automated,  field
                  hydrologic  monitoring  instruments;  software  (both
                  mainframe and IBM PC-compatible) for environmental,
                  hydrologic  and  engineering applications; time-sharing
                  on our VAX 11/780 computer with free access to In-Situ
                  proprietary  software;  geotechnical  and  hydrological
                  consulting services.

                  Inmetco                                       207
                  (International Metals Reclamation Company)
                  P.O. Box 720
                  EllwoodCity, PA16117                 412/758-5515
                  Waste Metal Processing and Refining • Hazardous and
                  Non-Hazardous  Waste Processing • Approved Hazar-
                  dous Waste Processing Facility — #PAD087561015
                                                                                                                                   EXHIBITORS' LIST     451

-------
International Technology Corporation         133-135
23456 Hawthorne Blvd.                      232-234
Torrance, CA 90505                    213/378-9933
International  Technology Corporation  is the nation's
largest firm devoted exclusively to the management of
hazardous materials,  with  expertise in such  areas as
nuclear decontamination  services, emergency  response
and  remedial action  involving dioxins  and PCBs, la-
goons and sludges, hazardous  material  transportation,
treatment, disposal, and resource recovery, geotechnical
profiling and modeling, analytical services, health and
safety training and environmental permitting.

Labelmasler                                    141
5724 N. Pulaski Rd.
Chicago, IL 60646                      312/478-0900
Supplier of labels, forms, placarding, publications and
packaging for the safe transport of hazardous malcrials.

Liquid Waste Technologies, Inc.                  423
990 North Main
Las Cruces, NM 88001                  505/523-3132
Stabilizing and  solidifying  polymers for organic and
petroleum waste and oil spills. Volume expansion of
treated wastes  is less than  lO^o.  Polymers effectively
solidify waste oils, paint and  oil  sludges,  transformer
and   other oils  contaminated with PCBs,  kerosene,
gasoline and solvents. Safe,  economical and easy to use.

 MUD CAT Div. National Car                    242
 Rental Systems  Inc.
 1411  Perimeter Clr. E., NE
 Atlanta, GA 30346                     404/394-3013
 Manufacture and marketing of MUD CAT Low  Tur-
 bidity Dredges • Marketers of Kenner Billy Goat Cutter-
head Dredges, Vaughan Chopper Pump Lagoon Pump-
ers,  AQUA Dozer Pollution Control Equipment and
 UMI Awuatic Weed  Harvesters.
 Mid-America Environment Service, Inc.           314
 17450 South Halsted St.
 Homewood, IL 60430                  312/957-7600
 Services  include:  hazardous  waste  transportation;
 chemical  treatment  and  neutralization;  soil borings;
 hydrogeologic surveys;  tank  and lagoon cleaning; in-
 cineration; resource reclamation; dewatering; industrial
 maintenance; solidification/fixation; water blasting; site
 closure plans; spill contingency plans; emergency spill
 response; and sampling and analysis.


 Mineral By-Products, Inc.                       446
 777 Franklin Rd., #100
 Marietta, GA 30067                    404/424-0247
 POZZALIME™  calcic  powder  is  used  for  sludge
 solidification  and pH adjustment.  POZZALIME™
 calcic  powder is available in  bag or bulk for truck or
 barge  shipment from various locations throughout the
 United States.

 National Lime Association                       211
 3601 N. Fairfax Dr.
 Arlington. VA 22201                    703/243-5463
 Effective use of  lime  for treating hazardous  waste.
 Neutralization of acids,  precipitation of heavy metals,
 solidification of toxic elements. Low cost and recom-
 mended by U.S. EPA.

 National Water Well Association                  102
 500 W. Wilson Bridge Rd.
 Worthington, OH 43085                 614/846-9355
 The  National Water  Well Association,  the  nation's
 leading organization committed  to  the protection of
groundwater, offers information on its 70 educational
 programs, its more than 100 scientific publications, its
own research studies and membership with its scientific
division,  the  Association of Ground Water Scientists
and Engineers.
                                                        O.K. Materials Co. (OHM)                       104
                                                        16406 U.S. Rt. 224 East
                                                        Findlay. OH 45840                     419/423-3526
                                                        OHM provides a wide range of services: groundwater
                                                        recovery,  on-site treatment,  facilities decontamination,
                                                        hazardous waste  cleanup, technical  advisory services,
                                                        sampling and analytical services, management of under-
                                                        ground storage tanks, emergency response, surface im-
                                                        poundment restoration—from response center in Mas-
                                                        sachusetts, New  Jersey,  Virginia,  Florida, Georgia,
                                                        Alabama, Louisiana,  Texas,  Ohio,  Michigan,  Minne-
                                                        sota, Missouri and California.

                                                        Orlando Laboratories, Inc./                      404
                                                        Global Resource Planning, Inc.
                                                        P.O. Box 19127
                                                        820 Humphries Ave.
                                                        Orlando, FL 32814                     305/896-6645
                                                        Orlando Laboratories is an  analytical laboratory  pro-
                                                        viding services to the public, engineering firms, govern-
                                                        ment agencies, industries and consultants. We endeavor
                                                        to utilize only the most modern analytical equipment in
                                                        the  analysis of our client's samples.  Our expertise in-
                                                        cludes hazardous waste facility monitoring, gas chroma-
                                                        tography/mass spectromctry confirmation and planning
                                                        services.

                                                        P.E. LaMoreaux & Associates                    147
                                                        P.O. Box 2310
                                                        Tuscaloosa, AL 35403                  205/752-5543
                                                        P.E. LaMoreaux  and  Associates,Inc.  (PELA),  hy-
                                                        drologists, geologists and environmental scientists, offer
                                                        hydrological, geological, environmental and hazardous
                                                        waste consultation services throughout the world. Other
                                                        services include  emergency  spill  response, sampling,
                                                        laboratory analysis,  development of monitoring  pro-
                                                        grams and installation of wells, reclamation, govern-
                                                        ment permitting, court testimony and graphic and com-
                                                        munication programs.
Photovac International, Inc.                      442
739B Park Ave.
Huntington, NY 11743                  516/351-5809
Photovac manufactures ultra high sensitivity analyzers
for the detection of trace levels of airborne and water-
borne contaminants; these have gained wide acceptance
as a practical screening tool in hazardous waste sites.
This year we shall  be demonstrating our new  10S70
model, which has fully automated sampling  with com-
pound  identification  and  quantification,  automatic
calibration. Time Weighted Averaging and Level Alarm.
The 10S70 has an RS232 interface with autodial modem
for telephone  communication. Also featured will be
Photovac's  TIP—our  very economical,  lightweight
photoionization survey device.
Poly-America
2000 W. Marshal]
Grand Prairie, TX 75051
                                               303
                                      800/527-3322
Poly-Flex  is a polyethylene  gcomcmbranc offering  a
cost-effective  method  for  lining  hazardous  waste
disposal facilities and preventing groundwaier pollution.
Poly-America, one of the largest and most technologi-
cally equipped film extrusion facilities in the United
Stales, has the world's largest  blown film line which pro-
duces a 25-foot wide seamless geomembrane.

Pressure Filtration Specialists, Inc.                 107
P.O. Box 686
Torrington, CT 06790        203/489-1221 or 485-2524
Pressure  Filtration  Specialists,  Inc.  (PFS)— Volume
Reduction of Liquid Waste. PFS' mobile, self-contained
trailer-mounted filter presses reduce generator  liquid
sludge volume by producing filter cake of 35-60% solids
by weight. This reduced volume offers substantial sav-
ings to generators  on transportation and disposal costs.
On-site sample testing available upon request.
                                                        R.E. Wright Associates, Inc.                  231-233
                                                        3240 Schoolhouse Rd.
                                                        Middletown, PA 17057                  717/944-5501
                                                        R.E. Wright Associates, Inc. designs and manufactures
                                                        groundwater cleanup and oil spill recovery equipment,
                                                        including  the  Auto-Skimmer,  Air-Stripping Towers,
                                                        Water  Table Depression  Pumps and  Liquid  Interface
                                                        Samplers.  The Auto-Skimmer  automatically recovers
                                                        subsurface spills of floating hydrocarbons from both
                                                        large and  small diameter wells.  This equipment  will be
                                                        exhibited along with a  working  scale model of an Air-
                                                        Stripping  Tower system for removing volatile organic
                                                        compounds from water.
                                                        RETROTEX
                                                        1700 Gateway Blvd., SE
                                                        Canton, OH 44707
                                                                                                       227
                                       216/453-4677
                                                        RETROTEX specializes in the turnkey replacement of
                                                        PCB transformers and related equipment. We also have
                                                        a complete substation maintenance organization to per-
                                                        form all  facets of transformer and switchgear repair,
                                                        testing, calibration and maintenance. We are a complete
                                                        organization devoted to the complete PCB elimination
                                                        through equipment replacement and disposal.
                                                        Radian Corporation
                                                        P.O. Box 9948
                                                        Austin. TX 78766
                                                                                                       122
                                                                                              512/454-4797
Technology-based company which provides professional
services and specialty products to government and in-
dustry.  Full range of services offered in environmental
services, and specifically in the area of solid and hazar-
dous waste management. These services include the
areas of permitting; remedial action planning/imple-
mentation; soils,  water and  waste analysis; and waste
management facilities design.

Recra Research, lac.                              110
4248 Ridge Lea Rd.
Amherst, NY 14226                     716/838-6200
Environmental services including: Geologic/Hydrogco-
logic Investigations,  chemical and biological, analyses
(including  bench  pilot and feasibility  studies),  field
sampling, monitoring and analysis (including experi-
enced remedial teams and  fleet of mobile laboratories),
hazardous  waste management, regulatory compliance,
environmental research and  development and concep-
tual engineering. Permit assistance and chemical control
management.

Rcxnord                                       144
5103 West Beloit Rd.. Bldg.  N.
Milwaukee, WI 53201                    414/643-3083
Rexnord offers broad services to solve RCRA problems
through  its Envirex  subsidiary and its  EnviroEnergy
Technology Center.  Services  include:   In-plant  En-
vironmental Assessments,  SPCC Plans, Contingency
Plans, Right-to-Know Programs, Laboratory-analytical
and Bench Treatability Services, Pilot Plant Systems and
Operations, Underground  Tank  Management/Closure
Programs,  Full-scale   systems-Phys/Chem   and
Biological,  Mobile Van Dewatering Services, "Van of
the '90s" Groundwaier Treatment System.
Rocky Mountain Analytical Laboratory            322
5530 Marshall St.
Arvada, CO 80002                      303/421-6611
Located  near  Denver, Colorado,  Rocky  Mountain
Analytical  Laboratory  (RMAL) provides  analytical
chemistry services in a variety of environmental areas,
but specializing in RCRA and CERCLA investigations.
While  RMAL is one of the major contract laboratories
supporting EPA's Supcrfund investigations, the majori-
ty of laboratory work is for industrial clients. RMAL
has received  national recognition for its work  in  two
areas: analysis of the Appendix Hazardous Constituents
and support of RCRA activities at petroleum refineries
(waste delisting, Part B, LTD, etc.). The laboratory is
452     EXHIBITORS' LIST

-------
staffed with over 100 professionals and has all of the
analytical instrumentation to support virtually any type
of environmental project.

Roy F. Weston, Inc.                              Ill
Weston Way
West Chester, PA 19380                 215/692-3030
Roy F, Weston, Inc. (Weston) provides hazardous waste
remedial action services in design, permitting, investiga-
tion, construction management  cleanup  and closures.
Support  services  include:  emission   monitoring,
analytical  laboratories,  data  management systems,
emergency  response,  health and safety  training, and
right-to-know training.

Schlegel Lining Technology, Inc.                   238
P.O. Box 23197
Rochester, NY 14692                    716/427-7200
Schlegel  Lining Technology, Inc. manufactures/installs
HOPE liner, SCHLEGEL®  sheet, available in thick-
nesses up to 100 mils.  A single SCHLEGEL® sheet: 34
ft wide, up to 900 ft long, reduces field seams. No fac-
tory seams. Sheets are joined with patented extrusion
welding producing 100% bond strength. Unmatched in
chemical resistance, puncture/tear resistance, strength
under physical loadings. Schlegel  Lining Technology has
15 years  experience in design/manufacture and installa-
tion of containment systems. Laboratory and design ser-
vices available.

Sevenson Containment Corporation              343
2749 Lockport Rd.
Niagara Falls, NY 14302                 716-284-0431
Sevenson Containment Corporation provides 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 design & construction; sludge solidification &
fixation; slurry wall & trench construction; waste isola-
tion & stabilization; waste recovery & treatment; drum
removal; and transportation & disposal.
 Shirco Infrared Systems
 1195 Empire Central
 Dallas, TX 75247
                                                316
                                       214/630-7511
 Shirco, Inc., incineration systems featuring the use of in-
 frared heating and conveyor  belt transport  of waste
 materials  through an  efficiently insulated modularly
 constructed waste disposal system. Since no fossil fuel is
 required, the reduced gas flow is economically treated to
 meet requisite emission standards. Systems are excellent
 for intermittent operation and have transportable cap-
 ability. Shirco Portable Pilot Test Unit is available for
 on-site testing at your facility.

 Soil & Material Engineers, Inc.                     114
 1903 Harrison Ave.
 P.O. Box 609
 Gary, NC27511                         919/481-0397
 Soil &  Material Engineers, Inc. is a multi-disciplinary en-
 vironmental engineering consulting firm with over 800
 employees  based  in the Southeast and Midwest. Our
 professional staff of engineers,  hydrogeologists and
 scientists provide environmental services to industry and
 government in the following areas: Groundwater Assess-
 ment   •  Underground  Tank  Management  •  RI/FS
 Studies • Regulatory Liaison & Permitting • Remedial
 Action--Design.

 SilidurN.A. Company                           339
 2200 N. Ridge Rd. East
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. If functions as a
flexible concrete retaining wall that turns  into a living
wall.
Soiltest, Inc.                                     440
(Soiltest/Environmental Division)
P.O. Box 931
Evanston, IL 60204                      800/323-1242
For sampling  of soils and water, the Environmental
Division of Soiltest, Inc. offers simple stainless steel soil
samplers up to sophisticated systems capable of remov-
ing a  specimen in a sealed container to retain volatile
gases and liquids. Water sampling equipment selection is
the largest in the U.S. Site investigation equipment in-
cludes  resistivity and  conductivity meters, lysimeters,
tensiometers and a full range of portable instruments.
SolidTek Systems, Inc.                            105
5371 Cook Rd., P.O. Box 888
Morrow, GA 30260                     404/361-6181
Full spectrum of  hazardous  waste  treatment  and
disposal services.  Custom chemical products and serv-
ices for chemical fixation and solidification to meet the
latest requirements.  Regional TSD  service.  Dedicated
systems  at  generator  locations.  Mobile  service  for
closures,  remediation projects.  Advanced  secure land-
fill.
Teledyne Geotech                                323
3401 Shiloh Rd.
Garland, TX 75041                      214/271-2561
Leakage Alarm System  for  Pipelines and  Pollutants
(LASP). The LASP represents a method for early detec-
tion of small subsurface leaks from storage tanks and
pipelines.  The location and size of a leak can be easily
determined. High  detection sensitivity with automatic
self-checking and  a  large signal-to-background  level
ratio provides a highly reliable system.
 The Tensar Corporation
 1210 Citizens Pkwy.
 Morrow, GA 30260
                                                345
404/968-3255
 Drainage Nets: Landfill Leachate Collection • Landfill
 or Surface Impoundment Leakage Detection • Landfill
 Cover Drainage • Landfill or Surface Impoundment Gas
 Drainage • Waste Pile Liquid Drainage • Landfill or
 Surface Impoundment Groundwater Underdrain. Slope
 Reinforcement Geogrids. Roadway and Base Reinforce-
 ment Geogrids.

 Trofe Incineration, Inc.                          332
 Trofe Industrial Park
 Mt. Laurel, NJ 08054                   609/235-3030
 Trofe Incineration, Inc. is an engineering and construc-
 tion firm that manufactures, markets and installs high-
 temperature incineration and  energy recovery systems.
 Three types of modular combustion systems are offered:
 a solids system, a liquids system or a single all-purpose
 system. The systems are designed with flexibility for ex-
 pansion upon need.

 U.S. Pollution Control, Inc.                  425-427
 2000 Classen Ctr.
 Oklahoma City, OK 73106               405/528-8371
 Waste disposal and transportation of RCRA and PCB
 waste. Remedial resonse and cleanup.  Commercial
 analytical  laboratory  and   state-of-the-art   recycling
 facilities.
                 WAPORA, Inc.                                 204
                 1555 Wilson Blvd., Suite 700
                 Rosslyn, VA 22209                     703/524-1171
                 WAPORA is a leading engineering and environmental
                 consulting firm that has served industry and government
                 for  over 15 years. The firm offers a wide range of ex-
                 perience  in  hazardous  waste site  investigation  and
                 remediation, including groundwater cleanup. Our cor-
                 porate  headquarters  are located in Washington,  DC,
                 and  we  maintain  regional  offices  in  New  York,
                 Philadelphia, Cincinnati, Chicago, Atlanta and Dallas.
                 Waste-Tech Services, Inc.
                 445 Union Blvd., #223
                 Lakewood, CO 80228
                                                                 421
                                       303/987-1790
                 Waste-Tech Services, Inc. provides a total service pro-
                 gram designed to destroy organic hazardous waste at a
                 client's site by using our proven and patented fluidized
                 bed  incineration technology.  Our  patented  system
                 handles gases, liquids, slurries, sludges and solids. WTS,
                 Inc. secures and holds all permits, provides all com-
                 pliance reports and interfaces with  all  regulatory agen-
                 cies.
                 Watersaver Company, Inc.
                 P.O. Box 16465
                 Denver, CO 80216
                                                                 334
                                       303/238-1818
                 Watersaver Company provides the world's most reliable
                 membrane  lining  systems.   Our  products  meet  all
                 regulatory requirements for zero seepage. With over 30
                 years' experience and 300 million square feet of suc-
                 cessful  installation,  Watersaver  leads  the  way  with
                 proven  products and systems that work correctly and
                 continuously. Linings and floating covers of Hypalon,
                 CPE and PVC up to 60 mils.
                 Wilson Laboratories
                 525 N. 8th St.
                 Salina, KS 67401
                                                                 217
                                       913/825-7186
Full service analytical laboratory specializing in en-
vironmental monitoring and the analyses of hazardous
waste samples. Expertise includes GC, GC/MS, HPLC
and Industrial Hygiene.

Woodward-Clyde Consultants                     305
P.O. Box 66317, 2822 O'Neal Ln.
Baton Rouge, LA 70815                 504/291-1873
Woodward-Clyde Consultants is  a nationwide profes-
sional services firm serving clients for over 30 years. Our
practice  includes  the  application  of  knowledge  in
engineering, the earth  sciences  and the environmental
and social sciences. Services  offered include:  waste
management; environmental assessments;  geology and
hydrogeology; hydrology;  site selection studies; oil spill
contingency planning; air and water quality studies; geo-
technical engineering, and risk and decision analysis.

YWC, Inc.                                      431
200 "Monroe Tpke.
Monroe, CT 06468                      203/261-4458
A  multidisciplinary  environmental   laboratory/engi-
neering  consulting firm, YWC, Inc. provides air, soil,
water and waste characterization for full-profile site
assessments. The York Laboratories Division is a par-
ticipant in EPA's Contract Laboratory Program. Addi-
tional services available include contract operation of
wastewater treatment facilities, interim sludge dewater-
ing services, hazardous waste-management and remedial
investigation/feasibility studies.  Growth  has  been
achieved through dedicated personal service with an em-
phasis on quality at most competitive prices.
                                                                                                                                   EXHIBITORS'  LIST     453

-------