-------
NATIONAL CONFERENCE ON
MANAGEMENT OF
UNCONTROLLED
HAZARDOUS
WASTE SITES
NOVEMBER 29-DECEMBER 1, 1982
WASHINGTON, O.C.
AFFILIATES:
U.S. Environmental Protection Agency
Hazardous Materials Control Research Institute
U.S. Corps of Engineers *
U.S. Geological Survey
American Society of Civil Engineers
Association of State and Territorial Solid Waste Management Officials
-------
ACKNOWLEDGEMENT
The National Conference and Exhibition on Management of Uncontrolled Hazardous
Waste Sites was made possible through the involvement of many individuals and organiza-
tions. We would like to express our thanks and appreciation to all who assisted.
U.S. Environmental Protection Agency
Hazardous Materials Control Research Institute
U.S. Corps of Engineers
U.S. Geological Survey
American Society of Civil Engineers
Association of State and Territorial Solid Waste Management Officials
The Program Committee is comprised of knowledgeable individuals cooperating to pro-
duce an effective and informative program. These individuals are:
Harold Bernard James Rollo
Hazardous Materials Control Research U.S. Geological Survey
Institute
Don Sanning
Brint Bixler U.S. Environmental Protection Agency
U.S. Environmental Protection Agency
Harold J. Snyder, Jr.
Doug Lament U.S. Environmental Protection Agency
U.S. Corps of Engineers
Wayne Tusa
Hugh Masters American Society of Civil Engineers
U.S. Environmental Protection Agency
The concentrated effort necessary to publish a Proceedings of this size and scope in the
time allotted is certainly 'above and beyond'. Our special thanks to Dr. Gary Bennett, Pro-
fessor of Biochemical Engineering, The University of Toledo, and Hal Bernard, HMCRI
whose editing allowed for a more uniform Proceedings; to the typesetters and graphics
team who managed to meet the impossible deadlines set; and to the staff of HMCRI, in
particular, Beverly Walcoff, Project Manager and Paula Geary, for coordinating the many
aspects and activities of this Conference.
-------
CONTENTS
SITE INVESTIGATION
Electromagnetic Resistivity Mapping of
Contaminant Plumes 1
J.D. McNeill
Field Investigation of an Abandoned
Pesticide Formulation Plant 7
Deborah A. Kopsick
Use of NOT Methods to Detect Buried Containers
in Saturated Silty Clay Soil 12
Robert M. Koerner, Ph.D., Arthur E. Lord, Jr., Ph.D.,
Somdev Tyagi, Ph.D. & John E. Brugger, Ph.D.
Systematic Hazardous Waste Site Assessments 17
R.B. Evans, Ph.D., R.C. Benson and J. Rizzo
Determination of Risk for Uncontrolled
Hazardous Waste Sites 23
Isabell S. Berger
Assessment of Hazardous Waste Mismanagement Cases 27
Wayne K. Tusa & Brian D. Gillen
Electrical Resistivity Techniques for Locating
Liner Leaks 31
Wendell R. Peters, David W. Shultz & Bob M. Duff
SAMPLING & MONITORING
Evaluation and Use of a Portable Gas Chromatograph
for Monitoring Hazardous Waste Sites
Jay M. Quimby, Robert W. Cibulskis &
Michael Gruenfeld
The use of Portable Instruments in Hazardous
Waste Site Characterizations
Paul F. Clay & Thomas M. Spittler, Ph.D.
Analytical and Quality Control Procedures for the
Uncontrolled Hazardous Waste Sites Contract Program ...
D.F. Gurka, Ph.D., E.P. Meier, Ph.D.,
W.F. Beckert, Ph.D. & A.F. Haeberer, Ph.D.
Biological Sampling at Abandoned Hazardous Waste Sites .
Amelia J. Janisz & W. Scott Butter field
Correlation Between Field GC Measurement of
Volatile Organics and Laboratory Confirmation of
Collected Field Samples Using the GC/MS
Extended Abstract
Thomas Spittler, Ph.D., Richard Siscanaw &
Moira Lataille
A Generalized Screening and Analysis Procedure for
Organic Emissions from Hazardous Waste Disposal Sites...
Robert D. Cox, Ph.D., Kenneth J. Baughman &
Ronald F. Earp
The Air Quality Impact Risk Assessment Aspects of
Remedial Action Planning
James F. Walsh & Kay H. Jones, Ph.D.
Air Monitoring of Hazardous Waste Sites
Richard W. Townsend
Air Emission Monitoring of Hazardous Waste Sites
Louis J. Thibodeaux, Charles Springer, Phillip Lunney,
Stephen C. James & Thomas T. Shen
Air Pollution Problems of Uncontrolled Hazardous
Waste Sites
Thomas T. Shen, Ph.D. & Granville H. Sewell, Ph.D.
The Investigation of Mercury Contamination in the
Vicinity of Berry's Creek
David Lipsky, Ph.D. & Paul Galuzzi
.36
.40
.45
.52
.57
.58
.63
.67
.70
.76
.81
GEOHYDROLOGY
Practical Interpretation of Groundwater
Monitoring Results
William A. Duvel, Jr., Ph.D.
Application of Geophysics to Hazardous
Waste Investigations
Robert M. White & Sidney S. Brandwein
Case Study of Contaminant Reversal and Groundwater
Restoration in a Fractured Bedrock
R.M. Schuller, W. W. Beck, Jr. & D.R. Price
Cost Effective Preliminary Leachate Monitoring at an
Uncontrolled Hazardous Waste Site
H. Dan Harmon, Jr. & Shane Hitchcock
Vadose Zone Monitoring Concepts at Landfills,
Impoundments, and Land Treatment Disposal Areas
L.G. Everett, Ph.D., E.W. Hoylman, L.G. McMillion
&L.G. Wilson, Ph.D.
Mitigation of Subsurface Contamination
by Hydrocarbons
W. Joseph Alexander, Donald G. Miller, Jr. &
Robert A. Seymour
An Approach to Investigating Groundwater Contaminant
Movement in Bedrock Aquifers: Case Histories
Richard G. DiNitto, William R. Norman &
M. Margaret Hanley
Evaluation of Remedial Action Alternatives—
Demonstration/Application of Groundwater
Modeling Technology
F. W. Bond, C.R. Cole. & D. Sanning
REMEDIAL RESPONSE
.86
.91
.94
.97
.100
.107
.111
.118
Remedial Action Master Plans
William M. Kaschak & Paul F. Nadeau
The Department of Defense's Installation
Restoration Program
Donald K. Emig, Ph.D.
Survey and Case Study Investigation of Remedial
Actions at Uncontrolled Hazardous Waste Sites
5. Robert Cochran, Marjorie Kaplan, Paul
Rogoshewski, Claudia Furman & Stephen C. James
Conceptual Designs and Cost Sensitivities of Fluid
Recovery Systems for Containment of Plumes of
Contaminated Groundwater
D.A. Lundy & J.S. Mahan
Planning Superfund Remedial Actions
Brint Bixler, Bill Hanson & Gilah Langner
Alternatives to Groundwater Pumping for Controlling
Hazardous Waste Leachates
Charles Kufs, Paul Rogoshewski, Edward Repa &
Naomi Barkley
The Exhumation Program for the SCA Wilsonville Site .
John J. DiNapoli
PCB's at Superfund Sites: Remedial Action Experiences
John W. Thorsen, Robert J. Schoenberger, Ph.D. &
Anthony S. Bartolomeo
Evaluation of Remedial Actions for Groundwater
Contamination at Love Canal, New York
Lyle R. Silka & James W. Mercer, Ph.D.
Monitoring Chlorinated Hydrocarbons in Groundwater
D.A. Palombo & J.H. Jacobs
.124
.128
.131
.136
.141
.146
.150
.156
.159
.165
-------
Drum Handling Practices at Abandoned Sites
Roger Wetzel, Kathleen Wagner & Anthony N. Tafuri
BARRIERS
Geotechnical Aspects of the Design and Construction
of Waste Containment Systems
Jeffrey C. Evans & Hsai-Yang Fang, Ph.D.
Coverings for Metal Contaminated Land
A.K. Jones, Ph.D., R.M. Bell, Ph.D., L.J. Barker &
A.D. Bradshaw, F.R.S.
Selection, Installation, and Post-Closure Monitoring
of a Low Permeability Cover Over a Hazardous Waste
Disposal Facility
Mark J. Dowiak, Robert A. Lucas, Andrzej Nazar &
Daniel Threlfall
Pollution Migration Cut-Off Using Slurry Trench
Construction
Philip A. Spooner, Roger S. Wetzel & Walter
E. Grube, Jr., Ph.D.
Gelatinous Soil Barrier for Reducing Contaminant
Emissions at Waste Disposal Sites
B.E. Opitz, W.J. Martin & D.R. Sherwood
.169
.175
.183
.187
.191
.198
.203
.209
.214
.220
.224
.228
TREATMENT
Treatment of High Strength Leach ate from
Industrial Landfills
R.C. Ahlert, P. Corbo & C.S. Slater
Treatment of TNT and RDX Contaminated Soils
by Composting
R.C. Doyle, Ph.D. & J.D. Isbister, Ph.D.
ULTIMATE DISPOSAL
Hazardous Waste Incineration: Current/Future Profile ...
C.C. Lee, Ph.D., Edwin L. Keitz & Gregory A. Vogel
The Use of Grout Chemistry and Technology in the
Containment of Hazardous Wastes
Philip G. Malone, Ph.D., Norman R. Francingues &
John A. Boa, Jr.
Criteria for Commercial Disposal of Hazardous Waste....
MiloG. Wuslich
Above Ground Storage of Hazardous Waste
Christopher J. Lough, Mark A. Gilbertson &
Stephen D. Riner
RESEARCH & DEVELOPMENT
Uncontrolled Hazardous Waste Site Control
Technology Evaluation Program 233
Ronald Hill, Norbert Schomaker & Ira Wilder
Applications of Soluble Silicates and Derivative
Materials in the Management of Hazardous Wastes 237
Robert W. Spencer, Richard H. Reifsnyder &
James C. Falcone, Jr., Ph.D.
Development and Demonstration of Systems to Retrofit
Existing Liquid Surface Impoundment Facilities with
Synthetic Membrane 244
John W. Cooper & David W. Schultz
A Block Displacement Technique to Isolate Uncontrolled
Hazardous Waste Sites 249
Thomas P. Brunsing, Ph.D. & Walter E.
Grube, Jr.. Ph.D.
CASE HISTORIES
Callahan Uncontrolled Hazardous Waste Site—During
Extreme Cold Weather Conditions 254
William E. Ritthaler
Operating Experiences in the Containment and
Purification of Groundwater at the Rocky Mountain
Arsenal
Donald G. Hager & Carl G. Loven
Cost Effective Management of an Abandoned
Hazardous Waste Site by a Staged Cleanup Approach
Kenneth F. Whit taker, Ph.D. & Robert Goltz
Picillo Farm, Coventry, Rhode Island: A Superfund
& State Fund Cleanup Case History
Barry W. Muller, Alan R. Brodd & John Leo
A Coordinated Cleanup of the Old Hardin County
Brickyard, West Point, Kentucky
Fred B. Stroud, Barry G. Burrus & John M. Gilbert
Silresim: A Hazardous Waste Case Study
John D. Tewhey, Ph.D., Josh E. Sevee &
Richard L. Fort in
Cleanup and Containment of PCBs—A Success Story
Brian D. Bracken & Hilary M. Theisen
Implementation of Remedial Actions at Abandoned
Hazardous Waste Disposal Sites
Dale S. Duffala & Paul B. MacRoberts
Proposed Cleanup of the Gilson Road Hazardous
Waste Disposal Site, Nashua, New Hampshire
Frederick J. McGarry & Bruce L. Lamarre
Remedial Activities at Florida's Uncontrolled
Hazardous Waste Sites
Vernon B. Myers, Ph.D., Dan DiDomenico &
Brent Hartsfield
PERSONNEL SAFETY
Safety and Air Monitoring Considerations at the
Cleanup of a Hazardous Waste Site
D.A. Buecker & M .L. Bradford
Uses and Limitations of Environmental Monitoring
Equipment for Assessing Worker Safety in the Field
Investigations of Abandoned and Uncontrolled
Hazardous Waste Sites
Christine L. McEnery
Worker Safety and Degree-of-Hazard Considerations
on Remedial Action Costs
J. Lippitt, J. Walsh, A. Di Puccio & M. Scott
Hazardous Waste Site Investigations: Safety Training,
How Much Is Enough?
Steven P. Maslansky
OFFSITE SAFETY
Addressing Citizen Health Concerns During Uncontrolled
Hazardous Waste Site Cleanup
Gregory A. Vanderlaan
Estimating Vapor and Odor Emission Rates from
Hazardous Waste Sites
Alice D. Astle, Richard A. Duffee & Alexander
R. Stankunas, Ph.D.
Air Modeling and Monitoring for Site Excavation
Brian L. Murphy, Ph.D.
Sampling Techniques for Emissions Measurement at
Hazardous Waste Sites
Charles E. Schmidt, Ph.D., W. David Balfour &
Robert D. Cox, Ph.D.
PUBLIC PARTICIPATION
Public Participation in Hazardous Waste Site Control-
Not "If" but "How"
Richard A. Ellis, Ph.D. & Robert W. Howe, Ph.D.
Public Involvement in Resolving Hazardous Waste
Site Problems
Hugo D. Freudenthal, Ph.D. & Jean A. Celender
.259
.262
.268
.274
.279
.285
.289
.291
.295
.299
.306
.311
.319
.321
.326
.331
.334
.340
.346
-------
Citizen Participation in the Superf und Program 350
Catherine Neumann & Ben Drake, Ph.D.
Progress in Meeting the Objectives of the Superfund
Community Relations Program 354
Barry H. Jordan, Anthony M. Diecidue &
James R. Janis
DATA BASE
Spill Incidents at Hazardous Material Storage Facilities:
An Analysis of Historical Data from the PIRS and
SPCC Data Bases 357
E. Hillenbrand & B. Burgher
The Hazardous Materials Technical Center 363
David A. Appier, Murray J. Brown & Torsten Rothman
Documentation for Cost Recovery Under CERCLA 366
R. Charles Morgan & Barbara Elkus
REMEDIAL COST
Development of a Framework for Evaluating Cost-
Effectiveness of Remedial Actions at Uncontrolled
Hazardous Waste Sites
Ann E. St. Clair, Michael H. McCloskey &
James S. Sherman
Negotiations: The Key to Cost Savings
William R. Adams, Jr.
RISK/DECISION
Application of Environmental Risk Techniques to
Uncontrolled Hazardous Waste Sites
Barney W. Cornaby, Ph.D., Kenneth M. Duke, Ph.D.,
L. Barry Goss, Ph.D. & John T. McGinnis, Ph.D.
Exposure-Response Analysis for Setting Site
Restoration Criteria
G. W. Dawson & D. Sanning
Perspectives of Risk Assessment for Uncontrolled
Hazardous Waste Sites
Terry Ess & Chia Shun Shih, Ph.D.
Abandoned Site Risk Assessment Modeling and
Sensitivity Analysis
Brian L. Murphy, Ph.D.
Assessing Soil Contamination at Love Canal
Glenn E. Schweitzer
Uses and Limitations of Risk Assessments in Decision-
Making on Hazardous Waste Sites
Ian C. T. Nisbet, Ph.D.
Multiattribute Decision-Making Imbedded with Risk
Assessment for Uncontrolled Hazardous Waste Sites
Chia Shun Shih, Ph.D. & Terry Ess
STATE PROGRAMS
U.S. Army Corps of Engineers Role in Remedial Response
Brigadier General Forrest T. Gay, HI, Noel W. Urban
& James D. Ballif
.372
.377
.380
.386
.390
.396
.399
.406
.408
.414
State Participation Under Superfund 418
Harry P. Butler
Federal/State Cooperation on Superfund: Is It Working? 420
Gail Tapscott
The TCE Response in Arizona 424
Pamela Jane Beilke
Implementation of a State Superfund Program:
California 428
Robert D. Stephens, Ph.D. & Thomas E. Bailey
INTERNATIONAL
An International Study of Contaminated Land
M.A. Smith & M.J. Beckett
Long Term Effectiveness of Remedial Measures
Klaus Stief
Leachate Treatment and Mathematical Modelling of
Pollutant Migration from Landfills and Contaminated
Sites
V.E. Niemele, K.A. Childs & G.B. Rivoche
Development of an Installation for On-Site Treatment
of Soil Contaminated with Organic Bromine
Compounds
W.H. Rulkens, Ph.D., J. W. Assink & W.J.Th.
Van Gemert, Ph.D.
Degraded and Contaminated Land Reuse—
Covering Systems
Graham D.R. Parry, Ph.D., Robert M. Bell, Ph.D. &
A.K. Jones, Ph.D.
In Situ Treatment of Uncontrolled Hazardous
Waste Sites
J. Bruce Truett, Richard L. Holberger &
Donald E. Sanning
.431
.434
.437
.442
.448
.451
LEGAL
Allocating Superfund Liability
Edward I. Selig, Esq. & Joanne Kadish, Esq.
Protection from Long-Term Liability as a Result of
Superfund Remedial Actions
D.E. Sanders, F. Sweeney & E. Hillenbrand
Liability and Insurance Aspects of Cleanup of
Uncontrolled Hazardous Waste Sites
Paul E. Bailey, J.D.
Negotiating Superfund Settlement Agreements
Lauren Stiller Rikleen, Esq.
Hazardous Waste and the Real Estate Transaction: A
Practical and Theoretical Guide for the Technical
Consultant, Real Estate Attorney, Business Person,
Investor, or Anyone Involved in Buying and Selling Land
Jeffrey T. Lawson & Barbara H. Cane, Esq.
.458
.461
.464
.470
.474
-------
ELECTROMAGNETIC RESISTIVITY MAPPING
OF CONTAMINANT PLUMES
J.D. McNEILL
Geonics Limited
Toronto, Ontario
FACTORS AFFECTING SOIL RESISTIVITY
The electrical resistivity of a soil is a measure of the relative dif-
ficulty encountered in causing an electrical current to flow in it; the
more resistive the soil, the smaller the current flow for a given
voltage. Surprisingly, most physical constituents of a soil are elec-
trical insulators of such high resistivity that no appreciable current
flows through them. What does allow significant current to flow is
the relatively conductive soil moisture; it is this parameter which
often controls the soil bulk resistivity.
An electrical model of soil where it is, considered to. consist of a
large number of insulating particles immersed in a conductive fluid
is shown in Fig. 1. The mixture resistivity should be affected both
by the resistivity of the conductive soil moisture and also by the fact
that the insulating particles act to impede the current flow. Em-
pirically it has been established that Archie's Law often gives the
correct behavior of soil resistivity:1'2
/sample = ? moisture x
1
\ (soil porosity)3
(1)
and, as expected, there is a linear relationship between soil resistivi-
ty and the resistivity of the included water. Now the water resistivi-
ty is determined mainly by the ionic content since it is the move-
ment of ions that carries the electrical current. For a given voltage
more ions permit greater current flow, i.e. reduced resistivity; it is
on this principle that the use of resistivity surveys to outline con-
taminated areas is based.
However, other factors also affect the measured soil resistivity.
For example, it is evident from Eq. 1 that soil porosity has a
somewhat greater effect on soil resistivity than the soil moisture, so
that variations in soil type, which result in changes in porosity can
cause incorrect interpretation of resistivity surveys carried out to
map contaminants. Clay content (and the type of clay) can addi-
tionally affect soil resistivity because of a "surface conduction"
Soil particles
(insulators)
Soil moisture
(conductive)
Lines of current flow
Figure 1.
Electric model of soil sample
phenomenon which occurs in clay. Furthermore since resistivity
measurements are influenced by the vertical distribution of
resistivity, which is in turn influenced by the vertical distribution of
soil moisture, variations in the moisture profile (such as changes in
the level of the water table) will affect survey results.
Since geological and hydrogeological factors can affect soil
resistivity, surveys intended to delineate a contaminant area must
include a sufficient density of measurements both over the suspect
region and also beyond into the surrounding area so that the possi-
ble influence of any of the above factors can be determined. Fur-
thermore, the survey interpreter must always bear in mind the
various factors other than soil water resistivity that can influence
the survey results.
CONVENTIONAL RESISTIVITY SURVEY TECHNIQUES
Conventional resistivity surveys are carried out by inserting four
metal electrodes in the ground in one of a number of arrays. The
theory of such techniques is well covered in the literature.2
In general, a voltage applied across two of the electrodes causes a
current to flow in the soil, and the resulting voltage measured
across the two other electrodes is a measure of the soil resistivity.
The Wenner array, commonly used for geotechnical surveys, is
shown in Fig. 2. The depth to which resistivity is sensed is deter-
mined essentially by the inter-electrode spacing, and for the Wen-
! Generator
•fl l^*^
^Voltmeter
r— (V)— J
*> _ n - ^> _ n _ i"i
• a "ip — a ' M • a *
II II II 1
|| M II
;: II II -
** V V _^~^
/? |S^ jy
' t V
1 '
/ \ V /
/ 1 \ *
1 \ ^^_ "X
/ \
! \
\
^ X-^ ^
1, Possible resistivity
inhomogeneity near
\\ voltage electrode
\ \
' \ N
\ \
i \
\ \
\
/ \
j \
/ Resistivity
/ =P (ohm meters)
= 27Ta y
Depth of investigation = a
Figure 2.
Conventional resistivity (Wenner array)
-------
2 SITE INVESTIGATION
ner array it is usually considered to be approximately a. The
resistivity of the ground for the Wenner array is given as:
f = 2raV (2)
I
where p = resistivity (Om)
a = inter-electrode spacing (m)
I = current flowing through outer electrodes
(amps)
V = voltage across inner electrodes (volts)
Although widely used for resistivity surveys, there are several disad-
vantages related to the use of electrodes:
•It may be difficult or impossible to drive electrodes into compact
earth. It is impossible to survey in the winter when the ground is
frozen.
•The presence of resistive inhomogeneities (for example, rocks)
near the voltage electrodes can cause large measurement errors,
even though the physical size of the inhomogeneity is much
smaller than the anticipated depth of exploration. Reconnais-
sance surveys are usually carried out by making a series of
measurements along the survey line at constant inter-electrode
spacing to achieve essentially constant depth of exploration. The
survey data are plotted as a profile of measured resistivity along
the survey line. Such profiles can be quite "noisy" due to
electrical inhomogeneities with the result that the presence of
subtle changes in resistivity caused by a contaminant, might be
missed by the interpreter.
•In order to make a measurement the four electrodes must be ac-
curately spaced (Fig. 2) and a total length of wire equal to 4a
must be connected between the electrodes and the instrumenta-
tion. Thus the process is laborious and progress is slow. Even for
the most organized surveys the survey costs on a line-mile basis
are high.
These high costs lead in turn to several important consequences:
•Within the presumed anomalous area, insufficient measure-
ments may be carried out to accurately characterize the plume.
•Outside of the anomalous area, insufficient measurements may
be carried out to accurately characterize the background against
which the plume is to be contrasted or to determine the ex-
istence and nature of other anomalous regions which may exist
and which may or may not be caused by contaminants.
•In areas of complex hydrogeology a time-consuming and ex-
pensive survey may be performed, to learn, at the conclusion,
that the data are inconclusive. The contaminant may simply not
be present in sufficient quantities to produce a marked and un-
ambiguous anomaly over the survey geological noise. Since such
cases do occur, the inclination to carry out further resistivity sur-
veys can be greatly tempered by a few such failures, which is un-
fortunate since resistivity measurement is often the single most
successful method in delineating contaminant plumes.
It was in recognition of the usefulness of resistivity and the high
cost of the conventional methods that inspired the research staff at
Geonics Limited to examine the application of electromagnetic
techniques for making resistivity measurements.
ELECTROMAGNETIC SURVEY TECHNIQUES
Let a small transmitter coil be situated on or close to the earth, as
indicated in Fig. 3. An alternating voltage, typically at an audio fre-
quency, is applied to the terminals of this coil, causing a current to
flow. This current generates an alternating magnetic field which,
through Faraday's Law, causes electrical currents to be induced in
the earth (no such current is induced in the air, which is effectively
infinitely resistive).
The induced currents in the earth generate a secondary magnetic
field. Both the primary and the secondary fields are detected by a
receiver coil located near the transmitter coil, as shown in Fig. 3,
and, in principle, measurement of the ratio of the secondary to the
Primary magnetic field
Transmitter
Secondary magnetic field
Ground of conductivity = a
Figure 3.
Inductive electromagnetic fields
primary magnetic field strength can be used to determine the elec-
trical resistivity of the earth. In general, however, this ratio is an ex-
tremely complicated function of the resistivity, the distance beween
the transmitter and receiver coils, and the frequency of the
transmitter current, so that interpretation of the results in terms of
the resistivity is quite involved.2-3 Furthermore, the depth of in-
vestigation is also a complicated function of the same three
parameters.
Fortunately a substantial simplification in the response function
results when the energizing frequency is chosen to be sufficiently
low so that the condition known technically as "operation at low
values of induction number" is fulfilled.3 For the remainder of this
paper all reference to the use of inductive electromagnetic techni-
ques assumes that the condition of a low induction number has
been met.
For example, for low values of induction number the ratio of the
secondary to primary magnetic field at the receiver becomes simply:
Hs -
"
(3)
operating frequency (Hz)
intercoil spacing (m)
ground conductivity which is the reciprocal of
resistivity (mho/m)
permeability of free space (a constant)
where f =
s =
a =
!>• =
and the quantity i (=\ ~1) indicates that the magnetic field aris-
ing from the induced currents in the ground is phase shifted by 90°
with respect to the primary magnetic field, greatly simplifying
measurement of the small ratio given by Eq. 3.
For the four-electrode resistivity measurement, Eq. 2 indicates
that the ratio of the voltage across the inner two electrodes, divided
by the current through the outer electrodes, is linearly proportional
to the terrain resistivity. For the electromagnetic measurement, Eq.
3 shows that the magnetic field ratio is linearly proportional to the
ground conductivity rather than resistivity, since, subject to the
constraint of operation at low induction number, the more conduc-
tive the ground the larger the current flow in the ground, and the
larger the resultant secondary field. Instruments based on this prin-
ciple are therefore called ground conductivity meters.
There are further advantages and some disadvantages to
measurement of terrain conductivity using the principles outlined.
The advantages fall into two main groups; ease of calculation of
system response to a layered earth and operational simplicity.
Calculation of Layered-Earth Response
In general, the resistivity or conductivity of the earth varies with
depth; for example in a typical vertical profile the conductivity will
initially increase with depth due to increasing soil moisture, becom-
ing essentially constant at the water table due to saturation. If the
underlying bedrock has very low porosity, the conductivity would
now decrease. Such a continuous conductivity profile, shown
-------
SITE INVESTIGATION
schematically in Fig. 4, would be approximated by the engineer/
geologist as a three-layer geometry also indicated in Fig. 4.
Suppose further, that in a certain region the groundwater may be
sufficiently contaminated to double the groundwater conductivity,
that is, through Eq. 1 to double the conductivity of the in-
termediate layer of Fig. 4. The question arises: "With a conven-
tional resistivity array of fixed interelectrode spacing 'a' or an elec-
tromagnetic system with fixed intercoil spacing V how is the in-
strumental response calculated over such 'layered earths' so that
the difference in response between the contaminated and uncon-
taminated areas can be determined?"
If the earth resistivity was uniform with depth, Eq. 2 shows that
the Wenner array, for a given current I, would give an inner elec-
trode voltage Vu related to the resistivity by:
Vu = p I (4)
If now the earth is layered, as indicated in Fig. 4, a different value
of voltage V^ will be measured for the same current I and inter-
electrode spacing a, and an apparent resistivity can be defined by:
/ a = 2ira
(5)
I
For a layered earth the apparent resistivity so defined will reflect
the influence of the various resistivities at the different depths.
To return to the contaminant problem, the question can now be
rephrased as "for a given fixed inter-electrode spacing how does
the apparent resistivity vary in going from the uncontaminated to
the contaminated region?" Unfortunately, for conventional
resistivity techniques, such a calculation requires a reasonably com-
plicated computer program (although it can now be performed on
the most advanced programmable pocket calculators). The calcula-
tion for an arbitrarily layered earth cannot be performed by hand.
For electromagnetic measurement of terrain conductivity at low
induction number, the concept of apparent conductivity is entirely
analogous. Equation (3) is inverted to yield:
"a. = _ 4_ Hs (6)
"H
which, for the case of a uniform earth gives the correct terrain con-
ductivity, and for the case of a layered earth gives an apparent con-
ductivity which also depends on the layering.
A major difference between the conventional and the elec-
tromagnetic survey techniques is that for the latter it is a simple
matter to calculate the apparent conductivity (by hand) for any
type of layering. The reason for this difference is that for conven-
tional resistivity measurements, the current distribution at any
point in the layered earth is a complicated function of the
parameters of all of the layers. In the case of the electromagnetic
surveys, the local current flow is determined by the local conduc-
tivity — changed in any given layer do not affect (to the low
induction-number approximation) the current flow in other layers.
It is thus possible to generate the curve in Fig. 5 which shows, for
Ca)
Co)
'. • Resistive material • .
•_': above . water table . ;
a 00
<• . Conductive . •
• . saturated zone •
1 1 1 1 1 -_L-
ix Very resistiveTI^
' 1 ' i bedrock ~^r"
V.
1
z,-
Zi
Hydro-geological
section
depth
Actual conductivity
Profile
depttl
Geo-electric model
2-0 Z
where z - depth / intercoi spacing
Figure 5.
Relative sensitivity to ground at various depths
100%
- 50%
2-0 Z
Figure 4.
Typical ground conductivity profile
where z = depth / intercoil spacing
Figure 6.
Cumulative sensitivity to ground at various depths
a uniform earth, the relative contribution to the meter reading from
a thin horizontal layer of thickness dz at any depth z, (where z is the
real depth normalized with respect to the intercoil spacing). The
figure shows that this relative response is very small near the sur-
face, that it increases with depth, becoming a maximum at 0.4 in-
tercoil spacings (i.e. at 4 m if the intercoil spacing is 10 m) and then
gradually decreases again. There is still appreciable response at 1.5
-------
o] - 5 mmho/m
Z| = 5 m
ffl 20 mmho/m
Z2 = 15m
(73 1 mmho/m
p\
ti
P2
12
/«3
= 2000m
5m
= 250m
10m
= 10000m
01 = 5 mmho/m
zi = 5 m
O2 = 40mmho/m
Z2 = 15 m
03 = 1 mmho/m
4 SITE INVESTIGATION
intercoil spacings. To construct such a response curve for conven-
tional resistivity techniques is not possible due to the interaction of
current flow at different depths.
Knowing the relative response to material at any given depth, it is
possible to generate from Fig. 5, the curve of Fig. 6 which gives the
cumulative contribution to the meter reading from all material
below any depth "z" (again normalized with respect to the intercoil
spacing). This curve can be used to quickly calculate the apparent
conductivity from a layered earth as follows,' using the data from
Table 1.
Table 1.
Postulated Geoeleclric Section (Fig. 4)
Uncontaminated Area Contaminated Area
p\ 2000m
tl = 5 m
pi 500m
12 = 10 m
/13 = 10000m
where a (mmho/tn) - tOOO/fl (Om)
and ij is the thickness of the ilh layer
and zj is the distance from the surface to the bottom of the ith layer.
Assuming the intercoil spacing is 10 m, all of the material below 0
meters produces 100% of the instrumental response, all the
material below 5 m produces 70% of the instrumental response and
all below 15 m produces 31% of the response. Therefore, the
material between 0 and 5 m produces 100-70 = 30% of the total
response and the material between 5 and 15 m produces 70-31 =
39% of the response. Since each layer produces its own contribu-
tion independently of that from the other layers, regardless of their
conductivity, to obtain the apparent conductivity one simply adds
the relative contribution from each layer, weighted according to its
own conductivity:
o = ffi x 0.30 + °2 x 0.39 + a- x 0.31
or, more generally :
aa = «i (l-R(zi)) + 02 (R(zi)-R(z2)) + as (R(z2)) (7)
where R(z) is the function given in Fig. 6.
For the uncontaminated area:
aa = 5(1-0.70) + 20(0.70-0.31) + 1 (0.31)
= 1.50 + 7.80 + 0.31
= 9.61 mmho/m
whereas for the contaminated area:
ffa = 5 (1-0.70) + 40 (0.70-0.31) + 1 (0.31)
= 1.50 + 15.60
= 17.41 mmho/m
Because of the contributions from the first and third layers, the ap-
parent conductivity has less than doubled.
The equation for R(z) is extremely simple (Fig. 6), so the calcula-
tion of the apparent conductivity for any number of layers can be
carried out with the simplest pocket calculator or, using the graph,
by hand. The contribution from each layer to the total is im-
mediately apparent, and it is simple to calculate the variation
caused by changes within any given layer.
In the example, a 2/1 change in the middle layer conductivity
produced a 1.8/1 change in the overall apparent conductivity. If,
however, the conductivity of the first layer increased from 5 to 25
mmho/m, fqr example, by encountering a region with high clay
content, the contribution from this layer would increase from 1.5
to 7.5 mmho/m which could be confused as an increase in the con-
taminant in the second layer. The provision for quick and simple
calculations of this type facilitates both the planning of surveys and
estimating the probability of their success.
Operational Advantages of Electromagnetic
Conductivity Measurement
•Resistivity inhomogeneities of a size much less than the antici-
pated depth of exploration would, if they were located near the
voltage electrodes, produce an anomalous measurement which is
truly "geological noise" since without further measurements it
is not possible to determine the resistivity contrast, the physical
location, or the size of the anomaly. In the case of the inductive
conductivity technique it can be shown4 that the current concen-
tration in the ground is highest in the vicinity of the transmitter
coil and one might anticipate that this technique would be es-
pecially sensitive to inhomogeneities in this location. However,
these high amplitude current loops have a small radius and their
effect on the relatively distant receiver is negligible. The net re-
sult is that in the inductive technique, once the intercoil spacing
has been selected to be approximately equal to the desired depth
of the exploration, the system is quite insensitive to small, local
variations in conductivity, and an accurate measurement of the
bulk conductivity is obtained. This is particularly important in
studies for groundwater contamination where the changes in the
apparent conductivity due to the presence of the contaminant
may be rather small. Fortunately variations in the apparent con-
ductivity of 20% are quickly and reliably measured.
•The presence of a highly resistive upper layer offers no barrier to
measurements with inductive electromagnetic systems and surveys
can be carried out when the upper layer is frozen, through
desert sand, and even through concrete (assuming that there are
no reinforcing bars).
•With the electromagnetic measurements, the effective depth of
exploration is given approximately by one and a half times the
intercoil spacing, whereas for conventional resistivity measure-
ments the exploration depth is only one third the array length.
There is no necessity to lay out lengths of wire on the ground
which are much greater than the exploration depth and there is no
requirement for electrodes.
•It is a simple matter to incorporate circuitry which automatically
indicates the correct intercoil spacing, thus doing away with the
requirement of physically measuring the distance.
•The equipment is lightweight and readily portable. A "two-
man" instrument achieves an exploration depth of up to 60 m.
Instrumental Disadvantages of Electromagnetic
Conductivity Measurement
The disadvantages of the inductive electomagnetic terrain con-
ductivity meters are instrumental in nature:
•At levels of conductivity below about 1 mmho/m, there simply
is not enough response from the small currents induced in the
ground to obtain an accurate measurement. At high levels of
conductivity, the "low induction number" approximation
breaks down and the instrument response becomes increasingly
non-linear with conductivity. This constraint also makes it diffi-
cult to design an instrument for large depths of exploration.
•The measured ratio of secondary to primary magnetic field is
typically 0.3% and often less (Eq. 3). To achieve precision at
these levels requires sophisticated electronic design, which re-
sults in instruments that are significantly more expensive to
manufacture than conventional resistivity equipment.
•Ideally the instrument "zero" would be set by removing the in-
strument from the influence of all conductive material, includ-
ing the earth. Obviously this is not possible and it is difficult to
establish and maintain this zero to better than a few tenths of a
mmho/m over the wide ranges of temperature, humidity, and
mechanical shock to which geophysical equipment is routinely
exposed. This feature further limits the accuracy in highly re-
sistive ground.
•In principle conductivity sounding with depth can be carried out
-------
SITE INVESTIGATION
in a manner completely analogous to that for conventional re-
sistivity equipment, i.e. measurement is made over a wide range
of intercoil spacings. Technical problems associated with the
dynamic range of the received signal make this difficult and ex-
pensive to do, and currently available instrumentation has a maxi-
mum of three switch-selectable intercoil spacings of 10, 20 and
40 m.
In summary, inductive electromagnetic techniques are most
suited to rapid reconnaissance-type surveys, where the relatively
high initial cost of the equipment can be offset by the speed and low
cost with which surveys can be carried out, and where the resolu-
tion in conductivity, whereby small variations can be accurately
mapped, is a prime consideration in the survey objectives.
For those situations where very high or very low conductivities
are to be mapped, or where an accurate profile of the vertical
distribution of resistivity is the objective, conventional resistivity
techniques will still be required.
SURVEY INSTRUMENTS
Instrument design conforming to the condition of operation at
low values of induction number, forms the basis of the patented
Geonics EM31, EM34-3 and EM38 terrain conductivity meters.
The EM31, a one-man portable instrument with a fixed intercoil
spacing of 3.7 m and a depth of exploration of about 6.0 m is
shown in Fig. 7. Basically designed as a rapid reconnaissance in-
strument the EM 31 can be effectively used with a chart recorder to
provide continuous profiles of ground conductivity. In addition
this instrument is very effective in detecting and mapping the loca-
tion of buried metallic drums.' Finally, by laying the instrument on
the ground and making two measurements, one with the device in
Figure 7.
EM 31
normal position and a second on its side (vertical and horizontal
dipole modes), it is possible to detect a two-layered earth and to as-
certain whether the more conductive material is near surface or at
depth.'
The EM34-3 (Fig. 8) is a two-man instrument with switch-
selectable intercoil spacings of 10, 20, or 40 m to permit maximum
depths of 15, 30 and 60 m. It too can be operated in either the ver-
tical or horizontal dipole mode to vary the instrumental sensitivity
with depth. The two coils are connected by a flexible cable: the
receiver console has two meters—one of which electronically in-
dicates the intercoil spacing.
To make a measurement the transmitter operator stops at the
survey mark: the receiver operator then moves his coil with respect
to the transmitter until this meter indicates that the correct intercoil
spacing has been achieved, whereupon he reads the terrain conduc-
tivity on the second meter. The whole procedure takes about 20 sec.
The EM 38 is a 1.0 m long instrument (depth about 1.5 m)
designed for soil salinity measurements.
SURVEY CASE HISTORY
A case history' will illustrate some of the features of surveys car-
ried out using inductive electromagnetic techniques.
The survey area, shown in Fig. 9, is described by Greenhouse and
Slaine6 as follows:
"A variety of waste chemicals from herbicide and pesticide
manufacturing were deposited in lined lagoons situated on
glacial overburden during the 1970s. One or more of the lagoon
liners has leaked into an unconfined aquifer, producing ground-
water conductivity anomalies proportional to total dissolved
solids (primarily chloride and sodium). The contamination
threatened a nearby creek but the pattern of movement was un-
known. Geophysical surveys were requested to assist in locat-
ing a drilling program."
This is a typical application for a geophysical survey.
During the planning stages of a conductivity survey, Greenhouse
and Slaine obtain all of the available hydrogeological data on the
CONTOUR INTERVAL • 5 ft
BURIED WASTE LAGOONS
Figure 8.
EM 34-3
Figure 9.
Survey case history area
-------
SITE INVESTIGATION
(Conductivities are in mmho/irO
ITUOUD
SECTION
Figure 10.
Hydro-geological and geo-electric section
site. These data are used to construct a hydrogeological model such
as that shown in Fig. 10. Alongside each geological unit an estimate
is made (generally from previous survey experience) of the electrical
conductivity of that unit in order to make a geo-electric section.
Next those formations whose conductivity might change due to
contamination are identified. For example, in Fig. 10 "the uncon-
fined sand aquifer is apparently isolated from the lower regional
sand aquifer by a clay till. Groundwater flow in the upper sand was
expected to be horizontal and towards the creek.'"
In order to determine which instrument, coil orientation, and in-
tercoil spacing is most suitable for the survey, calculations of the
type outlined above are performed to obtain the apparent conduc-
tivity measured by each instrument or spacing as the conductivity
of the postulated contaminated section is allowed to increase. From
such calculations the optimum instrument or spacing can be
selected, and, furthermore, estimates made of the survey success,
since as shown it is a simple matter to vary the conductivities in the
geo-electric section (to account for possible errors in their estimated
magnitude) and to determine whether these will mask the an-
ticipated anomaly. For this survey site, such calculations suggested
that for the EM31, any increase in conductivity of the silty sand
formation of more than a factor of about 1.8 would produce a
detectable anomaly over the usual background variations, and that
for the EM34-3 (used in the horizontal dipole mode) this factor
would be about 2.2, both of which would indicate a successful
survey since a higher change in conductivity could be anticipated
form the contaminant.
"Measurements were made on a basic 50 m grid covering a
500 x 500 m area centered on the southernmost lagoon...The
grid was refined to 25 m for much of the western half of the
survey, for a total of 150 stations per instrument. Establishing
the grid required 14 man hours; the...EM 31 and EM34 surveys
required 10 and 20 man hours respectively."6
Greenhouse and Slaine chose a logarithmic base for their data
presentation. More specifically, they plotted contours of decibels:
201ogloqa(x,y)
(8)
aa (background average)
where aa (x, y) = measured values of aa over the
survey area
ffa (background average) = average background con-
ductivity (i.e. average value
measured off the anomaly).
Thus, the zero db contour outlines the background, and their con-
tour interval of 4 db portrays successive factors of about 1.6 over
the background. They suggest three advantages for this technique:
•Logarithmic contours do not cluster close to the course and thus
do a better job of defining the plume
•Non-dimensional contour units with a zero background put all
instruments on an equal basis
•The procedure is easily automated once
-------
FIELD INVESTIGATION OF AN ABANDONED
PESTICIDE FORMULATION PLANT
DEBORAH A. KOPSICK
Ecology & Environment, Inc.
Kansas City, Kansas
INTRODUCTION
Between 1974 and 1981, a pesticide formulating plant operated
within the floodplain of the Missouri River, 11.3 km south of
Council Bluffs, Iowa. A sequence of events, beginning with a fire
in 1976, led to the investigation of the site by the Region VII United
States Environmental Protection Agency (USEPA). In 1981,
Ecology & Environment, Inc. (E&E) was authorized by Region VII
to perform a field investigation to determine the extent of con-
tamination at the site.
In this paper, the author discusses the design, implementation
and the data obtained in the various phases of this investigation. At
the time of the investigation, this site was being considered for in-
clusion on the first Interim Priorities List.
BACKGROUND
Physical Setting
The plant site is located at the eastern edge of the Missouri River
floodplain, directly adjacent to the loess covered bluffs that border
the valley (Fig. 1). The site has a minimal slope of 0-5 percent and is
located 3.0-6.1 m above the normal level of the Missouri River,
located 5 km to the southwest. The surface drainage in this
agricultural section of the floodplain is controlled through artificial
drainage ditches.
The alluvial sediments underlying the surface soils consist of
sands, silts and clays, with coarse sand and gravel near the base of
the alluvial sequence. The highly dissected loess bluffs bordering
I , COUNCIL BLUFFS SOUTH, IOWA—NEBR.
SCALE 1:24000
18
loop ;ooo 3000 aooo sooo eooo 7000 FEET
5 0 1 KILOMETRE
Figure 1.
Site Location Map
-------
SITE INVESTIGATION
MIHOUMI RrvEH FLOOO^LAIN
«ISIOfNTIAL//_-
HOUSC ^r. rj Northw»tt Drum Stor»Q« Ar««
• E Inoprop Spill Ar«»
r/-
t All«g«d Buried Wait*
g Loading Dock*
h South«wt Drum Storag* Ar«
tOO iOO •=
Figure 2.
Site Map
the Missouri River valley are composed of windblown silts of
glacial origin and rise 30.5-91.4 m above the floodplain.
The alluvial aquifer of the Missouri River provides the major
drinking and industrial water supply in the area. The drinking
water wells surveyed in the vicinity of the site range from 5.5-30.5
m in depth. Two wells located at the site average 27.4 m in depth.
Past History
Prior to 1974, this company operated a pesticide formulating
facility in Omaha, Nebraska. This facility was destroyed in a fire
and subsequently the operation was moved to its present site, south
of Council Bluffs, Iowa. This new facility (Fig. 2), which encom-
passes 8 ha, was located in a rural area to minimize exposure to the
population should a similar incident occur (Fig. 3).
The company operated as both formulators and packagers of
various organochlorine, organophosphate and s-triazine pesticides
and herbicides. Two divisions of the company were also located at
the site and handled the dry and wet formulations. The two major
products were ethoprop, an organophosphate pesticide, and
atra/ine, a triazine herbicide, along with small volume and pilot
project formulations.
In Nosember of 1976, a fire destroyed the atrazine formulation
plant at the Council Bluffs facility. Following this fire, production
at the plant was greatly reduced and in 1980 the firm filed for
reorganization under Chapter 1 1 of the Federal Bankruptcy Act. In
August 1981, a bankrupts sale was held at the site to liquidate the
assets of the company.
Contamination al the Site
During the process of extinguishing the 19^6 fire that destroyed
the atrazine torrnuldiion plant, it was estimated that 110 m of
water were used. Chemical contamination of the sediments and
water in the plant's drainage system was documented at this time by
Region VII, with prometon (60-357 mg/kg) and atrazine (13-840
mg/kg) being the major pesticides detected.
An investigation by Region VII in March 1980 showed concen-
trations of atrazine in the drainage sediments of 32 mg/kg. In
August of 1980, an investigation by the National Enforcement In-
vestigations Center (NEIC) of USEPA detected seven pesticides in
subsurface sediments with concentrations greater than 1 mg/kg:
atrazine, heptochlor, chlorodane, phorate, disulfoton, chlorpyrifos
and toxaphene. At the same time, a buried tank located north of
the former atrazine plant was sampled and found to contain lin-
dane, malathion, toxaphene, and methoxychlor in concentrations
of 20, 320, 1,200, and 3,200 mg/kg respectively. During each of the
sampling efforts described above, the two onsite wells were sampl-
ed and found not be contaminated with pesticides.
The final sampling effort occurred in August 1981, prior to the
E&E investigation and following the bankruptcy sale. During the
sale, a spill of ethoprop baghouse dust occurred. As a result of this
accident, two workers were admitted to the hospital with
organophosphate pesticide poisoning. At the same time, several
large metal atrazine storage tanks, sold as scrap, were drained into
a concrete basin at the site. Personnel from Region VII USEPA
coordinated the cleanup of this spill.
From the samples taken during these four investigations it
became clear that there were certain areas of concern that needed to
be addressed. The four investigations conducted at this site,
although limited in scope, documented contamination of the sur-
face soils, subsurface soils and drainage sediments. A buried tank
contained approximately 30 m3 of wastes from numerous pesticide
manufacturing processes. An open pit contained atrazine process
wastes. During those investigations, greater than 1,500 drums,
many in deteriorated condition, were being stored outdoors at the
site. An additional 1,200 fiber drums of baghouse dust were being
stored indoors. There were also allegations by former plant
employees and nearby residents of drummed waste buried at the
site. The E&E investigation was initiated to determine the extent of
the contamination and assess the potential for migration of con-
tamination via the air, soil and groundwater pathways.
FIELD AND LABORATORY INVESTIGATIONS
In order to assess the extent of pesticide contamination on and
around the site, a multi-phase analytical study was undertaken. A
description of a field investigation procedure similar to the one
followed at this site is found in a prior conference paper by Hagger
and Clay in 1981.'
Figure 3.
Aerial Photograph of Site
Isee Fig 2 for identifying features)
-------
SITE INVESTIGATION
Analyses of airborne participates, surface soil, subsurface soil,
subsurface sediments and groundwater samples were used to
delineate the most highly contaminated areas of the site. A metal
detector survey was also performed to locate areas of suspected
waste disposal. The following sections will describe the rationale
behind the design of each of these monitoring schemes.
Due to the public access to the site, Region VII USEPA person-
nel had a security fence constructed, financed by Superfund Im-
mediate Removal funds.
Air
As a Superfund candidate, it was necessary to determine the air
pollution release potential. Due to the chemical odors originating
from the site, an initial organic vapor survey was performed prior
to any field sampling.
Pesticides are not generally volatile and would therefore not be
recorded by an organic vapor analyzer. However, solvents used
during the formulation of pesticides are volatile and will be
detected, by the analyzer.
High-volume air sampling was also conducted to qualify the ex-
istence of airborne particulate pesticides on- and off-site. A quan-
tification of values was not possible, because pesticide may
volatilize while being pulled through the pressure differential
created in the high volume sampler.
Surface Soil
An extensive soil sampling study was conducted to produce data
on surface contamination resulting from spillage, leaking drums
and containers, and poor housekeeping practices at the plant. The
three grid sampling patterns chosen for the soil survey were design-
ed to cover all of the major areas of the plant used during its opera-
tion, as well as areas of waste disposal and alleged debris burial.
The major drainageways outside the site were sampled to document
any migration of contaminants offsite. Within the grid pattern, 89
samples were collected to'a maximum depth of 7.6 cm.
Because of the large number of samples to be analyzed, and time
and budgetary constraints, a method was used to identify those
samples which were most contaminated. This procedure of screen-
ing out non-contaminated samples consisted of analysis in a
capillary gas chromatograph with an electron capture (EC) detec-
tor. Total pesticide concentration estimates were calculated using a
standard mixture of several of the suspected organochlorine
pesticides. The concentrations measured may be substantially lower
than actual total concentrations of pesticides due to the greater
response of the EC to organochlorine compounds relative to or-
ganophosphate and s-triazine compounds. However, these concen-
trations are useful in comparing the concentrations of samples
from different locations and identifying those samples that were
significantly contaminated.
Based on the results of preliminary chemical analysis on the 89
samples, 20 samples were selected for further analysis. These
samples included uncontaminated background as well as onsite
contaminated samples. Thermionic specific detectors allowed
quantification of the organophosphorus and organonitrogen pesti-
cides. Gas chromatography/electron capture techniques were used
to quantify the organochlorine compounds. The pesticides which
were quantified are listed in Table 1.
Composite Soil Samples
Subsurface composite soil sampling was conducted to determine
the vertical extent of contamination in areas identified as having
surface contamination. Four composite samples were collected to a
depth of about 15.2-30.5 cm. These samples were analyzed using
the same techniques described for the surface samples.
Subsurface Sediments
During the drilling of the groundwater monitoring wells, subsur-
face sediment samples were collected at 1.5 m intervals with a split
spoon sampler. A total of 33 samples were collected to a depth of
10.7 m. The analysis of these samples, which was performed using
Table 1.
Pesticides Determined During Site Analysis
Additional Pesticides:
Atrazine
Chlorpyrifos
Disulfoton
Ethoprop
Methoxychlor
Phorate
Prometone
Priority Pollutants Pesticides:
Aldrin
Dieldrin
Chlorodane
4,4'—DDT
4,4'—DDE
4,4'—ODD
alpha-Endosulfan
beta-Endosulfan
Endosulfan sulfate
Endrin
Endrin aldehyde
Heptochlor
alpha-BHC
beta-BHC
gamma-BHC (Lindane)
delta-BHC
Toxaphene
the methods described above for surface soil samples, determined
the extent to which pesticides had migrated vertically into the sub-
surface.
Groundwater
A five-well groundwater monitoring system was installed to
monitor the shallow groundwater for pesticide contamination; in-
cluded were a background well and wells adjacent to areas of
buried wastes, the buried tank, and drum-storage areas. The wells
were installed in a blue-gray, silty, clay deposit overlying the
alluvial aquifer, with the purpose being to monitor the initial water-
bearing strata. They were constructed of 7.6 cm diameter PVC
threaded-joint pipe and designed to prevent any downward migra-
tion of surface contaminants. Each well was screened and gravel
packed for 6.1 m. A bentonite seal was placed above the gravel and
the remainder of the column was sealed with cement grout. Con-
crete pads at the surface and locking well covers completed the
wells. The groundwater samples were analyzed using electron cap-
ture and thermionic specific detection for the pesticides listed in
Table 1.
In conjunction with the onsite groundwater monitoring, 17 off-
site drinking wells in the vicinity of the site were sampled. Analyses
for the acid, base/neutral, pesticide, and volatile fractions of the
priority pollutant list were performed. Analyses for the seven addi-
tional pesticides were not determined.
PERSONNEL SAFETY AND DECONTAMINATION
The guidelines used to determine the appropriate levels of protec-
tion are outlined in the following manuals developed by E&E for
USEPA:
•Personnel Protection & Safety Training Manual2
•Hazardous Waste Site Investigation Training Manual3
Due to the odors at the site, the amount of pesticide spillage, and
the chlorinesterase-inhibiting nature of the organophosphate
pesticides, it was important to protect the sampling and drilling
crews. All personnel were required to wear air-purifying respirators
with combination cartridges for organic vapors and particulates.
Disposable coveralls, neoprene boots and neoprene gloves were
also worn.
In order to avoid cross-contamination between samples, strict
decontamination procedures were followed. Whenever possible, as
in the surface soil samples, disposable equipment was used. When
this was not possible, as during the composite soil sampling and the
subsurface sediment sampling, a five-step soap and water wash and
solvent rinse procedure was followed. All drill equipment was
steam cleaned between holes.
-------
10
SITE INVESTIGATION
FIELD AND LABORATORY RESULTS
Air
Only those readings taken in close proximity to drums which had
contained solvents had significantly higher concentrations of
organic vapors than background levels of 1 to 3 ppm. There were
odors present at the site, caused by mercaptan compounds used in
organophosphate formulations; however, mercaptans are detec-
table at concentrations less than 1 ppm, which is the lower limit of
the organic vapor analyzer in the survey mode.
Surface Soil
Based on the results of preliminary chemical screening of the 89
surface soil samples, 20 samples (Table 2) were further analyzed to
identify specific pesticides and develop background values for com-
parisons. The results of the analyses are presented in Table 3.
Table 2.
Locations of Surface Soil Samples
Sample No. Location
0 3 Ethoprop spill (east of building #1)
0 6 East of building #1
0 9 East of building tfl
I 6 East of Atrazine building (#4)
2 9 Southeast Drum Storage Area
3 6 Southeast Drum Storage Area
4 4 West of building #5
4 8 Loading dock at building #5
5 0 Northwest Drum Storage Area
5 3 Loading dock at building #5
5 9 North of Atrazine building #4
6 0 North of Plant (north of plant)
6 5 Along railroad tracks
7 9 West Drainage
Table 3.
Concentrations of Pesticides in
Surface Soil Samples (rag/kg)
Site Number
Pesticide 03 06 09 16 29 36 44 48 50 S3 59 60 65 79
190 400 200 120 140 1100
Promc-
lon(e)
Atrazine
Phorate
DisulToton
Chlorpy-
rifos
Heptachlor
Endosulfan
Sulfate
alpha-En-
dosulfan
Dicldrin
Aldrin
gamma-
BHC
(Lmdane)
TR 40 180 30 2 180
6 6
17 100
70
80
13.5 3600
670
940
I.I 0.6 550 1.0
0.5 3.2 O.I 1530 2.1
2.25
One soil sample collected near a loading dock showed high con-
centrations ( > 500 mg/kg) of aldrin, dieldrin, alpha-endosulfan,
endosulfan sulfate and heptachlor. The two major products of the
formulating plant, ethoprop and atrazine, were not detected in this
sample. Similarly, no ethoprop was detected in the area of the
ethoprop spill. However, several other pesticides were identified at
the following concentrations: disulfoton (70 mg/kg), prometone
(40 mg/kg) and 4,4'DDT (4.36 mg/kg). In the southeast drum
storage area, 100 mg/kg of phorate were detected near drums iden-
tified as containing phorate. These results agree with the data ob-
tained during the NE1C investigation.
Two piles of unidentified material dumped within the plant
grounds and thought to be pesticide products were found to be
primarily clay, with less than 1 mg/kg of pesticide. However, a red-
dish material dumped in the west drainage contained 110 mg/kg °»
prometone. This was the only significant contamination detected in
the plant drainages.
Composite Soil Samples
Pesticides were found in all four composite soil samples taken at
a depth of 15.2-30.5 cm below the surface (Table 4.) Heptachlor
was found in all four areas tested, while atrazine, ethoprop, and
phorate were detected in all locations except the northwest drum
storage area. In the southeast drum storage area, the major con-
taminant found was heptachlor (47 mg/kg). In the area of the
ethoprop spill, the following concentrations were detected:
ethoprop (23.8 mg/kg), phorate (44.1 mg/kg) and atrazine (1,600
mg/kg). The organochlorine pesticides heptachlor, methoxychlor,
and toxaphene were detected in the area north of the atrazine plant
and the northwest drum storage area.
Table 4.
Concentrations of Pesticides in
Subsurface Composite Samples (mg/kg)
Pesticide
Atrazine
Phorate
Ethoprop
Disulfoton
Toxaphene
Methoxychlor
Meptachlor
Dieldrin
Aldrin
alpha-BHC
beta-BHC
gamma-BHC
(Lindane)
Ethoprop
Spill Area
1600
44.1
23.8
2.44
1.29
1.40
0.055
0.12
0.76
North of
Atraz. Pt.
1.60
0.20
0.524
32.0
18.7
25.0
0.43
1.60
Northwest
Drum-Stg
35.0
16.7
16.0
4.0
1.55
Southeast
Drum-Stg
0.49
0.025
0.40
0.072
47.0
5.0
Subsurface Sediments
Subsurface sediment samples were collected at 1.5 m intervals to
a depth of 10.7 m in four locations. The results are given in Table 5.
Atrazine, heptachlor, and gamma-BHC were detected at depths of
10.7 m near the former atrazine formulating building. In this same
area, methoxychlor was detected to depths of 3.5 m.
Table 5.
Concentrations of Pesticides at Various Depths
in Monitoring Wells (mg/kg)
Monitoring Well #2 Monitoring Well #4
30
Pesticide
Atrazine
Disulfoton
Toxaphene
Methoxy-
chlor
Heptachlor
Endosulfan
Sulfate
Aldrin
glpha-BHC
gamma-
BHC
(Lindane)
5 10 15 20
18 8.5 6.4 5.3
0.3140.066
1.32 0.46
15 30 35 5 10
15.3 0.35 0.5
9.0 3.9
0.023
15 20
0.22 2.7
0.52 1.1
0.03
25
0.3
0.13
0.06
2.0 0.020.035
O.I500.I440.0630.I50
0.135
1.6 0.73
0.06
0.665
The sursurface data correlate with the composite soil data for
this area. Seven pesticides, including disulfoton, heptachlor,
atrazine, beta-BHC, aldrin, toxaphene and methoxychlor were
detected in the sediments underlying the southeast drum storage
-------
SITE INVESTIGATION
11
area. Neither ethoprop nor phorate were detected at depth. Except
for the concentrations of disulfoton in the southeast drum storage
area, there was no correlation between increasing depth and
pesticide concentrations.
Groundwater
Five wells were installed onsite: one well was up-gradient of the
plant operations and four wells were down-gradient. Atrazine was
detected in all down-gradient wells, with concentrations ranging
from 1.1 mg/1 in well MW #2, located near the former atrazine for-
mulation building, to 0.005 mg/1 in well MW 05, located furthest
down-gradient from the plant. No pesticide contamination was
detected in any of the off site residential wells. Therefore, ground-
water contamination is not a major concern at this site at this time.
However, continued monitoring is recommended.
Buried Waste
During the metal detector survey, two areas of buried metal were
located, both of these being located near the area marked by dead
trees. Following this survey, interviews with former plant
employees indicated that wastes had been buried in the areas in-
dicated by the metal detector. These wastes consisted of 2%
ethoprop baghouse dust, buried from 1976 to 1978.
Allegations have been made that drums of parathion were also
buried along with the ethoprop. Above background readings in the
air of 300 ppm were recorded during a subsurface survey of this
area with a photonionizer. These data suggest that volatile solvents
are also present. To date, these drums have not been excavated and
tested.
SUMMARY AND CONCLUSIONS
The organic vapor survey did not indicate a significant level of
volatile organic compounds in the ambient air. The results of high
volume air sampling did not indicate the presence of pesticide par-
ticulates in the ambient air. The odors at the site are probably due
to residual mercaptan compounds.
The analytical results from the different soil surveys revealed ex-
tensive soil contamination by numerous pesticides in several areas
of the plant. The areas which were identified as being most highly
contaminated were:
•Northwest drum storage area, particularly west of building #5
•Railroad tracks between buildings and north of buildings
•Loading dock area north of former atrazine plant #4
•Area east of building #2
•Area between buildings #1 and #2
•Area east of ethoprop spill (near building #1)
•Southeast drum storage area
Organophosphate, organochlorine and s-triazine pesticides were
detected in the soils of the plant. The compounds of greatest and
immediate concern are the organophosphates, due to their high
mammalian toxicity. Because of the colinesterase-inhibiting nature
of the organophosphates, a human health haard will exist at the site
in the form of discarded waste products and contaminated soil until
removal activities are completed. The organochlorine pesticides,
while being less toxic, are more persistent in the environment, and
less subject to biodegradation.
The on-site groundwater sample analyses revealed atrazine con-
tamination in the four down-gradient wells, the concentrations
ranging from 0.005 to 1.1 mg/liter. Contamination at these levels
does not constitute a serious hazard at the present time, though the
groundwater should be monitored periodically. Off-site residential
wells are not contaminated at the present time.
Any waste buried on-site will need to be excavated, analyzed,
and disposed of properly. Two areas have been identified as having
been areas of buried drummed waste. Since these drums are buried
below the high water table level, they may constitute a threat to
groundwater quality.
RECOMMENDATIONS
Based on the conclusions of this field investigation, it was recom-
mended that the following areas of contamination be addressed as
part of the first phase of cleanup activities:
•Containment of onsite drums
•Removal of concrete basin contents
•Removal of buried tank contents and disposal of tank
•Excavation of areas of buried waste
•Removal of contaminated soil
These activities are described below.
DContainment of drums—An estimated 500 drums are now
stored outdoors and may be in advanced stages of deterioration.
The cleanup tasks must include the characterization, segregation
and containment of these drums. The 1,200 drums stored indoors
must be either recycled or disposed of properly.
D Concrete basin—Two large metal tanks had been drained into
a concrete basin located in the former atrazine building. The main
components of this liquid are liquid atrazine and water. Removal of
the contents of this basin should be conducted in the first phase of
cleanup.
DBuried tank—In discussions with former employees of the
plant and review of Region VII USEPA records it was determined
that a buried tank filled with methoxychlor, malathion, lindane,
atrazine and toxaphene is located onsite. The exact dimensions are
unknown. However, its current capacity is estimated at 300m3. The
cleanup activities shall include removal of the tank's contents along
with removal and disposal of the tank.
D Trenches—With the information obtained during interviews
of former plant employees and local residents it was indicated that
trenches containing drums of ethoprop wastes and possibly
parathion are present on the site. The cleanup will have to include
the uncovering, characterization, removal, and disposal of the con-
tents of these trenches.
DSoil contamination—The extent of soil contamination was
determined from surface sampling and composite subsurface
sampling. The total volume being considered for removal is
estimated at 5400 m3. Removal of this soil will constitute removal
of the majority of the moderately to highly contaminated soil.
Reduced levels of contamination will still exist at the site. Back fill-
ing of these areas will reduce the pathway for contact with remain-
ing subsurface contamination. The backfill material should be cap-
ped with gravel to prevent erosion and allow the site to sustain traf-
fic.
ACKNOWLEDGEMENTS
The author wishes to acknowledge the assistance of G. Philip
Keary, Environmental Services Division, who was the on-site coor-
dinator for Region VII USEPA. Kerry Herndon, Air and Waste
Management Division, acted as regional coordinator for USEPA.
The technical report was prepared with the assistance of Gary
Mason and David Jackson, of E&E, Kansas City office.
REFERENCES
1. Hagger, C. and Clay, P.P., "Hydrogeological Investigations of an Un-
controlled Hazardous Waste Site," Proc. of National Conference on
Management of Uncontrolled Hazardous Waste Sites, Oct. 1981, Wash-
ington, D.C. 45-51.
2. USEPA, Personnel Protection and Safety, Course 165.2, 1980
3. USEPA, Hazardous Waste Site Investigation Training, 1980.
-------
USE OF NDT METHODS TO DETECT BURIED
CONTAINERS IN SATURATED SILTY CLAY SOIL
ROBERT M. KOERNER, Ph.D.
Department of Civil Engineering, Drexel University
Philadelphia, Pennsylvania
ARTHUR E. LORD, JR., Ph.D.
Department of Physics, Drexel University
Philadelphia, Pennsylvania
SOMDEV TYAGI, Ph.D.
Department of Physics, Drexel University
Philadelphia, Pennsylvania
JOHN E. BRUGGER, Ph.D.
U.S. EPA, Municipal Environmental Research Laboratory
Edison, New Jersey
INTRODUCTION
There are estimated to be 30,000 to 50,000 existing dump sites
in the United States containing various amounts and types of
hazardous materials. Furthermore, many new sites are discovered
on a regular basis. One of the first pieces of information needed
in the cleanup process is the physical extent of the dump site and
the resulting polluted area. This is very difficult to do when haz-
ardous materials (often in metal and plastic containers) are buried
beneath the ground surface. Since traditional methods of core bor-
ings and excavation of test pits are dangerous, discontinuous,
and expensive, the use of non-destructive testing (NDT) methods
is often suggested. Many of these methods, including those of the
authors,1'2 have been described in the literature.
In an earlier study,3 we evaluated use of the various NDT meth-
ods to locate buried metal and plastic containers in a uniform dry
sandy soil. In that study, the metal and plastic containers were bur-
ied at known locations and depths in various patterns and seven
NDT methods used for detection. The results indicated that the
metal detector, very low frequency electromagnetic, magnetometer
and ground probing radar techniques are of definite value in delin-
eating the drums. Continuous wave microwave techniques were less
successful, and seismic refraction and electrical resistivity were un-
successful under those particular conditions.
Since the soil and the site of that study5 represented nearly ideal
conditions, it was decided to repeat the entire project by burying
the metal and plastic containers in a saturated fine grained soil
which was eventually located in a construction contractor's storage
yard.
Following this section is a description of the NDT techniques
used, details of the site and specific results obtained.
DESCRIPTION OF EXPERIMENTAL METHODS
Since the continuous wave microwave technique was only mar-
ginally successful and seismic refraction and electrical resistivity
techniques were unsuccessful on the previously described nearly
ideal site,' they were not attempted for this more difficult situa-
tion. Commercially available metal detector (MD), very low fre-
quency electromagnetic (VLF-EM), magnetometer (MA) and
ground probing radar (GPR) were used.4"10
The metal detector (sometimes called a pipe locator or eddy cur-
rent method) and very low frequency electromagnetic methods
operate on essentially the same principle. They will be discussed to-
gether. Both of the instruments used had two coils; many of the less
expensive metal detectors are single coil/inductance change in-
struments. A transmit coil generated an electromagnetic field and a
receiving coil in the vicinity picks up the resulting field. Some of the
field arrives via the air and some via the subsurface material. The
field through the air is essentially constant for a given transmitter-
to-received distance, but the field arriving from the subsurface ma-
terials depends on the subsurface electrical conductivity and mag-
netic permeability. If a conducting body is present in the subsur-
face material between the two coils, the total detected field is al-
tered and the anomaly noted.
A magnetometer measures changes in the earth's magnetic field.
Any magnetic object, e.g., an iron ore deposit or a buried steel ob-
ject, will alter the earth's magnetic field locally and thus can poten-
tially be detected. The most commonly used magnetometer em-
ploys proton nuclear magnetic resonance. The nuclear spin of the
proton processes at a frequency which is linearly porportional to
the total magnetic field at the nucleus. If the total magnetic field
changes because of an anomaly, the precession frequency change
can be read accurately, and hence the magnetic field change can
be determined precisely.
In the ground probing radar technique, a few cycles of electro-
magnetic radiation (100 MHz to 900 MHz) are sent into the ground
from a highly damped antenna. A reflection occurs when a med-
ium of different dielectric constant is encountered. The time it
takes for the pulse to travel down and back given an indication of
the depth of the object. Lateral surveying gives an indication of the
spatial extent of the objects.
SITE DETAILS
The containers were buried in a heavy construction contractor's
storage yard in North Wales, Pa. The 150 ft by 120 ft area was
bounded by trees and a drainage ditch to the north, a chain link
fence to the east and south and miscellaneous forms, tanks and
trailers to the west.
Disturbed and undisturbed samples indicated that the soil was a
dense sandy silty clay of 128 lb/ff unit weight and 19% water con-
tent. This is equivalent to a 98% relative density (via standard
Proctor compaction test) and 100% saturation. Other physical
properties of the soil showed that the specific gravity was 2.54,
the liquid limit was 32%, the plastic limit was 23% and the shrink-
age limit was 11%. Regarding gradation characteristics, 21 % was in
the silt size range and 4% was the clay size range. Thus, the soil
is classified as ML-CL by the Unified Soil Classification system.
Being near the high point of the local topography, the soil was only
about 4 ft to 6 ft thick above bedrock which was observed to be de-
composed red shale, fractured at 4 ft but rapidly became sound at
a depth of about 6 ft.
12
-------
SITE INVESTIGATION
13
;r3*.iT;4ft
c.*
' S,.
Figure 1.
Photographs of Containers After Placement and During Backfilling.
Upper is a 30 Gallon Steel Drum; Lower is a 40 Gallon Plastic Drum.
Eighteen (18) steel and plastic containers were placed in back-
hoe excavated holes that varied from 2 ft to 6 ft deep (see Fig. 1).
The soil cover over the containers varied from 1 ft to 4 ft. The con-
tainer burial patterns were as follows:
•Four 30 gal steel containers buried with 1 ft, 2 ft, 3 ft and 4 ft of
cover.
•Three steel containers (5 gal, 30 gal and 55 gal) buried at 3 ft of
cover. (The 30 gal drum was included in the previous pattern.)
•Four 40 gal plastic containers buried with 1 ft, 2 ft, 3 ft and 4 ft of
cover.
•Four 30 gal steel containers buried with 3 ft of cover. Three of
the containers were adjacent to one another and the fourth was
separated by 16 ft.
•Four 30 gal steel containers with 1 ft of cover were each oriented
at 90 °, 60 °, 30 ° and 0 ° with the horizontal.
In general, all containers were cylindrical, placed with their long
axis horizontal (except where noted), placed empty, placed approx-
1
1
40PI
T
40
I
40
1
4O
•
'4
'
P2
'
P3
|
30S3
30!
1990-
f
30SIQ60*
<
30 S
1
1
I
\
\
TRAILER" 1
30 Sl®0*
L
i
L£
TRAILER" j T
TRAILER " 1 \
1
i
6END
3-30S3 First No. • Size in Gol.
r
1,
iteel (S) or Plastic IP)
.ast No. > Depth in Feet
...... Orientotion • Horiz. Unless
MC. 1 ML Mjil«H
rneuc Noted
>" i . unit Placed after Orumt
SCALE !'• 20'
5S3
Figure 2.
Plan View of Site Boundaries and General Conditions, Buried
Container Deployment and Survey Scan Lines for
Various NDT Techniques.
imately 25 ft from each other (except where noted) and back-
filled with the same soil that was excavated at the particular loca-
tion involved. After hand backfilling and tamping soil around the
drums, the site was further backfilled and compacted by a heavy
dozer, graded off to a level condition and allowed to stabilize for
approximately four months. During this time period the contrac-
tor brought additional equipment onto the site which made the area
more representative of conditions at an actual site. Fig. 2 is a plan
of the site, the containers and other relevant items.
RESULTS
Each of the aforementioned NDT techniques were used at the
site by making a series of seven parallel scans about 10 ft apart (see
Fig. 2), with data being taken at 2 ft intervals along each scan. Fig.
3 contains photographs of two of the techniques during monitor-
ing. Results were transferred from profiles along each scan (re-
sults not shown due to paper size limitations), to generalized plan
views. These generalized results are presented in this paper.
The metal detector (commercially available from Fisher Com-
pany, Model M-Scope)* gave results shown in Fig. 4. The cross
hatched areas along scan lines show where the device gave a pos-
itive indicate that each metal container was accurately located, but
that none of the four plastic containers was located. Positive loca-
tion of a buried object by the device is given by pinning of the
dial, indicator and by an audible signal. Positive signals were ob-
served on scan lines where no containers were purposely buried.
This is understandable since the site had numerous metal objects
(old cans, reinforcing bars, fencing, springs, etc.) very near to the
•Mention of trademarks or commercial products does not constitute en-
dorsement or recommendation for use by the U.S. Environmental Pro-
tection Agency or Drexel University.
-------
14 SITE INVESTIGATION
METAL DETECTOR RESULTS
Figure 3.
Photographs of Sile Showing Metal Detector (upper) and Very Low
Frequency Electromagnetic (lower) Techniques for Detecting Buried
Containers. Crosses Mark Locations Where Containers are Buried.
Figure 4.
Plan View of Site Showing Results of Metal Detector (MD) Survey.
See Fig. 2 for Actual Container Identification.
ground surface before the containers were buried. In the vicinity of
the trailers and metal forms, the metal detector remained pinned
continuously.
The results of the very low frequency electromagnetic device
(commercially available from Geonics Ltd., Model No. EM-31) are
given in Fig. 5. Similar to the previous results, the system accur-
ately located all metal containers, but no plastic ones (Fig. 5). The
possible exception was the plastic container buried 1 ft deep along
scan line #4. Anomalous spots were also seen along scan lines
where no containers were buried but, as described before, quite
possibly for the same reason. At a few of these locations the MD
and VLF-EM readings were in agreement, e.g., along scan line #5
at 90 ft west of the base line.
The magnetometer results (commercially available from EG and
G Geo-Metrics. Model No. F-856) are shown in Fig. 6. The in-
dividual scan lines showing magnetic field data were interpreted
and plotted for this figure. Correlation with actual containers
locations was very poor for the steel drums (which was unexpected)
and for the plastic drums (which was expected). The westerly por-
tion of each scan line became swamped due to the heavy mag-
netic metal (i.e., steel) concentration of the tanks, forms and trail-
ers on the ground surface.
Results from the ground probing radar (commercially available
from Geophysical Survey Systems, Inc., Model No. SIR-7) scans
are not shown in the same format as the preceding techniques
because results were very negative. A typical GPR trace along scan
line K2 is shown in Fig. 7. The trace was made over the four 30 gal
steel drums buried at I ft, 4 ft. 2 ft and 3 ft, respectively, and then
over the 55 gal drum at a 3 ft depth. No discernible return signal
was noted at the proper locations. This was typical of all GPR scans
over the site.
Four separate scan trips were made to the site, one before
drum placement and three afterwards. Perhaps a GPR system with
signal enhancement capabilities would have shown the expected
parabolic shapes indicating a curved object, but it was not ob-
vious in this situation.
CONCLUSIONS AND RECOMMENDATION
In contrast to the earlier study of container detection in a dry
sandy soil, most NOT techniques worked quite well; at this site
conditions were much more formidable. The major differences
between the sites were the:
•High clay content of the soil
•Complete saturation of the soil voice
•Closeness of the bedrock to the ground surface
•The fact that the bedrock surface was not abrupt but weathered
from highly decomposed to very hard within a 2 ft thickness
•Relatively confined area where background noise is present.
In spite of the above difficulties, this is typical of a real site hav-
ing buried containers.
For this situation, the metal detector and very low frequency
electromagnetic methods worked equally well in locating metal
containers. On the basis of equipment cost, the authors would
favor the metal detector ($600 versus $8,000). The VLF-EM has a
deeper penetration depth and lateral scan sensitivity as determined
-------
SITE INVESTIGATION
15
TRAILER*
i
A
K
Jv
^
TRAILER*
TRAILER *
VERY LOW FREQUENCY -
ELECTROMAGNETIC RESULTS^
Shaded Areas are Positive
Results ; Open are Questionable
Figure 5.
Plan View of Site Showing Results of Very Low Frequency
Electromagnetic (VLF-EM) Survey. See Fig. 2 for Actual
Container Identification.
from the earlier study.3 However, neither the MD nor the VLF-EM
will give the depth to the reflecting object.
The magnetometer was unsuccessful because of the abundance
of buried metal objects. This conclusion was expected and con-
firmed by this work.
Somewhat surprising was the failure of the ground probing radar
to detect even the shallow buried containers. Neither steel nor
plastic could be located—a marked departure from the earlier
study.3 Probably the saturated clay soil attenuated the signal before
any significant penetration occurred or the background conditions
submerged the signal in noise. GPR is the preferred technique
known to the authors to give a specific depth to a reflecting ob-
ject, but in this case was simply not successful.
Other NDT techniques that were marginally successful or un-
successful at the sandy soil site3 were not deployed at this satur-
ated clay soil site for the reasons stated earlier.
On the basis of these two studies (reference 3 in dry sandy soil
and this one in saturated clay soils), the authors recommend that
a high quality metal detector be the first NDT method to be used
at a suspect site. Only metal objects, at relatively shallow bur-
ials, can be detected, but this is very often the actual situation.
The device is inexpensive (about $600), can be obtained from a
number of equipment manufacturers, is easy to use, gives both
meter and audible inductions of a buried object and is instantan-
eous. It obviously has drawbacks. Such limitations as only be-
ing able to locate metal objects and poor penetration depth are the
most pronounced, however, its use as the first method to be de-
ployed is highly recommended.
Shaded Areas are
Positive Results
Figure 6.
Plan View of Site Showing Results of Magnetometer (MA) Survey.
See Fig. 2 for Actual Container Identification.
GROUND PROBING
RADAR RESULTS
FOR SCAN LINE 2
Known Locations of Buried Drums
Figure 7.
Ground Probing Radar (GPR) Trace Over Scan Line #2 Showing
Arrows Where Five Containers Are Actually Located.
-------
16 SITE INVESTIGATION
ACKNOWLEDGEMENTS
The authors would like to thank Francis A. Canuso, HI,
and Wallace A. Rutecki of Contec Construction Company, North
Wales, Pa., for allowing access to the site and providing equip-
ment and logistical support. Their help was indispensable. Thanks
are also due to the U.S. Environmental Protection Agency, Office
of Research and Development, Municipal Environmental Research
Laboratory, Edison, New Jersey, for financial support under Co-
operative Agreement No. CR80777710.
REFERENCES
1. Lord, A.E., Jr., Tyagi, S.D. and Koerner, R.M., "Nondestructive
Testing (NOT) Methods Applied to Environmental Problems Involv-
ing Hazardous Material Spills," Proc. of 1980 Conference on Con-
trol of Hazardous Material Spills, May 13-15, 1980, Louisville, Ky.
174-179.
2. Lord, A.E., Jr., Koerner, R.M. and Freestone, F.J., "The Identifi-
cation and Location of Buried Containers via Nondestructive Test-
ing Methods," Journal of Hazardous Materials, 5, 1982,221-233.
3. Lord, A.E., Jr., Tyagi, S., Koerner, R.M., Bowders, J.J., Sankey,
J.E. and Cohen, S., "Use of Nondestructive Testing Methods to
Detect and Locate Buried Containers Verified by Ground Truth
Measurements," Proc. 1982 Hazardous Materials Spills Conference,
April 19-22, 1982, Milwaukee, WI, 185-191.
4. Lord, A.E., Jr., Tyagi, S.D. and Koerner, R.M., "Use of Eddy Cur-
rents as a Potential NOT Method Applied to Hazardous Material
Spill Problems," Report 2 to EPA on Cooperative Agreement
CR80777710, January 6,1981 (available on request).
5. Lord, A.E., Jr., Tyagi, S.D. and Koerner, R.M., "Use of Very Low
Frequency-Electromagnetic (VLF-EM) Measurements as a possible
NOT Method Applied to Hazardous Material Spill Problems," Re-
port 8 to EPA on Cooperative Agreement CR80777710, July 4, 1981
(available on request).
6. Lord, A.E., Jr., Tyagi, S.D. and Koerner, R.M., "Use of Proton
Precession Magnetometer as a Potential NDT Method Applied to
Hazardous Material Spill Problems," Report 4 to EPA on Coop-
erative Agreement CR80777710, February 20, 1981 (available on re-
quest).
7. Cook, J.C., "Ground Probing Radar," Proc. of Symposium on Sub-
surface Exploration, Underground Excavation and Heavy Construc-
tion, American Society of Civil Engineers, 1974, 172-174.
8. Morey, R.M., "Continuous Subsurface Profiling by Impulse Radar,"
Proc. of Symposium on Subsurface Exploration for Underground
Excavation and Heavy Construction, American Society of Civil En-
gineers, 1874,213-232.
9. Rosetta, J.V., "Detection of Subsurface Cavities by Ground Probing
Radar," Proc. of Symposium on Detection of Subsurface Cavities,
U.S. Army Waterways Experiment Station, Vicksburg, MS, 1977,
120-127.
10. Lord, A.E., Jr., Tyagi, S.D. and Koerner, R.M., "Use of Ground
Probing Radar and CW Microwave Measurements as Possible NDT
Methods Applied to Hazardous Material Spill Problems," Report 13
to EPA, Edison, New Jersey, on Cooperative Agreement CR80777710,
October 1, 1981 (available on request).
-------
SYSTEMATIC HAZARDOUS WASTE SITE ASSESSMENTS
R.B. EVANS, Ph.D.
Environmental Monitoring Systems Laboratory
U.S. Environmental Protection Agency
Las Vegas, Nevada
R.C. BENSON
J. RIZZO
TechnosInc.
Miami, Florida
INTRODUCTION
The National Contingency Plan gives the objectives of prelim-
inary site assessment as: (1) evaluating the magnitude of the haz-
ard, (2) identifying the source and nature of the release, and (3)
evaluating the factors necessary to make the determination of
whether immediate removal is necessary.1 For this paper, these ob-
jectives can be restated in terms of the following goals for sub-
surface characterization: (1) determining the hydrogeologic char-
acteristics of the site, (2) locating and mapping leachate plumes,
and (3) locating and mapping buried wastes.
Systematic Assessments
In the past, investigations of hazardous waste sites were usually
dependent on drilling to obtain information about the geological
setting, on monitoring wells for samples of groundwater, and on
laboratory analyses of groundwater, soil, and waste samples. This
costly approach may fail to adequately characterize the site set-
ting, location of buried wastes, and the extent of groundwater
contamination.
There are several reasons why a network of monitoring wells
may poorly define a contaminant plume or produce an incomplete
picture of site hydrogeology, but probably the chief cause is that
the subsurface is rarely homogeneous. Substantial variations in
permeability or hydraulic conductivity can be produced by a
change of only a few percent in the silt or clay content of a sandy
aquifer. Fig. I2 illustrates the effect of small-scale heterogeneities
on the migration of a contaminant in a sandy medium. Such het-
erogeneities complicate the siting of monitoring wells. Other
sources of error include well construction, sampling techniques,
and even improper logging of soil and rock samples. Ironically, the
groundwater samples collected from well networks can be analyzed
for trace contaminants in the part-per-million or part-per-billion
range with great accuracy. In many groundwater contamination
investigations, the accuracy of such analyses may be much better
than the interpretation of the subsurface afforded by the well net-
works.
During the past decade, extensive development in geophysical
survey equipment, field methods, analytical techniques, and asso-
ciated computerized data processing has greatly improved the char-
acterization of hazardous waste sites. Some geophysical techniques
can detect contaminant plumes and flow direction. Some are ap-
plicable to measurements of contaminants and direction of flow
within the vadose zone; others offer detailed information about
subsurface geology. The capability to characterize the subsurface
directly without disturbing the site offers benefits in terms of less
cost and less risk. It can also achieve a more complete understand-
ing of the subsurface by filling gaps in data in the areas between
exploratory sampling wells.
Cost-effective preliminary assessments of hazardous waste sites
involves an integrated, three-phased approach:
•Review of available data, including the use of aerial photography,
on-site inspections, and review of readily available literature and
local information to characterize the site geology, hydrogeology,
and possible waste composition
•Geophysical surveys combined with exploratory drilling and
sampling
•Design and implementation of a monitoring program including
monitoring wells
This systematic approach defines the areas of contamination and
thus supports necessary planned removal of wastes or other remed-
ial action.
HIGHER
PERMEABILITY
LAYERS
Figure 1.
The Effect of Small Scale Heterogeneities on the Pattern of
Contaminant Migration
Table 1.
Some Typical Sources of Data Useful in a Systematic Assessment
of Waste Disposal Sites (from Le Grand, 1980)
Type of Data
Property Survey
Well Drillers Logs
Water Level Measurements
Topographic Maps
Air Photos
County Road Maps
Ground Water Reports
Soil Surveys of Counties
Geologic Maps
Climatological Data
Typical Sources
County Records, Property Owner
Well Driller, Property Owner, State
Records
Well Owners Observations, Well Drillers
Logs, Topographic Maps, Ground Water
Maps (Reports)
U.S. Geological Survey and Designed
State Sales Offices
U.S. Dept. of Agriculture, U.S. Forest
Service, etc.
State Agencies
U.S. Geological Survey, State Agencies
U.S. Dept. of Agriculture
U.S. Geological and State Surveys
U.S. National Oceanographic and
Atmospheric Agency
REVIEW OF AVAILABLE DATA
Much data are commonly available for sites in a variety of docu-
ments, reports, maps, etc. Some of the data sources are found in
Table 1.3 This information should be used first to gain an overview
of the site's relationship to the regional setting. This can be accom-
plished by a review of topographic maps, geologic maps, aerial
photographs and federal and state geohydrologic publications. Nel-
17
-------
18
SITE INVESTIGATION
son, et at.' described in detail a methodology for inventorying and
prioritizing possible uncontrolled waste disposal sites based on
available data. This methodology includes a systematic approach
to compiling, reviewing, and interpreting available data.
Aerial photography is a valuable tool for preliminary hazardous
waste site assessments; both current aerial photos and historical
imagery are useful. Historical (archival) photographs may date
back 40 years or more and can play a vital role in assessing spe-
cific sites. They can often trace the life of a waste disposal site
from its creation to the present. Locations of old landfills or dis-
posal ponds now covered by subsequent land use or vegetation
growth can be determined from historical photographs. The his-
tory of one disposal site is traced in Fig. 2 from 1945 through 1979
and shows the construction of residential neighborhoods over and
around the former site.'
Archival imagery can be obtained from the National Archives,
the data center of the U.S. Geological Survey, and other sources
such as city, county, or state agencies. Usually imagery taken at
three or more dates from the late 1930s to the present is examined.
If necessary, current photography is obtained from overflights as
part of the preliminary assessment. The history of the land use
around a site and drainage patterns should be carefully studied.
The potential exposures to nearby residences and direct and indi-
rect environmental pathways to them must be evaluated. Frequent-
ly, detailed imagery analysis is coupled with ground investigations
to provide a more complete picture of potential environmental
problems. The end product' of a detailed analysis is shown in
Fig. 3.
Waste Characterization
Waste characterization prior to beginning field work is highly de-
sirable. Valuable information about waste composition, assuming
legal considerations are not involved, can be obtained from the
property owner and/or those responsible for transporting or dis-
posing of the waste. Where a manufacturing or processing facility
was involved in the generation of the wastes of interest, the spe-
cifics of waste characteristics may sometimes be obtained from a
variety of corporate records, disposal logs, and conversations with
technical personnel. Wastes related to military bases can be char-
acterized as to their shape, size, construction and chemistry. The
military usually maintains this type of documentation.
SELECTING THE APPROPRIATE
GEOPHYSICAL TECHNIQUE
Contemporary geophysical techniques and their advantages and
limitations have been extensively described in many recent publica-
tions. A few of these key references are cited in the bibliography.
A brief description of the theory of operation and the applica-
tion of the primary geophysical techniques which have regularly
and successfully been applied for waste site assessments follows:
Ground Penetrating Radar (GPR) is designed to detect the pres-
ence and depth of subsurface features using radar waves trans-
mitted from a small antenna moved across the ground's surface.
This analysis results in a continuous cross-sectional "picture" or
profile of shallow, subsurface conditions. The responses are caused
by radar wave reflections from interfaces of materials having
different electrical properties. The reflections are often associated
with natural hydrogeologic conditions such as bedding, cementa-
tion, moisture and clay content, voids, fractures, and intrusions
as well as man-made objects. GPR has been used at numerous haz-
ardous waste sites to evaluate natural soil and rock conditions and
to detect buried wastes.
The Electromagnetics (EM) method measures the electrical con-
ductivity of subsurface soil, rock and groundwater. Electrical con-
ductivity is a function of the type of soil and rock, its porosity,
and the conductivity of the fluids which fill the pore space. In most
cases, the conductivity (or specific conductance) of the pore fluids
will dominate the measurement. The EM method is applicable to
both the assessment of natural geohydrologic conditions and map-
ping of many types of contaminant plumes. Trench boundaries,
buried wastes and drums, as well as metallic utility lines can also be
located with EM techniques.
HEAVY* j
SPILLA
IN DRUM
< STORAGE
' AREA
DRUM
STORAGE
^.^ rREWCHTJ, ^ ppfTA *
7 "Ntv^tewfr-^ «>5??
RECENT
CONTAINMENT
BERMS
^»*Sst
1966
Figure 2.
Overhead Imagery Used to Trace the History of One Site
Figure 3.
Typical End Product of an Aerial Site Investigation
-------
SITE INVESTIGATION
19
The Resistivity method measures the electrical resistivity of the
geohydrologic section which includes soil, rock and groundwater.
Interpretation of these measurements provides depth and thickness
of geologic strata as well as lateral changes in the subsurface.
Usually the presence, quantity and quality of groundwater are the
dominating factors influencing the resistivity value. The method
may be used to assess lateral and vertical changes in natural geo-
hydrologic settings as well as evaluate contaminant plumes at haz-
ardous waste sites. The method may also be used to locate buried
wastes.
Seismic Refraction techniques are used to measure the depth and
thickness of geologic strata using acoustic waves transmitted into
the ground. Seismic methods also provide measurements of rock
density and are often used to map depth to specific horizons such
as bedrock, clay layers and water table. Secondary applications of
the seismic method include location and definition of burial pits
and trenches at hazardous waste sites.
Metal Detectors (MD), commonly used by treasure hunters
searching for coins and by utility crews for locating buried pipes
and cables, are used to locate buried metallic objects. They can
detect metallic materials, including ferrous objects (iron and steel)
and non-ferrous metals such as aluminum and copper. In haz-
ardous waste site investigations, metal detectors are useful in de-
tecting buried drums and in delineating trench boundaries contain-
ing metallic containers at shallow depths.
A Magnetometer responds to distortion in the earth's natural
magnetic field. These distortions can be caused by the presence of
buried ferrous metals (iron or steel); a magnetometer will not re-
spond to non-ferrous metals. It will respond to variations of iron-
oxide concentrations in soil and rock as well as nearby cultural
features made of iron or steel. Magnetometers are regularly used
in locating buried drums, estimating their numbers, and in deter-
mining the boundaries of trenches containing drums.
When selecting the appropriate geophysical technique to meet a
mission objective, it should be recognized that more than one tech-
nique can usually be applied. The complementary nature of data
obtained from combining geophysical techniques can reduce the
errors in interpretation that have been common in the past.
The typical objectives of preliminary waste site assessments and
the geophysical techniques which can be applied are summarized
in Table 2. The reader is referred to "The Use of Selected Geo-
physical Remote Sensing Methods in Hazardous Waste Site Inves-
tigations'" for a more thorough treatment of the use of geophysics.
Interpretation
While geophysical surveys can sometimes stand alone, field in-
vestigations will usually require additional quantitative data. In-
Table 2.
Potential Applications of Geophysical Methods
Electro-
Application Radar Magnetics
Mapping of geology 1 1
and hydrogeologic
features
Mapping of con- 2 1
ductive leachates
and contaminant
plumes (e.g., land-
fills, acids, bases)
Location and bound- 1 1
ary definition of bur-
ied waste trenches —
without metal target
Location and bound- 1 1
ary definition of bur-
ied trenches — with
metal targets
Location and def- 2 2
inition of buried
metallic objects (e.g.,
drums, ordinance)
Metal Magneto-
Resistivity Seismic Detector Meter
1 1
1
2 2
2 222
1 1
formation from drilling logs and groundwater samples, obtained
from existing data and/or supplemental newly-acquired data can
provide the additional information needed for interpretation. The
following examples highlight the use of ground truthing informa-
tion.
The radar "picture" shown in Fig. 4 resulting from a field sur-
vey does not indicate the soil type—only that differences in soils
have been seen. A drilling log will provide the means to correlate
the radar data with actual soil type (Fig. 4).
Similarly, an EM record and the drilling logs which allowed
correlation of the EM data and permitted definition of the soils
in the area (Fig. 5).
The correlation of the direct sampling with the geophysical data
will allow the investigation to proceed with the design and subse-
quent installation of two monitoring welj[networks.
FINE
-QUARTZ
SAND
CLAY
LOAM
Figure 4.
Radar Profile of Soil Section'
2 -
I «-!
10-
T
PROFILE OF THICKNESS OF ORGANICS BASED ON
THREE BORINGS
2-
10-
POSITION IN FEET
1—Primary method—indicates the most effective method
2 Secondary method—indicates an alternate approach
THICKNESS OF ORGANICS BASED UPON CONTINUOUS EM
MEASUREMENTS
Figure 5.
Comparison of Continuous Electromagnetic Data with Three Test Borings
-------
20
SITE INVESTIGATION
GROUNDWATER MONITORING
The objective of groundwater monitoring in hazardous waste
site assessment is to provide long term information about the area!
extent of groundwater contamination and the concentrations of
contaminants, for the purposes of insuring the safety of water sup-
plies or measuring the effectiveness of corrective action. The de-
sign of a monitoring well network should consider many variables,
including direction and rate of groundwater flow, soil permeability,
and background water quality, all of which may interact in a com-
plex fashion.
Discussions of monitoring well placement are available in the
literature.' One set of guidelines states:
"In order to detect and evaluate potential or existing ground-
water contamination at a landfill, a minimally acceptable mon-
itoring well network should be implemented and consist of the
following:
•One line of three wells downgradient from the landfill and situated
at an angle perpendicular to groundwater flow, penetrating the
entire saturated thickness of the aquifer
•One well immediately adjacent to the downgradient edge of the
filled area, screened so that it intercepts the water table
•A well completed in an area upgradient from the landfill so that
it will not be affected by potential leachate migration
The size of the landfill, hydrogeologic environment, and
budgetary restrictions are factors which will dictate the actual
number of wells used. However, every effort should be made to
have a minimum of five wells at each landfill and no less than one
downgradient well for every 250 ft of landfill frontage."
Additional consideration must be given to the effects of the
aquifer characteristics. The reference publication does on to say:
"When considering the design of the monitoring system,
aquifers can be subdivided on the basis of permeability and
porosity. Although the design previously described could gen-
erally be applied to all aquifers, aquifer parameters should dic-
tate the following:
•monitoring well density, depth, construction, and drilling
methods
•probability of successful detection of the contaminant plume
•sampling methods
For the monitoring network to be effective, the basic network
design at a particular site will require modification according to
geologic conditions."
Ideally, the extent and location of the plume would be known
prior to construction of the well network. In practice, monitor-
ing well locations are generally chosen on the basis of limited
knowledge of the site geology and hydrogeology and the profes-
sional judgment of the investigator. The accuracy and effective-
ness of this approach is limited by the implicit assumption that sub-
surface conditions are uniform in the horizontal direction, which is
usually not true. This assumption can lead to poorly located wells,
with inaccurate interpretations of the subsurface conditions.
Geophysical surveys offer a means of obtaining beforehand es-
timates of the natural hydrogeologic conditions and/or the extent
of contaminant plumes. The vertical configuration of the plume to
be monitored and the subsurface geology defined in the previous
phases of the program should dictate well construction specifica-
tions. For example, clay lenses would not provide the optimum
location for well sampling screens. If present, permeable sand lay-
ers would be more suitable.
Limiting the vertical extent of well screening (zone of comple-
tion) to the zone of contamination would yield more accurate esti-
mates of actual contaminant concentrations in groundwater. As
geophysical techniques are capable to developing a comprehensive
picture of the hydrogeologic site conditions both laterally and ver-
tically, a well can be screened to the proper depth for sampling.
The location and completion depths of existing monitoring weUs
can also be put into perspective.
This is a brief description of the systematic approach to haz-
ardous waste site assessments. Implementation of this approach
must typically be modified to suit the actual conditions of each site.
Economics, legal implications, local reaction, schedule, site access,
and lack of existing information are some of the considerations
which can and do constrain and modify a systematic approach.
The following case studies illustrate a systematic approach ap-
plied to a number of site assessments and further show the typical
areas of compromise in actual .projects. Each example is based on a
project completed by the co-authors' firm in the past two years.
EXAMPLE 1—BULK CONTAMINANTS
A large playground was suspected of containing buried haz-
ardous waste materials. The type of contaminant materials, their
exact location and quantity were all unknown. Six soil sampling
and monitoring well locations were in place. The size of the total
area of investigation was 25 acres, resulting in a sample density of
approximately four acres per hole. There were indications of con-
taminants in the soil, groundwater, and a nearby stream. Aerial
photos provided the overall setting picture and showed old buried
stream channels in the immediate area of the site. No indication
of dumping was noted. The objectives of the subsequent survey
were to locate the precise boundaries of the contaminants, estimate
their volume and nature, provide a design for a monitoring well
program and, if possible, estimate local groundwater flow direction
and speed.
When the size of the target and location are unknown, discrete
sampling via drilling is a very uneconomical approach to the target
location problem. Therefore, non-destructive geophysical methods
were selected (see reference 8 for a more complete treatment of
drilling and geophysics costs).
Shallow EM measurements were used to locate the bulk contam-
inants and define their exact perimeter. Further EM soundings at
the site provided the approximate thickness of contaminants and
their maximum depths. Ground penetrating radar was used to eval-
uate the soil cover over the contaminants and further confirm the
contaminant perimeter. A magnetic survey indicated that only a
few steel objects were present in this rather large site. This informa-
tion suggsted that the waste material was bulk-dumped rather than
containerized within steel drums.
The survey indicated that the ratio of the total site area to the
area of the contaminants was approximately 10:1. This ratio re-
quires that more than ten drilling locations or sample points are
needed to yield a detection probability of greater than 95%.
Not surprisingly, the six existing exploratory holes had all missed
the contaminated area (Fig. 6). Additional holes would be required
to define both the boundary and the vertical extent of the contam-
inants. Fifteen to twenty sample points would only approximate
the geophysical data net shown in Fig. 6.
3-0 REPRESENTATION
OF CONDUCTIVITY DATA SHOWING
BURIED HAZARDOUS MATERIALS
WELL LOCATION
Figure 6.
Monitoring Wells Located Prior to Electromagnetic Survey Mi«
Buried Waste Site
-------
SITE INVESTIGATION
21
This survey provided the location, boundaries, depths and esti-
mated volumes of the bulk contaminants. The contaminants were
identified as highly electrically conductive and subsequently iden-
tified by drilling and sampling as a fly-ash material with a heavy
metal content.
Conductive plumes were detected in the immediate site area, in-
dicating minimum migration of conductive contaminants from the
site. Further, locations for the installation of monitoring wells were
suggested by the presence of a stream on one side and a buried
stream channel on the opposite side of the site. The most probable
migration paths were either into the adjoining stream channels or
downward into underlying strata.
This information improved the reliability of the evaluation of the
site location, contents, arid potential for contamination. It also
provided guidance for location of subsequent monitoring wells and
cost estimating and planning of remedial action.9
EXAMPLE 2—LANDFILL PLUME
A landfill approximately one square mile in area had been in use
for 29 years. An extensive network of near-field monitoring wells
was installed in and around a landfill to evaluate groundwater
contamination. The total depth of the aquifer in this location is
about 90 ft. The plume was sampled to a maximum depth of ap-
proximately 50 ft. The monitoring wells had characterized the
plume from the landfill along the vertical plane defined by the line
of monitoring wells, and detailed water chemistry was available
from groundwater samples. The initial 1974 data front the monitor-
ing wells are shown in Fig. 7.
The spatial extent of the plume, however, was unknown. There
was no information to the north or south or farther downgrad-
ient beyond the monitoring wells. The objectives of an initial geo-
physical survey in 1977 were to define and map the spatial extent
DISPOSAL SITE
V WELL LOCATION
Figure 7.
Isopleths of Specific Conductance (p mho/cm) from 1974 Well Data
(locations shown in Figure 8)
REGIONAL
GROUNDWATER FLOW
(EASTI
A INDICATES LOCATION OF WELLS
WELL FIELD
FOR MUNICIPAL
SUPPLY
Figure 8.
Plume Configuration Estimated from 1977 Resistivity Survey
(See resistivity cross-section A-A' on Figure 7)
of the plume and to provide recommendations for additional mon-
itoring wells if required. The results of the 1977 resistivity survey
are shown in Fig. 8.
A few years after the resistivity survey, a new auxiliary well field
was installed nearer the landfill—approximately 1.5 miles in an
approximate downgradient groundwater flow direction. This well
field had been pumping intermittently for approximately two years
(1979-1981) when a second geophysical survey was requested. The
landfill had been identified by USEPA as one of the major haz-
ardous waste sites in the country because of its proximity to the
local well field supplying drinking water to a large city.
The objectives of the subsequent 1981 survey were to evaluate
any changes in the plume and the influence of the new water well
field.
Monitoring well data were again reviewed. The contaminant
plume inferred from the 1981 well data alone is shown in Fig. 9.
Inspection of aerial photographs showed that during the four year
period between the first and second surveys, considerable develop-
ment had taken place in the area, and potential auxiliary sources
of contaminant were present. Available areas that could be readily
surveyed with geophysical techniques were identified at this time.
Resistivity was again selected as the survey method so that the
new survey measurements could be correlated with the earlier
measurements. The electromagnetic technique was used to provide
a more extensive and detailed data base and to improve the sta-
tistical validity of the survey. The new survey area was extended
beyond the earlier survey to cover more area down gradient. The
results of the new survey, plotted in Fig. 10, showed the plume after
a four-year period. There are important differences between the
plume inferred from the monitoring well data alone and the plume
inferred from the geophysical survey. The contamination observed
in the well to the south was found to originate from a source not re-
lated to the landfill.
A resistivity transect (line B-B' in Fig. 10) runs through the
plume and the well which was indicating contamination can put the
REGIONAL
GROUNDWATER FLOW
IEAST)
Figure 9.
Plume Configuration Estimated from 1981 Well Data Only
REGIONAL
GROUNDWATER FLOW
(EAST)
ACTIVE
(—) AUXILIARY
LJ WELL FIELD
MONITORING
WELL
UN IMPERMEABLE
ZONE I
WELL FIELD
FOR MUNICIPAL
SUPPLY
Figure 10.
Plume Configuration Estimated from 1981 Electromagnetics Survey
(B-B' indicates location of resistivity cross-section)
-------
22
SITE INVESTIGATION
Figure 11.
Resistivity Transect (Along cross-section B-B' in Figure 10)
existing well data and the geophysical data in perspective. The data
obtained along the transect are shown in Fig. 11. The main con-
taminant plume clearly does not reach the well in question. This
well is contaminated from other sources which were identified.
The node bisecting the outer edge of the plume was created by a
low permeability zone in which leachate was not penetrating. Most
importantly, the plume was clearly responding to the pumping of
the new auxiliary well field to the northeast.
The plume, as shown in Fig. 10, is representative of the conserva-
tive parameters measured by the electrical methods. For landfill
contaminants, these boundaries will typically represent a worst-
case condition. Most contaminants of a hazardous nature will lie
within these boundaries and often will have migrated less distance
than these conservative parameters. The locations for additional
monitoring wells can be precisely determined after these plume con-
figurations are defined. In addition to accomplishing this objec-
tive, the survey identified a number of other point sources which
were contributing to the contamination of the aquifer (not shown
in the data).
Probably, the most significant result of this work was the
demonstrated ability to measure, through successive geophysical
surveys, the migration of the contaminants over the four-year study
period.'°
APPLYING THE SYSTEMATIC APPROACH
The case examples have illustrated the systematic approach in
practice. It is important to understand the organizational struc-
ture and tools needed to correctly implement the systematic ap-
proach (Fig. 12). None of the technical tools or skills taken by it-
self is a panacea.
Project success will be achieved through the proper selection,
combination and balance of all of these elements. When these ele-
ments are combined properly, a synergism results which can signif-
icantly improve understanding of site conditions.
CONCLUSIONS
Relative costs and benefits of an integrated approach have been
discussed by Benson, et al.1 The use of geophysics considerably
reduces the total cost of most hazardous waste site assessments.
As site complexity increases and/or sites get larger, particularly
relative to the area of interest, geophysical-assisted projects become
increasingly more cost effective. Further geophysics may be the
only reasonable alternative when there is risk of puncturing waste
containers during drilling.
Knowledge gained in initial literature and aerial photography
review will aid in focusing a cost-effective systematic site assess-
ment.
Figure 12.
Technical Resources and Tools for Subsurface Investigations
of Hazardous Waste Sites
There is increasing recognition that hazardous waste site assess-
ments are complex. Shortcuts or elimination of essential elements
in a systematic approach will often result in failure to accurate-
ly or effectively characterize the site. Therefore, all available tech-
nical resources and tools should be considered in a subsurface in-
vestigation of a hazardous waste site.
REFERENCES
1. USEPA, "National Oil and Hazardous Substances Contingency
Plan," Federal Register, July 16, 1982, 31180-31243.
2. Freeze, R.A. and John A. Cherry, Groundwater, Prentice-Hall,
Englewood Cliffs, N.J., 1979.
3. LeGrand, H.E., "A Standardized System for Evaluating Waste Dis-
posal Sites," National Water Well Association, Worthington, Ohio,
1980.
4. Nelson, Ann B., and R.A. Young, "Location and Prioritizing of
Abandoned Dump Sites for Future Investigations," Proc. of the Na-
tional Conference on Management of Uncontrolled Hazardous Waste
Sites, Washington, D.C., Oct. 1981, 52-62.
5. Advanced Monitoring Systems Division, EMSL-LV, EPA, "Aerial
Reconnaissance of iVertac, Inc., Jacksonville, Arkansas," Las Vegas,
Nevada, June 1980.
6. USEPA, "The Use of Selected Geophysical Sensing Methods in Haz-
ardous Waste Site Investigations," prepared by Technos, Inc., draft
in review, to be published in Fall 1982.
7. USEPA, "Procedures Manual for Groundwater Monitoring at Solid
Waste Disposal Facilities," EPA 530/SW-ll, 1977.
8. Benson, R.C., R.A. Glaccum, and P. Beam, "Minimizing Cost and
Risk in Hazardous Waste Site Investigations Using Geophysics,"
Proc. of the National Conference on Management of Uncontrolled
Hazardous Waste Sites, Washington, D.C., Oct. 1981, 84-88.
9. Noel, M.R., R.C. Benson, and R.A. Glaccum, "The Use of Contem-
porary Geophysical Techniques to Aid Design of Cost-Effective Mon-
itoring Well Networks and Data Analysis," Proc. of the Second Na-
tional Symposium on Aquifer Restoration and Groundwater Monitor-
ing, Columbus, Ohio, 1982.
10. Glaccum, R.A., R.C. Benson, and M.R. Noel, "Improving Accur-
acy and Cost-Effectiveness of Hazardous Waste Site Investigations
with a New Generation of Geophysical Methods," Proc. of the Na-
tional Water Well Association, Groundwater Technology Division
Kansas City, 1981.
11. Johnson, R.W., R.A. Glaccum, and R. Wajtasinski, "Application of
Ground Penetrating Radar to Soil Survey," Proc. of the Soil and
Crop Science Society of Florida, 39, 1980.
-------
DETERMINATION OF RISK FOR UNCONTROLLED
HAZARDOUS WASTE SITES
ISABELL S. BERGER
RECRA Research, Inc.
Amherst, New York
INTRODUCTION
The development of risk assessment at both controlled and un-
controlled hazardous waste sites is perhaps the fastest growing area
of environmental investigation. Various federal and state agencies
as well as private consulting firms have become involved in the de-
termination of risk through initiating legislation, developing rules
and regulations, conducting site searches and investigations, as well
as developing new and more sophisticated risk assessment models.
Risk is usually defined as a function of the probability of an event
occurring and the magnitude or severity of the event should it oc-
cur,1 while risk assessment has been defined as, "the identifica-
tion of hazards, the allocation of cause, the estimation of prob-
ability that harm will result, and the balancing of harm with bene-
fit."2
Frequently, the assessments also include the cost of risk reduc-
tion compared to the public benefit. Risks associated with haz-
ardous waste disposal include threats to public health and safety
due to environmental degradation of air, water and land resources,
fire and explosion, exposure to carcinogenic, mutagenic or terato-
genic chemicals, as well as socioeconomic and legal risks due to the
loss of property value and potential liability suits.
The type and extent of a risk determination should be based
upon a general series of assessment criteria as well as an under-
standing of the ultimate purpose of the investigation. The nature
of the program will vary depending upon whether the assessment
will be used to distinguish priorities among numerous uncontrolled
sites, to initiate emergency removal, to evaluate sites for insurance
purposes, or to recommend long-term remedial measures. The pur-
pose of this paper is to illustrate how general risk assessment meth-
odology can be used in conjunction with a detailed analytical and
mass balance program in order to develop appropriate remedial
action recommendations.
THE NATIONAL CONTINGENCY PLAN
Section 105 of the Comprehensive Environmental Response,
Compensation and Liability Act, P.L. 96-510 (known as CERCLA
or Superfund) requires that the National Contingency Plan (NCP)
be revised to include the removal of oil and hazardous substances
and "shall establish procedures and standards for responding to
releases of hazardous substances, pollutants, and contaminants."3
According to the Act the plan shall include:
•Methods for discovering and investigating facilities
•Methods for evaluating and remedying releases or potential re-
leases
•Methods and criteria for determining extent of removal or remedy
•Means of assuring cost effective remedial action measures.
CERCLA goes on to say,
"Criteria and priorities under this paragraph shall be based upon
relative risk or danger to public health or welfare or the environ-
ment...taking into account, to the extent possible, the population
at risk, the hazard potential of the hazardous substances at such
facilities, the potential for contamination of drinking water sup-
plies, the potential for direct human contact, the potential for
destruction of sensitive ecosystems..."3
On July 16, 1982, the NCP Rules and Regulations were pub-
lished in the Federal Register.* In complaince with CERCLA, the
NCP provides direction on the criteria for conducting preliminary
assessment, undertaking immediate removal, evaluating planned
removal and undertaking remedial action including determination
of the appropriate extent of the response. Specific actions described
by the NCP for the preliminary assessments are: the evaluation
of the hazard magnitude, identification of the nature and source
of the release, determination of responsible parties, and evaluation
of background information.4
Once it has been determined that a site does not require immed-
iate remediation, the more detailed process of evaluation to deter-
mine appropriate response would begin. The general areas of con-
cern are: whether there is a population at risk, the amount and
form of hazardous substances, the hazardous properties of those
substances, the hydrogeologic factors affecting the site and the cli-
mate. Based upon the NCP, the appropriate response to such a
site upon discovering that a threat exists includes:
•Collecting and analyzing data
•Developing a limited number of alternatives
•Screening the alternatives based on cost, environmental, health,
and engineering criteria
•Refining the alternatives and performing detailed analysis (in-
cluding engineering implementation, constructibility, extent of
migration, detailed costs, and environmental impacts)
•Selecting the most cost-effective alternative that will mitigate the
danger and provide adequate protection
•Balancing the need for protection against the amount of money
available in the Superfund
Although the NCP was designed for uncontrolled sites where the
federal government has become involved, it can serve as an out-
line for appropriate response at numerous hazardous waste sites
which are not on the national priorities list. The assessment cri-
teria, however, should be considered minimum requirements. The
criterion of selecting the most cost-effective alternative which
"effectively minimizes and mitigates the danger and provides
adequate protection of public health, welfare and the environ-
ment" should be recognized as a general standard. Time and cost
constraints are usually the limiting factors for all types of remed-
ial action whether cleanup is being undertaken by industry or gov-
ernment agencies. Although the same type of balancing of protec-
tion against monies available in the fund would not apply, it should
be accepted that industries will apply the most cost-effective altern-
ative when designing and implementing remedial measures.
GENERAL RISK ASSESSMENT METHODOLOGY
The type and extent of risk determination undertaken will de-
pend upon the intended ultimate use of the assessment. The pur-
pose would shape the data requirements, the methodology, the
23
-------
24
SITE INVESTIGATION
rigor of the assessment, as well as the eventual value of the assess-
ment for secondary purposes. Potential objectives include estab-
lishing priorities among various uncontrolled waste sites, assess-
ing imminent health or environmental hazard, determining legal
and/or insurance liability, and evaluating remedial action require-
ments. Normally, the last of these objectives would require the
most extensive assessment.
General criteria have been developed which would apply to all
types of preliminary investigations. These are usually based upon
an investigation of available information related to essentially four
characteristics: receptors, pathways, waste characteristics, and
waste management practices.' The labels may differ depending
upon the particular risk assessment model, but the basic areas of
concern remain the same. In the Rating Methodology Model these
factors are defined as follows:
•Receptors—humans and other organisms that may be exposed to
hazards from the site
•Pathways—the routes or media by which hazardous materials
may escape from the site
•Waste characteristics—the hazardous properties of the waste in-
cluding quantity, mobility, toxicity, ignitability and persistence
•Waste management practices—the design characteristics and pro-
cedures used in managing and containing wastes'
The relationship between pathways and receptors in a simple
schematic diagram (Fig. I).6 The general information requirements
for a risk assessment at either a controlled or uncontrolled haz-
ardous waste site are listed in Table 1, which has been designed to
correspond to the receptor—pathway—waste characteristics—
waste management practices mode. Standard matrix forms are util-
ized to evaluate the risk associated with specific sites based on
available data. Each evaluation parameter is rated on a numerical
scale, and multiplier factors are applied to obtain the site parameter
score. The final site rating is obtained by adding the products of
all the factors and normalizing the results on a percentage basis.'
In actual fact, most of the information listed in Table 1 is not
readily available during a preliminary assessment. Frequently, in-
vestigators are compelled to rely upon previously gathered data
and the available literature in combination with a preliminary site
Table 1.
Risk Assessment Information Requirements
DIRECT TRANSFER
DIRECT TRANSFER
I. Identification of Facility
1. Owner and Operator
2. Type/Function of Facility
3. Location
II. Receptors
1. Description of Region
a. surrounding facilities
b. surrounding population
2. Land and Water Use
a. regional characteristics
b. surrounding land uses
c. drinking well data
3. Critical Habitats
a. endangered or threatened
species habitats
b. environmentally sensi-
tive areas (e.g., flood-
plains, earthquake
zones)
c. state or federal preserves
4. Off-Site Dose Assessment
a. dose exposure for sur-
rounding population
b. specific demographic and
agricultural data
c. biota monitoring
III. Pathways
1. General Description
a. site location and surround-
ing geography
b. site description
c. off-site discharges
d. potential natural and man-
made pathways
2. Meteorology
a. regional climatology
b. local climatology
c. air monitoring data
3. Geology
a. regional geology
b. site geology
c. soil sampling data and bor-
ing logs
4. Hydrology
a. surface water hydrology
b. groundwater hydrology
c. drainage patterns
d. water quality monitoring
data
IV. Waste Characteristics
1. Site Characteristics
a. quantity of waste
b. condition of waste con-
tainment
c. facility characteristics
d. facility capacity
2. Waste Characteristics
a. mobility
b. toxicity
c. ignitability
d. reactivity
e. corrosivity
f. carcinogenicity
g. volatility
h. radioactivity
i. solubility
j. persistence
V. Waste Management Practices
1. General Description of
Facility
a. location and layout
b. facility description
c. closure procedures
2. Structural and Mechanical
Design
3. Description of Operations
4. Training and Safety Pro-
cedures
5. Pollution Control Method-
ology
6. Monitoring and Record-
keeping Procedures
7. Quality Assurance
8. Government Relations
a. applicable regulations
b. permit applications
c. consent decrees
d. fines
Figure 1.
Major Physical, Biological Transport Pathways of Hazardous Waste
visit to complete the risk assessment. This is particularly true if the
investigator does not have the cooperation and approval of the
facility operator and/or owner.
Standard risk evaluation is a valuable tool for determining prior-
ities among numerous sites, for determining emergency situations
or for evaluating sites for insurance purposes; however, in order
to develop cost-effective remedial action alternatives for long term
cleanup of a site, a detailed analytical program must be under-
taken. The data gathered during the preliminary assessment should
be directed toward determining the optimum location, parameters
and frequency of sampling.
The first stage of the analytical program would involve the util-
ization of preliminary screening procedures such as determination
of pH, conductivity, total organic carbon, heavy metals, methy-
lene blue active substances and phenol. In addition, such tech-
niques as halogenated organic scans, organic nitrogen/phosphorus
scans, volatile halogenated/non-halogenated scans and organic
scans have proven extremely valuable in formulating a general con-
ception of site contamination.7 In addition, where some informa-
tion is available as to previous disposal activities analytical
parameters would be selected based upon:
•Chemicals which may have been deposited at the site in very large
quantities
•Chemicals which may be present in small quantities, but are
known to be extremely toxic even in low doses
•Chemicals commonly found at waste sites which behave in a man-
ner characteristic of many other contaminants'
This analytical program, combined with such tools as mass bal-
ance calculations and general risk assessment, can then be mean-
-------
SITE INVESTIGATION
25
ingfully applied to develop remedial action alternatives. Without
analytical results, the most a risk assessment can achieve is a qual-
itative ranking between sites and a general conception of the na-
ture of potential problems.
TECHNICAL APPROACH
Environmental Background
Selecting an effective remedial technique involves the balancing
of the need to contain contaminants within acceptable levels
against the costs associated with the cleanup measures. The ful-
crum of the balance is risk. The following example is intended to
illustrate the way in which remedial action alternatives can be de-
termined based upon a balancing of risk, environmental setting
and cost.
The particular site is an uncontrolled hazardous industrial land-
fill located in Niagara Falls, New York along the north shore of the
Niagara River.9 The site was utilized from mid-1948 until Sept.
1970 for the disposal of chemical waste by-products Almost the
entire landfill is covered by waste material ranging from 8 to 17 ft
and averaging 12 ft in thickness. Closure was originally accom-
plished by covering the waste with a final soil cover, establishing
vegetation, and constructing a retaining wall along the river.
A detailed preliminary examination of the geology, hydrology,
environmental setting, surrounding community, waste characteris-
tics and past industrial disposal practices was undertaken at this
site. Potential receptors of the contaminants included aquatic life
in the river and associated biota, recreational users and, to some ex-
tent, communities taking treated drinking water from the river.
General information concerning waste characteristics and past dis-
posal practices was supplied by the company involved in the remed-
iation.
Using the Rating Methodology Model, the risks associated with
the site were considered to be high. This determination was based
upon the waste characteristics of the material disposed of at the
site, the previous waste management practices and the hydrogeo-
logic pathways by which contaminants were leaving the site. How-
ever, the overall risk from the site was deemed to be medium when
the dilution factor of the river was included in the assessment be-
cause dilution effectively reduced the hazardous constituents reach-
ing potential receptors.
In order to determine remedial action alternatives, a detailed
hydrogeologic investigation, an analytical sampling program and
mass balance calculations were used to supplement previously
gathered information. These were the key elements to development
of alternatives and eventually remedial recommendations.
Briefly, the hydrogeologic investigation identified three water
bearing zones. First, the groundwater table, or unconfined satur-
ated one was located within the waste itself. Second, a semi-
confined aquifer was located within sandy Recent alluvial deposit.
Third, a confined to semiconfined aquifer existed within the Lock-
port Dolomite.
The analytical program suggested that chemical constituents
attributable to the landfill existed in the former, two water bear-
ing zones and to a lesser extent in the Lockport Dolomite. Down-
ward migration of groundwater from the landfill was restricted
by a stratum of highly impermeable glacial lake deposits (perme-
ability 2.1 x 10 = ' cm/sec). However, the stratum was not con-
tinuous across the site, and varied in thickness from 26 ft to less
than 2 ft. The geology and groundwater flow patterns are repre-
sented in Fig. 2.
The objective of the water balance was to quantify, to the ex-
tent possible, each of the mechanisms of groundwater inflow and
GROUNDWATER TABLE
IN LANDFILL
itUUSiiiiiii-i:^
QLACIOLACUSTRINE CLAY
•Jfc ••:•••• •!.-:-.".'r.T'-..1.; f'-.£?,;•'
^n^^Vv^V'^V'V
£%f^
'\'\'S\'.'^>.'^'S^''^'-'i^i'^':'^ii''?i'''?i''?'''?i'^''*'' ""' -"' -~~- '.^-' ~..U,U,' •
~~
_-^.^
"""/" 7
..
/ /
/ /
/
LOCKPORT DOLOMITE
Figure 2.
Generalized Groundwater Discharge Pathways
-------
26
SITE INVESTIGATION
outflow from the landfill. Further, by incorporating the chemical
character of the leachate into the assessment it became possible to
quantify each of the principal avenues of contamination in terms of
contaminant migration from the landfill. The finished water bal-
ance/contaminant analysis was an invaluable tool in evaluating po-
tential remedial measures in terms of their need and cost effec-
tiveness. The components of the water balance for the Niagara
Falls site are shown in Fig. 3. Calculations were based upon local
climatological and sampling data. The inflow component was
8,871 gal/day based upon inputs of 7,800 gal/day from precip-
itation percolation into the landfill and 1,071 gal/day from lateral
groundwater inflow. The outflow component was calculated to be
9,504 gal/day, with 70% derived from the migration of leachate
into the Recent alluvium aquifer, predominantly in the form of lat-
eral groundwater movement beneath the retaining wall to the
Niagara River. Other pathway components were infiltration to a
storm sewer, and lateral outflow around and through the con-
tainment structure.
Figure 3.
Schematic of Water Balance
Analysis of groundwater samples from piezometers was used in
conjunction with water balance information to calculate off-site
discharge loadings from the landfill. Samples were taken in the
saturated fill zone and the alluvial zone. Bedrock well data, sur-
face water data, and river sediment data were used for conceptual
confirmation of the calculated loadings. These data were then ap-
plied to the water volumes based upon the water balance predic-
tions. Loadings were estimated to be: 50 Ib/yr chlorinated organ-
ics, 6.0 Ib/yr volatile non-halogenated organics, 10 Ib/yr phenol
and 14 Ib/yr toxic heavy metals.
The distribution pattern of halogenated organics suggested two
discharge mechanisms. First, groundwater was migrating beneath
the containment structure and entering the river near the top of the
sediments along the base of the retaining wall. Second, off-site
sources and/or ground water via infiltration from the site was also
being carried into the river by the sewer line. After reaching the
river, the constituents appeared to move outward and downstream
from the landfill.
Remedial Action Alternatives
In order to reduce loadings of chemical constituents to the
Niagara River a corresponding reduction in the inflow component
of the water balance was necessary. Initially four remedial action
alternatives were suggested:
•Excavation
•Total encapsulation with leachate collection and treatment
•Three-sided cutoff wall with an impermeable clay cap
•Impermeable clay cap
Based upon the mass balance calculation, the environmental set-
ting, the risk associated with the site and the cost, alternative
four was selected as the most appropriate remedial action. The cal-
culated water balance indicated approximately 85 % of the inflow
component to be a function of precipitation percolation. An im-
permeable clay cap would reduce this component by at least 90%
and the resulting inflow component to 780 gal/day, combined
with other inflow constituents this equals a total of 1851 gal/day
or 20% of the present situation. Assuming the outflow com-
ponent would be proportionally reduced, the loading to the river
would then be: 10 Ib/yr chlorinated organics, 1.2 Ib/yr volatile
non-halogenated organics, 2.0 Ib/yr phenol and 2.8 Ib/yr heavy
metals.
Examining the environmental setting of the area, it was deter-
mined that the diminished loadings would reduce the risk to low/
acceptable levels. The additional inputs from this landfill site were
considerably lower than that of existing industrial loadings to the
Niagara River. Secondly, current water quality standards would
not be contravened by these inputs. Finally, the risk associated with
the landfill to public health and the environment was considered
minimal. Therefore, the costs associated with excavation and total
encapsulation was deemed to be prohibitive based upon the poten-
tial environmental benefit. A three-sided cutoff wall would assist
in the reduction of the remaining inflow parameter to the water
balance; however, costs combined with potential adverse environ-
mental impacts when the landfill groundwater level reached equilib-
rium with the Niagara River eliminated this alternative from con-
sideration. Consequently, it was recommended that an imperme-
able clay cap be placed over the landfill.
ACKNOWLEDGEMENTS
The author wishes to thank Mr. Robert K. Wyeth, Vice Pres-
ident for Environmental Services of Recra Research, Inc. for his
technical direction and review of this manuscript, and Ms. Mar-
garet J. Farrell for her technical assistance.
REFERENCES
1. Lowrance, W.W.; Of Acceptable Risk, William Kaufman, Inc., Los
Altos, Ca., 1976.
2. Harriss, R.C., Hohenemser, C. and Kates, R.W., "Our Hazardous
Environment," Environment, 20, Sept. 1978.
3. Comprehensive Environmental Response, Compensation and Liability
Act; P.L. 96-510, Dec. 11, 1980.
4. USEPA, National Oil and Hazardous Substances Contingency Plan;
U.S. Federal Register, Part V, 47, No. 137, July 16,1982.
5. JRB Associates, Inc. "How to Rank the Hazard Potential of Waste
Disposal Sites," McLean, Va., Dec. 10, 1979.
6. Barnhart, B.J., "The Disposal of Hazardous Wastes," Environ. Sci.
and Technol., 12, Oct. 1978, 1132-1136.
7. Wyeth, R.K., "The Use of Laboratory Screening Procedures in the
Chemical Evaluation of Uncontrolled Hazardous Waste Sites," Proc.
National Conference on Management of Uncontrolled Hazardous
Waste Sites, Oct. 1981, Washington, D.C., 107-109.
8. Schweitzer, G.E., "Risk Assessment Near Uncontrolled Hazardous
Waste Sites: Role of Monitoring Data," Proc. National Conference
on Management of Uncontrolled Hazardous Waste Sites Oct 1981
Washington, D.C., 238-247.
9. Recra Research, Inc. and Wehran Engineering, "Hydrogeologic In-
vestigation...Landfill, Niagara Falls, New York," Tonawanda New
York, Sept. 1981.
-------
ASSESSMENT OF HAZARDOUS WASTE
MISMANAGEMENT CASES
WAYNE K. TUSA
BRIAN D. GILLEN
Fred C. Hart Associates, Inc.,
New York, New York
INTRODUCTION
USEPA's Office of Solid Waste (OSW) is responsible for
promulgating hazardous waste management facilities regulations
under the Resource Conservation and Recovery Act (RCRA). In
addition to developing the necessary data base to support these
regulations, OSW must also develop Regulatory Impact Analyses
(RIAs), as required by Executive Order 12291, issued Feb. 18, 1981.
In order to achieve these objectives, USEPA identified the need to
develop and compile a data base on damage case histories asso-
ciated with land and non-land based hazardous waste disposal
facilities. Fred C. Hart Associates, under contract to the USEPA,
subsequently completed the study and analysis upon which this
paper is based.
The information generated by this project will provide: (1) a
compilation of damage information on a large number of both ac-
tive and inactive disposal sites meeting certain criteria, (2) informa-
tion of the kinds of environmental damage associated with certain
contamination events, and (3) some measure of the overall extent
of contamination and damage resulting from the mismanagement
of hazardous wastes. The OSW intends to submit the compiled
information for incorporation into the Administrative Record prior
to finalization of the RIA process. The data, as well, will be use-
ful in preparing regulations tailored to specific facility types, eval-
uating alternative regulatory scenarios, and for use as background
information in resolving a variety of other technical issues.
WORK OUTLINE
Fred C. Hart Associates, Inc. completed a number of specific
tasks in accomplishment of the overall project objectives:
•Identification, review and assessment of existing potential sources
of information on damage from hazardous waste disposal sites.
These sources include USEPA's Site Tracking System (STS), the
Superfund Notification System (NOTIS), the Hazardous Waste
Data Management System (HWDMS), the Surface Impoundment
Assessment files (SIA), the Center for Disease Control files, the
Eckhardt Report, Regional USEPA files, and the Regional Field
Investigation Team (FIT) files
•Selection of the FIT and USEPA Regional Office files as having
the data most suitable and readily available for the intended pur-
pose
•Development of site selection criteria to best meet the technical
information requirements of USEPA and to most efficiently util-
ize available contract dollars. These criteria, which were utilized
to select those active and inactive land disposal and non-land
disposal sites that were to be evaluated in this study, included:
-preferential selection of all sites scoring with the MITRE Super-
fund model (The interim list of 175 Mitre Scored sites, "re-
scored" under the direction of USEPA in September-October
1981), and
-preferential selection of any site associated with waste storage.
•Selection of a total of approximately 1,000 sites for preliminary
file review
•Development of a Damage Incident Summary Form (DISF) to
record the data to be collected from the individual site files
•Development of review criteria to insure uniformity of DISF re-
sponses regarding the identification of contamination and damage
events, rating of damage severity, determination of the level of
file documentation required to support given responses, etc.
•Review of the selected damage incident cases at each of the FIT
and USEPA regional offices to complete the DISF survey form
•Analysis of the data findings and preparation of the report en-
titled "Assessment of Hazardous Waste Mismanagement Damage
Case Histories" (Sept. 1982) upon which this paper is based
EVALUATION PROCESS
Case Selection Criteria
The final six criteria developed to select damage case histories
contained in the FIT and regional USEPA files are summarized in
Table 1. Files or sites conforming with these criteria were identified
as the most suitable from the perspective of the project goals. Since
site selection was not a random process, the reader is cautioned
against making generalizations based upon the findings of this spe-
cific analysis.
Files associated with sites for which sampling and analytical data
were available are identified in criteria 1 and 4. These files general-
ly were those sites inspected, investigated and sampled by FIT
and/or USEPA survey teams. FIT files for which sampling data
were not available usually were not sufficiently detailed to support
damage case assessments.
Criteria 2 and 5 identified files associated with hazardous waste
storage, such as tank and container facilities. Criteria 5 was further
refined to include only those sites for which there was preliminary
evidence of damage in order to generate a data base consistent with
project goals and resources.
Criteria 3 and 6 targeted sites identified under the Superfund
program as the 175 highest MITRE scored sites. These sites were
Table 1.
Summary of Case Selection Criteria for Evaluated Sites
for FIT and S&A Files
USEPA Field Investigation Team (FIT) Files
Criteria Number 1 2 3
Criteria Description Files having Files associated Files associated
sampling with waste storage with MITRE scored
data sites (a)
USEPA Survey and Analysis (S&A) Files
Criteria Number 45 6
Criteria Description Files having Files associated Files associated
sampling with waste storage with MITRE scored
data for which there is sites (a)
evidence of damage
(a)The interim list of 175 Mitre Scored sites, "rescored" on September/October 1982 under the
direction of USEPA.
27
-------
28
SITE INVESTIGATION
included in the survey based on the assumption that environmental
damage could potentially be most readily documented at these lo-
cations.
Evaluation Procedures
Files in each region were evaluated by a study team consisting of
a project director, team leader and four to five technical assistants.
Guidelines, definitions and criteria used by the study team in mak-
ing the interpretations and judgments needed to complete the
DISFs are discussed in detail in the project report.
In summary, the evaluation procedure consisted of a two-phase
effort. The first phase consisted of visiting regional FIT and Survey
and Analysis (S & A) offices, screening files according to the selec-
tion criteria, and transferring the appropriate information to the
DISFs. This effort was accomplished during a noncontinuous nine-
week period beginning in Nov. 1981 and ending Feb. 1982. The
second phase consisted of reviewing the completed DISFs for con-
sistency, format and editorial standards, tabulating the conformed
DISFs and summarizing the information in the report. This effort
was accomplished over a period of several weeks beginning in late
Mar. 1982 and ending with the submission of the report.
The DISF was used to assess damage case histories and asso-
ciated site characteristics. After the study team reviewed the file
information, appropriate responses were made on the DISF (Sec-
tions I through XII) and the case was summarized in a brief nar-
rative (Section XIII), which was attached to the DISF form.
I. Site Identification
II. Site Description
III. Date of Incident/Discovery
IV. Status of Operations
V. Exposed Media
VI. Affected Areas
VII. Epidemiological Studies
VIII. Event Causing Incident
IX. Waste Characterization
X. Status of Response
XI. Source of Information
XII. Severity of Damage
XIII. General Comments
Evaluation Criteria
DISF responses for Sections I, II, III, IV, VII, X and XI were
prepared from information available in the files according to the
definitions and instructions accompanying the DISF. DISF re-
sponses for Sections V, VI, VIII, IX and XII required value judg-
ments based on evaluation criteria developed for each value judg-
ment question. For example, the study team was frequently re-
quired to assess whether contamination had occurred, what media
had been exposed, what event caused the incident, and what waste
resulted in contamination. Finally, the study team was required to
assess the severity of damage which had occurred to either human
health and/or the environment. In order to ensure that the study
team rated sites uniformly, specific evaluation criteria were devel-
oped for use as guidance in:
•identifying contamination and damage events,
•rating the severity of damage, and
•determining the file documentation required to support a given
response (i.e., documented versus suspected).
For example, in this study, "contamination" is defined as the
presence of pollutants in groundwater, surface water, soil or air,
as identified by present standard sampling and analytical tech-
niques. "Pollutants" are defined as substances not naturally found
in the site-specific environment which may interfere with the best
use of, or cause environmental harm to, the affected resource.
"Identified" is defined as positive contaminant verification at con-
centrations above the detection limits of the sampling and analyti-
cal techniques applied. Verifiable concentration levels varied, but
in most cases were in the/ig/1 range.
Contamination was considered to be "documented" if the event
was substantiated by a direct investigative action by a regulatory
office or other recognized agency. File information required to sup-
port documentation included:
•Sampling data
•Excerpts from relevant documents (engineering reports, environ-
mental impact statements, NPDES, and RCRA permits, enforce-
ment actions, etc.
•Professional evaluations, expert witness testimony, etc.
"Damage" was defined as the presence of pollutants at con-
centrations causing interference with, loss in quality of or harm to
human health, drinking water, the food chain, flora, fauna or
property. Damage was considered to be documented according to
the same evaluation criteria discussed previously, with certain addi-
tional criteria:
•DISF responses indicating documented damage to human health
were to be based on authoritative references in the file correlat-
ing sickness, injury or death to contamination events occurring
at the site. These references would typically include hospital re-
ports, OSHA citations, regulatory agency reports, facility oper-
ating reports and, in certain limited cases, epidemiological data.
•DISF responses indicating documented damage to drinking water
were to be based on authoritative references in the file correlat-
ing excessive contaminant concentration levels in the water supply
with contamination events occurring at the site. Excessive con-
taminant concentration levels were defined as constituent con-
centrations exceeding USEPA National Interim Primary or
Secondary Drinking Water Standards or USEPA Human Health
Criteria for Maximum Contaminant Levels (MCLs) in water
supplies.
•DISF responses indicating damage to food chain and flora were
to be based on authoritative references correlating visible vege-
tation stress with contamination events occurring at the site.
•DISF responses indicating documented damage to fauna were to
be based on authoritative references, usually bioassay studies,
correlating fish and wildlife damage with contamination events
occurring at the site.
•DISF responses indicating documented property damage were to
be based on authoritative references correlating property damage
with contamination events occurring at the site. These references
would typically include insurance claims, regulatory reports,
OSHA citations and enforcement actions restricting residential
property, drinking water well or other site/facility usages.
Damage was considered to be "suspected" if the responses to
Section VI were based only on citizen allegations, newspaper re-
ports or inconclusive scientific studies.
The study team also rated each site according to the severity of
human health and environmental damage. The broad guidelines
developed by the study team to rate site severity of damage are
given in Table 2. High human health damage ratings were assigned
to sites where incidents resulted in deaths, whereas low damage rat-
ings were associates with minor, short-term injuries. High environ-
mental damage ratings were typically associated with sites corre-
lated with substantial fish or animal kills, and/or groundwater
contamination incidents in which contaminant concentrations ex-
ceeded ten times the drinking water criteria discussed previously.
Low environmental damage ratings were usually associated with
sites where soil or vegetation contamination were limited to rela-
tively restricted areas.
NATIONAL SUMMARY
The study team evaluated and completed IDSFs for a total of 929
sites nationwide. Many of the sites contained multiple facilities.
A total of 1,722 individual facility types were used in describing
the 929 sites in the ten regions. Of the 1,722 facility types eval-
uated, 23% were landfills, 22% were containers, 16% were sur-
face impoundments and 11% were tanks. The remaining 28% of
the facilities were described by various other categories (Fig. 1).
Contamination, either documented or suspected, was identified
in 834 sites, or at 90% of the sites evaluated. At 555 of the sites, or
60%, contamination was documented. Most of the contamination,
(Fig. 2) occurred in groundwater 32%, with the remaining inci-
dents occurring to soil (31%), surface water (29%) and air
-------
SITE INVESTIGATION
29
SITE DESCRIPTIONS BY FACILITY TYPE
TOTAL NUMBER OF FACILITY TYPES TABULATED
1722
CONTAINERS (385)
22%
LANDFILLS (396)^.
23%
SURFACE
IMPOUNDMENT (271)
16%
TANKS (197)
11%
'Sampled sites were not randomly selected. Site selection criteria and the implications of this cri-
teria are discussed in detail in the report.
Figure 1.
Hazardous Waste Sites DISF Summary of Evaluated Sites'
MEDIA CONTAMINATED
TOTAL NUMBER OF INCIDENTS TABULATED
2019
GROUND
WATER (646)
32%
SURFACE
WATER (589)
29%
(626) -INCIDENTS TABULATED
I I -DOCUMENTED INCIDENTS
'Sampled sites were not randomly selected. Site selection criteria and the implications of this cri-
teria are discussed in detail in the report.
Figure 2.
Hazardous Waste Sites DISF Summary of Evaluated Sites'
Media Contaminated
Table 2.
Summary of Guidelines Used in Rating Severity
of Damage at Evaluated Sites
Category
Human Health
High
Severity
Medium
Low
Environmental
ground water,
surface water & air
food chain, flora
fauna
soil
Damage incident to at least one person resulting in...
...death ...severe injury .. .minor injury.
Contamination of groundwater result-
ing in closure or restriction of drinking
water in a...
...community ...single private
water supply. well.
Contamination incident where sampling indicates the presence
of pollutants in concentrations...
...at levels ...at levels ...at detectable
greater than 10 equal to applic- levels, but less than
times applicable able standards. applicable
standards. standards.
Contamination incident resulting in stress to vegetated or
food crop area...
.. .greater than .. .greater than .. .in limited areas
one acre. Vi acre. only.
Damage incident confirmed by...
...massive kills ...limited kills ...bioassay studies
confirming tissue
contamination.
1 ' ...contamination
incident confirmed
by sampling data.
'Higher levels of damage were typically identified via use of evidence in the other categories.
TABULATION OF SITES
CONTAMINATED AND DAMAGED
SITES
1000-
900-
80O-
roo-
eoo-
soo-
4OO-
3OO-
200-
100-
0
SITES INDICATING DOCUMENTED OR SUSPECTED
CONTAMINATION TO AT LEAST ONE MEDIUM
SITES INDICATING DOCUMENTED OR SUSPECTED-
DAMAGE TO AT LEAST ONE AFFECTED AREA
SITES FOR WHICH CONTAMINATION OR DAMAGE
DOCUMENTED
929
834
544
834
565
T= 544
236
TOTAL NUMBER
OF SITES
EVALUATED
CONTAMINATED
SITES
DAMAGED
SITES
'Sampled sites were not randomly selected. Site selection criteria and the implications of this
criteria are discussed in detail in the report.
Figures.
Hazardous Waste Sites DISF Summary of Evaluated Sites'
Tabulation of Sites Contaminated and Damaged
Of the 2,019 responses originally indicating contamination, only
856 (42%) could be documented using the evaluation criteria
previously developed.
Each site was also evaluated for damage occurring to life, prop-
erty and various natural resources. This evaluation focused on six
potentially affected areas, including drinking water, food chain,
flora, fauna, human health and property. Damage (either docu-
mented or suspected), was identified in at least 544 sites, or 59%
of the sites evaluated. The total number of evaluated sites is com-
pared to the total number of sites rated as "contaminated" and/or
"damaged" in Fig. 3. ("Contaminated sites" means sites causing
contamination to at least one medium; "Damaged sites" are those
-------
30
SITE INVESTIGATION
AFFECTED AREAS DAMAGED
DOCUMENTED CASES
EVENTS CAUSING CONTAMINATION
TOTAL NUMBER OF AFFECTED AREAS DAMAGED
375
FOOD CHAW (15)
4%
HUMAN
HEALTH (30)
8%
FLORA (60)
16%
PROPERTY (105)
26%
(129) INCIDENTS TABULATED
'Sampled sites were not randomly selected. Site selection criteria and the implications of this
criteria are discussed in detail in the report.
Figure 4.
Hazardous Waste Sites DISF Summary of Evaluated Sites'
Affected Areas Damaged Documented Cases
sites resulting in damages to one affected area.) The fraction of
contaminated sites and damaged sites associated with the files hav-
ing the appropriate documentation are also shown in Fig. 3. Of the
1,171 affected areas indicating damage, only 375 (32%) could be
documented using the evaluation criteria.
Approximately 34% of the documented damage incidents oc-
curred to drinking water (Fig. 4), with the remaining incidents
occurring to property (28%), flora (16%), fauna (10%), human
health (8%) and food-chain (4%). The subsequent figure (Fig. 5),
indicates that 72% of the incidents causing the damage or con-
tamination described above were due to leachate (32%), leaks
(22%), or spills (17%). These incidents involved contamination
caused by metals, volatile halogenated organics, volatile non-
halogenated organics, acid compounds or base neutral extractables
in 70% of the incidents tabulated.
The test of the report provides additional data along these lines
on a facility by facility type basis, as well as on an USEPA region
by region basis. Specific information is also provided on the
sources of the data, facility type profiles, the events causing con-
tamination, the types and ranges of concentrations of chemicals
recorded in the files, the status of remedial responses indicated in
the files, etc.
CONCLUSIONS
As a result of the data base review, the site history reviews, and
the subsequent data analysis, the study team reached the following
conclusions:
•The FIT and regional USEPA files contain the most readily acces-
sible data base on potential damages from hazardous waste dis-
posal facilities of the data bases examined in this study.
•For those sites that met the selection criteria, the facility types
most commonly identified with potential contamination or
damage included landfills, containers, tanks and open pumps.
•Of the 929 sites evaluated, 41% were identified as active facil-
ities, and 43% as inactive facilities. The remaining 16% could
not be identified using the information available in the files.
TOTAL NUMBER OF EVENTS TABULATED
1671
LEACHATE (546)
33%
LEAKS (36 D*^
22%
SPU.SC292)
17%
•Sampled sites were not randomly selected. Site selection criteria and the implications of this
criteria are discussed in detail in the report.
Figure 5.
Hazardous Waste Sites DISF Summary of Evaluated Sites'
Events Causing Contamination
•Approximately 90% of these sites had evidence of suspected or
documented contamination.
•Approximately 60% of the sites indicating the potential presence
of contamination had documented evidence of contamination.
•Groundwater, surface water and soil were the media that were
contaminated most often and at approximately the same number
of sites.
•The events most often associated with contamination included
leachate, leaks, spills, fire/explosion, toxic gas emissions and
erosion.
•The most commonly identified contaminants included metals,
volatile halogenated organics and volatile non-halogenated
organics.
•Damage was suspected or documented at 59% of the sites eval-
uated, or at 63% of the sites involving contamination; in. total
approximately 25% of the sites evaluated have documented evi-
dence of damage to human health or the environment.
•Suspected damage was most often reported to drinking water,
human health, fauna and flora; documented damage was most
often reported to drinking water and property.
•While remedial programs varied on a case by case basis, various
legal actions and/or remedial activities have been initiated at a
significant number of sites evaluated in this study. For example,
legal or enforcement activities have occurred at 19% of the sites,
while 55% of the sites have had or are currently completing addi-
tional environmental investigations. At approximately 30% of the
sites, remedial activities of some type have been initiated.
•It is particularly important for the reader, throughout this
analysis, to be cognizant of the fact that the 929 sites evaluated
were selected based on specific criteria. This criteria included pre-
selection of active and inactive hazardous waste disposal sites
associated with cases having sampling data, cases associated with
waste storage and MITRE scored sites. In view of this preselec-
tion process, it should be noted that these cases are not neces-
sarily representative of all damage cases on file at USEPA or in
actual existence. As a consequence, it difficult to apply the find-
ings of this analysis to other data bases on hazardous waste facil-
ities, abandoned sites, etc., or to develop more generalized con-
clusions based upon the efforts of this study.
-------
ELECTRICAL RESISTIVITY TECHNIQUES FOR
LOCATING LINER LEAKS
WENDELL R. PETERS
DAVID W. SHULTZ
BOB M. DUFF
Department of Geosciences
Southwest Research Institute
San Antonio, Texas
INTRODUCTION
An important aspect of hazardous waste treatment and disposal
in landfills or surface impoundments is the prevention of surface
and groundwater contamination by fluids containing hazardous
constituents. Relatively impervious flexible membrane liners have
been used to establish the facility boundaries and to prevent fluids
from migrating into the surrounding water resources.
Research and evaluation projects are underway to investigate the
effectiveness of waste containment liner materials, including the
long-term deterioration of liners exposed to various waste pro-
ducts. Early landfill liners consisted of clay hardpans, and wood or
metal barriers. Improved liners have been developed more recently
using flexible polymeric materials to provide lower liquid
permeability and longer waste containment lifetimes.
While these newer liner materials are demonstrating improved
resistance to prolonged waste exposure, no methods are presently
available to nondestructively test their physical integrity in service
environments. A tear or rupture in a liner system will allow fluid to
migrate form the facility and thus violate the original intent of the
liner. The work reported herein has been directed toward develop-
ing a nondestructive electrical system which can be used for detec-
ting and locating leaks in waste containment liners while the facility
remains in service.
Since the philosophy of impervious liners is to contain rather
than absorb or filter contaminants, the physical characteristics of
the liner materials will usually differ significantly from the underly-
ing soil and the contained hazardous waste. In particular, liners
made of impervious plastics and rubbers will exhibit very high elec-
trical resistance which will act as an electrical insulator between the
internal and external liner surfaces. If the liner is physically punc-
tured or separated so that fluid passes through the liner, the elec-
trical conductivity of the fluid and saturated underlying soil will
form a detectable electric current path through the liner by which
the leak may be revealed. Electrical methods are attractive because
they can be applied on the surface of the landfill or fluid impound-
ment. The methods are completely nondestructive in operation
and, by means of automated field equipment, can be made very ef-
ficient and cost-effective for facility monitoring operations. The
resulting method, when fully developed, will be useful for survey-
ing existing as well as new facilities where electrically resistive mem-
brane liners are installed.
Two- and three-dimensional computer models were used to ex-
amine the current distribution in the cross sections of several
simulated liner geometries having different leak locations. In addi-
tion to these analytical model studies, a three-dimensional physical
scale model was also constructed using a plastic liner in a wooden
frame. Soil was placed inside the model to simulate a landfill.
Water was used to simulate a fluid impoundment. Electrical poten-
tial distributions on the surface of the soil and water were measured
to determine the effects of punctures in the liner. The effects of
multiple leaks as well as those caused by subsurface anomalies con-
tained within the liner were studied. Based upon these application
concept studies, the most promising electrical testing methods have
been identified and subsequent project activities are now directed
toward the design and fabrication of an appropriate experimental
field instrumentation system. Computer software is also being
developed and tested to allow on-site analysis of field data.
EXPERIMENTAL METHOD
The basic electrical testing concept where one current electrode is
located away from the facility and a current path exists over
through a leak penetration in the liner as well as over the buried
edge of the liner is shown in Fig. 1. This figure shows that when a
leak penetration is present in the liner, current flow between elec-
trodes located inside and outside the facility will follow two paths,
namely, through the leak and over the buried edges of the liner.
Figure 1.
Leaking Liner. Arbitrary Search Electrode Position
Since surface potentials are directly related to the current distribu-
tions in the vicinity of the search electrode, they can be used to
locate the leak.
Fig. 1 represents an idealized case of a non-conducting liner. Ac-
tual liner materials have high but finite volume resistivity. The
voltage to current ratio at the source electrode can be expressed as a
resistance due to the earth in the absence of the liner in series with a
total resistance across the liner. If the current density is assumed to
be uniform over the entire area of the liner, then this liner
resistance may be expressed as:
RL =
where:
RL = The total liner resistance
PL = the volume resistivity of the liner material
t = thickness of the liner
A = total surface area of the liner
(1)
31
-------
32
SITE INVESTIGATION
For a liner 0.030 inches (30 mil) thick and five acres in area, and a
material with a volume resistivity of 2 x 10'° ohm-meter, Equation
(1) gives a total liner resistance of 750 ohms. Thus, even though the
resistivity of the liner material is high, the total series resistance due
to the liner can be relatively small for a large facility. While Equa-
tion (1) is useful for estimates of liner resistance, it is inaccurate
since the current density is very non-uniform over the liner surface
in an actual facility.
An accurate analytical solution for potential and current flow is
not possible for a realistic facility geometry and numerical methods
must be used. An examination of the nature of solutions is,
however, very useful in the development of the leak detection
technique. The surface potential within the facility can be expressed
in terms of the current density across the liner as:
2ir
2r
dn R
(2)
where:
r = the distance from the current injection electrode (inside the
facility) to the point on the surface at which the potential
is measured
V(r) = the potential at point r on the surface
I = the total injected current
r = the distance from the current injection electrode to the point
of integration on the liner
R = the distance between the potential measuring point r and
the integration point r'
Jn(r') = the component of current density perpendicular to the
surface of integration, i.e., liner
SL = the surface defined by the liner
SL = the surface SL with leak areas removed
PI = the volume resistivity of material within the facility, i.e.,
liquid or solid waste
p2 = the volume resistivity of the earth surrounding the facility
PL = the volume resistivity of the liner material
t = the thickness of the liner
n = a unit vector normal to the surface SL
The current density Jn which appears in the integrands of Equation
(2) is unknown and is dependent upon the geometry of the facility,
the resistivities of the materials inside and outside of the liner, and
the position of the current injection electrode. Examination of
Equation (2) reveals two important features of the electrical
method. First, the influence of the liner on the surface potential
decreases with distance from the measurement point if the same
current density is assumed. Second, the influence of any portion of
the liner on the surface potential is proportional to the current den-
sity crossing the liner at that point.
The current density crossing the surface SL is dependent upon the
effective resistance across that portion of SL. The effective
resistance across the surface, now assumed to be of thickness t,
may be computed from Equation (1) if the area A is taken suffi-
ciently small for the current density to be essentially constant. Con-
sidering two small areas, one containing liner material of resistivity
PL, the other a leak for which the resistivity is P\, then, from Equa-
tion (I), the ratio of resistances is the ratio of the resistivities,
PL/PL- The resistivities of typical liner materials are in the range 1 x
10* ohm-meters to 1 x 10M ohm-meters, while the resistivity of
material within a facility will generally be 1 to 10 ohm-meters or
less. It may. therefore, be concluded that in the vicinity of a leak,
the current density crossing the liner will be many orders of
magnitude less than the current density at the leak.
The surface potential at points above a leak will be significantly
influenced by the shunting effect of the leak on the current density
across the liner (Eq. 2). To obtain more precise information a
numerical solution of the problem is required.
The first numerical model study used a general purpose circuit
simulation computer program (SPICE). Circuits containing
resistors, capacitors, inductors, and voltage and current sources
Figure 2.
Computer-Modelled Vertical Cross Section of
Waste Liner with No Leaks
Figure 3.
The Waste Liner Physical Scale Model and
Instrumentation Van
may be simulated with the program. A two-dimensional resistor
network model designed to simulate a membrane liner is modelled
using the program. The resistivity of the liner fill and the surroun-
ding earth is modelled using normalized resistance values of one
ohm. The insulating effect of the liner is represented in the model
by using parallel 10,000-ohm resistors along the path of the liner. A
leak (conductive path) in the liner is simulated by replacing one of
the 10,000-ohm resistors in the liner with a 1-ohm resistor. The
shunting effect of the surrounding soil across the liner edges is
simulated by using two-ohm resistors at the surface. Current and
voltage sources are used to inject current into the surface of the
model. t
The output of the model analysis is the voltage at each of the
nodes connecting the sensor elements. These values are stored in in
array which is then used in a second program which plots the results
in the form of equipotential contours. A contour plot showing the
two-dimensional potential distribution around a liner without leaks
is shown in Fig. 2. The network used to model this cross section is a
rectilinear array of 21 by 11 resistors for a total of 494 elements.
The current injection point is at the top center of the figure with the
other reference electrode connected to a conducting path along the
bottom and sides of the cross section. This reference is far enough
from the liner to be located at effective infinity. The outline of the
liner has been sketched in the figure and is represented by the three-
sided trapezoidal figure in the center of the plots. The equipotential
lines showing the voltage distribution patterns were computer
generated. The current flow paths are at right angles to the
equipotential lines and were sketched in by hand. The current flow
-------
SITE INVESTIGATION
33
over the edges of the liner and into the surrounding earth is il-
lustrated in Fig. 2.
In parallel with the computer analysis work described above, a
three-dimensional physical scale model was constructed in an out-
door environment where ground is used as the soil underlying the
physical scale model. Fig. 3 is a photograph of the model and the
instrumentation van. The 11 ft2 framework is lined with a 6-mil
polyethylene sheet. Shunt resistors on the edges couple the interior
of the liner to the surrounding earth. To facilitate voltage
measurements, most of the work to date has been done with water
in the liner. Current is injected into the liner using a single electrode
with the return electrode located approximately 300 ft away from
the model. Various liner fill depths could be easily modelled by
varying the depth of the water.
Surface voltages were measured with both potential measure-
ment electrodes located in the basin. The potential reference elec-
trode was fixed in position near one corner of the model. The ex-
ploration measurement electrode was composed of many electrodes
mounted on a fiberglass beam at 2.5 in. spacings. Initially, data
were taken on a rectangular grid but it was determined that better
quality data could be acquired using a polar grid with the current
injection electrode at the center.
Initially, a constant current source was used in the experiments.
However, this device did not have the range of current needed for
the measurements. Subsequently, a laboratory oscillator was used
to supply a 25-Hz signal which was amplified by an audio power
amplifier. The output of the amplifier was transformer-coupled to
the current injection electrodes. A digital ammeter was used to
monitor the 50-ma injection current.
The leaks in the bottom of the liner were generated by driving a
0.5 in. diameter copper-clad steel rod through the liner and into the
soil. This rod provided a good conducting path between the water
and the soil without loss of water.
In Fig. 4, a close-up of the fiberglass beam with the measurement
and current injection electrodes in place is shown. Wave action
around the current injection electrodes was found to cause pro-
blems in maintaining a stable constant current. For this reason, a
PVC still-well was placed around the current injection electrode as
seen in the photograph. Monofilament nylon lines were used to ac-
curately locate the current injection electrode.
EXPERIMENTAL RESULTS
In developing and employing the technique, it is important to
know the electrical properties of the liner materials being used in
landfills. Since the electrical parameters of liner materials are not
provided by their manufacturers, a laboratory electrical test of
samples of liner materials was performed. A sample of the
measurements made on the materials is shown in Table 1. The
resistivity values were calculated using measured values of
resistance.
Table 1.
Electrical Properties of Landfill Liner Materials*
Sample Type
High Density
Polyethylene (A)
High Density
Polyethylene (B)
Polyethylene
Chlorosulfonated
Polyethylene (A)
Chlorosulfonated
Polyethylene (B)
Chlorosulfonated
Polyethylene - nylon
reinforced
Polyvinyl Chloride
Polyvinyl Chloride - oil
resistant
Chlorinated
Polyethylene
Chlorinated
Polyethylene-
reinforced
Urethane Asphalt
Volume
Thickness Resistance Resistivity
(mils) (ohms) (ohm-cm)
55.0 4.00 x 10" 2.86 x 10"
121.0 3.25 x 10" 4.81 x 10"
6.0 1.80x10" 1.18x10"
40.0 3.20x10' 3.14x10"
36.0 3.75x10'" 4.10x10"
34.0 3.60x10' 4.16x10"
29.0 1.52x10' 2.06x10"
30.0 1.70x10' 2.23x10"
32.5 7.20 x 10'° 7.84 x 10"
36.0 6.70x10' 7.32x10"
69.0 2.20x10" 1.25x10"
•Area of samples = 100 cm1 except High Density Polyethylene 45.5 cm1 and Chlorinated
Polyethylene (9.0 cm')
Figure 4.
Close-up of the Current Injection Electrode and the
Still-Well Used to Dampen Wave Effects
For one placement of the beam, data were acquired by making
measurements at increasing radial distances away from the current
injection electrode. The beam was then rotated by a specified angle
about the current electrode position and another series of radial
readings taken. Proceeding in this manner, measurements covering
the entire model surface were obtained.
The results of one of the computer generated two-dimensional
analyses of a liner having a leak in the bottom are given in Fig. 5.
The equipotential lines which appear in the figure were generated
and plotted automatically by the computer program described
earlier.
This result shows that, as predicted, when a leak penetration is
present in the liner, current flow between electrodes located inside
and outside the liner will follow two paths, namely, through the
leak and around the buried edges of the liner. Note the non-
symmetry of the equipotential lines terminating on the surface of
Figure 5.
Computer Modelled Vertical Cross Section of a Waste Liner
with a Leak in the Bottom
-------
34
SITE INVESTIGATION
the landfill above the liner. The voltage gradient is clearly steeper
along the surface on the side of the current injection electrode
above the leak. Asymmetries such as this were investigated in subse-
quent work to develop field measurements which could locate liner
leaks.
The relatively inexpensive nature of the two-dimensional com-
puter model and its precise control make it a very productive
method of investigating electrical testing concepts related to liner
leaks. Even though it is a two-dimensional model that does not ac-
curately simulate the three-dimensional volume of landfill or sur-
face impoundment, the ability to accurately predict the potential
distributions in various simulations of liner conditions makes it a
valuable tool. The sensitivity of this model to a simulated leak is
greater than for a true landfill condition because of the two-
dimensional character of the model. Nevertheless, the results ob-
tained are, in general, indicative of the full-scale equipment sen-
sitivity requirements.
Examples of equipotential contours measured using the physical
scale model are shown in Figs. 6 and 7. For both cases, a single leak
was located at an arbitrary position in the bottom. For all plots, the
current injection point is located at the center of the polar coor-
dinate system.
The distortion of the equipotential lines in the neighborhood of
the leak which is on the 340° radial is clearly shown in Fig. 6. A
steep voltage gradient in the radial direction crossing over the leak
distorts the equipotential lines.
A similar plot for a different current injection point is shown in
Fig. 7. Again, the current injection point is at the center of the
polar coordinate plot. An aluminum block was also placed between
the current injection point and the leak. The same pattern of per-
turbed equipotential lines as observed in Fig. 6 can be seen in Fig.
7 about the leak. No observable distortion was generated by the in-
troduction of the conducting block. Many data acquisition runs
120P i
-------
SITE INVESTIGATION 35
results illustrated in Fig. 7. The X(I) column is the distance in in-
ches from the current injection point. The corresponding voltage
measurements (millivolts) are given in the column labeled Y(I). A
logarithmic curve fit was made to these data with the general equa-
tion Y = a + b In (X), resulting in the empirical relationship
Y = 8374-1897 In X. (3)
A plot of this curve and the measured data are shown in Fig. 8. The
effect of the leak is discernable at a radial distance of 72.5 in.
In the third column of Table 2, labeled Y(I), are the calculated
values of voltage based on the fitted curve. The data in the column
labeled RESIDUALS are calculated by taking the difference bet-
ween Y(I) and (YI). At a radial distance of 72.5 in., a very
noticeable perturbation caused by the leak occurs in the residuals
column where it reaches a minimum value of - 156.91. This data
reduction technique will be used in the equipment design and data
processing to extract the leak generated signatures from the field
data.
Using the residual analysis described above on more than one
radial line and plotting the analyzed results along the radials gives a
contour plot as shown in Fig. 9. Here the physical scale model was
configured with two leaks located on radials which were 40° apart.
As observed in this figure, the region near the two leaks is clearly
discernable from the closed contours which surround the two leak
positions. The center of the contour on the left was above one of
the leaks and the contour on the right missed the leak by about 5 in.
In both cases, the leak targets were in the correct azimuth direction
with only small errors in the radial distance. Additional measured
values along other radials crossing the first will result in a more
precise determination of the leak location.
FUTURE WORK
The analytical and laboratory model studies of electrical liner
testing completed to date have revealed some promising approaches
for leak detection. Based on these results, specifications for field
instrumentation are being established. Sufficient model testing has
been completed to permit full specification of equipment operating
characteristics.field requirements, data recording and data process-
ing software.
To supplement the data base obtained to date using the water-
filled physical scale model, similar tests are in progress using a soil-
filled model. The inhomogeneities associated with this type of fill
will better approximate the conditions anticipated at full-scale land-
fill sites.
At the conclusion of the work in progress, an experimental field
system will be assembled for performance demonstration testing
and verification of data acquisition precision. The system operating
CURRENT INJECTION
7000
6000
5000
4000
3000
2000
1000
0
\
••*_.
i i i i i i i i
20
40
INCHES
60
80
Figure 8.
Logarithmic Curve Fit for the Data Taken on a
Radial Crossing Over the Leak
Figure 9.
Polar Plot of the Residuals Calculated by Taking the Difference
Between the Field Data and Logarithmic Fitted Curves. The Leaks
Shown Are on Radials Separated by 40 °
functions will be implemented using a desktop computer and a
digital data recording system. The system will be semi-automatic in
that several electrodes will be laid out in advance and automatically
scanned and read by the computer. Software written for the com-
puter will be used to process the data in the field to generate plots
similar to those obtained in the model studies.
After completion of the system assembly, a field test plan will be
developed based upon information describing existing landfill sites
and surface impoundments utilizing polymeric membrane liners.
Candidate sites will include liners with and without leaks if possi-
ble.
CONCLUSIONS
Computer modelling of electrical techniques of leak detection in
liners was successful in proving the fundamental concepts needed
to develop a practical system for finding liner leaks in hazardous
waste facilities. Two-dimensional analysis provided valuable in-
sights into current flow patterns and equipotential patterns
associated with liner geometry and leaks.
The computer work also provided the foundation for the
physical scale model studies which were used to generate scaled
measurements simulating ideal facility liners with leaks. Equipoten-
tial plots of the voltages measured on the surface of the model
showed leak-dependent patterns which may be used to locate the
leaks. The model studies showed that multiple leaks may also be
resolved and that anomalies such as blocks of conductive material
buried near the search area do not have a serious effect on the field
data. Parametric studies of surface effects versus depth showed
that the surface perturbations are reduced for greater fill depths.
However, measurable potential patterns in the leak regions still
generate useful information which can be used to locate the leaks.
The results of this investigation are very encouraging in regard to
the feasibility and usefulness of a practical leak detection and loca-
tion system. This technique appears to show promise for detecting
and locating membrane liner leak paths for new as well as existing
landfills and surface impoundments. Future field tests will help
verify the system capabilities and the conditions where the most
reliable data are produced.
ACKNOWLEDGEMENTS
The research which is reported in this paper is being performed
under Contract No. 68-03-3033 "Investigation of Electrical Tech-
niques for Leak Detection in Landfill Liners" with the USEPA,
Municipal Environmental Research Laboratory, Cincinnati, Ohio!
The support and guidance of Mr. Carlton Wiles, Project Officer, is
gratefully acknowledged.
-------
EVALUATION AND USE OF A PORTABLE GAS
CHROMATOGRAPH FOR MONITORING HAZARDOUS WASTE
SITES
JAY M. QUIMBY
ROBERT W. CIBULSKIS
U.S. Environmental Protection Agency, Environmental Response Team
Edison, New Jersey
MICHAEL GRUENFELD
U.S. Environmental Protection Agency, Municipal Environmental Research Laboratory
Oil and Hazardous Materials Spills Branch
Edison, New Jersey
INTRODUCTION
The imminent threat posed by the continuing discovery of aban-
doned hazardous waste dumpsites requires the development of rap-
id and effective on-site chemical analysis techniques to assess the
extent of environmental contamination and monitor ensuing clean-
up efforts. The USEPA recently evaluated the practicality of using
a light weight shoulder-borne gas chromatograph (GC) to monitor
potentially hazardous atmospheres at chemical spills and waste
chemical dumpsites. The instrument evaluated was the Model 128
Century Organic Vapor Analyzer (OVA) equipped with a flame
ionization detector (FID), which is manufactured by the Foxboro
Corporation, Burlington, Mass.1'2'3'4'5 The technical evaluation of
this portable unit was a joint venture on the part of USEPA's En-
vironmental Response Team (ERT) and the Oil and Hazardous
Materials Spills (OHMS) Branch of the Municipal Environmental
Research Laboratory.
In this paper, the authors address several important aspects of
their evaluation of the OVA-GC including: instrument operating
performance parameters (i.e., detection limits, column efficiency),
QA/QC considerations (i.e., accuracy, reproducibility, linear dy-
namic range), associated operational difficulties and recommended
field uses. Analytical methods and specialized techniques devel-
oped in the OHMS Branch for the specific use with the OVA-GC
were essential to this instrument's evaluation and are contained
herein.
The ERT has employed this portable unit during numerous field
activations involving hazardous waste storage sites and chemically
contaminated lagoons. Successful field applications of the OVA-
GC has also involved support to mobile laboratory activations in-
volved with monitoring the effectiveness of waste treatment sys-
tems for removing organic contaminants from surface waters,
leachates and sediments.'
DESCRIPTION OF UNIT EVALUATED
Several important features of the Model 128 OVA-GC include its
ability to: (1) continuously measure the total level of FID detec-
table organic vapors in ambient air, (2) screen high levels of organ-
ic vapor concentrations in environmental samples to protect sensi-
tive analytical instrumentation (i.e., GC/MS) from detector con-
tamination, and (3) generate integrated sample analyses from
chromatograms of organic vapors in air mixtures.
The OVA-GC offers both manual and automatic injection cap-
abilities. The manual injection mode requires direct, gas-tight
syringe injection through a septum, while automatic injections
make use of a positive displacement injection valve incorporated
within the instrument. An instrument backflush valve functions to
flush from the GC column any high molecular weight organics that
may be trapped after a sample run. The column is externally
mounted, thus being exposed to ambient temperature conditions
during normal use. An accessory for maintaining constant column
temperature is available from the manufacturer of the OVA unit.
An instrument output meter serves to indicate, in ppm units, the
concentration of total organic vapors in contaminated air mixtures.
In the GC mode of operation, a strip chart recorder supplies a
chromatogram of the detector output signal.
INSTRUMENT START-UP
The initial start-up of the OVA-GC requires only 15 min, and
this operation includes recharging of the hydrogen supply tank.
Hydrogen functions as a fuel for detector combustion and serves as
the carrier gas for the GC mode of operation. The instrument is
operable in either a portable mode, drawing from a built in bat-
tery power source (DC voltage), or in a stationary mode, draw-
ing from a 120V, AC power supply. The instrument is also
equipped with a battery charging unit which requires a 24 hr period
to fully charge the internal battery pack. A complete battery charge
will allow 8 hr of OVA operating time.
Since the detector flame has been ignited, time must be allowed
for the output meter attachment and the recorder to stabilize.
Usually this can be achieved within one hour, although several
days may be required depending on the frequency of use and the
degree of instrument and/or detector contamination. A stable
baseline in an atmosphere free from organic solvents and com-
bustion processes (i.e., automobile exhaust) will read between 3
and 4 ppm on the meter attachment.
TEST ATMOSPHERES
Test atmospheres are used to generate data on chromatographic
resolution, retention times, linear dynamic ranges, and detection
limits of compounds for which the OVA-GC will be used to an-
alyze. There are many techniques by which these test atmospheres
can be obtained.7 Comparisons between a variety of techniques
(i.e., dynamic systems vs. static systems) were made in order to
judge the most suitable methods for use with the OVA. These
comparisons were based on: accuracy and precision, time, cost,
complexity, portability, and space requirements.
Vial Dilution
Test atmospheres can be obtained by using methods of prepa-
ration in the field or laboratory. During the course of this instru-
ment's evaluation, the OHMS Branch developed a method whereby
known concentrations of materials in the gaseous phase could be
accurately prepared. This method is that of vial dilution.
Initially, a gaseous stock vapor solution must be prepared. This
is done by dispensing a small volume (i.e., 5-10 pi) of an organic
liquid contaminant into an empty septum sealed 43 ml glass vial.
A small volume of liquid is used to insure that the contaminant
will fully vaporize in the vial's air space. Known parameters of
saturation coefficients, vial volume, specific gravity of material,
volume of dispensed material, temperature and pressure are used
to accurately determine the concentration of material in the vial in
»ig/mj or ppm.1 A dilution sequence may then be prepared by
36
-------
extracting known volumes of vapor via gas tight syringe from the
stock vial's air space. This method is extremely rapid, inexpen-
sive, easy, portable, accurate, and reproducible.
Premixed Canisters
Test atmospheres can be purchased in commercially available,
multi-component canisters at known and accurate concentrations.
A vendor will prepare desired mixtures of volatile organic com-
ponents at prescribed concentrations. Simple mixtures of volatile
organics can provide multiple internal standards for field use if the
individual components contained in the canister have been prev-
iously analyzed to insure concentration accuracy. A syringe adap-
ter can be fitted to the canister to allow withdrawal of contam-
inants in microliter quantities (via syringe) thus avoiding personnel
exposure to hazardous components. This method offers high
accuracy and precision, is simple, rapid, and portable but is the
most expensive of the three techniques discussed.
Headspace Analysis
Headspace analysis was a technique introduced to the OHMS
Branch by Dr. Thomas Spittler (USEPA Region I).2 Analytical-
ly, head space analysis is used to indirectly determine the level of
water contamination due to volatile organics by directly determin-
ing the concentration of vapors in the air space above a contam-
inated water sample.9
The headspace technique can be utilized for the preparation of
test atmospheres. This is accomplished by first preparing a sat-
urated solution consisting of pure water and one known liquid con-
taminant at equal volumes. This solution can then be stored in a
40 ml crimp sealed glass bottle. A series of saturated water solu-
tions, each containing a different organic compound, is prepared in
the same manner.
A dilution of the saturated water solution(s) is made by with-
drawing a known volume of its water phase via microliter syringe
and injecting the syringe contents into a second known volume of
pure water. The contaminant(s) then imparts(s) a gas phase/liquid
distribution coefficient specific to each contaminant diluted into
the second vial. The air space of the diluted vial is then analyzed
by the OVA-GC. Quantification indirectly related the concentra-
tion of the contaminant in air to its concentration in water.
The accuracy of this method is dependent upon: (1) the accur-
acy of the published saturation coefficients, (2) the ability of each
material in a multi-component mixture to exert its own vapor pres-
sure and not effect the other materials present (i.e., Henry's
Law), and (3) the human accuracy of dispensing accurate volumes
of vapor and liquid via syringe.
The technique of headspace analysis can also be utilized when
monitoring soil samples for suspected contamination. However,
difficulties in preparing standards of known accuracy and precis-
ion in a soil media exist, thus the method results in less than accur-
ate quantitative measures of volatile organics in the soil.
CALIBRATION
The OVA-GC must be calibrated prior to each field usage to
insure the validity and precision of obtained results. Calibration is
accomplished by measuring at least three accurately known concen-
trations of methane in air across the GC attenuation settings (i.e.,
IX, 10X, 100X). These concentrations of methane can be ob-
tained in small (i.e., hand held) commercially available, pressur-
ized canisters. A gas sampling bag (i.e., a Mylar or Tedlar bag) is
then attached to the OVA sampling probe using an inert flexible
tubing. The standard mixture is drawn from the bag and a direct
continuous reading (in ppm) is obtained from the output meter.
Adjustment of the OVA's potentiometers calibrates the instru-
ment to the desired output meter reading.
Internal standards can be used to identify day to day changes
in detection limits and retention times when the OVA-GC is used
in the field. Any of the three methods mentioned under the Test
Atmospheres heading on page 4 may be used. The most practical
method is that of vial dilution.
SAMPLING AND MONITORING 37
INSTRUMENT PERFORMANCE—GC MODE
Reproducibility
Reproducibility of GC peak heights and retention times was in-
vestigated for both the syringe and the automatic injection valve
modes of the VOA. Difficulties were encountered when the instru-
ment was operated in the automatic injection valve mode, with
nori-reproducible peak retention times (i.e., CV*= 6-12%) re-
sulting.8 These unacceptable retention time (RT) results were traced
to an instrument malfunction. Syringe injection, however, yielded
acceptable peak retention time reproducibility (i.e., CV = l-3%)
when using the same "malfunctioning" instrument. Following in-
strument repair by the manufacturer, both automatic valve and
direct syringe injection yielded acceptable reproducibility (i.e.,
CV = l-4%). These findings suggest that: 1) the automatic injec-
tion valve mode should be used with caution when attempting
component identification, and 2) adequate GC peak retention time
reproducibility should be established prior to operation in either
sampling mode.'
Early results in this evaluation revealed that the reproducibility
of peak heights is affected by whether the backflush valve is in the
up or the down position. Investigation of the significance of the
backflush value position indicated that a backflush valve in the up
position correlates with the largest peak height values. Thus, in this
evaluation, the backflush valve was fixed in the up position for all
sample runs. Verification of GC peak height reproducibility for
quantification purposes using both sample injection modes yeilded
equivalent results (i.e., CV less than 5%).
The stability of relative retention times (RRTs) and absolute re-
tention times (ARTs) when subject to moderate temperature
change was also evaluated. A comparison of results revealed that,
over a 30 ° C temperature range, the reproducibility of RRTs for
several test compounds was in the range of 5-10% CV, while CV
values for ARTs of the same test compounds were in excess of
50%.'
Accuracy
Determinations on the accuracy of the linear dynamic Range
(i.e., concentration vs. response) of the OVA detector was per-
formed using benzene and carbon tetrachloride.10 Benzene gave the
largest detector response of the hazardous compounds studied in
terms of peak heights per concentration unit. Carbon tetrachloride
yielded the smallest peak height per concentration unit (Table 1).
The linear calibration range (0.7 ppm-165 ppm) and the asso-
ciated calibration errors determined for benzene are shown in
Table 2. The highest calibration solution concentration used was
165 ppm which is sufficient for ambient air analyses. The largest
calibration error for benzene was found to be 25% at the detec-
tion limit and a calibration error of 14% was determined at the
highest concentration value. Calibration error values between these
limits did not exceed 9%.
Table 1.
OVA Gas Chromatographic Analysis Parameters for Test Compounds
Benzene and Carbon Tetrachloride.
Analysis Parameters Benzene Carbon Tetrachloride
Ambient Temperature 25 °C 28 °C
Retention Time 1.77min 1.53min
Column G-24" * G-24" *
Injection Mode Syringe Syringe
Injection Volume 250 jil 250 jil
Injection Valve Up Up
Backflush Valve Up Up
Attenuation 0-3 ppm @ IX 0-67 ppm @ IX
73-166 ppm @ 10X 100-548 ppm @ 1 OX
Flow Rate Factory Set @ 2 (1/min) Factory Set @ 2 (1/min)
•The packing material for a G-24" column is 10% SP-2100 on 60/80 mesh supelcoport.
*ACV (coefficient of variation) of 5% or less is considered acceptable by
the OHMS Branch for the purpose of validating reproducible data.
-------
38
SAMPLING AND MONITORING
Table 2.
Linear Calibration Range and Detection
Limit Established for Benzene
Detection Limit
Linear Calibration
Range
Calibration Solution
Concentrations (ppm) 0.74
Mean Peak Height (mm) 3.0
%C.V. 0.0
%Calibration Error 25
0.70 ppm
0.74 ppm-164.50 ppm
1.47
5.0
0.0
5
2.94
9.8
3.3
73.57 164.5
22.5 45.8
4.2 0.5
9 14
Table 3.
Linear Calibration Range and Detection Limit
Established for Carbon Tetrachloride
Detection Limit 8.0 ppm
Linear Calibration
Range 8.0 ppm-547.5 ppm
Calibration Solution
Concentrations (ppm) 8.0 39.72 67.11 273.90 547.80
Mean Peak Height (mm) 5.83 22.83 42.75 166.75 308.13
%C.V. 4.9 2.5 1.8 1.7 1.4
^Calibration Error 19 7 — 3 10
Linear calibration range (8 ppm-547 ppm) and the associated
calibration errors for carbon tetrachloride are given in Table 3.
An upper concentration limit of 547 ppm is more than sufficient
for ambient air analysis. The largest calibration error was 19%
at the detection limit and a calibration error of 10% was deter-
mined at the highest concentration value. Calibration error values
. between these limits did not exceed 7 %.
Detection Limit
The detection limit of the OVA-GC is a function of the test com-
pounds detector response relative to that of the calibrant gas, which
in this case is methane.When analyzing for a compound other than
methane, the resulting detector response will be either higher or
lower than that response generated by an equal concentration of
methane.
Detection limits of many hazardous volatile organics have been
determined for the OVA, and were found to range from 700 ppb
for benzene to 8 ppm for carbon tetrachloride (Tables 1 and 2).
Unbranched alkanes were found to yield an even more sensitive
detector response than did benzene.
Column Efficiency: Ambient Conditions
The types of columns used in the evaluation of the OVA were
the G-24" (Table 1), T-12", G-8", and T-8". The shorter col-
umns (i.e., 8 in) were incorporated with a temperature control pack
assembly that is available from the manufacturer. The G-24" col-
umn had the widest range of application based on the materials
studied in this evaluation. Column selection would be based on the
desired application of the OVA unit, in terms of compound resolu-
tion and retention times (RTs).
The manufacturer of the OVA-GC does not test column per-
formance by using quantifiable techniques, such as theoretical
plates and resolution data. Therefore data accumulated on reten-
tion times, detection limits, and chromatogram resolution could
differ for columns having equivalent parameters, and also from
data published by the manufacturer.
When analyzing for a simple mixture of volatile organics in an
unknown atmosphere, problems may arise with regard to the resol-
ving ability of the OVA-GC column. For example, compounds
with the same or nearly the same RTs could not be resolved from
one another, but rather a blending of retention times resulted.
Thus compound identification, both qualitatively and quantita-
tively, become extremely difficult. In addition, when analyzing a
mixture with two compounds having nearly the same RTs, the
RT values were found to shift from the expected values (i.e., the
RTs became closer to one another). Again, compound identifica-
tion becomes confounded.
Variation of column length did not significantly improve com-
pound resolution. The maximum column length that can be used
with the OVA is 4 ft; use of a column of greater length will re-
sult in failure of the detector flame to ignite.
Column Efficiency: Thermally Controlled
Constant column temperature can be maintained by using an
accessory which consists of a screw-top Nalgene canister, inside
of which is an 8 in column encased in an insulating, solidified foam.
Column temperature can be controlled by using a metal slug which
has been oven heated to a known temperature above ambient con-
ditions. For subambient temperature conditions the manufac-
turer supplies a screw top plastic cap to be used in place of the
metal slug. The cup can be filled with ice, which is then placed into
the holding compartment of the Nalgene canister. The ice cup theo-
retically provides a column temperature of O °C.
The disadvantages of the thermal control accessory are: 1) ac-
curate measurements of column temperature over time cannot be
performed, 2) teflon line connectors are attached from the insu-
lated column to the column connector fittings outside of the OVA
unit. These teflon connectors are easily crimped when field sur-
veys are performed, aside from bench use, 3) due to the short, 8 in.
column there are 'a greater number of compounds that exhibit
similar RTs.
PERFORMANCE: TOTAL ORGANIC MODE
Interpretation of a total organic vapor concentration from the
OVA must be exercised with caution, however it appears possible
to use the total organic vapor monitor for the purpose of assur-
ing that airborne organic levels do not exceed prescribed concen-
trations .4'9'"
Concentrations, in ppm, of non-methane compounds in the am-
bient air are registered on the output meter assembly of the instru-
ment. These concentrations are not direct summations of all in-
dividual organic contaminants in the air being sampled; rather, the
concentration (in ppm) represents a summation of the percent rela-
tive response values characteristic of each individual organic com-
pound in the sample. A reasonable estimate of the total organic
vapor concentration in the air can be determined if the qualita-
tive composition of organic vapors in the atmosphere is known,
and if the detector weighted response for each compound in an air
mixture is known.
RECOMMENDED USES FOR THE OVA-GC
The ERT has used the OVA-GC during field activations at many
incidents where hazardous wastes were involved. The portable GC
has been applied to various situations, ranging from chemically
contaminated lagoons, to waste chemical (55 gal) drum storage
sites. Based on a variety of field experiences, the ERT recommends
the following limited applications for the OVA.
The OVA can be applied as a means of monitoring water and
soils for low molecular weight organics by using the method of
headspace analysis." Various ERT activities (e.g., Epping, NH)
have involved treatment of the contents of chemically contami-
nated lagoons using a mobile activated carbon treatment trailer
and other treatment systems. Here, the headspace method func-
tioned as a means of screening influent and effluent samples prior
to their analysis by sensitive bench instruments (i.e., GC.GC/MS),
thus helping to prevent laboratory instrument downtime as a result
of detector overload. This monitoring method can be rapidly per-
formed, and has proven to be a helpful and inexpensive means of
assisting in laboratory analyses.
Monitoring of ambient air at 55 gal drum waste storage dump-
sites yielded unexpected results. This application did not record sig-
nificant levels of volatile organic contaminants unless the OVA
-------
SAMPLING AND MONITORING
39
sampling probe was placed within inches of an opened drums bung
hole.' When monitoring soils contaminated by chemical spills, the
same placement of the OVA probe was necessary.
CONCLUSIONS
Application of the OVA-GC for ambient air monitoring yields
best results if the atmosphere being tested is of known volatile
organic composition." When the composition of the atmosphere is
unknown, both qualitative and quantitative interpretations suffer
severely.
To insure proper instrument operation, the reproducibility of
RTs must be evaluated through the use of the automatic injector
valve. This is accomplished by using any static test atmosphere
preparation where dilution of the test atmosphere is not a problem.
For example, use of Mylar or Tedlar bags is ideal for this type of
instrument check.
An excellent means of preparing internal standards for field ap-
plication is to generate static test atmospheres using the vial dilu-
tion technique. Another way to obtain internal standards is to pur-
chase commercially available multi-component canisters at pre-
scribed concentrations in an air medium. Both methods are accur-
ate, and in addition, the risk of personnel exposure to hazardous
compounds is minimized.
Headspace analysis offers an excellent technique for monitoring
contaminated waters and soils for low molecular weight organ-
ics.2'3'" This method functions adequately as a screening device,
protecting sensitive analytical instruments from detector overload.
Also, headspace analysis can be used as a method to monitor the
removal efficiency of low molecular weight organics as applied to
the use of carbon treatment systems or other systems designed to
remove organics.
It is recommended that, when the OVA-GC is to be applied for
qualitative analysis, RRTs be used rather than ARTs. It was found
the RRTs yield more precise results in the event of temperature
changes and instrument instabilities.1
ACKNOWLEDGEMENTS
The authors wish to express grateful appreciation to Ms. Kathy
Vasile and Mr. David P. Remeta for their assistance in the prepara-
tion of the manuscript.
REFERENCES
1. Gruenfeld, M., Quimby, J., and DeMaine, B., "Limited Evaluation
of a Portable Gas Chromatograph," EPA Quality Assurance News-
letter, 4, Jan. 1981.
2. Spittler, T.M., "Use of Portable Organic Vapor Detectors for Haz-
ardous Waste Site Investigations," Second Oil and Hazardous Ma-
terial Spills Conference and Exhibition. Philadelphia, Pennsylvania.
Dec. 1980.
3. Hagger, C., and Clay, P., "Hydrogeological Investigation of an Un-
controlled Hazardous Waste Site." Proc. 1981 National Conference
on the Management of Uncontrolled Hazardous Waste Sites, Wash-
ington, D.C., Oct. 1981, 45.
4. Turpin, R., "Initial Site Personnel Protection Based on Total Vapor
Readings." Proc. 1981 National Conference on the Management of
Uncontrolled Hazardous Waste Sites, Washington, D.C., Oct. 1981,
277.
5. Melvold, R., Gibson, S., and Royer, M., "Safety Procedures for
Hazardous Materials Cleanup." Proc. 1981 National Conference on
the Management of Uncontrolled Hazardous Waste Sites, Washing-
ton, D.C., Oct. 1981,269.
6. Gruenfeld, M., Frank, U., and Remeta, D.P., "Rapid Methods of
Chemical Analysis Used in Emergency Response Mobile Laboratory
Activities." Proc. 1980 National Conference on Management of Un-
controlled Hazardous Waste Sites, Washington, D.C., Oct. 1980,
165.
7. Nelson, G., Controlled Test Atmospheres, Principles and Techniques.
Ann Arbor Science Publishers, Inc. P.O. Box 1425, Ann Arbor, Mi.,
1979,p.59-154.
8. Gruenfeld, M. and DeMaine, B., "Availability of Computer Pro-
grams," EPA Quality Assurance Newsletter, 4, Jan. 1981.
9. Turpin, R., Lafornara, J., Allen, H., and Frank, U., "Compatibil-
ity of Field Testing Procedures for Unidentified Hazardous Wastes."
Proc. 1981 National Conference on Management of Uncontrolled
Hazardous Waste Sites, Washington, D.C., Oct. 1981, 110.
10. Gruenfeld, M. and Remeta, D., "Selection of a Measurement Range
for Quantitative Analyses Using Single Point Instrument Caiibra-
tion," EPA Quality Assurance Newsletter, 3, Apr. 1980.
11. Turpin, R., "Environmental Response Team's Air Monitoring Pro-
gram for Multimedia Hazardous Material Incidences," Proc. 1982
National Symposium American Chemical Society, Division of Chem-
ical Health and Safety, Kansas City, Mo., 1982.
12. "Air Pollution Training Institute (APTI) Course 435, Atmospheric
Sampling, Student Manual," EPA 450/2-80-004, Environmental Re-
search Center, RTF, NC, Sept. 1980.
13. Hachenberg, H., and Schmidt, A., "Gas Chromatographic Head-
space Analysis." Hey den & Son Ltd., Alderton Crescent, London,
England, 1979.
-------
THE USE OF PORTABLE INSTRUMENTS IN
HAZARDOUS WASTE SITE CHARACTERIZATIONS
PAUL F. CLAY
Ecology and Environment, Inc.
Woburn, Massachusetts
THOMAS M. SPITTLER, Ph.D.
U.S. Environmental Protection Agency
Lexington, Massachusetts
INTRODUCTION
Over the past several years, USEPA, state agency and private
contractor personnel involved in hazardous waste site investiga-
tions have been employing portable field instruments to assist them
in their work. In the New England region, the USEPA and a na-
tional EPA contractor, Ecology and Environment, Inc. (E & E),
have applied the use of several portable instruments to a wide va-
riety of hazardous waste site situations and activities. This variety
of field experience has provided an opportunity to thoroughly eval-
uate the utility and reliability of the instruments used, and further,
has stimulated the development of several innovations which have
enhanced the capabilities of the instruments.
The utility of the instruments described herein is primarily based
upon their capability for detecting volatile organic compounds,
which, because of their wide industrial, commercial and household
use, are found to be associated with a large majority of hazardous
waste sites. The instruments for which a number of applications
will be described include the following: the HNu Systems PI 101
Portable Photoionizer, the Century Systems (Foxboro) Organic
Vapor Analyzer (OVA) and the Photovac, Inc. 10A10 Portable
Photoionization Gas Chromatograph.
As the focus of work at hazardous waste sites across the coun-
try has progressed from the site discovery and investigation phase
to the remedial investigation and clean-up phase, the application of
portable field instruments has evolved as well. Where the initial
uses of portable monitoring instruments were focused upon per-
sonnel safety and gross site characterization, the development of
some new techniques has enabled the use of portable instruments
to become invaluable during remedial investigations and a variety
of remedial response activities.
Two areas of remedial activities which are greatly enhanced by
portable instrument use are data base acquisition and cost-effec-
tiveness. By employing a variety of field analytical techniques, the
amount of data, especially real-time data gathered during remedial
activities, can be significant and in some cases, even crucial. Also,
cost-effectiveness is achieved since portable instrument use reduces
laboratory costs and by providing virtually instantaneous data, can
reduce the costs of associated remedial investigation techniques,
such as groundwater monitoring well installation. The wide variety
of applications to which field instruments have been put by USEPA
and E & E investigators in the New England region is summarized
in Table 1.
For the majority of applications listed in Table 1, field analysis
is where the development of innovative techniques has been most
significant. The development of field analytical techniques adapted
to instrument design has enabled the list of instrument applica-
tions to be greatly expanded. Qualitative and quantitative analy-
tical techniques have been developed and tested under a variety of
field conditions and have proven to be reliable, cost-effective and
invaluable.
In addition, efforts have been directed at providing a high de-
gree of support and information to other field personnel in order to
Table 1.
Summary of Portable Field Instrument Applications for Hazardous
Waste Site Investigations
Type of
Investigative
Activity
Spills/Incident
Response
Site Discovery/
Initial
Characterization
Remedial
Investigation/
Response
Applications
•Determination of level of personnel/respiratory
protection
•Determination of extent of impact on soil, water
and air
•Evaluation of clean-up/containment effectiveness
•Determination of level of personnel/respiratory
protection
•Determination of extent of contamination in soil,
water and air
•Identification of potential sampling points and
waste container contents
•Preliminary screening of multi-media samples;
provision of rapid documentation of site-related
problems
•Rapid identification of immediate or imminent
threats to the public health
•Determination of level of personnel/respiratory
protection during a variety of remedial activities
including excavation, transfer, removal, etc.
•Hydrogeologic investigations—development of
depth vs. contaminant concentration profiles,
prevention of accidental well contamination
•Monitoring of performance standard levels for a
variety of remedial activities, such as ground-
water treatment, e.g., monitoring effluent from
carbon filtration, monitoring air quality near air
stripping towers, etc.
•Development of real-time data to assist in place-
ment of time-integrated sample stations
•Development of real-time data during time-inte-
grated sampling to refine interpretation of time-
weighted data
•Generation of data under a variety of meteoro-
logical conditions without costly laboratory
analysis.
increase the utilization of these field analytical techniques. To this
end, a comprehensive field applications manual has been pub-
lished and several week-long national training courses have been
conducted in order to provide the degree of knowledge required
to utilize the instruments beyond their "classical", or more ob-
vious applications.
To illustrate the range of applications to which portable field
instruments can be put, the authors briefly describe in the next sec-
tion the field analytical techniques which have been developed.
This section will be followed by several site case studies in which
Ambient Air
Monitoring:
Real-Time and
Time-Integrated
40
-------
SAMPLING AND MONITORING
41
the techniques were applied. The presentation of the techniques
prior to the case studies is intended to foster a clearer understand-
ing of the full capabilities of the instruments as applied to the case
studies.
INNOVATIONS FOR PORTABLE INSTRUMENT USE
Several innovations which augment the capabilities of portable
field instruments have been developed and field-tested. These inno-
vations include techniques for preparing and using qualitative/
quantitative field standards, rapid screening of samples of a va-
riety of media, and the development of a field operator support
program which includes an applications manual and a training
program, The following sections will summarize these innova-
tions.
Field Standards
Two of the portable instruments which have been thoroughly
evaluated and field tested are portable gas chromatographs (GCs).
In order to identify and quantify compounds detected by the GCs,
it is necessary to have standards available. One must also be able to
prepare aqueous or vapor standards, depending upon the medium
being analyzed.
Quantitative standards for many organic solvents can be pre-
pared quite rapidly from a saturated aqueous solution. The satur-
ated solutions are prepared by placing several milliliters of the pure
solvent with an approximately equivalent volume of water into a
septum vial. A quantity of mercury may also be placed in the vial
to provide a seal for multiple punctures of the septum. The solu-
tions are then allowed to reach equilibrium. Depending upon the
density of the solvent, the saturated aqueous layer will occupy the
upper or lower portion of the container (Figure 1).
AQUEOUS
Headspace
25%
Aqueous
COS
STOCK SOLUTIONS
5%
H20- ORGANIC So,utlon
Equilibrium , ^ ^^==^^
Organic
Aqueous
Hg
Figure 1.
GC Standards
"A" (Fig. 1) represents a vapor standard which can be pre-
pared either by placing a small quantity of pure solvent (taken from
the pure organic layer of the original standard) in the vial and cal-
culating the vapor concentration for the evaporated solvent, or by
removing headspace vapor from a stock bottle of pure solvent and
calculating the desired vapor concentration from the equilibrium
vapor pressure value. "B" represents an aqueous secondary stand-
ard which can be prepared for headspace analysis by diluting ali-
quots of the saturated aqueous layers taken from stock solutions
(standards) "C, D and E." "C and D" represent stock solutions
of an organic solvent having a density less than one, and a density
greater than one, respectively. "E" represents a stock solution for
a solvent which is readily soluble in water, such as methyl ethyl
ketone.
By using the solubility value of the solvent in water as the con-
centration of the aqueous layer, aliquots of the aqueous layer can
be used to prepare standards of the desired concentration. Refer-
ences such as the CRC Handbook of Chemistry and Physics re-
port that many of the solvents are insoluble in water. A search of
other literature reveals that the actual solubilities may range from a
few hundred parts per million (ppm) to several thousand ppm.
Since the portable GCs employ columns which are at ambient
temperature, only vapor injections may be made. Both the aqueous
samples obtained in the field and the standards prepared in the
manner described above are analyzed by using the headspace vapor
technique. The technique is based upon the fact that an equilibrium
is reached between the solvent dissolved in the aqueous phase and
the solvent present in the vapor phase (headspace), with the con-
centrations being directly proportional. Since the quantitative
standards are also prepared for headspace vapor analysis, the use
of this technique allows easy standard preparation and enables
concentrations to be established in the field. The reliability of this
standard preparation method has been supported by laboratory
trials in which standards prepared in this manner were used to an-
alyze spiked samples by gas chromatograph/mass spectrometer
(GC/MS). The concentrations determined were well within accep-
table limits. In addition, a large number of samples analyzed under
field conditions have been compared to duplicate samples analyzed
in the laboratory by GC/MS. Agreement among the data has been
excellent.
Qualitative standards for compound identification in the field
can also be easily prepared for use in headspace analysis. For ex-
ample, a mixture of commonly encountered chlorinated solvents
can be prepared from the stock saturated aqueous solutions. Prior
to going into the field, the retention time order of the compounds
in the mixture can be established on several GC columns. Once in
the field, a single injection of headspace vapor from this mixture
can provide retention times under field conditions for the com-
pounds in the mixture, and the retention times compared to those
of compounds which are found in field samples. By using mixtures
of other commonly encountered solvents, e.g., a mixture of aro-
matics, it is possible to very quickly establish tentative identifica-
tions of compounds present in the samples. Confirmation is
achieved by running the samples and standards on several different
columns. A schematic representation of a chromatogram obtained
by injecting headspace vapor from a chlorinated solvent mixture is
shown in Fig. 2.
Sample Screening
It has been the experience of New England region USEPA and
E & E personnel that volatile organic solvents are associated with
Figure 2.
CHCI Mix on Col. T-24 50 1 x 2v c. s. 2cm/min
-------
42
SAMPLING AND MONITORING
the large majority of hazardous waste sites. Even if volatiles are
not the predominant waste type at a site, the detection of volatiles
may be the first indication that a waste disposal problem exists,
since their physical properties cause them to migrate in soil, air
and ground water more rapidly than other types of contaminants.
Obviously, where specific information indicates the presence of
non-volatile wastes (e.g., oil, PCBs, creosote, inorganics, etc.) a
different approach is indicated. However, the widespread presence
of volatile wastes presents an opportunity to use portable instru-
ments capable of detecting these compounds to obtain a large
amount of data very quickly and without the costs associated with
laboratory analysis.
The two portable instruments which lend themselves exception-
ally well to being used for screening samples for volatile organic
compounds are the OVA and the Photovac, Inc. portable GC. The
OVA is equipped with a valve arrangement which allows a head-
space vapor sample to be injected directly to the flame ionization
detector (FID). The response of the instrument allows a rapid de-
termination if volatiles are present in the sample. Use of this tech-
nique will permit 30-40 samples per hour to be processed. Samples
showing the presence of volatiles are then analyzed by changing the
valve arrangement so that injections are made onto the GC col-
umn. Chromatograms are recorded and the analyses are carried
out using the qualitative/quantitative standards described in the
previous section. Samples of a variety of media, including air,
soil and water, may be screened in this manner.
The Photovac, Inc. portable GC is best applied to screening
samples where low jig/1 concentrations of volatile compounds are
anticipated. This would include the screening of drinking water
samples obtained downgradient from a contamination source
where dilution may reduce concentration; or where the leading
edge of a contamination plume may be present. Ambient air
samples obtained off-site may also typically show contaminants in
the low ppb concentration range.
The Photovac GC utilizes a highly-sensitive photoionizer detec-
tor (PID) which will respond to a wide variety of organic com-
pounds. This unit may be fitted with two options which permit rap-
id screening of samples prior to analysis. One option is a back-
flush valve which allows carrier gas flow to be reversed. Thus, a
sample aliquot may be directed straight to the detector upon in-
jection, so that the response to the presence of total volatiles may
be recorded. A second option is a dual column arrangement which
permits samples to be initially screened on a short column to de-
termine the general nature of contaminants present, without re-
quiring an inordinate amount of time between injections. Samples
thus screened are analyzed further on an appropriate longer col-
umn, if the preliminary screening indicates the presence of vola-
tile contaminants.
The HNu portable photoionizer may also be used to a lim-
ited degree to screen samples. The probe of the instrument may be
inserted into the headspace of a sample jar and the total response
noted. Compound identification as not possible and the total re-
sponse is usually reported as benzene equivalent.
In summary, sample screening with portable instruments can
provide a great deal of data in a short period of time and at low
cost. The data obtained from this process can be used to select
sampling points for detailed laboratory analysis. In addition, the
availability of data to field personnel can enable them to narrow or
broaden the scope of their efforts as the work progresses. This is
especially useful during such work as hydrogeologic investigations
—an application which will be illustrated in a case study.
Field Chemist Support Program
The portable instruments described herein are widely used
throughout the country by USEPA, state and contractor person-
nel. This is especially true of the HNu photoionizer and the OVA,
which had been used for several years in industrial applications
prior to adaptation to hazardous waste site work. As part of an
effort to disseminate the information about the techniques devel-
oped in New England, a week long training program was de-
signed jointly by EPA and E & E. The program provided prac-
tical, hands-on instruction in the techniques described in the prev-
ious sections. The primary objective was to increase the number of
field chemists who are aware of the full range of capabilities of
portable instruments. The program was conducted on two occa-
sions and included USEPA, state and contractor personnel.
To provide ongoing additional support, a comprehensive man-
ual describing all aspects of the applications of the HNu and OVA
was prepared. In addition to applications information, sections on
qualitative/quantitative standard preparation, sample screening
and field analysis techniques are included. Additional sections on
other instruments such as the Photovac GC and updated informa-
tion on techniques will be added as the information is developed.
It is believed that the effort to provide this type of support has ex-
panded the field investigative capabilities of the personnel who par-
ticipated and who have continued to be involved in hazardous
waste site work.
CASE STUDIES
In order to illustrate the application of the techniques described
in the previous section, three case studies involving the use of port-
able instrumentation will be presented. These studies were per-
formed by USEPA and/or E & E personnel during hazardous
waste site investigation activities in the northeast. Since litigation
is pending for two of the sites, the names of the sites will not be dis-
closed. Although some of the more "classical" applications of por-
table instruments are briefly mentioned, the primary focus will be
upon the presentation of data obtained through the more innova-
tive applications.
Case Study One
This site was an abandoned drum recycling/storage facility
where there was evidence that large quantities of waste organic
solvents had been discharged onto the ground. A hydrogeologic
investigation was conducted in order to determine the extent of
groundwater contamination that had occurred in the site vicinity.
Locations for the proposed monitoring wells were selected on the
basis of the site history, background information on regional hy-
drogeology and some geophysical studies. During the monitoring
well installation, two portable instruments capable of detecting vol-
atile organic compounds were used for several purposes.
First, it was determined through site records that several of the
on-site monitoring wells would be installed in areas where highly
contaminated soil could be encountered. Therefore, penetration
into contaminant-saturated soil and groundwater by drilling equip-
ment could release organic vapors. Personnel working in the direct
vicinity of the well-hole would then be subjected to an organic
vapor respiratory hazard. The HNu portable photoionizer was used
to survey for total organic vapor concentrations in the immediate
vicinity of each well-hole.
The HNu instrument is particularly adapted for this type of ap-
plication since it does not respond to ground methane, which is
often encountered during drilling. Responses shown by the in-
strument are most likely due to the presence of volatile organic
contaminants. Also, the instrument can be run continuously with-
out recharging for a normal work day.
At the locations of two on-site monitoring wells, total organic
vapor concentrations intermittently approached 1000 ppm, meas-
ured at the top of the well-hole. Although concentrations were still
in the range of ambient background in the breathing zone of per-
sonnel and favorable weather conditions existed (low temperature
and strong, steady winds), all personnel working near the well-hole
donned Self-Contained Breathing Apparatus (SCBA) to complete
well drilling at these two locations. Additionally, SCBA was used
during the withdrawal of the drilling auger sections, as highly con-
taminated soil was brought to the surface during this operation.
Second, an OVA was used as a field GC to analyze several types
of samples as the monitoring wells were installed. Fig. 3 is a photo-
graph of the OVA set up in the back of a field equipment van
where the analyses were conducted. To the left of the instrument
-------
SAMPLING AND MONITORING
43
Figures.
OVA with field standards set up in field laboratory
is a portable strip chart recorder used to record chromatograms.
Just behind the instrument is a box containing a variety of field
analytical standards similar to those described in the section on
field analytical techniques. During the drilling of monitoring wells,
soil samples were collected using a 2 in O.D. split-spoon sampler.
Samples were obtained every 5 ft or change in stratum. In addi-
tion to samples for geologic evaluation, a small quantity of soil
was placed in a septum vial and screened for the presence of vol-
atile contaminants.
Samples showing the presence of volatile compounds were then
analyzed to determine the identity and concentration of contami-
nants. A variety of volatile organic contaminants were identified,
including 1,1,1-trichloroethane, trichloroethylene, tetrachloroethy-
lene, toluene and ethylbenzene. In this manner, a depth versus
contaminant concentration profile was established for each bore-
hole.
The availability of this information enabled field personnel to
make a variety of adjustments to the project plan as work pro-
gressed. For example, it was ensured that monitoring well place-
ment, as selected from background and geophysical data, was in-
deed in the pathway of contaminant migration. On the basis of the
soil sample screening data, adjustments were made in the place-
ment of subsequent wells.
Further, an evaluation of geologic strata in comparison to con-
tamination present enabled decisions to be made about the appro-
priate depths for well screen placement and whether multi-level
well clusters were warranted at certain locations. For this partic-
ular study, it was determined from the field generated data that
multi-level well clusters, which had been proposed for several loca-
tions, were unnecessary. Thus, some cost savings were realized
from this approach.
In addition, the OVA was used to analyze wash water used dur-
ing the drilling process. Since the only convenient source of wash
water was a nearby well thought to be unaffected by contaminants,
a sample of each load of water was screened for volatiles prior to
its use in the drilling. This ensured that the drilling process was
not the source of contamination in a given well-hole.
The analysis of soil samples in the manner described resulted in
the detection of a variety of volatile organic contaminants in sev-
eral of the downgradient boreholes. The use of field standards
allowed the identification of the compounds. Field identification
agreed with the waste-type records available for the site and subse-
quent groundwater samples obtained from the wells and analyzed
by GC/MS confirmed the presence of the compounds identified in
the field. Obviously, the same benefits could not be realized if
samples were not analyzed or if samples were sent to a laboratory
for analysis, with the latter alternative being totally unacceptable
in terms of the time and expense involved.
As a footnote to this case study, one further benefit was ob-
tained from the field analysis technique. When groundwater
samples from all of the wells were obtained for laboratory analy-
sis following installation, duplicate samples from each well were
obtained for headspace analysis on the OVA. The object was to
provide preliminary data for further scope of work development
while the laboratory samples were being run. When the laboratory
results were reported back and compared to the OVA data, it was
found that for several samples, the laboratory failed to report a
compound that was found with the OVA. When the laboratory was
queried about the discrepancy, it was determined that the labor-
atory had indeed experienced difficulty with the sample analysis.
Subsequently, another round of samples was obtained and the
analysis confirmed the original data obtained by field analysis.
Case Study Two
This site involved an emergency response to the threat that a
lagoon containing millions of gallons of liquid waste would over-
flow into a nearby river at a point upstream from a municipal
water supply intake. Samples obtained from the upper aqueous lay-
er of the lagoon contained a variety of volatile organic compounds.
It was decided that the appropriate emergency response was to
pump off several feet of this top layer, run the pumped ma-
terial through a portable carbon filtration unit and discharge the
filtered effluent into the nearby river. A portable carbon filtration
unit was dispatched from the EPA's Environmental Response
Team (ERT).
In order to determine when a carbon bed became spent, per-
iodic samples were taken of the bed effluent and analyzed for vol-
atile organic compounds using an OVA which was set up at the
site. Qualitative and quantitative analytical standards were used to
identify and quantify volatiles in the effluent. This procedure elim-
inated any guesswork about when it was appropriate to switch to
a new filtration bed. It also ensured that the discharged effluent
contained no significant concentrations of volatile organic com-
pounds.
Once again, the timeliness which can be achieved through the
use of field analytical techniques is illustrated. Since there was min-
imal delay between sampling and analysis and virtually no cost,
which would have been significant if a laboratory had been em-
ployed, the entire emergency response procedure took place in a
timely and cost-effective manner.
Case Study Three
This site is a municipal landfill in a large metropolitan area.
Records indicate that there was no evidence of chemical waste dis-
posal there. As portions of the site were filled, a number of pipes
were installed to provide for venting of vapors and gases and to re-
duce the lateral migration of ground methane into the basements
of some nearby buildings. Preliminary field tests using an OVA
confirmed the presence of methane in vent pipe emissions, and
further field tests using the OVA, HNu and colorimetric detector
tubes indicated that a variety of other substances were present in
addition to methane. Of particular concern was the preliminary in-
dication that vinyl chloride was present.
In order to confirm the presence of vinyl chloride in the vent pipe
emissions and to further define the extent of air contamination
caused by the emissions, an initial, characterization was performed
using a Photovac, Inc. 10A10 portable photoionization GC. Use
of this instrument allows grab samples of air to be injected directly
onto a GC column for analysis. The need for pre-concentration of
the sample is virtually eliminated because the detector of the in-
strument is sensitive to a few ppb of many compounds, includ-
ing vinyl chloride.
Using a gas tight syringe, grab samples were taken directly from
the emission stream of several vents. The samples were injected
directly into portable GC. In addition, grab samples of ambient
air at various locations upwind and downwind of the more active
bent pipes were obtained. Several vinyl chloride standards were
used to establish retention time matches and to determine concen-
-------
44 SAMPLING AND MONITORING
INSTRUMENT: Pholovac Inc. 10AIO
Portable PiO GC
COLUMN SE 30, 5% on 60 80mesh
Chromosorb G; 4'
CARRIER GAS: Zero Air
FLOW RATE" 10cc mm
ATTENUATION 100 on A
20onBlC
CHART SPEED: 1cm mm
Q.
u
1
Figure 4.
Schematic Representations of Vinyl Chloride Chromatograms
trations of vinyl chloride in the samples. Reproductions of sev-
eral chromatolgrams recorded during the study are shown in Fig.
4. "A" is a 10 microliter (ul) injection of a sample obtained di-
rectly from one of the vents. The first large off-scale peak was iden-
tified as a C-4 hydrocarbon and the smaller, sharp peak is vinyl
chloride. The retention time of the vinyl chloride under the oper-
ating conditions was about 40 sec. "B" is a 10 ul injection of a 11.0
ppm vinyl chloride standard (balance as nitrogen). The first, small
peak is actually the nitrogen. (There are several theories for the
reason the instrument will respond to the nitrogen in a standard,
but not the nitrogen in air). "C" is a 1 Cm3 injection of an ambient
air grab sample obtained approximately 30 ft downwind from an
active vent. Here, the sensitivity of the instrument was increased
five-fold and the injection size was much larger than those of
samples taken directly from the vents.
Concentrations of vinyl chloride in the vent emissions, as de-
termined by field analysis, ranged from 10-50 ppm. Subsequently,
samples of the vent emissions were obtained for laboratory analy-
sis. Laboratory results agreed well with the field generated data.
Concentrations of vinyl chloride in the ambient air in the vicinity of
the vents varied. Obviously, wind direction with respect to the vents
and the sampling locations was a major factor. Another factor was
the air temperature, which approached 90 °F on one day during the
study, but was 70-75 °F during another phase of the study. The
concentrations of vinyl chloride in the ambient air ranged from a
few ppb up to 100 ppb.
In addition, the field analysis identified the presence of several
other compounds in the vent pipe emissions including trichloro-
ethylene and toluene. These were present at generally lower con-
centrations than the vinyl chloride. The presence of these com-
pounds was also confirmed by laboratory analysis. Through the use
of a backflush valve and using the technique described earlier, it
was possible to determine if heavier molecular weight substances
were present. None were detected.
The first phase of this study took place over two days, during
which approximately 35 air samples were obtained and analyzed.
On the basis of the field analytical data generated during this time,
several time-integrated samples were obtained by drawing ambient
air through activated charcoal tubes. The tubes were solvent de-
sorbed in a laboratory and subsequently analyzed. No vinyl
chloride was detected; however, the detection limit for this method
is generally 1 ppm. This illustrates that the use of direct injec-
tions into a portable GC can provide a more sensitive method than
current laboratory procedures for detecting vinyl chloride in am-
bient air.
During the next phase of the study, the same approach was used
to obtain some data for determining the nature of the dilution of
vinyl chloride as it is emitted from a vent into the ambient air.
In addition, several time-integrated sampling stations were set up
with activated charcoal tubes to be thermally, rather than sol-
vent, desorbed. Several instantaneous grab samples were obtained
by gas-tight syringe at a point 30 ft directly downwind from a vent
and at the location of a time-integrated sampling station. The max-
imum vinyl chloride concentration was 20 ppb. Analysis of the
time-integrated sample showed a time-weighted vinyl chloride con-
centration of < 1 ppb.
Using a variety of factors, which had been determined for this
particular vent, some calculations were performed using simple
Gaussian dispersion equations. The calculations predicted the max-
imum vinyl chloride concentration at the sampling point to be 2.8
ppb. This result agrees quite well with the field data, although sev-
eral assumptions were made for purposes of the calculations.
Additional work is needed to answer several questions which have
been raised about the vent pipe emissions.
A number of 1 Cm3 grab samples of ambient air were also ob-
tained off-site. These were injected directly into the Photovac
10A10 and analyzed for vinyl chloride. No vinyl chloride was de-
tected under the conditions present during this limited study.
Counting the samples analyzed from on-site, a total of almost 50
air samples were obtained and analyzed over a three-day period.
CONCLUSIONS
The use of portable instruments and field analytical techniques
have been applied to a variety of hazardous waste site investiga-
tive activities. USEPA and E & E personnel in the New England
region have found that the use of these instruments greatly in-
creases the amount of data which are available for evaluating the
extent and degree of volatile organic compound contamination at
sites. Where several of these instruments have previously seen most
of their use in helping to determine levels of personnel protection
and to perform initial site characterization, the new and relatively
simple techniques which have been developed have expanded in-
strument potential.
The primary advantages which can be realized by using these
techniques are time and cost savings. The use of field analytical
techniques provides data more rapidly and allows the data to be
factored into field investigations while the work is actually taking
place. The data also enable investigators to develolp additional
plans for field activities while awaiting laboratory confirmation.
This contributes to the more rapid completion of projects.
In situations where public health might be threatened, the ability
to acquire reliable data rapidly is also an obvious advantage. By
using the sample screening techniques described in the first section
and a portable GC, a USEPA chemist was able to analyze 30 pri-
vate drinking water samples for volatile organic contamination
within 24 hours. This enabled authorities to reassure residents un-
til laboratory results confirmed the initial screening conclusions.
Cost savings are realized because expensive laboratory time is not
wasted on unnecessary samples. Investigators are afforded a luxury
in that the number of field analyses need not be limited. Once the
results of field analyses with portable instruments have been eval-
uated, a much smaller, select number of samples can be obtained
and analyzed in the laboratory for further confirmation. This ob-
viously reduces laboratory costs.
Finally, as illustrated in case study three, the improved technol-
ogy that some portable instruments offer provides greater sensitiv-
ity than currently accepted laboratory methods.
-------
ANALYTICAL AND QUALITY CONTROL PROCEDURES FOR
THE UNCONTROLLED HAZARDOUS WASTE SITES
CONTRACT PROGRAM
D.F. GURKA, Ph.D.
E.P. MEIER, Ph.D.
W.F. BECKERT, Ph.D.
U.S. Environmental Protection Agency
Las Vegas, Nevada
A.F. HAEBERER, Ph.D.
U.S. Environmental Protection Agency
Hazardous Response Support Division
Washington, B.C.
LEGISLATIVE BASIS
USEPA has begun a major effort to identify and clean up uncon-
trolled and abandoned waste disposal sites that are hazardous to
the environment. This effort includes the investigation and
monitoring of several thousand potentially hazardous uncontrolled
sites throughout the nation. Recent legislation entitled "Compre-
hensive Environmental Response, Compensation and Liability Act
of 1980 (CERCLA)", commonly referred to as Superfund,1 has
given increased emphasis to this part of the Agency's activities.
SCOPE OF THE HAZARDOUS WASTE PROBLEM
Implementation of the Superfund legislation requires reliable
analytical tools and methodology. Of special concern are the re-
quirements for identification and delineation of waste sites,
characterization of waste composition and waste sites, and detec-
tion of environmental contamination due to hazardous waste
operations. These areas present a range of multimedia problems
that are not easily addressed with current technology. For example,
the waste itself can include complex mixtures of hazardous
substances at concentrations ranging from nearly pure material to
dilute solutions. The waste may occur as a solid, semi-solid, or li-
quid material. It may be stored in enclosures or simply dumped in
landfills, pits, open fields, ponds, or lagoons.
THE NATIONAL CONTRACTS LABORATORY
ANALYTICAL PROGRAM (NCLAP)
In order to accomplish the difficult task of characterizing these
uncontrolled waste sites the USEPA has found it necessary to sup-
plement its in-house analytical resources with contractor-operated
analytical laboratories. This need has resulted in the National Con-
tracts Laboratory Analytical Program (NCLAP) which provides
the Agency with a variety of capabilities for the analysis of low,
medium and high concentration waste site samples for toxic organic
and inorganic compounds. Sample matrices to be analyzed may in-
clude soils, sediments, sludges and aqueous or non-aqueous media.
The NCLAP offers a variety of other analytical services to Agency
clients but in this paper the authors will only discuss the analysis for
organics at low and medium levels in mixed media.
This paper has been reviewed in accordance with the U.S. Environmental Protection
Agency's peer and administrative review policies and approved for presentation and
publication.
Analytical services are obtained by first advertising the Agency's
requirements in Commerce and Business Daily. Interested parties
are advised that the complete Invitation for Bid (IFB) is available
from the USEPA Procurement and Contracts Division. Sealed bids
are then submitted to that office. Responsive offerers submitting
the lowest bids are notified of their potential eligibility and that a
performance evaluation (PE) sample is being sent to them for
analysis. The analytical results submitted by the laboratories are
then evaluated by the Environmental Monitoring Systems
Laboratory at Las Vegas (EMSL-LV) using criteria which are in-
cluded in the IFB package and therefore known to the offerers. An
example of the PE evaluation criteria is provided in Table 1.
Those laboratories judged to have acceptably analyzed the PE
samples are subsequently visited by USEPA personnel. The pur-
pose of these on-site visits is to evaluate the bidders' personnel,
facilities, equipment, and written procedures. Since these are the
requisite tools for successful completion of the contractual re-
quirements, a contract will not be awarded unless these re-
quirements are in place and of sufficiently high quality to meet the
Agency's standards.
SELECTION OF ANALYTICAL METHODOLOGY
The analytical methodology utilized for monitoring hazardous
waste sites must be facile, yet capable of identifying and quantify-
ing a diversity of organic and inorganic chemicals in multimedia
wastes. Current estimates of the number of chemicals manufac-
tured in the United States range from 50,000 to 75,000
compounds,2 while the number of chemicals registered with
Chemical Abstracts is reported as five million.3 The difficult
analytical task of identifying large numbers of chemical com-
pounds in wastes is aggravated by the fact that waste site concentra-
tion levels may range from parts-per-billion (ppb) levels to neat
chemicals. Notwithstanding these difficult requirements, the
analytical methods of the NCALP must product data of
demonstrable quality which is sufficient for enforcement actions.
To aid in the selection of the analytical methods to be used for
environmental monitoring, the EMSL-LV is compiling manuals of
the best available sampling and the best available methods." As a
minimum requirement any analysis procedure used in the NCALP
has been evaluated, in at least one laboratory, for accuracy, preci-
sion and sensitivity. Ideally every NCALP method should have
undergone an interlaboratory comparison test utilizing the
45
-------
46
SAMPLING AND MONITORING
Table 1.
Performance Evaluation Sample Data Scoring
Laboratory.
Dale
II. Sample A.2—Soil/Sediment Screening Sample:
A. Volatile fraction evaluated correctly.
B. B/A/N/ fraction evaluated correctly.
C. Pesticide fraction evaluated correctly.
10.
.10 .
10 .
Analytical Procedure Scoring
(Maximum = 100 points per sample set)
Sample Set B: Sample Set C:
1. Hazardous Substances List (HSL) Compound Results
A. Identifications (Maximum of 20 points)
1. All HSL compounds correctly identified.
2. One HSL compound not correctly identified.
3. Two to four HSL compounds not correctly identified.
4. More than four HSL compounds not correctly identified.
B. Quantitation (Maximum of 20 points)
I All HSL compounds within performance control window
setbyUSEPA.
2. One to two HSL compounds not within performance
control window.
3. Three to five HSL compounds not within performance
control window.
4. More than five HSL compounds not within performance
control window.
C. Matrix Spike and Duplicate
I. Matrix Spike Recoveries (Maximum of 10 points)
a. All within performance control window.
b. One outside performance control window.
c. Two outside performance control window.
d. More than two outside performance control window.
2. Precision of Duplicate Results (Maximum of 10 points)
a. All HSL compounds within performance control window.
b. One to two HSL compounds outside performance control
window.
c. Three to five HSL compounds outside performance control
window.
d. More than five HSL compounds outside performance
control window.
D. Quality Assurance Data
I. Surrogate Spikes (Maximum of 5 points)
a. All within performance control window.
b. One outside performance control window.
c. Two outside performance control window.
d. More than two outside performance control window.
2. Reagent Blank HSL Compounds (Maximum of 5 points)
a. None reported.
b. None reported at greater than 50% of detection limits.
c. One to four reported at greater than 50% of detection
limits.
d. More than four reported at greater than 50% of
detection limits.
II. Non-HSL Compound Detection Results
A. Identifications (Maximum of 20 points)
I. All detected.
2. One not detected.
3. Two to four not detected.
4. More than four not detected.
B. Non-HSL Compounds in Reagent Blank
(Maximum of 10 points)
1. None reported.
2. None reported at greater than 25% of internal standard
area.
3. One to two reported at greater than 25% of internal
standard area.
4. Three or more reported at greater than 25% of internal
standard area.
TOTAL ANALYSIS SCORES _
(Maximum of 100 points per sample set)
20
18
15
0
20
18
15
0
10
8
5
0
10
20
18
15
0
10
8
5
0
Screening Procedure Scoring
(Maximum = 60 points)
Sample Set A:
I. Sample A.I—Aqueous Screening Sample:
A. Volatile fraction evaluated correctly.
B. B A N fraction evaluated correctly.
C. Pesticide fraction evaluated correctly.
Points Set A
10
10
10
TOTAL SCREENING SCORE
(Maximum of 60 points)
TOTAL COMBINED SCORE
(Sample Sets A, B and C)
(Maximum of 260 points)
guidelines established by either the ASTM3 or the AOAC.6 A few
environmental analytical methods have undergone interlaboratory
comparison tests,7'8 but since these intercomparison tests are both
costly and time-consuming, most NCALP procedures have
undergone only single-laboratory evaluations.
The EMSL-LV is currently directing an intercomparison study of
the IFB procedures for determining volatile and semivolatile
organics in solid wastes.9- 10 In addition, the Environmental
Monitoring Support Laboratory in Cincinnati (EMSL-CI) is sub-
jecting the IFB procedure for dioxin in water (Method 613) to an
interlaboratory comparison evaluation."
ANALYTICAL METHODS SPECIFIED IN THE IFB FOR
THE DETERMINATION OF MEDIUM LEVEL/LOW LEVEL
ORGANICS IN MIXED MEDIA
As an example of the types of analytical services offered by the
Agency, via the NCALP to its clients, the analytical and quality
control requirements specified in the IFB for the determination of
medium level/low level organics in mixed media will be described in
some detail.
The tasks specified for the contractor in this IFB are:
•To receive and prepare hazardous waste site samples under strict
chain-of-custody and sample preparation procedures which are
defined in the IFB
•To screen sample extracts for IFB target compounds using gas
chromatography (GC) and gas chromatography/mass spec-
trometry (GC/MS). The purpose of this GC screen is to provide
an estimate of analyte concentrations so that extract volumes
may be suitably adjusted to protect the GC/MS detector
•To identify the target compounds by comparison of the mass
spectra and GC relative retention times of unknowns in the ex-
tract with the corresponding data derived from standards
•To quantify the identified target compounds
•To identify the non-target compounds of greatest apparent con-
centration (up to 10 unknowns in the volatile fraction and up to
20 unknowns in the semivolatile fraction)
The analytical methods specified by the IFB to accomplish these
tasks are:
•Water Methods:
Method 608 (pesticides)
Method 613 (dioxin)
Method 624 (volatile compounds boiling up to 150°C)
Method 625 (semivolatile compounds)
•Medium Level Solid Waste Methods:
EMSL-LV #1 (volatiles)
EMSL-LV #2 (semivolatiles)
•Low Level Solid Waste Methods:
Modified EMSL-CI procedures for volatiles and semivolatiles
•GC Screen for Organics in Water:
NEIC ffl procedure for volatiles and semivolatiles in water
The "600 series" water methods have been developed by the
EMSL-CI and have been published.12'15 The procedures for volatile
and semivolatile organics at medium levels (10 ppm to 15%) in solid
wastes have been prepared and evaluated by the Battelle Columbus
Laboratories under contract to the EMSL-LV.' The low-level pro-
cedures (less than 10 ppm) for volatile and semivolatile organics In
solid wastes have been developed by the EMSL-CI, subsequently
modified, and have been used by the USEPA in its Love Canal ef-
fort,16'17 and elsewhere. The GC screening procedure for organics,
-------
SAMPLING AND MONITORING
47
in water is a modified form of a method developed by USEPA's
National Center for Enforcement and Investigation (NEIC) and
has seen widespread usage in USEPA Regional laboratories.
The IFB analytical schemes for volatiles, semivolatiles and
pesticides in mixed media are illustrated in Figs. 1, 2 and 3. Each
sample is assumed to contain medium levels of organics which must
undergo a GC screen. The results of the GC screen dictate whether
the sample must undergo a medium- or low-level workup pro-
cedure. The screen result is also used to determine any adjustments
in extract volume required to optimize the subsequent GC/MS
analysis. It is particularly imiportant to keep the concentrations of
individual extract components sufficiently low to prevent satura-
tion of the mass spectral detector. Detector saturation which may
result in a contaminated mass spectrometer source is not unusual
with hazardous waste samples and can be a major cause of in-
strumental down time.
QUALITY ASSURANCE/QUALITY
CONTROL REQUIREMENTS
To ensure the reliability of all analytical data derived from the
NCALP, the IFB requires adherence to stringent quality
assurance/quality control (QA/QC) requirements by all contrac-
tors. These QA/QC requirements fall into the twin categories of
contractor in-house QA/QC and USEPA external QA/QC.
Scheme
VOA GC Screen and GC/MS Analysis
Scheme
Semivolatile (Base/Acid/Neutral) GC Screen
and GC/MS Analysis
Sampto
Yes
«/-— —
Dilution and
Woifcup by
NEIC #1
Analysis by
624
Figure 1.
VOA GC Screen and GC/MS Analysis
Included in contractor in-house QA/QC are daily calibration
and maintenance of instruments, maintenance of instrument logs
and adherence to standard operating procedures (SOPs), adherence
to chain-of-custody procedures, and preparation of control charts
to monitor the sample workup efficiency and the reproducibility of
GC/MS response factors (used for analyte quantitation). In addi-
tion, the contractor must analyze a minimum of one matrix spike,
one duplicate sample and one blank for each case of 20 samples.
QA/QC externally imposed by the USEPA on its contractors in-
cludes pre-award and quarterly on-site inspections, analysis of pre-
award and quarterly performance evaluation samples and blind
check samples, and the checking of contractor-reported data
(deliverables) which consist of data forms and magnetic tapes. The
data package required from all IFB laboratories is summarized in
Figure 2.
Semivolatile (Base/Acid/Neutral) GC Screen and GC/MS Analysis
Scheme
Pesticide GC Screen and GC/MS Analysis
s
Check
B/A/N GC/MS
Chromatogram
Figure 3.
Pesticide GC Screen and GC/MS Analysis
Table 2. The monitoring of matrix spikes, surrogate spikes,
duplicates, and PE samples provides the USEPA with an ongoing
check of the method efficiency. If a check of contractor data sug-
gests poor performance, sample deliveries to that contractor are
halted or reduced until the problem is corrected to the satisfaction
of the USEPA.
-------
48
SAMPLING AND MONITORING
SURROGATE RECOVERY RESULTS
Surrogates are chemical compounds which are added to samples
or extracts to monitor the method's efficiency. The IFB defines ac-
ceptable surrogate recovery windows; recovery values outside these
windows may provide the first indication that a method is out of
control. The lower limit of the formal performance control window
for surrogate spike recoveries are 50% for acidic semivolatiles,
70% for base/neutral semivolatiles and 80% for volatiles. The
mean values of the surrogate recoveries from a randomly selected
set of 26 samples are shown in Table 3. All samples were submitted
to contractors for low-level analysis; the 26 samples consisted of 17
aqueous and 9 soil matrices. The data in this table illustrate the
recoveries that are expected and an overview of what was actually
achieved by the laboratories. Results averaging less than the
minimum performance criteria, or greater than 100%, may not
necessarily indicate poor quality data but may reflect the interven-
tion of other factors such as matrix effects or chromatographic in-
terferences.
Table 2.
Data Package
QA/QC DATA
For Each Sample
Calibration Data
•DFTPP, Spectrum and Tabulation
•BFB, Spectrum and Tabulation
Standards Reference Data
•Acid/Base/Neutral
RIC and Spectra
•Volatiles
RIC and Spectra
•Pesticides and PCBs
GC-ECD Chromatograms, RIC
and Spectra
Sensitivity Tests and Tailing Factors
•Acids/Base/Neutrals
•Volatiles
•Pesticides and PCBs
For Each Set of Samples (20)
Method Blank Analysis
•Data Summaries
•RIC and Spectra
•Quantitation Report
Matrix Spike Analysis
•Matrix Spike Result Summaries
•QC Spike Results
•RIC and Spectra
•Quantitation Report
Duplicate Analysis
•Data Summaries
•Duplicate Analysis Results
•Relative Percent Difference Calculations
Table 3.
Surrogate Recoveries
Laboratory Performance,
Compound
Volatiles
Base/Neutral
Semivolatiles
Acid/Semi-
volatiles
Lower End
of Perform-
ance Window
80%
75%
50%
Mean*
111
69
57
Standard
Deviation
9
16
17
Min.
96
35
2
Max.
132
103
79
•Based on 26 sample analyses.
The average recoveries of volatile surrogates from both water
and soil/sediment samples are excellent (Table 4). Average
recoveries of base/neutral surrogates from water samples are also
excellent, however, their average recovery from soil/sediment
samples is not as high. The average recoveries of acidic surrogates
are low regardless of the sample matrix type. There are at least
three factors which may explain these variations between volatile
and base/neutral surrogate recoveries: (1) the greater efficiency of a
purging extraction technique over solvent extraction, (2) the greater
polarity of the semivolatile with respect to the volatile surrogate
standards, and (3) the possibility of irreversible adsorption or
decomposition of surrogate standards on soil or sediment samples.
However, these conclusions are tentative as this data base of 26
samples is too small. More definitive conclusions regarding sur-
rogate recoveries will be available as sample data accumulate in the
EMSL-LV data bank.
Table 4.
Recoveries of Volatile and Semivolatile Surrogate Compounds
from Water, Soil and Sediment Samples
Sorafat*
Lral
/L)
(Zl
Sraochlo
1.4-WcUorabMaM
Oft UOMM
Dg-ToluM*
123 103 121 103
101 104 110 94
89 99 111 97
98 100 100 117
2-riuoroblptMBTl
Dg-tapbchtUM
Dj-nuaol.
La*
Dg-TolwM
2-Tluorablphnyl
101
101
101
11
72
SO"
50
23
23
92
96
99
27
10
100
102
14
36
119 113 103
92 109 106
100 116 110
36
18
96
98
69
74
43
11
116
116
&8
50
36
17
138
126
74
47
98
97
89
98
72
87
87
54
21
112
108
62
43
63
84
87
• Surrot»c« !•*•! 90-100
*" MicroPro* p«r fru.
The individual surrogate recoveries from eight of the 26 samples
are shown in Table 5. For comparison purposes the recovery results
from an organics-free method blank are included. An idea of the
recovery precision may be gained from the results of triplicate
analyses for some of the surrogates listed in this table. With the ex-
ception of ds-phenol the absolute recovery and recovery duplica-
tion is quite good.
Table S.
Surrogate Recovery from a Low Level Aqueous Sample Matrix
Surrogate
Compound
Dt-B^.ne
l-Chloro-
2-bromo-
propane
06- Toluene
2-Fluoro-
phenol
Dr- Phenol
05-H1tro
benzene
2-riuoro-
biphenyl
l-CVei
(ug/L)
50
50
50
103
104
101
101
Dg-Haphtha- |Q(
lene
* Sample not analy
1
82
96
82
57
47
86
95
95
'zed
2
96
102
79
86
79
81
73
3
110
104
114
72
76
71
88
85
for voUttl
4
100
102
104
86
82
99
90
93
• ou
S»pU Hunber
5"
104
98
98
75 81
42 35
83 62
91 70
94 67
rrogite*.
100
•96
100
64 60
35 35
60 57
73 65
66 60
6" 7
98 100 98
94 96 92
102 106 100
60
33
63
77
69
"
98
94
104
59
26
77
89
88
81. nk
96
to
98
78
14
92
96
99
^ Sample not analyzed for ««lvol«tile •urrogatei.
PERFORMANCE EVALUATION (PE) SAMPLE RESULTS
An indication of the relative performance of analytical
laboratories can be obtained by inspecting the results obtained
from a blind check analysis or performance evaluation (PE) sample
(PE samples must be analyzed on a quarterly basis by all NCALP
contractors). A recently distributed PE aqueous sample (Table 6)
contained 20 target compounds spiked at levels ranging from 80 to
500 /tg/1. The spike compounds selected for the PE sample were
target compounds which must be analyzed for, by all NCALP
laboratories, under the IFB terms. A complete list of IFB target
compounds is provided in Table 7 (non-target compounds must
also be identified if they fulfill the IFB criteria). This PE sample
was analyzed by 16 laboratories for the volatile and base/neutral
-------
SAMPLING AND MONITORING
49
Table 6.
Composition of Performance Evaluation Sample
Fraction
Volatiles
Base/Neutrals
Acids
Pesticides
acenaphthene
acenaphthylene
acetone
acrolein
acrylonitrile
aldrin
alpha-BHC
aniline
anthracene
benzene
benzidine
benzo(a)anthracene
benzo(a)pyrene
benzo(b)fluoranthene
benzo{ghi)perylene
benzoic acid
benzo(k)fluoranthene
benzyl alcohol
benzyl butyl phthalate
beta-BHC
bis(2-chloroethyl)ether
bis(2-chloroisopropyl)-
ether
bis(2-ethylhexyl)-
phthalate
bis(2-chloroethoxy)-
methane
bromodichloromethane
bromomethane
bromoform
4-bromophenyl phenyl
ether
2-butanone
carbon disulfide
carbon tetrachloride
chlordane
4-chloroaniline
chlorobenzene
chlorodibromomethane
chloroethane
2-chloroethyl vinyl ether
chloroform
chloromethane
4-chloro-3-methyl phenol
(para-chloro-meta-cresol)
2-chloronaphthalene
4-chlorophenyl phenyl
ether
Compound
Chlorobenzene
1 , 1 ,2,2-Tetrachloroethane
Methylene chloride
1,1,2-Trichloroethane
Chloroform
1,4-Dichlorobenzene
Naphthalene
Acenaphthylene
Hexachlorocyclopentadiene
Dibenz(a,h)anthracene
N-Nitrosodiphenylamine
2-Chloronaphthalene
Benzidine
4-Nitrophenol
Pentachlorophenol
Phenol
(3-BHC
p,p'-DDD
Endosulfan II
Endrin
Table 7.
Hazardous Substances List
44. chrysene
45. 4,4'-DDD
46. 4,4' -DDE
47. 4,4'-DDT
48. delta-BHC
49. dibenzo(ah)anthracene
50. dibenzofuran
51. 1 ,2-dichIorobenzene
52. 1,3-dichlorobenzene
53. 1 ,4-dichlorobenzene
54. 3,3'-dichlorobenzidine
55. 1,1-dichloroethane
56. 1,2-dichloroethane
le 57. 1,1-dichloroethene
58. trans-1, 2-dichloroethene
59. 2,4-dichlorophenol
ic 60. 1 ,2-dichloropropane
61. cis-l,3-dichloropropene
ite 62. trans-1, 3-dichloro-
propene
ler 63. dieldrin
/!)- 64. diethyl phthalate
65. 2,4-dimethylphenol
66. dimethyl phthalate
67. di-n-butyl phthalate
68. 4,6-dinitro-2-methyl-
phenol
ane ^ ~..~.
69. 2,4-dinitrophenol
70. 2,4-dinitrotoluene
lyl 71. 2,6-dinitrotoluene
72. di-n-octyl phthalate
73. 1,2-diphenylhydrazine
74. endosulfan I
75. endosulfan II
76. endosulfan sulfate
77. endrin
78. endrin aldehyde
ane 79. ethylbenzene
80. fluoranthene
. 81. fluorene
82. gamma-BHC (lindane)
83. heptachlor
henol 84' hePtachl°r epoxide
resol) 85. hexachlorobenzene
: 86. hexachlorobutadiene
iyl 87. hexachlorocyclopenta-
diene
Concentration
O^g/L)
80
100
160
150
120
260
200
95
500
175
500
150
400
400
350
200
150
200
175
250
88. hexachloroethane
89. 2-hexanone
90. indeno(l,2,3-cd)pyrene
91. isophorone
92. methylene chloride
93. 2-methylnaphthalene
94. 4-methyI-2-pentanone
95. 2-methylphenol
96. 4-methylphenol
97. naphthalene
98. 2-nitroaniline
99. 3-nitroaniline
100. 4-nitroaniline
101. nitrobenzene
102. 2-nitrophenol
103. 4-nitrophenol
104. N-nitrosodiphenylamine
105. N-nitrosodipropylamine
106. o-xylene
107. PCB-1016
108. PCB-1221
109. PCB-1232
110. PCB-1242
111. PCB-1248
112. PCB-1254
113. PCB-1260
114. phenanthrene
115. phenol
116. pentachlorophenol
117. pyrene
118. styrene
119. 2,3,7, 8-tetrachlorodi-
benzo-p-dioxin
120. tetrachloroethene
121. 1,1,2,2-tetrachloro-
ethane
122. toluene
123. toxaphene
124. 1,2,4-trichlorobenzene
125. 1,1,1-trichloroethane
126. 1 , 1 ,2-trichloroethane
127. trichloroethene
128. 2,4,5-trichlorophenol
129. 2,4,6-trichlorophenol
130. vinyl acetate
131. vinyl chloride
compounds and by 17 laboratories for the pesticides. Twelve of
these laboratories were contractor laboratories and five were
USEPA Regional Laboratories. The PE results for this sample are
summarized in Table 8.
Some trends are immediately apparent in Table 8. Nine of twelve
contractor laboratories were unable to detect bendizine. Benzidine
is known to be toxic and it is also known to pose many analytical
problems.18 Endosulfan II, a cyclic sulfite ester of a chlorinated
bicyclic pesticide, went undetected by 5 of 12 contractor
laboratories. This compound may be undergoing a slow hydrolysis
under the pH conditions of the analysis. Beta-BHC, a chlorinated
cyclohexane, was not detected by three laboratories. This com-
pound may undergo a slow pH-catalyzed dechlorination." Because
of the problems encountered, benzidine has been excluded from
subsequent PE samples, however, a solution to the pesticides
degradation problem is difficult since many pesticides are known to
undergo decomposition during analytical workup or chromato-
graphic analysis.20
In particular, Table 8 reveals a very poor performance by
Laboratory No. 9. This laboratory failed to detect 5 of 20 com-
pounds, and some of its quantitation results appear to be an order
of magnitude too high. After inspecting its PE results, the sample
shipments to this laboratory were terminated. This is an example of
the IFB QA/QC checks detecting a potential problem with the con-
tractor laboratories. The EMSL-LV is currently compiling all PE
data in a computer bank. The data are undergoing statistical treat-
ment for the purpose of establishing precision and accuracy
guidelines for this analytical methodology.
The true spike values with the recoveries obtained by four of the
contractor laboratories are compared in Table 9. These laboratories
were selected to compare the performance of laboratories operating
in control to that of Laboratory No. 9, and to compare their in-
dividual performances with those of the statistically treated data
from all 16 laboratories. The mean and the standard deviation of
the recoveries for 16 laboratories (the data from Laboratory 9 are
excluded) are tabulated and, for the purpose of this study, perfor-
mance windows were defined as 0.5-3 times the true spike levels
(T). Data within the performance windows are not necessarily
viewed as acceptable by the Agency, however, data outside the win-
dows are viewed as unacceptable and will trigger an appropriate
QA/QC response such as termination of sample shipments or an
on-site audit and action to correct potential problems. An addi-
tional function of these windows is to indicate to the contractor
laboratory the efficiency and the possible necessity for im-
provements of their in-house QA/QC program.
The EMSL-LV is currently compiling PE data for soil, sediment
and water samples in a computer data base with the ultimate pur-
pose of refining the performance windows for each matrix type. It
is hoped that analytical difficulties resulting from the uniqueness of
individual sample matrices will not preclude the narrowing of per-
formance windows for general matrix types.
IMPROVING ANALYTICAL CHEMICAL DATA
DERIVED FROM THE NCALP
In a recent article21 the American Chemical Society's Ad Hoc
Committee Dealing with the Scientific Aspects of Regulatory
Measurements has suggested some potential problems with the col-
lection and use of analytical data by regulatory agencies. This arti-
cle concludes (in part), "Assuring the validity of quantitative in-
terlaboratory measurements at the parts-per-million level and
below is an extremely difficult technical problem. Additional
demands must be met when the data are to be used for regulatory
purposes." The USEPA is ensuring the reliabilty of its en-
vironmental analytical data by the following measures:
•The imposition and implementation of strict QA/QC procedures
•The usage of analytical protocols which have been subjected to at
least a single-laboratory evaluation
•The subjection of those analytical protocols in widest usage to
interlaboratory comparison tests
•The refinement of analytical protocols by input from EPA sci-
entists, EPA contractors and from external sources
•The development of confirmatory techniques to supplement exist-
ing analytical methods. The potential of gas chromatography/
Fourier transform infrared spectroscopy (GC/FT-IR)22*23 for or-
ganics confirmation and inductively coupled plasma emission
spectroscopy (ICP) for inorganics confirmation are currently
under active investigation. ICP has recently been employed by
the EMSL-LV as a screening tool for metals.24'25
•Utilization of external and internal peer review procedures in
those instances where time permits
•Development of, and increased reliance on additional standard
reference materials26'27'28
-------
50
SAMPLING AND MONITORING
Table 8.
Performance Evaluation Sample Results
Contractor Laboratories
Compound
Chlorobenzene
1,1,2,2,-Tetrachloro-
echane
Methylene chloride
1,1,2-TrIchloro-
e thane
Chlorof ora
1 , 4 -Die hlorobenzene
Napthalene
Acenapthalene
Isophorone
llexachlorocyc lo-
pentadlene
Benzldtne
Dtbenz(a,h)anthracene
N-Nltroao-dlphenyl
anlne
2-Chloronapthalene
4-Nltropheool
Pentachlorophenol
Phenol
t-BHC
p.p'-DDD
Endoaulfan II
Endrln
'
46
NDb
20
10
14
160
too
40
100
380
ND
180
350
50
375
210
170
HD
140
ND
HD
95
59
99
150
69
ND
41
60
3200
280
ND
ND
1300
220
160
220
98
130
213
ND
180
57
110
31
160
42
100
200
100
590
710
42
120
830
170
279
190
82
120
150
300
220
32
63
ND
100
37
140
140
70
660
200
160
80
580
100
360
230
110
100
150
170
300
41
10
65
121
43
ND
133
ND
ND
1105
ND
29
926
130
246
429
80
148
ND
224
0.06
51
94
54
130
36
160
170
96
370
ND
10
62
520
130
ND
520
69
130
150
50
180
7
60
107
58
152
31
116
98
86
ND
747
ND
10
1016
132
443
285
248
ND
218
ND
222
g
64
18
88
148
61
214
223
109
ND
1818
ND
ND
894
175
150
304
113
61
136
150
159
9
39
ND
35
148
303
ND
2710
1220
ND
4600
ND
120
13400
1800
2050
1840
1330
512
167
ND
212
10
29
76
16
120
39
140
130
75
290
530
ND
160
610
100
285
280
130
ND
200
ND
ND
||
38
63
69
100
37
96
131
70
222
ND
ND
95
493
113
168
238
6
140
130
160
160
12
42
90
55
112
40
285
285
169
ND
ND
ND
ND
1118
220
754
1080
160
99
197
197
217
A
40
68
68
85
50
130
150
84
280
380
20
50
510
120
180
360
120
160
240
20
240
EPA Laboratories
B
39
73
64
115
51
121
129
83
280
441
ND
76
396
103
344
454
104
113
119
190
189
c
37
99
17
94
40
120
130
100
200
440
33
190
10
130
190
360
75
130
290
ND
220
D
a
a
a
a
a
186
180
143
328
461
5
237
722
123
144
462
87
5
231
5
345
E
42
90
11
103
40
152
168
106
220
190
8
345
780
157
262
326
82
117
174
23
340
True
Value
80
100
160
150
120
260
200
95
300
500
400
175
600
150
400
350
200
150
200
175
250
* Laboratory "D" did not analyze the aanple for volatile organic!.
b ND - Not detected.
Table 9.
Analysis of Performance Evaluation Sample Data
Laboratory
No.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
Compound
Chlorobenzene
1,1,2,2-Tetra-
chloroethane
Methylene chloride
1,1,2-Trlchloro-
ethane
Chloroform.
1 ,4-Dlc hlorobenzene
Napthalene
Acenapthalene
Hexachlorocyclo-
pentadlene
Benzldlne
Olbenzanthracene
N-Nitroso-
dlphenylamine
2-Chloronapthalene
4-Nltrophenol
Pentachlorophenol
Phenol
beta-BHC
p.p'-DDD
Endosulfan 11
Endrln
1
46
ND
20
10
41
160
100
40
380
ND
180
350
50
375
210
170
ND
140
ND
ND
3
57
110
31
160
42
180
200
110
710
42
120
830
170
270
190
82
120
150
300
200
6
51
94
54
130
36
160
170
58
ND
10
62
520
130
ND
520
69
130
150
50
180
No.
9
39.5
ND
35.1
148
30.3
ND
2710
1220
4600
ND
120
13400
1800
2050
1840
1330
512
167
ND
212
True
Value
80
100
160
150
120
260
200
95
500
400
175
600
150
400
360
200
150
200
175
250
Mean3
49.5
75
54.2
116.0
44,3
165
160
96
617
78.3
109
851
143
294
368
122
132
186
179
219
Standard Performance Performance
Deviation Window Window
(S.D.) 0.5T - 3T Mean +2 S.D.
17.4
29.6
26.3
42
10.7
53.4
69.5
35.8
450
70.9
74.4
309.0
48
169
237
48
33
50
84
87
40 -
50 -
80 -
75 -
60 -
130 -
100 -
47.5 -
250 -
200 -
87.5 -
300 -
75 -
200 -
175 -
100 -
75 -
100 -
87.5 -
125 -
240
300
480
450
360
780
600
285
1500
1200
525
1800
450
1200
1050
600
450
600
525
750
14.7
15.8
1.6
32.3
22.9
58.4
20.2
26.4
0
0
0
249
47
0
0
26
46
86
11
45
- 84.3
- 134.0
- 107.0
- 200.0
- 66.7
- 272.0
- 300.0
- 170.0
- 1517
- 220
- 258
- 1468
- 239
- 632
- 842
- 218
- 178
- 286
- 247
- 393
The listed mean values for compounds No. 1 to 5 are based on 16 laboratories. The mean
values for compounds No. 6 to 13 are based on 16 laboratories and do not include values
reported by laboratory No. 9. Mean values for compounds No. 14 to 20 are based on 17
laboratories and do not Include values reported by laboratory No. 9.
-------
SAMPLING AND MONITORING
51
DEFINITIONS
CERCLA—Comprehensive Environmental Response, Compensation and Liability
Act of 1980
NCALP—National Contracts Laboratory Analytical Program
GC/MS—Gas Chromatography/Mass Spectrometry
IFB—Invitation for Bids
PE—Performance Evaluation
GC—Gas Chromatography
VGA—Volatile Organics Analysis
SOP—Standard Operating Procedure
DFTPP—Decafluorotriphenylphosphine
BFB—p-Bromo-fluoro-benzene
RIC—Reconstructed Ion Chromatogram
PCB—Polychlorobiphenyl
BHC—Benzene hexachloride
HSL—Hazardous Substance List
ASTM—American Society for Testing and Materials
AOAC—Association of Official Analytical Chemists
REFERENCES
1. Public Law 96-510. 96th Congress, 94 Stat. 2767, Dec. 11, 1980.
2. April, R. and Harvey, C., "Chemical Inventory Information Tape."
EPA/DF-79005. May, 1979.
3. Martinsen, D. "Survey of Computer Aided Methods for Mass Spec-
tral Interpretation." Applied Spectroscopy, 35, 1981, 255.
4. Beckert, W., Project Officer, "Compendium of Procedures for the
Analysis of Hazardous Wastes." USEPA Contract No. 68-03-3050,
Technical Directive No. 74.06. Fitzsimmons, C., Project Officer,
"Compendium of Procedures for the Sampling of Hazardous
Wastes."
5. "Standard Practice for Conducting an Interlaboratory Test Program
to Determine the Precision of Test Methods." ASTM, Part 41, E691,
1980, 959-992.
6. "Collaborative Study Procedures of the AOAC." Prepared for the
Joint International Symposium, "The Harmonization of Collabora-
tive Studies," March, 1978, London, England. Published by the
American Chemical Society, 1978.
7. Hilpert, L.R., May, W.E., Wise, S.A., Chester, S.N., and Hertz,
H.S., "Interlaboratory Comparison of Determinations of Trace Level
Petroleum Hydrocarbons in Marine Sediments." Analytical Chem-
istry, 50, 1978, 458.
8. Wise, S.A., Chesler, S.N., Guenther, F.R., Hertz, H .S., Hilpert,
L.R., May, W.E., and Parris, R.M., "Interlaboratory Comparison
of Determinations of Trace Level Hydrocarbons in Mussels". Analy-
tical Chemistry, 52, 1980, 1828.
9. Gurka, D.F., Project Officer, "Evaluation of Methods for Hazardous
Waste Analysis." USEPA Contract No. 68-03-3098.
10. Warner, J.S., Slivon, L.E., Meehan, P.W., Landes, M.C. and
Bishop, T.A., "Interlaboratory Comparison of a GC/MS Method to
Determine Semivolatile Organic Compounds in Solid Waste." Paper
presented at the Division of Environmental Chemistry of the Ameri-
can Chemical Society in Las Vegas, Nevada, Mar. 1982.
11. McMillan, C.R., Hileman, F.D., Kirk, D.E., Mazer, T., Warner,
B.J., Longbottom, J. and Wesselman, R., "Determination of 2,3,
7,8-TCDD in Industrial and Municipal Wastewater, Method 613—
Part 2—Performance Evaluation and Method Study Results." Paper
presented at the Division of Environmental Chemistry of the Ameri-
can Chemical Society in Kansas City, Missouri, Sept. 1982.
12. Bellar, T.A. and Eichelberger, J.W., "Determining Volatile Organics
at Microgram per Liter Levels by Gas Chromatography." /. Ameri-
can Water Works Assoc., 66, 1974, 739.
13. U.S. Environmental Protection Agency. "Guidelines Establishing
Test Procedures for the Analysis of Pollutants; Proposed Regula-
tions." Federal Register, 44, pp. 69464-69575.
14. Olynk, P., Budde, W.L. and Eichelberger, J.W., " Simultaneous
Qualitative and Quantitative Analyses, I. Precision Study of Com-
pounds Amenable to the Inert Gas-Purge-and-Trap Method." J.
Chromatographic Science, 19, 1981, 377.
15. Thomas, Q.V., Stark, J.R. and Lammert, S.L., "The Chromato-
graphic and GC/MS Analysis of Organic Priority Pollutants in
Water." J. Chromatographic Science, 18,. 1980, 583.
16. "Sampling and Analysis Procedures for the Love Canal Study."
USEPA Contract No. 68-02-3J68.
17. "Environmental Monitoring at-Love Canal Volume I." EPA-600/
4-82-030a, May 1982.
18. Riggin, R.M. and Howard, C.C., "Determination of Benzidine and
Diphenylhydrazine in Aqueous Media by High Performance Liquid
Chromatography." Analytical Chemistry, 51, 1979, 210.
19. Eliel, E.L., Allinger, N.L., Angyal, S.J., and Morrison, G.A.,
"Conformational Analysis", John Wiley and Sons, Inc., New York,
N.Y. 1965, 96.
20. "Manual of Analytical Quality Control for Pesticides and Related
Compounds," EPA-600/2-81-059, Apr. 1981.
21. "Improving Analytical Chemical Data Used for Public Purposes."
Chemical and Engineering News, 44, June 7, 1982.
22. Gurka, D.F., Laska, P.R., and Titus, R., "The Capability of GC/
FT-IR to Identify Toxic Substances in Environmental Sample Ex-
tracts." J. Chromatographic Science, 20, 1982, 145.
23. Gurka, D.F. and Betowski, L.D., "Gas Chromatography/Fourier
Transform Infrared Spectrometric Identification of Hazardous Waste
Extract Components." Analytical Chemistry, 54, 1982, 1819.
24. Beckert, W.F., Hinners, T.A., Williams, L.R., Meier, E.P., and
Gran, T.E., "Sampling and Analysis of Wastes Generated by Gray
Iron Foundries." EPA-600/4-81-028, Apr. 1981.
25. Williams, L.R., Meier, E.P., Hinners, T.A., Yfantis, E.A., and Beck-
ert, W.F., "Evaluation of Procedures for Identification of Hazardous
Wastes." Part 1—Sampling, Extraction and Inorganic Analytical
Procedures. EPA-600/4-81-027, Apr. 1981.
26. May, W.E. and Stemmle, "The Development of an Aqueous Trace
Organic Standard Reference Material for Energy Related Applica-
tions: Investigation of the Aqueous Solubility Behavior of Polycyclic
Aromatic Hydrocarbons." EPA-600/7-80-031, Feb. 1980.
27. Rasberry, S.R., "Reference Materials." American Laboratory, 14,
Aug. 1982, 76.
28. Rasberry, S.R., "Reference Materials." American Laboratory, 14,
Sept. 1982, 140.
-------
BIOLOGICAL SAMPLING AT ABANDONED HAZARDOUS
WASTE SITES
AMELIA J.JANISZ
W. SCOTT BUTTERFIELD
Fred C. Hart Associates, Inc.
Newark, New Jersey
INTRODUCTION
The Field Investigation Team (FIT) for the USEPA Region II
has investigated over 200 uncontrolled hazardous waste sites. Ac-
tivities at these sites included extensive sampling programs to de-
termine the extent of contamination. Standard approaches to char-
acterizing a site included the sampling of soil/sediment, surface
waters, groundwater, leachate, drums, oil, and gas vents, the drill-
ing of monitor wells, and the monitoring of ambient air. At sites
where extensive surface water contamination was suspected, bio-
logical sampling was part of an integrated approach to site inves-
tigation.
Many studies of water pollution have concentrated on chem-
ical or physical parameters, however, the principal effect of water
pollution is on biological organisms.1 Most aquatic organisms have
very narrow tolerance limits for any environmental changes.
Changes in pH as a result of acid rain, minimal increases in metal
concentrations, or organochlorine insecticides have affected aquat-
ic organisms at all trophic levels.2'3'4 Aquatic organisms can act as
natural monitors and even during short-term exposure to water
pollution will exhibit changes in their community structure. Tra-
ditionally, biologists will sample an area of an impacted community
and an area of a control or unaffected community.
Organisms are identified to a species level and compared numer-
ically using a diversity index. A healthy community, generally, is
characterized by having a large number of different kinds of organ-
isms—high diversity index—rather than a large number of a few
species. Unfortunately, this method is time-consuming and expen-
sive. Alternate diversity indices such as the Sequential Comparison
Index (SCI)! which do not require any taxomonic expertise since
specimens are compared one-to-one, are still time consuming.
While diversity indices will reveal whether the aquatic commun-
ity has been exposed to pollutants, it does not identify specific
pollutants or in what concentrations they are present. Since aban-
doned hazardous waste site investigations are directed toward spe-
cifically characterizing possible impacts tissue analyses of aquatic
organisms are necessary. Programs initiated by the states6 and
federal government' have monitored priority pollutants present in
the species of aquatic organisms human beings consume for several
years. Very few studies exist, however, which investigate the ef-
fects of specific hazardous waste sites on aquatic organisms.
Biologists routinely conduct aquatic bioassays to test the toxicity
of various pollutants. However, it is difficult to predict in a natural
environment that a given concentration of chemical will cause a
given amount of stress. The interaction of favorable and unfav-
orable stresses may result in the organisms concentrating pollu-
tants at levels which are toxic to human beings' but not to the
animals. If a particular site is contributing to an aquatic environ-
ment's contamination, tissue analyses will be required to deter-
mine what chemicals are being accumulated by the organisms and
in what concentrations.
In this paper, the authors discuss the design and execution of
biological sampling programs and present a detailed case study.
The names, specific locations, and identifying characteristics of
sites mentioned in this paper have been changed.
DESIGN AND EXECUTION OF
BIOLOGICAL SAMPLING PROGRAMS
During 1981 and 1982, the Field Investigation Team concluded
extensive biological sampling efforts on five sites, one in New Jer-
sey (site A), one in New York (site B) and three in Puerto Rico
(sites C,D, and E) to determine the effect of priority pollutants on
aquatic fauna. The criteria for the selection of these five sites, the
design of the sampling programs, sampling equipment, and sample
handling procedures for each are discussed below.
Criteria For Site Selection
Three criteria were considered to determine if the potential ex-
isted for aquatic ecosystem impact. The first criterion was loca-
tion. All five sites were located sufficiently close to water bodies to
permit direct surface and/or subsurface discharge of contaminants.
Direct discharge to surface water was either observed or the poten-
tial for subsurface discharge was substantiated by staff hydrogeol-
ogists.
The second criterion was a history of uncontrolled disposal of
hazardous materials at the site. The existence of compounds read-
ily accumulated by aquatic organisms, such as polychlorinated bi-
phenyls (PCBs) and mercury, were either documented through
previous testing or disposal records or were alleged by knowledge-
able individuals within local environmental agencies.
The final criterion was the regular utilization of the adjacent
water bodies by sport and/or sustenance fishermen. The three
water bodies in Puerto Rico were extensively utilized by local resi-
dents as a major source of food.
In addition, consideration was given to the usefulness of the re-
sulting data in the development of an enforcement case. This was a
major factor at the New Jersey and New York sites. Additional in-
formation demonstrating the detrimental impact of the sites was
beneficial to the appropriate environmental agency's case against
the site operators.
Sampling Program Design
The biological sampling programs were designed to be site spe-
cific. Each program considered accessibility, expected aquatic
fauna, sampling equipment and potential health hazards to
sampling personnel.
Sampling stations were tentatively identified after preliminary
investigations. The initial selection of the sampling stations was
designed to permit statistical comparisons between aquatic fauna
from contaminated and uncontaminated areas. In the cases of sites
A, B, and C, sampling stations were located upstream, adja-
cent to, and downstream of the points of known or suspected dis-
charge, permitting classical upstream-downstream comparisons.
Site A is described in greater detail in the case study section. Site D
was located near the head of the potentially impacted stream thus
permitting only downstream observations. Site E had potentially
52
-------
SAMPLING AND MONITORING 53
impacted two large lagoons and a stream. A model of ideal loca-
tions for sampling stations is presented in Fig. 1. Station loca-
tions were occasionally modified during sampling operations due to
unforeseen accessibility problems and also to utilize local sampling
equipment.
POM) STATIONS,
ADJACENT STATION
DOWNSTREAM
STATION
Figure 1.
Model of Ideal Locations for Biological Sampling Stations
An attempt was made to collect a variety of aquatic organisms
from the water bodies near each site. Species selection included
such considerations as organisms most apt to concentrate contam-
inants, organisms with limited mobility and thus greater potential
for exposure, organisms sought by local fishermen and organisms
within the various trophic levels to determine where pollutants were
accumulating in the ecosystem. Organisms most likely to accum-
ulate pollutants included predacious fish and organisms with high
body lipid or fat concentrations such as eels and carp. Less mobile
species included small fish and sediment dwelling invertebrates.
Species sought by anglers were determined by observation and dis-
cussion with government fishery biologists and local residents.
Additional modifications to the sampling programs were then
made during implementation because all desired species were either
not caught or were not caught in sufficient numbers for laboratory
analysis.
Sampling Equipment
The selection of sampling equipment was determined by charac-
teristics of each site such as water depths, stream versus pond,
water conductivity, stream bed consistency and expected organ-
isms. A brief description of equipment used at each site, including
methodologies, is presented below. Examples of equipment utilized
are presented in Fig. 2.
Site A—The water bodies were narrow, steep-banked, soft-
bottomed, tidal creeks which precluded the use of large nets. The
creeks were slightly brackish so electrofishing units could not be
used. Minnow traps were selected because residential small fish
were known to inhabit the creeks. Traps were baited and set for
approximately 24 hr. The stream banks were inhabited by burrow-
ing crabs. Shovels were used to dig these organisms out of bur-
rows. Site A is discussed in further detail in the case study.
NETS
VNYIOHMMM
4* OtAMlEfl FUML - r LONQ
Figure 2.
Examples of Equipment Used to Catch Aquatic Organisms
Site B—The stream adjacent to this site had a firm bottom and
was shallow enough to be waded. A portable back pack type elec-
trofishing unit was used since it permitted efficient sampling of
pooling areas adjacent to stumps and undercut banks. Sweeps of
approximately 200 ft stream sections were made at the three
sampling stations. Stunned fish were collected using dip nets.
Site C—The stream adjacent to this site also permitted the use of
a portable electrofishing unit and was sampled similarly to site B.
Large freshwater shrimp which also inhabited the stream were col-
lected using the electrofishing unit. Galvanized steel minnow traps
and local fish traps were baited and set.
Site D—A variety of sampling equipment was utilized at this site
because of the different types of aquatic habitats. A small seine
was used along the banks of the lagoons. Local fishermen also
assisted with setting circular throw nets and setting and hauling an
80 ft gill net by boat. Hooks-and-lines were also utilized because of
a recommendation from local fishery biologists.
Site E—Biological sampling at this site was not originally sched-
uled during the trip to sample Sites C and D. However, conver-
sations with local environmental officials revealed potential mer-
cury contamination. Oysters were then observed in the area of the
suspected discharge. These organisms were chipped from the sides
of a concrete dike using a hammer and screwdriver. To determine
if contaminants were reaching a nearby bay, fresh fish caught with-
in the bay were purchased from local fishermen.
Consideration also had to be given to the protection of personnel
from exposure to chemical and biological hazards. Several of the
sampling stations at Site A were in an area that required respira-
tory and skin protection. Cartridge respirators, waders and protec-
tive gloves were worn during the sampling operations. Two sites
in Puerto Rico, Site C and D, were located in freshwater areas of
the island where the parasitic trematode, Schistosoma masoni,
occurs. This parasite is transmitted to humans, its final host, via
contact with water. Contact with the surface water at Sites c'and
-------
54 SAMPLING AND MONITORING
D was minimized. Waders were worn at all times as well as elbow-
length water proof gloves. Rain gear and hard hats with face shields
were occasionally worn. When contact did occur, the areas of the
body were flushed with rubbing alcohol and abraded with rough
cloths. Aquatic organisms collected at these sites were frozen be-
fore processing to kill any remaining parasites.
Sampling Handling
An important aspect of the biological sampling programs was
handling and processing samples for tissue analysis. Care was taken
to prevent tissue decomposition which could result in the break-
down of organic contaminants. Cross-contamination of samples
from different species and stations was also avoided. Procedures
followed were generally consistent with those prescribed by
USEPA, NY Department of Environmental Conservation and
NJ Department of Environmental Protection.10'"1'2
Tissue decomposition was minimized by icing the organisms un-
til they could be processed or frozen for processing at a later date.
Dry ice was utilized for quick freezing when available. The samples
taken in Puerto Rico presented unique problems because of high
temperatures, difficulties in procuring ice and dry ice in remote
areas, and parasite contamination. The collected organisms were
packed in ice in coolers and shipped overnight to a laboratory facil-
ity where they were frozen.
The preparation of tissue samples for laboratory analyses was
conducted under both field and laboratory conditions. All equip-
ment coming in contact with the samples was detergent washed and
solvent rinsed with methylene chloride. This cleaning procedure
was repeated between samples from each station and between
samples of each species.
The size of the fish collected determined the processing tech-
nique. Larger fishes were filleted while smaller fishes were scaled,
deheaded, detailed and eviscerated. The shells of the oysters were
removed while whole body samples of crabs and shrimp were used.
The prepared tissue from each sample was either minced or
homogenized in a blender and placed in detergent washed, solvent
rinsed glass jars with Teflon lined lids. The prepared samples were
frozen, packed in ice or dry ice in coolers and snipped overnight to
contract laboratories for analysis.
Field processing of samples presents logistical problems and
should only be attempted when a large number of samples is not
planned. Field processing requires establishing a temporary work
area with equipment such as scales, measuring boards, glass trays,
and knives. The work area must provide a clean environment to
prevent contamination of samples during processing. When con-
ducted out of doors consideration must be given to possible ad-
verse weather conditions such as extremely high or low tempera-
tures, precipitation and winds which may carry contaminants. Pro-
visions for cleaning equipment between samples, discarding used
solvent, and disposing of unused portions of the organisms must
also be made.
CASE STUDY
Site A landfill is located adjacent to a tidal wetlands near a
heavily populated area. During the 1970s several thousand drums
were buried at the site. A lagoon formed east of the landfill and
was observed to be flowing into a tidal creek, Creek A, at a point
north of the landfill (Fig. 3). The USEPA initiated action and to in-
tercept, store and remove the leachate for proper disposal on an in-
terim basis. The total quantities of leachate which drained into
Creek A are unknown. Samples collected from 1978 to 1980 by the
USEPA from leachate seeps surrounding the landfill identified
numerous hazardous organics and levels of PCBs ranging from 1 to
200mg/l.
In 1981, Fred C. Hart Associates, Inc., acting as the Field In-
vestigation Team (FIT), was directed by the USEPA to conduct a
biological sampling program to determine if an impact of the la-
goon discharge on aquatic organisms could be domonstrated. A re-
connaissance of the area was conducted in late fall to determine the
accessibility of the creeks.
Figure 3.
Site Map and Sampling Station Locations
Seven sampling stations were initially selected (Fig. 3). Sta-
tions 1 and 2 were located at the head and near the mouth of Creek
B, respectively. Three stations were initially established on Creek
A: 3A, 4A, 5. Station 3A was thought to be located at the discharge
point of the lagoon. During the actual sampling, station 3A was
determined to be a minor discharge point. The primary point of
entry of lagoon leachate was at station 3B. Station 4B was added
during the invertebrate sampling program to provide a point
approximately midway between the leachate discharge and the
mouth of Creek A. Stations 6 and 7 were located on control Creek
C, near the mouth and upstream, respectively.
The study design was based on "after-impact" data and analy-
ses of the lagoon's impact were complicated by the location of the
town-operated municipal sanitary landfill. Creek B was selected to
assess separately any effect of the municipal landfill on aquatic
organisms.
Tissue samples were collected from minnows, eels, fiddler crabs,
shrimps, mud crabs, and amphipods. Sediment samples were
collected from the channel bottom and marsh soils on the creek
banks adjacent to the sampling stations. Water samples were also
collected.
Two 12 ft aluminium boats with small outboard motors each
manned by two FIT members were used to approach the sampling
stations since the channels were surrounded by a tidal march and
steep, soft creek banks made it very difficult to conduct land oper-
ations. FIT members wore cartridge respirators in Creek A, and
waders and neoprene gloves in all creeks. Each boat had one
walkie-talkie to maintain contact with a land-based observer on top
of the municipal landfill. Line-of-sight observation was not pos-
sible for the land-based observer when the boats were in Creek C.
The boats, however, maintained sight contact with each other at
all times.
Sampling in Creek A was complicated by the presence of three oil
booms, contaminated sediments and the proximity to the la-
goon. FIT personnel occasionally became stuck in the mud during
-------
SAMPLING AND MONITORING
55
sample collection. Strenuous physical exertion was necessary to re-
move the waders and FIT member from the mud. Cartridge res-
pirators, although necessary because of the close proximity to the
lagoon, were difficult to use while breathing heavily.
Galvanized steel minnow traps were selected for fish collection
based on the species captured during short seine hauls taken dur-
ing the preliminary reconnaissance. The traps were baited with
bread and raw chicken, tied to bricks and floats, and placed in the
deepest area of the channel at stations 1, 2, 3A, 4A, 5,6, and 7.
The traps were retrieved after approximately 24 hr. All fish col-
lected were the common mummichog Fundulus heteroclitus. The
fish from each trap were placed in prepared 16 oz. wide-mouth
jars. The jars were capped with Teflon-lined lids and iced in a cool-
er. Excess fish from the traps were returned to the water. Two eels
were captured, one from station 3A and one from station 7. They
were placed in separate sample jars and iced. The iced fish were
transported to the laboratory and frozen.
Fiddler crabs, Uca Minax, were dug from the creek banks where
they were hibernating, at stations 1, 2, 3B, 4B, 5, 6, and 7. Dip nets
were used to capture shrimps, Palaemonetes sp., wherever suffic-
ient weights for analysis of crabs were not collected. A few mud
crabs, Rhithropanopeus harrisi and amphipods, Gammarus sp.,
were also included in many of the invertebrate samples. The fid-
dler crabs accounted for at least 90% of the sample weight at all
stations except 6 and 7, where the crabs accounted for approx-
imately 50%. The crabs and other invertebrates were thoroughly
rinsed with creek water at each station to remove adhering sed-
iments. Rinsed invertebrates were placed in plastic bags and iced.
The samples were transported to the laboratory and frozen.
Trap retrieval and invertebrate sampling were done simultan-
eously with sediments and water sampling. Problems during trap
retrieval included difficulties in maintaining a stationary boat
position due to swift currents, and motor failure due to extreme-
ly cold weather. Maneuvering the booms in Creek A also proved
difficult due to the strong current.
Fish samples were prepared for laboratory analysis by the follow-
ing procedure. The fish from each station were partially thawed
and then placed in a glass tray. After it was measured and weighed,
each fish was beheaded, detailed and eviscerated. The fish bodies
were then minced with a stainless steel knife. Appropriate modif-
ications of the protocols developed by the New Jersey Depart-
ment of Environmental Protection12 for cleaning equipment and
preparation of fish flesh for tissue analyses were used to prevent
cross-contamination between stations. Composites for the fish
from each station were placed in two prepared sample jars, one
for organics analysis and one for metals analysis. The jars were
capped with Teflon-lined tops and refrozen. The two eels were sim-
ilarly prepared, but only for organics analysis. Whole body com-
posites of the invertebrates were prepared similarly. During evis-
ceration of the fish, the incidence of internal parasitization was
also recorded.
Table 1.
PCB Concentrations (ppm) in Vertebrates,
Invertebrates, and Channel Sediments
Location
Station 1
Station 2
Station 3A
Vertebrates
0.94
2.0
5.5
1.8 (eel)
Station 3B
Station 4A
Station 4B
Station 5
Station 6
Station 7
6.0
8.6
0.64
0.71
1.6 (eel)
•Detection limit was 0.1 ppb
Invertebrates
0.27
0.36
6.30
2.92
2.25
0.24
0.61
Channel Sediments
ND*
1.42
0.08
67.80
9.78
2.08
ND
ND
Results
The results of the organics analyses are shown in Table 1. Verte-
brates, unless otherwise specified, are the common mummichog,
Fundulus heteroclitus, and were collected at all stations sampled.
More eels were not collected since large adults move downstream in
autumn. It is clear that animals collected at stations 3 through 5,
Creek A, had levels of PCBs significantly above the values found in
Creeks B and C. Elevated levels of PCBs were also present in the
channel sediment of Creek A. Metal analyses of the tissue samples
revealed elevated levels of aluminum in both fish and invertebrates
in Creek C. Cadmium was detected at higher levels in the inverte-
brates tissue samples from stations 4B and 5 (Creek A) and was de-
tected in a water sample at station 3A.
Before the fish were prepared for tissue analysis, each fish was
measured for total length, weighed and examined for gross deform-
ities, abnormalities = , and parasites. The mean total lengths, con-
dition indices, and percent of fish parasitized in the three weeks are
shown in Table 2. The condition index, a measure of the robust-
ness (relative health)13 for total length Ktj was calculated for each
specimen. An analysis of variance (ANOVA) was used to compare
the total lengths in the three creeks. Statistically significant differ-
ences were noted in the length of Creek A mummichogs compared
to the other two creeks; mummichogs in Creek A were significantly
smaller. Examination of the correlation coefficients suggested that
PCB concentration may be responsible for the difference in total
lengths. The condition factor also exhibited the pattern of less heal-
thy fish in Creek A.
Table 2.
Comparison of Mean Total Length (TL), Condition Index
(K,|), and Percent of Fish Parasitized
Location TL(cm) Ktl % Parasitized N
Station 1 7.19 .55 5 58
Station 2 8.25 .38 0 46
Stations 6.83 .08 50 62
Station 4 7.53 .30 33 72
Stations 7.60 .41 3 59
Station 6 8.55 .38 0 44
Station? 8.17 .47 0 45
Note:
TL = total length (cm)
K,,=(Wx 10')/TL>, where W = fish weight(g)
N = number of fish
Parasites were found during evisceration within the abdominal
cavity of the fish and appeared to be a single species, probably a
monogenetic trematode, commonly found in mummichogs. No ex-
ternal parasitizations, fin-rot, or other fungal diseases were re-
corded. The incidence of parasitization did not correspond to prev-
iously established baselines on seasonal parasite burden (which for
winter should be zero) or the relative size of the mummichogs.'"
Gross deformities were recorded, however, on the eel collected
from Creek A. The eel had ulcerations and tumors on its skin as
well as internal parasites. In contrast, the eel collected from Creek
C had no parasites or deformities.
The concentration of PCBs found in the tissue analyses did not
appear to be acutely toxic to the mummichog. However, the strik-
ing incidence of parasitization in the fish and statistically signifi-
cant lower total lengths and Ktjs of the organisms in Creek A sug-
gest sublethal effects that may be related to the leachate discharging
from the lagoon. Sublethal reproductive, physiological, and mor-
phological effects of low concentrations of PCBs (1 to 5 ppb) have
been documented for fish and fish which were not adversely af-
fected at the time of initial exposure later succumbed to fungal
disease." Additionally, negative interactive and synergistic effects
of salinity, temperature, and PCB concentration on survival and
developmental rate of chronically exposed fish are documented.16
The high levels of PCBs found in the aquatic organisms led to
questions by the USEPA on the local population's fishing habits,
since the mummichog is a forage fish often consumed by piscivor-
-------
56
SAMPLING AND MONITORING
eaten & Effort I
OUT:
LOCM-1CN:
Trip Type.
Pier, aoc*
Jetty, breajo-ater
Brulge, causeway
Otter man-wade
Beacn/Bank
Party boat
Charter boat
Private or rental boat
Departure
Betum
Hours fished
No. of Anglers
Residence: State
Cdunty
Species
No. Caught
Are you going to sell any of your catch? ,
Boat typet
TtKal I of trlpa/seaaon
Type of Fishing Bqulpiwnti
Cbserved Level of Fishermen Actlvltyr
High
Average
La*
Hone
unknown
No. of Fishermen
Figure 4
Catch-and-Effort Form
ous species. The FIT was then directed to find out what and where
people were catching fish in the tidal river. The FIT developed a
catch-and-effort form (Fig. 4) and then conducted a half-day car
survey of both banks of the tidal river within 10 miles of Creek A
to determine where local fishermen had access to the river. The
municipal dock next to the municipal landfill was one of two with-
in the car survey radius that provided easy and convenient access
for local fishermen. Fishermen crabbed, and caught, and ate sev-
eral species of fish directly off the boat municipal dock. A large
part of the information gathered in interviewing the fishermen was
anedotal or historical in nature. The FIT limited data to reports of
fish caught on the survey data or within the previous week.
REFERENCES
1. Gaufin, A.R., "Use of aquatic invertebrates in the assessment of
water quality," Cairns, J. Jr. and K.L. Dickson [eds.J, Biological
Methods for the Assessment of Water Quality, American Society for
Testing and Materials, Special Technical Publication 528. Philadel-
phia, Pa., 1973,96-116.
2. Haines, T.A., "Acid precipitation and its consequences for aquatic
ecosystems: a review," Transactions of the American Fisheries Society
110, 1980,669-707.
3. Butler, P.A. and Schutzmann, R.L., "Bioaccumulation of DDT and
PCB in tissues of marine fishes." In: Marking, L.L. and R.A. Kimerle
[eds.], Aquatic Toxicology, American Society for Testing and Ma-
terials, Special Technical Publication 667. Philadelphia, Pa., 1979,
212-220.
4. Prosi, F., "HeaVy metals in aquatic organisms," In Forstner, U. and
G.T.W. Wittman. In: Metal Pollution in the Aquatic Environment.
Springer-Verlag, New York, N.Y., 1979, 271-321.
5. Cairns, J. Jr., Dickson, K.L. and Lanza, G., "Rapid biological mon-
itoring systems for determining aquatic community structure in Re-
ceiving Systems," In: Cairns, J. Jr. and K.L. Dickson [eds.], Biolog-
ical Methods for the Assessment of Water Quality, American Society
for Testing and Materials, Special Technical Publication 528. Phil-
adelphia, Pa., 1973, 148-163.
6. PCB Monitoring Program, Fish and Game Division. New Jersey De-
partment of Environmental Protection. Personal Communication—
D. Lipsky, 1979.
7. Toxic Substances in Fish and Wildlife: November 1, 1981 to April
30, 1982, 5, No. 2. Bureau of Environmental Protection. Division of
Fish and Wildlife. New York State Department of Environmental
Conservation, 25 p.
8. Schmitt, C.J., Ludke, J.L. and Walsh, D.F. 1981, "Organochlorine
residues in fish: National Pesticide Monitoring Program, 1970-74."
Pesticides Monitoring J. 4(4), 1981, 136-204.
9. Sheffy, T.B., "PCB monitoring program in Wisconsin—surveillance
strategy and use of data," In: Eaton, J.G., Parrish, P.R. and Hen-
dricks, A.C. [eds.], Aquatic Toxicology. American Society for Testing
and Materials, Special Technical Publication 707, Philadelphia, Pa.,
1980,155-163.
10. USEPA, "Interim Methods for the Sampling and Analysis of Priority
Pollutants in Sediments and Fish Tissue," U.S. Environmental Pro-
tection Agency, Environmental Monitoring and Support Laboratory.
Cincinnati, Ohio, 1980, 54 pp.
11. Anonymous. Undated. "Fish preparation procedures for contami-
nants analysis." New York State Department of Environmental Con-
servation, Bureau of Environmental Protection. 2 p.
12. Jacangelo, D.J. Undated. "Technique for processing fish and shellfish
specimens collected for polychlorinated biphenyl (PCB) analysis in
New Jersey," State of New Jersey Department of Environmental
Protection. Division of Fish, Game and Shellfisheries. Bureau of
Fisheries. 7 p.
13. Bennett, G.W., Management of Lakes and Ponds 3rd ed. Van Nos-
trand Reinhold Co. New York, N.Y., 1971, 375 p.
14. Barkman, L.L., and Jame, H.A., "The incidence of monogenetic
trematodes in the common mummichog, Fundulus heteroclitus," Iowa
State J. of Research 54(1), 1979,77-81.
15. Anonymous, "Polychlorinated biphenyls—environmental impact," A
review by the panel on Hazardous Trace Substances. Environmental.
Research 5, 1972, 248-362.
16. Linden, O., Sharp, J.R., Laughlin, R. Jr. and Neff, J.M., "Inter-
active effects of salinity, temperature and chronic exposure to oil on
the Survival and developmental rate of embryos of the estuarine killi-
fish, Fundulus heteroclitus," Marine Biology 51(1), 1979, 101-110.
-------
CORRELATION BETWEEN FIELD GC MEASUREMENT OF
VOLATILE ORGANICS AND LABORATORY CONFIRMATION
OF COLLECTED FIELD SAMPLES USING THE GC/MS
EXTENDED ABSTRACT
THOMAS SPITTLER, Ph.D.
RICHARD SISCANAW
MOIRA LATAILLE
U.S. Environmental Protection Agency Regional Laboratory
Lexington, Massachusetts
INTRODUCTION
A procedure has been developed involving a two phase mon-
itoring program to monitor low levels of volatile organics in am-
bient air. The first phase involves field measurements using a sen-
sitive portable gas chromatograph with a photoionization detector
capable of determining most common organic solvents in the range
of 0.1 to20ppb.
Variations in sensitivity depend on the electronic nature of the
molecule. Sensitivities vary in the general order of chlorinated
ethylenes> aromatics> oxygenated solvents i=> chlorinated al-
kanes> saturated hydrocarbons. Field measurements are typically
performed over a 1 to 2 day period using syringe-collected air
samples directly injected into the GC. Sample size can range up
to 10 cm3 and down to 1 jil.
In order to determine the overall concentration range of vol-
atiles, two other field instruments are employed: (1) a non-selec-
tive organic vapor detector (with sensitivity of around 0.5 ppm)
and, (2) a portable FID GC with sensitivity in the 0.2 ppm range
for some volatiles. Differentiation between ambient hydrocarbons
(Methanes and automotive exhaust) and solvents is easily achieved
on either GC.
Once the areas of concern are defined, a few laboratory samples
are collected over 4 to 8 hr periods for laboratory GC and GC/MS
analysis. Samples are collected on pre-cleaned activated charcoal
and/or Tenax packed in stainless steel tubes. Typical volumes for
integrated samples are 5-30 liters.
The laboratory analysis is performed using a Programmed
Thermal Desorber which desorbes the contents of the field samp-
ler into a 300 cm3 teflon lined piston from which aliquots can be
withdrawn for GC analysis or GC/MS confirmation.
DETAILED STUDY PLAN
Based on the results of this first phase, a detailed field study
plan is formulated. At this point, the field and laboratory data
from phase 1 have delineated the volatile organics of concern and
the general areas of significant impact around the site.
Field collection of the second phase samples is again done on pre-
cleaned charcoal and/or Tenax tubes. Laboratory standards are
prepared for any of the solvents detected in phase 1 and collec-
tion-desorption efficiency studies are inaugurated for any com-
pounds encountered in phase 1 for which the laboratory has no
prior data.
Field sample tubes are carefully preserved against contamination
after pre-cleaning by storage in individually sealed containers and
field blanks are included with each study. Samples are transported
in metal cannisters containing charcoal and are refrigerated after
collection and in the laboratory until analysis.
In the laboratory, analysis is performed using the Programmed
Thermal Desorber, GC-PID and GC/MS. Appropriate standards
are prepared fresh for all volatiles found in phase 1 and both the
GC and GC/MS are calibrated using these standards.
For GC/MS analysis of aliquots from the PID, a large syringe
is used to enable up to 60 cm3 of vapor from the 300 cm3 piston to
be injected into the GC/MS. A dry purge and trap technique was
developed and each sample is spiked with 1 to 3 internal stand-
ards that were not found in the phase 1 study. The internal stand-
ard is introduced into the syringe used to withdraw samples from
the PTD just prior to injection into the dry P&T setup.
RESULTS
To date, the correlation between initial phase 1 field data and
final GC/MS confirmation quantitation and qualitative identifica-
tion has been excellent. Often the entire problem is outlined in the
first day of field work and tentative identifications are made using
a set of field standards.
The overall technique described here is relatively simple to em-
ploy, produces answers in a rapid turnaround time and has been
taught in two courses to Field Investigation Teams (FIT) through-
out the ten USEPA regions. Several New England states are also
getting ready to use the technique in their investigations of haz-
ardous waste sites and chemical industry odor complaints.
The procedures described and the equipment used, especially the
sensitive PID GC, will require a chemist with experience in gas
chromotography. This instrument is sensitive to picogram quan-
tities of volatile organics and hence requires careful work and
scrupulous technique to prevent errors due to low level contam-
ination.
Equipment* described is supplied by the following: PID (Photo-
ionization GC); Photovac, Inc. FID and PTD (Flame ionization
GC and programmed thermal desorber); Foxboro Analytical.
'Mention of trade names or commercial products does not constitute endorsement
or recommendation for use by USEPA.
57
-------
A GENERALIZED SCREENING AND ANALYSIS PROCEDURE
FOR ORGANIC EMISSIONS FROM HAZARDOUS
WASTE DISPOSAL SITES
ROBERT D. COX, Ph.D.
KENNETH J. BAUGHMAN
RONALD F. EARP
Radian Corporation
Austin, Texas
INTRODUCTION
Emissions of organic compounds from hazardous waste disposal
sites (HWDS) can take place by a variety of routes, including
gaseous emissions, liquid runoff, and solid emissions such as dust
and other respirable particles. Analyses of each of these emissions
are required to determine the environmental and health related im-
pacts of HWDS. Also, analyses of solid and liquid materials at
HWDS are important for assessing potential emissions and impacts
of site disturbances. Normally, a wide variety of techniques is used
to analyze solid, liquid, and gaseous environmental samples. When
assessing emissions in all these phases from one site, the use of dif-
ferent types of analytical techniques can result in ambiguities when
interpreting data, as well as increasing costs.
The purpose of this paper is to describe a versatile analytical
technique capable of analyzing samples of air, water, or soil
matrices. This technique is primarily for the determination of emis-
sions of volatile compounds with a boiling point range of - 100°C
to over 200°C covered during a single analysis. In addition, the
technique responds well to a wide variety of compounds, including
halogenated organics, sulfur-containing organics, aromatics,
aldehydes, alkenes and alkanes. A variety of detectors have been
used with this technique to provide generalized screening data,
specific data on certain classes of compounds, or confirmatory
data.
This technique basically consists of the following steps:
•Separation of volatile organics from the sample matrix
•Cryogenic concentration and focusing of organics
•Gas chromatographic separation using fused silica capillary col-
umns
•Detection and screening using FID, PID, and HECD, in single
or dual detector mode
•Sample component confirmation by mass spectrometry
The fact that the same technique can be used to analyze air,
water and soil samples greatly simplifies the task of data interpreta-
tion. Application of this technique to prediction of volatile organic
emissions from chemical wastewaters and soils will be
demonstrated. In addition, good versatility accomplished through
the use of multiple chromatographic detectors with respect to
screening and/or confirmatory information for a variety of organic
compounds was obtained. Applications of several combinations of
detectors will be demonstrated.
EXPERIMENTAL
A block diagram of the technique is presented in Fig. 1. The
system was designed to accept air samples either directly from
pressurized canisters, or from solid sorbents via thermal desorp-
tion. The system was also designed to transfer volatile organics in
water samples to the gas phase via a liquid purge, and in soils
samples via heated or ambient temperature solid purges.
After volatile organics from any of the waste site matrices were
transferred to the gas phase, the purge gas (and organic consti-
tuents) was passed through a single tube Perma Pure® drier to
remove moisture. Typical sample flow rates through the drier were
100 ml/min and purge rates were maintained at 1000 ml/min. The
drier was heated to 60 ° and purged for 10-15 min between analyses.
All organics were then concentrated and focused on a IS cm. by 0.3
cm. o.d. nickel trap packed with 80/100 mesh glass beads and
cooled in liquid oxygen (— 183°C).
The design of the valving associated with the traps, and the effi-
ciency of hydrocarbon collection have previously been described.1
One modification of the design for this work was replacing the high
precision pressure gauge with a high precision vacuum gauge
(Wallace and Tiernan Model 6IC-1A-0015) and evacuating the gas
reservoir. The volume of gas passed through the cryogenic trap was
calculated by monitoring the pressure differential of the evacuated
gas reservoir.
Water
Samples
X
s
Gas
Purge
Soil
Samples
Thermal
Desorption
Figure I.
Block Diagram of the Organic Vapor System
Data
Integration
J_L
[ Data System [
Auxilliary
Data
Processing
58
-------
SAMPLING AND MONITORING
59
Samples containing ppm level organic content were analyzed by
direct injection to the cryogenic trap using a 5 ml gas tight syringe.
ap loading and injection flows were controlled using a two-
position eight-port valve (Valco #V-8-HTa). The valve and
associated stainless steel connecting tubing were heated to 60 °C to
avoid sample condensation.
Organic species were rapidly desorbed by immersing the trap in
boiling water until the temperature of the trap reached 90 °C. At
that time (approximately 20 sees), the water bath was removed and
the trap was heated to 180°C with a 750 watt heating cartridge
(Wallow #J4A198). Trap temperatures were controlled with an
Omega Model 4201 RTD Temperature Controller. Desorbed com-
pounds were swept from the trap to the chromatographic column
by the carrier gas. The transfer lines to the column were heat traced
(60°C) 0.16 cm. o.d. stainless steel tubing.
For gas chromatographic analyses, separation of organic species
was achieved on a 60 m SE-30 wide bore, thin film fused silica
capillary column (J & W Scientific). The column was operated at an
initial temperature of -50°C for 2 min, then temperature pro-
grammed to +100 °C at a rate of 6 °/min. The carrier gas was UHP
grade helium at a rate of approximately 2 ml/min and a column
head pressure of approximately ISOKPa.
The output of the capillary column was split to two detectors
simultaneously using a capillary splitter (Scientific Glass Engineer-
ing) which has been described. The instrument used was a Varian
3700 gas chromatograph with subambient temperature program-
ming capabilities, dual flame ionization detectors (FID), a photo-
ionization detector (PID), and a Hall electrolytic conductivity de-
fector (HECD). For gas chromatographic/mass spectrometric anal-
yses, separations were obtained on a 50 m DB-5 banded phase
fused silica capillary column (J & W Scientific). This column was
operated at an initial temperature of -20°C for 2 min, then
temperature programmed to 60 °C at 30°/min, and to 250 °C at
4°/min. The instrument used was a Finnigan Model 4023 GC/MS.
The spectrometer interface was direct source coupled, the scan
range was 35-300 amu, and the ionization voltage was 70 eV.
Chromatographic data were processed using a Varian 401
Chromatography Data System. This served primarily as an in-
tegrator and transferred integrated data to an Apple II Plus
Microcomputer which was used for quantitative and qualitative
peak identification and calculation of relative response ratios for
the various detector combinations. GC/MS data were processed by
an INCOS computer utilizing searchable NBS libraries and manual
spectral interpretation.
RESULTS AND DISCUSSION
System Performance
The unique aspects of these techniques are the ability to analyze
air, water and soil samples using the same analytical system, and
the wide range of compounds which can be determined through the
various detector combinations. These aspects provide very cost ef-
fective techniques, as well as greatly simplifying data interpretation
for predicting volatile organic emissions from various matrices.
Analyses of air, water, and soil samples on the same analytical
system were accomplished by transferring volatile organics to the
vapor phase, removal of moisture, and cryogenic concentration
and focusing. The transfer of volatile organics from solid sorbents
was accomplished via thermal desorption. For water samples, the
transfer was affected using a low-volume (5-10 ml) Bellar and
Lichtenbergh apparatus. For soil samples, small quantities of soil
(50-100 mg) were placed in an all glass sparging chamber and purg-
ed with UHP N2 at ambient or elevated temperatures.
Analysis of soil and water in this manner provided an indication
of organic vapor emissions which could occur during treatment or
disturbances of the site. Analysis of air samples by the same techni-
que provides a ready confirmation of emissions during treatment
procedures or site disturbances.
The problem of moisture in the sample gas stream has mandated
the use of a variety of different analytical techniques for air, water
and soil analyses in the past. The use of capillary columns and
subambient temperatures have been hampered by moisture, which
causes freezing at the head of the column. The single-tube Perma
Pure® drier used for this work functioned very well for the
removal of moisture and in the throughput of organic vapors.
Recoveries of a variety of organic classes through Perma Pure®
driers are presented in Table 1.
In general, good recoveries were obtained for low and medium
molecular weight aliphatic and aromatic hydrocarbons, chlorinated
hydrocarbons, aldehydes, and sulfur containing hydrocarbons.
Low recoveries of alcohols and variable recoveries of ketones were
obtained with Perma Pure® driers.
Fused silica capillary columns and subambient temperature pro-
gramming were used to cover a wide range of organic compound
volatilities and functionalities. With the SE-30 column, compounds
with boiling points ranging from -100 °C to + 200°C could be de-
termined in a single analysis. With the DB-5 column used for
GC/MS, compounds with boiling points ranging from 36 °C to
220°C were normally determined in a single analysis. Organic com-
pound elution orders on both columns were essentially based on
boiling points, which aided in chromatographic data interpreta-
tion, and in comparisons of GC and GC/MS data.
Table 1.
Summary of Organic Vapor Recoveries Using Perma Pure8 Driers
Class
Carbon
No.
Range
Alkanes
Alkenes
Aromatics
Chlorinated
Hydrocarbons
Aldehydes
Ketones
Alcohols
2-10
2- 5
6-10
1-6
2-5
3-5
1-5
No.
Species
Tested
19
8
12
6
6
4
5
Percent
Recovry
(%)
95-106
100-101
90- 99
99-104
84- 95
0-70*
0
Concntra.
Range
(ppb-C)
16- 280
141- 204
167- 505
66- 206
140-1400
28- 66
25-1500
•Change in drier permeability observed
Correlation of Emissions
Correlation of organic vapor emissions from liquid and solid
phases at HWDS can be difficult when different techniques are
used to analyze various sample phases. Leachate from a HWDS
was collected and tested using laboratory scale biotreatment cham-
bers to determine if the leachate could be disposed of in this man-
ner. Normally, these tests are performed by only monitoring influ-
ent and effluent concentrations of pollutants in the aqueous phase.
In this case, the system aeration was carefully controlled to simulate
large scale treatment facilities, and air purging through the system
was collected and analyzed.
An analysis of the water influent to the treatment system (1/20
dilution of leachate with artifical sewage) using the purge technique
which was described, and an analysis of air passed through the
treatment system is shown in Fig. 2. Volatile chlorinated hydrocar-
bons as well as aromatics were observed in the air.
Even with the nonselective detector which was used in this case
(FID), the correlation of organic components in air with those in
the water is apparent. Species identifications in the water were con-
firmed by GC/MS. It is apparent from this study that during
biotreatment of chemical wastes, air emissions as well as water ef-
fluent emissions should be monitored to properly determine en-
vironmental impact of the treatment.
Multiple Detector Applications
The use of multiple gas chromatographic detectors with this
technique provides generalized screening information as well as
confirmatory data, depending on the application. Examples of
various detector combinations on the analysis of different types of
waste samples will be presented.
-------
60
SAMPLING AND MONITORING
FID
Allen 16
260 mL of Air
FID
Allen 4
1 mLol Water
1:20 Dilution
1 i
i ,
i. .1,
I
6
i !
2 c
c
UJ
Figure 2.
Analysis of the Aqueous Influent and Treatment Air from a Bench
Scale Biological Treatment Study of a Hazardous Waste Site Leachate
PID
Sensitivity: 10-"
Attn: 4
Makeup: 30mL/mln
Split: 75/25 (F/P)
f
i I
1 1
o
11 .
u
^LK
«
s
ll.
L-
,1
1 FID
| Sensitivity: 10-1 1
1 Attn: 4
£ Makeup
S Sample:
j.
Inl
1 1 1 1 1 1
0 5 10 15 20 25
30mL/mln N2
Air, Waste Site
(/"SOOmL)
ill
yuijuLuiJL^.
30 35
Minutes
Figure 3.
Analysis of Air Collected Over a Hazardous Waste Disposal Site
by Simultaneous PID/FID Detection
The combination of flame ionization and photoionization detec-
tors provides a good generalized screening technique for organics
present in ambient air. The flame ionization detector in general
responds to compounds on the basis of carbon content; hence, it
serves as a universal detector for hydrocarbons and is good for
quantitative purposes. The photoionization detector responds to
compounds based on their ionization potentials, which in turn is
dependent on the degree of unsaturation in the molecule among
other factors. By calculating a ratio of the response between the
two detectors, it was possible to estimate the degree of unsaturation
(and chemical class) for a compound producing a given chromato-
graphic peak.
The chemical class information provided by the dual detector
technique proved to be a valuable tool when used in conjunction
with GC retention time data for species identifications. Normalized
response data for a variety of volatile organic compounds on the
two detectors have been developed.2 Average PID/FID responses
normalized to that of toluene (TNR) are presented in Table 2 for a
variety of organic classes. These data were separated for those com-
pounds eluting before and after 17 min (the time at which
aromatics begin to elute).
This detector combination provides identification information
for a wide variety of compounds, including alkanes, alkenes,
aldehydes, ketones, aromatics, saturated and unsaturated
halogenated compounds, and sulfur containing hydrocarbons. The
versatility of this detector combination in combination with high
resolution chromatography makes it ideal for screening air, water
and soil samples from HWDS. An example of the analysis of air
collected at a HWDS by this technique is presented in Fig. 3. The
identification of a number of aldehydes, aromatics, and
chlorinated aromatics is illustrated.
The Hall electrolytic conductivity detector (HECD) can be used
in modes selective for halogenated, nitrogen, or sulfur containing
compounds. When used in conjunction with the PID, the two de-
tectors provide a range of unique types of information. Application
of the HECD-halogen detector in conjunction with the photoioni-
zation detector for analysis of a mixture of halogenated and
aromatic compounds is illustrated in Fig. 4. This figure
demonstrates that with this technique, all of the US EPA 601 and
602 purgeable compounds3 can be determined in a single analysis.
All of the halogenated compounds were detected by the HECD,
whereas nonhalogenated aromatics were detected only by the PID,
In addition, the PID provides discriminatory information for
halogenated compounds, since this detector only responds to the
unsaturated halogenated organics.
Application of this technique to a chemical wastewater sample is
shown in Fig. 5. Low concentrations of aromatics and higher levels
of oxygenated compounds were detected by the PID. A number of
different halogenated compounds were detected by the HECD. No
response for the halogenated compounds on the PID confirmed
that these species were saturated.
Table 2.
Hydrocarbon Class PID/FID Normalized Responses (Ref. 2)
Class
Retention Times <17 mln
Species
Tested TNR
Retention Times >17 mln
Species
Tested TNR
(Mean ± SD) (Mean ± SD)
Halogenated
alkanes
Simple alkanes
Cyloalkanes
+ Trimethyl-
alkanes
Alkynes
Alkenes
Aldehydes
Ketones
Aromatics
Chlorinated
Aromatics
Chlorinated
Alkenes
Sulfur Hydro-
carbons
9
13
2
3
23
4
2
0
0
3
2
0
3
3
3
70*
69f
157
218
500
±
±
±
±
±
±
±
±
±
0
3
1
3
11
10
5
180
210
6
10
5
0
20
1
2
21
7
4
2
1 ±
12 ±
27 ±
—
55 ±
56 *
123 ±
—
141 ±
211 ±
129 ±
1
4
8
10
25
12
91
37
•Does not include ethylene
tDoes not include acetaldehyde
-------
SAMPLING AND MONITORING
61
Hall-Halogen
Range: 100
Altn: 16
Makeup: 30mL/min H2
Split: 50/50 (H/P)
LiLU
Hall-Halogen
Range 100
Atln: 64
Makeup Gas: 30ml_/min H2
„ Split: 50/50 (H/P)
PID
Sensitivity: 10-1'
Attn: 32
Makeup: 30mL/min N2
Sample: 601, 602 VOC Standard
MOOng)
11
c g>
S |
y g «
£ ° S
"I
"§
ft
S E
10 15
25
30 Minules
PID
Sensitivity: 101'
Attn: 64
Makeup: 30mL/min N2
Sample: Wastewater (1.0mL)
10
i
15
20
25
30
Minutes
I
35
Figure 4.
Analysis of 601, 602 Purgeable Standards by
Simultaneous HECD-Halogen and PID Detectors
Figure 5.
Analysis of a Chemical Wastewater Sample by
Simultaneous HECD-Halogen and PID Detection
The HECD can also be used in a mode selective for sulfur con-
taining compounds. In addition, the PID gives a very strong
response for sulfur containing compounds. The two detectors in
conjunction provide confirmatory information for a variety of
sulfur containing compounds. The PID does not respond to
hydrogen sulfide or carbonyl sulfide, but gives very good responses
for the mercaptans, organic sulfides, and thiophenes. The analysis
of air collected during trenching operations at a HWDS by this
technique is presented in Fig. 6. A large amount of sulfur dioxide
was detected by the HECD, along with small quantities of several
mercaptans and dimethyl thiophene. The PID did not give a
response to SO2, but did give responses for the other sulfur contain-
ing compounds and a number of aromatics.
When samples from HWDS become too complex, or when a high
degree of confirmation is required, GC/MS techniques must be
used. The similarities in chromatography of he SE-30 column used
for GC analyses and the DB-5 column used for GC/MS analyses
simplify interpretation of data between these techniques.
CONCLUSIONS
The techniques described in this paper provide a wide range of
information concerning volatile organic emissions from hazardous
waste disposal sites.
•The use of the same technique for analyzing air, water and soil
samples simplifies the tracking of volatile emissions from soil or
liquid wastes, as well as providing data necessary for prediction of
emissions from HWDS.
•The use of multiple chromatographic detectors provides unique
screening and class selective data, as well as cost effective anal-
yses.
•Determinations of a variety of different compounds and organic
classes from samples collected at HWDS were demonstrated.
•When complex samples or a high degree of confirmation is re-
Hall-Sulfut
Range: 100
Attn: 8
Makeup: 30mL/min Air
Split: 50/50 (H/P)
PID
Sensitivity: 10'11
Attn: 8
Makeup: 30mL/min N2
Sample: Air-Waste Site 2
(1.0mL)
/I
l
i
I
UJ
jjl
10
15
20
25
30
Minutes
Figure 6.
Analysis of Air Collected During Trenching Operations at a HWDF
by Simultaneous HECD-Sulfur and PID Detectors
-------
62
SAMPLING AND MONITORING
quired, GC/MS techniques were used.
•Comparisons of GC and GC/MS data were simplified by using
similar columns and the same sample component separation and
injection techniques.
REFERENCES
1. Cox, R.D., McDevitt, M.A., Lee, K.W. and Tannahill, G.H., "De-
termination of Low Levels of Total Nonmethane Hydrocarbon Con-
tent in Ambient Air," Environ. Sci. Technol. 16, 1982, 57.
2. Cox, R.D. and Earp, R.F., "Determination of Trace Level Organics
in Ambient Air by High Resolution Gas Chromatography with
Simultaneous Photoionization and Flame lonization Detection,"
Anal. Chem., in press, Nov., 1982.
3. USEPA, "Guidelines Establishing Test Procedures for the Analysis of
Pollutants," Federal Register, 44 (233):69468-69476 (1979).
Compound Identification Scan No.
1. Butane 171
2. Trichlorofluoromethane 239
3. Pentane 247
4. Benzene 359
5. Toluene 442
6. Tetrahydrothiophene 487
7. MethyltetrahydrothiopheneJ 539
8. Ethyl benzene 555
9. Xylenet 567
10. XyleneJ 602
11. Dimethyltetrahydrothiophenet 626
12. Dimethyltetrahydrothiophenej 632
13. Dimethyltetrahydrothiophenei 759
14. Ethyl tetrahydrothiophene 698
15. Trimethyl benzenej 831
JOne of two or more structural isomers reported in this sample.
Figure 7.
Reconstructed Ion Chromatogram of 100 mg Soil Sample
Analyzed by GC/MS on 50 m SE-54 fscc.
-------
THE AIR QUALITY IMPACT RISK ASSESSMENT ASPECTS
OF REMEDIAL ACTION PLANNING
JAMES F. WALSH
KAY H. JONES, Ph.D.
Roy F. Weston, Inc.
West Chester, Pennsylvania
INTRODUCTION
As part of the current federal activities aimed at cleaning up
hazardous waste sites under Superfund, Roy F. Weston, Inc.
(WESTON) has had remedial action planning responsibility for
USEPA Regions III and V. To date, WESTON has had experience
with a significant number of sites including both Superfund and
municipally funded landfill problems. These sites have all had
diversified air exposure risks associated with them, and have re-
quired differing levels of analysis. Some recent sites analyzed in-
clude:
•Chem-Dyne, Hamilton, Ohio—surface storage of drum and bulk
waste, possible surface and subsurface disposal problems.
•Sylvester Landfill, Nashua, New Hampshire—contaminated
groundwater migration to surface stream.
•Lehigh Electric, Old Forge, Pennsylvania—PCB contaminated
soil, including coal fines, from leakage of stored transformers and
capacitors.
At each of these sites, the total array of potential air exposure
risks were not recognized because of the primary focus on
hydrogeological concerns. Historically, the only air pollution em-
phasis involved: (1) the safety protocols set to protect on-site
workers and (2) emergency planning to protect the surrounding
community in case of an accidental release of hazardous chemicals.
The benchmarks being used for such assessments were limited to
Threshold Limit Values (TLVs) and Permissible Exposure Limit
(PELs). Little or no consideration has been given to the off-site
migration of low levels of potentially carcinogenic compounds dur-
ing disturbed or undisturbed site conditions.
There are three steps in remedial action planning in which air ex-
posure risks must be considered, both from a health impact and a
liability point of view. The purpose of the authors in writing this
paper is to identify and explain the necessary elements of such
assessments as well as present several case studies where the air risk
became a remedial action design criteria.
SCOPE OF REQUIRED RISK ANALYSIS
There are many factors that must be considered when one is
evaluating the air quality risk considerations of a hazardous waste
site. The potential long-term carcinogenic risks associated with low
pollutant concentrations emphasize the need for a careful analysis.
As mentioned above, there are three logical divisions in this type
of evaluation. The first is an assessment of the present off-site risk
before any type of action is considered since air risk considerations
may be a remedial action design factor. This assessment would in-
clude an analysis of any existing air quality impact data. In each of
the three sites noted in the introduction, there was no information
concerning off-site pollutant migration simply because this
pathway was seldom considered.
Oftentimes, the only air data available from hazardous sites con-
sists of total organic concentrations given in the parts per million
range, which is approximately the concentration that causes
subacute health effects. Because of this data scarcity, the monitor-
ing requirements of the particular site must be assessed. In areas
where the contents of the landfill or storage area are known, limita-
tions may be placed on the number of toxics that may be sampled
for and analyzed. This must be determined on a case-by-case basis.
The second major task is associated with the potential air quality
impacts of remedial action alternatives. An air risk assessment of
the candidate alternatives must be undertaken to predict the
ramifications of whichever cleanup alternative is chosen. Technical
alternatives which have potential air risks may include:
•Physical removal operations
•Air or steam stripping operations
•Chemical stabilization
•On-site combustion
Most remedial action alternatives are likely to increase the ex-
posure risk in the short-term, so this increase must be monitored
and controlled by instituting operational practices to minimize off-
site exposures. For example, at the Lehigh site, dust suppression
may be required to limit off-site transport of PCB-contaminated
coal fines.
Regardless of the mitigating measures taken, the authors believe
that a credible monitoring program should be conducted during the
remedial activity to fully document any off-site exposures in order
to preclude future liability. This monitoring program should be
designed to quantify exposures at the most critical receptor areas.
Body burden calculations can be made to estimate the potential im-
pact of these exposures from a long-term, carcinogenic, point of
view. The compressed short-term nature of this exposure must be
taken into consideration when assessing the body burden.'
In addition to the low level body burden considerations of
remedial action, the subacute off-site emergency type exposure
potential must also be considered. This would include steps to be
taken in the event of any significant release due to drum rupture,
explosion, etc. Dispersion modelling should be employed to deter-
mine what on-site concentrations might cause potential problems
off-site. Evacuation/notification plans should be developed accor-
dingly.
The final phase of a complete air quality risk assessment deals
with post'dosure considerations. This would primarily consist of
continued area monitoring over a short period of time where air
risks were clearly identified as a remedial action problem. The pur-
pose of this monitoring is to document the fact that air risks have
been reduced to acceptable levels. This is an additional liability
prevention step.
HEALTH EFFECTS CRITERIA
In all segments of this assessment, certain criteria must be used to
determine when a risk level is, or is not present. There is a good
deal of uncertainty as to which effects criteria should be used
especially when it relates to cancer risk. This controversy is one of
the major hurdles, which must be quickly overcome if the preven-
tion of hazardous pollutant exposure is going to develop in a
63
-------
64 SAMPLING AND MONITORING
uniform and rational fashion. WESTON has been using the follow-
ing criteria to evaluate the range of risk levels of interest:
•Cancer Assessment Group Values (CAGs)!
These are recommended lifetime exposure limits to known car-
cinogens which have been developed by the USEPA for a limited
number of toxic compounds. The CAG number represents max-
imum allowable concentrations that may result in incremental risk
of human health over the short-term or long-term at an assumed
risk. This assumed risk is the expected number of increased in-
cidences of cancer in the effected population when the concentra-
tion over a lifetime equals the specified value.
The CAG values are listed in the "Land Disposal Toxic Air
Emissions Evaluation Guideline" published by the USEPA in Dec.
1980. They represent a further refinement of carcinogenic assess-
ment beyond the MEG values because of the compound specific
nature of the toxicological evaluation involved with CAG
documentation. Therefore, Cancer Assessment Group values
should be used in all cases of conflict between MEGs and CAGs.
•Multimedia Environmental Goals (MEGs)3
The MEGs were developed in recent years by the Research
Triangle Institute for the USEPA to meet the need for a workable
system of evaluating and ranking pollutants for the purpose of
multimedia environmental impact assessment. Consideration in ar-
riving at these ambient level goals was given to existing Federal
standards or criteria, established or estimated human threshold
levels, and acceptable risk levels for lifetime human exposure to
suspected carcinogens or teratogens, among others.
•Threshold Limit Value (TLV)4
TLVs are established by the American Conference of Govern-
mental Industrial Hygienists as guidelines for prevention of adverse
occupational exposures and are based on both animal studies and
epidemiological findings and inferences. They refer to airborne
concentrations of substances and represent conditions under which
it is believed that nearly all workers may be repeatedly exposed for
eight hours a day, five days a week, without adverse effects. Many
TLVs also protect against short-term aggravations such as eye ir-
ritation, odor impacts, headache, etc. However, TLVs do not
represent the hypersusceptible minority of individuals. Additional-
ly, by definition, they assume a daily period (non-working hours)
of non-exposure time.
Many states have developed or are currently developing ambient
air hazardous waste guidelines and regulations based on these
TLVs. A common practice is to adjust the TLV for a 24 hr ex-
posure and then divide by some large uncertainty factor. Common
limits are TLV/300 - TLV/420.5'6'7
CASE STUDIES
Chem-Dyne: Hamilton, Ohio
The Chem-Dyne site consisted of approximately 40,000 waste-
containing drums and 15 to 20 bulk storage tanks containing an
unknown array of chemical substances. The site borders a residen-
tial neighborhood with a large adjacent recreational area complete
with several ballfields and a swimming pool. Although not quan-
tified, there was a concern that on-site contaminants were
migrating off-site and impacting on the surrounding residents.
Prior to any remedial action undertaken by WESTON, state of-
ficials arranged for a large quantity of drummed wastes to be
removed by the generators, leaving approximately 10,000 drums
and the remaining bulk wastes.
The first step from an ambient air assessment point of view was
to determine what the existing concentrations were before any on-
site remedial activity began. This determination would help quan-
tify existing levels and serve as a baseline. A comparison of baseline
values with concentrations determined during operational phases
would document increases due to on-site activity and also
demonstrate ambient air quality improvement after the remedial
activity is concluded and the site closed.
The background sampling program consisted of a two-day, five
sample per day perimeter monitoring operation using USEPA
recommended methodology.' One sampler was placed upwind and
four spanned the appropriate downwind areas. During the actual
sampling process, two samples per day were collected at downwind
locations. This sampling protocol was limited by project budgetary
considerations.
The monitoring identified 17 different contaminants leaving the
site both prior to, and during the on-site remedial activities. Three
of the most significant contaminants and the peak levels identified
during the entire five-day operation are shown in Table 1.
Table 1.
Selected Air Contaminants found at Chem-Dyne
Air Contaminant
Benzene
Trichloroethene
Tetrachloroethene
Peak Concen-
tration (ppb)
19.2
1.8
0.62
CAG Level
(ppb)
0.065
0.44
0.194
Ratio Concen-
tration/CAG
295
4.1
3.2
Benzene presents the most danger to the surrounding popula-
tion. Exposure concentrations of the magnitude determined, for
even a short period of time (months to years), may cause an ex-
ceedance of the body burden calculations on which the CAG value
is based, therefore, theoretically increasing the cancer risk in the
area. The ambient air exposure route in the case of the Chem-Dyne
site should be a determining factor in deciding upon various
remedial action alternatives.
Furthermore, these documented levels will allow a comparison
between pre- and post-closure site conditions. When all the wastes
have been removed, a significant improvement in air quality should
resujt. If, in fact, this does not occur the other emission sources
such as contaminated soil must be analyzed to insure the health of
the surrounding population.
Nashua, New Hampshire
The problem at Nashua resulted from illicit dumping of approx-
imately three million gallons of hazardous waste onto an illegally
operated disposal site. Groundwater contamination was extensive
and the groundwater plume was migrating toward a small nearby
stream.
The air quality concerns involved the eventual interception of the
groundwater plume with the stream and the subsequent volatiliza-
tion of the organics. The air quality assessment dealt with the alter-
native of a no-action scenario, in other words, allowing the con-
taminated groundwater to meet the stream without any type of
remedial action. The assessment of the potential ambient air impact
associated with this site necessitated a completely different
analytical protocol (Fig. 1).
Using various emission estimates9'10 in conjunction with known
contamination levels in the groundwater, a rate and period of
volatilization was predicted above the stream. The expected at-
mospheric concentrations in a nearby trailer park were modelled"
based on these emission rates. The predicted exposures for the most
critical contaminants based on health effects considerations are
shown in Table 2. The table shows the comparison of the permissi-
ble body burden exposures to the predicted exposures. Chloroform
was the most critical contaminant, being 85 times above the
criteria.
Table 2.
Predicted Concentrations Downwind from the Lyle Reed Brook
Criteria
CAG
CAG
MEG
MEG
Contaminant
Chloroform
Trichloroelhylene
Dimethyl Sulfide
Methylene Choride
CAG
or
MEG
(ppb)
0.033
0.440
20
200
Off-Site
Concen-
tration
(ppb)
10.0
1.54
3.7
95.0
Long-
Term Body
Burden
Exposure*
2.8
.44
1.04
27.0
Ratio
BB/CAG
or MEG
85.0
1.0
.05
.14
-------
SAMPLING AND MONITORING
65
DISPOSAL OF TOXICS ON
SYLVESTER SITE
LEACHATE
CONTAMINATION
OF GROUNDWATER
GROUNDWATER
MIGRATION
TO
LYLE REED BROOK
VOLATILIZATION
OF
TOXICS
FROM BROOK
SHORT-TERM
AND
LIFETIME
IMPACTS ON
NEARBY RESIDENTS
Critical Parameters
• estimate volume of
contaminants
• range of measured
concentrations of each
contaminant in the
groundwater
• variating groundwater
flow rate
• contaminant flushing
time
• variating concentrations
in groundwater over
time
contaminant evapora-
tion rates, i.e. hourly,
daily, annual due to
temperature, wind speed,
etc.
dilution rate of
contaminant
wind direction
spacial distribution
of exposed population
downwind from the brook
health impact criteria
Figure 1.
A schematic of the basic processes which may cause short- and long-term
ambient air impacts on residents living near the Sylvester Site.
The Nashua case is unique. No actual monitoring was feasible
due to the nature of the problem, or rather the predicted problem,
so no definitive information was available to determine ambient air
concentrations. However, the emission estimation and the
predicted off-site concentrations became a controlling factor in the
particular remedial action alternative chosen, i.e., pump and treat-
ment of the contaminated groundwater.
Lehigh Electric, Wilkes-Barre, Pennsylvania
This case represents another situation where the relevant ambient
air pathway was initially ignored, but where subsequent modelling
analysis showed potential off-site impacts.
The Lehigh Electric site had been an electrical equipment service
and storage facility since the mid-1960s. In 1980, it was found that
some of the site's 3,000 transformers, capacitors, and other electric
apparatus had been leaking polychlorinated biphenyl (PCB) con-
taminated fluids. Subsequent on-site soil sampling and analysis
revealed concentrations of up to 65,000 ppm.
The surface soil at the Lehigh site is dominated by coal fines with
some ash and shale to the one-foot depth. The predominance of
carbon in the fines is ideal for significant PCB adsorption. The
PCB contamination as a fraction of coal particle size is shown in
Table 3. There was not a sufficient sample fraction in the less than
45 n size range to determine PCB contamination, but one would ex-
pect to find similar or higher PCB concentrations because of the
higher surface area to mass ratio in the smaller particulate size
range.
Table 3.
PCB Concentration on Coal Fines at Lehigh Electric
Particle Size (ft) PCB Concentration (ppm)
>105 2500
45 to 105 3700
< 45 Insufficient sample
In the absence of appropriate low level particulate monitoring,
an analysis of the ambient air PCB levels was determined through
the use of a USEPA-approved model.12 This model predicted par-
ticulate concentrations in; the air from wind erosion, which cor-
responded to no on-site activity, and from on-site truck traffic,
which would necessarily result from some remedial activity. This
predicted particulate level was translated into a PCB concentration
conservatively based on the 3,700 ppm PCB-contamination of coal
fines in the mid-range.
The results of this modelling analysis indicated that simple on-
site wind erosion would not yield significant concentrations of
PCB's off-site. This assessment was based on the USEPA CAG
value of 0.0075 ng/m3 of PCB. When dilution and wind direction
were taken into consideration, it was predicted that no off-site
residential exposure would exceed the applicable CAG value.
However, when on-site activity which would include a moderate
number of trucks and other equipment was considered, the model
indicated that a potentially significant off-site concentration of
PCBs might result. This assessment again was based on the CAG
value, and in this case, further modified to account for the short
time frame of the on-site activity compared to a lifetime exposure.
This analysis developed two recommendations for safeguarding
the public health and eliminating future liability. First, an ap-
propriate low level monitoring program should be implemented to
confirm the model's predicted impacts from both a status quo and
a remedial action scenario. Second, any on-site remedial activity
should be accompanied by a comprehensive dust suppression pro-
gram to reduce the possibility of off-site, PCB-laden, particulate
transfer.
CONCLUSIONS
It is evident from the evaluation of these three hazardous waste
sites that a wide range of potential problems may occur in properly
evaluating the off-site risk to the nearby population. These pro-
blems must be dealt with on a case-by-case basis.
Certain common elements, however, should be .included in all
ambient air risk assessments. Initially, an evaluation of the status
quo situation should be made. This is best accomplished through
the use of a well-designed monitoring program. The use of modell-
ing analyses can be used to assist, but should not be relied upon
totally, in most cases. The Lehigh site is an example of a case where
modelling was relied on totally because appropriate background
monitoring was not conducted. The Nashua site is an example
where modelling was the most appropriate analytical path to follow
because of the particular circumstances of the site, i.e., ground-
water contamination with latent surface elution.
The second component of an ambient air risk assessment of a
hazardous site involves the monitoring of the concentrations of
toxic pollutants during on-site activity. This accomplishes two main
objectives. First, it will determine whether any significant concen-
trations are being released during the remedial action from a short-
term health protection point of view. High readings would direct
management personnel to take measures to prevent or reduce this
exposure. Additionally, any ambient air contamination will be
documented, thereby preventing unwarranted liability from future
claims of health impairment.
The final step should include post-remedial action monitoring.
This step will document the actual improvement in air quality due
to the remedial action or, conversely, emphasize the need for fur-
ther investigation and cleanup if levels have not been reduced.
Again, this step is important from both a health and liability point
of view.
All hazardous waste site ambient air evaluations must follow this
basic assessment methodology, although the emphasis will vary as
is evidenced by the three examples presented in this paper. A failure
to follow the system in a thorough fashion may result in:
•Poorly designed site safety considerations
•Excessive population exposures to carcinogens from a long-term
chronic perspective '
•Possible liability cases against the contractor, consultant
USEPA/state/local agencies '
-------
66
SAMPLING AND MONITORING
REFERENCES
1. Crump, K.S., "The Scientific Basis for Health Risk Assessment."
Presented at a seminar sponsored by George Washington University
and by USEPA, Washington, D.C., Mar. 1982.
2. USEPA Cancer Assessment Group. Published in Land Disposal Toxic
Air Emissions Evaluation Guideline. Office of Solid Waste, EPA,
Washington, D.C., Dec. 1980.
3. Research Triangle Institute. "Multimedia Environmental Goals for
Environmental Assessment." Prepared for USEPA, Mar. 1980.
4. American Conference of Governmental and Industrial Hygienists.
Threshold Limit Values for Chemical Substances in Workroom Air,
Cincinnati, OH, 1981.
5. Philadelphia Ad Hoc Committee for Toxic Air Pollutants. Ongoing
committee which is establishing toxic emission guidelines for Phila-
delphia.
6. Michigan Air Pollution Control Commission. "A Proposed Frame-
work for Processing Air Quality Permit Applications for New Emis-
sion Sources of Non-Criteria Pollutants." Final Report to Special
Air Advisory Committee. Nov. 1981.
7. New York State Department of Environmental Conservation. "Air
Guide 1, Application of 6 NYCRR 212—Toxic Air Contaminants."
Revised Dec. 1981.
8. USEPA, "Ambient Air Monitoring of Organic Compounds: Analy-
tical Protocol", Washington, D.C. 1981.
9. Shen, T.T., "Emission Estimation of Hazardous Organic Com-
pounds from Waste Disposal Sites" Presented at 73rd meeting of
the Air Pollution Control Association. Montreal, P.Q., June 1980.
10. Hwang, S.T., "Hazardous Air Emissions from Land Disposal/
Treatment Facilities". Presented at 74th meeting of the Air Pollu-
tion Control Association. Philadelphia, PA, June 1981.
11. Zimmerman, J.R. and Thompson, R.S., "Users Guide for HIWAY:
A Highway Air Pollution Model." USEPA, Washington, D.C. Feb.
1975.
12. Cramer Co., H.E., "Industrial Source Complex (ISC) Dispersion
Model", Developed for USEPA. Dec. 1979.
-------
AIR MONITORING OF HAZARDOUS WASTE SITES
RICHARD W. TOWNSEND
The Bendix Corporation, Environmental & Process Instruments Division
Largo, Florida
INTRODUCTION
Air monitoring of hazardous waste sites is done for numerous
reasons: to identify and quantify specific hazardous contaminants
present in the environment, to determine how waste site workers
are exposed, to investigate community complaints, to determine
if the compliance with various health standards is being followed,
and to evaluate the effectiveness of disposal and storage methods.
The purpose of the sampling will dictate the type of sampling
strategy that will be used. If the waste site is a large dump that cov-
ers many acres of an area then a grid type pattern may be con-
ducted to determine what is present in different areas. This same
strategy can be used if a contaminant was spilled or dispersed over
a large area. A grid pattern is also effective to evaluate the effec-
tiveness of clean-up operations and containment of spills. In this
method of area sampling, the area is marked with tape or by physi-
cal landmarks such as streets or city block areas. These areas are
then systematically numbered, sampled and the results recorded in
the field notebook. Record keeping is an extremely important step
in a successful contaminant sampling program. Once an area is set
up for sampling careful identification of potentially hazardous
substances may be initiated. An understanding of general contam-
inant classification will aid the identification process. Airborne
contaminants can be divided into two major categories: (1) particu-
lates and aerosols, (2) and gasses and vapors. Particulates and
aerosols are particles of matter small enough to be dispersed in the
air, either in a solid or liquid (droplet) form. They can be divided
into subcategories.
Dusts are solid particles which become airborne as the result of
grinding or crushing of solid materials, or through the distur-
bance of bulk masses of powders. Examples of this would be asbes-
tos that has been disposed in a waste site and cyanide dusts.
Other particulates are fumes, smokes and mists. Smokes may be
found at burning landfills or dump sites. Fumes are not normally
present at dump sites since they are dispersions of solid particles
formed by condensation of a vapor. The vapor is formed by heat-
ing a material which is normally a solid at room temperature. These
are most commonly found in welding. Mists are droplets of liquid
substances and could be found in containers that are under pres-
sure, which are leaking.
Of major concern at a waste site is whether or not particu-
lates are respirable or non-respirable. Particulates are classified as
non-respirable if they are either too large (75 microns) or two small
(less than 0.1 micron in diameter). These particles would not
normally be retained in the lungs if they are inhaled.
Gases are substances which are gaseous at normal room tempera-
ture and air pressure. Vapors are substances which are liquid (or
solid) at normal room temperature and air pressure, but which vol-
atize under extreme environmental conditions (as through evapora-
tion). There are currently numerous sources of information which
provided detailed information concerning dangerous contaminant
levels:
1. Industry booklets—such as "The Hazardous Materials Guide"
by Sun Oil Company.
2. "Handbook of Organic Industrial Solvents" by the Alliance of
American Insurers.
3. "Handbook of Hazardous Materials" by the American Mutual
Insurance Alliance.
4. ''Job Health Series'' by OSHA.
5. "NIOSHHealth & Safety Guides."
6. "NIOSH Criteria Documents."
7. ' 'NIOSH Recommended Standards.''
8. "AGGIH— TLVBooklet."
9. "AIHA— Hygienic Guide Series."
PRELIMINARY SURVEY
At a hazardous waste site, it will usually be necessary to con-
duct a preliminary survey. An experienced environmental special-
ist or professional industrial hygienist can, in many cases, eval-
uate quite accurately, the magnitude of chemical hazards asso-
ciated with a waste site without the benefit of any instrumenta-
tion. The first step is to make a careful examination of the site.
It may be possible to screen the area for empty containers which
still have labels on them. This does not guarantee that the con-
tainers actually contain these materials. Are any bulk powders or
dusts present, are drums of materials intact, or are they leaking
due to deterioration of containers? Are there any underground fires
from pockets of materials that are producing smoke?
A preliminary inventory should be made. The list should in-
clude the suspected contaminants, their composition and any by-
products which may be associated with them.
This means that the investigator should obtain complete in-
formation on the composition of the various commercial pro-
ducts. This information may be obtained from the Material Safety
Data Sheets of their manufacturers. Manufacturers of chemical
agents often sell the same substance under several different trade
names, each with slightly different formulas; this can make it diffi-
cult to identify all the chemical components of trademarked
substances. Material Safety Data Sheets (MSDS's) help solve this
problem.
The Material Safety Data Sheet (Form OSHA-20) is divided into
nine (9) sections containing detailed technical information about a
substance, including hazardous ingredients, physical data, fire and
explosion hazard data, health hazard data, reactivity data, spill
or leak procedures, special protection information, special pre-
cautions, and manufacturing info (name and address). 29 CFR
1915, 1916, and 1917, Public Law 85-742 and Public Law 92-596
are the applicable regulations.
During a preliminary walk-through investigation, many poten-
tially hazardous areas can be visually detected. Are large amounts
of dusts present? The senses of sight, smell, are used in this pre-
liminary investigation. The entire effort in this case is to add to
information about a waste site. In some cases, the workers at a
dump site may be able to provide some information if they have
67
-------
68
SAMPLING AND MONITORING
kept any type of records, or if they have noticed any unusual
odors. For many substances the odor threshold concentration is
greater than the permissible safe exposure level. However, many
substances such as hydrogen sulfide produce olfactory fatigue.
Ask workers whether they have experienced any symptoms such
as eye or skin irritation, dizziness, nausea, etc., when working
in specific areas. Try to be as specific as possible, exact symptom
identification can provide important clues about the identity of an
unknown contaminant.
Certain contaminants have distinctive odors which can aid in
their identification. For example, it is possible to distinguish
between the hay-like odor of phosgene and the fish-like odor of
trimethylamine. Many contaminants are odorless, however, and
some may actually act as olfactory anaesthetics as has been prev-
iously mentioned. Hydrogen sulfide, for example, smells distinct-
ly like eggs, but prolonged exposure to it can dull a worker's
sense of smell so severely that he may not notice heavy concen-
trations of other contaminants. The odor threshold of H2S is
0.00047 ppm; the TLV is 10 ppm. A list such as this is prepared by
the American Industrial Hygiene Association Journal and it is use-
ful in the education of workers and industrial hygiene students.
AREA THAT SHOULD BE SAMPLED
There are at least three general locations in which air samples
may be collected: (1) at a specific container within the waste
site, (2) ambient air, or (3) in a worker's breathing zone. The
choice of the location is dictated by the type of information needed,
and often the use of all three methods will be necessary to provide
detailed information.
Most frequently, the sampling is performed to determine the
level of contamination present in a specific container. In this in-
stance a grab sample or instantaneous concentration is all that is
needed. A portable instrument or detector tube may be used. If
a worker's exposure is needed, it may be necessary to collect
samples in the worker's breathing zone. If area samples are needed,
then they should be collected using area monitors.
SAMPLING DURATION
The sampling time is usually the minimum sampling time neces-
sary to obtain an amount of material sufficient for accurate
analysis. The duration of the sampling period is therefore based on
the following considerations: the sensitivity of the analytical pro-
cedure, the permissible exposure limit or TLV of the particular
substance and the anticipated concentration of the contaminant
in the air being sampled. The sampling period should represent
some period of time if relative to worker exposure; i.e., the amount
of time that the worker spends at the waste site or particular grid
area for one day. It is desirable in this case to sample the work-
er's breathing zone for his full work shift at the waste site. This
is important if sampling is being conducted to determine com-
pliance status relative to OSHA standards.
Evaluation of a worker's daily-time weighted average exposure
is best accomplished by personal sampling. For personal sampling,
the sampling train consisting of a pump, drawing air through a
collection medium, which separates and collects contaminants for
later laboratory analysis. For personal sampling, the sampling train
assembly is attached to the individual exposed, with its collec-
tion medium positioned so that the air sampled is collected from
the worker's breathing zone. For area sampling, the sampling train
assembly is placed at a fixed point.
The ceiling exposure and time-weighted average must both be
considered when planning a sampling program. The ceiling ex-
posure is determined by taking a 16 minute sample at the place of
heaviest anticipated contaminant concentration.
The choice of a particular sampling instrument depends on a
number of factors which include portability and ease of usage,
efficiency of the instrument or sampling method, reliability, avail-
ability of the instrument, and whether instantaneous readings are
desired, and cost. Presently there is no one universal sampling in-
strument or method in use today.
One of the most useful type of instruments for evaluating haz-
ardous spills and unknown contaminants at hazardous waste
sites is the use of direct reading colorimetric detector tubes.
DETECTOR TUBES
The advantages of having direct reading detector tubes are
rather obvious. On-site evaluations of atmospheric concentrations
of hazardous substances may be made immediately. The size of a
detector tube system is very light and highly portable, typical
systems may weigh under 36 ounces. Direct reading tubes are avail-
able for over 200 different substances, and many of the tubes may
be used to screen for groups of materials such as the use of an
amines detector tube which will detect any one of 30 different
amines.
The use of detector tubes permits rapid estimation of the con-
centration of a contaminant, permitting on site evaluations and im-
plementation of immediate corrective measures. The use of these
tubes may save time and expense of laboratory analyses and may
be used as evidence in court proceedings. The cost of detector
tubes is not high considering comparable methods. Because they
are so lightweight it is easy to carry a shoulder bag with 40 or 50
boxes of detector tubes to screen for a variety of substances. If
the investigator wants to make his own field kit 4 or 5 tubes for
100 different substances can be carried.
Detector tubes are glass tubes that are packed with a chemically
treated substance (usually silica gel, or alumina) which gives an
immediate, direct reading of a contaminant level. In colorimetric
tubes, the degree of concentration of the contaminant causes the
hue of the tubes chemically treated substance to take on varying
intensities. The exposed table must be compared to a color stand-
ard or chart.
In length of stain tubes, the substance develops a stain the
length of which is proportional to the contaminant level.
One should always read the complete instructions found in each
box of detector tubes. Some tubes require humidity, altitude, and
temperature compensation factors for accurate readings. These
special requirements will be described in the instructions.
Detector tubes should be stored in a refrigerator to prolong
their shelf life, but must always be allowed to reach room tempera-
ture before being used. These tubes are marked with an expira-
tion date which should always be noted.
Gas
Table 1.
Gas Detector Tube Unit Certification Compounds 11/20/81
Certified Range (ppm)
Acetone
Ammonia
Benzene
Carbon Dioxide
Carbon Disulfide
Carbon Monoxide (Range A)
Carbon Monoxide (Range B)
Carbon Tetrachloride
Chlorine
Ethyl Benzene
Ethylene Dichloride
Formaldehyde
Hexane
Hydrogen Chloride
Hydrogen Cyanide
Hydrogen Sulfide (Range A)
Hydrogen Sulfide (Range B)
Methyl Bromide
Methylene Chloride
Nitric Oxide
Nitrogen Dioxide
Perchloroethylene
Styrene
Sulfur Dioxide
Toluene
Trichloroethylene
500 to 5000
25 to 250
5 to 50
2500 to 25000
10 to 100
25 to 250
250 to 2500
5 to 50
0.5 to 50
50 to 500
25 to 250
ItolO
250 to 2500
2.5 to 25
5 to 50
5 to 50
50 to 500
7.5 to 75
250 to 2500
12.5 to 125
2.5 to 25
SOtoSOO
50 to 500
2.5 to 25
SOtoSOO
SOtoSOO
-------
SAMPLING AND MONITORING
69
While it is true that the operating procedures for colorimetric
indicator tubes are simple, rapid and convenient, there are dis-
tinct limitations and potential errors inherent in this method of
assessing concentrations of toxic gases and vapors.
They may lead to dangerously misleading results unless they
are used by an adequately trained person who: (1) enforces
periodic testing and calibration of individual batches of each spe-
cific type of tube for its response against known concentrations
of contaminants as well as refrigeration of the tubes, (2) under-
stands the physical and chemical interferences associated with their
use, (3) uses other independent sampling and analytical procedures
to back up results. The American Industrial Hygiene Association
publishes an excellent manual entitled "Direct Reading Colori-
metric Indicator Tubes Manual." This manual described in detail a
recommended practice for the use of colorimetric indicator tubes,
and their limitations.
Procedures for Use
To use a detector tube, its tips are broken off, the tube placed in
the manufacturer's pump and the recommended volume of air is
drawn through the tube. Each air moving pump or device must be
calibrated frequently.
The pump can be calibrated using a standard bubble buret and
by performing a soap bubble test. This test uses a soap film which
is drawn by the air moving device. The distance the soap bubble
travels may be measured in cc's marked on the buret. An accep-
table pump should be correct to within ±5% by volume. The
pump should also be checked for valve leakage and inlet plug-
ging frequently. A calibration sticker should be placed on the
pump.
An accessory which may be used with detector tubes includes
extension hoses where the detector tube is placed at the end of a
length of hose. This permits sampling in holes, tunnels, or in areas
where contaminants may have entered such as a manhole or under-
ground tunnel. If hot gases are to be checked such as those com-
ing from underground landfill fires, a hot probe may be used to
cool the sampled air, prior to its entry into the detector tube.
Examples of detector tubes utilization have been numerous
throughout the past 20 years. Some of these uses were in Florida
to check empty drums of chemicals found dumped beside the
highway. These drums were suspected of containing hydrogen cy-
anide and acetic acid. Detector tubes were used to verify what the
labels on the container listed. This method was rapid and allowed
for immediate protective clothing to be worn when handling the
drums containing the cyanide. The area was also checked for spill-
age. Soil samples would, in this instance, also be submitted to a
laboratory for analysis. Detector tubes may also be used to check
burning materials at dump sites which in many cases contain haz-
ardous substances. Detector tubes are available for over 200 sub-
stances. Table 1 contains a list of those tubes which are certified
by the National Institute for Occupational Safety and Health.
There are numerous analytical methods for monitoring haz-
ardous waste sites. Each should be used, depending upon the local
situation. The primary purpose of this paper has been to make in-
vestigators of hazardous waste sites aware that detector tubes do
offer a quick and reliable method for screening. While they are
far from perfect they do provide a useful method to assist in the
evaluation of uncontrolled hazardous waste sites.
REFERENCES
1. TLV® Booklet", American Conference of Governmental Industrial
Hygienists, Cincinnati, Oh.
2. "Hazardous Material Guide", Sun Oil Company, Philadelphia, Pa.
3. "Handbook of Hazardous Materials", American Mutual Insurance
Alliance, Chicago, 111.
4. Clayton, G.D. and Clayton, F.E., "Patty's Industrial Hygiene and
Toxicology", 3rd ed., John Wiley & Sons, New York, NY, 1978.
5. "The Industrial Environment—its Evaluation and Control", NIOSH.
-------
EMISSION MONITORING OF HAZARDOUS WASTE SITES
LOUIS J. TfflBODEAUX
CHARLES SPRINGER
PHILLIP LUNNEY
Department of Chemical Engineering
University of Arkansas
Fayetteville, Arkansas
STEPHEN C. JAMES
Solid and Hazardous Waste Research Division
U.S. Environmental Protection Agency
Cincinnati, Ohio
THOMAS T. SHEN
New York State Department of Environmental Conservation
Albany, New York
INTRODUCTION
Volatile chemicals in surface impoundments can readily escape
into the air and result in an area-source of emissions. Surface im-
poundments serve as aqueous waste storage basins, pretreatment
basins and treatment basins. In all cases the natural water surface
area plus any mechanically generated interfacial area is in contact
with the atmosphere, thus providing a direct pathway for dissolved
chemical species vaporization.
Means of assessing the air emissions from such wastewater
sources are needed. The assessment methods should have a two-
fold approach: 1) field techniques for monitoring the air emissions
from measurements taken at or near the site and, 2) verified
mathematical models with a predictive capability that allow ac-
curate estimation of the emissions based upon the chemicals pre-
sent in the impoundment. The latter approach is needed by permit
writers while the proposed impoundment is in the planning and
design stages.
The work presented here is primarily concerned with chemical
emissions from two surface impoundments of a hazardous waste
disposal facility. The identity of the volatile chemicals involved, the
concentrations in air and water, and measurements of the flux rate
of emission are reported. Emission calculations based on the
mathematical model are also presented. Comparison of the
measured and model predicted emission rates are presented and
discussed.
CONCENTRATION PROFILE TECHNIQUE
A field method of measuring volatile chemical emission rates
from area sources has been developed.' The method, the so-called
concentration profile (CP) technique, employs profiles of chemical
concentrations, temperature and wind speed to quantify the
chemical emission rate. One significant feature of the method is
that it does not disturb the natural transport processes in effect on
the water surface. The method is limited with respect as to sample
time, wind speed, fetch-to-height ratio of upwind disturbances, etc.
These limitations, along with the theory of operation, are con-
tained in the above referenced document. Only the final working
algorithm of the CP technique will be presented here.
Only field measurements of the concentration of the chemical in
air- PM (g/L): wind speed, vx (cm/s), and temperature Tb (°K),
within the turbulent boundary layer (no more than two to three
meters above the water surface) are made. Six observations of each
parameter, distributed in a logarithmic fashion from the water sur-
face and taken well downwind of any wind velocity disturbances,
should be made. Based on these measurements, the following equa-
tion is used to determine the vertical flux of the chemical species
(i.e., species B) from the water surface:
nB = - (DB1/DAI)2
cA1
«>
(1)
where nB is the chemical flux rate (ng/cm2-s), S/o is the slope of a
line from a graphical plot (or linear regression) oip A1 vs. In y,
where y is the sample height above the water surface in cm.
If S,^ is negative then chemical emission is occurring from the
source whereas if S p is positive, then the water surface is absorb-
ing the chemical species. Sv is the slope of a line from a similar rela-
tion between vx vs. In y. This slope should always be positive and
have units of cm/s. The combination of the units of S/? in g/1 and
Sv in cm/s is the units of flux in ng/cm2 • s. k = 0.4 is the von Kar-
man constant. Since the air sampling is done close to the water sur-
face, only the vertical flux is needed to assess the emission rate.
Note that Sv is related to the friction velocity at the surface, V*, by
Sv = VVk.
The chemical concentration and velocity vs. In-height profiles
should be linear over the full range of the six observation heights
only if the air boundary layer is neutral (i.e., essentially
isothermal). The profiles will likely be non-linear under stable and
unstabale micrometeorological conditions and thus display some
curvature. In such a case, the slopes of tangent lines drawn to the
profiles in the boundary layer region nearest the water surface are
used for S/J and Sv.
The stability-turbulent Schmidt number correction factor (i.e.,
<£2m ScA1(1)) in Equation 1 is obtained from the following empirical
equations developed for water (i.e., species A) vapor:
0m!ScA1 = (1 + 50 Ri)-
for stable conditions and
Ri
4>m2ScA1 = (1 - 50 Ri)*
for unstable conditions and
number defined by:
Ri = g(Ti2 -T"u)(y2 -
Ri
(2a)
(2b)
< 4.0. Ri is the Richardson
- Vxl)'T,
(3)
where g = 980.7 cm/s!, T12, Tn, v^, and vxl and dry bulb air
temperatures and wind speeds at heights y2 and yt above the sur-
face. T! is the average of the temperature. Ri is dimensionless and if
Ri <0, then the air is unstable, whereas if Ri = 0, the air is neutral
or if Rb" 0, the air is stable. The 2/3 power ratio of .molecular dif-
fusivities of the chemical of interest to that of water vapor corrects
the equation for the chemical for which the flux is desired.
The three equations presented above are the working formulas of
the CP technique. Claith, el a/.,2 employing a similar field techni-
que referred to as the aerodynamic technique, report measurements
of the vapor flux rate of S-ethyl N.N-dipropylthiocarbamate
(EPTC) from irrigation water in a flooded alfalfa field at Brawley,
70
-------
SAMPLING AND MONITORING
71
CA.2>3 EPTC was applied at a rate of 3.0 kg/ha at an average con-
centration of 2.17 ppm to irrigation water. The process is called
herbigation. An equivalent of 13 cm (i.e., depth of water) of irriga-
tion water was applied to the 2 ha field surface. The alfalfa plants
were approximately 15 to 35 cm high.
Beginning at 0840h on the day of application, wind speed, water
samples and air samples were taken. The EPTC emission rate was
measured from equipment similar to the CP technique located near
the center of the field. Table 1 contains pertinent field data col-
lected during the flooded period and Table 2 contains the chemical
and physical data on EPTC for this well documented field test.
Water temperature during the test apparently was not recorded.
Table 1.
EPTC Field Data (3)
Time
(h)
0840
1300
1430
1450
1500
1600
1630
1711
1800
1830
1915
2000
2100
2200
2300
2400
0100
•In head ditch.
tin tailwater.
chemical name
molecular weight
diffusivity in air
vapor pressure
solubility in water
Henry's constant
Wind Speed '
1 m
V»
(m/s)
1.6
1.0
3.0
2.5
1.9
1.6
Cone, in Water
p A2
(ppm)
2.14*
2.30*
2.08*
1.92*
1.97f
1.76t
1.44f
Flux, nR
(g/ha«h)
35
140
260
65
50
Table 2.
EPTC Properties at 30 °C (2)
S-ethyl N,N-dipropylthiocarbamate
189.3
.0582 cmVs
29.7E-emmHg
320 ppm
7.29E-3(g cm-3 in air/g cm-3 in water)
MODELS FOR EMISSION RATE PREDICTION
It is highly desirable to have available mathematical models for
predicting, in an a priori sense, chemical emission rates from ex-
isting or planned surface impoundments based on chemical content
of the water and known environmental parameters, such as wind
speed and temperature. Such models have been proposed and are in
various stages of verification. The following review outlines the
present status of the model and the extent of verification.
The basic ideas of using the two resistance theory for interphase
mass-transfer of volatile chemicals in waste water basins were
presented by Thibodeaux and Parker in 1974." The model was
modified by Freeman5 and successfully tested in a laboratory reac-
tor/stripper with acrylonitrile by Freeman and Schroy.6 Recently,
Hwang7 and Shen" have summarized and simplified the elements of
the emission model so that it can be applied easily to practical
engineering problems. The reader should refer to the above cited
works for details of the model.
The model has been field tested for methanol emissions from
four aerated basins treating pulp and paper.' Methanol is gas-phase
controlled, and this aspect of the model has been verified with field
measurements using the CP technique. This paper extends the field
test verification of the model to include chemicals that are liquid-
phase controlled.
ANALYSIS OF THE EPTC DATA
As noted above, except for the lack of an exact water
temperature, the field data collected and documented by Claith, et
a/.,2'3 provides sufficient information for testing the emission
model. The gas and liquid phase coefficients as a function of wind
speed were estimated based on the compiled literature information
available on water vapor and benzene.' The individual coefficients
were then transformed for EPTC using Graham's law of inverse
square root of molecular weights. The two resistance theory was us-
ed to obtain the overall coefficient, 'K^ and the flux rate was
calculated by:
nB = 1001
(4)
with !KA2 in cm/h, n A2, the concentration in water, in mg/L and
NB in g/ha»h,
1/1KA2 = l/ikA2 + l/H2kA, (5)
with H being Henry's constant.
The calculations appear in Table 3. The individual gas-phase and
liquid phase coefficients, 2kAl and lkA2, respectively, are
presented in columns one and two. Comparison of the overall coef-
ficients in column three with those in columns one and two
demonstrate that EPTC is dominated by processes on the liquid-
phase side of the interface. The water temperature was assumed to
be 30 °C. The model calculated EPTC flux is compared to the
measured flux in Fig. 1 . In all cases, the calculated flux exceeds the
measured flux. However, considering the uncertainties in
temperature, and that the mass-transfer coefficients used to not ac-
count for plant stems protruding through the air-water interface,
the model does provide a reasonable estimate of the volatilization
process. Plants in the water would reduce the effective transport
coefficients, thus resulting in low estimations of the flux.
400
300
EMISSION
FLUX
(g/ha hr)
200
100
T
0
CALC.
MEAS.
-o
13GO2DOO22002400
HOURS OF MAY 25, 1977
Figure 1.
Volatilization of S-Ethyl, N.N-Dipropylthiocarbamate
from a Flooded Alfalfa Field
-------
72
SAMPLING AND MONITORING
Table 3.
Calculated Emission Coefficients and Rates for EPTC
Gu-pfcnUqahl-phsie
coefl. co»tr.
time
(kl
1300
1600
1800
2000
2200
2JOO
2400
IfcAi
Icro/h)
419
247
765
666
493
469
440
l|Aj
(cm /hi
083
0.83
2.79
2.14
0.83
0.83
0.83
Overall
cotff.
(cm h)
0.65
0.57
1.86
1.49
0.67
0.67
0.66
Cone, in
water
PM
(mg/L)
2.1
2.0
2.0
1.76
1.44
1.44
1.44
Evaporative
flux
OB
(g/hafh)
137
114
372
261
97.1
96.2
95.00
DESCRIPTION OF HAZARDOUS WASTE
FACILITY SURFACE IMPOUNDMENTS
In Aug. 1981, the CP technique was used to measure the emis-
sion of volatile organic chemicals from two surface impoundments
in Western New York State. These impoundments routinely
receive, store and treat aqueous waste delivered by tanker trucks.
In July 1981, the facility received 720mJ of aqueous waste,
neutralized 1790m' gal and processed 7900m1 through the water
treatment system. A typical waste liquid analysis appears in Table
4.
Table 4.
Waste Liquid Analysis
Halogenated organics
Non-halogenated organics
Organic salts
Organic acids
Metals
Metal salts
Inorganics
Water
Weight %
0.15
1.15
0.87
0.07
1.24
5.20
4.00
87.32
The aqueous waste treatment system includes chemical reduction,
chemical oxidation, neutralization, activated carbon and biological
oxidation treatment. Depending upon the nature of the aqueous
waste, selected operations are chosen for treatment. All operations
are batchwise. The aqueous treatment process is only a portion of
the operations at the facility, which includes a hazardous waste
landfill, drum recovery and solvent recovery operations.
Characteristics of the two surface impoundments studied are
listed in Table 5. Pond 1, 2 is an earthen basin with a 1.2 to 1.5 m
berm. The plan dimensions of the pond are 85m wide and 253 m
long. A partial dike separates the rectangular shaped pond into two
sections. A 2m opening, however, allows free exchange of water
between the two sections. Section 1 contains 50,000m! and section 2
contains 18,400m'.
Pond 1,2 is a wastewater biochemical oxidation reactor. Surface
aerators maintain the oxygen content to keep the process aerobic.
Aerobic conditions seemed to be present during the sample period.
The water was light green in color, suggesting algae growth was oc-
curring. Foam generated by the surface aerators was also present.
Table 5.
Surface Impoundment Data
Characteristic Pond 1,2 Pond 6
Surface area, nV 21,600 4,650
Water volume, m' 68,400 4,630
Water depth, m 3.2 1.0
Shape rectangle rectangle
No. of surface aerators operating 8 0
Aerator power, hp per aerator 15
The non-persistent foam covered 2 to 5% of the surface. A drum
storage and drum cleaning operation was situated directly north of
Pond 1,2.
Pond 6 is a temporary storage and pre-treatment pond. This
pond is rectangular in shape with a 1,8m berm and a plastic liner on
bottom. The water was yellowish in color and more odorous than
pond 1,2.
SAMPLING AND ANALYSIS METHOD
The CP technique apparatus consisted of devices installed onto a
boat for the simultaneous measurement of dry-bulb temperature
and wind speed and obtaining air constituent samples. Placement
of the apparatus on a boat allows facing the devices into the wind
and proper location on the impoundment surface. The apparatus
placement above the water surface is shown in Fig. 2. Temperature
measurements were taken at 39, 52, 69, 93, 142, and 211 cm above
the water surface. Wind speed was measured at 47, 67, 107, 187,
267 and 347 cm. One liter of air drawn through adsorbent tubes
over a 15 min sample time were positioned at 7.0, 14, 24, 55, 104,
and 232 cm. A weather vane, atop the wind speed mast, allowed
continuous indication of wind direction.
The micrometeorological mast was manufactured by C.W.
Thornthwaite Associates (wind profile register system model 106).
The temperature and air sampling mast were designed and built at
the University of Arkansas. The temperature mast consisted of
open-ended cylindrical tubes that (radiation) shielded a clay-in-cup
sensor. An electronic tip-thermocouple probe with a 2 sec response
time and LED digital display was inserted into each clay sensor for
temperature measurements. The sample tube mast consisted of six
tube height positioners, six extension tubes and six air rotameters
connected to a hand evacuated vacuum tank.
Air samplers were 15 cm stainless steel, 3.2mm O.D. and 2.2mm
I.D. tubes. Each tube was packed with 11.5 cm of Tenax, 60/80
mesh and 2.5 cm of silica gel, 40/60 mesh. Air flow was through the
Tenax then through the silica gel. The prepared tubes were pre-
conditioned by heating to 270-300 °C and purged with helium for
one hour. The tubes had Swagelok plugs on each end and were
Figure 2.
Concentration Profile Apparatus
-------
SAMPLING AND MONITORING
73
sealed prior to and after sampling. Surface water samples were ob-
tained from each impoundment in one liter, amber teflon-lined cap
bottles. Each sample was acidified with HC1 immediately to sup-
press biological activity.
A Varian 3700 Gas Chromatograph was used for all flame
ionization detection (FID) analysis. Water samples were analyzed
on Chromosorb 102, Chromosorb 101, and Bentonone 34 using
various flow rates and temperature programs to obtain optimum
separation. Chromosorb 1(M was best with temperature held at
45°C for 6 min during trap desorption (@180°C), elevated to
125 °C and held for 6 min, then temperature programmed at
4°/min up to 200°C. With a flow of 53 ml/min Nitrogen carrier,
the Chromosorb 101 produced the greatest number of well resolved
peaks. The resulting peaks were compared to standards of
purgeable priority pollutants. The use of three different standards
and spikes of these aided in identification. The results were
reported at total FID organics in iig as Benzene. The individual
compounds were reported in us individual compound. Water sam-
ple results were reported in /xg/1.
A Varian 2800 gas chromatograph equipped with a Hall detector
was used for chlorinated organics. The Carbopac B packing served
well for the Hall detector analysis. The temperature was held at
35 °C for 8 min during trap desorption (at 180°C), then elevated to
220 °C at 8 Vmin and held for a total of 52 min. Samples were again
compared to three purgeable priority pollutant standards for iden-
tification and quantification. Chlorinated organics were reported
as total chlorinated organics in fig as chloroform. Individual com-
pounds were reported in /ig individual compound. Water sample
results were reported in jig/1.
Gas Chromatographic analysis of air and water samples was per-
formed at the University of Cincinnati.10 Two water samples were
analyzed on GC/MS by Environmental Consultants, Inc. of Cin-
cinnati." Air sample tubes and water samples were provided to the
site operator. GC/MS analysis were performed on the air samples
and TC, TOC and TOD on the water samples in the operators
laboratory.
FIELD MEASUREMENT RESULTS
AND DISCUSSION
A total of 90 air samples and 12 water samples were obtained
from the two impoundments. Fifteen concentration profiles were
obtained, nine above pond 1, 2 and six above pond 6. Temperature
and wind speed profiles were also obtained. Nine profile tube-sets
(i.e., six tubes make a set) were analyzed by GC/FID and six by
GC/Hall detector. Four profile tube-sets and the corresponding
water samples were delivered to the site operator for split-sample
analysis.' Remaining tube-sets and water samples were analyzed by
the University of Cincinnati, Department of Civil and
Environmental Engineering. Selected water samples were also
analyzed by Environmental Consultants, Inc."
The major chemicals (identified and confirmed) and concentra-
tions in the water phase of the two surface impoundments are listed
in Table 6. These concentrations are a summary of the results of the
three laboratories. Approximately 50 other compounds were iden-
tified but were not confirmed with standards. The chemical make-
up of the water is consistent with tanker truck waste liquid
discharged to the impoundments as reported in Table 4. In general
the concentration levels in pond 1,2 are lower than in pond 6. This
is likely due to "treatment" occurring in pond 1,2 since it was
designed for biochemical oxidation. Pond 6 serves as a storage
pond and biological treatment is not encouraged.
Several chemicals were identified in the air boundary layer above
both surface impoundments; the ranges of concentrations detected
near the water surface and at two meters above the surface of both
ponds are shown in Table 7. The concentration gradients suggest
that the water is a source of the air contaminants. The detailed pro-
file structure for benzene, 1,1, dichlorethane and total hydrocar-
bon concentration in the air boundary layer above pond 6 is shown
in Fig. 3. Such a linear relationship of decreasing concentration
with natural logarithm of height above the water indicates that the
water is an area source and the boundary layer is turbulent.1
Jable 6.
Major Chemicals in Water of Surface Impoundments
Concentration and s.d. (/ig/l)
0.31
4.1
8.4
34.
207.
235,000.
73,000.
64,000.
±
±
16.
43.
125.
22. ±
267.
124.
8.3
144.
6.8
33.2
3.3
10.3
9.0
6,300,000.
2,150,000.
2,000,000.
0.31
4.7
2.8
7.7
37.
10,000.
2,300.
0.0
9.5
3.0
19.
Chemical**
POND 1,2
Benzene
Toluene
Total hydrocarbon*
1,1 dichloroethane
Total chlorinated hydrocarbon D
Total oxygen demand
Total carbon
Total organic carbon
POND 6
Benzene
Toluene
Total hydrocarbon*
1,1 dichloroethane
Total chlorinated hydrocarbon D
Methylene chloride
Chloroform
1,2 dichloroethane
1,1,1 trichloroethane
Trichloroethene
Tetrachloroethene
Chlorobenzene
Ethylbenzene
Total oxygen demand
Total carbon
Total organic carbon
**GC/MS results, confirmed with standards.
•Flame ionization detector, reported as benzene.
DHall detector, reported as chloroform.
—indicates only single values available.
Table 7.
Chemicals and Concentration Ranges Above Surface Impoundments
Chemical Concentration in"Air Sampler (/ig/l)*
Benzene
Toluene
Total Hydrocarbon as
Benzene
Ethylbenzene
1,1 Dichlorethane
Total Chlorinated Hydro-
carbons as CHC13
Near Water
Surface
0.2
0.04
2.0
0.01
0.07
0.5
•Results of University of Cincinnati and site operator laboratories.
Two meters above
Surface
0.01
0.002
0.04
0.002
0.02
0.2
Hydrocarbon
1.0
.05 0.1 .15 1.0
£A1, Concentration In Air'.^g/l
Figure 3.
Concentration Profiles for Pond 6
1.5
2.0
-------
74
SAMPLING AND MONITORING
Table 8.
Profile Measurement Summary
Profile
no.
1
1
2
2
3
3
4
4
5
5
648
648
7
7
9
9
10
10
11,13,14
11,13,14
11,13,14
12,15
12,15
3T
t
Pond
no.
2
2
2
2
2
2
2
2
2
2
1&2
H2
1
1
1
1
6
6
6
6
6
6
6
Chemical
n
Flux* Corr. Friction
(obs) (ng/cm2-s) Coeff.
Benzene
Total HC
Benzene
Toluene
1,1 DCE
Total CLH
1,1 DCE
Total CtH
Benzene
Toluene
1,1 DCE
Total CLH
Total HC
Benzene
Benzene
Toluene
Benzene
Toluene
Benzene
Toluene
Total HC
Total CLH
1,1 OCE
Positive denotes
6
6
5
4
6
6
5
5
5
4
6
6
6
6
5
5
4 +
5
5
6
6
3
5
emission
UC denotes University of
+0.060
+1.05
-0.38 +
+0.015
—
-0.21 +
-0.022 +
-0.33 +
-0.54 +
-0.019 +
+0.0404
+0.153
-1.02 +
-0.290 +
-0.30 +
-0.038 +
58.0
+3.6
+0.095
+0.014
+1.30
+0.28
+0.028
and negative
.878
.545
.176
.426
—
.390
.713
.783
.376
.678
.764
.412
.763
.711
.593
.685
.776
.687
.980
.222
.958
.998
.990
denotes
Cincinnati Laboratory
Vel.
(cra/s)
20.4
20.4
11.5
11.5
11.9
11.9
8.74
8.74
19.9
19.9
22.0
22.0
22.2
22.2
21.2
21.2
8.9
8.9
12.0
12.0
12.0
11.0
11.0
Sv
Corr.
Coeff.
+ .996
+ .996
+ .998
+ .998
+ .987
+ .987
+ .986
+ .986
+ .979
+ .979
+ .993
+.993
+ .993
+ .993
+ .993
+ .993
+ .985
+ .985
+ .985
+ .985
+ .985
+ .983
+ .983
Vx
W1nd@3.5m
(cm/s)
282
282
196
196
191
191
151
151
232
232
360
360
335
335
352
352
159
159
325
325
325
275
275
absorption.
and SO
denotes
the site
Comments Lab.t
Aug. 26, 10:45AM, sunny and clear, wind
variable from S, SE, SW.
Wind SSW, changing to S. No wind for
brief periods.
No chemical detected to all six levels.
Wind very low, direction variable from
S to E. No wind for brief periods.
3:OOPH, wind shift to north. Good
steady wind.
North wind, good and steady.
North wind, good and steady.
North wind, good and steady.
North wind, good and steady.
North wind, good and steady.
North wind, good and steady.
Aug. 27, 12:12PM Chemical analysis
faulty.
Wind N, NE with drizzle rain.
Wind N, NE with drizzle rain.
Wind N, NE with drizzle rain.
Wind N, NE with drizzle rain.
Wind changing to east at 2: 35PM.
operator laboratory.
UC
UC
SO
SO
UC
UC
UC
UC
SO
SO
UC
UC
UC
UC
SO
so
so
so
UC
UC
UC
UC
UC
A data summary of the CP technique measurements appears in
Table 8. This table contains: the profile number (Col. 1), the pond
number ( Col. 2), the chemical (Col. 3), the number of usable data
in the profile (Col. 4), the flux rate of the chemical in the boundary
layer (Col. 5), the correlation coefficient for S/o (Col. 6), the fric-
tion velocity at the surface (Col. 7), the correlation coefficient for
Sv (Col. 8), wind speed ( Col. 9), comments (Col. 10), and the
laboratory which performed the analysis (Col. 11). This data repre-
sent field results for Aug. 26 and 27, 1981. Pond 1,2 was surveyed
the 26th and pond 6 was surveyed the 27th.
The flux measurements presented in Table 8 need to be inter-
preted in light of the operation of the ponds and other local sources
of volatiles. Both positive (emission) and negative (absorption)
fluxes were observed. This occurred for pond 1,2 only. It appears
that volatile chemical concentrations were low in the water of pond
1,2 on Aug. 26. Low levels are the result of biological treatment
and/or volatilization immediately after placement of a batch of
wastewater in the pond. There are other sources of VOC at the
facility. A landfill area is located to the fast and northeast of pond
1,2. Waste receiving, process area, solvent recovery areas as well as
pond 6 were also located northeast of pond 1,2. A site containing
waste in drums, covering approximately two acres, was located
about 90 m north of pond 1,2.
The emission and absorption measurements values in Table 8 are
the result of changing winds. Apparently, winds from the north,
northeast and east caused absorption, whereas winds from the
south, southwest and west caused emission. Strong organic
chemical odors were evident when the wind was from the north.
Drum clean-out was occurring at the time and profiles 7 and 9 in-
dicate hydrocarbon absorption. When wind was from the south the
background concentration of organics in the air was low. If a shift
to (he north occurred at sometime within the 15 min sample period
air with high background concentration of organics swept over the
tubes. The net result was very erratic chemical concentration pro-
files and low correlation coefficients.
The conditions of the next day (Aug. 27) were ideal for emission
measurements on pond 6. Winds were constant and always from
the north and northeast with no apparent upwind sources of VOC.
These conditions resulted in high correlation coefficients for pro-
files 10 through 15. The chemical analysis performed on tubes of
profile 10 appears to be faulty. Nevertheless emission is indicated.
Profiles 11 through 15 were averaged resulting in high correlation
coefficients. Averaging is appropriate due to constant
micrometeorological conditions throughout the period. Table 10
contains a summary of the ranges of the chemical flux rates observ-
ed above the ponds. Except for 1,1 dichloroethane, the emission
rates from pond 6 were greater than pond 1,2. This is to be ex-
pected, since the concentration of VOCs in pond 6 is greater.
Table 9.
Surface Impoundment Transport Coefficient Calculations
Chemical
benzene
1,1 OCE
benzene
1,1 DCE
Pond Temp. Henry's Transport Coefficients (cm/h) tj
°C Constant Hq. gas Hq. gas overall half-
coef. coef. coef. coef. life
(gcnr^/gcnrJ) 'k^ ^Al kA2 kAl *A2
1,2 24
1,2 24
6 21
6 21
.232
.191
.215
.193
3670
3810
148,000
199,000
31.4
18.7
11.4
10.0
792
981
741
909
59.0
50.8
10.6
9.5
3.7
4.3
6.5
7.3
•Evaporation half-life 11/2 - .69 x depth + 'KA2
Table 10.
Comparison of Calculated and Measured Flux Rates
Chemical
Benzene
Toluene
Total HC
1,1 DCE
Total CLH
Benzene
Toluene
Total HC
1.1 DCE
Total CLH
Pond
1,2
1,2
1,2
1,2
1,2
6
6
6
6
6
Calculated Flux
dig/cm'*!1
+ 0.0051
+ 0.062
+ 0.14
+ 0.48
+ 2.7
+ 0.047
+ 0.12
+ 0.37
+ 0.058
+ 0.064
Measured Flux
(ng/cm'«»)
-0.29 to +0.06
-0.038 to +0.015*
-1.0 to +1.1 *
-0.02210 +0.04
-0.33 to +0.15 •
+ 0.095
+ 0.014 •
+ 1.3
+ 0.028
+ 0.28
•denoted low correlation coefficient pn concentration profile slope.
Based upon the observed concentrations of chemicals in the
ponds (Table 6) and suggested modeling guidelines for air emis-
sions,' flux rates were calculated. The method of calculation is
basically outlined by Equations 4 and 5. This calculation was per-
-------
SAMPLING AND MONITORING
75
formed as part of a continuing effort to "field test" proposed emis-
sion models. Table 9 contains calculated transport parameters and
other calculated results for benzene and 1,1 dichloroethane. Pond
1,2 contained eight, 15 hp (operating) surface aerators at the time
the sampling was performed. The effective liquid-phase and gas-
phase coefficients for the aerators appear in columns 5 and 6 of
Table 9. The natural, wind-induced coefficients appear in columns
7 and 8. Pond 6 had no operating aerators at the time. The
calculated overall transport coefficients are in column 9. Inspection
of the coefficients suggest that benzene and 1,1 dichloroethane are
liquid phase controlled. The overall coefficient in pond 1,2 is five
times higher than that of pond 6 due to the surface aerators; see
1KA2 values in next to last column Table 9. Emission half-life
calculation result appears in the last column of Table 9. The times
suggest fairly rapid volatilization rates are operative in the ponds.
A comparison of the (CP technique) measured VOC emission
rates and model calculated emission rates appear in Table 10. For
the flux calculations of total HC (hydrocarbon), benzene properties
were used and for total CLH (chlorinated hydrocarbon), 1,1
dichloroethane properties were used. In cases where the measure-
ment statistics of correlation are high the model is in general agree-
ment with the measured values. The range of measured emission
rates (with high correlation coefficients) were from 0.028 to 1.3
ng/cm2»s. The calculated emission rates ranged from 0.0051 to 27
ng/cm2»s. Inspection of the measured vs calculated rates in Table
10 for pond 6 indicates agreement within an average factor of ±
2.5. The agreement is + 2.0 if only the data for 1,1 dichloroethane
and benzene are considered.
Vaporization rates of VOCs from the two surface impoundments
are in Table 11. Since the profile measurements on Pond 1,2 were
hampered by unfavorable wind and high background concentra-
tions the flux rates were calculated using the effective transport
coefficients and the measured concentrations of the volatiles in the
water. Fifty kilograms/day of total chlorinated hydrocarbon is in-
dicated for pond 1,2 with 9 kg/d being contributed by 1,1 dichloro-
ethane. Only five kg/d of total hydrocarbon is being emitted from
pond 6 with benzene and toluene combined for 0.9 kg/d of the total.
CONCLUSIONS
Conclusions drawn from this study are generalized with respect
to the CP technique and emission modeling and are also of a site
specific nature with respect to the hazardous waste facility.
Site Specific Conditions
• Levels of chlorinated hydrocarbons, including 1,1
dichloroethane, and total hydrocarbons, including benzene, were
observed (identified and confirmed) in both the water and air boun-
dary layer of two surface impoundments of a hazardous waste
disposal facility.
•Concentration gradients in the air boundary layer above the
surface impoundments indicate that the ponds are a source of
VOCs; however, periodic windshifts move air with high concentra-
tion of VOC from other areas of the facility over the impound-
ments so that chemical absorption onto the water is indicated.
•VOC emissions from the two surface impoundments surveyed
include approximately 51 kg/d of total chlorinated hydrocarbon
compounds and 8 kg/d of total hydrocarbon compounds. 1,1
dichloroethane accounted for 9 kg/d and benzene accounted for
0.4 kg/d. Odors were present at the facility and had their origin, in
part, from volatile chemicals in the water. Due to the batch nature
of the operation and the very short evaporation half-life (4 to 7 hr)
of the chemicals, intense odors and higher VOC emission rates are
likely immediately after waste containing volatiles are discharged
into the impoundments.
•The transport process for the chemicals vaporizing from the im-
poundments is liquid phase controlled. The surface aerators
enhance the natural (wind dominated) transport process by a factor
of five in the case of 1,1 DCE and benzene.
Chemical
Toluene
Total HC
1,1 DCE
Total CLH
Benzene
Toluene
Total HC
1,1 DCE
Total CLH
Table 11.
VOC Emission Rates
Flux Basis
Pond (Tabale 10)
calculated
1,2 calculated
1,2 calculated
1,2 calculated
1,2 calculated
6 measured
6 calculated
6 measured
6 measured
6 measured
Emission Rate
(kg/d)
0.095
1.2
2.6
9.0
50.
0.38
0.48
5.2
0.11
1.1
General Conclusions
•The CP technique, originally designed for assessing emissions
from area sources, is apparently also capable of detecting and
quantifying absorption onto area sources. Both vaporization and
absorption were observed on pond 1,2. Only vaporization was
detected on pond 6.
•Measurements at this site provided the first data to "field test"
proposed VOC emission models for chemicals in which the
transport process is liquid-phase controlled. Comparison of
calculated emission flux rate and the measured flux rate data for
1,1 DCE and benzene indicates that the model predicts within a fac-
tor of ± 2. Model predictions of the EPTC data, which is also
liquid-phase controlled, were +3 times higher than the field
measurements. Both these model versus field measurement com-
parisons are better than that reported for methanol emission from
pulp and paper wastewater treatment lagoons.1
REFERENCES
1. Thibodeaux, L.J., Parker, D.O. and Heck, H.H., "Measurement of
Volatile Chemical Emissions from Wastewater Basins", Final Report,
USEPA, Industrial Environmental Research Laboratory, Grant No.
R-805534, Cincinnati, OH, Dec. 1981.
2. Claith, M.M., "Vapor Behavior of EPTC in Aqueous Systems",
Ph.D. Thesis, University of California, Riverside, CA., June 1978.
3. Claith, M.M., Spencer, W.F., Farmer, W.J., Shoup, T.D. and
Grover, R., "Volatilization of S-Ethyl N,N-Dipropylthiocarbamate
from Water and Wet Soil during and after Flood Irrigation of an
Alfalfa Field", J. Agric. Food Chem., 28, 1980, 610-613.
4. Thibodeaux, L.J. and Parker, D.G., "Desorption Limits of Selected
Industrial Gases and Liquids from Aerated Basins", AIChE Symp.
Ser. 156, 72, 1976.
5. Freeman, R.A., "Stripping of Hazardous Chemicals from Surface
Aerated Waste Treatment Basins", Presented at APCA/WPCF
Speciality Conference on Control of Specific (Toxic) Pollutants,
Gainesville, FL, 1979 , 13-16.
6. Freeman, R.A., Schroy, J.M., Klieve, J.R., and Archer, S.R., "Ex-
perimental Studies on the Rate of Air Stripping of Hazardous Chemi-
cals from Waste Treatment Systems", APCA meeting, Montreal,
Canada, June 1980.
7. Hwang, S.T., "Air Emission Monitoring—Evaluation Guideline
for Land Disposal Toxic Air Emissions", USEPA Guidance Docu-
ment for Subpart F, USEPA Office of Solid Waste, Dec. 1980, 5-12.
9. Thibodeaux, L.J., Chemodynamics, John Wiley and Sons Inc New
York, N.Y., 1979, 183-189.
10. Rickabaugh, J., Report on Hazardous Waste Lagoon Samples, Uni-
versity of Cincinnati, Department of Civil Engineering, Jan. 13, 1982.
11. Furnish, T.S., Analysis Report, Environmental Consultants, Inc.,
Clarksville, Indiana, Nov. 4, 1982.
-------
AIR POLLUTION PROBLEMS OF UNCONTROLLED
HAZARDOUS WASTE SITES
THOMAS T. SHEN, Ph.D
GRANVILLE H. SEWELL, Ph.D.
Columbia University
Division of Environmental Sciences
New York, New York
INTRODUCTION
In 1980, about 41 million tons of hazardous wastes were gen-
erated by various sources in the United States' and almost 80%
of these wastes were deposited in lagoons, landfills and open
dumps.2 The cumulative effect of these environmentally unsound
practices has been the contamination of many sites across the
United States. Numerous case examples of this contamination
spreading to local community water supplies and air sheds demon-
strate that public health and welfare have been unnecessarily
threatened.
Historically, the strategy for management of hazardous wastes
has only focused on preventing contamination of water supplies
by surface runoff and underground leachate. Only recently has
awareness grown that hazardous emissions at the disposal sites can
also be a severe hazard. In many sites, air pollution problems are
often still not recognized because ambient air monitoring data are
lacking.
In this paper, the authors discuss causative factors and control
problems inherent in the management of hazardous waste sites.
Major health effects of toxic emissions are identified, current regu-
latory requirements for control are described, methods for pre-
dicting toxic emission rates are presented, and available control
techniques for minimizing hazardous emissions from those sites
are discussed.
PROBLEMS
Dust emission is one of the more significant but less dramatic
air pollution problems at hazardous waste treatment and disposal
sites. Many air quality control regions have not met the ambient
air quality standards for particulate matter, and dust particles can
consist of both soluble and insoluble hazardous materials and can
have adsorbed on them molecules of toxic substances thereby creat-
ing toxic mixtures. The respirable size of many dust particles also
enhances their hazard. Dust emissions from the waste disposal sites
may result from:
(a) Wind erosion of the waste materials
(b) Re-entrainment of particulate matter by vehicle traffic
(c) Excavation of waste materials
(d) Wind erosion of the soil cover
Another air pollution problem at hazardous waste treatment and
disposal sites is waste volatilization, the process of conversion from
a liquid or solid state to a gaseous or vapor state. This process
occurs at all landfills and waste lagoons.3' The rate of waste vol-
atilization is highly dependent upon the physical and chemical
properties of the waste, site design and operation, surrounding en-
vironment and meteorological conditions.
The vaporized contaminants of particular environmental con-
cern are halogenated organics and aromatic hydrocarbons. Unlike
fugitive dusts, a hazardous vapor may not be removed from the
atmosphere for a relatively long period of time, and the toxicolog-
ical properties can represent an unusually severe health threat. A
study of physical and chemical removal processes of 43 chemicals
revealed that for a volatile chemical, such as acrylonitrile, viny-
lidene chloride, ethylene dichloride, perchloroethylene and ben-
zo(a)pyrene, chemical removal residence times were estimated to
range between 3 to 70 days based on reaction with hydroxyl rad-
icals and ozone(5). That means vaporized contaminants may travel
from rural areas to metropolitan areas, causing more than a local-
ized problem.
Moreover, the hazardous properties of most organic compounds
will probably remain unless destroyed by reactions. The undes-
troyed hazardous vapors and gases may be adsorbed or absorbed
on small particles in the atmosphere and then eventually will fall
out on land and in waters. Subsequently, they can be re-emitted
into the atmosphere again through a cyclic process illustrated in
Fig. 1.
Figure 1.
Hazardous Constituent Cycle
HEALTH EFFECTS
The human health effect of fugitive dust and toxic vapors can
be acute or chronic. An acute effect is a sudden, recognizable ill-
ness directly attributable to dust and chemical exposures. For ex-
ample, dust can irritate the respiratory system or occur in suffic-
ient quantity to physically overwhelm the respiratory system's
natural defense mechanisms. Similarly, some gases, such as meth-
ane or carbon dioxide, can dilute the oxygen available to the lungs
sufficiently to cause asphyxiation. In other cases, such as inhala-
tion of carbon monoxide, the mechanism of asphxiation is bio-
chemical. Other materials can be severe irritants or cause serious
76
-------
SAMPLING AND MONITORING
77
responses involving the immunological system with subsequent
systemic shock, swelling of limbs, or various other neurological
disorders.
More frequently, however, the dust and vapors will contribute to
the occurrence of chronic diseases, the accelerated failure of organ
systems or development of genetic disorders, such as cancer and
teratogenesis. This illness can develop over long periods of time or
the onset can be delayed for years or even decades, a time span
known as the latency period. Lung cancers attributable to asbes-
tos exposures during World War II, for example, have been re-
ported in increasing numbers during recent years. Occurrence of
the diseases will depend upon numerous risk factors, such as
genetic characteristics, smoking or dietary habits, subsequent or
previous environmental exposures, and the type and intensity of
exposure.
Usually, however, generalizations can be made. Dust is asso-
ciated with chronic bronchitus, emphysema, and, in some cases,
lung cancer. Toxic vapors or soluble particulates will also affect
specific organ systems. For example, mercury vapors will impair
the central nervous system; benzene will suppress the capacity of
bone marrow to form blood cells and can eventually contribute
to occurrence of leukemia; cadmium and other heavy metals can
affect the kidneys.
The possible roles of toxic vapors and hazardous particles in
causation of cancer and cardiovascular disease are more complex.
Carcinogens, for example, are generally classified into three cate-
gories:
•The direct-acting or primary carcinogens will produce tumors in
a specific organ or where exposure occurred. However, these car-
cinogens in small quantities can usually be detoxified and ex-
creted before cancers develop.
•The secondary or pro-carcinogens normally have target organs,
such as the liver, and often will be transformed into the active
agent by metabolic processes. Often these compounds require pro-
moters.
•Promoters activate or stimulate the carcinogenetic effects of car-
cinogens absorbed by tissues from previous exposures. Examples
would be croton oil or phorbol esters. Thus, is a person were ex-
posed to croton oil but had not been previously exposed to a car-
cinogen, one would assume that no tumor would develop. If a
person were exposed to a recognized secondary carcinogen, such
as benzo(a)pyrene that is present in trace amounts in almost all
industrial and urban environments, and then the person is ex-
posed to croton oil, rapid development of tumors can be ex-
pected.
Considerable progress is being made in identifying the carcino-
gens that may be present in hazardous wastes but identification
of promoters is considerably more difficult.
The prediction of health effects from chemical exposures is al-
ways complicated by the differences in responses between in-
dividuals. Each person will be more resistant or susceptible than
another person because risk factors will differ. Furthermore, haz-
ardous waste vapors and dusts are mixtures of chemicals, and
are not pure, individual chemicals like those used in research or
even in industrial environments that have provided the locale for
most of the current toxicology knowledge. At this stage, esti-
mates of human health effects from hazardous waste sites must be
based upon identification of specific toxic chemicals, estimates of
human exposure to these chemicals, and then some comparison
with levels measured in toxicology research or found in epidemio-
logical investigations. However, these estimates of possible effects
represent crude, often incomplete approximations and additional
effects may often be found.
REGULATORY REQUIREMENTS
Thousands of landfills and surface impoundments used for dis-
posal or treatment of hazardous wastes are operating with interim
status under Section 3005(e) of the Resources Conservation and Re-
covery Act (RCRA). These interim regulations and standards
promulgated on May 9, 1980, and Jan. 12, 1981, primarily attempt
to protect against contamination of surface and underground water
supplies. Air pollution problems associated with hazardous waste
facilities are not adequately addressed. RCRA regulatory require-
ments6 promulgated to reduce hazardous air emissions from haz-
ardous waste sites include:
•Waste piles containing hazardous waste must be covered or other-
wise managed to prevent wind erosion
•Migration of air contaminants from the site must be controlled
•The vertical and horizontal escape of gases must be controlled
by a gas collection and control system if one is present in the
landfill
•Bulk or noncontainerized liquid waste must not be placed in a
landfill with few exceptions, such as use of liners or pretreatment
of the waste
The air monitoring requirements for new hazardous waste sites
under RCRA include:
•An air monitoring system to yield air samples for analysis and
to provide sufficient ambient air quality data to perform the re-
quired comparison and evaluations
•An air emission sampling and analysis plan which describes the
sampling and analytical techniques and procedures
•An air emission evaluation to compare the anticipated effect of
the waste site on ambient air quality with the provision of the
site permit
•Records of all analyses and evaluations of ambient air quality,
wind direction, and speed
The number and location of the monitors will depend upon the
size of a site, meteorological conditions, prevalent winds, and the
surrounding population density and distribution. Meterological
data are needed to facilitate the interpretation of the ambient mon-
itoring data.
Air monitoring is primarily concerned with the detection of in-
dividual hazardous contaminants; however, indicators for toxic
compounds, such as total hydrocarbons, total hologenated com-
pounds, or total chlorine content, are used for air quality assess-
ments. Further analysis for composition of hazardous air emissions
at the location is required if monitoring detects the release of a
significant amount of such indicator compounds. Measurements at
the time of a maximum emission rate are preferable but they
should reasonably be representative overall of emission levels.
However, both average and worse case approaches to analyze the
confidence in making evaluations would be useful.
The New York State Department of Environmental Conserva-
tion (DEC) currently requires hazardous waste land disposal fa-
cilities to have five1 cells based on chemical properties of the waste
to be disposed.7 These cells are to be hydrologically isolated:
•General cell
•Flammable waste cell
•Pseudometal cell
•Heavy metal cell
•Halogenated organic cell.
The primary reason for subdividing the landfill into cells is to
match the construction material used in the receiving cell with the
corrosive and other physical characteristics of the waste. Sub-
division in the landfill also promotes safe handling during dis-
posal operations, permits accurate record keeping of waste de-
position and facilitates environmental monitoring for leakage and
emission. Furthermore, use of cells can reduce the rate of gas gen-
eration within the entire landfill and, correspondingly diminish
overall hazardous air emissions.
AMBIENT AIR QUALITY ASSESSMENT
The purpose of ambient air quality assessment is to assure that
emission levels of air contaminants into the atmosphere will be
within acceptable limits for protection of public health and the
environment. Ambient concentrations of air contaminants at or
near a source of emissions are generally obtained by actual mea-
surement or monitoring. However, under current logistic con-
straints, such as limited resources of money, manpower, and time
-------
78
SAMPLING AND MONITORING
required for answers, prediction methods for toxic emissions from
hazardous waste sites may be applied as a screening tool for am-
bient air quality assessment in the absence of monitoring data.
Prediction methods for average emission rates of major organic
contaminants are available and are discussed in the next section.
To convert emission rates (mass/time) to ambient concentrations
(mass/volume), an appropriate atmospheric dispersion model must
be applied. Numerous atmospheric dispersion models are avail-
able, but none were designed for estimating ambient concentra-
tions from area sources, such as hazardous waste disposal sites.
Among the available atmospheric dispersion models, the com-
puter programmed PAL model' appears to be the best for this ap-
plication. PAL is a multisource Gaussian-Plume atmospheric dis-
persion model used directly in the computation for point, line,
and curved path sources. It is suitable for computing ambient con-
centrations nearby a receptor located at least 100 meters downwind
of area sources. Concentration calculations are based on hourly
meterology, and averages can be computed for averaging times
from one to 24 hours.
EMISSION PREDICTION METHODS
For fugitive dust emissions produced by wind erosion of waste
piles, about 2.5 to 10% of the waste may become airborne de-
pending on the waste type and properties, wind velocity, and sur-
face geometry. This range of air emission was originally developed
for estimating emissions generated by agricultural soils.'
Dust emissions from unpaved haul roads by vehicle traffic are
affected by the road surface texture on the road, road material,
surface moisture, vehicle speed and vehicle type. Emissions can
be predicted using the following equation described elsewhere:9'10
E = (0.81S)(v/30)(365-N)/365 (1)
where:
E = Emission factor (Ib/vehicle-mile);
S = Silt content of road surface material (%by weight of
particles less than 75 /t diameter);
V = Average vehicle speed (miles/hr); and
N = Mean annual number of days with 0.01 in. or more of rain-
fall.
This equation valid for four-wheeled vehicles with speeds in the
range of 30-50 miles/hour. It seems reasonable to adjust Equa-
tion (1) by a multiplier factor of W/4 applied to vehicles with more
than four wheels where W = number of wheels on a vehicle.
Waste volatilization from landfills and surface impoundments
have been studied by a number of researchers.""19 They have
developed several prediction models for emission estimation by in-
corporating numerous variables and extensive input data into com-
plicated calculations with multiple unit conversions. Based on their
studies and experimental data, Equations (2-6) were recommended
as screening tools after many variables were consolidated into few
so that the input data could be kept to minimum and restricted
to those data that could be readily obtained.
Emission rates of organic compounds from industrial waste
buried sites can be predicted, assuming that diffusion in the vapor
phase is the only transport process operating. If transport in mov-
ing water and degradation of the organic compound in the site are
considered insignificant, the emission rate of a specific organic
compound based on Pick's law can be estimated" using the follow-
ing equation:
Ei = DiCsiAPt4/3Wi(l/L)
where:
(2)
EJ = emission rate of a specific compound in the wasteJg/sec)
DJ = diffusion coefficient of a specific compound (cm /sec)
GSJ = saturation vapor concentration of the compound (g/cm )
A = exposed area (cm )
P( = total soil porosity (dimensionless)
L = effective depth of soil cover (cm)
\V j = weight fraction of a specific compolund in the waste
The soil porosity can be calculated by the following equation:
Pt-l-Vp (3)
where:
d^ = soil bulk density (g/cm3)
dp = particle density (g/cm5).
Dump sites are landfills which have no covering material and
cause greater volatilization problems. Based on Arnold's"
diffusion equation, the volume of vapor generation of a pure
organic compound under steady state conditions can be calcu-
lated " using the following equation:
dV/dt = 2CeW(DLv/7rFv)1/2Wj (4)
where:
dV/dt = emission rate (cmVsec)
C = equilibrium vapor pressure
W = width of the dump site (cm)
D = diffusion coefficient (cmVsec)
L = longest dimension of the dump site (cm)
v = wind speed (cm/sec)
Fy = correction factor (see Fig. 2)
W- = weight fraction of a specific compound in the waste
100
Data Source • Reference 19
Figure 2.
To Find Correction Factor Fv
A strong wind can increase the emission rate; however, it also
increases the dilution factor. Thus, the net-effect of wind speed on
ambient concentration becomes compensative and depends on
location of the receptor.
The rate of waste volatilization from industrial lagoons or ponds
can be estimated4 using the following empirical equations:
(ERP)j = (18x10"*) (Kj ):A C; (5)
i_ LJ 1 I
where:
(ERP)j = emission rate potential of a compound (g/sec)
(KL)j = liquid-phase mass transfer coefficient of a compound
(g-mol/cm'-sec)
A = lagoon surface area (cm2)
Cj = concentration of the compound in wastewater (tag/I).
Although Equation (5) appears relatively simply, its solution is
difficult primarily because we have only limited experimental data
for determining the liquid-phase mass transfer coefficient (Ki)
of organic compounds. However, there are methods to determine
-------
SAMPLING AND MONITORING
79
of organic compounds. Most of these methods involve com-
plex equations and mathematical models which are too complicated
to be applied by field and practicing engineers for quick and prac-
tical solutions. The following is a simplified equation recently
developed:4
(6)
(KL)j = 4.45 x 10-3(Mi)-°-5(1.024)t-20(u)0-67(H)-°-85
where:
= molecular weight of the compound (g/mole)
= surface water temperature (OC)
u = surface velocity = 0.035 wind velocity (cm/sec)
H = water depth of the lagoon (cm)
The above six emission-related prediction equations should be
applied cautiously since they are suitable only for screening pur-
poses in extreme situations where emission rates and risks are clear-
ly acceptable or unacceptable. Therefore, it would be advisable to
analyze the level of confidence by making estimations using both
average and worse case approaches.
CONTROL TECHNIQUES
As with all air pollution problems, dilution and dispersion of
hazardous air emissions into the atmosphere is still a vital method
of control. But for less-reactive hazardous air contaminants, this
approach of dilution and dispersion should be applied with ex-
treme caution to meet the acceptable ambient air quality guide-
lines.
Fugitive dust emissions from haul roads used to transport wastes
may be controlled by watering, chemical stabilizing, reducing
vehicle speed, or paving these roads. Of course, dust emissions can
effectively be controlled by paving these roads with concrete or as-
phalt. This approach is expensive and, therefore, has not been used
on temporary roads at most waste treatment and disposal facil-
ities. Main haul roads intended for long-term use are normally
paved, however. Control efficiencies of available methods for fug-
itive dust emissions are shown in Table i:10
Table 1.
Control Methods for Unpaved Roads.
Control Method Efficiency, %
Paving 85
Treating surface with penetrating chemicals 50
Working soil stabilizing chemicals into roadbed 50
Vehicle speed control®
30miles/hr 25
20miles/hr 65
15 miles/hr 80
©Based on the assumption that uncontrolled speed is typically 40 miles/hr.
Between 30 to 50 miles/hr emissions are linearly proportional to vehicle
speed.
Emissions from industrial waste lagoons can be minimized by
design considerations, such as increasing lagoon depth and de-
creasing lagoon surface area or constructing wind barriers at the
upward location where the prevailing wind occurs in summer and
autumn seasons. The best approach would be to remove volatiles
from the waste entering into the lagoon. Steam stripping is one
of the control alternatives for removal of volatiles from waste-
water. For hydrocarbon mixtures, control alternatives may be re-
cycling or recovery of volatiles by conventional distillation pro-
cesses or disposal by incineration. Air emissions can be sub-
stantially reduced by covering the waste lagoon with an oil film.
However, this approach can interfere with aerobic processes which
are important for biological treatment in waste lagoons.
Control of landfill gam emissions is more difficult than control
of leachate because prediction of the migration route for landfill
gas is more complicated and difficult than for leachate. At present,
landfill gas is generally vented directly to the atmosphere. This
practice may cause significant adverse effect on air quality. Un-
less the quantity and composition of gaseous emissions are known
to be insignificant, gas collection and control systems should be
installed to prevent hazardous air emissions."
To determine if gas collection and control systems will operate
efficiently, one should first investigate whether the gas migra-
tion is primarily a pressure or diffusion problem. If the predom-
inant mechanism is pressure flow, one would normally choose a
passive venting system with adequate control to remove the
pressure gradient, assuming the cover is relatively impermeable.
On the other hand, if the primary gas migration process is a
diffusion flow, a passive venting system would probably not be
effective and the gas will eventually escape from the landfill cracks
unless a forced venting system is installed to maintain a slightly
negative pressure within the landfill.3'19
A practical and effective control of gas generation and emission
for new landfills is pre-treatment of the wastes. The usual objec-
tives of pretreatment are to destroy, recover, or convert hazardous
components of the wastes into forms suitable for reuse or innoc-
uous forms that are acceptable for land disposal. Gas generation
can also be reduced by eliminating all volatile organic wastes and
liquids from landfilling. Keeping the liquids from entering and
leaving landfills would eventually minimize gas generation and gas
emissions. This can be accomplished by using liners that may be
concrete, asphalt, certain plastics, or a mixture of natural soil and
sodium bentonite. Sodium bentonite swells more than ten times
its original size when it comes into contact with water. When com-
bined with natural soil, it creates an impermeable layer.
As a remedial action at existing landfills, the use of covering
materials known as "capping," has been proven to be an effec-
tive method for temporary solution of a problem that needs
immediate action. An example of such remedial action was the
Caputo disposal site located near South Glens Falls, N.Y. Assorted
materials, such as topsoil, papermill sludges and manure, were
used to cover the site and prevent PCB volatilization during the
summer months in 1980. These materials were readily available,
economical, and easy to apply. Furthermore, papermill sludge
and manure have a high capacity for gas and vapor adsorption.
Because they are organic material, they are also relatively combus-
tible and can be later incinerated.
One of the important measures in controlling hazardous air emis-
sions is to avoid co-disposal of certain wastes in the same land-
fill cell or lagoon. Examples of such wastes are acids with metals
or solvents with certain highly toxic aromatic hydrocarbons. In
some cases, waste stabilization or solidification may be necessary to
reduce their mobility and emissions.
SUMMARY AND CONCLUSIONS
•Dust emission and waste volatilization have been identified as
the major air pollution problems at hazardous waste sites. But the
degree and extent of such problems have not been well defined
because of lacking air monitoring data.
•Emissions of fugitive dust may cause a localized problem but toxic
vapor releasing from waste disposal sites may travel in the atmos-
phere a long distance from rural to urban areas. Effects of such
emission on human health and the environment may be significant.
•Current regulations are not adequate to control hazardous air
emissions. Air pollution control seems to be a weak linkage of
hazardous waste legislation.
•In the absence of air monitoring data, emission prediction
methods and atmospheric dispersion models may be used for am-
bient air quality assessment of waste disposal sites. The accuracy
of the emission prediction methods and atmospheric dispersion
models, however, depends on the availability of experimental data
because the theoretical basis and analytical tools are insufficient.
•Both dust emission and waste volatilization can be minimized by
available technology at a cost that is widely accepted in public-
works projects. Thus, air quality should not be degraded unless
the degree of degradation is determined, evaluated, and found to
be acceptable. To do otherwise would be to blindly allow haz-
ardous emissions without knowing the consequences.
-------
80
SAMPLING AND MONITORING
REFERENCES
1. USEPA, EPA Activities Under The Resources Conservation and Re-
covery Act of 1976, Annual Report to the President and the Con-
gress for Fiscal Year 1978, SW-755, March 1979.
2. USEPA, Hazardous Waste Generation and Commercial Hazardous
Waste Management Capacity, P. 111-3 and Appendix C-2, SW-894,
Dec. 1980.
3. Shen, T.T. and Tofflemire, T.J., "Air Pollution Aspects of Land
Disposal of Toxic Waste" /. Of I he Environmental Engineering Divis-
ion ofASCE, - °t, No. EE1, 1980, 211-226.
4. Shen, T.T., "A Simplified Method for Estimating of Hazardous
Emissions from Waste Lagoons" Pre-print paper No. 82-46.2, pre-
sented at the 75th APCA Annual Meeting in New Orleans, Louis-
iana, June 20-25, 1982.
5. Cupitt, L.T., "Fate of Toxic and Hazardous Materials in Air Environ-
* mem," USEPA Publication No 600/S3-80-084, December 1980.
6. USEPA, The Interim Status Standards for Hazardous Waste In-
cinerators and Hazardous Waste and Consolidated Permit Regula-
tions, Part 265 of Federal Registers Dec. 18, 1978; May 19, 1980;
Jan. 23, 1981; and June 24, 1982.
7. New York State Environmental Facilities Corporation, "Technical
Marketing, and Financial Findings for the N.Y. State Hazardous
Waste Management Program", prepared by Camp Dresser & McKee
Environmental Consultants, P. 111-12, Mar. 1980.
8. USEPA, User's Guide for PAL—A Gaussian—Plume Algorithm
for Point, Area, and Line Source, EPA-600/4-78-013, February 1978.
9. PEDCO Environmental Inc., Technical guidance for control of In-
dustrial Process Fugitive Paniculate Emissions. USEPA Publication
No. 450/3-77-010, Mar. 1977.
10. USEPA, Compilation of Air Pollutant Emission Factors, Supple-
ment No. 5 of EPA Report AP-42, December 1975.
11. Thibodeaux, L.J., Chemodynamics, Chapter 4, John Wiley and Sons,
Inc., 1979, 139-189.
12. Cohen, Y., Cocchio, W., and Mackay, D., "Laboratory Study of
Liquid-Phase Controlled Volatilization Rates in Presence of Wind
Waves," Env. Sci& Tech., 12, 1978, 553-558.
13. Owens, M., Edwards, R.W. and Gibbs, J.W., "Some Reseration
Studies in Streams," Inter. J. Air and Water Pollution, 8, 1964,496.
14. Smith, J.H., et al., "Prediction of Volatilization Rates of High-
Volatility Chemicals from Natural Water Bodies," Environ. Sci. and
Tech., 14, N'"°, ""—".
15. Neely, W.B., "Predicting the Flux of Organics Across the Air/Water
Interface," Proceedings of the 1976 National Conference on Con-
trol of Hazardous Materials Spills, New Orleans, Apr. 1976, 197-
200.
16. Hwang, S.T., U.S. EPA Draft Guidance Document for Subpart F.,
Air Emission Monitoring—Evaluation guideline for Land Disposal
Toxic Air Emission, Office of Solid Waste, Washington, D.C., Dec
1980.
17. Farmer, W.J., et al., "Land Disposal of Hexachlorobenzene Wastes-
Controlling Vapor Movement in Soil", USEPA-600/2-80-119, August
1980.
18. Arnold, J.H., "Unsteady-State Vaporization and Absorption", Trans-
action of American Inst. ofChem. Engineers, 40, 1944, 361-379.
19. Shen, T.T., "Control Techniques for Gas Emissions from Hazardous
Waste Landfills" J. APCA 31, Feb. 1981, 132-135.
-------
THE INVESTIGATION OF MERCURY CONTAMINATION
IN THE VICINITY OF BERRY'S CREEK
DAVID LIPSKY, Ph.D.
Fred C. Hart Associates
Newark, New Jersey
PAUL GALUZZI
Hackensack Meadowlands Development Commission
Lyndhurst, New Jersey
INTRODUCTION
One of the largest anthropogenic sources of mercury in the world
can be traced to an uncontrolled hazardous waste site located in the
Hackensack Meadowlands of New Jersey (Fig. 1). From both a
technical and legal perspective, this site has been one of the most in-
tensely studied hazardous waste sites, thereby providing a useful
model for the study of landfill investigation and remedial action
programs.
Enforcement and investigation activities, completed to date, en-
compass most of the tasks that must be completed for a typical un-
controlled hazardous waste site case. These include:
•Site identification/site history
•Site investigation
•Initial assessment of potential site related environmental/
public health effects
•Initiation of litigation against past/present landowners
Figure 1.
Location of mercury site in boroughs of Wood Ridge and Carlstadt, Bergen County, in the Hackensack Meadowlands, N.J.
81
-------
82
SAMPLING AND MONITORING
•Rendering of a court decision
The remaining tasks in this investigation include:
•Obtaining permits for remedial action work
•Implementation of a remedial action plan
A summary of the work completed to date is described below.
SITE IDENTIFICATION AND SITE HISTORY
The mercury dump site is located on a 40 acre tract of land in the
New Jersey Hackensack Meadowlands. A more detailed site loca-
tion map is provided as Fig. 2. The site is bounded on the east by
Berry's Creek, a tidal tributary of the Hackensack River. As shown
in the figure, Berry's Creek is sectioned near the site by a tidegate
installed prior to 1930 and replaced in 1978. The surrounding area
is zoned for light and heavy industry, with residential areas located
Figure 2.
Detailed map of Berry'* Creek in the Hackensack Meadowlands District
within 1,000 ft to the north and west of the site.
From 1930-1974 a mercury processing plant operated on a 7 acre
portion of the 40 acres. During those years, property and plant
ownership changed hands, but plant operations remained directed
to the manufacture of fungicides, insecticides, red and yellow ox-
ides of mercury, phenyl mercuric acetate, and other organic and in-
organic forms of mercury. The plant also had a distilling operation
in which contaminated mercury was purified and recovered from
both plant wastes and from outside customers.
Although a major portion of the tract was originally wetland, ap-
proximately 19 acres between Berry's Creek and the plant site was
used as a dump site primarily for the disposal of the plant's in-
dustrial and chemical wastes.
The history of state/federal regulatory and enforcement actions
against the operators of the plant is a long and detailed one. Begin-
ning in 1956 with state enforcement actions, the site came under in-
creasing federal scrutiny with the creation of the USEPA in 1970.
At that time, it was thought that the major problem associated with
the site was the active discharge of mercury into Berry's Creek sur-
face waters. Subsequent to the reduction of these discharge levels,
and the discovery that the soils near the site were highly con-
taminated with mercury, lead enforcement responsibilities were
returned to the state.
Plans by a local developer to construct two warehouse buildings
(which have since been constructed) on the site where the process
building was located, resulted in the collection of a series of soil
samples by regulatory agencies.' Concentrations of mercury above
1 % were detected in some areas.2 This discovery, and the failure of
local parties to agree to terms for the entombment and/or cleanup
of the site, led the State to commence a law suit in APril 1976,
against all past and present land owners, to determine the liability
for site cleanup. A decision in favor of the State was rendered on
Aug. 27, 1979.
SITE INVESTIGATION
Site investigation activities included the collection of air, water,
soil, and sediment samples to determine the magnitude and extent
of mercury contamination at the site and within surrounding
wetlands. Additionally, since mercury particularly in the
methylated form, can move into the food chain, samples of biota
were collected to insure the lack of an immediate threat to public
health.
The investigation was coordinated and directed by the N. J. Dept.
of Environmental Protection (NJ-DEP). As indicated, the data
that will be presented were collected by several institutions in-
cluding NJ-DEP, USEPA, Hackensack Meadowlands Develop-
ment Commission (HMDC), N.J. Marine Sciences Consortium
(NJMSC), and Jack McCormick Associates, Inc. (JMA).
Extensive sampling of air, water, soil and sediment indicated that
the soil on and adjacent to the 40 acre tract of land is highly con-
taminated with mercury. Furthermore, a zone of mercury con-
tamination was found to extend approximately 13,000 ft southward
from the site, within the Upper Berry's Creek tidal basin, to the
Route 3 bridge. The increasing degree of contamination with prox-
imity to the site indicates that the site is the source of mercury con-
tamination in the adjacent areas.
A summary of the soil and sediment sampling data is given in
Table 1. In a series of 31 core samples collected by JMA on the 40
acre site, concentrations of mercury ranged up to 123,000 ppm. In
at least 13% of the samples, concentrations greater than 1000 ppm
were found.2 Similarly, high levels of mercury were found by
HMDC within the tidal marshes adjacent to Berry's Creek. Among
13 soil sore samples collected north of the Route 3 bridge, four
samples (31 %) had peak mercury concentrations of over 1000 ppm.
In contrast, among 36 core samples collected elsewhere in the
Meadowlands district, only one soil sample had greater than 100
ppm as a peak mercury concentration.3-4
-------
SAMPLING AND MONITORING
83
Table 1.
Data Summary-Soil and Sediment Sample Collection
% samples within Hg
concentration range
Sample Desc.
Description
Soil core sam-
ples collected
40 acre Hg site
to depth of
18ft
Peak Hg concen-
trations in
marsh core sam-
ples collected to
depth of 18 in.
Peak Hg concen-
tration in chan-
nel sediments to
depth of 18 in.
No. of Peak Hg
Sam- Cone.
pies (ppm)
0.1-100
(ppm)
101-1000 1000 +
(ppm) (ppm)
a. Within Upper
Berry's Ck tidal
basis north of
Rt. 3 bridge
b. All other
Meadowiands
locations
a. Within Upper
Berry's Ck 28
tidal basin
North of Rt. 3
bridge
b. All other
Meadowiands
locations
123,000 55% 32%
2,006 15% 54% 31%
36 158 97% 3%
1,730 46% 39%
30 97 100%
The correlation of mercury concentrations in marsh soils with
proximity to the site is illustrated by Fig. 3, where peak mercury
values within three different ranges of mercury concentrations are
plotted for each of the sampling locations.'
Results from an HMDC study of channel sediments collected
from Berry's Creek and other creeks in the Meadowiands are con-
sistent with the data obtained from marsh soils. The concentration
of mercury within Berry's Creek ranged to 1730 ppm, with greater
concentrations found nearer to the site. Among 28 core sediment
samples collected between the tidegate and the Route 3 bridge, over
53% of the samples had peak mercury values in excess of 100 ppm.
In contrast, none of the sediment samples collected from other
creeks had peak concentrations exceeding 100 ppm.
The data indicate that the concentration of mercury in Berry's
Creek channel sediments falls off rapidly with distance from the
site. This is shown in Fig. 4 where the concentration of mercury in 0
to 4 in. and 4 to 8 in. deep core sample intervals are plotted against
distance from the site.
BERRYS CREEK
SAMPLING SITES
fl ti 23 14 IB
ROUTE 3 M1IX2
0-4"CORE SAMPLE
f-i'cam MMPI.C
Figure 4.
Mercury concentrations in 25 channel core samples collected
from Berry's Creek (Route 3 bridge to tidegate)
Figure 3.
Geographical distribution of peak mercury concentrations in the
marsh sediments of the Hackensack Meadowiands District
Above background levels of mercury were found in the surface
water of Berry's Creek." Filtered and unfiltered water samples were
collected monthly by HMDC at high and low tides, at eight
separate Meadowiands locations, with three locations in the Upper
Berry's Creek tidal basin. Essentially no mercury was found in the
filtered water samples. However, among the unfiltered samples,
36% of the Berry's Creek samples were in excess of 1.2 jig/1. In
contrast, in the remaining sampling locations only 6% of the
samples exceeded this level. The surface data are summarized and
compared to data collected from other New Jersey locations in
Table 2.
A study of organic" and inorganic mercury concentrations in the
ambient air above the 40 acre site was also performed.7 Ambient air
quality was monitored by USEPA over a 5 day period at three on-
site locations and within one adjacent warehouse building. Samples
-------
84
SAMPLING AND MONITORING
were collected for 8, 12, and 24 hr periods using silver foil columns
to trap elemental mercury (via amalgamation). No organic mercury
was found in this study. However, elemental mercury concentra-
tions ranged from 0.29 ;ig/m3 to 3.3 /i/m3 (Table 3).
Table 2.
Frequency Distribution of Mercury Concentrations (mg/l) in
New Jersey Surface Waters Compared with Those Found in the
Hackensack River and Berry's Creek.9
Range Hg
(ppb)
n.d.-O.l
0.1-0.2
0.2-0.3
0.3-0.4
0.4-0.5
0.5-0.8
0.8-1.2
1.2 +
Sampling Sites
N.E. N.J.
103 (70%)
16(11%)
22 (15%)
2 (1 %)
1(1%)
1 (1%)
0
0
Hackensack R.
(excluding
Berry's Creek)
33 (30%)
33 (30%)
15 (13%)
9 (8%)
6(5%)
7 (6%)
1 (1 %)
6 (6%)
Berry's Creek
3 (3%)
1 (2%)
2(3%)
5 (8%)
6 (9%)
9 (14%)
17 (26%)
24 (36%)
These average mercury concentrations found in the ambient air
were not judged high enough to indicate any immediate threat to
nearby residents or workers. Nonetheless, when compared with
recommended occupational exposure limits of (50 /ig/m3 for a 40
hour work week) and adding safety factors for 24 hour potentially
continuous exposure of non-worker populations, a precautionary
study of mercury levels in local residents and workers was under-
taken by the N.J. Dept. of Health (described below).
MERCURY CONTAMINATION IN BIOTA
Between 1978 and 1980, as part of the site investigation, the DEP
sponsored several studies of mercury contamination in aquatic and
terrestrial biota.3'7 The purpose of these studies was to assess the im-
pact of mercury contamination on local biota, particularly those
species trapped and consumed by area residents.
Table 3.
Concentrations of Mercury in Ambient Air (ug/m1) at Four Locations
at a Mercury Dump Site In Woodridge, Bergen County, New Jersey
During August 1978
Station &
Dales
Site I
8/14
8/15
Site 2
8/10-11
8/11-12
8/12-13
8/13-14
8/14
8/14
•mean
Site 3
8/10-11
8/11-12
8/12-13
8/13-14
8/14-15
•mean
Site 4
8/12-13
8/13-14
8/14-15
•mean
24-hr values
(MR/ID*)
1.02
0.46
0.52
1.01
0.76
0.29
0.29
0.39
0.27
1.02
0.75
2.85
1.03
0.20
2.09
2.21
1.50
12-hr values
0.47
0.79
1.65
1.54
0.21
0.18
0.38
1.00
8-hr values
Ctg/rn1)
0.38
0.74
0.55
0.38
0.44
0.72
1.68
0.25
0.50
1.37
3.26
nean 1.50
In a large study conducted in 1978 by HMDC, 479 specimens of
sh comprising 14 species, 94 specimens of aquatic invertebrates,
36 mammals, and 35 avian specimens were analyzed. Average mer-
cury concentrations for each vertebrate class and for selected in-
vertebrate species are reported in Table 4. Fish and invertebrates in
the Meadowlands district exhibit slightly higher mercury con-
tamination relative to background levels.' However, the average
mercury concentrations were well within levels acceptable for
human consumption (1 ppm). Similar results were obtained in two
later studies of finfish and shellfish completed by the NJMSC in
1980.
The only permanent resident finfish that can be found in abun-
dance in the waters of the Upper Berry's Creek tidal basin is the
Common Mumichog of Killifish. Although these organisms are not
consumed by man, they are a useful indicator of relative con-
tamination in that they can be trapped easily and spend most of the
lives swimming within the same general area. In the study under-
taken by HMDC, Killifish collected downstream on Berry's Creek
had greater average concentrations of mercury than those collected
upstream. In two separate Killifish studies performed by NJMSC,
conflicting data were received regarding the relationship of mercury
concentrations with proximity to the mercury site. Only one
species, the fiddler crab, was found to consistently exhibit a body
burden gradient with proximity to the site. Monitoring of aquatic
organisms is continuing.
Table 4.
Concentrations of Mercury (ppm) in Tissues of Organisms
Collected in the Hackensack Meadowlands, New Jersey during 1978
Vertebrate
Class Samples Muscle Liver Kidney Fur
Fish
27 0.566
0.696
Mammals 27 0.725
No. of
Samples
479
27
23
27
34
36
34
33
35
34
34
28
10
54
Muscle
0.451
0.027
0.358
0.231
0.253
0.234
Liver
0.566
0.168
1.1
Feather
0.462
1.1
3.2
Birds
Invertebrates
Fiddler crab
Blue claw crab
Grass shrimp
Tissue concentrations of mercury in the mammals and birds col-
lected in the Hackensack Meadowlands compare favorably with
levels observed in terrestrial herbivores and omnivores.' Generally,
lower values were observed than those reported in most other
studies of mercury contamination.10 Among the birds and water-
fowl, mercury concentrations were greater in the aquatic birds than
those restricted to a terrestrial habitat. This would indicate that the
movement and transport of mercury through the Meadowlands oc-
curs primarily in the aquatic environment. Hence, those ^pecies
most contaminated by mercury will be those who depend on the
aquatic food web.
Examination of mercury burdens in specific body tissues in-
dicates that mercury accumulation occurs in the kidney and fur of
mammals. Similarly, levels of mercury in avian species were found
in decreasing concentrations in the feather, kidney, liver and mus-
cle tissue. Generally the amount of mercury observed in feathers
was three times that found in other tissues.
INITIAL HEALTH ASSESSMENT
It is clear that mercury has contaminated the physical environ-
ment surrounding the Berry's Creek ecosystem. In order to
minimize the potential for health risks to humans, several routes of
exposure have been examined. Because local residents utilize a
public supply water and because the physical/chemical
characteristics of mercury limits the migration of mercury con-
taminated groundwater, water supply is not considered an impor-
-------
SAMPLING AND MONITORING
85
tant risk. Exposure due to contaminated soils have been minimized
through the fencing of the property and through public awareness.
Although above background levels of mercury were found in the
local biota, the levels are within acceptable federal standards.
Exposure to unsafe levels of mercury vapor in surrounding
residences was also thought to be improbable by both the DEP and
the NJ Department of Health. Nonetheless, blood, hair, and urine
samples were collected for mercury analysis by the NJ-DOH from
over 300 individuals in nearby residences and businesses." The
results of these tests in 1979 indicated that there was no acute health
hazard to the population residing in the area from the mercury con-
tamination. Only low levels of mercury, typical of the U.S. popula-
tion at large, were detected in the blood, hair, and urine samples.
For example, it can be expected that less than 10 /xg/1 of mercury
can be found in the average person's blood, but levels as high as 30
/tg/1 can be found in up to 5% of the U.S. population. Near the
mercury site, ninety (90%) percent of the tested individuals in the
area had blood mercury levels below 10 /ig/1 with the high level
observed at 15 /ig/1. Similarly, with the exception of four in-
dividuals, the levels found in urine samples were well within an ac-
ceptable and non-hazardous range. (The aforementioned in-
dividuals are all members of the same family in which a bottle of
metallic mercury brought home from school was thought to have
been the major source of exposure.) Nevertheless, within this
general range of acceptable and non-hazardous low levels of mer-
cury detected in the urine samples, the NJ-DOH was able to ascer-
tain that residents who have frequent exposure to the site have
slightly higher urine mercury levels than residents with no exposure.
HISTORY OF LITIGATION
During 1972, an investigation of marsh sediments in Berry's
Creek tidal marsh revealed unusually high concentrations of mer-
cury and other heavy metals.12 Subsequent studies in 1974 and 1976
sponsored by the N.J. Sports and Exposition Authority (NJSEA)
confirmed the earlier findings. Furthermore, elevated mercury
levels were found to exist in areas outside Berry's Creek.13'14
During 1974, the demolition of the mercury processing plant
created an oily slick on the surface waters of Berry's Creek.
Analysis of water samples by HMDC, NJ-DEP, and USEPA
revealed concentrations of mercury more than 57,000 times that
allowed in surface waters. The completion of two studies by JMA
and HMDC in 1977 and 1978, funded by NJ-DEP, confirmed
heavy contamination at the mercury site and adjacent marshes in
Upper Berry's Creek. The accumulated evidence of mercury con-
tamination, some of which has been summarized above, spurred a
law suit, State of New Jersey vs. Ventron Corp. et al. which went to
trial in May 1978, and on Aug. 27, 1979, a decision in favor of the
State of New Jersey was rendered. On Nov. 17, 1979, an order and
judgment was issued which declared the defendants liability for
pollution abatement on the contaminated properties and cleanup of
Berry's Creek.
A cleanup plan was submitted to the court by the state during
Feb., 1980. It was recommended that the creek be dredged
downstream to the Route 3 bridge. For the disposal of the dredged
sediments, the state recommended the use of the contaminated pro-
perties as a dewatering and disposal facility. On Dec. 10, 1981, the
Appellate Division of the Superior Court of New Jersey
unanimously affirmed the lower court's ruling that the defendants
caused the mercury contamination and that they were responsible
for the cleanup of the present contamination found in the area.
During June 1981, the DEP submitted its application to the New
York District, Army Corps of Engineers for a permit to dredge. On
Apr. 16, 1982, following the Corps's public notice of Feb. 1, 1982,
a session was held to discuss the preliminary scope of work for the
required Environmental Impact Statement (EIS). Hence, the EIS is
presently being prepared to determine the environmental feasibility
of the dredging of Berry's Creek.
The lawsuit regarding the contamination of its mercury site and
Berry's Creek began in 1976 and still continues. The manpower and
financial resources required for this lawsuit have been enormous,
and have brought the issue to a decision of liability and cleanup,
with an appeal before the Supreme Court of New Jersey still re-
maining, and the permit process just beginning. Because of the
many facets of this case, it should be examined more closely by
those who will be cleaning up other similar sites under
"Superfund".
REFERENCES
1. "Report on the Investigation of the Ventron/Velsicol Properties
and Berry's Creek System", Reed, R. and Hutchinson. DEP Re-
port 1977.
2. Jack McCormick and Associates. "Investigations of Aquatic and
Terrestrial mercury Contamination in the Vicinity of the Former Lo-
cation of the Wood-Ridge Chemical Corporation Processing Plant,
Boroughs of Wood-Ridge and Carlstadt, Bergen County, N.J.,
1977, 97 p.
3. Hackensack Meadowlands Development Commission. "Concentra-
tions of Mercury in the Hackensack Meadowlands Ecosystem" Var.
pp. Final Report of HMDC to NJDEP, 1978.
4. Galluzzi, P. and Sabounjian, E. The Distribution of Mercury Con-
tamination in Marsh Sediments, Channel Sediments, and Surface
Waters of the Hackensack Meadowlands, N.J., Presented at the N.J.
Academy of Science, 1980.
5. Lipsky, D.R., and Harkov, R. Mercury Levels in Berry's Creek. N.J.
DEP Report, 1979, 54 p.
6. USEPA, "Air Monitoring Report to DEP", Oct. 1978.
7. "Biomonitoring and Assessment for mercury in Aquatic Fauna of
Berry's Creek Tidal Marsh and Adjacent Biozones." Report to DEP,
New Jersey Marine Sciences Consortium. Aug. 1980.
8. Sabounjian E. and Galluzzi, P. Mercury Concentrations in Fish and
Aquatic Invertebrates from the Hackensack Meadowlands, New
Jersey. Presented at N.J. Acad. of Science, 1980.
9. Galluzzi, P. Mercury Concentrations in Mammals, Reptiles, Birds,
and Waterfowl Collected in the Hackensack Meadowlands, New
Jersey. Presented at N.J. Acad. of Science, 1981, 24 p.
10. Cumbie, P .M. and Jenkins, J.H. Mercury Accumulation in Native
Mammals of the Southeast. Paper presented at 28th Annual Confer-
ence of the Southeastern Association of Game and Fish Commis-
sioners, White Sulfur Springs, W.Va., 1974, 20 p.
11. "Mercury Concentrations in Moonachie/Wood Ridge Residents".
NJDOH Report (unpublished) Sept., 1979.
12. Jack McCormick and Associates, Inc. "Draft Assessment of the Po-
tential Environmental Impact of the Construction and Operation of a
New Jersey Sports and Exposition Complex at a Site in East Ruther-
ford, Bergen County, N.J., 1971, 91 p.
13. Jack McCormick and Associates, Inc. "Supplemental Report: Mer-
cury Concentrations in Berry's Creek Marsh", Monitoring Report
No. 14, Apr. 1974, 8 p.
14. Jack McCormick and Associates, Inc. "Summary Report on Analyses
for Mercury in Sediments and Waters of the Hackensack Meadow-
lands District", 1976, 44 p.
-------
PRACTICAL INTERPRETATION OF GROUNDWATER
MONITORING RESULTS
WILLIAM A. DUVEL, JR., Ph.D.
Environmental Research & Technology, Inc.
Concord, Massachusetts
INTRODUCTION
Normally, groundwater investigations involve the installation of
monitoring wells on and around the site, collection of samples from
these wells, analysis of the samples for selected constituents, and
comparison of the results to determine the extent of the contamina-
tion. The usual approach is to compare results from upgradient
(off-site or background) wells with downgradient wells.
The problem is that many times the results of the analyses are not
clear. This is particularly true for organic compounds (like priority
pollutants) measured by GC/MS at the /ig/1 level. Commonly, dif-
ferences and anomolies are experienced which are difficult to inter-
pret and explain. For example: Suppose 10 /tg/I of trichlorethylene
(TCE) is found in Well A and 22 /ig/1 TCE is found in Well B. Is
the water in Well A really different from that in Well B?
Given that the analytical results of the groundwater investigation
form the basis for public health risk assessments and the basis for
costly remedial action, it is absolutely essential that the analytical
interpretation and results be correct. This point cannot be overem-
phasized. The results of the groundwater investigation are often
used to determine whether there is a risk to the public health and
also to determine what remedial actions are necessary. These are
important decisions and must be made on the basis of sound data
and clear interpretations.
Statistical methods are powerful tools which greatly enhance the
ability to understand and interpret the results of groundwater
analyses and allow the investigator to define the confidence and
certainty of the results. Most scientists and engineers are aware that
statistical tools are available. Yet, these tools are seldom utilized
because workers do not know or understand how to use them, do
not believe they add any real value, and/or do not believe the added
cost (for replicate analysis) is justified. In this paper, the author ex-
plains the practical application of statistical methods in the design
of groundwater investigation programs and interpretation of the
results of those programs.
OBJECTIVE
The principal objective of a groundwater investigation at a
known or suspected hazardous waste site is to design and imple-
ment a cost-effective program that will reveal, with a known degree
of certainty, whether and where the groundwater is contaminated.
In the context of the investigation, this means that one wants to
know "if the water in Well A is truly different from Well B." One
wants to be confident the results are correct and spend no more
money than is really necessary. Statistical design and interpretation
is the only way to meet this objective.
If one takes water samples from the same well on two separate
occasions (or from two separate wells at the same time) and
analyzes the samples, the measured concentration will nearly
always be different in the two samples. There are three principal
sources of this variability:
•Variability resulting from actual changes in concentration of the
analyte in the groundwater (or differences in concentration due
to differences in location)
•Variability introduced as a result of the sampling procedure
•Variability in the analytical procedure
To determine whether the groundwater is contaminated, one must
somehow subtract or otherwise take into account the variability
from the sampling and analysis and simply compare the actual con-
centrations. Proper application of statistics makes this distinction
possible.
THE "t" TEST
There are various statistical tools that can be used to design in-
vestigations and analyze experimental results. Many of these are
relevant to groundwater systems. However, the Student's t test or,
simply, t-test, has been widely used. It is popular because it is sim-
ple, applicable to many situations, easily understood and inter-
preted. Also, and perhaps most important, it has recently become
part of the Resource Conservation and Recovery Act of 1976
(RCRA) regulations relating to groundwater monitoring and repor-
ting systems.
The t-test is used to compare the concentration of an analyte in
one well (say Well A) with that in another well (say Well B). The
first step is to establish a hypothesis regarding whether the water in
these wells is the same or different. For this example, assume that
the water in Well A is the same as that in B. The next step is to take
and analyze some samples from these wells for the analyte of in-
terest. Now calculate the t-statistic using the following formula:
(1)
where
A =
nB =
mean concentration of analyte in Well A
mean concentration of analyte in Well B
the variance of analyte concentration in Well A
the variance of analyte concentration in Well B
number of samples taken from Well A
number of samples taken from Well B
This calculated t value is compared to a tabulated value of t (Table
1) at a known probability level, usually 5% or 1 % (equivalent to a
95% or 99% confidence level). If t-calculated is less than
Table 1.
Values of t
Probability Level
Degrees of
Freedom
1
2
3
4
5
6
7
8
9
10
15
20
25
30
60
120
Infinity
0.2
3.078
1.886
1.638
1.533
1.476
1.440
1.415
1.397
1.385
1.372
1.341
1.325
1.316
1.310
1.296
1.289
1.282
0.1
6.314
2.920
2.353
2.132
2.015
1.943
1.895
1.860
1.833
1.812
1.753
1.725
1.708
1.697
1.671
1.658
1.645
0.05
12.706
4.303
3.182
2.776
2.571
2.447
2.365
2.306
2.262
2.278
2.131
2.086
2.060
2.042
2.000
1.980
1.960
0.01
63.657
9.925
5.841
4.604
4.032
3.707
3.499
3.355
3.250
3.169
2.947
2.845
2.787
2.750
2.660
2.617
2.576
Source: Abridged from Sled and Tonic'
86
-------
GEOHYDROLOGY
87
t-tabulated, then the original hypothesis is correct and one con-
cludes that the water in Well A is the same as the water in Well B. If
t-calculated is greater than t-tabulated, then the original hypothesis
is incorrect: Well A is different from Well B. An example calcula-
tion is shown in Table 2.
Table 2.
Example Calculation of t-Statistic
Benzene Cone.
in Well B (Cone, in B)'
1,521
2,025
6,400
2,500
£X!B = 12,446
Benzene Cone.
in Well A
(Mg/D
22
52
35
42
£XA = 151
XA = 37.8
(Cone,
484
2,704
1,225
1,764
£X'A
484
2,704
1,225
1,764
£X'A = 6177
39
45
80
50
£XB = 214
XB = 53.5
"A'1
2 . 12,446 - (214)V4 _ ...
8B 4=1 332
2 _ 6177 - (151) /4 . ,-
A 4-1 L3y o
SA = 12.6 6B - 18.2
Test the hypothesis that Benzene in A = Benzene in B at 5 % level
t . *B' "A ^ i ^ . 53.5 - 37.8 . 1-42?
[nA nB J L •*
df " degrees of freedom «u. +n_-2»6
From Table 1, the value of t at df = 6 and 0.05 probability is 2.447;
hence, t-calculated (1.427) is less than t-tablulated (2.447).
Conclude: Original hypothesis is correct. The benzene concentra-
tion in Well A is the same as the benzene concentration in Well B.
In this example, four samples are taken from two different wells:
Well A is upgradient of the source of contamination and Well B is
downgradient from the source of contamination. The benzene con-
centration has been measured in these four samples. The calculated
t is 1.427; the tabulated t is 2.447. Since the calculated t is less than
the tabulated t, one concludes that the original hypothesis is cor-
rect: the benzene concentration in Well A is the same as that in Well
B. This statistical conclusion is especially important for this exam-
ple because, simple visual comparison of the mean benzene concen-
tration in Well A (38 fig/1) with that in Well B (54 jtg/1) leads to the.
opposite (and incorrect) conclusion that Well B contains more
benzene. Thus, the t-test allows one to state with a known degree of
confidence whether or not there is a real difference between the
water in Well A and the water in Well B.
TYPES OF ERRORS
Because statistical tests are based on probability, there is some
chance that one will reach the wrong conclusion. It is therefore im-
portant to know and understand the types of errors which can oc-
cur. There are two kinds of errors: Type I and Type II.
Type I Error
The Type I error is made when one rejects the original hypothesis
and that is in fact correct. The probability of such an error is
designated by a (Greek alpha). In the example, a is 0.05, so there is
only a 5% chance of a Type I error.
Type II Error
The Type II error is made when one accepts the original
hypothesis and it is not true. The probability of such an error is
designated by /3 (Greek beta). In the example, j3, which is a function
of the number of samples and the sample variance, was not
calculated. It is possible that a Type II error has been committed in
the example problem because the sample size is too small to show
the significant difference that may exist.
The relationship between the t statistic, the conclusion reached,
and the types of error involved is shown in Table 3. In hazardous
waste site investigations, avoiding both types of error is important.
One needs to know, with a'high degree of certainty, whether or not
there are real differences in the quality of the groundwater.
Since ft is related to the number of samples needed and the cost
of the investigation and monitoring program are also related to the
number of samples, it is important to consider this parameter.
NUMBER OF SAMPLES
In choosing the sample size, one must guard against both the
Type I and Type II errors. One would like the probability of com-
mitting these errors to be small. Typically the a-level, the probabili-
ty of committing a Type I error (concluding the two wells are dif-
ferent when, in fact, they are the same) is chosen to be 0.05. One
would also like the same level of protection against the probability
of committing a Type II error (concluding the two wells are the
same when, in fact, they are different). It is reasonable therefore to
also set /3 equal to 0.05.
Table 3.
Decisions and Associated Errors
Original Hypothesis: Analyte concentration in Well A is the same as that in WeU B.
I statistic
t-calc> t-tab
t-calc > t-tab
t-calc< t-tab
t-calc < t-tab
One Con-
cludes
A * B
A 4 B
A = B
A = B
True Situ-
ation
A * B
A = B
A ± B
A = B
Decision
is
Correct
Wrong
Wrong
Correct
Error
Type
None
Typel
Type II
None
Probability
of Error
The sample size can be determined by using the following equa-
tion:2
where:
n = number of samples needed
Ur/3 = standard normal deviate for probability 1-/3
(Table 4)
Ura/2 = standard normal deviate for probability l-a/2
(Table 4)
s = standard deviation of the analyte in the wells
XA = mean concentration of analyte in Well A
XB = mean concentration of analyte in Well B
Having established levels of a and /3, the values of U are obtained
from Table 4. If values of U are known, n depends on values in the
right hand expression of equation 2:
XA - XB
If one lets d = XA - XB, then the number of samples is a function of
the ratio s/d.
The problem is how to determine s and d so that one can com-
plete the calculation and find the number of samples needed. The
quantity d is especially important. One must select the minimum
difference between XA and XB which is to be considered significant.
In other words, how small a difference between the chemical con-
centrations in each well does one wish to be able to detect? This is
the margin the analyst wishes to be protected against for both Type
I and Type II errors.
-------
88
GEOHYDROLOGY
Ideally, one should conduct a small pilot sampling program to
obtain estimates of the means and variances of the chemicals of in-
terest. Usually the pilot program contains three or four samples
from the same well, enough to calculate the mean and variance.
Such a pilot program would provide estimates of s and d. One can
then calculate n more precisely and conduct a full-scale sampling
program based on the results of the first stage.
Table 4.
Standard Normal Deviate, U, as a Function of Probability
of the standard deviation, s. If d is defined in terms of s, then one
can easily calculate n.
For example, suppose one wants to be able to detect a difference
equal to one standard deviation. In this example, s = d, so s/d =
1. If a = 0.05 and /3 = 0.05, then substituting into equation 2 gives
the following:
)2 #2
(1.645
1.96)2 (U2
Probability
(1-0 or l-a/2)
0.80
0.90
0.95
0.975
0.99
0.995
Corresponding
y
0.842
1.282
1.645
1.960
2.326
2.576
12.99
or
By way of illustration, assume the example problem in Table 2
provides the results of the pilot program. In this example one wants
to use the expression from Eq. 2:
-) or
As shown in Table 2, »\ does not equal §j so one needs to
calculate a pooled s (Table 5). With the pooled sj = 281, one
calculates n = 15. Now double n and distribute the number of
samples between A and B in proportion to the standard deviations
(Table 5). For this example nA = 12 and nB = 18. This means to be
able to distinguish a difference of 16 /tg/1 between Well A and Well
B and be 95 % certain one has done so correctly, one needs to take a
total of 12 samples from Well A and 18 samples from Well B.
Table 5.
Calculation of the Number of Samples Required to Distinguish a
Significant Difference in Benzene Concentration Between
Well A and Well B at a = 0 = 0.05
1. Using the data from Table 2, calculate the pooled variance
I("
pooled s
pooled a
nA + n,, - 1
16117 + 12,446) - 1(151 * 212.) )
2 . 8
4 + 4-1
pooled a - 281
2. Calculate n using equation (2).
n - (1.645 * 1.96)2 [f£f]
N " 14.8 round to 15
3. Double n to account for two wells, 2n = 30
4. Distribute 2n between wells in proportion to their standard deviation
SA(30) 12.6(30)
A °8
30 - 12
12.6 * 18.2
12
18
Now, it may not be possible to conduct a pilot program and one
must estimate s and d without having data from the site under con-
sideration. If this situation exists, it is easier to consider d (the
minimum difference one wishes to be able to detect) as a percentage
n = 13
Similarly, if one wants to determine a difference equal to half of
the standard deviation (d = s/2) with the same at and 0, then n =
(1.645 + 1.96)2 (2)2 = 52. Obviously, as the minimum detectable
difference decreases, the sample size increases dramatically. If 52
samples are too many, but one still wants to be able to detect a dif-
ference equal to half of the standard deviation, then one can in-
crease a = 0.2 and 0 = 0.2, the sample size is now 18 instead of 52.
The sample size calculated as a function of different values of o, (3,
and s/d is shown in Table 6.
Common sense should be used in selecting the minimum detec-
table difference. At a hazardous waste site, one could say one
wants to know if there is any difference at all between Well A and
Well B. If one assumes any difference means detecting 1 /tg/1 dif-
ference, then one quickly arrives at an astronomical number of
samples being required. For example, if s, the standard deviation
for an analyte, were 10 /tg/1, and d = 1 /tg/1 and a and 0 both equal
0.05; then the number of samples required would be 1300. Clearly
this is an impractically large number. One must therefore consider
some larger difference as the minimum acceptable difference, or
some larger probability of reaching the wrong conclusion.
What level of difference would be acceptable for organic priori-
ty pollutants at levels between about 10 and 200/tg/1 and how many
samples would be needed to show differences between two wells?
At these low levels, the variability due to sampling and analysis may
be about the same order of magnitude as the actual variability of
the analyte in the water. Therefore if one sets d = s/2, one should
be able to distinguish any variability due to the analyte. Where d =
s/2 and a and j3 = 0.05, n = 52. This is probably too large a
number. If one lets d = s (that is d/s = 1) and a and 0 = 0.05,
then n = 13 (Table 6). This is still a large number, especially if no
pilot program has been conducted. Again using data in Table 6, d
= s, and a and 0 = 0.1, then n = 7. This may be a reasonable
number. Obviously other combinations are possible. The impor-
tant point is to select a difference which has some importance or
significance.
WHERE AND WHEN TO SAMPLE
Common questions asked in the design of an investigation are:
"how many wells do I need; how often should I sample; and how
long do I need to continue a monitoring program?" In some sense,
these questions all relate to that just addressed: how many samples
do I need?"
The number of wells depends upon the homogeneity of the site.
The statistical analysis presented here applies to any portion of an
aquifer that is reasonably uniform with respect to its geological and
chemical properties. As long as the geological and chemical obser-
vations demonstrate that water quality conditions are reasonably
uniform, the number of wells required will be independent of the
size of the area studied.
The number of wells for a homogeneous area is not a function of
land size or area. Thus a large area of 100 acres with a
homogeneous aquifer can be adequately sampled by one well.
Repeated sampling from that one well over time will be adequate to
provide the necessary number of samples needed to accomplish the
-------
GEOHYDROLOGY
89
Table 6.
Sample Size, n, for Various Values of a, (3, and s/d
s/d
0.25
0.33
0.5
0.67
0.75
1.0
1.25
1.5
0.2
0.2
1
2
3
5
7
10
0.1
I
2
3
4
7
10
14
0.05
1
1
2
4
5
8
13
18
0.01
1
2
3
6
7
12
19
27
0.2
1
2
3
4
7
13
15
0.1
0.1
1
1
7
4
5
9
14
20
0.05
1
1
3
5
6
11
17
24
0.01
1
2
4
7
9
15
24
34
0.2
1
1
2
4
5
9
14
20
0.05
0.1
1
2
3
5
6
11
17
25
0.05
1
2
4
6
8
13
21
30
0.01
1
2
5
8
10
18
28
40
0.2
1
2
4
6
8
13
21
30
0.01
0.1
1
2
4
7
9
16
25
36
0.05
2
3
6
10
12
21
33
47
0.01
2
3
6
11
14
24
38
54
statistical analysis. Stated differently, 12 wells sampled once from a
homogeneous unit will allow the same statistical interpretation as
one well sampled 12 times. In the latter case, the sampling must be
done at intervals sufficient to allow the water in the well to ex-
change with other water in the aquifer. Drawing 12 samples con-
secutively from the same well in the same day will give one informa-
tion about the sampling and analytical variability, but will not tell
one about the variability of water quality in the aquifer. Therefore
one "background" well may be sufficient. In practice, more than
one well may be installed in a single homogeneous unit. Installing
more than one well allows one to evaluate the validity of assuming a
homogeneous unit. It also allows one to get more data from the
same unit without having to wait long periods of time between
sampling.
More than one well is usually necessary in the area downgradient
from a hazardous waste source. Wells are installed to identify
where there are significant differences or discontinuities in water
quality as compared to water quality in the upgradient or
background wells. More wells are required because the location of
contaminant paths are unknown. In this case, the number and loca-
tion of wells required is clearly a function of how big one thinks the
plume is, how far one thinks it has travelled, and where one thinks
it is located. The placement of these wells is not directly related to
any statistical analysis. Finding the path of contaminated ground-
water by installing wells is somewhat like finding the studs in a wall
by pounding nails in the wall until you hit something solid.
How does one tell whether the downgradient well is con-
taminated? It is accomlished by taking n number of samples over
time and comparing the results of this well to the results of the
background well(s) using the t-test. It is perfectly acceptable prac-
tice to combine data sets from different wells and from different
time periods if they are demonstrated to be samples drawn from the
same homogeneous unit. For example, it was suggested earlier that
a pilot study is useful to get an estimate of sample means and sam-
ple variances as well as test out the sampling protocol. Consider the
example problem used in Table 2 and Table 5. The number of
samples required was 12 from Well A and 18 from Well B. But 4
samples already have been taken from each so only 8 additional
samples from Well A and 14 additional samples from Well B are
needed.
A pilot program has other advantages besides providing an
estimate of means and variances. The pilot program allows one the
opportunity to confirm the appropriate sampling methods and to
modify procedures to improve the ease and efficiency of sampling.
COST CONSIDERATIONS
The overall cost of a sampling and monitoring program is a func-
tion of the number of samples taken and it is possible to estimate
the number of required samples necessary to provide statistical
reliability. But what will such a program cost? Is the cost
reasonable? Suppose the cost is too high. How can a less costly pro-
gram be designed yet still have a known degree of statistical
reliability?
The cost of the sampling and analytical program is given by the
following equation:
CP = W + K + A (3)
where:
W = Cost of installing wells
K = Cost of sampling from wells
A = Cost of analyzing samples
Each of these terms is defined by the following expression:
W = Cm + wCw (4)
K = kCkm + kwr Ck (5)
A = kwrCa (6)
where:
Cm = mobilization cost associated with installing the wells
Cw = the unit cost of installing each well
w = the number of wells
k = the number of sampling trips
r = the number of samples per well
Ckm = tne mobilization cost associated with going out to
take samples
Ck = the unit cost to take one sample
Ca = the unit* cost to analyze one sample
Substituting equations (4), (5), and (6) into equation (3) gives:
CP = Cm + w Cw + k Ckm + kwrCk + kwr Ca
The number of samples, n, is given by the expression:
n = kwr
(7)
If r = 1 (one sample per well per trip) and substitute n in equation
7, one obtains:
C = C + n [,
p m k
w
w
+ C + C ]
k a
(8)
Equation 8 is an expression that relates the cost of the sampling
program to the number of samples. A plot of the cost of the
monitoring program as a function of the number of samples taken
is given in Fig. 1. The lines in this plot show that he cost increases
very rapidly as the sample size increases and show that (for this ex-
ample) it is more cost effective to have fewer wells and sample
longer, than to have more wells and sample for a short period of
time.
Consider now the example problem, discussed earlier in Table 2
and Table 5. To be able to distinguish a difference of 16 /tg/1 be-
tween Well A and Well B with 95 % confidence, one needs t« take
12 samples from Well A and 18 samples from Well B. What will
-------
90
GEOHYDROLOGY
50
40
2 30
0
o
E
i
20
10
7 *-W:*.
Where: Cm = S500
Cw = S2400
Ckm = 8600
ck = sioo
Ca = 8600
w = number of wells
k = number of sampling trips
n = kw
I i I
0 10 20 30 40 50
n Total Number of Simple* from All Well*
Figure 1.
Program Cost as a Function of the Total Number of Samples Taken
this program cost? Using the line in Fig. 1 for which w = 2 and n =
30, the program cost is about $35,000. Assume this cost is too high
and one is prepared to pay about $20,000, what happens? With a
program cost of $20,000, one can take a maximum of 14 samples
with two wells (say 7 from each well). What does that do to the
statistics?
Taking seven samples from each well will obviously change the
interpretatioan of the data. Initially, a difference of 16 /ig/1, s/d =
1.05, would be determined using Table 6, n = 15 at s/d = 1.0 when
or = 0 = 0.05. But now n = 7. If s/d = 1.05, then one must in-
crease (mathematically) the chance of making an error. For exam-
ple, at s/d = 1.0 (Table 6), there are two places where n = 7. One
where a = 0.1 and 0 = 0.2 and the other where a = 0.2 and j3 =
0.1. Thus one can still see a difference of 16 ng/\ but one will not be
as confident as one was originally.
One can also consider the alternative situation. Hold a = 0 =
0.05, but accept a larger difference between Well A and Well B.
Enter Table 6 at a = 0 = 0.05 and n = 7 (between 6 and 8) to find
s/d = 0.70. This means one can distinguish a difference of d =
s/0.70 = 16.8/0.70 = 24/ig/I with 95% confidence.
Thus, for a savings of $15,000 one can either lessen the certainty
of the conclusions or broaden the significant difference which one
is willing to accept. In this example it is probably reasonable to pro-
ceed with the less expensive program.
One must also ask, "why do we need any additional sampling
beyond that done in the pilot program?" In the example, the 8
samples collected cost about $13,500 (Fig. 1). Why should an addi-
tional $6,500 be spent to increase the number of samples to 14? The
answer is found in Table 6. There are several places where n = 4,
but not at s/d = 1.05. One simply cannot with any certainty
distinguish a difference as small as 16 pg/1 with these few samples.
the best one can do is distinguish 24 /ig/1 at a = 0.1 and /8 = 0.2 or
34 /ig/1 at a = 0 = .05. Both are likely not to be useful or mean-
ingful comparisons. As a result, the additional $6,500 for the extra
samples allows meaningful differences to be distinguished between
Well A and Well B with a high degree of confidence for a
reasonable amount of money.
SUMMARY
The foregoing discussion explains the value of statistics in con-
ducting groundwater investigations at hazardous waste sites, par-
ticularly where low levels of organic priority pollutants are of con-
cern. In these investigations, the t-test is an appropriate tool which
allows the investigator to make comparisons between two wells (or
two groups/sets of wells) with a known level of confidence and
known probability of being wrong. The generic steps in doing an
investigation using the t-test are these:
1. Conduct a small pilot program to estimate the mean concentra-
tion of the analyte in each well, to calculate the variance of
each well, and to adjust the sampling protocol. Usually 3 or 4
samples over time are necessary from each well.
2. Choose the minimum acceptable probability of being wrong
(i.e., select a and |8).
3. Establish d, the minimum acceptable difference in mean con-
centration between the two wells which is considered significant
(i.e., considered real, meaningful, and useful.)
4. Calculate the estimated number of samples required.
5. Develop the cost equation or cost curve which relates program
costs to the number of samples.
6. Using the original estimate of the number of samples, calculate
the anticipated cost of the program.
7. Determine if the estimated cost is reasonable. If not, reevalu-
ate the choice of a, /3, and d until a reasonable balance is
achieved between sampling costs and useful results.
8. Do the additional sampling required, pooling the results of the
pilot program with the full-scale program.
9. Calculate the t-statistic for each analyte and determine whether
there are significant differences.
REFERENCES
Steel, R.G.D. and Torrie, J.H. Principles and Procedures of Statistics,
McGraw-Hill Book Co., New York, N.Y., 1960.
Grossman, M.A., Goodwin, B.A., and Brenner, P.M. "Statistical
Analysis of Trace Metal Concentrations in Soils at Selected Land Treat-
ment Sites," Proc. of the Seventh Annual Research Symposium on
Land Disposal: Hazardous Waste. EPA-600/9/81-0026, Marc. 1981.
-------
APPLICATION OF GEOPHYSICS TO
HAZARDOUS WASTE INVESTIGATIONS
ROBERT M. WHITE
SIDNEY S. BRANDWEIN
Law Engineering Testing Company
Marietta, Georgia
INTRODUCTION
Geophysical techniques, originally developed from mineral ex-
ploration, have become the primary investigative methods in many
hazardous waste studies. The types of studies include the location
of dumps, trenches and spills, etc., the characterization of subsur-
face contamination, primarily leachate plumes, and the evaluation
of the geohydrologic environment in the vicinity of- existing or
planned hazardous waste facilities. Field investigations typically in-
clude remote sensing, drilling and the installation of monitoring
wells, as well as geophysics.
In this paper, the authors describe, through case histories and ex-
amples, the use of surface and borehole geophysics including elec-
trical, seismic, and magnetic techniques.
Surface geophysical techniques, due to their non-penetration of
the subsurface and their rapid execution, are ideally suited to the
detection of 'dumps, trenches, spills, etc. Spills usually alter
groundwater conductivity near the surface and are, therefore,
detectable with surface electrical surveys. Waste trenches can be
detected due to either the presence of fill replacing the natural soil,
the presence of trash such as barrels which may or may not be
related to the contaminant, or by the presence of the contaminant
itself. Geophysical techniques most commonly used to locate sub-
surface contamination sources are:
•Electrical surveys—
Resistivity surveys
Electromagnetic (EM) Surveys (conductivity)
•Ground penetrating radar
•Magnetic surveys
Electrical surveys are usually designed to measure the electrical
resistivity of subsurface materials by making measurements at the
surface. The resistivity method imposes an electrical field on the
survey area and, by measuring the surface expression of the
resulting potential field, calculates the resistivity of the subsurface
material.
The electromagnetic method induces an electric field in the sub-
surface from a cycled magnetic field at the surface. The resistivity
of the subsurface material is then measured by recording the
magnetic field induced at a surface coil by the time varying subsur-
face electric field and comparing it with the original magnetic field.
In both techniques, subsurface resistivities are computed from sur-
face measurements.
Resistivity and electromagnetic equipment used in waste in-
vestigations are portable and capable of rapid surveying. Practical
considerations dictate the choice of techniques (resistivity or EM)
and often both are used in the same survey.
Ground penetrating radar is a relatively new technique which has
strong application to the detection of waste containers, pipes and
trenches in hazardous waste investigations. Ground penetrating
radar offers the highest level of detail available from any surface
geophysical technique due to the high frequency energy used. Its
depth of penetration is shallow relative to other geophysical techni-
ques. Clay and conductive groundwater limit penetration depths.
Portable magnetometers are very useful in locating magnetic ob-
jects such as steel drums, car bodies, and pipes. The magnetic
survey measures the earth's magnetic field over a relatively small
area. Near surface magnetic objects are located by their perturbac-
tion of the earth's magnetic field.
CASE1
A recent project in the Coastal Plain involved evaluating an
abandoned oil reclamation facility for its impact on the ground-
water and designing a cleanup program. Some of the waste material
was exposed at the surface (drums and other trash), and oil covered
vegetation and lagoon surfaces.
The location of areas of fill where similar material has been
buried and the locatioan of barrels within that fill were investigated
by surface geophysics. Electrical resistivity and electromagnetic
surveys were used to delineate fill areas, and magnetometer and
metal detector surveys in conjunction with the surface electrical
surveys delineated concentrations of steel drums within the fill. The
existence of clayey surficial soils and high conductivity ground-
water precluded the use of ground penetrating radar at this loca-
tion. Two distinct areas were located: one area had abundant
metallic objects but did not appear to be a source of groundwater
contamination, and one area of no metallic objects but a very
definite source of groundwater contamination.
CASE 2
In the western U.S., in the Basin and Range Province, electrical
resistivity surveys were used to delineate waste trenches whose exact
lengths were not known. The target of the survey was not the con-
taminant itself, but the contrast of the electrical properties between
the in-place and re-worked intermontane sediments. A profile
perpendicular to the trenches that exhibit the apparent resistivity
values obtained by a gradient method that was used to outline the
trenches is shown in Fig. 1. Radar would have been successful here,
but at an increased cost.
C
11
\4
a
a
a
v/A\//A u---""— r°"
Trench Locations
Figure 1.
Resistivity profile across waste trenches in Basin and Range Province show-
ing location of major trenches. A gradient method was used to differentiate
between the in-place and reworked intermontane sediments. In this case,
the contaminant was not a good geophysical target; the difference between
disturbed and non-disturbed material provided the geophysical contrast.
91
-------
GEOHYDROLOGY
Figure 2.
Truck mounted geophysical logger which is capable of sophisticated func-
tions such as acoustic, density, porosity, and temperature logging as well as
conventional electric gamma and SP logging.
LiTMOLoar
• ••ICTANCC
7« T« »
A III
SCHEMATIC RESPONSE OF GAMMA RAY , S.P. AND
RESISTIVITY LOGS TO A COMMON GROUND WATER
ENVIRONMENT
Figure 3.
Typical down-hole log responses to hydrogeologic situations differentiating
sands from clays, and fresh water from salt water.
The evaluation of the hydrologic environment in the vicinity of
an existing or planned hazardous or municipal solid waste facility
or a waste impoundment, requires the evaluation of permeability,
porosity and water quality in the vicinity of the facility for input to
groundwater modeling or other analysis. Surface and subsurface
geophysics can optimally locate facilities by determining the con-
tinuity of confining layers and permeable zones and through ex-
trapolation of permeability and water quality. These techniques
reduce the cost of and the amount of borings needed and provide a
positive basis for extrapolation between them. Surface electrical
and seismic surveys and borehole electrical, nuclear, and acoustic
logging are commonly used.
Borehole geophysics has been widely used in the exploration for
groundwater during the last 20 years and in groundwater con-
tamination studies more recently. Correlation vertically within
borings and horizontally between them, estimation of formation
lithologies and the quantitative measurement of porosity,
permeability, and water quality are the prime groundwater uses of
borehole geophysics. Geophysical logging units in the groundwater
industry are designed for shallow' small diameter holes primarily
drilled in fresh water. The equipment is usually mounted in a small
truck and some very basic units are hand portable.
Figure 4.
Interpreted seismic refraction profile in the Piedmont showing cross-
sectional area of groundwater flow channel.
The seismic refraction technique determines the acoustic
velocities of subsurface materials which are generally a function of
the elastic moduli, saturation and porosity of subsurface materials.
A common groundwater exploration situation is that of low
permeability bedrock overlain by unconsolidated material. The ex-
ploration targets are the depth and configuration of bedrock and
depth of the water table, as well as an estimate of aquifer porosity.
Refraction seismology can be used in this situation because the
geologic condition yields a multi-layered velocity profile with
velocity increasing with depth.
CASE 3
The selection of geophysical techniques to characterize
hydrologic environments is primarily a function of the geologic set-
ting. In Fig. 4, a hydrologic cross section through a stream in the
Georgia Piedmont south of Atlanta is shown. In this type of
geologic setting, seismic refraction is best suited to characterize the
hydrologic environment because the degree of weathering and
water saturation strongly alters the seismic velocity of the material.
Electrical techniques would not be optimal here due to the relative-
ly high resistivity of the subsurface in general.
CASE 4 AND S
The hydrogeologic setting of shallow aquifers in the Gulf Coast
is characterized by unconsolidated sediments. The geologic cross
sections from the Gulf Coast (Fig. 5) can be considered reasonably
representative of coastal plain geologic settings in general. Surface
electrical methods are well suited in this geologic setting as the elec-
trical resistivity of the low permeability clays and the high
permeability sands contrast strongly.
In Case 4, in Louisiana, a blanket sand obscured the faulted
geology below and drilling without the aid of geophysics would
have been expensive and possibly inadequate. Case S, in Mississip-
pi, was one of highly permeable buried channels with no surface ex-
pression. Surface electrical surveys allowed them to be mapped
rapidly and avoided. Electrical and other borehole geophysical lop
were run in all borings made in conjunction with these surveys and
allowed for interpreted results which were better than those which
would have been obtained from surface or subsurface surveys
alone.
Borehole geophysical data allows the computation of hydrologic
parameters continuously in the borehole at a cost usually less than
that associated with sampling and laboratory testing. Other
hydrogeologic settings which have traditionally been explored by
means of geophysics are bedrock channels, salt water intrusion and
the location and water quality of deep groundwater in arid areas.
The monitoring of leachate plumes from groundwater con-
-------
GEOHYDROLOGY 93
a!
H
50 FOOT A SPACINGS
a.
30-
QUATERNARY SANO
WILCOX
SANO
WILCOX
SANO
CAIN
CLAY
NORTHERN LOUISIANA
10O FOOT A SPACINGS
b.
QUATERNARY SAND. SILT. AND CLAY 1
__, oo--:
"^CHANNEL SANDS AND GRAVELS/ J
loo-ioo-
TERTIARY CLAY
SOUTHERN MISSISSIPPI
Figure 5.
a. Resistivity profile outlining subsurface, faulted, sand and clay units in
Louisiana. No surface trace of faulting was visible.
b. Resistivity profile showing locations of buried channels in Mississippi.
No surface indication of these channels was evident.
tamination sources can best be carried out by surface conductivity
of groundwater, the configuration and concentration of the con-
taminant plume can be determined with surface resistivity or elec-
tromagnetic surveys combined with a few borings whose locations
are guided by the surface resistivity results.
Repeating an existing surface electrical survey in the vicinity of a
groundwater contaminant source can be used to monitor the move-
ment of the contaminated groundwater with time. This can provide
a valuable addition to long term monitoring of groundwater con-
tamination.
Figure 6.
Contaminant plume, outlined by electrical resistivity or electromagnetic
conductivity techniques, which might occur in an alluvial environment.
Figure 7.
a. Exposure of limestone in the Appalachian Plateau province exhibiting
large vertical joints which may affect subsurface flow.
b. Apparent Resistivity contour map produced at a site covered with 20 feet
of alluvial material, which outlines 2 vertical joints and a sinkhole. The big-
ger fracture had contaminant in it.
An idealized leachate plume characterized by a surface electrical
survey is shown in Fig. 6. The contaminant has lowered the
resistivity (or conversely raised the conductivity) of the ground-
water and thus the underlying sand itself. The result is that con-
tours of interpreted resistivity of the subsurface sand are also con-
tours of the degree of its contamination.
CASE 6
In the Appalachian Plateau Province, in an area of limestone
bedrock, underground contaminant transport may be controlled by
vertical joints in the rock. In Fig. 7, an exposure of limestone with
vertical joints and an apparent resistivity contour map produced
nearby is shown. Two fractures and one sinkhole were located
beneath about 20 ft of overburden. Subsequent drilling into one of
the fractures encountered contaminant and isolated the transport
mechanism of a plume.
CONCLUSIONS
Geophysics can play an important role in hazardous waste or
groundwater contamination studies. The advantages offered by
geophysics are its measurement of continuous in-situ subsurface
hydrologic properties and the economy derived from the reduction
in required drilling and associated laboratory testing. To achieve
these advantages, the overall exploration program should be
designed to use the exploration elements, typically including drill-
ing, remote sensing and geophysics, in a complementary way.
-------
CASE STUDY OF CONTAMINANT REVERSAL AND
GROUNDWATER RESTORATION IN A FRACTURED BEDROCK
R.M. SCHULLER
W.W. BECK, JR.
D.R. PRICE
SMC Martin Inc.
Valley Forge, Pennsylvania
INTRODUCTION
Groundwater contaminated with trichloroethylene (TCE) was
discovered in the vicinity of a manufacturing plant in southeastern
Pennsylvania where contamination had resulted from a series of
spills. Subsequent to this discovery, the plant contracted with SMC
Martin to study the extent and magnitude of the TCE contamina-
tion in the vicinity of the plant and to develop an appropriate
method for cleanup. SMC Martin's approach to this study included
the following principal objectives:
•Assessment of the extent and magnitude of TCE contamination
in the area
•Definition of the groundwater flow system controlling the trans-
port of TCE in the vicinity of the plant
•Development and implementation of a groundwater recovery
and restoration program
•Maintenance of a monitoring system to determine the effective-
ness of the recovery and restoration program
A rapid response to cleanup of the groundwater was desirable
due to the potential for contamination of residential wells in the
vicinity of the plant. However, the study was constrained by the
availability of locations for well placement and limited client
resources. As is typical of most industrial projects, the client and
the regulatory agency were interested in a cost effective and effi-
cient solution to the problem, which is rarely possible in the com-
plex hydrogeologic conditions of southeastern Pennsylvania.
However, the uniquely simplistic fracture system controlling
groundwater flow in the area resulted in a well-defined contami-
nant migration path which permitted a direct approach to contami-
nant reversal and groundwater restoration.
PHYSICAL SETTING
The plant is located in southeastern Pennsylvania and lies within
the Triassic aged, Newark-Gettysburg Basin. The property is drain-
ed by an intermittent tributary of a moderately-sized creek.
The area is underlain by the Lockatong Formation, a gray
argillite. The Lockatong is considered to be a poor aquifer due to
its low yield (4 to 40 gal/min with an average yield of approximately
7 gal/min).
The plant primarily depends on public water for its water supply
but also uses an in-plant well rated at 2 gal/min for cooling system
make-up water. Domestic water supplies adjacent to the plant are,
for the most part, low-yielding private wells.
FIELD INVESTIGATION
Prior to installing a monitoring well system, water samples were
collected from nearby residential wells (sampling points #27, #28,
#30, and #37 on Fig. 1), surface water (#24 and #26), the plant
discharge (#20), the plant well, and sumps used by the company to
remove groundwater that has infiltrated through the basement
floor of the plant. Analysis of these samples indicated that TCE ap-
peared to be restricted to the plant property.
A sample from the plant well had 4,700 /tg/1 TCE, considerably
above the Pennsylvania Department of Environmental Resources
(PaDER) recommended drinking water limit of 4.5 pg/l. Samples
collected from the domestic wells exhibited no evidence of TCE
contamination. It appeared that the plant well and the sumps
located in the basement of the plant were creating a cone of depres-
sion in the vicinity of the plant preventing the migration of TCE off
site.
At this time, inspection of topographic maps and aerial
photographs of the area indicated that a fracture trace existed in
the area extending northeast under the creek below stream sampl-
ing point #24 and southwest under the plant in the direction of
sampling point #37, a residential well (Fig. 1). To verify whether the
fracture extended under the plant, Monitor Wells #3 and #6 were
installed into the fracture zone and Monitor Wells #1 and #4 along
the flanks of the suspected trace. The wells were drilled in Oct. 1979
at the locations shown on Fig. 1. Due to problems encountered dur-
ing drilling, Monitor Well #5 was abandoned before completion
and Well #2 could not be drilled where desired due to problems of
accessibility.
KEY
MONITOR WELL
O STREAM .SAMPLING POINT
fj RESIDENTIAL WELL
fc-FBACTURE TRACE
Figure 1.
Map of the plant, sampling points, and the fracture trace
94
-------
GEOHYDROLOGY
95
All wells are open-hole completions, with 20 ft of 6 in. steel casing,
grouted to prevent surface water infiltration. Monitor Well #1
was drilled to a depth of 90 ft. Monitor Wells #3, #4, and #6 were
all drilled to a depth of 200 ft for potential use as recovery wells.
The first samples were collected from the monitoring wells in
Nov. 1979. A sample taken from Monitor Well #6 contained
230,000 /tg/1 TCE. Well #6 was installed at the location where TCE
had previously been delivered. This location is the principal source
of TCE contamination. TCE concentrations in samples from
Monitor Wells #1, #3, and #4 were 150, 430, and 2.1 /tg/1, respec-
tively.
During the Nov. sampling, two additional residences were
sampled (#31 and #37). They were selected due to their proximity
along the fracture trace and the likelihood that they would be the
most readily contaminated residential wells if contaminant migra-
tion was occurring along the fracture. The sample from Well #37
had no TCE; however, the sample from Well #31 contained 3,000
/ig/1 of TCE. The presence of TCE in Well #31 indicated a
hydraulic connection between this well and the plant as no other
possible source of TCE could be found in the area.
The residents using this well were immediately notified of the
contamination in their well and a duplex carbon filter was installed
in Dec. 1979. The residents were also provided with bottled water
for drinking between the period of notification and the date of in-
stallation of the carbon filter system.
TCE was also detected in stream samples collected in the vicinity
of the plant. The plant discharge at sampling point #20 contained
1,100 /tg/1 of TCE. The creek upstream from the plant (sampling
point #26) contained no TCE. The creek at sampling point #24
downstream from the plant discharge showed a decrease in TCE
levels (150 /tg/1) due to dilution by the stream.
Prior to initiation of a recovery program, Monitor Well #7 was
installed between the plant and Well #31 as near to the fracture
trace as possible to further define the hydraulic connection between
Monitor Well #6 and Residential Well #31. Routine monitoring was
initiated at a total of 18 sampling points to assess the effectiveness
of TCE recovery. Samples were collected weekly during the first
year, biweekly in the second year, and are currently being taken
once a month. Water levels in the monitor wells were taken during
every sampling round.
GROUNDWATER RECOVERY AND
RESTORATION
The potential for contamination of water supplies mandated that
a groundwater recovery and treatment program be initiated at the
plant. Treatment alternatives included granulated activated car-
bon, direct aeration, and counter-current air flow stripping. The
latter was chosen for its low operating cost and high efficiency.
: An air stripper is an engineered device which enhances the effi-
ciency of stripping volatile compounds from liquid streams. The
contaminated water flows downward by gravity over a porous bed
of ceramic, plastic or metal packing (typically 1/4 to 1 in.
diameter) and is contacted by an upward-flowing stream of air.
The stripper installation was completed with Monitor Well #6
used as the recovery well, and the system became operational in
May 1980 with an average pumping rate of 11 gal/min. Prior to
start up of recovery, depth to water measuresments in Monitors
Wells #1, #3, #4, #6, and #7 were 1.3, 14.5, 14.1, 12.4, and 1.74 ft,
respectively. It appeared from these measurements that Monitor
Well #3 was downgradient from the recovery well (Monitor Well
#6). However, the almost continuous pumping of the sumps in the
basement of the plant disturbs the natural hydraulic gradient and
makes it difficult to determine the actual direction of flow in the
immediate area of the plant. As mentioned earlier, it appeared that
the basement sumps were creating a cone of depression centered on
the plant.
After start up of pumping of the recovery well the depths to
water for Monitor Wells #1, #3, #4, #6, and #7 were 4.4, 67.2,18.1,
86.8, and 2.7 ft, respectively. Pumping from the recovery well af-
fected the water levels in Monitor Wells #1 and #4, but the effect
was only slight when considering the 52.7 ft drop in the water table
at Monitor Well #3. Also, Monitor Wells #1 and #4 are spaced only
50 ft to either side of Monitor Well #3.
Therefore, the fracture system as it is encountered by the
monitor wells is quite simple and Monitor Well1 #3 is in direct
hydraulic connection with the recovery well. The effect of pumping
from the recovery well on Monitor Well #7 was slight if not negligi-
ble. This was a result of not being able to place this well on the frac-
ture trace due to the inaccessible nature of the area along the trace
between the plant and the contaminated residential well. In fact,
the water level in this well indicates shallow groundwater is
discharging to the creek.
The water table elevations show that pumping from the recovery
well has resulted in a elongated cone of depression extending along
the fracture. To date, the pattern of water table elevations has con-
tinued with only minor overall changes due to prolonged drought
conditions in the area.
As mentioned previously, surface and groundwater samples were
collected to determine the effectiveness of the groundwater
recovery program. The results of TCE analyses on samples taken
from Monitor Wells #3, #6, and #7, the contaminated residential
well (#31) and the stream (sampling point #24) are plotted on Figs. 2
and 3. The TCE concentrations in samples from Monitor Wells #1
and #4 showed little change throughout the recovery operation and
KEY
- • RECOVERY WELL
--- • RESIDENTIAL WELL 31
TIME ran
Figure 2.
Plot of TCE concentrations versus time for samples from the
Recovery Well (#6) and Residential Well 031
KEY
Figure 3.
Plot of TCE concentrations versus time for Monitor Wells #7 and #3
and the stream at sampling point #24
-------
% GEOHYDROLOGY
were not plotted. The TCE concentrations for these samples re-
mained about 100 and 400 /ig/1, respectively, throughout the study
period.
Water samples from Well #31 have been obtained before, be-
tween, and after carbon filtration. This filter system effectively
removed the TCE from the water. The well water before filtration
contained in excess of 3,000 /tg/1 TCE. Subsequent to initiation of
pumping from the recovery well, TCE levels in Well #31 decreased
to 1,500 /ig/1 in Apr. 1980 and then increased to a peak value of
11,000 /ig/1 in Dec. 1980 (Fig. 2). From Dec. 1980 to May 1981,
TCE concentrations in Well #31 decreased to 4,500 /tg/1 then
steadily rose again to 7,500 /ig/1 in Sept. 1981. Since that time, the
TCE concentrations have decreased to their present levels.
The initial TCE concentration in the recovery well after pumping
was initiated was 68,000 /ig/1 in May 1980 and decreased to 2,000
/ig/1 by July 1980. With the exception of two periods of increasing
TCE concentrations in the recovery well—one peaking in Feb. 1981
and the other in Dec. 1981—the TCE concentrations in the
recovery well have steadily decreased.
The increases in TCE concentration in samples from the recovery
well can be correlated to the two prominent increases which occur-
red in samples taken from the residential well (#31). This shows that
prior to start-up of the recovery program, the most severe TCE
contamination had already migrated downgradient along the frac-
ture in a series of pulses past Well #31. The delay between the peaks
in the graphed results on Fig. 2 are a result of the time needed for
the TCE-contaminated groundwater to pass along the fracture
from Well #31 to the recovery well. The lesser magnitude of the
peaks in the graph of the recovery well is most likely due to dilution
by water being pulled into the recovery well from the vicinity of
Monitor Well #3.
TCE concentrations in samples taken from Monitor Wells #7 and
#3 and from samples collected from the creek at point #24 are plot-
ted versus time on Fig. 3. The plot of the TCE concentrations from
the creek approximately mirror those from Well #31 and the
recovery well with peak concentrations in Jan. and Dec. 1981. The
creek at this location serves as a discharge point along the fracture
and further substantiates the presence of an elongate cone of
depression drawing the plume back toward the recovery well. Dilu-
tion from the creek is responsible for the significantly reduced con-
centrations in these samples versus those taken from Well #31 and
the recovery well.
The plot of TCE concentrations for samples from Monitor Well
#7 show that the groundwater at this point had a TCE concentra-
tion of 2,700 /ig/1 prior to start up of the recovery program. Upon
initiation of recovery, the TCE concentrations in the samples from
this well decreased to about 20 /ig/1 in Aug. 1980 and then rose to a
series of peak concentrations in Jan. and Mar. 1981. Since that
time, the TCE concentration has decreased and remained at 20
/tg/1. Data from this well are difficult to interpret due to the well's
proximity to the creek. Although its peak concentration in Jan. and
Mar. 1981 reflects the peak seen in the plot of the concentrations
for the creek at point #24, it is difficult to determine if the variabili-
ty in the data is in response of the well's proximity to the creek.
The plot of the TCE concentrations from samples taken from
Monitor Well #3 shows an increase from initiation of pumping at
the recovery well until July 1981 when a peak TCE concentration
was observed. Since that time, the TCE concentration has de-
creased to its present level. This pattern indicates that prior to
recovery an additional plume had extended beyond Monitor Well
#3 indicates that groundwater is migrating in both directions along
the fracture from the plant. This may be due to the presence of a
groundwater divide under the plant or pumping in the vicinity of
Residential Well #37.
The air stripper had a TCE removal efficiency which continually
exceeded 99% up to June 1981. Since that time, no TCE has been
detected in the stripper effluent or in samples taken from the plant
discharge at stream sampling point #20.
CONCLUSIONS
Recovery and restoration of TCE-contaminated groundwater is
continuing at a plant site in southeastern Pennsylvania. The lateral
extent of groundwater contamination has been controlled by a sim-
ple fracture system extending from the plant. By pumping ground-
water from a recovery well which penetrates the fracture, it has
been possible to draw the contaminated groundwater back toward
the recovery well where TCE is being removed by a counter-current
air stripper.
The presence of two pronounced peaks of TCE concentration in
samples collected from several points along the fracture indicates
that the contaminant plume had migrated away from the plant in a
series of pulses. To date, cleanup and recovery of the contamina-
tion is continuing successfully.
-------
COST EFFECTIVE PRELIMINARY LEACHATE MONITORING
AT AN UNCONTROLLED HAZARDOUS WASTE SITE*
H. DAN HARMAN, JR.
Ecology and Environment, Inc.
Decatur, Georgia
SHANE HITCHCOCK
U.S. Environmental Protection Agency, Region IV,
Atlanta, Georgia
INTRODUCTION
The passage of the Comprehensive Environmental Response
Compensation and Liability Act of 1980, commonly referred to
as Superfund, brought about much needed authority for the mit-
igation of uncontrolled hazardous waste sites. The notification
requirements of this legislation concerning past waste disposal
practices caused a deluge of potential hazardous waste site iden-
tifications. In Region IV alone, there were in excess of 1300 notif-
ications submitted. This vastly increased the number of sites to be
evaluated under Suprefund accentuating the need for a quick,
reliable means of screening hazardous waste sites. The screening
process involves many criteria including a determination of the oc-
currence and movements of leachate. Leachate detection can be
accomplished by using indirect methods such as electrical resistiv-
ity or by using the direct method of drilling and sampling.
The constraints of the drilling and sampling method of leachate
monitoring are very apparent when the intent of the screening pro-
cess is to quickly identify the more serious waste sites. The sub-
contracting capability for well drilling operations, although avail-
able through Ecology and Environment, Inc., USEPA's Field In-
vestigation contractor, is often prohibitive for screening efforts
due to: (1) the time associated with subcontracting procedures,
(2) the high cost of well installation, and (3) the response time in
receiving analytical results. Electrical resistivity, an indirect subsur-
face exploration method, then becomes an attractive means of cost
effective preliminary leachate monitoring.
The advantages of the resistivity technique are: (1) the minimal
equipment cost, (2) the data produced are immediately avail-
able, and (3) with sufficient background information, the inter-
pretation can provide a high degree of reliability. These advan-
tages are illustrated by the recent application of resistivity tech-
niques at a hazardous waste site.
SITE HISTORY
The site was used as an industrial manufacturing facility from
the 1950s to the mid-1970s. After manufacturing operations ceased,
the site was leased and used as an unpermitted storage site for in-
dustrial wastes. When the site was discovered by USEPA there were
in excess of 2000 55-gal drums stored above ground at- the site
(Fig. 1). Analytical data from drum samples revealed the waste to
be spent solvents, sludges, and heavy metals. The drums, in various
states of deterioration, were scattered over several acres. Some
drums appeared full and sound while others were leaking or empty.
A partial listing of the numerous contaminants detected in drum
samples included the following compounds:
1,1-dichloroethane Naphthalene
1,1,1-trichloroethane Barium
*The authors are solely responsible for the views and opinions contained herein.
This paper is not an official statement by the U.S. Environmental Protection Agen-
cy nor is it an endorsement of the views and opinions expressed or implied on the
part of the authors.
Figure 1.
Location of Waste Storage
Chloroform Cadmium
Benzene Chromium
Phenol Mercury
Toluene Lead
Most of the runoff from the leaking drums followed a south-
westerly course away from the site along a gently sloping draw, evi-
denced by the stained soil and stressed vegetation observed along
the flow path. An isolated area of contamination also existed at
the abandoned homestead where approximately 100 drums had
been perforated and the contents allowed to drain out on the
ground.
Of potential significance to groundwater quality is the existence
of an abandoned oil and gas well. The well proved unsuccessful but
did produce, under flowing conditions, water reportedly contain-
ing hydrogen sulfide gas. The well was finally plugged with cem-
ent grout prior to any drum storage on the site. The well water is
suspected of impacting the groundwater quality of this site.
After the enactment of Superfund, federal authority was granted
for a planned removal action at the site. All drums and a majority
of the contaminated soil were removed and disposed of at an ap-
proved facility. However, there remained the concern that contam-
inants from the leaking drums had contaminated the local aquifer.
A highly productive well field, serving from 65,000 to 75,000 peo-
ple in the surrounding area, withdraws water from the aquifer less
than 5000 ft from the site.
HYDROGEOLOGY
The hydrogeology of the case site is relatively simple. The alluvial
type deposits which are common to the area are composed of clay,
97
-------
98
GEOHYDROLOGY
silt and fine-grained quartz sand which overlay quartz sand and
gravel with clay lenses. Shale bedrock is approximately 90 ft
deep.
The regional groundwater flow direction is generally toward the
southwest. Groundwater levels at the site encountered during lim-
ited USEPA shallow drilling and pit digging were 5 to 10 ft deep.
Localized groundwater flow directions may vary due to past ex-
cavations and fill operations and the outwash stream valley north-
east of the site.
RESISTIVITY SURVEYS
To efficiently determine the presence, and extent, of the sus-
pected leachate plume at the case site, resistivity surveys were per-
formed at various data points (Fig. 2). The first survey, conducted
while drums were on site, consisted of five soundings ranging in
proposed depths of investigation between 30 and 150 ft deep. To
Figure 2,
Location of Resistivity Data Points
insure that the resistivity measurements could be reliably inter-
preted, control readings were made at a previously logged well. A
good correlation was obtained between the modified Wenner array
sounding method,1 the driller's log, and the geophysical logs of the
abandoned oil and gas well (Fig. 3). For example, the sounding
indicated major cumulative resistivity slope changes at 20, 55, and
92 ft. Both the driller's log and the geophysical logs indicated
changes at comparable depths.
During the first survey profiles were conducted at 31 locations.
The standard Wenner array profile method2 was used with elec-
trode "A" spacings of 10 and 25 ft. Apparent resistivity values
outside the topographic draw southwest of the buildings ranged
from 116 to 257 ohm-ft whereas inside the draw the values ranged
from 11 to 105 ohm-ft. These low values indicate groundwater
contamination (Fig. 4). By comparing Fig. 1 and Fig. 4, the corre-
lation can be seen between waste storage at the head of the draw,
the once flowing sulfur well and low apparent resistivity values in
the draw. Assumed groundwater flow and surface-water drainage
from the head of the draw are also southwest along the draw. These
factors contribute to the preliminary evaluation that leachate is
moving from the waste storage area at the head of the draw south-
Figure 4.
Apparent Resistivity Map First Survey
Figure 3.
Comparison of Sounding, Geophysical Logs, and Driller's Log
Figure 5.
Extent of Groundwater Contamination 20 foot "A"
Spacing Second Survey
-------
GEOHYDROLOGY
99
west along the draw. This preliminary evaluation was supported by
a second resistivity survey after the drums were removed.
The second resistivity survey consisted of additional soundings
and profiles at 126 locations (Fig. 2). In order to detail the hor-
izontal and vertical extent of interpreted groundwater contamina-
tion, electrode "A" spacings of 20, 40, 60, and 80 ft were chosen.
The extent of contamination using electrode "A" spacings of 20,
40, 60, and 80 ft is shown in Figs. 5, 6, 7 and 8. Again, by com-
paring Fig. 1 with Fig. 5 the correlation between waste storage
areas and interpreted shallow groundwater contamination is very
apparent.
Using the 40 ft "A" spacing profile the interpreted contam-
ination corresponds with that at the 20 ft "A" spacing but is less
Figure 6.
Extent of Groundwater Contamination 40 Foot "A'
Spacing Second Survey
Figure 7.
Extent of Groundwater Contamination 60 Foot "A"
Spacing Second Survey
extensive. Using the 60 ft "A" spacing the interpreted contamina-
tion varies in size and direction due to probable variations in
alluvial lithology, permeability and localized groundwater flow
NOTEl THI» MAP HEPHEaENTS AN
INTEBPBETATION OP APPARENT
HESIITIVITV DATA.
ICOLOOT AND ENVIRONMENT
FIELD INVESTIGATION TEAM
• EOION IV ATLANTA
Figure 8
Extent of Groundwater Contamination 80 Foot "A"
Spacing Second Survey
patterns. Using the 80 ft "A" spacing the interpreted contamina-
tion is limited to zones directly underneath the homestead drum
storage area and the sulfur well.
CONCLUSIONS
The study described in this paper consisted initially of an evalua-
tion of site history, waste analyses, and hydrogeology. During
the study, no control wells were drilled, requiring that the resis-
tivity data interpretation be based upon existing subsurface data
and experience at other sites similar in history and waste types. The
first survey was completed in two days by a crew of three people.
After the data interpretation and drum removal another survey
was planned to detail the extent of the leachate. The second sur-
vey was completed in four days by a crew of three people.
As a preliminary method of leachate monitoring, resistivity
proved to be a very cost-effective method in screening this site.
By comparison, a drilling and sampling method to achieve the same
results would probably have cost several thousand dollars more.
The time delay in requesting drilling proposals, contract ap-
provals, actual drilling work and laboratory analyses would prob-
ably have delayed management decisions on future work by several
weeks. In addition, a second series of monitoring wells would
probably be required for the most effective monitoring program.
The locations of waste sources and interpreted leachate occur-
rences and movement based on the resistivity studies correlated
very well. The influence of the water from the flowing sulfur well
on the resistivity data is not definitely known but is believed to be
of great significance. By using the resistivity method the need for
monitoring wells was established and a limited number of wells can
be placed effectively. Once monitoring wells have been drilled and
sampled, further management decisions can be made to determine
the need for future tracking of the leachate.
REFERENCES
Carrington, T.J. and Watson, D.A., "Preliminary Evaluation of an Al-
ternate Electrode Array for Use in Shallow-Subsurface Electrical Re-
sistivity Studies," Ground Water. 19, 1981, 48-57.
Bison Instruments Incorporated. Bison Instruments Earth Resistivity
Meters Instruction Manual., Minneapolis, Minn., 1975.
-------
VADOSE ZONE MONITORING CONCEPTS AT LANDFILLS,
IMPOUNDMENTS, AND LAND TREATMENT DISPOSAL AREAS
L.G. EVERETT, Ph.D.
E.W. HOYLMAN
Kaman Tempo, Natural Resources Program
Santa Barbara, California
L.G. MC MILLION
U.S. Environmental Protection Agency, Environmental Monitoring Systems Laboratory
Las Vegas, Nevada
L.G. WILSON, Ph.D.
Water Resources Research Center, University of Arizona
Tucson, Arizona
INTRODUCTION
The Resource Conservation and Recovery Act of 1976 was en-
acted to promote the protection of public health and the environ-
ment through various regulatory, technical assistance and train-
ing programs in dealing with hazardous waste disposal site opera-
tions. Subpart M, 265.278, entitled "Unsaturated Zone Monitor-
ing," specifically defines vadose zone monitoring requirements for
land treatment.
A major concern at all hazardous waste disposal sites, includ-
ing abandoned, active, and planned sites, is the possibility of
polluting an underlying groundwater system. Because of this con-
cern, requirements were included in the Hazardous Waste and
Consolidated Permit Regulations, issued on May 19, 1980. The
regulations require a minimum of four groundwater sampling wells
at impoundment, landfill, waste pile, and land treatment sites.
The regulations also stipulate the location of such wells relative to
site boundaries and specify the parameters to be determined on
water samples. Vadose zone monitoring (i.e., "leachate" monitor-
ing, as defined in the regulations) is required only at land treat-
ment areas. The rationale generally used for excluding vadose zone
monitoring at impoundments, landfills, and waste piles is as
follows: (1) the primary monitoring tool in the vadose zone is
the suction lysimeter, (2) suction lysimeters provide only point
samples, (3) suction lysimeters cannot be installed in existing fa-
cilities without removing the waste deposits, and (4) suction lysi-
meters tend to clog.
The regulations do not at this time address the possibility of
alternative vadose zone monitoring methods at hazardous waste
impoundments, landfills, waste piles, and land treatment areas.
In actuality, a host of alternative methods are available.' By judic-
iously selecting from these methods, an effective vadose zone
monitoring system could be assembled at impoundments, land-
fills, and olther waste disposal sites. Operation of such a system
would lead to an "early warning" of potential pollution and allow
for the initiation of remedial measures. An early warning system
is desirable in regions where the vadose zone may be hundreds of
feet thick and the travel time of pollutants may be in the tens of
hundreds of years. In such regions, when samples from monitor
wells indicate the presence of pollutants, the groundwater system
will have been essentially destroyed. An early warning system is of
equal or even greater importance for regions underlain by shallow
potable systems because of the short travel time and reduced po-
tential for pollutant attenuation.
VADOSE ZONE DESCRIPTION
The geological profile extending from ground surface to the
upper surface of the principal water-bearing formation is called
the vadose zone. The term "vadose zone" is preferable to the
often-used term "unsaturated zone" because saturated regions are
frequently present in some vadose zones.2
The topsoil is the region that manifests the effects of weather-
ing of geological materials, together with the processes of eluvia-
tion and illuviation of colloidal materials, to form more or less
well-developed profiles.3 Water movement in the topsoil usually oc-
curs in the unsaturated state, where soil water exists under less-
than-atmospheric pressures. A great deal of literature on the sub-
ject is available in periodicals and textbooks. Within the topsoil,
saturated zones may develop over horizons of low permeability.
A number of references on the theory of flow in perched water
tables are available.4'5 Soil chemists and soil microbiologists have
also attempted to quantify chemical-microbiological transforma-
tions during soil-water movement.6'7
1
~ I
lAZAHOOUS WASTE DISPOSAL
SATURATION LAVIfl
-I !
1
I I
Figure 1.
Vadose zone including saturated and unsaturated flow
Weathered topsoil materials gradually merge with the underlying
earth materials. The zone beneath the topsoil and overlying the
water table, in which water in pore spaces coexists with air, or in
which the geological materials are unsaturated, is known as the
vadose zone (Fig. 1). Perched water tables may develop above in-
terfaces between layers having greatly different textures. Saturated
conditions may also develop beneath recharge sites as a result of
prolonged infiltration. In contrast to the large number of studies on
water movement in the topsoil, parallel studies in the vadose zone
have been few. The term "no-man's land of hydrology" was
coined to describe the limited knowledge of this zone.'
CATEGORIZATION OF VADOSE ZONE
MONITORING METHODS
A vadose zone monitoring program for a waste disposal site in-
cludes premonitoring (preoperational) activities followed by active
(operational) and post-closure monitoring programs. Basically,
premonitoring activities consist of assessing hydraulic properties
of the vadose zone, specifically storage and transmissive prop-
100
-------
GEOHYDROLOGY
101
erties, and the geochemieal properties relating to pollutant mobil-
ity (Table 1). The results of the premonitoring program will pro-
vide clues on the potential mobility rates of liquid-borne pollu-
tants through the vadose zone, the storage potential of the region
for liquid wastes, and the likelihood that specific pollutants will be-
come attenuated. Premonitoring activities are discussed in detail in
a separate report.' A premonitoring program will also provide val-
uable information for the design of a vadose zone monitoring
system.
The active and post-closure monitoring programs will comprise a
package of sampling and nonsampling methods selected from an
array of possible methods. Sampling methods provide actual liquid
or solid samples from the vadose zone, whereas nonsampling
methods provide inferential evidence about the movement of
liquid-borne pollutants. Complete descriptions of sampling and
nonsampling methods for active and post-closure monitoring pro-
grams have been developed.'"
CRITERIA FOR SELECTING ALTERNATIVE
VADOSE ZONE MONITORING METHODS
A guiding principle for selecting methods for a monitoring pro-
gram (premonitoring/active/post-closure monitoring) was aptly
stated as follows: "...for an efficient, long-term operation of an
I
operational monitoring network, the devices to be used must be
simple enough to be used by trained but not educationally skilled
personnel."" If the device meets these criteria and also if it is in-
expensive, one does not need to look further. In practice, selec-
tion of a method from a group of alternatives is governed by addi-
tional site-specific and function-specific requirements. The 14 cri-
teria for selecting alternative vadose zone monitoring methods are
given in Table 2.
CONCEPTUAL VADOSE ZONE
MONITORING DESCRIPTIONS
Development of a groundwater monitoring program for a haz-
ardous waste disposal facility is an integrative process. The pri-
mary elements to be considered, e.g., categorization of waste,
waste disposal methods, hydrogeologic setting, and monitoring
equipments, have been discussed.10
In general, site and waste characteristics will dictate the disposal
method and thereby suggest the most effective monitoring program
for a given location.
LANDFILLS
A typical landfill is constructed by using either the area or trench
method. With the area method, waste is deposited directly on the
Table 1.
Premonitoring of the Vadose Zone at Hazardous Waste Disposal Sites.
Property
Purpose of Monitoring
Approach
Alternative Methods
1. Storage
Z. Transmission of liquid wastes.
a. Flux.
b. Velocity
3. Pollutant mobility.
1. To determine overall
storage capacity of
the vadose zone.
2. To determine regions
of potential liquid
accumulation (perched
groundwater).
1. To determine Infil-
tration potential.
Z. To estimate percola-
tion rates in vadose
zone.
1. To estimate the flow
rate of liquid pollu-
tants in the vadose
zone.
1. To estimate the mo-
bility of potential
pollutants in the
vadose zone.
1. Relate storage capacity to
depth of water table or depth
to confining layer.
2. Estimate available porosity.
1. Characterize subsurface
stratigraphy.
Z. Locate existing perched
groundwater zones.
Measure infiltration rate in
the field.
Measure unsaturated hydraulic
conductivity for use in
Darcy's equation.
Measure or estimate saturated
hydraulic conductivity for use
in Darcy's equation.
a. Use core samples or grain-
size data.
b. Measure saturated hydraulic
conductivity in shallow
regions.
c. Measure saturated hydraulic
conductivity in deep
regions.
l". Estimate from field data on
flux.
Z. Tracer studies.
1. Characterize solids' samples for
properties affecting pollutant
mobility: cation exchange ca-
pacity, clay content, content of
hydrous oxides of iron, pH and
content of free lime, and sur-
face area.
2. Estimate from laboratory or
field testing using liquid
wastes.
1. Examine groundwater level maps.
2. Measure water levels in wells.
3. Examine drillers' logs for depth
to water table or confining layer.
4. Drill test wells.
1. Estimate from grain-size data.
2. Drill test wells and obtain drill
cuttings.
3. Neutron moisture logging (avail-
able porosity « total porosity-
water content, by volume).
1. Examine drillers' logs.
2. Drill test wells and obtain sam-
ples for grain-size analyses.
3. Natural gamma logging.
1. Examine drillers' logs.
2. Drill test wells.
3. Neutron moisture logging.
1. Inflltrometers.
2. Test plots.
1. Instantaneous rate method.
2. Laboratory column studies.
1. Permeameters
2. Estimate from grain-size data
using a catalogue of hydraulic
properties of soils.
1. Pump in method.
2. Air entry permeameter.
3. Infiltration gradient.
4. Double tube method.
1. USBR open end casing test.
2. USBR open hole method.
3. Stephens-Neuman method.
1. Use flux values obtained as above;
divide flux values by water con-
tent values at field capacity.
1. Field plots, coupled with a depth-
wise sequence of suction samplers,
using conservative tracer.
1. Obtain solids' samples (e.g., by
drilling test holes) and conduct
standard laboratory analyses.
1. Batch testing.
2. Column studies.
3. Field plots.
-------
102
GEOHYDROLOGY
Table 2.
Crilerti for Selecting Alternative Vadose Zone Monitoring Methods.
Item Criteria
I Applicability to new, active, or abandoned sites
2 Applicability to laboratory or Held usage
3 Power requirements
4 Depth limitations
5 Multiple use capabilities
6 Data Collection system
7 Possibility of continuous sampling
8 Sample/measurement volume
9 Reliability and life expectancy
10 Degree of complexity
11 Direct versus indirect sampling/measurement
12 Type of media
13 Effect of sampling/measurement on flow regime
14 Effect of hazardous waste type on results
ground surface. Run-on must be diverted away from the active por-
tion of the landfill. Run-off from the facility must be collected and,
if it is a hazardous waste, treated accordingly as specified under
the RCRA, Part 261. If the landfill material is subjected to dis-
persal by wind, it must be covered or otherwise managed so that
hazardous waste is controlled. With the trench method, waste is de-
posited at one end of a trench and covered at the end of the day as
required. For this example, it is assumed that hazardous wastes are
disposed of using the trench method.
During construction of the facility, data on the vadose zone at
the site may have been compiled and should be reviewed as part of
the premonitoring effort. This information, along with Soil Con-
servation Service (SCS) soil maps and well cuttings from geologic
formations penetrated by monitoring wells, should be examined to
determine the thickness, structure, and chemical characteristics of
the vadose zone. Water table levels should be plotted to determine
existing hydraulic gradients and the thickness of the vadose zone.
If vadose zone samples have been saved from site development,
several laboratory techniques are available for estimating hydraulic
conductivity (K). Saturated K values can be determined using
permeameters and from the resultant K values, assuming the hy-
draulic gradients are unity, the flux at the site can be estimated
from Darcy's equation. Saturated K values for different layers
can also be determined using the relationship between grain size
and K developed from grain-size distribution curves. Additional
laboratory methods for measuring unsaturated K values include the
"long-soil column," pressure plate methods, and other column
techniques.
Storage potential for water-borne pollutants can be inferred
from storage coefficients developed from pump tests in perched
water systems within the vadose zone. These data can be gener-
ated at minimal cost if wells and observation piezolmeters are
available from preliminary work at the site. Aerial photographs
should be reviewed for evidence of springs and seeps caused by
modifications of the water table under landfills and potential
threats to surface water quality at the site. Rainfall data and am-
bient surface water quality for the site should be known. Sur-
face structures designed to control run-on and run-off from the
waste material should be inspected to ensure compliance with the
RCRA.
Data on the chemical characteristics of the vadose zone are vital
in developing and understanding pollutant attenuation. In particu-
lar, the percentage of colloidal-sized particles, e.g., clay minerals,
and pH of a representative soil-moisture extract should be deter-
mined. In porous geologic materials, these particles can exchange
ionic constituents absorbed on the particle surfaces. The nature of
the surface charge of these particles is a function of the pH. At high
pH, a negatively charged surface occurs, while at a low pH, a
positively charged surface is developed. The tendency for adsorp-
tion of anions or cations is therefore dependent on the pH of the
soil water solution found in the vadose zone. This information will
be useful in estimating pollutant attenuation based on expected
leachate constituents. In addition, the redox (oxidation-reduction)
potential or Eh of perched groundwater, soil-moisture extract,
and/or leachate should be measured. With the pH and Eh of the
vadose zone known, Eh-pH diagrams can be constructed showing
stability fields for major dissolved species and solid phases. These
diagrams are useful in understanding the occurrence and mobility
of minor and trace elements. Construction of the Eh-pH diagram
has been described in detail.12'13
Ambient groundwater quality is an important data base to estab-
lish during the premonitoring effort. At old existing sites, care must
be taken to ensure that water data unaffected by the landfill have
been sampled to determine background quality. Of particular inter-
est as indicator parameters are TDS, COD, conductivity and BOD5
that have been found in high concentrations in solid waste leach-
ate studies. In addition, temperature, color, Cl, and Fe are listed as
indicator parameters."
The continued operation of a landfill either by the area or trench
method requires additional land for disposal of new waste material.
Therefore, each site is composed of a combination of existing and
projected waste cells. Monitoring activities for these areas will
differ in that the cost effective placement of a specific type of mon-
itoring equipment for new cells may not be possible for existing
cells. For example, a resistivity network installed under the protec-
tive liner of the landfill to determine leachate migration through the
liner could be easily incorporated in the earthwork required for
development of the new waste cell. This type of installation would
not be possible under an existing cell. Segregation of wastes by tox-
icity may dictate alternate monitoring networks around selected
waste cells. Depending on the toxicity level of the waste material
and its vertical proximity with groundwater aquifers, the use of
nondestructive aquifer monitoring methods (i.e., geophysical tech-
niques) may be preferable to sampling methods that require bore-
holes through the potentially polluted strata. These boreholes have
been known to short circuit liquid-borne wastes to the water table
through the annular well space of an improperly completed mon-
itor well.
Figure 2.
Generic monitoring design for existing hazardous waste landfill
A generic monitoring design for an active/new hazardous waste
landfill is shown in Fig. 2. Elements of the design include utiliza-
tion of both nonsampling and sampling methods. Nonsampling
methods include: (1) neutron moderation probes, (2) tensiometers,
(3) a resistivity network underlying new waste cells, and (4) surface
and borehole. All of these geophysical methods could be used as
required for detection or definition of pollutant plumes. Sampling
methods include: (1) multiple completion wells, (2) multilevel
samplers, (3) suction samplers, (4) piezometers, and (5) gas samp-
lers. Technical description, field implementation, and range of ap-
plication and limitations of these techniques have been discussed.15
-------
GEOHYDROLOGY
103
SURFACE IMPOUNDMENT
Unlike the preceding generic landfill model where existing and
projected waste cells were evaluated for one facility, two separate
cases for generic surface impoundments were examined: Case 1, a
new unused facility (see Fig. 3) and Case 2, an active site." This
format was necessary because surface impoundment operations do
not include the subsequent disposal of new waste materials on addi-
tional lands. Following construction of the impoundment, no addi-
tional land is required until the facility is closed and a new site
developed. Therefore, access for instrument emplacement differs
between the two sites. A premonitoring effort was undertaken to
characterize the waste water and vadose zone properties. The con-
ceptual program included surface water, vadose zone and saturated
zone monitoring techniques.
To accommodate measurement of intake rates in the lagoon, a
stilling well with water stage recorder was mounted on the end of a
platform. During operation of the lagoon, intake rates can be
determined by the instantaneous rate methods, in which water-
level declines, measured in the lagoon during a brief shutdown
period, are related to volumetric relationships for the lagoon.
Figures.
Water quality monitoring design for a new surface impoundment
Sampling units were installed beneath the base of the lagoon, as
shown in Fig. 3, to sample percolating water in case of failure at
the liner. The sampling units were, of course, installed before lay-
ing down the liner. The sampling units selected for this package
were filter candles, laid horizontally. Filter candles contact a larger
area than point samplers, such as lysimeters. The individual units
were laid within sheet metal troughs. Because of the importance of
detecting wastewater movement below a failed liner, back-up units
were installed beneath the shallower units. In addition, other units
were installed at the same depths at other locations in the pond.
Such duplication can be regarded as "planned redundancy," to
avoid the sampling problems described earlier'in this paper. Inlet
and outlet lines from the samplers were terminated in an above-
ground shelter containing sources of vacuum-pressure, sample bot-
tles and other appurtenances. Provisions are included to permit ob-
taining either discrete or continuous samples.
The second "line of defense" for direct sampling of percolating
wastewater consists of monitoring wells installed in the perched
groundwater body. These wells are particularly valuable in the
event that water movement beneath the pond occurs at matric pres-
sures below the limit of the filter candles (i.e., about -0.8 atm).
Ideally, these monitor wells should be of large enough diameter to
permit the installation of permanent submersible pumps.17 Samp-
ling should be initiated as soon as the wells are completed to ob-
tain baseline water quality values.
Indirect methods selected for detecting the percolation of waste-
water from the lagoon consist of tensiometers, heat-dissipation sen-
sors, and access wells for neutron moisture logging. The tensio-
meters and heat dissipation units are useful in estimating both stor-
age changes and hydraulic head gradients. In addition, the tensio-
meters will indicate the appropriate negative pressure to apply to
the filter candles during sampling, to avoid affecting unsaturated
flow paths. The heat-dissipation sensors provide information on
hydraulic gradients at negative pressures below the failure point
of tensiometers. The selection of heat-dissipation sensors in this ex-
ample was arbitrary. To conform with the principle of "planned
redundancy," batteries of psychrometers/hydrometers and/or
electric resistance blocks also could be installed. The tensiometer
readings will be recorded manually by a field technician. However,
the signal from the other matric-potential sensors could be auto-
matically recorded.
Neutron moisture logging in the access wells will show the lateral
spread of wastewater in the vadose zone in case of liner failure,
provided that storage changes occur in the geologic profile. The
wells also can be used to monitor changes in groundwater levels
and for obtaining water samples from the vicinity of the water
table.
For active (and abandoned) lagoons it will not be possible to in-
stall monitoring units beneath the base of the impoundment. Con-
sequently, such units must be installed on the periphery of the facil-
ity. A stilling well and water storage recorder are installed to aid in
determining intake rates by the instantaneous rate method. A
cheaper technique would be to position stilling wells on the sides of
the lagoon. For active ponds it will be possible to estimate flux and
velocity in the vadose zone. Dividing this value by representative
water content values gives an estimate of velocity.
Sampling techniques consist of clustered suction cup lysimeters,
in a common borehole, and a perched groundwater well. Non-
sampling techniques consist of tensiometers and access wells for
neutron logging. As in the case of the new lagoon, it is advisable
to install other nonsampling units such as psychrometers/hygro-
meters for monitoring in the dry range, where tensiometers be-
come inoperative. A complete discussion of the rationale and meth-
ods used for Case 1 and Case 2 has been developed in an EPA re-
port."
LAND TREATMENT FACILITIES
As defined in the Hazardous Waste and Consolidated Permit
Regulations," a land treatment facility is "...that part of a facility
at which hazardous waste is applied onto or incorporated into the
soil surface." The objective of land treatment is to enhance the
microbial decomposition of waste pollutants, or to otherwise re-
tard their mobility by soil physical/chemical reactions. According
to some authors," land treatment of wastes involves the following
three steps following application and incorporation: (1) mixing the
waste wsith surface soil to aerate the mass and expose waste to soil
microorganisms, (2) adding nutrients or amendments (optional),
and (3) remixing the soil and waste periodically to maintain aerobic
conditions.
Techniques for selecting, managing, and operating land treat-
ment facilities were reviewed by several authors.20'19' |2Z USEPA
requirements for surface water control, record keeping, waste
analyses, monitoring, use at food chain crops, and closure are in-
cluded in "Standards Applicable to Generators of Hazardous
Waste".2' In contrast to the monitoring requirements for haz-
ardous waste impoundments and landfills, vadose zone monitoring
is required at land treatment areas. The designated techniques are
pore-water sampling and solids sampling.
-------
104
GEOHYDROLOGY
A typical land treatment facility is illustrated in Fig. 4. As shown,
the facility contains a lagoon for storing incoming wastes. Con-
ceivably, such a lagoon would be lined to minimize seepage and
would include monitoring facilities, such as described previously
for impoundments. The field is diked to prevent uncontrollable
run-off. Whatever run-off occurs is collected into a sump.
A IMETHUMINT IHf LTEfl
• MULTILIVCL i»M»L*lt
O ACCESS WCLL
Ct OKSEfWATION HELL
(2) installing nonsampling units. The water budget approach us-
ing soil moisture accounting is a simple, rapid method for estimat-
ing the volume of deep percolation below a selected soil depth.
The method developed by two researchers" is used most often and
a computer program designated WATBUG is available to simplify
calculations." Inflow components that must be measured include
precipitation and waste water application. Outflow components re-
quiring measurement include runoff and crop evapotranspiration.
The change in storage equals water content change in the depth of
interest. All terms are equated to deep percolation.
A generic assortment of monitoring units at a station at a land
treatment area is shown in cross section in Fig. 5. Elements of the •
design include: (1) an access well for neutron moisture logging,
(2) implantable electrical conductivity probes for detecting changes
in salinity, (3) tensiometers for measuring matric potential down
to -0.8 atm, and (4) thermal dissipation sensors for measuring
matric potential below -0.8 atm. Technical descriptions, limita-
tions, and field applications of these techniques have been des-
cribed."
To avoid interfering with field operations and to facilitate access,
each station should be located on a beam extending across the field
as shown in a plan view of the facility in Fig. 4.
Two access wells have been installed at the station, one shallow
and one deep (Fig. 5). In practice, a large number of shallow
( <10ft) access wells could be installed throughout the site to de-
termine changes in water content during application and drying
cycles. If it appears that all of the water content changes are oc-
curring in the shallow depth, deeper units may not be required,
However, if extensive deep percolation is occurring, deeper access
wells will be needed, perhaps extending to the water table. This
method is useful in determining site-specific water balances and
soil water flux. Moisture logging only accounts for changes in
water content, a capacity factor. However, water (and pollutants)
may flow through specific zones without a change in water con-
tent being registered on moisture logs. Thus, backup facilities are
required.
Figure 4.
Plan view of monitoring units for land treatment area
Land Treatment Premonitoring
In general, the premonitoring activities relative to land treat-
ment of hazardous wastes will be identical to those for site selec-
tion, e.g., characterization of wastes and vadose zone properties.
Ideally, these activities should be staged as follows: (1) identify
properties of wastes that will affect the mobility of pollutants in the
vadose zone, (2) identify properties of vadose zone solids that will
affect the mobility of waste pollutants, and (3) conduct tests to
evaluate waste/soil interactions promoting pollutant attenuation
in the vadose zone.
As indicated, the techniques specified by USEPA for vadose
zone monitoring at land treatment sites includes soil core and
soil-pore water monitoring. The monitoring plan requires that the
owner/operator must specify details on the depth of monitoring,
number of samples, the frequency of sampling, and the timing of
sampling when using these techniques. The generic monitoring pro-
gram presented here is based on the premise that the owner/oper-
ator has also chosen a mixture of nonsampling methods to assist in
determining sampling depth, frequency, and timing of samples us-
ing the two basic techniques. In view of the great cost involved in
analyzing for pollutants, this approach would be cost effective.
In addition, it is assumed that the owner or operator is suffic-
iently environmentally conscious that he has elected to use alterna-
tive sampling methods to back up the suction cup samplers that
are inoperative at soil-water pressures below -0.8 atmosphere.
Nonsampling Methods
Nonsampling approaches to monitoring at an active land treat-
ment area include: (1) conducting a water budget analysis, and
Figures.
Generic assortment of monitoring units at land treatment area
A depth-wise battery of implantable conductivity probes is in-
cluded. These units are based on the four-probe Wenner array for
detecting electrical resistivity in the field of measurement. Changes
in resistivity indicate that conductive fluids are moving past the
units. Alternative techniques that could have been used for detect-
ing changes in salinity include salinity sensors, portable conduc-
tivity probes, and a portable four-electrode array.
Tensiometer units are included for sensing the soil water matric
potential. Tensiometer readings are useful in determining the cor-
rect vacuum to be applied to suction cup samples (to avoid in-
fluencing the flow field). For soil-water pressure gradients that are
-------
GEOHYDROLOGY
105
detectable by a battery of units where the gravitational component
does not limit the matric potential readings, inexpensive hydraulic
switches can be used to attach many tensiometers to a single pres-
sure transducer." Deeper units will require the use of integral pres-
sure transducers. The latter units also have the advantage of rapid
response time and no aboveground components, except lead wires
running to a shelter.
Inasmuch as tensiometer units fail at negative pressures below
-0.8 atm, the generic system also includes a battery of thermal dis-
sipation sensors, operative in the dry range. Other units that could
have been used to extend the range of tensiometers for detecting
changes in matric potential include electrical resistance/capacit-
ance blocks, thermocouple psychrometers/hydrometers, and os-
motic tensiometers.26
An identical array of nonsampling methods to that shown in Fig.
5 could be installed at a new site. The principal difference would
be that the installation depth at access wells, conductivity probes,
tensiometers, and thermal dissipation sensors would be shallow,
say within 10 feet of the land surface. This approach would min-
imize the cost of installation. However, if it becomes evident that
fluids are moving beyond the sensing depths, installation of mon-
itoring equipment at greater depths will be required.
Sampling Methods
Sampling methods at an active site include soil sampling, pur-
suant to the requirements of USEPA,23 and the generic methods
depicted in Fig. 5. Soil sampling includes using hand augers and
samplers such as the Veihmeyer tube. Deeper methods require the
use of power equipment including flight augers and hollow stem
augers with wire-line samplers.27
Generic methods depicted in Fig. 5 include: (1) suction cup
samplers, (2) multilevel samplers in perched groundwater, (3) a
depthwise array of piezometer units, (4) an observation well, and
(5) a screened well point at the end of the access well.
Suction cup lysimeters are installed in a depth-wide battery to
ensure detecting the vertical movement of pollutants. For shallow
sampling depths, the simple vacuum-operated unit will be ade-
quate.
Deeper unit will require vacuum pressure or high-pressure
vacuum units. In lieu of suction cup units, filter-candle type
samplers can be used.
Multilevel samplers are installed in a shallow body of perched
groundwater. These units are useful for depth-wise sampling to de-
termine the vertical extent of a plume. A horizontal transect of
such units is useful in detecting the lateral dimensions of a plume.
They are also used to measure hydraulic gradients. As shown in
Fig. 5, some of the sampling points extend above the water table
to facilitate sampling during a water table rise (useful for collect-
ing pollutants "hung up" in the vadose zone). A battery of piez-
ometer units shown in the figure could be used in place of or to sup-
plement the multilevel samplers. Piezometers are useful for con-
ducting the so-called piezometer tests for determining saturated
hydraulic conductivity values.
A standard observation well is shown installed in a body of
perched groundwater. Such wells permit extracting large volumes
of perched groundwater for analysis. Results may be indicative of
the integrated water quality flowing through the vadose zone. As
such, these units may be more useful in estimating mass flux than
point samplers, such as suction cup lysimeters. Pumps installed in
the observation wells facilitate sampling. Pumps are also useful for
conducting pumping tests when determining the hydraulic proper-
ties of the vadose zone.
Finally, the deep access well shown in the figure includes a well
point for obtaining a water sample near the water table. A suitable
batch is required for sampling these units. The representative-
ness of such a point sample is perhaps questionable. However,
the results are useful in a qualitative sense.
The sampling methods shown in Fig. 5 are also applicable to a
new facility. However, a staged approach should be taken for in-
stalling sampling units. In other words, deeper units should be in-
stalled when evidence from shallow sampling (and nonsampling)
units shows that pollutants are moving deeper in the profile.
REFERENCES
1. Everett, L.G., Schmidt, K.D., Tinlin, R.M., and Todd, D.K.,
Monitoring Groundwater Quality: Methods and Costs, EPA-600/4-
76-023, USEPA, Environmental Monitoring and Support Laboratory,
Las Vegas, Nv., 1976.
2. Bower, H., Groundwater Hydrology, McGraw-Hill Book Co., New
York, N.Y., 1978, 480 p.
3. Simonson, R.W., "What Soils Are," Soil, The Yearbook of Agricul-
ture, The U.S. Dept. of Agric., 1957.
4. Luthin, J.M., ed., Drainage of Agricultural Lands, American Society
of Agronomy, Madison, Wi., 1957.
5. van Schilfgaarde, J., "Theory of Flow to Drains", Advances in
Hydroscience, 6, 1970, 43-106.
6. Rhoades, J.D., and Bernstein, L., Chemical, Physical and Biolog-
ical Characteristics of Irrigation and Soil Water, Water and Water
Pollution Handbook, Vol. 1, L.L. Ciaccio, ed., Marcel Dekker, Inc.,
New York, N.Y., 1971,141-222.
7. Dunlap, W.J., and McNabb, J.F., Subsurface Biological Activity in
Relation to Ground Water Pollution, EPA-660/2-73-014, USEPA,
Corvallis.Ore., 1973.
8. Meinzer, O.E., "Ground Water," Hydrology, Oscar E. Meinzer,
ed., Dover Publications, Inc., New York, N.Y., 1942, 385-477.
9. Fuller, W.H., "Premonitoring Waste Disposal Sites," Establishment
of Water Quality Monitoring Programs, L.G. Everett and K.D.
Schmidt, eds., 1979,85-95.
10. Everett, L.G., Wilson, L.G., and McMillion, L.G., "Vadose Zone
Monitoring Concepts for Hazardous Waste Sites," Ground Water,
20, May-June, 1982.
11. Vanhof, J.A., Weyer, K.U., and Whitaker, S.H., Discussion of "A
Multilevel Device for Ground-Water Sampling and Piezometric
Monitoring by J.F. Pickens, J.A. Cherry, G.E. Grisak, W.F. Mer-
ritt, and B.A. Rizto", Ground Water, 17, 1979, 391-393.
12. Cloke, P.L., "The Geochemical Application of Eh-pH Diagrams,"
J. Geol. Educ., 4, 1966, 140-148.
13. Guenther, W.B., Chemical Equilibrium: A Practical Introduction for
the Physical and Life Sciences, Plenum Press, New York, N.Y.,
1975.
14. USEPA, Procedures Manual for Ground Water Monitoring at Solid
Waste Disposal Facilities, Office of Solid Waste, EPA SW-611, 1977.
15. Everett, L.G., Groundwater Monitoring, General Electric Co. Tech-
nology Marketing Operations, Schenectady, N.Y., 1980.
16. Everett, L.G., Hoylman, E.W., and Wilson, L.G., "Vadose Zone
Monitoring Manual," Interim Report, EPA, Las Vegas, In press.
17. Schmidt, K.D., Personal Communication, 1982.
18. USEPA, Treatability Manual, Volume II: Industrial Descriptions,
Office of Research and Development, EPA-600/8-80-042b, 1980.
19. Ross, D.E., and Phung, H.T., "Soil Incorporation (Land Farming)
of Industrial Wastes," Toxic and Hazardous Waste Disposal, Volume
Four, R.P. Pojasek, ed., Ann Arbor Science, Ann Arbor, Mi., 1980,
291-308.
20. Phung, T., Barker, L., Ross, D., and Bauer, D., Land Cultivation
of Industrial Wastes and Municipal Solid Wastes: State of the Art
Study, Volume 1: Technical Summary and Literature Review, EPA-
600/2-78-140a, USEPA, .1978.
21. Huddleston, R.L., "Solid Waste Disposal: Land Farming," Indus-
trial Wastewater and Solid Waste Engineering, V. Cavaseno, ed
McGraw-Hill Co., New York, N.Y., 1980, 275-280.
22. Miller, D.W., Waste Disposal Effects on Ground Water, Premier
Press, Berkeley, Ca., 1980.
23. USEPA, Hazardous Waste and Consolidated Hermit Regulations
Federal Register, 45, 1980, 33066-33588.
-------
106
GEOHYDROLOGY
24. Thornwaite, C.W., and Mather, J.R., Instructions and Tables for
Computing Potential Evapotranspiration and Water Balance, Drexel
Institute of Technology Laboratory of Climatology, Publications in
Climatology, X, No. 3. Centerton, N.J.. 1957.
25. Willmott, C.J., WATBUC: A FORTRAN IV Algorithm for Cal-
culating the Climatic Water Budget, University of Delaware Water
Resources Center, Contribution No. 23, 1977.
26. Rawlins, S.L., Measurement of Water Content and State of Water in
Soils, Water Deficits and Plant Growth, T.T. Kozlowski, ed., Vol. 4,
Soil Water Measurement, Plant Responses, and Grading of Drought
Resistance, Academic Press, New York, N.Y., 1976, 1-55.
27. Kaufman, R.F., Gleason, T.A., EUwood, R.B., and Sinsey, G.P.,
"Ground Water Monitoring Techniques for Arid Zone Hazardous
Waste Disposal Sites", Ground Water Monitoring Review, 1981.
-------
MITIGATION OF SUBSURFACE CONTAMINATION
BY HYDROCARBONS
W. JOSEPH ALEXANDER
DONALD G. MILLER, JR.
ROBERT A. SEYMOUR
Law Engineering Testing Company
Marietta, Georgia
INTRODUCTION
Subsurface contamination by hydrocarbons is a recurring prob-
lem. Sources range from home-fuel tanks to terminals and trans-
continental pipelines. Subsurface releases may go undetected for
months or years. The technical literature from the United States
and Europe contains numerous references regarding the fate of
hydrocarbon in the subsurface.
Overviews of subsurface hydrocarbon contamination are pre-
sented by Schwillel and Dietz.2 A theoretical description of hydro-
carbon migration is given by J. Van Dam.3 Duffy" and M. Van Der
Waarden5 present analytical procedures and the results of labor-
atory experiments dealing with percolation and leaching of hydro-
carbons. Subsurface biodegradation of hydrocarbons is addressed
by Vanloocke.' Case histories involving hydrocarbons in ground-
water are presented by McKee,7 Williams and Wilder,8 Matis,9 and
Moein.10
An understanding of the various transport phenomena is bene-
ficial in developing initial working hypotheses for a given site.
Site-specific definition of the extent of contamination is ultimately
necessary, however, for implementing cost-effective mitigative
measures. In this paper, the authors describe two cases involving
hydrocarbon leaks from terminals which have substantially differ-
ent hydrogeological settings and consequently required different
approaches for evaluation and mitigation.
CASE STUDY—A
Background
Case A involved the subsurface movement of hydrocarbons in
a multiple-terminal area. The source of the hydrocarbon, a dis-
colored and apparently old gasoline, was not obvious. The
terminals are in a mixed industrial-residential setting where the
hydrocarbon was first detected as petroleum vapors from a sewer
line. Seepage of the hydrocarbon first appeared near a commer-
cial operation, causing a partial interruption of business. The
hydrocarbon subsequently entered a nearby stream.
"Terminal A" was held responsible for the clean up of the
stream, determining the source of the leak, and resolving the
problem. Terminal A is at a lower elevation than the other termi-
nals which reportedly had some leaks in the past. The sewer line
also entered the vicinity of Terminal A from these and other poten-
tial upgradient sources.
This problem occurred in an area which is characterized by
rolling hills, numerous streams, and a thick sequence of relative-
ly flat-lying sedimentary rocks. Exposed bedrock in the vicinity
is chiefly limestone of Ordovician age. The regional dip of rock
units in this area is generally to the northwest, but is locally mod-
ified by secondary folds. Flexing of the rocks during the forma-
tion of secondary folds has resulted in numerous joints and small-
displacement faults.
Methodology
The objectives of this project were to determine the potential
source(s) of the contamination and to implement corrective action
to control the hydrocarbons. The project evolved in the following
sequential activities.
The initial activities at the site concentrated on recovery of the
hydrocarbon from the streams. This cleanup operation was
followed by the installation by the owner of approximately 40
observation wells in an unsuccessful attempt to determine the
source of the contamination. Air percussion drilling techniques
were used and the wells were of a simple construction.
Subsequently a geologic reconnaissance was conducted in the
terminal area and a preliminary evaluation was made of the avail-
able data. Interviews were conducted with terminal personnel,
sewer authorities, and the owners of the commercial property.
The site was surveyed and a base map was prepared.
Test pits were excavated near some of the observation wells in
which hydrocarbon was encountered to obtain a better understand-
ing of the subsurface conditions. This activity was followed by a
resistivity survey and the placement of supplemental observation
wells and test pits in anomalous areas depicted by the survey.
These data were incorporated into the final site evaluation and
subsequent design of the mitigative measures.
In an area with well-lithified, fractured rock, the presence and
orientation of fractures are significant with respect to the flow of
fluids in the ground. Although fractures provide conduits for flow,
they are very thin and therefore difficult to assess through con-
ventional drilling. Geophysical methods which measure a total-
field condition within a known volume of ground are more effec-
tive in locating such features. The electrical resistivity method was
selected for the site to supplement the data provided by observa-
tion wells. Four resistivity profiles were aligned parallel to the
southern pr.operty boundary. Six profiles were parallel to the east-
ern property boundary.
The test pits used for observation of hydrocarbon occurrence
were excavated with a medium-sized backhoe. Because of the rela-
tively thin limestone beds, it was possible to break the rock and
rip up slabs of the limestone within the uppermost-weathered
zone. Deeper excavations for the recovery trenches required pre-
drilling on close spacings before excavation of rock could be ac-
complished.
Findings
Subsurface material beneath the site consists of two rock units
which are overlain by a thin layer of residual clayey soil and occa-
sional fill material. The contact between the two rock units ap-
pears to exist in the vicinity of Terminal A.
The lower rock unit is predominant at the site and consists of
blue-gray, medium-grained, moderately fossiliferous limestone.
These rocks have beds from one to eight inches thick, with mod-
erately well defined bedding planes. Thin layers of calcareous
shale or shaley limestone are also present.
107
-------
108 GEOHYDROLOGY
The overlying rock unit is finer grained and more thinly bedded
than the lower rock unit with little or no fossil content. The beds,
although generally planar, are very uneven. The rock material
within the individual beds is very dense, but the bedding planes
frequently contain thin layers of clay or shale.
The dip of observed joints in the rock units is nearly vertical.
The orientation of the jointing is consistent with regional patterns.
A zone of rock weakness was also observed in excavations
oriented approximately parallel to the strike of the rock. This zone
appeared to coincide with the contact of the rock units.
Overlying the consolidated rock is a layer of clayey soil which
represents residual material from in situ chemical weathering of the
underlying limestone. The soil ranges from less than two feet to
more than ten feet in thickness. The boundary between the soil
and the underlying rock is very distinct in areas where layers of
dense limestone form the uppermost rock and is gradational in
shaley zones.
Ground water beneath the site primarily occurs within bedding
planes and joints in the limestone bedrock. There is also a lim-
ited amount of water within the clayey soil, mostly at the inter-
face between the clay and the rock. The upper surface of the sat-
urated zone is under water table conditions. In the bedrock beneath
the site, the water flows through the more porous zones (bedding
planes, or joints), and the exact flow path of individual water
particles is influenced by the geometry of these discontinuities.
Flow paths may therefore be more contorted than in a homo-
geneous material.
The configuration of the water table is similar to that of the
ground surface at the site, with a general slope downward from the
northeast toward the southwest (Fig. 1). Groundwater discharges
into a nearby stream during periods of high groundwater levels.
The hydrocarbon has been observed in many of the wells and
excavations in the southern and eastern portions of the site. In all
locations where hydrocarbon could be observed it was flowing
from the bedding planes within the rock. Its occurrence within the
ground appeared to be limited vertically by the water table, upon
which the thin layer of hydrocarbon was floating. The overall
direction of contaminant movement was controlled by the config-
uration of the water table. The exact flow paths appeared to be in-
fluenced by the occurrence and orientation of discontinuities in
th«rock.
Apparent resistivity was calculated and plotted versus distance
along the profiles. A strong anomaly (high apparent resistivity)
was located in the southern profiles and suggested an alignment
roughly perpendicular to the southern property boundary. A high
apparent resistivity anomaly was also located on the profiles along
the eastern side of the property. The locations of peak readings
suggest a possible linear alignment trending almost perpendicular
to the eastern property boundary. These anomalies appear to be
associated with structural geologic features and coincide with the
occurrence of hydrocarbon.
Excavations were performed near the sewer line to determine if
hydrocarbons existed adjacent to these structures. The sewer line
consisted of a concrete pipe, buried at an approximate depth of
3ft below the road's upper surface. The line had apparently been
placed in a trench which was backfilled with the natural soil ma-
terials. Some of the southern sections of the line had been exca-
vated within bedrock.
Groundwater was observed in the sewer line backfill with no in-
dication of hydrocarbon in the northern section of the line.
Groundwater and hydrocarbons were observed in the soil backfill
and in the adjacent walls of the rock trench along the southern
section of the line. In plain view, the occurrence of hydrocarbon
forms a "U"-shaped pattern, concave to the south and generally
consistent with the directions of groundwater flow (Fig. 1). The
pattern of hydrocarbon contamination can be explained by the
presence of rock discontinuities (observed in excavations and in-
dicated by geophysical methods) and by the presence of the sewer
line. These subsurface features could provide a vertical and hori-
zontal path with a higher permeability than the surrounding rock
for contaminant flow.
The interpretation of the hydrocarbon occurrence indicates that
an on-site loading area was the likely source area for the product.
This interpretation does not preclude the possibility of other on or
off-site sources under different conditions. Hydrocarbons orig-
inating from this area could potentially move southeast and south-
west from a localized groundwater high, with preferential move-
ment within the higher permeability zones.
Solution
The installation of relatively shallow recovery trenches within
the bedrock at two locations was the most feasible technique for
preventing additional seepage exiting, or entering, Terminal A.
The trenches were also the most feasible technique for recover-
ing the existing petroleum products beneath Terminal A and the
adjacent property to the southwest.
The recovery trenches were constructed in general accordance
with the schematic drawing presented in Fig! 2. Variations in the
subsurface conditions (geology, groundwater, and the occurrence
of hydrocarbons) were anticipated at the site. These variations
resulted in minor modifications of the corrective action, plan. The
final details of the recovery trenches were established during the
field implementation to allow for on-site modifications. The im-
plementation was supervised by a hydrogeologist who was familiar
with the site conditions.
« MITEOHSCHEEHTO • -. tMIELUUFII1
ALLOW LIQUIDS TO FUSS •> ' • , '
Figure 1.
Layout of Case A
Figure 2.
Schematic Drawing of Recovery Trench
-------
GEOHYDROLOGY
109
Trench A was 20 to 30 ft long; its depth is from 6 to 8 ft with a
central sump 2 ft deeper (Fig. 2). The primary purpose of this
trench was to intercept hydrocarbons which were believed to exit
the site along localized geologic discontinuities observed in the
test pits. This trench location and orientation is also based upon
the results of the resistivity anomaly encountered in this area of
the site.
Trench B was 30 to 50 ft long and 12 to 14 ft deep. Its
primary purpose was to help recover hydrocarbons moving down-
gradient from the source area and from the adjacent commercial
property. The water level in the center of the trench will remain
at a specified elevation under pumping conditions to reverse the
flow of groundwater beneath the northern portion of the adjacent
property.
The overall effectiveness of the recovery trenches will be ascer-
tained by routine measurements of water levels in existing wells
and springs. Provisions for monitoring were also made for several
of the original excavations by the installation of standpipes prior
to backfilling.
CASE STUDY—B
Background
Case B involved the loss of hydrocarbons from a known source
at "Terminal B." More than 200,000 gal of a primarily non-leaded
gasoline entered the subsurface by an underground pipe line that
was damaged during routine maintenance (Fig. 3). The contamina-
tion has not migrated from the owner's property.
The site is located in a belt of Pre-Cambrian age metamorphic
rocks. The facility is underlain by deep residual soils that appear
to be derived from the in-place weathering of a mica-quartz
schist. The silty soils are relatively uniform and no structural
features are apparent from the boring data.
Methodology
The objectives of this project were to recover the usable hydro-
carbons and to prevent off-site migration. In the first phase of
study, nine observation wells were installed around the source area.
The purpose of the wells was to allow a preliminary estimate of
the area! extent and thickness of the contamination and to indicate
fluid-flow directions.
The wells were constructed of 4 in diameter PVC using man-
ufactured well screens with a slot size of 0.010 in. A sand pack was
placed between the borehole wall and the well screen. A bentonite
seal was placed above the sand pack and the remaining annular
space filled with grout. The screen lengths varied from 5 to 25 ft,
and well depths varied from 20 to 35 ft. Split-spoon samples and/or
auger cuttings were obtained from the borings to determine sub-
surface soil conditions.
The wells were developed with a pump or bailer and allowed to
return to static conditions before measuring the fluid levels in the
wells. In-situ permeability tests were then performed in selected
wells for preliminary estimates of contaminant migration rates.
During this period of study the owner independently installed a
12 in diameter well immediately adjacent to the hydrocarbon
source area to begin the recovery operation.
After completion of the first phase, it was determined that addi-
tional wells and aquifer testing would be required to more accur-
ately determine hydrocarbon extent, flow direction and aquifer
properties. A preliminary recommendation was to install a recovery
system near the downgradient hydrocarbon front. The data to be
obtained from the second phase would help determine the location
and spacing of the recovery system.
In the second phase, six additional observation wells were in-
stalled closer to the suspected product limits. The wells were con-
structed in a manner similar to that described in the first phase.
After analysis of the data obtained from the additional observa-
tion wells, a 6 in diameter well was installed just upgradient of the
contaminant front. This well was installed as a prototype recovery
well and used to perform an aquifer test in an area where the hydro-
carbon had a considerable thickness. Prior to testing, static fluid
levels were obtained in all the wells and a constant fluid-
level recorder installed in the observation well closest to the
pumped well.
The aquifer test was performed in several steps to obtain as much
useful information as possible within the time constraints of the
project. Eight observation wells were monitored at selected inter-
vals throughout the testing. During the first 48 hr the test well was
pumped at a constant rate of 0.25 gal/min. The well was then
pumped at 0.33 gal/min for 8 hr while the response of the eight
observation wells and the pumped well to the increased discharge
rate was monitored. The pump was then turned off and a recov-
ery test was performed over a period of 16 hr. After the recovery
test the pump was raised and lowered in the test well and pumped
at various rates to observe if any significant changes occurred in
the hydrocarbon to water ratio.
Findings
The subsurface at the site consists of residual soils that vary
from red-brown to buff, micaceous fine sandy to clayey silts.
The sand content is estimated to vary from 5 to 30% and the clay
content is generally less than 10%. Unweathered rock was not en-
countered in any of the borings.
The fluid depths (hydrocarbon or groundwater) varied from 10
to 20 ft below land surface. The thickness of hydrocarbon varied
from 15 ft near the source area to 12 ft in the well located 200 ft
downgradient from the source area (Fig. 4).
• WELL
'RELATIVEFLUID LEVEL • OCCURRENCE OF
CONTOUR. IN FEET PETROLEUM PRODUCT
MO .
RELATIVE ELEVATION
IN FEET
EXPLANATION
WELL LOCATION
i OCCURRENCE OF PETROLEUM PRODUCT
_ MO
RELATIVE ELEVATION
IN FEET
Figure 3.
Layout of Case B
Figure 4.
Profile of Subsurface Conditions
-------
110
GEOHYDROLOGY
A potentiometric surface map was constructed from the observa-
tion well data. Contaminant flow is from the southern end of the
site to the north-northeast. The shape of the hydrocarbon plume is
elongate and orientated to the northeast. The shape and preferred
orientation suggest that the physical properties of the soils are
anisotropic, which is supported by analysis of the aquifer test
data.
A sustained yield of slightly less than 0.25 gal/min could have
been maintained in the pumped well. The specific capacity of the
well was estimated to be less than 0.03 gal/min/ft. The transmiss-
ivity of the aquifer was estimated to average 350 gal/day/ft with a
storage coefficient of 0.09. The fluid levels in observation wells
within 25 ft of the pumped well were influenced during the test.
The fluid level in a well 70 ft from the pumped well was only slight-
ly affected. After the recovery test it was observed that intermit-
tent pumping in the test well produced a significantly greater
hydrocarbon to water ratio than by constant pumping.
Solution
A recovery well system consisting of five 6 in diameter wells was
installed in the vicinity of the test well. The test well was also util-
ized in the system. Recovery wells were spaced on centers 20 ft
apart on the basis of the hydraulic properties obtained from the
aquifer test, with an average depth of 30 ft. The wells will be
pumped to approximately 15 ft below static level and allowed to
recover within 10 ft of static before resumption of pumping. This
will maintain a minimum drawdown of 10 ft and allow for intermit-
tent pumping to obtain the optimum hydrocarbon to water ratio.
A recharge well system consisting of seven 4 in diameter wells
was installed downgradient of the recovery wells. Two of the exist-
ing observation wells were utilized in the system. The wells are
spaced on centers 20 ft apart with an average depth of 25 ft. The
purpose of these wells is to form a hydraulic barrier by injecting
water and thereby minimize downgradient migration of hydrocar-
bons.
The entire system will be observed for six months. Any modifica-
tions to the system or further corrective measures will be considered
at that time.
SUMMARY
The two cases of hydrocarbon contamination described in this
paper, as well as those of previous publications, illustrate that sev-
eral factors impact the approach to problem assessment and mit-
igative planning. These factors include: timing of leakage discov-
ery, type and age of hydrocarbons, rate of contaminant migra-
tion, hydrogeological setting, general site layout, and proximity
of adjacent utilities and structures to the contaminated area.
If the extent of hydrocarbon migration is such that the contam-
inant is an imminent threat to the environment, then the assess-
ment is directed toward immediate mitigation (as illustrated by
Case A). If product migration is slower there is often time for more
detailed evaluation and development of optimal mitigative meas-
ures (as in Case B).
The detailed in-situ testing utilized in Case B was not appro-
priate in Case A because of the shallow occurrence of the hydro-
carbons in a heterogeneous setting. Also, the wells and excava-
tions used in Case A were relatively inexpensive. Excavations to
examine the occurrence of hydrocarbon were not practical for
Case B. Likewise, the geophysical techniques utilized in Case A
were not appropriate for Case B because of the extensive network
of transfer pipes and electrical conduit in the area of interest.
The site conditions occasionally preclude the use of one or more
mitigative measures. For example, in Case A the heterogeneity of
the rock units and the necessity to minimize increases in ground-
water levels did not permit the use of hydraulic barriers. The use
of continuous trenches was unacceptable in Case B because of the
depth and thickness of hydrocarbons, the presence of buried util-
ities, and the proximity of settlement-sensitive structures to the
contaminated area.
REFERENCES
1. Schwille, Friedrich, "Petroleum Contamination of the Subsoil—A
Hydrological Problem", Joint Problems of Oil and Water Industries,
London: Institute of Petroleum, 1967.
2. Dietz, D.N., "Pollution of Permeable Strata by Oil Components",
Water Pollution of Oil, London: Institute of Petroleum, 1971.
3. Van Dam, J., "The Migration of Hydrocarbons in a Water-Bearing
Stratum", Joint Problems of Oil and Water Industries, London:
Institute of Petroleum, 1967.
4. Duffy, J.J., Peake, E., and Mohtadi, M.F., "Subsurface Persis-
tence of Crude Oil Spilled On Land And Its Transport In Ground-
Water", 1977 Oil Spill Conference Proceedings, New Orleans, La.,
Mar. 1977,475.
5. Van Der M. Waarden, Bridie, A.L.A.M., and Groenewoud, W.M.,
"Transport of Mineral Oil Components To Groundwater-I", Water
Research,!, 1971.
6. Vanloocke, R. Borger, R. De, Voets, J.P., and Verstraete, W., "Soil
And Groundwater Contamination By Oil Spills; Problems and Rem-
edies"; Intern. J. Environmental Studies, 8, 1975.
7. McKee, Jack E., Laverty, Finley B., Hertel, Raymond M., "Gaso-
line in Groundwater", Journal of the Water Pollution Control Fed-
eration, 44, 1972,293.
8. Williams, Dennis E. and Wilder, Dale G., "Gasoline Pollution
of a Ground-Water Reservoir—A Case History", Ground Water, 9,
Nov./Dec. 1971.
9. Matis, John R., "Petroleum Contamination of Ground Water in
Maryland", Ground Water, 9, Nov./Dec. 1971.
10. Moein, George J., "Containment, Treatment, Removal, Disposal and
Restoration of Large Volumes of Oil and Hazardous Substances in a
Land Site in Chattanooga, Tennessee", Proc. of 1980 National Con-
ference on Control of Hazardous Material Spills, Louisville, Ky, May
1980, 46.
-------
AN APPROACH TO INVESTIGATING GROUNDWATER
CONTAMINANT MOVEMENT IN BEDROCK AQUIFERS:
CASE HISTORIES
RICHARD G. DINITTO
WILLIAM R. NORMAN
M. MARGRET HANLEY
Ecology and Environment, Inc.
Woburn, Massachusetts
INTRODUCTION
Many hydrogeologic investigations of hazardous waste sites
have been focused on contamination in the unconsolidated ma-
terials above bedrock. A common investigative approach involves
the installation of groundwater monitoring wells to a depth just
below the water table. This approach has been utilized by investi-
gators who accept the hypothesis that organic chemical contam-
ination remains near the upper surface of an aquifer.
However, several studies have indicated that organic compounds
readily migrate to the bottom of the overburden aquifer regard-
less of their specific gravity or solubility. Furthermore, shallow
wells may be inappropriately located to take into account the
configuration of the bedrock surface which may play an important
role in channeling contamination.
With the discovery of thousands of hazardous waste sites across
the United States, the need to assess bedrock aquifers and their
effect on groundwater movement has taken on a new importance.
In New England, where many drinking water wells are screened in
bedrock, the bedrock aquifer cannot be ignored when investigating
the impact of a hazardous waste site. When planning remedial
measures for hazardous waste sites an adequate assessment of the
bedrock aquifer is necessary to properly design a program that is
capable of containing or removing contaminated groundwater.
Additionally, bedrock aquifers underlying proposed sites for
hazardous waste treatment facilities or disposal areas should be
properly assessed to determine the possible impact the site may
have on groundwater resources. Therefore a thorough assessment
of the entire groundwater regime is required when investigating
hazardous waste sites or planning for future hazardous waste treat-
ment or storage facilities.
This paper was written in response to the need to more ade-
quately assess the relationships of bedrock aquifers to unconsol-
idated overburden aquifers. The approach presented here is based
upon Ecology and Equipment, Inc.'s (E & E) field experience at
more than 100 hazardous waste sites in England. This experience
has shown that bedrock plays a significant role in the migration
of contaminated groundwater.
The purpose of this paper is to illustrate the failure of some
common approaches to investigating hazardous waste sites, to
stress the importance of assessing the potential for bedrock con-
tamination, to outline an approach for assessing the bedrock
aquifer regime, and to present brief case histories which illus-
trate the successful employment of E & E's approach to investi-
gating bedrock aquifer contamination.
EVALUATION OF SOME COMMON APPROACHES
TO GROUNDWATER MONITORING
Commonly, hydrogeologic investigations of hazardous waste
sites have been conducted with little or no emphasis on assessing
contaminant migration at depth and within bedrock or the hydraul-
ic relationship between the bedrock and overburden aquifers. Mon-
itoring wells are often screened at depths that are too shallow or
in unsuitable locations to detect contaminants within the, lower
portion of an aquifer system. Several reasons why wells are im-
properly installed or misplaced are discussed here. _
In the past, monitoring well networks have been developed based
on the theory that volatile organic compounds, some of which are
less dense than water, remain at the top of the aquifer. However,
E & E's field experience and analytical data, as well as those of
other investigators,' indicate that contaminants are often forced
lower into the aquifer as the contamination migrates from its
source. Therefore shallow wells installed a distance from the source
may not intersect a contaminant plume.
A common practice for installing wells has been to use driller's
"refusal" as an approximation of the bedrock surface and the
depth for the well. In glaciated terrain, refusal commonly repre-
sents till or boulders. Although till is frequently located at or near
the bedrock surface, resistant tills can occur at any depth within
the overburden aquifer. Boulders in excess of one meter in diameter
are also common and may be mistaken for bedrock. Failure to
identify the true bedrock surface may result in wells that are too
shallow to monitor groundwater at the bedrock-overburden inter-
face.
Wells are also installed at inadequate depths when the presence
of till or a clay layer is assumed to act as an aquiclude that may
inhibit the migration of contamination to underlying portions of
the aquifer. Numerous studies1'2 have shown that under the proper
hydraulic conditions contamination can migrate through or around
till and clay. Therefore deep or fully penetrating bedrock wells are
needed when these geologic conditions exist in order to monitor
contaminant movement at all levels of the aquifer.
Inadequate assessment of the bedrock surface configuration and
structure (i.e., bedrock troughs or fractures that may channel con-
tamination) has led to the installation of deep or bedrock wells in
areas that are not in the proper location for detecting contamina-
tion. Finally the costs associated with the proper construction and
placement of deep bedrock wells have often influenced investiga-
tors to install shallow overburden wells.
APPROACH TO INVESTIGATING
BEDROCK CONTAMINATION
Common approaches used in establishing a monitoring well net-
work to investigate hazardous waste sites often do not accurately
assess the hydrogeologic conditions that may influence contam-
inant behavior within or near the bedrock aquifer. For this rea-
son, E & E has developed an approach which enables an assess-
ment of characteristics of the bedrock and overburden aquifers
which is cost-effective and allows a successful monitoring network
to be developed and implemented.
Aspects of this approach that are discussed below involve the
assessments of : 1) hydraulic relationships between the overburden
and bedrock aquifers, 2) the hydraulic characteristics of the bed-
rock aquifer, 3) the structure and configuration of bedrock, 4) the
natural groundwater quality in bedrock and 5) considerations for
well installation designed to assess bedrock contamination.
Ill
-------
112
GEOHYDROLOGY
Hydraulic Relationship Between the
Overburden and Bedrock Aquifers
In areas where an overburden aquifer overlies a bedrock aquifer,
the hydraulic relationship between the two aquifer systems needs to
be established. Three characteristics of the aquifer systems should
be considered: 1) the hydraulic conductivity of the overburden
aquifer mantling the bedrock, 2) the extent and thickness of the
overburden aquifer and 3) groundwater flow directions between
the bedrock and overburden aquifer systems.
The hydraulic conductivity of the overburden dictates the ease
with which contaminants will migrate into bedrock. A deposit ex-
hibiting low conductivity, such as glacial till or clays in contact with
the bedrock surface, may impede or prohibit contaminant move-
ment into bedrock.
The hydraulic conductivity(ies) of the deposit(s) at a site can be
extrapolated from existing information or measured directly in the
field. If the amount of hydrogeologic information available for a
specific site is minimal, hydrogeologic data published by the USGS,
state and local agencies can be reviewed prior to field work. Often
these publications give general conductivity values, or values can be
assigned from published figures for a particular deposit. Well logs
of nearby existing wells or for wells installed to monitor a site can
be reviewed to establish types of subsurface deposits. Approximate
hydraulic conductivities can then be assigned to each deposit using
published figures.
A more accurate determination of the hydraulic conductivity is
made during field work by one or a combination of field permea-
bility determinations, such as the slug test, the auger hole or piez-
ometer method. If possible the conductivity of the formation that
covers the bedrock should be determined. Conductivity values ob-
tained from field measurements may be verified to some extent in
the laboratory using permeability tests.
The lateral extent and thickness of the unconsolidated deposits
covering the bedrock determine if contaminants in the unconsoli-
dated aquifer are likely to move into the bedrock aquifer. Discon-
tinuous or thin deposits will allow contaminants to reach bedrock,
whereas deposits that are continuous and adequately thick through-
out the area of groundwater contamination will likely prevent con-
taminant movement into bedrock.
The extent and relative thickness of deposits can be determined
in many cases by a preliminary review of existing information con-
cerning the study area, and then verified by field work. In the field,
the extent and thickness of deposits are generally determined using
geophysical methods such as seismic refraction and/or resistivity
and direct methods such as soil borings/well installations.
Information generated from the preliminary literature review
and field work is used to construct geologic cross-sections through
the study area. The cross-sections can be used to predict the char-
acter and extent of the unconsolidated deposits.
Regardless of the hydraulic conductivity, and the extent and
thickness of deposits lying on the bedrock, an interchange of con-
taminated groundwater between the unconsolidated and bedrock
aquifers is not possible unless there is a net vertical downward
component of groundwater flow between the two aquifer systems.
Flow components are predicted prior to field work from a review of
topographic maps and existing information.
Hydraulic Characteristics of the Bedrock Aquifer
The hydraulic characteristics of the bedrock aquifer are deter-
mined to establish if the bedrock is water-bearing and capable of
transmitting contaminants and also to predict the velocity with
which contaminants will move through the bedrock. Hydraulic
characteristics of the bedrock that need to be determined are the
relative total porosity (primary, genetic and secondary) and hy-
draulic conductivity of the bedrock unit. Review of existing infor-
mation will provide preliminary data and additional inferences con-
cerning porosity and hydraulic conductivity values can be made by:
•Observing bedrock outcrops in the area of the site; noting bed-
rock type and frequency of fractures/faults in the bedrock
•Noting yeilds of bedrock wells in the vicinity of the study area
•Noting Rock Quality Designations (RQDs) measured during engi-
neering studies in the vicinity of the study area
Values derived from such observations may at best be rough es-
timates since the rock types observed or those in which measure-
ments were taken may not be continuous under the entire study
area. More accurate assessments of bedrock porosity and hydraulic
conductivity are determined in the field by measuring RQDs dur-
ing well installation and performing permeability tests in monitor-
ing wells installed and sealed in bedrock.
Assessment of the Structure and Configuration
of the Bedrock Aquifer
To insure the proper placement of bedrock monitoring wells, it
is important to first access the bedrock surface configuration and
structure. Bedrock wells which are placed without regard to these
features may fail to intercept plume migration along fault zones or
within bedrock troughs.
Both classical field techniques and more contemporary remote
sensing techniques are used to assess bedrock structures and surface
configuration. The methods include:
•A field observation and statistical evaluation of bedrock frac-
tures, frequency and orientation in outcrops
•Fracture trace analysis and aerial photographic interpretation to
determine predominant fault and fracture orientations
•Seismic refraction and electrical resistivity methods to determine
depth and orientation of the bedrock overburden interface, and
the presence of clay or till layer within an overburden aquifer
•Ground penetrating radar to determine the presence or absence
of boulders in the overburden at proposed well locations
Assessment of Natural Groundwater Quality in Bedrock
Concentrations of naturally occurring inorganic compounds and
metals within the overburden and bedrock aquifers must be
assessed to understand what effect a hazardous waste site has had
on the quality of the groundwater. Naturally-occurring organic
compounds (other than methane) are rare except in areas where
natural deposits of oil, gas or coal exist. The presence of organics
in groundwater is therefore assumed to be derived from a source
other than soil or bedrock in regions lacking fossil fuel deposits.
The primary source of inorganic compounds and metals is from
bedrock. The amount of any given compound will depend on rock
type, fault zones, ore deposits and degree of weathering. In New
England, where many sulfide mineral deposits occur along major
fault zones and in pegmatites, there may be naturally elevated con-
centrations of arsenic, antimony, cadmiurri, copper, lead and mer-
cury in the groundwater. A preliminary literature search should
provide information about an area's subsurface geology, faulting,
existence of ore deposits and mineralogy of the rock types present.
Determination of the extent of faulting and fracturing (see
Assessment of Bedrock Structure and Configuration) can give an
indication of: 1) the extent to which groundwater may flow through
the bedrock and thereby increase the weathering process, and 2) in-
dicate those areas where enhanced heavy metal deposits may exist.
The presence of these deposits can be assessed by field examination
of bedrock outcrops and thin sectioning of rock samples from out-
crops or well cores. Thin section analysis can also assist the inves-
tigator in determining the existence of major faulting if cataclasis
is present.
Comparison of groundwater analyses from existing upgradient
and downgradient off-site wells will provide data on the natural or
background levels of inorganics and metals in a given area. When
comparing analyses, it is important that the sampled wells be
screened over the same depths and in the same rock types and geo-
logic structures.
Analyses of groundwater from newly-installed monitoring wells
must also be carefully evaluated. Recent E & E analyses of both
new and old wells within similar rock types and geologic struc-
tures have shown higher concentrations (by 50 to 2000%) of the
-------
GEOHYDROLOGY
113
trace elements and heavy metals in the newest wells. The increased
levels in new wells may be attributed to the suspension of pulver-
ized bedrock and overburden materials as a result of drilling. The
use of hardened steel tools may have contributed to the increased
concentration of such compounds as vanadium and iron.
Well Installation for Assessing Bedrock Contamination
There is a wealth of information currently available on well in-
stallation procedures and techniques; therefore, an in-depth exam-
ination and review of existing procedures will not be discussed.
The most critical aspects of well installations related to evaluating
groundwater in bedrock will, however, be discussed here.
In designing a monitoring network that includes bedrock wells,
the purpose of the bedrock wells will determine the type and meth-
od of construction of the well.
Various criteria for locating bedrock wells must be evaluated so
that the objectives of the monitoring program can be met. These
criteria should include, but not be limited to the following:
•The monitoring well network should insure detection of contam-
inants that may be migrating along the bedrock surface or in bed-
rock fractures. This can be accomplished by one or more wells in-
stalled within possible pathways such as bedrock troughs or fault
zones. This requires prior knowledge of the bedrock surface con-
figuration and structure to assess possible contaminant migra-
tion directions.
•The bedrock monitoring well network should allow the natural
or background levels of inorganic compounds to be determined.
A minimum of one well should be installed in the same rock type
and geologic structure (i.e., fault zone) as where the contamina-
tion is.
•A minimum of one bedrock well, in conjunction with a co-lo-
cated overburden well, is necessary to establish hydraulic head
relationships between the bedrock and overburden aquifers, en-
abling vertical hydraulic gradients to be determined.
•Bedrock wells should be installed both within and out of major
fracture zones to adequately assess the variability of bedrock hy-
draulic characteristics.
A variety of drilling methods can be employed when drilling to
bedrock and subsequently into bedrock. Certain drilling considera-
tions to take into account include:
•The utilization of driven casing rather than hollow stem augers
enables the driller to telescope through boulders (after coring)
•A minimum of 10 ft of bedrock, coring should be allowed to es-
tablish bedrock competency and rock type
•The drive casing should be advanced as far into bedrock as pos-
sible to prevent caving in of the cored borehole. A roller bit can
be employed to assist in flushing out any bedrock materials that
may have fallen into the hole.
Three types of bedrock wells or well systems are suggested (Fig.
1). Each well should incorporate the use of Schedule 80PVC well
casing with threaded-flush joints to prevent bowing of the cas-
ing, to provide a smooth casing surface to alleviate problems when
removing the drive casing and to eliminate the use of solvent join-
ing compounds. Screens should have 0.010 in maximum slots to
prevent silting and should be set at depths depicted in Fig. 1. Back-
filling of the screen is accomplished by using Ottawa sand or a
similar silica material with a grain size greater than 0.010 in. The
annular space around the well casing should not be less than one
in.
CASE HISTORIES
The phased approach for assessing bedrock aquifer contam-
ination has been successfully employed by E & E in their inves-
tigation of hazardous waste sites in New England. The extent to
which the approach is utilized at any given site is determined by
the quantity and quality of existing information and the complex-
ity and size of the site.
Two case histories which illustrate E & E's overall approach and
the effect of bedrock on contaminant migration are presented here.
While each case history is purposefully brief, the discussions center
on the use of E & E's approach, the lack of existing data to fully
assess the problem and the effect of the bedrock aquifer in con-
trolling contaminant migration.
5' STEEL PROTECTIVE CASING AND LOCKING CAP
SURFACE fj
1.5" NOMINAL
DIAMETER SCHEDULE
80 PVC SOLID
5'-10' OF
BENTONITE >•
M
\
I
|
_
E
—
-
i
^
/ N
^
^
J
.
E
—
-
—
—
—
r
I
_^ CONCRETE
— ¥-
1.5" NOMINAL '
DIAMETER SCBEDUL!
to PVC o.oi"
SLOTTED CASING
BEDROCK
SURFACE
*, ^W,
1 \
Y, THREADED BOTTOM PLUG
'
A. FULLY SCREENED OVERBURDEN WELL A1TO BEDROCK WELL
This two well system can monitor overall
contaminant concentrations within the two
aquifers. Vertical hydrsulic gradients can
'also be determined.
5' STEEL PROTECTIVE CASING A8D LOCKUG CAP
8S
77**\
|
l.i" TCMINAL 5
DIAMETER SCHEDULE >
to PVC o.oi" <
SLOTTED CASING X
-
-
E
—
I
r' \
'/9 5*
. — *"
S
^
•
£^Csl V
"""
CONCRETE
^S±^fAP
^^^
'^^''^
5'-10' OP
•• ~ BENTONITE —
V
\
\
THREADED BOTTOM PLUG _
'wyWx^
B. HIED LEVEL OVERBURDEN AMD BEDROCK WELL
This two well system csn monitor contaminant
concentrations within bedrock versus the base
of the overburden aquifer. This system provides
more detslled information concerning contaminant
infiltration into bedrock.
• 5§
1.5" NOMINAL
DIAMETER SCHEDULE
» 80 PVC SOLID CASING —
~~ — -H.
BEDROCK SURFACE
\
X
$
— — —.____!
—
i
sp —
E
E
—
•
1.5" NOMINAL DIAMETER
J SCHEDULE 80 PVC 0.01"
SLOTTED CASING
f2^
^
s
\
^^Y
C. COMBINATION BEDROCK AHb FIXED OVERBURDEN WELL
Thin well tr ats the base of the overburden
and the bedr ck SB one regime. It Is parti-
cularly uaef 11 In areas where bedrock la not
highly fract red or productive and the
contamlnatio migrates primarily along the
Figure 1.
Monitoring Well Systems
-------
114
GEOHYDROLOGY
Case History I—Picillo Farm Site, Rhode Island
Site History and Description
The Picillo Site, a former pig farm in Coventry, Rhode Island,
operated as an illegal hazardous waste disposal site until 1977 when
a fire resulted in the discovery of containers of various chemical
wastes by the Rhode Island Department of Environmental Man-
agement.' Subsequent geophysical and excavation work revealed
that approximately 10,000 containers of wastes were buried in four
trenches on the site.
The disposal site consists of 7.5 acres of clear land surrounded by
terrain that slopes to the north, west and south. A swamp located
1200 ft northwest of the disposal site is a major discharge area for
surface and groundwater flowing from the site. Surface and
groundwater flow from the swamp is toward the southwest. A
stream which drains the swamp carries surface water to Whit-
ford Pond which is an irrigation source for a cranberry bog one
mile directly southwest of the disposal site (Fig. 2). Several miles to
the east of the site is the Quidnick Reservoir, which is not cur-
rently a source of drinking water. Private bedrock wells are located
approximately one-half mile north of the site.
Previous Investigation
Prior to E & E's involvement, the State of Rhode Island and
USEPA, through several subcontractors, had installed a total of 30
groundwater monitoring wells on and around the disposal site. This
initial work included well installation, bedrock coring, seismic re-
fraction survey, field permeability tests, and sampling and analy-
sis of soil, surface water and groundwater. The focus was upon
areas immediately adjacent to the disposal site and as such did not
address the impact the site might have on environmentally sensi-
tive areas nearby. The locations and depths of monitoring wells
installed during the initial work made it impossible to assess the
extent of contaminant movement.
Hydrogeologic Investigation
The major objective of E & E's investigation was to deter-
mine the possible existence of a hydraulic connection between the
disposal site. Quidnick Reservoir, the cranberry irrigation pond
and private wells. The approach included:
•Establishment of major plume directions using existing wells
•A preliminary evaluation of bedrock characteristics
•Field assessment of bedrock characteristics
•Installation of monitoring wells
•Sampling and analysis
Analytical data from existing monitoring wells were incomplete
and an additional sampling program involving all existing wells was
required. All samples were analyzed on a portable gas chromato-
graph and several key samples were analyzed for USEPA priority
pollutants.
The analyses confirmed the presence of contamination in
groundwater and defined the major contaminant plume directions,
one of which was toward the swamp. A seismic study was then
conducted to define the bedrock surface configuration and a frac-
ture pattern analysis to determine the bedrock structure. From
these data a bedrock surface map (Fig. 3) was prepared which in-
Figure 2.
Location Map of the Picillo Site
LEGEND
EMt 8«ltmlC — filon* Uflft*
=1- UlilliifKntt tM
A UMIttorlM Mil
Approilmiti B«drock Surtec*
~EI*villonUbow HSLI
Pradomlnau Sbltomd D4p ol
Bedrock Fraclurn
Figure 3.
Bedrock Geology
-------
GEOHYDROLOGY
115
Figure 4.
Concentration of Total Volatile Organics (ppb)
dicated that an east-west-trending bedrock trough existed be-
neath the swamp and that bedrock fractures strike north-south with
a moderate dip toward the west.
Nine monitoring wells were then installed (Fig. 3) in overburden
and bedrock to assess the full extent of contamination and pos-
sible migration through the bedrock trough and fractures. An addi-
tional sampling and analysis round was performed to include the
newly-installed wells.
Hydrologic and analytical data indicated that major contam-
inant plumes originating from the disposal site were moving in two
directions through the overburden aquifer. Contaminants from the
disposal site migrated predominantly in a fan-shaped configura-
tion northwestward toward the meadow and swamp (Fig. 4). A
portion of the contamination discharged into the swamp, and the
balance migrated westward in and under the swamp, toward Whit-
ford Pond.
Hydrologic and analytical data generated from monitoring well
installations, existing data, and groundwater sampling indicated a
slight vertical downward component of groundwater flow from
the overburden to the bedrock aquifer in recharge areas, and vice
versa in discharge areas. However, vertical interflow between aqui-
fers was believed to be minimal when compared with lateral flow,
since lateral hydraulic gradients exceeded vertical gradients. There-
fore, the lateral flow in the bedrock aquifer was found to be in the
same general direction as flow in the overburden aquifer.
Based on bedrock data gathered during field studies and analyti-
cal data generated from the sampling efforts, E & E concluded that
the major contaminant flow direction in the bedrock aquifer is
from east to west and is controlled by bedrock structure (orien-
tation of fractures and slope of bedrock trough).
As a result of the comprehensive assessment of the bedrock as
well as the overburden aquifer performed during this investigation,
E & E concluded that Whitford Pond and the cranberry bogs are
in danger of being contaminated. Contamination was not detected
nor is expected to reach Quidnick Reservoir and the residential
bedrock wells in the vicinity of the disposal site.
Case History II—Woburn, Massachusetts Site
Site History
Potentially hazardous materials have been generated, stored
and disposed of in the City of Woburn, Massachusetts for over
120 years." These activities have contaminated a large and pro-
ductive aquifer that has supplied as many as 100 industrial, pri-
vate and municipal wells.
In 1979 two of Woburn's municipal drinking water wells, lo-
cated within this aquifer, were closed when significant amounts of
chlorinated organic solvents were detected in the water. An inves-
tigation was required to assess the full extent of contamination and
the sources of contamination so that an evaluation of the impact on
human health and the environment could be made.
Site Description
The study area (Fig. 5), approximately 10 mi2, includes most of
East and North Woburn and is centered within the Aberjona
River Watershed. The Aberjona River, which flows toward the
south, forms a moderate north-south topographic valley along the
axis of the study area. The valley forms a productive aquifer where
two of Woburn's drinking water wells were located. The water
from these wells serviced as many as 20,000 residents.
The study area also incorporates several large industrial com-
plexes and is the region in which many tanning and chemical man-
ufacturing firms once existed. Presently high technology firms
dominate the region in addition to food service, trucking, barrel
cleaning, tanning and plastics manufacturing industries.
Previous Investigations
Several investigations have been performed within the study area
concerning the hydrologic characteristics of the aquifer and the
hazardous waste problems by the Commonwealth of Massachu-
setts, USGS and USEPA subcontractors. Emphasis was placed on
surface deposits of heavy metals that typically do not migrate in
groundwater and on surface water contamination where volatile or-
ganics for the most part will rapidly volatilize. Preliminary ground-
water evaluations utilized water level readings from different types
of wells taken during different seasons of different years.
The delineation of bedrock aquifers and overburden aquifers
was never fully addressed. The lack of sufficient data to properly
assess the contamination problem required further hydrogeologic
investigations.
-------
116
GEOHYDROLOGY
Figure 5.
Locus Map—North and East Woburn, Ma. Study Area Base Map from USGS Topographic Sheets—Wilmington, Reading, Lexington and Boston North
Quadrangles
Hydrogeologic Investigation
The presence of chlorinated organic solvents within well water
from Woburn's municipal wells indicated that a contaminant
plume existed within and was migrating through the aquifer. E & E
followed the approach outlined below to determine the extent of
that contamination, to identify possible sources and to provide
data useful for evaluating remedial options:
•Review existing data
•Determine extent of contamination from existing wells
•Assess geologic materials of the aquifer including bedrock char-
acteristics that may affect flow
•Install monitoring wells
•Sampling and Analysis
All available information regarding the geology and hydrology
of the study area was reviewed including driller's logs for 123 wells.
These data indicated that the Aberjona River Valley was under-
lain by a complex interstratified mixture of glacial deposits with-
in a deep buried bedrock valley. In many places till or other suit-
able confining materials did not exist suggesting that a hydraulic
connection between the overburden and bedrock aquifers may ex-
tend over wide areas.
Thirty-two accessible existing wells within the study area were
sampled and analyzed for USEPA priority pollutants. There were
over 60 present in the groundwater. Some of the wells screened in
bedrock or near the bedrock surface were contaminated with chlor-
inated volatile organics indicating that the bedrock aquifer had
been affected.
-------
GEOHYDROLOGY
117
CONTOUR DUOIUUl or TW KUS TO 3> HURT*
M*. IMTCCNT«l.*Mum,IU.
Contour Diagrams for bedrock joints
and observable faults from surface
exposures. The joint pattern does
not exhibit any major fracture
orientations. The faults do have a
preferred orientation with a trend
of N55E and a dip of 85' NW.
Figure 6.
Contour Diagrams for bedrock joints and observable faults from surface
exposures. The joint pattern does not exhibit any major fracture orienta-
tions. The faults do have a preferred orientation with.a trend of N55E
and a dip of 85 °NW.
Because contamination existed within bedrock an accurate map
of the bedrock surface topography and its structure was required
to evaluate contamination migration. A seismic survey was per-
formed to gather the necessary depth to bedrock data, and defined
regions within the bedrock that exhibited extensive fracturing.
Field analysis of bedrock outcrops including fracture pattern analy-
sis of faults and joints indicated preferred directions of these frac-
tures (Fig. 6).
Using the available bedrock and contamination data 22 moni-
toring wells were installed and fully screened both in overburden
and bedrock. Bedrock cores examined for fracture frequency and
petrographically for evidence of faulting confirmed the existence
of several fault zones and minor offsetting faults. Soil sampling
during drilling also indicated that most of the volatile organic con-
tamination existed near the bottom of the overburden aquifer,
even for compounds that are less dense than water.
A second round of sampling and analysis on groundwater from
both newly installed wells and existing wells provided a more ac-
curate extent of contamination. When the data for total volatile
organics was plotted as an isoconcentration map, a pattern of con-
tamination emerged that is consistent with the hydrogeologic data
Figure 7.
Areal Distribution of Trichloroethyle in Woburn
(Fig. 7). In fact, many of the known and suspected faults of the
region closely follow the pattern of contaminant migration. This
indicates that the faulted bedrock can play a significant role in
controlling the movement of contamination. The presence of most
of the contamination near the base of the aquifer shows that mi-
gration will probably be along bedrock troughs developed by faults
and/or that migration may be through the bedrock fractures them-
selves.
REFERENCES
1. Freeze, R.A. and Cherry, J.A., Groundwater, Prentice-Hall, Inc.,
Englewood Cliffs, N.J., 1979.
2. Josephson, J., "Protecting public groundwater supplies," Environ.
Sci. Tech., 16, 1982, 502A-505A.
3. Clay, P.P. and Norman, W.R., "Hydrogeologic Investigation for the
Picillo Farm Site, Coventry, Rhode Island," Ecology and Environ-
ment, Inc., Woburn, Massachusetts, 1981,250 p.
4. Cook, O.K. and DiNitto, R.Q., "Evaluation of the Hydrogeology
and Groundwater Quality of East and North Woburn," Ecology and
Environment, Inc., Woburn, Massachusetts, 1982, 750 p.
-------
EVALUATION OF REMEDIAL ACTION ALTERNATIVES—
DEMONSTRATION/APPLICATION OF GROUNDWATER
MODELING TECHNOLOGY
F.W. BOND
C.R. COLE
Batelle, Pacific Northwest Laboratories
Richland, Washington
D. SANNING
U.S. Environmental Protection Agency
Cincinnati, Ohio
INTRODUCTION
The LaBounty landfill was an active chemical waste disposal
site from 1953 to 1977. During this period, it is estimated that over
181,000 m' of chemical wasted were disposed in this 4.86 ha site lo-
cated in the flood plain of the Cedar River in Charles City, Iowa.
Compounds that are known to have been dumped in significant
quantities include arsenic, orthonitroaniline (ONA), 1,1,2-Trich-
loroethane (TCE), phenols, and nitrobenzene. Disposal ceased in
December 1977 following the discovery that wastes from the site
were entering the river. Since that discovery, the LaBounty land-
fill has been under intensive investigation.
One aspect of these investigations was the groundwater modeling
effort described in this paper. The overall objective of this model-
ing effort was to demonstrate the use of modeling technology for
evaluating the effectiveness of existing and proposed remedial ac-
tion alternatives at the site. The primary emphasis of the project
was on technology demonstration which included:
•Modeling of the groundwater system and prediction of the move-
ment of contaminants
•Development of criteria to determine which water sources im-
pacted by the site would require remedial action and the level of
action necessary to insure risks are at acceptable risk levels
•The evaluation of the costs and effectiveness of various remedial
action alternatives
DESCRIPTION OF THE STUDY AREA
The LaBounty landfill site is located on the floodplain of the
Cedar River just south of Charles City, Iowa. Charles City is in
Floyd County situated in northeastern Iowa. The site sits on alluv-
ial sand and gravel overlying a bedrock formation (Cedar Valley
formation). Glacial tills have been found along the west side of the
site above the Cedar Valley formation. The Cedar Valley forma-
tion can be divided into upper and lower units separated by a shale
layer which acts as a confining layer at the site.1'2
The hydrology of the LaBounty site is somewhat complex in that
regional potential contours for the upper Cedar Valley aquifer and
the shallow alluvial groundwater system indicate that the river acts
as the major discharge area. In order to accurately represent the site
with a model it is necessary to simulate this vertical movement.
MODEL DEVELOPMENT
Two computer codes were used to model the LaBounty landfill
study area: FE3DGW and CFEST.' The FE3DGW code is a three-
dimensional, finite-element, groundwater flow code that can be
used to predict ground water potentials, water flow paths and
travel times. The CFEST (Coupled Flow, Energy, and Solute
Transport) code is an extension of FE3DGW that predicts the
movement of contaminants with the convective dispersion equa-
tion for contaminant transport.
In order to better characterize the groundwater system at the
LaBounty site, a two-stage modeling approach was used. The first
stage consisted of developing a coarse grid regional hydrologic flow
model, and the second stage consisted of developing a flow and
contaminant transport model for the immediate vicinity of the
LaBounty site. The purpose of the regional model was to estab-
lish hydrologic boundary conditions for the local model.
Regional Flow Model Implementation
The coarse grid regional model covers an area of 6.3 km by
4.8 km. The regional boundaries were chosen so that the equal
potential contours could be assumed nearly vertical with depth, or
at a distance far enough away that errors in the assigned poten-
tials did not influence local model boundary conditions.
The finite element grid pattern used to represent the large reg-
ion is shown in Fig. 1. The local model finite element grid boun-
daries superimpose onto the regional element pattern at the site.
Five different material types (layers) were simulated in the reg-
ional model. The lower layers, upper and lower bedrock and shale,
are present over the entire model region. The alluvial layer exists
along the Cedar River from a point north of Charles City south-
east to the model boundary (Fig. 1). A combined fill and till layer
was simulated in the area of the LaBounty Site. The structural top
and ground-water potential for the layers were taken from data re-
ported by Munter's paper.2 Hydraulic conductivities (K) of the
layers were initially set at average values obtained from several
sources1'2'5'6'7 and the vertical to horizontal permeability ratio
(K/Kx ) was set at 0.1.
ALLUVIAL LAYER BOUNDARY
LOCAL REGION BOUNDARY
Figure 1.
Large Region Model Finite Element Grid
118
-------
GEOHYDROLOGY
119
Initially it was assumed that 10% of the total precipitation of
900 mm8 recharged the groundwater system. This was the only
source of recharge considered.
Local Flow Model Implementation
The local model covers the area just in the immediate vicinity of
the LaBounty site (Fig.. 2) and is 685 m on a side. The Cedar River
forms the boundary to the north, east, and south while the western
boundary was arbitrarily set to the west of the fill material.
The finite element grid pattern used to represent the small region
is also shown in Fig. 2. The same materials simulated in the large
region were simulated in the small region except that the fill and till
were modeled separately.
The structural top and bottom and potentiometric surfaces for
the shale and bedrock layers were defined the same as in the reg-
ional model. More detailed structure and potential surfaces were
prepared for the local model alluvium, fill and till layers from data
obtained by Munter.2
Stage data for the Cedar River were extrapolated to obtain held
river elevations along the model boundary. The same permeabilities
and recharge rate used in the regional model were initially used in
the local model.
Local Contaminant Transport Model Implementation
The CFEST code was used to simulate the movement of arsenic
from the landfill to the Cedar River in the local model region and
to test the effectiveness of the proposed remedial action alterna-
tives in controlling the arsenic plume. Since the groundwater flow
portion of CFEST is identical to FE3DGW, the final calibrated
FE3DGW data set was used directly in CFEST to predict ground-
water flow. The additional inputs required by the CFEST code to
simulate contaminant transport were longitudinal (DjJ and trans-
verse (Dj) dispersivity and initial contaminant concentration lev-
els. Initially DL and Dj were set at 30.0 m and 0.5 m, respec-
tively. Arsenic was chosen for use in the calibration and remedial
action simulations because it has been extensively monitored and its
source is fairly uniformly distributed over the area of the fill ma-
terial.
Arsenic probably enters the groundwater flow system from two
sources: 1) precipitation infiltrating through the fill material, and
2) leaching due to fill material being below the water table. The
location of the saturated portion of the bill was determined from
the literature2'6 to be in two places (Fig. 3). For modeling pur-
poses these two sources were simulated by holding the downgrad-
ient surface nodes of the saturated fill layer at an arsenic concen-
tration of 500 mg/1. The 500 mg/1 concentration is based on the
estimates of the arsenic solubility. Water infiltrating through the
cap was also assigned an arsenic concentration of 500 mg/1.
MODEL CALIBRATION
The flow model calibration process consisted of matching ob-
served potentials with model predicted potentials for both the reg-
ional and local models. Calibration was accomplished by first cali-
brating the regional model and then testing the local model with a
consistent set of data. This alternating process was carried out un-
til a satisfactory agreement between measured and model-pre-
dicted heads was achieved in both models.
The final input parameters obtained for the local model were
then used in the contaminant transport model. These parameters
were not changed in the contaminant transport calibration process.
Parameters related to contaminant transport were adjusted to ob-
tain a reasonable match between field-measured and model-pre-
dicted contaminant concentrations.
Flow Model Calibration
The parameters adjusted in the flow model calibration process
were the hydraulic conductivities (K), the ratio of vertical to hori-
zontal hydraulic conductivities (KZ/KX), the specific storage, and
the rate of recharge. After several calibration runs, a good match
was achieved between measured and model-predicted potentials for
both the regional and local models. The final calibrated hydraulic
conductivities and storages used in the flow models are listed in
Table 1. More detailed recharge calculations estimated that about
14% of the precipitation (126 mm/yr) reached the ground water.
Figure 2.
Local Groundwater Flow Model Finite Element Grid
Figures.
Contaminant Transpolrt Model Finite Element Grid. Cross-hatched area
indicates area determined to be below the water table
-------
120
GEOHYDROLOGY
Table 1.
Input Values Used In the Final Calibrated Flow Models
Uyer
(Small)
I*
2
31
4
5
6
Mi Intel
Fill
Aluvium
Till
Upper
Cedar Valley
Shale
Lower
K'
(m/d«r>
0.5
10 and 2
5*10~4
2.0 and 4.0
2.0and5xlO~3
2.0
Specific
Storage
(I/ml
0.0025
0.01
0.0025
,„-•>
10
V*
(Udo
O.l
O.l
O.I
O.I
O.l
O.l
Cedar Valley
•Where the two values of K are given, ihc first number listed covered the largest percentage of
the total area covered by that material.
tThe fill and till layers were combined into one layer in the regional model with a K of 10
m/day.
Contaminant Transport Model Calibration
One of the first changes made was to change the node and ele-
ment grid to allow for a more accurate description of the loca-
tion of the arsenic source and better represent the spread of the re-
sulting arsenic plume. The finite element grid used in all subse-
quent calibration runs and in the remedial action cases is shown in
Fig. 3.
Different values of longitudinal and transverse dispersivity were
tested in order to match model-predicted with measured arsenic
concentration levels. The results of these test runs indicated that
longitudinal and transverse dispersivities of 5.0 m and 0.5 m, re-
spectively, gave the best results. The initial, concentration level for
arsenic of 500 mg/1 was not changed in the calibration process.
The last step in the calibration process was to determine the loca-
tion of the arsenic source.' Arsenic could enter the flow system
from either infiltration alone, or a combination of infiltration and
groundwater leaching where fill material is thought to be below the
water table. Both cases were tested by comparing model-predicted
arsenic concentrations and river loading quantities against field-
measured levels.
After running both cases for 6000 days (16.4 yrs), it became ap-
parent that the case of infiltration alone would never achieve the
arsenic levels that have been measured at the site. The results indi-
cate that part of the fill material must indeed be below the water
table and that the majority of the contaminant transport is orig-
inating in the saturated groundwater system and not in the unsatur-
ated flow system as a result of infiltration.
Based on these findings, the final calibrated model, hereafter
referred to as the base case, included both sources of arsenic, in-
filtration and groundwater leaching. A contour map of the base
case arsenic concentrations after 6000 days (near equilibrium) is
shown in Fig. 4. This final calibrated contaminant transport model
base case became the initial condition from which all the remedial
action cases were initiated.
The final calibrated transport model predicted arsenic mass out-
flow rate to the river was 25.2 kg/day. The average mass outflow
rate as measured by the monitoring system at the site is 22.7
kg/day.
ASSESSMENT OF REMEDIAL ACTION ALTERNATIVES
Seven remedial action alternatives were identified for potential
application to the LaBounty landfill site and were modeled with
the CFEST model; clay cap, upgradient cut-off wall, downgrad-
ient cut off wall, limited excavation, limited bottom lining, stabil-
ization, and pump and treat. The clay cap alternative has already
been implemented at the site.
Contaminant transport model parameters were adjusted to simu-
late arsenic migration to the river for each alternative. The alterna-
tives were assessed by first quantifying their effectiveness on reduc-
ing contaminant concentrations and Cedar River loading rates.
CONCENTRATION
75.QQ— CONTOUR
Figure 4.
Arsenic Concentrations (mg/1) at the Top of the Alluvium for the Base
Case of Combined Infiltration and Disposal Below the Water Table
These results were compared with the results for the clay cap al-
ternative to obtain a measure of effectiveness.
The next step was to estimate the capital and annual operating
and maintenance costs associated with each alternative. Given the
effectiveness and cost, the final step was to make a cost-effec-
tiveness comparison of the alternatives.
The base case, as discussed earlier, was run for 6000 days. Each
remedial action case was run for anf additional 4500 days begin-
ning at the end of the 6000-day period.
Clay Cap
The clay cap alternative was modeled by reducing the infiltra-
tion rate by a factor of 100 in the area of the cap. The magnitude
of the decrease was estimated using a simple compartmental model
and a permeability for clay of 3 x 10 ~ ^ m/day. The rate of arsenic
injection from infiltration was also reduced by a factor of 100 to
0.08 kg/day.
The results of this run indicated a continued but slight rise in pre-
dicted arsenic loading of the Cedar River to 27.5 kg/day after
4500 days, again illustrating that the portion of the waste dis-
posed below the water table is the major source of the contam-
ination problem at LaBounty. The clay cap was assumed to be in
place for simulation of the remaining six remedial action alterna-
tives.
Upgradient Cut-Off Wall
The objective of the upgradient cut-off wall is to reroute in-
coming ground water such that it bypasses waste contaminated soil.
This can only work if the wall extends downward to an imperm-
eable layer or incorporates some means of transport to cany off
water mounded behind the wall. Since the dolomite base rock at
Charles City is transmissive, the later would be required. For the
purposes of this evaluation, it was assumed that a subsurface drain
is constructed upgradient from the cut-off wall to carry uncon-
taminated water away from the site.
-------
GEOHYDROLOGY
121
The effectiveness of the upgradient cut-off wall was simulated
by introducing a line of elements with low permeability upgrad-
ient of the landfill (Fig. 5). The upstream set of surface nodes along
the low permeability cut-off wall were treated as held potential
boundary nodes to simulate the upgradient drain. The flux to these
held nodes was calculated to determine the required capacity of
the subsurface drain associated with this alternative.
SCALE (m)
200
UNCRAOIENT CUTOFF WALL
SATURATED FILL MATERIAL
DIRECTION OF GROUND-
WATER FLOW
SLURRY TRENCH
SUBSURFACE DRAIN
TREATMENT SYSTEM
Figure 5.
Upgradient Cut-Off Wall Remedial Action Alternative
DOWNCRADIENT CUTOFF WALL
SATURATED FILL MATERIAL
DIRECTION OF GROUND-
•" WATER FLOW
— SLURRY TRENCH
SUBSURFACE DRAIN
• TREATMENT SYSTEM
Figure 6.
Downgradient Cut-Off Wall Remedial Action Alternative
The results of this analysis indicate that new equilibrium is
reached in about 5 years with Cedar River loading rate reduced to
14.9 kg/day after 4500 days.
Downgradient Cut-Off Wall
The downgradient cut-off wall is intended to intercept contami-
nated groundwater and then route it via drains for treatment. The
effectiveness of the downgradient cut-off wall was simulated in the
model with a line of impermeable elements downgradient from the
site (Fig. 6). The upstream set of surface nodes formed the simu-
lated drain and were treated as a held potential outflow boundary.
The flex to these nodes and arsenic concentration levels were pre-
dicted to determine the capacity of the subsurface drain and the
treatment facility.
Results of this simulation indicate that new equilibrium river
loading rates were achieved after approximately 3 years, 2 years
faster than for the upgradient case. The new Cedar River loading
rate predicted at the end of 4500 days was 7.66 kg/day.
Limited Excavation
The intent of the limited excavation approach is to remove those
portions of the waste below the water table and to backfill with
clean material.
Simulation of the limited excavation case with the model in-
volved removing the source term associated with wastes disposed
below the water table and changing the permeability in the area
where the fill material was removed to that of the alluvium. The
approximate size of the areas below the water table is shown in
Fig. 3.
Results of this simulation indicate that this alternative provides
a more delayed response in terms of river loading rates but that
with time, much lower arrival rates could possibly be achieved.
Model results indicated that after 4500 days predicted Cedar River
loading has decreased to 8.0 kg/day.
Limited Bottom Lining and Stabilization
These two approaches are very similar in that they are both iso-
lation techniques, either through the use of a liner or fixation meth-
od. Only those portions of the waste below the water table would
be lined or stabilized. Presumably, this would reduce water flux
and hence minimize further contributions from these portions of
the landfill to the overall contaminant plume. Stabilization pre-
sents a challenge to current technology since in situ fixation has
not been perfected and may not be technically feasible at the La-
Bounty landfill. Both of these alternatives were modeled in an
identical manner.
The total source mass of waste remained the same as the base
case but the leach rate was reduced. This was simulated in the
model by changing the permeability in the saturated fill area to that
of a clay (10 ~ ^ m/day).
The results for these alternatives are very similar to those of the
limited excavation alternative. As with limited excavation, the re-
duction in Cedar River loading is more gradual but it has the pos-
sibility of eventually reaching much lower river loading rates.
After 4500 days a Cedar River loading of 8.14 kg/day was pre-
dicted.
Pump and Treat
In this isolation method, removal is accomplished through use of
a series of withdrawal wells (Fig. 7). The water is subsequently
treated to eliminate contaminants and either discharged back to the
river or reinjected to assist in directing the contaminated plume
toward the wells.
Pumping rates were estimates at 1806 mVday. As with the
downgradient cut-off wall, predicted concentrations for treatment
during the first year are higher (71 mg/1) but soon levels off at
about 34 mg/1. Predicted Cedar River loading rates have pretty
much equilibrated after 5 years and were predicted to reach 10 12
kg/day after 4500 days.
-------
122
GEOHYDROLOGY
0 SCALE (m) 200
PUMP AND TREAT
SATURATED FILL MATERIAL
DIRECTION OF GROUND-
~~r WATER FLOW
• PUMPING WELL
SUBSURFACE DRAIN
• TREATMENT SYSTEM
Figure?.
Pump and Treat Remedial Action Alternative
The capital cost and annual operating cost for all the alterna-
tives are given in Table 2. For all cases, these costs should be con-
sidered only as rough estimates, since they are based on prelim-
inary conceptual designs for each alternative and they do not re-
flect local cost differences.
DISCUSSION
A comparison of arsenic mass outflow rates to the river for the
base case and the seven remedial action alternatives as predicted
by the model is shown in Fig. 8. The contaminant transport model
very clearly indicated that the major source of groundwater con-
Table 2.
Estimated Cost for Each of the Seven Remediation Alternatives
(1982 dollars)
Remediation Alternative
Upgradienl cut-off wall*
and treatment
Limited excavation"
(on site reburial)
Limited bottom lining
Pump and treatment
Oowngradient cut-off
wall and treatment
Limited excavation
(off site reburial)
Stabilization
Capital
Cost ($)
835,000
1,000,000
1,348,000
1,431,000
2,569,000
16,000,000
16,787,000
Annual
Operating
Cost ($)
99,000
35,000
67,000
168,000
114,000
35,000
35,000
Limited Bottom Lining/
stabiliiation
19619
1972 19819 19919
HYPOTreTIW. TIME (TEWSJ
Ml optimised for additional contamineni removed, costs for this alternative are likely to be
higher vince diversion of more water will most likely be required.
••A»ume^ made here is that »asle removed from below the water table would be rebuncd on
tut above the water table and capped with clay
Figure 8.
Predicted Effective of the Various Remediation Alternatives in Reducing
the Arsenic Loading of the Cedar River
lamination arises from that portion of the arsenical sludge buried
below the water table.
Model results indicate that only the clay cap and the upgrad-
ient cut-off wall would probably fail to bring river concentrations
(under low flow conditions) below drinking water standards. The
upgradient cut-off wall, however, could probably be significantly
improved through optimization studies with the transport model.
Optimization studies were not performed for any of the alterna-
tives.
The remaining five alternatives all lowered river concentrations
to less than drinking water standards for Cedar River low flow
conditions; however, the selection of the best alternative is still
subjective. The downgradient cut-off wall and pump and treat al-
ternatives show a quick response in reducing contamination levels
but they approach a higher Cedar River loading limit. On the other
hand, the bottom lining, excavation, and stabilization alternatives
do not respond as quickly but have the potential to achieve much
lower Cedar River loading rates and aquifer contamination levels.
Cost effectiveness of the various remediation alternatives is an
important part of any decision making process. As indicated in
Table 2, limited excavation (on site reburial) and bottom lining
are the least cost alternatives (excluding the upgradient cut-off
wall which failed to meet the level of contamination criteria). When
comparing costs per unit reduction after 4500 days in predicted
Cedar River arsenic concentrations (low river flow conditions) or
the cost per unit reduction in arsenic contaminant mass in the satur-
ated aquifer system, the limited excavation (on site reburial) is most
cost-effective, followed by limited bottom lining and pump and
treat. In both cases, stabilization is the most expensive alterna-
tive, followed by limited excavation (offsite reburial) and the down-
gradient cut-off wall.
CONCLUSIONS
A hydrologic flow and transport model of the Charles City
LaBounty landfill was developed, calibrated and used to evaluate
seven remedial action alternatives for the LaBounty site. This mod-
eling study has reaffirmed the importance of considering the reg-
ional hydrologic system when modeling the hydrology and trans-
port for a more local system like LaBounty, and illustrates the
value of groundwater modeling in developing an understanding of
a complex flow and transport system.
In the process of modeling the LaBounty site, it was demon-
strated that boundary conditions for the local model cannot be
-------
GEOHYDROLOGY
123
arbitrarily set and still obtain reasonable estimates for travel time
and river discharge rates. The contaminant transport model devel-
oped for the LaBounty site, while not a perfect indicator of ob-
served contamination movement, is certainly an adequate indicator
for the purpose of general understanding and for the evaluation
of the effectiveness of various remedial action alternatives.
REFERENCES
1. Eugene A. Hickok and Associates, "Hydrologic Investigation, La-
Bounty Site, Salsbury Laboratories, Charles City, Iowa." For: De-
partment of Environmental Quality, Des Moines, Iowa, 1981.
2. Munter, J.A. Evaluation of the Extent of Hazardous Waste Con-
tamination in the Charles City Area. Iowa Geologic Survey, Iowa
City, Iowa, 1980.
3. Gupta, S.K., Cole, C.R. and Bond, F.W., "Methodology for Re-
lease Consequence Analysis—Part III, Finite-Element Three Dimen-
sional Groundwater (FE3DGW) Flow Model. Formulation Program
Listings and User's Manual," PNL-2939, Pacific Northwest Labora-
tory, Richland, Wash., 1979.
4. Gupta, S.K., Cole, C.R., Kincaid, C.T. and Kaszeta, G.E., "Des-
cription and Application of the FEZ3DGW and CFEST Three-Di-
mensional Finite-Element Models," LBL-11566, Lawrence Berkeley
Laboratory, University of California, Berkeley, Ca., 1980.
5. May, J.H. Evaluation of Methods for In Situ Stabilization of Con-
taminants at the LaBounty Landfill. For: Municipal Environmental
Research Laboratory, USEPA, Cincinnati, Ohio, 1981.
6. Eugene A. Hickok and Associates, "Soil Characteristics, LaBounty
Site, Salsbury Laboratories, Charles City, Iowa." For: Department of
Environmental Quality, Des Moines, Iowa, 1977.
7. Layne-Western Company, Inc., Monitoring Well System Hydrologic
Evaluation, LaBounty Landfill Site. Charles City, Iowa, 1980.
8. NOAA (National Oceanic and Atmospheric Administration). Local
Climatological Data. Monthly Summary for Charles City and Water-
loo, Iowa. Department of Commerce, 1979,1981.
-------
REMEDIAL ACTION MASTER PLANS
WILLIAM M. KASCHAK
PAUL F. NADEAU
U.S. Environmental Protection Agency
Hazardous Site Control Division
Washington, D.C.
INTRODUCTION
The Remedial Action Program has been developed to respond
to releases of hazardous substances from the 400 sites comprising
the National Priorities List under the commonly known Super-
fund Program. These sites may require long-term cleanup actions
to provide adequate protection of public health, welfare and the
environment.
The specifics of the Remedial Action Program are described in
Phase VI, Section 300.68 of the National Contingency Plan
(NCP), published in Federal Register, Vol. 47, No. 137, July 16,
1982, pursuant to the Comprehensive Environmental Response,
Compensation, and Liability Act of 1980 (CERCLA). The
Remedial Action Program is complex even with the limited
number of sites currently being addressed.
The Interim Priority List, which identified 115 sites eligible for
remedial funding, was published on Oct. 23, 1981. This list was
expanded to 160 sites on July 23, 1982. As the program expands
and the National Priorities List of 400 sites is published, an effec-
tive long-range planning mechanism becomes essential.
An effective technical and financial planning document which
has been developed to assist with the long-range planning needs of
the USEPA is the Remedial Action Master Plan or RAMP. In this
paper, the authors discuss the development of the RAMP and its
current and future uses as a planning tool for conducting remedial
activities at uncontrolled hazardous waste sites, and outlines the
basic structure and contents of the RAMP with the use of an il-
lustrative example.
BACKGROUND
The conduct of work in the Remedial Action Program is
governed by certain provisions of the NCP and CERCLA.
Engineering, economic and environmental factors must be con-
sidered in developing the "cost-effective" remedy for a hazard-
ous waste site. A brief description of the remedial action process
and the key provisions of the NCP and CERCLA that affect the
financial aspects of the program are highlighted below.
The NCP identifies sequences of activities that must be under-
taken prior to remedial action implementation. These phases are:
a remedial investigation, a feasibility study to select the cost-
effective action, and design of the selected remedy. The eventual
remedial action selected may be one of the three types: initial
remedial, source control and off-site measures. Certain sites will
undergo combinations of these measures, depending on the need
for that particular site.
Several provisions of CERCLA affect the financial aspects of
the Remedial Action Program. The first is the State cost share re-
quirements, which are 10% for privately owned sites and at least
50% if the State or one of its political subdivisions owned the site
at the time of disposal. The second is the need to balance the
funding requirements of a site against the amount of money
available in the CERCLA Trust Fund to respond to other sites
which present or may present a threat to public health, welfare or
the environment. The third is the potential for responsible parties
to provide funds or undertake the activities themselves for all or
part of the necessary response actions.
The primary purpose of the RAMP is to assess available data for
a site and identify the type, scope, sequence and schedule of
remedial projects which may be appropriate. Budget level cost
estimates are prepared for the first projected phase of activity,
along with a detailed statement of work. Brief project descriptions
and order of magnitude cost estimates for future projects are also
included. Other key components of the RAMP are the evaluation
of existing data and any limitations associated with the data such as
sampling/analysis protocols and chain of custody requirements;
assessment of the need for additional data to evaluate remedial
alternatives; community relations requirements; and a discussion of
administrative and procedural requirements including any special
problems that may be encountered during project implementation.
Ramp Development
The concept of the RAMP is not new to the remedial program.
Upon the passage of CERCLA, remedial action plans were
developed for several sites to establish a priority for funding under
a special Resource Conservation and Recovery Act (RCRA) ap-
propriation which was authorized to conduct remedial planning ac-
tivities at hazardous waste sites. The sites selected comprise what is
known as "the initial 20 list". Remedial action plans were prepared
on short notice and were primarily limited in scope and content,
but were useful for initiating work at these sites. It was decided to
expand this concept into the Remedial Action Master Plan and
develop a document that would assist with the long-range planning
needs in the Superfund program.
The starting point was defining what a RAMP should include
and what the desired products should be. With the assistance of
USEPA Region II, Region X and URS Company, Inc. (a subcon-
tractor to Booz-Allen and 'Hamilton, Inc.), several sites were
selected to have RAMPs prepared under a trial project. The objec-
tive was to determine the timing and level of effort to complete a
RAMP. This trial project was completed in March 1982 and pro-
vided a basis for preparing RAMPs at other sites. The current
estimate for the completion of an average RAMP is about 300 work
hours of effort over a duration of three months. These numbers
should decline as the procedures for completing RAMPs become
more refined. The responsibility for preparing RAMPs rests with
each USEPA Regional Office, in close consultation with the States.
There are currently 68 RAMPs completed or under development at
this time excluding sites where remedial activities have already been
initiated.
RAMP SUPPORT FUNCTIONS
The major planning and administrative tools of the RAMP are
the site master schedules, the project cost estimates and the detailed
statements of work. These tools are of use for technical and finan-
cial planning purposes, the necessary ingredients of a cost-effective
remedial action program.
124
-------
REMEDIAL RESPONSE
125
Technical Support
Remedial activities will be conducted by any one of the following
methods: (1) the responsible party through voluntary or judiciary
actions, (2) the State through a cooperative agreement with
USEPA, or (3) USEPA in accordance with a State Superfund con-
tract. The RAMP provides a technical guide for conducting
remedial activities at a hazardous waste site following the pro-
cedures in the NCP, eventually leading to the recommendation of
the cost-effective remedial action for the site. These activities
should be followed regardless of who performs them, thereby en-
suring compliance with the intent of the NCP and CERCLA.
Once a decision is made to use Superfund monies to clean up a
site, the work schedules, cost estimates and statement of work are
used to prepare the cooperative agreement or State Superfund con-
tract. For State lead projects, the statement of work provides ample
information for the development of a Request for Proposals for the
State to procure engineering services. For USEPA lead projects,
the statement of work would be formulated into either a work
assignment for a USEPA contractor or an interagency agreement
with the U.S. Army Corps of Engineers. The RAMP will also pro-
vide the USEPA Regions with a guide for overseeing activities con-
ducted by other parties.
The RAMP will be updated when events cause significant
changes in the scoping of remedial activities at the site. By keeping
the RAMPs up-to-date with the current technical and financial re-
quirements of sites, USEPA and the States can periodically reassess
their priorities and make changes as necessary.
Financial Planning
The Superfund program is almost two years old and has been
operational (i.e., funded by Congressional appropriations) for one
and one-half years. The RAMP has demonstrated its short-term
capabilities for scoping out remedial activities and providing a basis
for the development of cooperative agreements and State Super-
fund contracts. The long-term Financial planning capabilities of the
RAMP are yet to be tested.
State Financial Planning Requirements
States are required to contribute 10% of the remedial costs or at
least 50% of the costs if the State or one of its political subdivisions
owned the site at the time of disposal of hazardous substances.
States are also required to assume all future maintenance of
remedial actions. Some states have already passed their own
minisuperfunds, appropriations, and bond issues to cover the costs
for their share of remedial activities under the Superfund program.
The project schedules and cost estimates of the RAMP can be
used to develop the overall financial needs of the program as well as
cash flow projections for both USEPA and the States. This is im-
portant for establishing and managing any program, not just
Superfund. States can use the RAMPs developed for their sites to
establish their own priorities. The information will assist the States
in identifying State financial requirements and it can be used to
justify and obtain funds from State legislatures to cover the cost
share requirements.
USEPA Financial Planning Requirements
USEPA has the responsibility of administering the Trust Fund
for the entire Superfund program including emergency response ac-
tions, site discovery and investigation efforts, other Federal agency
programs, research and development efforts, and remedial
response activities. The key aspects of CERCLA that affect the
Remedial Action Program are the cost sharing provisions, the
development of cost effective remedial actions, and the need to
balance Trust Fund monies at the facility under consideration with
the ability to respond to other sites.
Remedial activities at sites are beginning to pick up at a rapid
pace. Until now, emphasis has been on initial project startups.
However, with the expected release of the 400 list and the expected
completion of many feasibility studies recommending large expen-
ditures for remedial implementation projects, emphasis is shifting
toward the planning and phasing of projects at sites.
The RAMP will be a useful tool for this purpose because it con-
tains project costs which are based upon actual conditions at the
site. Current practice is to use an average unit cost to conduct a
remedial investigation, feasibility study, design or implementation
project until more definitive costs are developed on a site by site
basis. The cost estimates in the RAMP, either the budget level for
the first phase of activity or the order of magnitude estimates for
future projects, will be more accurate for the purpose of developing
long-range plans.
RAMP STRUCTURE
The basic structure of a RAMP includes an executive summary
and three major sections: (1) compilation and evaluation of existing
data, (2) remedial activities, and (3) appendices.
Compilation and Evaluation of Existing Data
The compilation and evaluation of existing data is a necessary
first step in approaching site remediation. Probably one of the
most difficult questions to answer is what additional information is
necessary to be able to identify and evaluate remedial alternatives
for a hazardous waste site without "studying the site to death."
This problem became apparent as USEPA and the States began to
work on the initial 20 Superfund sites. Some of these sites had an
extensive array of reports available, while others had little if any in-
formation about them. Much of the data was compiled years ago,
raising several important questions such as: how good are these
data today; can an engineering solution be properly developed and
designed; and will the data be defensible in court?
The first step of a RAMP is to collect, analyze and evaluate
available data and identify any data gaps or other limitations. This
is not intended to be a detailed assessment of all existing data,
which could, in some situations, be very costly and time consum-
ing. For instance, there is one particular site about which more than
100 reports have been written, all of which contain potentially
useful information for evaluating remedial alternatives. The brief
review of data for the RAMP is designed to prevent the costly
duplication of previous efforts and to ensure that the information
from which decisions are made is technically and legally sound.
To minimize costs, RAMPs are prepared exclusively from ex-
isting information. The site data base is likely to include
assessments and site inspections performed by USEPA, studies per-
formed by the State, county or municipality, and records contained
in regulatory and licensing agency files. The actual volume and
quality of data will vary considerably from site to site and,
therefore, influence the level of detail contained in the RAMP. The
review of existing data focuses on the quality assurance/quality
control procedures, chain of custody procedures, and sampling and
analytical protocols.
A preliminary determination is then made on the adequacy of the
information, to support the development and evaluation of
remedial alternatives, as well as support cost recovery actions. The
available data serve as a baseline to be used in assessing the site
situation and provides a mechanism for determining data limita-
tions and the need for obtaining additional data. Appropriate
techniques on how to best collect the required information are
discussed in the remedial activities section of the RAMP.
The first section also includes a description of the location and
history of the site, the environmental setting, previous community
relations efforts, enforcement actions and other pertinent informa-
tion. The previous uses of the site are summarized as a means of
providing a general understanding of the current site situation. This
description outlines previous waste disposal practices or site opera-
tions which have led to the contamination of the site.
Although a complete chronological description of all past site ac-
tivities is seldom possible, the site historical description outlines
major actions at the site. The description provides a chronological
listing of site owner(s), the operations which have taken place at the
site, and the periods during which these operations have taken
place. Of particular importance are the types of operations and the
-------
126 REMEDIAL RESPONSE
types of raw and waste materials used or processed at the site. Any
permits issued or regulatory actions taken against the site or its
previous owners are also identified and briefly described. The en-
vironmental setting is often a key to identifying potential pathways
of exposure to the hazards posed by a particular site. In most situa-
tions, a site visit is conducted to verify the information gathered
and to note significant changes at the site.
Remedial Activities
The remedial activities section is the key section of the RAMP. In
this section, the sequence, timing, estimated cost and correlations
of projects needed to clean up the site are outlined.
The major element of this section is a preliminary "scoping deci-
sion" which establishes a basis for conducting all future remedial
activities at the site. The preliminary scoping decision is based upon
the existing data and identifies the potential for initial remedial
measures and the need to evaluate source control and/or off-site
measures. Such initial measures might involve the removal of
drums and/or visibly contaminated soils. Options could also in-
clude drainage controls, lagoon stabilization, temporary potable
water supplies and temporary capping. In this way, appropriate ac-
tions can be taken based on available information while
simultaneously collecting additional information and evaluating
alternatives for source control or off-site measures. Parallel ac-
tivities are practical at most sites where imminent hazards are taken
care of as an initial remedial measure while more extensive source
control and off-site actions are evaluated.
With the necessary scoping decision made, the engineer is able to
prepare a detailed statement of work, project work schedule and
cost estimate to complete the next phase of remedial activity.
Future remedial activities are also identified and are displayed on a
site master schedule. Order of magnitude cost estimates are provid-
ed, as appropriate, to assist with the financial planning re-
quirements of the Superfund program.
In order to illustrate "scoping" more clearly, a typical RAMP is
described below. The RAMP is for an actual site but the site name
and location are withheld as the contents and concept of the
RAMP are the primary concern. The particular site is selected
because it portrays the potential for "fast tracking" an initial
remedial action and scopes out the full remedial investigation and
feasibility study to evaluate source control and off-site remedial
alternatives.
Site Description
The site has been inactive for several years. There are several
thousand drums of various waste stored on the site. The drums
have deteriorated over the years and, as a result, the contents of the
drums have leaked. Since there is evidence of the area being regrad-
ed, one would expect the presence of buried wastes. Previous aerial
surveillance of the site also supports this assumption. There are no
surface drainage controls present at the site, and off-site migration
by run-off is evident from stressed vegetation. Water samples from
a nearby creek and groundwater samples from monitoring wells
show high levels of several priority pollutants. The ability to iden-
tify buried wastes is hindered by the cramped conditions in the
regraded area caused by drums stored on the surface.
RAMP Recommendations
Through the compilation and evaluation of existing site data, the
recommendations in the RAMP include a limited remedial in-
vestigation and feasibility study, and the design and implementa-
tion for a fast track initial remedial action to remove surface drums
and visibly contaminated soils. The RAMP also proposes a more
detailed remedial investigation and feasibility study to evaluate
source control and off-site control measures. The comprehensive
work schedule covers 40 months to conduct the remedial activities.
The first 24 months of the site activity are shown in Fig. 1.
The State in which this site is located has elected to enter into a
cooperative agreement with the USEPA. The State will, therefore,
have the lead in the projects and conduct the remedial activities
through State contractors. The time sequence of the schedule
begins with the signing of the cooperative agreement with the State.
The normal sequence of activities (bottom track) can be initiated at
almost any time. This is annotated by the dashed line as "float
time" in the schedules.
The separate site activities continue on parallel schedules (Fig. 1).
The limited remedial investigation and feasibility study provides the
information necessary to prepare contract bid documents and to
determine the cost-effective disposal alternatives for the various
wastes present at the site. Various reviews and decision points are
also scheduled. For this particular site, most of the remedial in-
vestigation activities (bottom track) will have to wait due to worker
safety considerations and the difficulty with determining the extent
of buried materials until the surficial cleanup is completed.
The cost estimates for all phases of work at the site as well as the
TIME IN 0 2 4
MONTHS
6 8 10 12 14 16 18 20 22 24
PROCUREMENT 0
LIMITED REMEDIAL
INVESTIGATION/
:EASIB1LITT STUDY
(DESIGN 1
0 REVO 0 REVO PROCUREMENT 0
. PERMITS
IMPLEMENTATION
REMOVAL OF
HAZARDOUS
IASTE MATERIALS
0
F
A
S
T
T
R
A
C
K
0 COOPERATIVE C
NOTE: IF
MATERIAL
SHALL BE
FAST
TRACK-REMOVAL OF HAZARDOUS
IS IMPLEMENTED
DELAYED SO AS
. THE REMEDIAL
TO COMMENCE AT
MASTE
INVESTIGATION
THE
COMPLETION
OF THE REMOVAL
PROCUREMENT
.
REMEDIAL INVESTIGATION
NORMAL SEQUENCE FEASIBILITY
0 CONTRACTOR
AGREEMENT
STUDY
0 —
0
REMEDIAL
INVESTIGATION
0-
COMMUNITY DEVELOP
RELATIONS FINAL WORK
PLAN 0 PLAN 0
< | IMPLEMENT COMMUNITY RELATIONS PLAN I >
Figure 1.
Master Site Schedule
-------
REMEDIAL RESPONSE
127
estimated work schedule in calendar days are summarized in Table
1. Cost ranges are provided to account for any unknown site condi-
tions that may be encountered during the limited remedial in-
vestigation stage. The cost estimates, work schedules, and a detail-
ed work statement (prepared as part of the third section of the
,RAMP) provide the basis for an allocation of Superfund monies to
conduct the work at the site.
Table 1.
Summary of Remedial Projects
Schedules and Cost Estimates
Estimated Cost
Schedule
Project (Cal. Days)
Community Relations Plan As Needed
•Limited Remedial
Investigation/Feasibility
Study 90
•Remedial Design 90
•Implementation: Removal
of Hazardous Waste
Materials 180
Remedial Investigation 120
Feasibility Study 120
Remedial Design
Implementation of Source
Control and/or Off-Site
Measures
Post-Closure Monitoring
•Fast Track
Low High
Lead Agency Lead Agency
$ 30,900
$ 25,800
$1,800,000
$ 122,000
$ 116,800
$ 44,600
$ 39,000
$2,428,000
$ 161,400
$ 163,700
In this case, the State and USEPA have used the RAMP infor-
mation to develop a cooperative agreement and the State is now
preparing to conduct the limited remedial investigation and
feasibility study at the site. The cost estimates developed in the
RAMP established the amount of the funding request and helped
to identify the direct project management costs of the State.
Appendices
The final section of the RAMP includes the appendices. The
most important appendix is the detailed statement of work for the
first phase of activity that needs to be completed according to the
procedures established in the NCP. The statement of work outlines
the major activities of the remedial investigation, feasibility study,
initial remedial measure or any phase of activity which can be pro-
perly scoped with the use of available data. In the example
previously illustrated, the statement of work provided the basis for
the work plan of the cooperative agreement with the State.
Other appendices may include summaries of sampling and
analytical data, references, draft community relations plan, a
detailed site chronology and other pertinent information.
CONCLUSIONS
The RAMP has been developed as a management tool to assist
with the technical and financial responsibilities of the remedial ac-
tion program. The simplistic structure facilitates quick revision as
site conditions change and as more definitive objectives for each
site are established. States will be conducting remedial activities at
some sites and USEPA at others. 'Since the State and Federal
resources to conduct these activities are limited, these resources
must be used in a prudent and consistent manner at sites across the
country. The RAMP is one vehicle to meet these objectives.
-------
THE DEPARTMENT OF DEFENSE'S INSTALLATION
RESTORATION PROGRAM
DONALD K. EMIG, Ph.D.
Office of the Assistant Secretary of Defense
Washington, D.C.
BACKGROUND
The Department of Defense (DoD) began its Installation Restor-
ation (1R) program in 1975 before the passage of CERCLA. The
IR program is a comprehensive program to identify and evaluate
past DoD hazardous waste disposal sites on DoD installations,
and to control the migration of contamination resulting from such
operations. The IR program also is applicable to property that
is excess to DoD requirements and which might be made available
for other public or private use.
DoD initiated the IR program prior to any public outcry or legis-
lative mandate because of its concern for the public health and wel-
fare and environmental quality. The agency is proud of the pro-
gram and the leadership it has demonstrated to federal, state, and
local governments and to the private sector.
On Aug. 14, 1981, in Executive Order 12316, the President dele-
gated certain authority specified in the Comprehensive Environ-
mental Response, Compensation, and Liability Act (CERCLA)
to the Secretary of Defense. The Secretary of Defense was given
responsibility for:
•Response actions (i.e., removal and remedial actions)
•Investigation, monitoring, survey, and testing
•Such planning, legal, fiscal, economic, engineering, architectural,
and any other studies or investigations as necessary for response
actions, cost recovery, and to enforce the provisions of CERCLA,
for DoD facilities and vessels
The National Contingency Plan goes on to further recognize
DoD on-scene coordinators for DoD facilities and vessels.
Within DoD, the Secretary of Defense's authority in Executive
Order 12316 has been re-delegated to the Secretaries of the Army,
Navy, and Air Force. The Assistant Secretary of Defense for Man-
power, Reserve Affairs and Logistics on Nov. 20, 1981, formally
identified DoD's functioning IR program as the DoD Super-
fund program.
The objectives of the DoD restoration program are these:
•To identify and evaluate past hazardous material disposal sites
on DoD facilities, to control contamination migration, and haz-
ards to health or welfare
•To review and decontaminate as necessary land and facilities
excess to DoD's mission
These objectives go beyond the Superfund mandate.
DoD has required that the military departments and the Defense
Logistics Agency establish and operate installation restoration pro-
grams, and complete records searches at every installation listed
on their priority lists by the end of FY 1985. They have also been
required to develop and maintain a priority list of contaminated
installations and facilities requiring remedial action. In the IR pro-
gram, priority is given to control of migrating contamination that
may pose a threat to the public health and welfare of surround-
ing communities or to on-post personnel.
Other considerations in establishing the priority lists are:
•The installation's mission
•Suspected or known contamination hazard
•Possible land excessing action
•Environmental sensitivity of area
•Known disposal sites
•Public interest and any other factors considered appropriate by
each DoD component—Army, Navy, Air Force, and the Defense
Logistics Agency.
Each of the military departments has assigned a principal role
to in-house environmental organizations to coordinate or accom-
plish installation restoration program actions (Fig. 1). The Army's
Toxic and Hazardous Materials Agency is located at the Edge-
wood area of Aberdeen Proving Ground, Maryland. The Navy's
Energy and Environmental Support Activity is located at Port
Hueneme, California, and is responsible for the Navy and Marine
Corps program. The Air Force's Engineering and Services Center
is located at Tyndall Air Force Base, Florida, and is responsible
for the installation assessment phase of the Air Force program.
The Occupational and Environmental Health Laboratory is located
at Brooks AFB, Texas, and is responsible for the confirmation, or
field survey, phase of the Air Force program.
Air Force
Army Materiel
Development
and Readiness
Command
Naval Facilities
Engineering
Command
Air Force
Engineering
and
Services Center
Air Force
Occupational
and
Environmental
Health
Laboratory
Army Toxic and
Hazardous
Materials
Agency
Naval Energy
and
Environmental
Support Activity
Figure 1.
DoD Installation Restoration (Superfund) Organization
128
-------
REMEDIAL RESPONSE
129
THE IR PROGRAM: IN CONCEPT
Installation Assessment Phase
The first phase in the IR program (Table 1) is an installation
assessment. In this phase, installation files are examined, current
employees and key retirees are interviewed, and the terrain and
facilities are examined. Additionally, all available information on
past mission, current operations, waste generation, disposal, and
geohydrology of the area are collected. Limited soil and water
sampling may also be conducted to determine if contaminants are
present.
A preliminary survey may be conducted to produce field data
which will either confirm or rule out the presence and/or migra-
tion of contaminants, and assess the degree of risk associated with
the identified condition. If contamination exceeding safe levels is
found, interim containment of the migration is accomplished as
soon as possible. If not, no further action is required.
Confirmation Phase
The second phase in the IR program is referred to as the con-
firmation phase. In this phase, a comprehensive survey is con-
ducted to fully define the problem through environmental sampl-
ing and analyses. Data are developed to fill identified informa-
tion gaps revealed during the installation assessment phase, and
survey data from all technical areas is interpreted and interrelated.
The survey will normally be sufficiently detailed to determine the
locations of contaminant sources, quantify the contamination pres-
ent, and define the extent of contaminant migration and the con-
taminant boundaries.
Depending on the nature of contamination being investigated,
an ecological survey may be used to collect baseline data on plant,
animal, and aquatic species to determine the presence of target
contaminants throughout the ecosystem.
Technology Base Development
In the third phase, referred to as Technology Base Develop-
ment, control technology is matched with specific contamination
problems at a given site to determine the most economical solu-
tion. If control technologies do not exist, they are developed in
this phase. Technical specifications and corrective project pro-
gramming documents may also be developed. Additional ground-
water monitoring wells and sampling could be required to address
specific technology application considerations and fill in remain-
ing data gaps. Technology development may also support further
project planning to control migration, or to support actual restor-
ation efforts on contaminated properties.
A DoD IR Technology Coordinating Committee, chaired by the
Army, has been established recently to facilitate the exchange of
technical information among the military departments and to re-
view control technology or other criteria that may be useful.
Also, in this phase decontamination criteria are identified, or
developed if they do not exist. Criteria development defines the
Table 1.
DoD IR Program Concept
I. Installation Assessment
1. Records Search
2. Preliminary Survey
II. Confirmation
1. Comprehensive Survey
2. Ecology Survey
III. Technology Base Development
1. Containment/Decontamination Technology Development
2. Criteria Development
3. Cost/Benefit Analysis
IV. Operations
1. Project Plan
2. Survey
contamination levels judged acceptable for protection of the pub-
lic health and welfare, based on best available information, in the
absence of federal or state standards. Criteria development efforts
will generally include: (1) problem definition studies, (2) chemistry
studies, and (3) toxicological studies.
Problem definition studies provide baseline information for data
development through literature searches. Problem definition stud-
ies also identify the research necessary for an objective assessment
of adverse effects. When the contaminant situation is unique to the
DoD (i.e., a DoD-unique pollutant), the need for conducting chem-
istry and toxicology studies will be evaluated to decide whether to
assess the risk associated with various concentration levels by con-
trolled laboratory experiments on various animals or plant species.
Since the criteria to be developed are intended for widespread
use, extensive coordination of presumptive criteria, research proto-
cols, and results is accomplished. The results of these toxicolog-
ical and chemistry studies are made available for peer review.
The National Academy of Science may be requested to review
presumptive criteria, research protocols and results. Additionally,
experts from the Environmental Protection Agency, Department of
Health and Human Services, Department of Agriculture, Depart-
ment of the Interior, and other concerned agencies may be asked
to comment when appropriate. Criteria developed by the Navy and
Air Force are provided to the Army for incorporation into a DoD
database.
After collecting data on standards, contaminant control tech-
nology, ecological effects, and current environmental levels of con-
tamination, DoD evaluates the costs and benefits of applying
available control technologies to reduce the hazard from contam-
inants. The options considered would normally include a no-action
alternative. This option is of particular interest when property is
to be excessed to non-DoD ownership for a like-usage, and low-
levels of contaminants are present, and not migrating.
Operational Phase
The final phase of the IR program, when required, is what DoD
refers to as the operations phase. This phase includes design,
construction, and operation of pollution abatement facilities, and
the completion of remedial actions. This phase could include the
construction of containment facilities or decontamination pro-
cesses, and associated monitoring systems. Also included are com-
pleted project phase-out, data storage, and publication of a final
project report.
A project plan describes the survey results, analyzes feasible
control options, and presents a proposed approach for controlling
the contamination. An estimate of resources, research require-
ments (if any) associated with the proposed solution, and a plan-
ning schedule for design, construction, and operations are also
included.
A survey will be conducted prior to, and following, the opera-
tions phase to ensure that target standards or criteria are achieved.
IR PROGRAM ACCOMPLISHMENTS
DoD's goal is to complete records searches at every installa-
tion currently identified on Army, Navy, Air Force, and the De-
fense Logistics Agency's priority lists by the end of FY 1985. The
Deputy Assistant Secretary of Defense for Installations monitors
semi-annually the progress of the military departments toward this
goal. The IR program results to date are impressive (Table 2).
As of June 30, 1982, 210 Army installations have been iden-
tified which require a records search, 156 installations have been,
or are being assessed largely by contract effort. 104 records
searches have been completed and published at an approximate
cost of $50,000 per installation. Of these 104, 49 surveys are re-
quired, 35 surveys are underway and 14 have been completed. Typ-
ical costs of the surveys, which are done by contract, is around
$300,000 per installation.
To date, eight installations have been determined to require re-
medial action. They are: Rocky Mountain Arsenal, Colorado, Red-
stone Arsenal, Alabama, Alabama Army Ammunition Plant, Ala-
-------
130 REMEDIAL RESPONSE
Table 2.
DoD Installation Restoration (Superfund)
Program Accomplishments
Records Search
Required (Total)
Completed
Survey
Required (Total)
Completed
Operation
Required (Total)
Completed
Army
( + DLA)
210
104
49
14
8
2
Navy
80
1
0
0
0
0
Air Force
82
26
26
4
2
0
bama, Milan Army Ammunition Plant, Alabama, Anniston Army
Depot, Alabama, and Savannah Army Depot, Illinois. Work at
two installations has been completed. Pine Bluff Arsenal, Arkan-
sas; and Frankford Arsenal, Pennsylvania. Costs associated for
this phase are site specific and vary considerably, ranging to date
from $ 1.6 to $44 million.
Eighty Navy installations require records searches. Twenty-
eight have been initiated to date, and one final report has been
published. The Navy is currently evaluating their preliminary re-
sults to determine how many surveys are required. To date, two
installations have been determined to require remedial action.
These are the Jacksonville Naval Air Station, Florida, and a site at
Pearl Harbor, Hawaii.
Eighty-two Air Force installations require a records search.
As of June 30, 1982, 26 records searches have been completed and
published. All 26 installations require surveys, and 4 are in pro-
gress. Two installations have remedial actions underway. They are
Wurtsmith Air Force Base, Michigan, and Air Force Plant 44,
Arizona.
Probably the most widely reported DoD installation restoration
project is the Army's effort at Rocky Mountain Arsenal, Colorado.
Rocky Mountain Arsenal is adjacent to the City of Denver,
with Stapleton International Airport lying directly south of the Ar-
senal. The Arsenal complex, some 17,000 acres, has groundwater
contaminated as a result of the military production of chemical
warfare agents and Shell Chemical production of commercial
pesticides. The Army's project at Rocky Mountain Arsenal has
been active since 1976. To date, over $44 million has been spent to
define contaminant migration and provide remedial actions at the
Arsenal, with the end not yet in sight.
The data collection and analytical effort at the Arsenal has been
extensive. Over 1500 wells have been drilled on the 25 square mile
site. 4,000 to 6,000 analyses are completed every month, with over
270,000 data points on record, and over 900 technical reports pub-
lished. Since this project has been discussed extensively elsewhere,
suffice it to say, though, that the Army has demonstrated at Rocky
Mountain Arsenal the commitment of the Defense Department
to aggressively pursue its installation restoration program.
CONCLUSION
DoD has developed substantial expertise in the identification,
characterization, and control of environmental contamination.
DoD will continue to actively pursue its Installation Restoration
program over the next several years.
However, the IR program is only one important part of theDoD
environmental quality programs. DoD has demonstrated signifi-
cant initiatives and accomplishments in other important areas of
concern. DoD believes, though, that its Installation Restoration
program has demonstrated significant leadership, and serves as a
model for other agency Superfund programs. Hazardous waste dis-
posal site problems will remain a major area of emphasis within
DoD for the 1980s.
-------
SURVEY AND CASE STUDY INVESTIGATION OF REMEDIAL
ACTIONS AT UNCONTROLLED HAZARDOUS WASTE SITES
S. ROBERT COCHRAN
MARJORIE KAPLAN
PAUL ROGOSHEWSKI
CLAUDIA FURMAN
JRB Associates
McLean, Virginia
STEPHEN C. JAMES
Municipal Environmental Research Laboratory
U.S. Environmental Protection Agency
Cincinnati, Ohio
INTRODUCTION
With the passage of the Comprehensive Environmental
Response, Compensation, and Liability Act (Superfund) and the
release of the National Contingency Plan, the USEPA has de-
veloped a systematic methodology for assigning liability and con-
ducting remedial actions at uncontrolled hazardous waste manage-
ment facilities. Prior to the passage of Superfund, USEPA relied
on two laws for providing assistance for remedial actions at uncon-
trolled hazards waste sites, the Clean Water Act (Section 311) and
the Resource Conservation and Recovery Act.
It is clear from past experiences such as Love Canal that short
term and long term environmental and health related hazards exist
when inadequate technologies are used during the handling and
disposal of hazardous materials. Presently the Solid and Hazardous
Waste Research Division USEPA (Cincinnati, OH) and the Oil and
Hazardous Spills Branch USEPA (Edison, NJ) are involved in the
research and development of existing and novel technologies for
use in the remediation of hazardous materials released to the en-
vironment.
Because new remediation techniques are continually evolving
and known technologies are constantly being retrofitted and
adapted for remedial actions use, the alternatives available for
cleanup are continuing to expand. In order to assess the effective-
ness and limitations of these remedial actions techniques, USEPA
is conducting case study examinations of these technologies which
have been employed in field situations.
PROJECT DESCRIPTION
JRB, under contract with the USEPA, is in the process of con-
ducting a nationwide survey of uncontrolled hazardous waste sites
to identify and examine the various types of remedial action
technologies which have been implemented or which are in progress
or are proposed as cleanup techniques. At selected sites, detailed
case studies will be performed to document the specific reasons for
the success or failure of applied remediations technologies and to
determine the limitations and applicability of these technologies in
other restoration activities or situations.
The results of the case study investigation will be used to provide
a transfer of information between those who have experienced the
implementation of a remedial action and those contemplating or
presently assessing or pursuing some form of site cleanup. The an-
ticipated audience to be assisted by the survey findings and case
study reports includes members of industry and commerce, State
Agencies, Local Authorities, and the USEPA.
Case Study reports will be structured so that they will provide
detailed data on the remediation techniques employed, the cir-
cumstances and conditions in which they were implemented, their
apparent effectiveness in correcting or controlling the problem, and
their potential uses in other natural environments and remedial ac-
tion situations. Ultimately the survey and case study results will
quantify the number and type of reported uncontrolled releases of
hazardous substances that have undergone a certain degree of
remediation, will provide a standard for comparison when assess-
ing or deciding on a plan for remediation, will identify cleanup
technologies which may warrant further research, and most impor-
tantly will provide a forum in which others can learn from past
miscalculations and successes.
Survey Method
The initial task was to compile an updated list of uncontrolled
hazardous waste sites which included both ongoing and completed
remedial actions. The sources of information used to compile this
list included: in-house literature review of selected documents, data
from USEPA on present and past remedial action programs and
site studies, including work at Superfund sites, Department of
Defense contacts with knowledge of restoration work at military
bases, State environmental and health agencies involved in site
remedial actions, and industrial and pertinent trade association
contacts involved in spill responses or hazardous substances
cleanups.
The first step in conducting the survey began by conducting a
review and updating data on the 199 sites identified by SCS
Engineers' and the data contained within the list of 114 Top Priori-
ty Superfund Sites released Oct. 23,198.1. Simultaneously, an inter-
nal review was conducted of in-house files and publications, in-
cluding the Groundwater Newsletter, Hazardous Waste Reporter,
and Hazardous Materials Report, to identify any sites not contain-
ed in the SCS Report or Superfund list.
Once the data review was completed, USEPA Regional, State
and Local parties identified as being knowledgeable about hazar-
dous site remediation activities were contacted to supplement infor-
mation collected from the internal survey and to expand the site
data base. Knowledgeable parties contacted included USEPA
Regional Emergency Response Coordinators, Regional Land
Disposal Branch personnel, Regional On-Scene Coordinators,
State On-Scene Coordinators, consulting contractors, site manager
and operators, Department of Defense, and State and Local of-
ficials.
Cooperation was requested from trade associations representing
industries which may have been involved in hazardous waste
management, generation or transportation. Contact was also made
with the firms involved with the design and implementation of
remedial actions at hazardous materials spills sites and waste
management facilities.
Approximately 300 sites were identified. Data collected on the
sites included the name and location of the site, the remedial ac-
tions implemented or planned, the type of waste management prac-
tices used at the facility, the waste types/contaminants present, the
availability of engineering cost data, and the ease of access for case
study. A clear majority of the sites surveyed were obtained from
two data sources, the SCS Report and the 114 top-priority Super-
fund sites. Contacts made during discussion with responsible par-
ties involved in remediation activities at the Superfund sites and
SCS sites often led to discovery of sites not previously listed.
131
-------
132
REMEDIAL RESPONSE
The results of the DOD survey effort indicate that remediation
activities within the armed forces are in the initial stages. DOD has
established a phased approached for conducting site restoration ac-
tivities within all branches of the Armed Forces. Presently, each
branch of DOD is proceeding at individual rates relative to the
phase program, and in most cases have conducted initial site
assessments but have not initiated any site restoration activities.
Contacts with several cleanup firms, consultants specializing in
remedial alternative design, and trade associations led to the
discovery of several sites. However, in most cases client confiden-
tiality agreements hindered full cooperation in identifying sites. It is
reasonable to assume the many private sector cleanups are com-
pleted and not reported and therefore are not identifiable and are
not included in this population of surveyed sites.
Once the site survey activities were completed a series of criteria
were used to select candidate sites for detailed case study analysis.
These criteria included:
•Availability for field survey activities
•Availability, accessibility, and completeness of remedial action
cost and engineering data
•Type of remedial action technology implemented, so that a range
of remedial action techniques were investigated
•Type of waste management practice, so that a wide range of
technologies common to hazardous waste management were
studied
•Types of waste and contaminants present at the facility to ensure
that a variety of waste streams and pollutants were included
•Hydrogeologic setting, so that a variety of settings were repre-
sented
•Geographic location to provide a nationwide distribution of sites
Approximately 20 sites have been identified for case study analysis
and field surveys for these sites are presently being performed.
Survey Results
Three hundred and sixteen sites have been identified as being
associated with some form of completed, ongoing or planned
remediation activity relative to the uncontrolled releases of hazar-
dous substances to the environment (Table 1). This is a substantial
increase in the number of sites identified in a similar survey con-
ducted in 1980. Two key factors can be attributed to the increased
number of identified sites:
•The data base used in the 1982 survey was an expanded version
of the 1980 data base, including contact with the Department of
Defense, pertinent trade associations, industrial contacts, and
cleanup contractors
•Additional data relative to hazardous waste sites have been made
available through the USEPA Field Investigation Team activi-
ties over the past two years and there has been increased aware-
ness by state and local officials relative to site discovery and
identification
Even though the general population of identified sites has in-
creased by 147, the percentage of sites located in the individual
states has not changed. Also, a majority of sites identified are con-
centrated in the heavily industrialized regions of the country. This
corresponds well with the economics and convenience of disposing
of waste materials near the generating source.
A compilation of the various types of disposal methods
employed at the sites identified during the survey is found in Table
2. The data collected from approximately 30 of the sites surveyed
have not been accumulated in sufficient detail to enable an accurate
assessment of the disposal method used. Therefore, these sites were
eliminated from consideration when constructing Table 2.
Waste management practices and contamination releases iden-
tified during the survey included land disposal, drum and tank
storage, incineration and treatment, subsurface injection, spills
and illegal dumps. Seventy eight percent of the facilities where
remedial actions were either completed, ongoing, or planned can be
associated with three waste management technologies: landfilling,
Table 1.
State Location of Remedial Action Sites in 1980 and 1982'
State
Alabama
Alaska
Arizona
Arkansas
California
Colorado
Connecticut
Delaware
Florida
Georgia
Hawaii
Idaho
Illinois
Indiana
Iowa
Kansas
Kentucky
Louisiana
Maine
Maryland
Massachusetts
Michigan
Minnesota
Mississippi
Missouri
Montana
Nebraska
Nevada
New Hampshire
New Jersey
New Mexico
New York
North Carolina
North Dakota
Ohio
Oklahoma
Oregon
Pennsylvania
Rhode Island
South Carolina
South Dakota
Tennessee
Texas
Utah
Vermont
Virginia
Washington
West Virginia
Wisconsin
Wyoming
Total 50 States
•Percent of total identified nationwide.
No. of
Sites '80
2
0
3
2
3
3
4
2
7
4
0
0
8
3
1
2
5
3
2
1
5
11
3
0
4
5
0
0
1
10
0
14
7
6
4
0
0
16
4
3
0
10
3
1
0
2
0
1
3
1
169
% of
Total*
1.2
0.0
1.8
1.2
1.8
1.2
2.4
1.2
4.1
2.4
0.0
0.0
4.7
1.8
0.5
1.2
3.0
1.8
1.2
0.5
3.0
6.5
1.8
0.0
2.4
3.0
0.0
0.0
0.5
5.9
0.0
8.3
4.1
3.6
2.4
0.0
0.0
9.5
2.4
1.8
0.0
5.9
1.8
0.5
0.0
1.2
0.0
0.5
1.8
0.5
No. of
Sites '82
2
0
6
4
9
5
11
4
26
5
0
0
14
4
2
3
6
9
3
3
11
13
8
3
10
5
0
1
6
17
3
29
8
6
5
3
0
23
5
5
2
10
7
2
1
4
5
2
5
1
316
%of
Total*
0.6
0.0
1.9
1.3
2.8
1.6
3.5
1.3
8.2
1.6
0.0
0.0
4.4
1.3
0.6
1.0
1.9
2.8
1.0
1.0
3.5
4.1
2.5
1.0
3.2
1.6
0.0
0.3
1.9
5.3
1.0
9.2
2.5
1.9
1.6
1.0
0.0
7.2
1.6
1.6
0.6
3.2
2.2
0.6
0.3
1.3
1.6
0.6
1.6
0.3
Table 2.
Types of Disposal Methods
Disposal Methods
Landfill
Drum Storage
Surface Impoundment
Spills
Incinerator
Injection Wells
Illegal Dumps
Total
No. of Sites Percent of Total
121
67
104
25
10
7
39
373
In some cases, more than one disposal method has been used at an individual site.
32
18
28
7
3
2
10
100%
-------
REMEDIAL RESPONSE
133
drum storage, or surface impounding. This association is to be ex-
pected based on the fact that these technologies have been the most
common methods over the years for managing and disposing of
hazardous substances.
The relatively low percentage of incinerators and injection wells
identified in the survey can be attributed to the limited applicability
of these technologies for treating and disposing of a wide spectrum
of hazardous materials, and to their limitations for use in broad
geographic and environmental regions. The low percentages of il-
legal dumps identified can be associated with the expected
unavailability of data necessary to sufficiently characterize these
facilities.
The remedial technologies used at the surveyed sites included a
variety of methods such as capping and grading, removal and off-
site burial, groundwater pumping, chemical and biological treat-
ment, and containment and encapsulation (Table 3). However,
76% of the remediation efforts identified involved only four
Table 3.
Types of Remedial Action Employed
Remedial Action
Capping/Grading
Drum Removal
Contaminant Removal
Groundwater Pumping
Groundwater Containment
Contaminant Treatment
Encapsulation
Dredging
Gas Control
Incineration
Lining
Total
No. of Sites Percent of Total
59
56
70
22
23
48
8
5
3
3
7
304
19
18
23
7
8
16
3
2
1
1
2
100%
In some cases, more than one remedial action technique has been used at an individual site.
techniques, capping/grading, drum removal, contaminant removal
or treatment. In most instances, these techniques correlate well
with the high percentage of facilities identified as practicing land-
filling, drum storage, or surface impounding as a waste manage-
ment method. This correlation is based on the following factors:
•Most remedial actions to date have been directed at controlling
the immediate threat, i.e. removal of the waste material by land-
fill and contaminated soil removal, surface impoundment pump-
ing and removal, or drum removal.
•Technologies such as grading/capping, contaminant removal,
and drum removal are in most cases relatively unsophisticated
and economic remedial activities when compared with other re-
medial options.
•In the natural order of implementing remedial actions, removal
of the contaminant source is the most likely initial step in per-
forming a staged facility cleanup.
•Complete removal of the source of contamination is the most
effective and direct method of reducing or eliminating continued
releases of contaminants to the environment.
Less often used methods of site remediation include encapsula-
tion, dredging, incineration, groundwater containment or pump-
ing, and subsurface gas control. Several reasons exist for the
restricted use of these methods as remediation techniques, these in-
clude:
•Constraints based on site specific conditions such as waste type,
area of contamination, and media contaminated
•Present level of technological development relative to proven
field use and successful application in real world situation
•Effects of economic and institutional factors which limit the
availability of funds, or requirements for specific approaches to
be used in resolving the cleanup problem (i.e., emergency clean-
up vs. long term cleanup)
Based on the information gathered on the restoration activities
of relatively new sites, it can be said that there has been a shift in
the approach historically used to conceptualize and implement
remedial actions. In more recent times, an emphasis has been
placed on thoroughly understanding the dynamics of the con-
tamination and then developing and implementing the remedial ac-
tion program. The shift to more systematic approaches in planning
remedial actions can be attributed to more comprehensive regula-
tions relative to hazardous waste management and the assignment
of strict site cleanup liabilities and responsibilities (i.e., Superfund
and the Resource Conservation and Recovery Act), and also to
lessons learned from successful and unsuccessful cleanup methods
and techniques employed in the past.
CASE STUDIES
Once JRB had selected the case study sites, field activities were
conducted to collect additional information to ensure the develop-
ment of accurate and complete case history reports. Field visits in-
cluded trips to the sites as well as meetings with the appropriate
Federal, State, and private parties involved in the site remediation.
Presently, JRB is in the process of completing case study investiga-
tions and a companion report on these investigations. The follow-
ing .is a brief synopsis of three case study investigations.
Trammel Crow Site—Dallas, Texas
The first case study was a former Texaco oil refinery which
operated in Dallas, TX from 1915-1945. When the refinery ceased
operation, it was sold to a metal recycling firm which reclaimed any
valuable scrap metal on site. The property was left vacant for 20
years after the scrapping operation until it was sold in 1980 to the
Trammell Crow Corporation for industrial development.
At the time of purchase, five waste sludge pits with over five
million gallons of still bottoms and other petroleum refinery wastes
were located on the site. The five ponds were not all the same size
and did not contain the same materials. One pond, the largest, con-
tained approximately 3.5 (16,600 yd3) million gal. of crude oil tank
bottoms consisting of 50% carbonaceous material, 35% water, and
15% ash. The carbonaceous portion of the sludge was made up of
equal portions of asphaltenes and paraffins. A second contained
approximately 10,000 yd3 of hard-coke/slag material believes to be
coke cinders from the petroleum refinery cracking process. The re-
maining three ponds contained approximately 1.5 million gals of
sedimentation or oxidation pond oily residues.
The waste ponds were located above a Trinity Clay soil which ex-
tends to a depth of 20 to 45 ft below the ground surface. The Trini-
ty Clay is a moderately alkaline soil of low permeability. This in
turn, overlies in Eagle Ford shale which extends to a depth of ap-
proximately 400 ft below the surface. The Trinity Clay and Eagle
Ford shale formations form an impermeable barrier between the
surface and the underlying aquifer, therefore posing little threat of
groundwater contamination.
The wastes were tested for toxicity using the USEPA required ex-
traction procedure test as defined by the Resource Conservation
Recovery Act (RCRA). The sludges were not found to be RCRA
hazardous waste, however, the sludges in the ponds had to be
treated, removed or both before development could take place.
Although the waste oil was not RCRA hazardous waste, it was
classified under Texas Law as a Class II industrial waste and could
not be put into a municipal landfill. The costs of transporting the
sludge to the closest industrial waste landfill were much greater
than the alternative devised, which was to solidify and dispose of
the sludge in a site with the approval of the Texas Department of
Water Resources. Although the remedial approach was used for a
non-hazardous waste oil, it is apparent that the same approach
could be applied to similar sites where oily waste sludges are also
defined as hazardous wastes.
The technology chosen was a sludge solidification process using
cement kiln dust available locally. Instead of using only fresh kiln
-------
134
REMEDIAL RESPONSE
dust (which is limited in supply because of outstanding contracts
for the material) to mix and solidfy with the sludge, a stale kiln dust
was also used. The stale dust, which contains 38% moisture, is
usually stockpiled in large quantities, is less expensive, and is a very
effective solidifying agent.
The solidification procedure began with the three smaller ponds
using stale cement kiln dust. A landfill was excavated next to the
three ponds. Stale cement kiln dust was delivered to the bottom of
the landfill and leveled by bulldozer into a 6 to 12 in. layer. A
backhoe lifted the sludge out of the pits and on top of the kiln dust
in the landfill. The bulldozer then mixed the dust and the sludge in-
to one foot layers 3:1 and 2:1 dust to sludge ratios. A pulverizing
mixer was then driven over the layer to completely homogenize the
mixture. Each layer was air dried for approximately one day and
then compacted and field tested to ensure proper compaction. This
procedure continued until the contents of all three ponds were
solidified in the landfill.
The largest pond solidification was somewhat difference because
the sludge was more liquid and the pond itself was several thousand
feet from the landfill. Because of its more liquid nature, the sludge
was solidified using both fresh and stale kiln dust. The procedure
for solidifying the sludge incorporated the same layering process
used for the three smaller ponds, however, the sludge was treated
prior to placement in the on-site landfill.
Fresh cement kiln dust was blown into the sludge pond using a
ratio of 1.5:1 dust to sludge. A backhoe mixed the dust into the
sludge which semi-solidified the sludge. The semi-solidified sludge
was loaded, transported to the on-site landfill, and unloaded
onto a bed of stale kiln dust and mixed as previously described. As
each side of the pond was excavated it was backfilled with clean dirt
until all of the sludge had been removed and solidified in the on-site
landfill. The coke/slag material was also removed from the remain-
ing pond and mixed in with the sludge in the on-site landfill.
This entire process took approximately 75 working days using
approximately 41,00 tons of kiln dust. This was much less than
originally anticipated because the stale kiln dust solidified with the
sludge better in the field than in the laboratory. The projected cost
of $500,000 had been based on an estimated use of 75,000 tons of
kiln dust, however, the project cost only $377,528. This was much
less than the alternative of off-site disposal which was estimated to
cost $1,500,000.
Presently, the site is completely solidified, graded over, and
covered with vegetation. It is awaiting sale for future development.
Goose Farm Site—Plumsted Township, New Jersey
The Goose Farm site is an abandoned hazardous waste drum
burial site that was partially cleaned up in about a one year period
from Aug. 1980 to Nov. 1981. The drum burial site is located in
Plusted Township, Ocean County, NJ in a region known as the
Pinelands, a unique ecological area characterized by acidic sandy
soils and low lying forests of pitch pine and oak.
From about 1945 to in the mid 1970s, Thiokol Corporation (a
manufacturer of ammunition, rocket fuel, plastics, and organic
fibers) dumped bulk and containerized hazardous wastes into a pit
about 300 ft long by 50 ft wide by 15 ft deep, under contract with
the site owner. In January 1980, township officials informed the
New Jersey Department of Environmental Protection (DEP) of the
existence of hazardous wastes at the Goose Farm site. Shortly
thereafter, DEP conducted a preliminary investigation of the site,
which revealed the discharge of contaiminated groundwater as a
surface seep to a nearby stream. During the next six months, site
activities included sampling of the stream and groundwater from 17
monitoring wells, and reviewing regional geological data and well
driller logs.
A report assessing the situation at Goose Farm was submitted in
June 1980, which described a contaminated groundwater plume
about 200 ft wide originating from the drum burial area and mov-
ing into the stream where it was discharging. The plume was also
believed to be moving downward because of the geology in the
area. Preliminary testing was inconclusive, but suggested potential
contamination of up to 60 ft in depth beneath the site, with the ma-
jority of contamination occurring within a depth of 40 ft.
Monitoring of the upper groundwater and surface seepage in-
dicated that a large variety of inorganic and organic chemicals had
been dumped, including chlorinated compounds, solvents, and
pesticides. Levels of contaminants were highest for methylene
chloride, benzene, and toluene, in the 100 mg/1 range. Total
organic carbon (TOC) concentration ranged from 1,600 to 17,000
mg/1. The report also indicated that the plume was not an im-
mediate danger to any drinking water wells in the area, but if un-
corrected could cause widespread contamination of the lower
strata, a moderately used aquifer.
The state established the following objectives and hired an oil
and hazardous waste cleanup contractor to carry out the following
tasks:
•Obtain data to confirm contamination of the surface water by the
Goose Farm site and to better define the extent of groundwater
contamination
•Contain the discharge to the stream by pumping and treating the
groundwater
•Remove the source of pollution, i.e. buried wastes and contam-
inated soil
•Further groundwater removal and treatment near the source area
Temporary containment measures such as the installation of an
open or gravel filled cut-off trench were suggested but were not im-
plemented.
The contractor proceeded with cleanup efforts, consisting of
constructing a well-point treatment system to prevent the contami-
nant plume from discharging into the stream, and excavating,
segregating, and treating the buried waste materials in the disposal
area. The well-point system was located between the disposal area
and the stream and was oriented in such a way as to intercept the
contaminant plume. The well-point system consisted of about 400
ft of 6 in. aluminum header pipe with 100 well-points spaced about
every 7.5 ft. The well-points were installed to a depth of 22 ft by
water jetting. The system was pumped at a rate of about 50,000 to
75,000 gal/day to contain migration of the contaminants.
The collected groundwater was routed to a treatment system con-
sisting of the following:
•A vacuum receiver which volatilized about 20% of the TOC in the
stream
•Activated carbon adsorbers to treat the gaseous stream of the
vacuum receiver
•A clarifier which reduced suspended solids and removed about
10% of TOC in the stream
•Multi-stage activated carbon adsorbers which removed about
70% of the initial TOC.
•After treatment, the effluent was spray irrigated
The final effluent from treatment had a TOC concentration of ap-
proximately 54 mg/1, a reduction of 96.6% to 99.7%.
Concurrently, waste removal operations were carried out in the
pit area. During a 45 day period, over 4,880 drums and containers
were excavated, analyzed, secured, and segregated. A backhoe and
two specially designed drum grapplers were used to complete
removal operations. Salvagable drums were overpacked and stored
on-site. Badly degraded drums were tested for acid-base com-
patibility and emptied into concrete holding tanks prior to disposal.
In addition, about 3,500 tons of contaminated soil were excavated
and temporarily stored, awaiting disposal.
After the drum pit area had been excavated, a second well-point
system was installed in the pit area. Groundwater was pumped to
the treatment system and the effluent was injected into the pit
area via well points to flush contaminants from the underlying
soils. The extraction well-points were first installed in the un-
saturated zone to a depth of 7 to 13 ft and later lowered into the
water table to collect groundwater contamination.
The groundwater treatment system was shut down in March 1981
before treatment was complete because of operating expenses.
-------
REMEDIAL RESPONSE
135
Secured drums, bulked liquids, and contaminated earth were stored
on site from March 1981 to Nov. 1981, awaiting funding for final
disposal operations. In Nov. 1981, 4,400 tons of waste were
transported in 200 truck loads to an approved waste disposal site.
In addition, 12 drums of PCBs were transported to an approved
hazardous waste incinerator.
Stroudsburg Site—Stroudsburg, Pennsylvania
The Stroudsburg site is located in the Borough of Stroudsburg,
Monroe County, Pennsylvania, along the shores of Broadhead
Creek at the site of a historical coal gasification plant once operated
by Stroudsburg Gas Co. From the late 1800s to the early 1900s, one
of the disposal methods practiced on site for the disposal of coal tar
residuals was to inject the waste material into the ground through
an injection well, located in the northwestern quadrant of the plant
property. The well was constructed such that the waste products
were injected into the gravel alluvium that underlies the plant area,
approximately 20 ft below the surface.
In 1947, Pennsylvania Power and Light Company (PP&L) pur-
chased several parcels of land from Stroudsburg Gas Co., most of
which were located along the shores of Broadhead Creek. In Oct.
1980, during maintenance construction of the flood control levees
along the western shore of Broadhead Creek, a black, odorous
material was observed seeping from the base of the dike at several
locations along the side of the stream.
The incident was reported to the Pennsylvania Department of
Environmental Resources (DER), Bureau of Water Quality and in-
vestigations commenced to determine the nature and extent of the
contamination. A preliminary assessment of the situation was made
in March 1981 and at this time the DER requested the assistance of
the USEPA in the further investigation of the problem. The in-
vestigative field studies that followed were conducted by DER,
USEPA and the Pennsylvania Fish Commission (PFC) and involv-
ed the following areas:
•The hydrogeology of the site area
•The impact of the coal tar on stream quality and its biological
community, and
•The erosional behavior of the stream
The remedial actions that were taken at the Stroudsburg site in-
volved several technologies. The initial remedial response to the
problem was conducted by USEPA in April 1981 under section 311
of the Clean Water Act (Public Law 92-500) and involved the in-
stallation of filter fences, sorbent booms and inverted dams. These
installations were temporary structures constructed to prevent
direct and immediate coal tar seepage into Broadhead Creek.
As field investigations continued, it was decided that actions per-
missible under section 311 (i.e., emergency, oil spill response ac-
tions) would not sufficiently remedy the pollution problem at the
Stroudsburg site. A more reliable and permanent remedial system
was necessary. In Nov. 1981, funds were appropriated under
Superfund, establishing Stroudsburg as the first site to receive
Superfund monies.
Also in Nov. 1981, PP&L began installation of a recovery well
system to remove the coal tar from the stratigraphic depression. An
estimated 35,000 gal of coal tar had accumulated in the depression.
The project, completed in late spring of 1982, consists of four well
clusters each containing four wells.
Presently, only one central well of one of the clusters is in opera-
tion and the recovery rate is less than had been anticipated. The
original recovery rate predicted was approximately 100 gal/day.
The present rate is between 20-25 gal/day. This situation was
primarily caused by an over calculation of the quantity of concen-
trated coal tar present. Much of what exists is actually a mixture of
coal tar and water, which was not realized until the wells had
already been installed. Approximately 7,500 gal of pure coal tar
(95% have been recovered to date). The recovered residue is sold to
Allied Chemical in Detroit, MI, where it is used as fuel.
Concurrent with PP&L's project, USEPA constructed a cement-
bentonite slurry trench cut-off wall along the west bank levee. The
containment wall is 700 ft long, one foot wide and 17 ft deep. The
downstream end of the barrier is horizontally keyed into an im-
permeable curtain formed by pressure grouting. The upstream end
is keyed into the sheet piling wall that exists below the concrete
flood wall. The slurry wall extends over the area of observed
seepage and passes vertically through the gravel layer that bears the
contaminant and is keyed two feet into the underlying sand
stratum. A total of eight monitoring wells were installed for
monitoring slurry wall performance and sampling groundwater in
the area. Four wells are located on either side of the wall.
In Jan. 1982, following slurry wall construction and the installa-
tion of the monitoring well system, the contaminated material in
the back water channel was excavated, drummed and disposed of at
a secure landfill. Throughout all the activities undertaken at the
Stroudsburg site, materials meant for disposal were packed in
drums and stored on-site until appropriate disposal sites were
located.
REFERENCES
1. SCS Engineers. "Survey Results and Recommended Case Studies—
Study of On-Going and Completed Remedial Action Projects." Pre-
pared for USEPA. June 9, 1980.
2. "EPA Announces First 114 Top-Priority Superfund Sites." Environ-
mental News, USEPA, Washington, D.C., Oct. 23, 1981.
3. Neely, N.S., Gillespie, D.P., Schauf, F.J., and Walsh, J.J., "Remed-
ial Actions at Hazardous Waste Sites: Survey and Case Studies." Oil
and Special Materials Control Division, USEPA, Cincinnati, OH,
January 1981.
4. Morgan, D.J., Novoa, A., and Halff, A.A., "Solidification of Oil
Sludge Surface Impoundments with Cements Kiln Dust." Preliminary
Draft. Albert A. Halff Associates, Dallas, TX.
5. Neely, N.S., Gillespie, D.P., Schauf, F. and Walsh, J.J., "Survey of
On-Going and Completed Remedial Action Projects." Proc. Manage-
ment of Uncontrolled Hazardous Waste Sites. Washington, D.C.,
October 1980, 125.
-------
CONCEPTUAL DESIGNS AND COST SENSITIVITIES OF FLUID
RECOVERY SYSTEMS FOR CONTAINMENT OF PLUMES OF
CONTAMINATED GROUNDWATER
D.A. LUNDY
J.S. MAHAN
Geraghty & Miller, Inc.
Annapolis, Maryland
INTRODUCTION
At least one nationwide study' has shown that removal and
treatment of contaminated fluids from wells or drains is the most
commonly used method for controlling the movement of a plume
of contaminated groundwater. Other methods, such as fluid en-
capsulation within subsurface physical barriers and in-situ treat-
ment or fixation, are being researched and applied, but those in-
volving fluid removal probably will continue to be the most pop-
ular on the basis of their cost and reliability.
The purpose of the authors in writing this paper is: (1) to des-
cribe simple hydraulic models for making conceptual designs and
cost estimates for recovery-well systems, and (2) to examine the
general sensitivities of recovery system costs to selected plume and
aquifer characteristics. The costs presented are for simplified
scenarios of contamination, which represent a foundation upon
which more complex models can be built. Although the more com-
plex models are increasingly being applied in the detailed design
and evaluation of systems of recovery wells, there is and will con-
tinue to be a place for simpler mathematical models that can pro-
vide rough estimates of system discharge and number/size of re-
covery wells.
The costs apply only to engineered structures and related serv-
ices for a containment system, with other cost elements such as
source removal, land purchase, legal services, etc., not included.
Also, the paper is concerned not only with containment and treat-
ment of contaminated groundwater, but also with total cleanup of
the site and the aquifer. The findings are derived from background
work performed for Sobotka & Company, Inc., in conjunction
with the Economics and Policy Analysis Branch Office of the
Office of Solid Waste, USEPA. The contents do not necessarily
reflect the views and policies of the USEPA.
Fundamental Concepts of Containment
"Plume containment" is defined in this paper as stopping
further migration of a plume of contaminated groundwater by
removing the contaminated water through wells located inside the
plume boundaries. Complete hydrodynamic containment can be
achieved only when limiting flowlines or groundwater divides are
maintained outside the plume boundary.
An idealized elliptical-shaped plume occupying part of a ground-
water flow field is shown in Fig. 1A) and how a discharging well
located at the downgradient end of the plume creates limiting flow-
lines around the plume is shown in Fig. IB. All contaminated water
within the plume is thereby induced to move toward the well.
Depending upon the hydraulic properties of the aquifer and the
dimensions and depths of a plume, more than one well may be
needed to achieve total containment.
The approach illustrated is feasible only when the plume is mov-
ing in essentially one direction in response to the local hydraulic
gradient. In more complex situations where, for example, a plume
originating from a source of contamination is spreading radially,
the need is to create a groundwater divide outside the plume
boundary by operating multiple recovery wells properly located in-
side the plume.
Factors Affecting Design and Cost
The interrelationships of the technical factors or elements that
can significantly affect the design and costs of a recovery-well
system are shown in Fig. 2. These factors are grouped, as indicated,
under three major categories pertaining: (1) to the plume itself,
(2) the aquifer in which the plume is moving, and (3) the engi-
neering aspects of the well system. The factors in the lowest tier of
the diagram are the ones to be quantified in order to make an over-
all evaluation of the design and costs.
The analysis is restricted to the major driving variables relating
to the dimensions and flux of the plume and to the hydraulic prop-
erties of the aquifer. All other factors along the lowest tier were
held constant except for well discharge which was computed with a
hydraulic model for each cost scenario.
Figure 1.
Control of plume movement with a single recovery well
136
-------
REMEDIAL RESPONSE
137
MODEL ASSUMPTIONS
The strategy illustrated in Fig. 1 is applicable to all aquifer and
plume conditions regardless of their complexity. For this study,
however, aquifers are assumed to be single layers of homogeneous
and isotropic porous media, and are considered to be infinite in
areal extent and underlain by impermeable rocks. No distinction
is made between confined and unconfined aquifers, and flow is
assumed to be essentially steady and horizontal.
Plumes are assumed to be single-phase homogeneous mixtures of
groundwater and leachate, with an elliptical shape in plan view.
In the vertical dimension, plumes are assumed to occupy the full
aquifer thickness. Contaminant concentrations are assumed to be
diluted in order to apply hydraulic models for fluids having a
density equal to that of water. Finally, contaminant transport
mechanisms that involve mechanical dispersion and chemical
diffusion are ignored.
As shown on Fig. 3, an optimum solution is determined with
successive approximations. Total discharge, Q, is initially calcu-
lated with Eq. 1,* and is assigned to a single hypothetical recov-
ery well.' The hypothetical well is located inside the downgrad-
ient limit of the plume at a distance, xo, from the stagnation point
downgradient of the well, as determined with Eq. 2 which is based
on the work of Forchheimer.2 A two-dimensional Cartesian co-
ordinate system is established with the recovery well at the origin.
The plume boundary can be mapped and the limiting flowlines
can be plotted with Eq. 3 from Forchheimer.2 Discharge from the
well is increased in small increments until the flowlines bound the
plume up to the widest point, beyond which the plume is narrower
and remains within the limiting flowlines.
•For all equations, see Fig. 3.
Driving Variables For This Paper
Figure 2.
Factors that affect hydrodynamic containment, design requirements and costs
CONTAINMENT STRATEGY AND HYDRAULIC MODELS
The containment strategy illustrated in Fig. 1 translates into an
algorithm that uses hydraulic models for determining the loca-
tion, number, and discharge of wells to contain a simple plume
within limiting flowlines. The algorithm is shown as a generalized
flowchart in Fig. 3 with a sequence of equations that simulate
conditions at various steps. Variables used in the equations are de-
fined in the Appendix to this paper. Data needed for the applica-
tion of the strategy include plume dimensions (width, length, and
general shape), the hydraulic gradient across the plume, and the
transmissivity of the contaminated aquifer.
A feasible solution to the containment strategy must satisfy
three constraints:
•The recovery system is located near enough to the downgrad-
ient plume boundary to reverse the hydraulic gradient at that
boundary
•Total discharge is large enough to create limiting flowlines that
bound the plume up to its widest part upgradient of the well
system
•Drawdowns of water levels resulting from withdrawals from the
system do not exceed limits that are a function of aquifer satur-
ated thickness
An "optimum" feasible solution is defined as the minimum
number of wells pumping the smallest discharge that meets the
above constraints. These data may be modified by engineering
safety factors to become the "design requirements" of the well
system.
Having calculated the total discharge for a single-well system, the
system discharge is apportioned among several wells in order to
accommodate a practical limit on water-level drawdowns in the
wells. In an array of wells, the drawdown will be greatest at the
most central well, owing to well interference. Eq. 5 is one of sev-
eral analytical solutions to the groundwater flow equation that
can be used to compute drawdowns.' Wells• are added on Fig. 1
along the y-axis of the coordinate system, in a line orthogonal to
the hydraulic gradient. As more wells are added, the calculated
discharge per well decreases and eventually the drawdown con-
straint is satisfied. In some cases, as in formations of very low
transmissivity, the number of hypothetical wells becomes so large
that the most practical design solution becomes a drain. When
more than one recovery well is required, the shape of the limiting
flowlines is altered slightly, however, this has a negligible effect
on conceptual designs and costs of the recovery system.
COST SENSITIVITIES
The sensitivity of recovery-well system costs to varying aquifer
and plume conditions is qualitatively determined with selected
cost scenarios. Non-complex conditions were assumed for the site,
the aquifer, and the contaminant plume. As mentioned previously,
other factors such as complex waste streams, conflicting property
boundaries, litigation requirements and public opposition are not
considered. These situations, however, can result in order-of-mag-
nitude increases in total containment costs. In some instances,
legal fees and public relations costs alone can be in the millions
of dollars. Thus, the ranges presented are selected to demonstrate
-------
138 REMEDIAL RESPONSE
FLOW CHART
CALCULATE PLUME DISCHARGE
LOCATE RECOVERY WELL
WHERE WELL Q = PLUME 0
EQUATIONS
1) O-TIW
2) x0»-0/(21TTI)
3) y«-x til
4) O-O+AO
2TTTI
-
CALCULATE DRAWDOWN
AT MOST CENTRAL WELL
6)
6) Q.
T) n*n
41TT
W(u)
Figure 3.
Flow chart and equations for basic containment strategy
technical cost sensitivities and not to indicate total costs of remed-
iation.
Cost Components
The principal capital (K) cost components of fluid recovery
systems include: plume delineation, system design, wells/drains,
additions to surface infrastructure, and water-treatment facilities.
Both capital and operation and maintenance (O&M) cost elem-
ents for each component are listed in Table 1.
Plume delineation includes determination of the areal and ver-
tical extent of the plume, and the rate and direction of local
groundwater flow. System design includes hydraulic modeling and
the specification of the number, location, and yeild of the wells.
Pump capacity, well design, and completion method are also spe-
cified. Additions to the surface infrastructure include access roads,
power transmission lines, and fluid handling (pipes) capabilities,
which must be provided and integrated with the water-treatment
facility. The recovered water is assumed to contain levels of less
than 100 ug/1 of a single organic contaminant such as TCE (Tri-
chloroethylene). This frequently occurring contaminant is assumed
to be treatable by filtration through activated carbon.
Associated with these K-cost components are annual operating
and maintenance costs for the well/pump system, the treatment
system, and to a lesser extent surface piping, power, and access
roads. Monitoring costs will also be incurred for verification of
plume containment through water-level measurements and analysis
of water samples.
Plume delineation, system design, well system installation, and
monitoring costs are based on several published sources,3'4'7'10
manufacturer's catalogues, quotes from distributors, and profess-
ional experience. Infrastructure costs were based on recent work
performed by SCS Engineers,1 whereas treatment costs are based
on data from Gumerman, Gulp, and Hansen.'
Table 1.
Overview of system components
Cost Component
(K,0(M)
Plume
Delineation (K)
System
Design (K)
Well
System (K)
Surface
Infrastruc-
ture (K)
Treatment
Facility (K)
Well
System (O&M)
Treatment
Facility (O&M)
Monitoring
Plume
Delineation (K)
System
Design (K)
Well
System (K)
Surface
Infrastruc-
ture (K)
Treatment
Facility (K)
Well
System (O&M)
Treatment
Facility (O&M)
Monitoring
Cost Elements
Soil borings
Monitor wells
Data analysis
Laboratory analysis
Reporting
Consulting fees
Computer time
Construction engineering
Well construction
Materials
Access roads
Power transmission
Piping
Construction engineering
Treatment system
Pump operation (power)
System maintenance
Chemicals
Labor
Power
Sampling
Analysis
Reporting
Soil borings
Monitor wells
Data analysis
Laboratory analysis
Reporting
Consulting fees
Computer time
Construction engineering
Well construction
Materials
Access roads
Power transmission
Piping
Construction engineering
Treatment system
Pump operation (power)
System maintenance
Chemicals
Labor
Power
Sampling
Analysis
Reporting
Principal Factors
Affecting Cost
Plume area & depth
Complexity of hydrogeology
Complexity of hydrogeology
Size of containment system
Plume size
Aquifer flux
Trans missivity
Plume size
Configuration of wells
System discharge
Composition/ concentration of recovered water
Aquifer flux/pumping depth
Fluid corrosivity
System discharge
Composition/concentration of recovered water
Complexity of system
Plume area & depth
Complexity of hydrogeology
Complexity of hydrogeology
Size of containment system
Plume size
Aquifer flux
Transmissivity
Plume size
Configuration of wells
System discharge
Composition/ concentration of recovered water
Aquifer flux/pumping depth
Fluid corrosivity
System discharge
Composition/concentration of recovered water
Complexity of system
Cost Scenarios
To show the relative importance of well-system costs to other
cost elements of the system, eight hypothetical scenarios have
been evaluated. In these scenarios, plumes are assumed to be ellip-
tical in plan view with the following range of dimensions:
•250 to 2500 ft wide
•500 to 5000 ft long
•25 to 250 ft deep
Plumes are assumed to be moving in unidirectional flow fields
representing low to high aquifer flux (0.05 to 1.0 mgd/mile). A
relatively small transmissivity (5,000 gal/day/ft) is assigned to four
high-flux scenarios and a large transmissivity (100,000 gal/day/ft)
is assigned to four low-flux scenarios. Outputs of the analysis con-
cerning numbers of wells and pumping rates are summarized in
Table 2 while the costs of each element in the analysis are given in
TableS.
Discussion of Sensitivities
How the cost components are effected by order-of-magnitude
changes in plume size and aquifer flux is shown in Table 3. Plume
delineation costs tend to be the largest cost component for all
scenarios, but are only influenced by size variables. Delineation
costs and both K- and O&M-costs of wells are moderately sensitive
to depth. System discharge is directly proportional to plume flux
and hence increases directly with plume width as indicated by Eq.
1. Although the data in Table 2 indicate that system discharge
varies by a factor of 200 (from 2 to 400 gal/min), corresponding
-------
REMEDIAL RESPONSE
139
Table 2.
Summary of design parameters
Flux = 0.05 mgd/mlle width
Transmissivity = 100,000 gal/day/f I
250x500x25
250x500x250
2500x5000x25
2400x5000x250
Flux = 1.0 mgd/mile width
Transmissivity = 5,000 gal/day/ft
250x500x25
250x500x250
2500x5000x25
2500x5000x250
Number
of Wells
2
2
2
2
Well Discharge
Discharge
(gal/min)
2
2
20
20
40
40
4oa
400
aDrawdown constraint forces use of drain instead of well.
total K- and O&M-costs vary by factors of less than 5. The plume
delineation costs and the assumption of a non-complex waste
stream for treatment together prevent system discharge from hav-
ing a more significant influence on total costs.
Capital costs of well systems range from about 7 to 30% of the
total capital costs, while corresponding O&M-costs for wells range
from 37 to 57% of the total O&M-costs. Although well system
costs are relatively small, discharge from the system influences in-
frastructure and treatment costs. In a few scenarios these two
elements may become significant and can account for up to 40%
of K-costs and up to 67 % of O&M-costs.
The principal design parameters for each plume in the low-flux
case indicate minimal variability in the number of wells and the
discharge needed to contain both the smaller and larger plumes.
Total capital costs vary from $180,000 to a high of $750,000.
Operating and maintenance costs vary from about $30,000 to
$35,000 per annum.
In the low-flux scenarios, delineation and modifications to sur-
face infrastructure represent the dominant difference in cost among
plume sizes. Changes in delineation costs are attributed to the
larger number and increased depth of monitor wells required
for delineation of a larger and deeper plume. Surface infrastruc-
ture costs increase in proportion to the surface area of the plume
and to system discharge.
The high-flux scenarios (Table 2) require a greater number of
wells to handle larger plume fluxes; this is the result of combin-
ing the higher flux with the lower transmissivity. Also, a hy-
draulic drain is required for the high flux scenario involving a large
plume in a relatively shallow aquifer, in order to meet the limit on
drawdown, which is set at 50% of saturated thickness. Due to
greater depth of aquifer, and hence greater available drawdown,
no additional wells are needed in the deeper large plume.
In the high-flux scenarios, the well system and treatment costs
become more significant compared with costs of delineation, due
to the larger volumes of recovered water. Annual operation and
maintenance costs are still minimal compared with the capital cost
components.
CONCLUSIONS
In this paper the authors have presented a method for determin-
ing the conceptual design and cost of a fluid recovery system
needed to provide hydrodynamic containment of a plume of con-
taminated ground water. The method utilizes existing hydraulic
models and is applicable to elliptical plumes moving in response
to unidirectional gradients within an aquifer.
Cost scenarios developed for this paper indicate that recovery well
systems represent a relatively small part of total capital costs but
as much as 57% of total O&M costs. Discharge from the recovery
system affects infrastructure and treatment costs and, in rounded
figures, can control as much as 60% of capital costs and 90% of
O&M costs. For more complex waste streams than that assumed
for the cost scenarios, system discharge will have an even greater
influence on costs.
After delineation costs have been incurred at a site, recovery sys-
Table 3.
Summary of Eight Recovery System Cost Scenarios
AQUIFER AND PLUME
CHARACTERISTICS
Low Flux, High Transmissivity
(plume width x length x depth, ft)
(250 x 500 x 25)
(250 x 500 x 250)
(2500 x 5000 x 25)
(2500 x 5000 x 250)
High Flux, Low Transmissivity
(250 x 500 x 25)
(250 x 500 x 250)
(2500 x 5000 x 25)
(2500 x 5000 x 250)
/
75 a
150
200
400
75
150
200
400
25-100
25-100
25-100
25-100
25-100
25-100
25-100
25-100
15
50
15
60
30
110
45
130
35
35
150
150
35
35
150
150
30
30
40
40
50
50
110
110
15
20
15
20
15
20
15
45
<5
<5
5
5
15
15
50
50
10
10
10
10
10
10
10
1U
180-255
290-365
430-505
675-75U
215-290
370-445
530-605
815-890
3U
35
30
35
40
45
75
105
a Cost in thousands of dollars
-------
140
REMEDIAL RESPONSE
tern discharge is the most critical design criterion needed for esti-
mating costs of remedial options that involve hydrodynamic con-
tainment of a plume. System discharge is related to aquifer trans-
missivity, hydraulic gradient, and plume width. A recovery sys-
tem must intercept the plume discharge; however, when wells are
located inside the plume boundaries, modeling of the limiting flow-
lines shows that recovery system discharge is generally larger than
plume discharge. The increase over plume discharge is controlled
by plume shape, and varies with the ratio of plume width to length.
Plume depth and changes in the aquifer saturated thickness were
treated as having no affect on discharge. Rather, modeling showed
that they greatly influence the depth, number, and spacings of re-
covery wells.
For the idealized scenarios tested, total capital costs range from
$180,000 to $890,000 and annual operating and maintenance costs
range from $30,000 to $105,000. Because of the common occur-
rence of complex hydrogeologic and/or contaminant-quality con-
ditions, it is felt that the costs represent the low end of costs typ-
ically encountered in the field. These parameters, plus factors
such as public and legal pressures, could easily increase total cap-
ital costs an order of magnitude higher than the basic costs quoted
in this paper. Continued cost estimation work for more complex
site conditions will result in an improved methodology for estimat-
ing costs for a wider variety of scenarios.
APPENDIX—DEFINITION OF VARIABLES
I —Hydraulic gradient across plume [dimensionless]
n —Number of recovery wells [dimensionless]
Q —Volumetric discharge through a [LVT]
part of an aquifer that encom-
passes the plume
AQ —Incremental change in discharge [LVT]
from the recovery system
Qw —Discharge of an average well in a [LVT]
multiple-well recovery system
S —Drawdown caused by pumpage from [L]
recovery well(s)
T —Aquifer transmissivity
W —Maximum width of plume
measured at right angles to
gradient
W(u)—Well function of u
x, y —Cartesian coordinates
XQ —Distance from stagnation point
to single recovery well creating
limiting flowline
REFERENCES
[LVT]
[L]
[dimensionless]
[L]
[L]
Victor
1. Darcy, H., "Les Fontaines Publiques de la Ville de Dijon,"
Dalmont, Paris, 1856.
2. Forchheimer, P., "Grundwassenbewegung," Hydraulik, 3rd Edition,
B.C. Teubner, Leipzig, 1930, 51-111.
3. Gibb, J.P., "Cost of Domestic Wells and Water Treatment in Illi-
nois," Illinois State Water Survey Circ., 104, 1971, 23 p.
4. Gibb, J.P. and Sanderson, E.W., "Cost of Municipal and Industrial
Wells in Illinois, 1964-1966," Illinois State Water Survey Cir 98
1969.
5. Gumerman, R.C., Culp, R.C., and Hansen, S.P., Estimating Water
Treatment Costs, Vol. 2 and 3, EPA-600/2-79-162b and c, USEPA,
Cincinnati, Oh., 1979.
6. Lindorff, D.E., and Cartwright, K., "Groundwater Contamina-
tion: Problems and Remedial Actions," Env. Geol. Notes, No. 81,
Illinois State Geol. Survey, Urbana, Illinois, 1977, 58 p.
7. Means, S., 1981 Means Guide to Construction Cost Estimating,
R.S. Godfrey, ed., Robert Snow Means Company, Inc., 19SI, 330p.
8. SCS Engineers, Written communication performed under Contract
68-01-5838 for the USEPA, Reston, Va., 1982.
9. Theis, C.V., "The Relation Between the Lowering of the Piezometric
Surface and the Rate and Duration of Discharge of a Well Using
Groundwater Storage," Trans. Amer. Geophys. Union. 2. 1935, 519-
524.
10. Anon., "The Water Well Industry: A Study," Water Well J., 35,
(1981)73-97.
-------
PLANNING SUPERFUND REMEDIAL ACTIONS
BRINT BIXLER
BILL HANSON
Hazardous Site Control Division, Office of Emergency and Remedial Response
U.S. Environmental Protection Agency
GILAH LANGNER
ICF Incorporated
Washington, D.C.
INTRODUCTION
The remedial action program is a key part of the Superfund
mandate to clean up uncontrolled hazardous waste sites across the
nation. The USEPA has developed a remedial planning process em-
bodied in the National Contingency Plan (NCP) that ensures rapid,
consistent, and rational decision-making on the appropriate extent
of remedy at priority hazardous waste sites. The process will be
applied to individual sites to ensure that projects focus on remedial
action in an effective manner. A flow diagram illustrating the
major steps in the remedial action process is shown in Fig. 1.
In this paper, the authors review five elements of the remedial
planning process: (1) the remedial action master plan, (2) remed-
ial investigations, (3) feasibility studies, (4) selection of a remedial
alternative, and (5) remedial design. For each of these activities,
the authors discuss both the technical considerations and the pro-
cedural requirements for conducting and submitting a complete
product. The emphasis in this paper is on remedial investigations
and feasibility studies, which form the basis for determining the
appropriate extent of remedy and the cost-effective remedial al-
ternative at an uncontrolled hazardous waste site.
REMEDIAL ACTION MASTER PLAN
A remedial action master plan (RAMP) is generally prepared for
sites that have been ranked as a priority on the national Priorities
List (or Interim List) and selected for remedial action. The RAMP
acts as both a general planning document and an effective site
management tool. It contains the information necessary for plan-
ning a coherent strategy and for assisting in the selection of an
appropriate course of action.
The preparation of the RAMP is the responsibility of the Reg-
ional EPA office working closely with the State. The RAMP con-
tains available site information, such as site inspection sampling
data, maps and topographical information, previous corrective ac-
tions, and available cost estimates, indicates where data gaps ex-
ist, and presents and justifies the scoping decision on which type or
types of remedial action should be initiated or studied.
The scoping decision is a preliminary determination of the gen-
eral category of remedial action needed, based on the complexity,
imminence, and extent of the hazard at the site. Three types of
remedial action are identified in the NCP: (1) initial remedial
measures, (2) source control remedial actions, and (3) off-site
remedial actions. The scoping decision selects which of these three
actions or combinations of actions may be appropriate for the site,
either for rapid implementation (initial, remedial measures) or for
further study and design (source control and/or off-site remedial
actions).
Initial Remedial Measures
Initial remedial measures may be appropriate when straight-
forward solutions are available for relatively simple problems.
These measures must be able to limit either actual or potential ex-
posure to a significant health or environmental problem. Examples
include construction of fences, stabilization of dikes or waste im-
poundments, temporary provision of alternative water supplies,
and removal of drums stored above ground.
Source Control Measures
Source control remedial actions may be appropriate if a sub-
stantial concentration of hazardous substances remains at or near
REMEDIAL ACTION PROCESS
Is
there a
Ihreal to public
health, welfare.
or the environ
ment?
Initial
Measures
7 ls \
/requeslX
lor lundinnXYes
No
Action
,approvedy
Cooperative
Agreement
or Slate
Contract
Reme
No
Action
dial investigation
Feasibility study
Figure 1.
Remedial Action Process
141
-------
142 REMEDIAL RESPONSE
Remedial Action Process (continued) Remedial Investigation/Feasibility Study Phases
Initial
Screening
Are alternatives
reasonable in cost:
»ety 10 substanlBJry
mitigate threat to put*
health, weave, or the
environment: technicaly
feasble: and not
fcely to mpact
health or
environment
adversely?
Is
Fund- \ Yes
Balancing
test met?
Remedial Invesbgation
FeasiWity Study
the area where they originally were located and the substances are
inadequately contained from migration into the environment. Ex-
amples of source control actions include installation of grout cur-
tains, trenches and drains, closure of surface impoundments, place-
ment of caps over contaminated areas, leachate collection and
treatment systems, and excavation and off-site disposal of contam-
inated soil or buried drums.
Off-Site Measures
Off-site remedial actions may be appropriate in some situations
to minimize and mitigate the migration of hazardous substance
and the effects of such migration. Off-site remedial actions may,
for example, include provision of permanent water supplies, con-
trol of a contaminated aquifer, treatment of a contaminated drink-
ing water source, dredging of contaminated river sediments, or re-
location of affected population.
If the scoping decision indicates that initial remedial measures
are called for, based on the criteria listed in the NCR, then a
"fast-track" implementation is possible. Planning for Fund-fi-
nanced initial remedial measures will generally not run longer than
six months; and in many cases can be much shorter. Planning will
include: (1) negotiation of a state cooperative agreement or con-
tract, (2) a remedial investigation focused on the initial measures,
(3) a limited analysis of alternatives which demonstrates the cost-
effectiveness of the measures proposed, and (4) a limited remedial
design to prepare contract documents needed for competitive
bidding.
In essence, the process for implementing initial remedial meas-
ures is a limited and expedited version of source control and off-
site remedial actions. Because of their limited nature, initial remed-
ial measures can be implemented while planning for source con-
trol or off-site actions is underway. However, the determination
must be- made that an initial remedial measure is cost-effective,
or will be a necessary part of any possible cost-effective source
control or off-site remedy. More detailed and extensive remedial in-
vestigations and feasibility studies are needed for most source con-
trol and off-site actions because of their increased complexity and
cost.
Following development of the RAMP, a request for funding an
initial remedial measure or a remedial investigation/feasibility
study for source control or off-site actions is submitted. Usually,
the remedial investigation and feasibility study are conducted as
one project, although they are discussed separately below. These
studies begin after the negotiation and signing of a state contract or
cooperative agreement between the State and USEPA, which out-
lines the responsibilities of the Federal and state governments with
respect to management of the remedial action at a particular site.
REMEDIAL INVESTIGATION
The remedial investigation is conducted to assess the problem at
a site and collect data necessary for its resolution. Investigation
activities must be carefully planned to obtain essential information
while minimizing costs. During the investigation phase, activities
are continually assessed to determine whether all the planned in-
vestigation activities are actually needed in light of new informa-
tion as it is obtained. The output resulting from the remedial in-
vestigation is a data base adequate to justify the need for site re-
medial action, and to support the selection and analysis of alterna-
tives in the feasibility study.
Scope
The following provides examples of the scope of each type of
remedial investigation that would be associated with the three types
of remedial action.
•Remedial investigations for initial remedial measures may in-
volve sampling wastes contained in drums to test for compatibil-
ity, as well as other limited sampling and monitoring efforts spe-
cifically needed to select or make effective use of initial remed-
ial measures.
•Remedial investigations for source control remedial actions may
involve more intensive sampling, surveys, and monitoring efforts,
focused at or near the area where the hazardous substances orig-
inally were located. Limited off-site investigations may be in-
cluded when necessary to document the migration of contam-
ination to other areas.
•Remedial investigations for off-site remedial actions should use
all existing information and may involve extensive data collec-
tion activities. These remedial investigations may be more exten-
sive in terms of the area covered and the extent of monitoring and
Sampling.
In special situations, a remedial investigation for off-site actions
may be combined with remedial investigations for initial remedial
measures or source control action. A combination of source con-
trol and off-site action investigations would be appropriate when
wastes have migrated off-site, but a significant'amount still re-
mains on site, or when available data are not adequate to deter-
mine whether either type of action, on its own, will provide the
solution to site problems. The scoping decision about the type of
remedial action needed may be revised as additional information is
gathered during the remedial investigation, and the remaining in-
vestigation and subsequent feasibility study may be revised to focus
-------
REMEDIAL RESPONSE
143
on source control or off-site actions, as appropriate.
Normally, however, remedial investigations for off-site action
are treated separately from other remedial investigations and fund-
ed only if there is substantial evidence of a threat to public health
from contamination that has migrated beyond the area where the
hazardous substances were originally located, and significant rea-
son to believe that Fund-financed action can achieve adequate pro-
tection of public health, welfare, and the environment.
Activities and Tasks
Typically, remedial investigations involve a sequence of activ-
ities and tasks, such as the following:
•Preliminary work to prepare for site investigations. This may in-
clude site visits, definition of boundary conditions, preparation
of a site map, and establishment of a site office if necessary.
•Site investigations. These may include, as appropriate waste char-
acterization, hydrogeologic investigations, soils and sediments in-
vestigations, surface water investigations, and air investigations.
These can include magnetometer, resistivity, and other remote
sensing activities.
•Identification of preliminary remedial technologies and categories
of remedies that may be appropriate.
Data Needed
A major purpose of a remedial investigation will be to provide
data to support the selection of a remedial action. An investiga-
tion for source control actions should provide answers to questions
such as the following:
•Should an impermeable barrier and clay cap be used to prevent
contamination of groundwater?
•Is incineration or reclamation a viable option?
•Is on-site treatment a viable option, and if so, what category of
treatment (e.g., biological, physical, chemical, thermal) should
be investigated?
•Will substances continue to migrate off-site if no action is taken?
A remedial investigation for off-site measures should address
questions such as:
•Does the volume of contaminated groundwater make treatment
impracticable?
•Are reliable technologies available to treat the identified contam-
inants at the site?
•Do technologies exist to effectively remove contaminated sedi-
ments from the site?
•Will the off-site contamination continue to pose a threat if no ac-
tion is taken?
•Will the action assure the future use of the affected resource (e.g.,
continued supply of drinking water from a threatened aquifer)?
The final report of the remedial investigation presents the data
collected and the results of the site investigations in relation to the
preliminary remedial technologies developed.
Additional requirements during both the remedial investigation
and feasibility study include the submission of periodic technical
progress reports and financial management reports by the contrac-
tors; assistance in development and implementation of community
relations plans, conduct and documentation of sampling and
analyses in accordance with chain-of-custody procedures, develop-
ment of a safety plan for personnel on site, and a quality assur-
ance/quality control plan for all sampling, analysis, and data
handling.
FEASIBILITY STUDY
The feasibility study is conducted for the purposes of develop-
ing and evaluating alternastives, recommending the appropriate
cost-effective remedial action, preparing an environmental assess-
ment, and developing a conceptual design for the recommended ac-
tion. The NCP decision process involves the development of al-
ternatives, initial screening, and detailed analysis of the remaining
alternatives. For some sites, this process can be accomplished ex-
peditiously. In other cases, the process will require many iterations
of alternatives development, screening, and refinement. Described
below is the basic sequence of tasks undertaken in the feasibility
study.
Development of Alternatives
This task involves the establishment of remedial response objec-
tives, the identification of appropriate remedial technologies, and
incorporation of objectives and technologies into site-specific
remedial alternatives. Alternatives should include non-cleanup op-
tions (e.g., alternative water supply, relocation) as well as a no-ac-
tion option.
Initial Screening of Alternatives
The alternatives are screened on the basis of the following con-
siderations: cost, effects on health and the environment, and engi-
neering feasibility. During this screening process, USEPA encour-
ages project review meetings, with the participation of USEPA and
state representatives, and the engineer conducting the feasibility
study, if appropriate. USEPA believes these working meetings will
be useful for a variety of purposes: to solve any problems that have
arisen in developing and screening alternatives; to ensure that the
alternatives under consideration are reasonably likely to represent
final alternatives; and to enable planning for necessary reviews by
USEPA and other federal agencies.
Detailed Analysis of Alternatives
This task involved a detailed development of remaining alterna-
tives, a cost analysis, an environmental assessment, and an evalua-
tion and recommendation of the cost-effective alternative. Alterna-
tives remaining after screening should be developed to include the
following factors, as appropriate:
•Description of appropriate treatment and disposal technologies
•Special engineering considerations required to implement the al-
ternative (e.g., pilot treatment facility)
•Environmental impacts and proposed methods for mitigating any
adverse effects
•Operation, maintenance, and monitoring requirements of the
completed remedy
•Off-site disposal needs and transportation plans
•Temporary storage requirements
•Safety requirements for remedial implementation
•A description of how the alternative could be phased into in-
dividual operable units
•A description of how the alternative could be segmented into
areas to allow implementation of differing phases of the alterna-
tive
Each alternative should be described using these factors or other
descriptive information needed to complete'the cost-effective eval-
uation described at the end of this section.
Analysis of Costs
Cost estimation is an essential part of the detailed analysis of al-
ternatives. In performing the cost analysis, both monetary costs of
the remedial alternatives and associated non-monetary costs should
be evaluated.
Monetary costs should be calculated in terms of the present
worth of the remedial alternatives over the planning period. Cost
estimates should be adequate to cover the effective and dependable
operation of the remedial measure during the remedial action plan-
ning period. This period is the lesser of: (1) the period of poten-
tial exposure to the contaminated materials in the absence of re-
medial action; or (2) 20 years. The same planning period is used
for each remedial action alternative considered.
Cost Estimation. Monetary costs of a remedial action alterna-
tive include direct and indirect capital costs and direct and indirect
operating costs. Direct capital costs might include acquisition of
land, right-of-ways, or easements; acquisition and installation of
facilities, structures, equipment, initial supplies, or other assets,
-------
144
REMEDIAL RESPONSE
and relocation costs. Indirect capital costs would include costs of
design engineering, field exploration and engineering services,
start-up costs such as operator training, overhead and profit, and
contingency allowances consistent with the cost estimates' level of
precision and detail. Direct operating costs include annual recur-
ring costs for operation and maintenance such as labor, materials,
utilities, and transportation and disposal. Indirect operating costs
include administrative overhead and profit on direct operating
costs.
The various components of costs are calculated on the basis of
market prices prevailing at the time of the cost analysis (i.e., the en-
tire analysis is performed in real dollars denominated in the year of
the analysis). Revenues generated through recovery of energy or
other resources, and accruing to the party or parties financing the
remedial action, are deducted from the costs of the remedial action.
The analysis should also include out-of-pocket costs to the local
community resulting from alternatives such as relocation or provis-
ion of alternate water supplies.
Based on available information at the time of the study, the cost
analysis should also clearly show estimates of costs to the Fund
over the planning period for each remedial action alternative.
Discount rate. The most recent rate mandated by the Office of
Management and Budget—currently 10%—should be used in the
feasibility study.
Salvage Value. Land purchased for remedial measures, includ-
ing land used as part of the treatment process or for ultimate dis-
posal or residues, where complete restoration of the land to orig-
inal conditions is not expected, may be assumed to have a salvage
value at the end of the planning period less than or equal to its pre-
vailing market value at the time of analysis. Estimated diminution
in market value must be substantiated and will be subject to ap-
approval by USEPA. Otherwise, in calculating the salvage value of
land, the land value should be appreciated at a compound rate of
3% annually (in real terms) over the planning period, unless the use
of a greater or lesser percentage can be justified based on historical
differences between local land cost escalation and construction cost
escalation. Right-of-way easements may be considered to have a
salvage value not greater than the prevailing market value at the
time of analysis.
Structures and other site improvements may have a salvage value
if there is an identified use for them at the end of the planning per-
iod. In this case, salvage value can be estimated using straight-line
depreciation during the useful life of the facilities. This method
may also be used to estimate salvage value at the end of the plan-
ning period for phased additions of process or auxiliary equip-
ment.
Documentation should be provided in the cost analysis if it is de-
termined that the useful life of any major component of the rem-
edy will be less than the planning period. When the anticipated
useful life of a facility is greater than the designated planning per-
iod for the remedial action, salvage value can be claimed for equip-
ment if it can be clearly demonstrated that a specific market or
reuse opportunity will exist.
Remedial action alternatives may have associated costs to public
health, the environment, or public welfare, such as urban and com-
munity effects or use of scarce resources and energy. When alterna-
tives appear to differ significantly in their associated non-mone-
tary costs, the specific cost elements should be presented and eval-
uated. Particularly important in this regard is the assessment of any
variations in public health and environmental costs among remedial
alternatives.
Analysis of Effects
Remedial alternatives in any category of remedial action con-
sidered should be analyzed for possible adverse environmental im-
pacts and for possible adverse effects on worker safety and health.
At a minimum, an analysis of effects should include an evaluation
of each alternative's environmental effects, an analysis of measures
to mitigate adverse effects, physical or legal constraints, and com-
pliance with CERCLA or other regulatory requirements. Addi-
tional analysis of effects of source control and off-site remedial
actions should be done, as described below.
Source Control Remedial Actions. The analysis of source con-
trol alternatives investigates whether, for the life of the remedial al-
ternative, the affected population is effectively protected from ex-
posure to hazardous substances that could threaten public health,
welfare, and the environment. This investigation is presented in the
form of a comparison between the situation currently existing (or
that might exist in the future if the action under consideration is
not taken), and the situation expected to occur following imple-
mentation of each alternative source control action.
Alternatives for source control action that are considered in a
detailed analysis should be reasonably expected to achieve the con-
trol and containment of the source of contamination. The analysis
must indicate the degree of containment expected from implemen-
tation of each of the alternatives. All methods of likely migration
of substances into the environment should be investigated to the
extent feasible. To the extent possible, the analysis should also con-
sider the fate of particular substances in the environment and the
suitability of the proposed remedial aternatives for control of the
substances.
Off-Site Remedial Actions. The analysis of remedial alternatives
must explicitly consider the extent to which each alternative miti-
gates or minimizes the threat of harm to public health, welfare,
and the environment. Alternatives which cannot be shown to effec-
tively reduce or eliminate the hazard to an acceptable level should
be rejected. If no technologies afe available to adequately mitigate
the threat (for example, to produce drinking water of an accep-
table quality), the plans for remedial action at a site may be recon-
sidered to determine if non-cleanup methods, such as provision of
alternative water supplies, is possible.
When analyzing off-site remedial alternatives, a comparison
should be made of the existing situation (and reasonable projec-
tions of what might be expected to occur without remedial action)
with the situation expected following implementation of off-site re-
medial alternatives.
Cost-Effective Evaluation of Alternatives
The evaluation of alternatives to determine cost effectiveness
should include the analysis of cost and effects described above.
Specific criteria should be developed in these and other appropriate
areas for use in evaluating each alternative. Typical criteria used
for evaluation of alternatives includes an assessment in each of the
following areas:
•Reliability of the remedial technology
•Flexibility during implementation of alternatives, including phas-
ing of alternatives into operable units and segmenting alternatives
into project areas on the site
•Institutional requirements to implement the alternative
•Operation and maintenance requirements
•Public acceptance
•Environmental effects and mitigation needs
•Safety requirements
•Cost of the remedial alternative
SELECTING THE REMEDIAL ALTERNATIVE
After completion of the feasibility study, a recommended plan is
transmitted from the State and USEPA Regional office to USEPA
headquarters for selection of the alternative that represents an
appropriate extent of remedy. Where appropriate, USEPA obtains
scientific expertise from other agencies that have a role in the re-
medial action program. The local community is also offered the
opportunity to review and comment on the recommended remedial
action.
The Assistant Administrator for USEPA's Office of Solid Waste
and Emergency Response considers the recommended plan, feas-
ibility study and other documentation to select the cost-effective
remedial action. An important part of this process is the require-
-------
ment of section 104(c)(4) of CERCLA to balance the need for
protection of public health, welfare, and the environment at the
site where remedial action is being considered and the availability
of Fund monies to respond to other sites. USEPA may choose to
exercise this "Fund-balancing" requirement in selecting the remedy
for a site.
Following selection of a remedial alternative, a conceptual de-
sign is prepared; it includes the engineering approach, implemen-
tation schedule, special implementation requirements, institu-
tional requirements, phasing and segmenting considerations, design
criteria, preliminary site and facility layouts, budget cost estimates
(including operation and maintenance costs), outline safety plans
and impacts on cost.
REMEDIAL DESIGN
Following completion of the feasibility study and approval of
the remedy by USEPA and the State, a remedial design will be
prepared. Funding will be provided through the same mechanisms
used in the remedial investigation and feasibility study phase (i.e.,
cooperative agreement or Superfund State contract).
Where USEPA has the lead, the final design and implementa-
tion activities will be managed by the U.S. Army Corps of Engi-
neers in accordance with the interagency agreement of Feb. 3,1982.
States assuming lead management roles will be responsible for
undertaking design and implementation activities.
Formal advertisement for contracts is the preferred method for
implementation of remedial actions. Therefore, the remedial de-
sign will usually result in a set of contract documents, including
REMEDIAL RESPONSE 145
plans and specifications, that describe the remedy in sufficient de-
tail to allow preparation of competitive bids.
CONCLUSIONS
USEPA has developed a rational decision making process for
selecting and implementing the appropriate extent of remedy at
priority hazardous waste sites. The process consists of five steps:
(1) a remedial action master plan, (2) remedial investigation, (3)
feasibility study, (4) selection of a remedial alternative, and (5)
remedial design. The effort required to develop each of these steps
will be evaluated individually for each site. In this way, projects
will address only the activities needed to focus on remedial ac-
tions. Three categories of remedial action have been established to
correspond to typical site conditions: (1) initial remedial measures,
(2) source control actions, and (3) off-site remedial actions. Use of
these categories facilitates quick action where needed and allows
USEPA to tailor remedies to particular sites.
Data collection efforts (remedial investigations) and feasibility
studies must be focused on specific remedial alternatives. Remed-
ial alternatives are analyzed for engineering feasibility, health and
environmental effects, and present worth costs (including opera-
tion and maintenance costs and salvage values, as appropriate).
The success of this remedial planning process depends on a close
working relationship between USEPA and the State, and on the
contributions made by local communities, contractors responsible
for conducting the studies, and other Federal agencies.
-------
ALTERNATIVES TO GROUND WATER PUMPING
FOR CONTROLLING HAZARDOUS WASTE LEACHATES
CHARLES KUFS
PAUL ROGOSHEWSKI
EDWARD REPA
JRB Associates
McLean, Virginia
NAOMI BARKLEY
U.S. Environmental Protection Agency
Municipal Environmental Research Laboratory
Cincinnati, Ohio
INTRODUCTION
Many techniques are available for controlling groundwater con-
tamination from hazardous waste management facilities including
pumping, removal (excavation), subsurface drains, low permeabili-
ty barriers, and in situ treatment. Groundwater pumping systems
are very commonly used because of their large range of applicabili-
ty and low installation costs.
Pumping systems can be designed to perform almost any func-
tion such as adjusting potentiometric surfaces, containing leach-
ate plumes to prevent further migration, and removing contam-
inated ground water. The chief technical advantages of pumping
systems are:
•Applicability—usable in confined and unconfined aquifers of
rock or unconsolidated materials under any conditions of homo-
geneity and isotropy
•Design flexibility—usable for injection as well as extraction re-
gardless of the depth of contamination
•Construction flexibility—able to be installed using a variety of
readily available materials by most qualified well drillers
•Operational flexibility—easy to repair most systems components
or modify the system or its operation as site conditions dictate
Although there are many advantages to using ground water
pumping techniques, some distinct disadvantages to their use exist.
The main technical disadvantages of pumping are that it requires:
•Extensive design data—on the site's hydrogeology and leachate
characteristics so that the system's components and operating
conditions can be designed or selected properly
•Effluent treatment—or some other means of managing well dis-
charges
•Continuous monitoring—to verify adequacy of system design and
function and to safeguard against component (e.g., pump) fail-
ure which can result in contaminant escape
•Permeable aquifers—otherwise the system cannot function effic-
iently or effectively
•Leachate compatibility—in terms of the leachate's ability to move
in ground water to the wells and not deteriorate well materials.
•Dilution—leachate may significantly dilute with groundwater
making it necessary to pump larger volumes of groundwater to re-
move contaminant.
Of these requirements, aquifer permeability and leachate com-
patibility are the most common reasons for seeking an alterna-
tive to ground water pumping.
Probably the single most significant drawback to using ground
water pumping is the high operation and maintenance (O&M)
costs typically associated with the system after installation. Many
times this is overlooked in remedial action planning especially when
capital (installation) costs are much higher for other options.
Minimizing long-term costs is particularly crucial to site remedia-
tion under Superfund, because states must pay all O&M costs.
Key O&M costs for pumping systems include: treating the con-
taminated discharge, providing electricity for the pumps, main-
taining and repairing system components, and monitoring system
performance. While other remedial actions also have these O&M
costs associated with them, pumping costs are frequently higher
because system operation is more demanding. For example, a sub-
surface drain may require the same amount of effluent treatment,
maintenance, and monitoring as a pumping system. However, elec-
trical requirements will be lower because the leachate is collected
by gravity flow. Low permeability barriers will have very low O&M
costs, especially if leachate treatment is not required. Consequent-
ly, these techniques may be more cost effective than pumping in
the long run even though their installation costs are higher.
In the remaining sections of this paper, the authors describe
alternate technologies for controlling ground water contamination
in cases where pumping is inappropriate. Also discussed are the
conditions under which an alternate technology might be a more
effective remedial action than pumping.
ALTERNATIVES TO GROUND WATER PUMPING
A number of alternative remedial techniques are available in the
event ground water pumping is deemed technically or economical-
ly impractical. An often selected option involves excavating the
source of contamination with subsequent disposal or treatment of
the excavated wastes on- or off-site.
Subsurface drains can be installed to form a continuous
hydraulic barrier in which contaminated water seeping into the
drain is collected and pumped to a treatment facility. A low
permeability barrier wall can be installed around the site to prevent
the flow of contaminated groundwater into or out of the surround-
ed area.
Finally, in situ treatment techniques may be applicable in some
cases where the waste can be neutralized, detoxified, flushed, or
treated directly while in-place. These four alternatives to ground
water pumping are described briefly below.
Removal and Treatment/Disposal
Many times the source of the pollution is removed. Contam-
inated earth materials may be excavated using conventional con-
struction equipment such as backhoes, draglines and front-end
loaders. The excavation of drums is more complex and usually in-
volves the use of specialized drum handling equipment such as
backhoe mounted drum grapplers. Following proper safety pro-
cedures is paramount in removal operations.
Excavated waste materials may be disposed of in an approved
chemical landfill, they may be incinerated on-or off-site, or they
may be treated using other techniques such as solidification or
encapsulation. Solidification techniques incorporate the waste ma-
terials in a chemical matrix such as cement, lime, or a polymer
so that the wastes cannot leach out of the matrix. Disposal costs
can then be reduced if the solidified waste can be considered a
non-hazardous material. Encapsulation techniques form an im-
146
-------
REMEDIAL RESPONSE
147
permeable polymeric capsule around the waste container, prevent-
ing leakage of waste out of the container.
Removal and disposal/treatment techniques are short-term ac-
tions that are applicable to many sites where the wastes are acces-
sible. Waste removal is ineffective, however, for contaminants
that have already leached from the site. Excavation of all contam-
inated substances (including groundwater and the strata through
which it flows) is rarely performed because of the high cost of treat-
ment or disposal of large volumes of material. Consequently, some
additional remedial measure is usually required to complete site
restoration.
Subsurface Drains
A subsurface drain consists of a narrow trench dug to a desig-
nated depth below the ground water table. A perforated drain-
age pipe is installed on the floor of the trench and the trench is
backfilled with gravel or crushed rock. This system forms a con-
tinuous permeable wall in which ground water seeps into the gravel
packing, enters the drainage pipe, and is carried by gravity flow to
a central collection point. The contaminated water is then pumped
from the collection area to a treatment facility. Subsurface drains
are most applicable to areas of low and/or varying permeabilities
and groundwater flow rates.
Subsurface drainage is often used in conjunction with low per-
meability barrier walls. It can be used inside of a downgradient bar-
rier wall to collect contaminated water prior to its reaching the wall.
It can be constructed within barriers that encircle sites to prevent
the "bathtub effect." It can also be used outside an upgradient
wall to retard seepage of clean water through the wall and into the
site.
The use of subsurface drains is similar to ground water pump-
ing in that the collected contaminated water must be treated and
disposed of properly. Subsurface drainage as a remedial approach
suffers from some of the same problems as pumping technologies
(i.e., long-term operation and maintenance costs). However, sub-
surface drainage can be used in low permeability strata where
ground water pumping would not form a continuous hydraulic
barrier. Depending on its placement, a subsurface drain may be
able to intercept a concentrated leachate prior to dilution with
groundwater and therefore avoid pumping large volumes of pol-
luted groundwater. However, handling the concentrated, aggres-
sive leachate may require the use of special materials. Operation
and maintenance requirements are typically lower with drains than
with wells because most of the flow in drains is caused by gravity.
Barriers
An underground, vertical barrier wall can be installed around a
polluting disposal site to impede or completely cut off groundwater
flow into and out of the site. The most effective application of bar-
riers is to completely encircle the site and key the bottom of the bar-
rier wall into an underlying impermeable formation.
There are a number of barrier wall materials that have been used
successfully. All form relatively impermeable structures when pro-
perly installed, and thus, cut-off any substantial flows through the
barrier. The most common technique is to use a clay slurry, which
forms a relatively low permeability barrier because of the swelling
of the clay particles in water. Slurry walls are presently being used
at many hazardous waste disposal sites to cut-off flow of con-
taminated groundwater plumes. A variation of the clay slurry wall
is to use an asphaltic emulsion to form the continuous barrier. Con-
crete is also used as a barrier material, particularly when a greater
degree of structural strength is required. The barrier wall system to
be installed at the Love Canal site will be constructed of concrete.
Where applicable, local materials can also be used to form low
permeability barrier walls. In California, San Francisco Bay muds
having hydraulic conductivities as low as 1 x 10 ~8 cm/sec have been
used to form barrier walls at waste disposal sites.
Sheet piling and grout injection have been used to form low
permeability barriers, however, their use is not common. Sheet pil-
ing is expensive and often cannot be used in strata containing
boulders where the possibility of pile deflection and subsequent
misalignment and gapping is high. Similarly, grout injection (in
which a wall is formed by injecting grout into adjacent boreholes to
produce interconnected soil/grout cylinders) frequently does not
form a continuous low permeability seal. Particular attention must
be paid to the compatibility of the barrier material and to the waste
being contained to avoid the material being rendered useless
because the waste causes a significant change in permeability
characteristics.
Barrier walls are often used with other techniques such as clay
capping and subsurface drainage to reduce infiltration and prevent
the accumulation of groundwater behind the barrier wall. While in-
stallation costs for barrier walls are much higher than for pumping
systems, O&M costs are generally much lower. However, if subsur-
face drainage or pumping is used in conjunction with the wall,
O&M costs will be significant.
In-Situ Technologies
In-situ techniques have been used on a limited basis in the past to
clean-up underground contamination caused by hazardous
material spills. They are generally short-term approaches which in-
clude biodegradation, soil flushing/solution mining, and
neutralization/detoxification. In certain situations, they have been
or could have been used as remedial techniques to treat contami-
nant plumes from waste disposal sites.
Biodegradation (or bioreclamation) is a microbiological treat-
ment technique originally developed to detoxify leakage from
underground gasoline storage tanks. In the bioreclamation techni-
que, contaminated groundwater is pumped to the surface where
nutrients and a bacteria (cultured to degrade the specific contami-
nant) are added. Then, the contaminated water is aerated and rein-
jected into the contaminant plume. Aerobic microbial degradation
of the contaminant plume can then take place underground.
Assuming the source of the contamination is removed, contami-
nant levels are gradually reduced to a residual level in the ground-
water.
Soil flushing or solution mining is a technique in which water, a
surfactant, or a solvent, is injected near an area of contamination
to wash hazardous chemicals from a contaminated soil. The con-
taminated solvent is then pumped from the subsurface and sent to a
treatment or disposal facility. If the solvent is treated to remove the
contaminants, it can then be reinjected into the area of contamina-
tion. Generally water is used so that new pollutants are not in-
troduced. This technique was used at the Goose Farm Site,
Plumsted Township, New Jersey, to flush contaminants with water
from contaminated soil beneath a drum disposal pit.
Neutralization/detoxification is an in-situ technique where a
chemical is added to contaminated water to neutralize the adverse
characteristics of the contaminant. In this technique alteration
rather than removal is the goal. Underground injection of a sodium
hypochlorite solution has been used to successfully treat cyanide
contaminated groundwater resulting from the indiscriminate dum-
ping of hazardous wastes. Dilute solutions of acids or bases can be
used for pH neutralization. Other potential neutralization/detoxi-
fication agents may include calcium salts (for phosphates),
hydrogen peroxide (chemical oxidation), and ferrous sulfate
(chemical reducing agent). Another in-situ treatment technique is
to inject catalysts into the subsurface that will cause organic con-
taminants to polymerize and become relatively immobile. This
technique has been used to control the contaminant plume from an
underground leak of acrylate monomors.
As with removal operations, in-situ treatment is generally con-
ducted on a one-time basis. Consequently, there are no long-term
O&M costs. The applicability of in-situ treatment is limited by the
contaminants involved and the treatment techniques currently
available, and thus, is not used widely. The cost of in-situ treatment
varies widely with site conditions and the technology to be used.
Selection of Alternatives
The conditions that prohibit the use of pumping for groundwater
restoration will, in part, determine which alternatives may be ap-
-------
148
REMEDIAL RESPONSE
Table 1.
Relative Effectiveness and Cost of Alternatives to Ground Water Pumping.
Alternative
Applied) ility
Relative Effectiveness
(1)
Relative Cost
(1)
Comments
REMOVAL
BARRIERS
SUBSURFACE
DRAINS
ItJ-SITU
TREATMENT
Low permeability
materials
Immobile or aggrea-
ive contaminants
• Eliminates source of
leachate but does not
affect off-site or
deep contamination
Higher in the short e Proper treatment and
term but may equalize disposal required
if long-term actions • Special safety pre-
would have been re-
quired
• Immobile or some
aggressive contami-
nants
• Where long-term
activities would be
required by a pumping
system
• As effective if not
more so provided bar-
rier is designed and
maintained properly
and barrier material
la resistant to the
contaminants
• Low permeability
materials
• Where long-term
activities would be
required by a pumping
system
Some immobile or
aggressive contami-
nants
Where long-term ac-
tivities would be re
quired by a pumping
system
• As effective if not
more so.provided drain
is designed and main-
ttined properly, and
drain materials are
resistant to waste ma-
terials are compatible
• Hay avoid dilution of
leachate with ground-
water
• Variable depending on
the nature of the con-
tamination and the
site's hydrogeology
Higher in the short
term but lower over
time
Could be much less
if little treatment
is needed
e Higher in the short
term but could equa-
lize over time de-
pending on pumping
and treating requiri
ments and duration
of activities
• Variable, but could
be less in some
Instances
cautions oust be taken
during excavation
Most appropriate for
small sites where a
high hazerd to
drinking water supplies
exists, where insoluble
wastes could not be
removed by pumping
alone, or where long
term treatment would
be too costly
Generally wall base oust
be keyed into an imper-
meable layer
Site capping and other
measures generally
required depending
on wall placement
e Proper treatment of
drainage required
• Not generally
applicable to conta-
- mination in deep bed-
rock
e Typically leas diffi-
cult to operate and
maintain than a pump-
ing system
• Applications somewhat
limited
e Little performance
data available
(I),
Relative to ground water pumping systems
plicable. When the contaminated layers are of very low permeabili-
ty, removal or subsurface drainage may be feasible. When the con-
taminants in the groundwater are either relatively immobile or will
aggressively attack well materials, removal, low permeability bar-
riers, and in-situ treatment may be alternatives.
In some cases, pumping will be technically feasible but not cost
effective because it may be required for a long duration and result
in pumping unreasonable volumes of water. Subsurface drains,
barriers, and in-situ treatment may be applicable in these instances.
The relative effectiveness and cost of these alternatives compared
to pumping are outlined in Table 1.
The appropriateness of the alternatives to groundwater pumping
that are available will depend on a variety of factors including:
•Site requirements (e.g., appropriate hydrogeologic setting)
•Areal extent of the contamination
•Nature of the contamination
•Capital and O&M costs
-------
REMEDIAL RESPONSE
149
Table 2.
Relative Comparison of Several Factors in Remedial Action Selection.
Factors In Selection
Remedial
Action
Pumping
Key Site
Requirements
Permeable earth
materials
Areal Extent of
Contamination
Any
Depth of Nature of
Contamination Contamination
Capital
Costs
Any Must be mobile Relative low
ground water and
not aggressively
deteriorate con-
struction materials
Long-term
O&H Costs
High because of
pumping and treat-
Ing requirements
Removal
Extent of conta-
mination limited
to easily remov-
able (I.e., un-
consolIdated)
earth materials
Relatively small Relative shallow
Practically any High
provided proper
safety precautions
are taken
None
Subsurface
Drains
None
Any
Limited only the Same as pumping
cost of excava»
ting bedrock and
by the capabilities
of the trenching
equipment
Moderate to Moderate to high
high because of pumping
and treating
requirements
Barriers
Generally Imper-
meable layer
required to key
wall Into
Any
Same as for sub- Must not sggres- Moderate to
surface drains slvely deterlor- high
rate wall materials
Relatively low
In-Sltu
Treatment
Same as pumping More applicable
to smaller sites
Limited by the
requirements of
the treatment
technique to be
used
Limited by the
applicability of
the treatment
technique to be
used
Variable depen-
ding on the
treatment tech-
nique to be usec
None
moved
ment
If source re-
before treat-
The limitations of the five technologies in terms of these factors are
found in Table 2.
There are many significant factors in remedial action selection.
To select the best remedial action scenario for a given site, it is
essential to identify and evaluate all factors for each of the viable
options. In the case of groundwater pumping, it is critical to assess
costs as well as technical factors. This is especially true of long-term
O&M costs, which may reduce pumping to an economically unat-
tractive alternative.
SUMMARY
Pumping is an effective and widely used technology for
remediating groundwater contamination. Nevertheless, in some
cases pumping is either not technically feasible or not cost effective.
A variety of alternatives to pumping are available for these cases.
However, the applicability of the alternative may also be limited.
Careful consideration must be given to all significant factors in
selecting the best remedial action.
-------
THE EXHUMATION PROGRAM FOR THE SCA
WILSONVILLE SITE
JOHN J. DINAPOLI
SCA Chemical Services, Inc.
Wilsonville, Illinois
INTRODUCTION
Once the agreement to exhumate wastes from Wilsonville was
signed, the task of putting this multifaceted project together be-
gan in earnest. What had happened in the past was no longer an
issue to be dealt with. The wastes were to be removed in the most
safe and expeditious manner.
ORGANIZATION
A project of this nature required the assembling of a team of in-
dividuals with various specialties who could work toward a com-
mon goal. Management positions were filled by personnel from
several SCA Chemical Services facilities who had experience in
Project Management, Landfill Operations, Safety Procedures,
Laboratory Operations, and Accounting Procedures. The various
staff positions were to be filled locally by equipment operators,
material handlers, laborers, lab technicians, and other support in-
dividuals. An organizational chart for the project is shown in
Fig. 1.
PRE-OPERATIONAL PREPARATIONS
Prior to the start of exhumation activities, a series of con-
struction and purchasing activities took place at the site. The pur-
pose of such operations was to transform a closed landfill which
had not been operated for four years into a full-fledged landfill
facility operating in the reverse mode. The preparations involved
included the following:
•Renovation of the existing office building to house the adminis-
trative staff
•Renovation of the existing equipment warehouse into a person-
nel building which housed the various employee support func-
tions such as:
-Shower facilities
-Lockers
-Safety equipment storage
-Lunch facilities
-Equipment repair
•Upgrading of the site roadways which had been used during
filling operations along with construction of new roadways to the
trench areas and staging areas
•Construction of staging areas for drums, trailers, and mobile
equipment
•Upgrading and expanding of the electrical, telephone, and water
distribution systems to provide services to the exhumation
trenches and the mobile laboratory
•Equipment Procurement—This consisted of the purchase of ap-
proximately 1.5 million dollars of operations, laboratory, safety,
administrative, and transportation equipment to be used through-
out the project (Table 1).
•Personnel Hiring—The hiring of all employees below the man-
ager level began approximately three months before operations
were scheduled to begin. Skilled labor was located fairly easily
due to the high rate of unemployment in the area and the prox-
imity to other chemical plants.
•Employee Training—The training of all operational and safety
personnel began approximately two weeks before operations be-
gan. The training programs undertaken were as follows:
Figure 1.
Wilsonville Site Exhumation Program, Organization Chart
Table 1.
Equipment List
1. Backhoe/Loader
2. Front End Loader
3. Supersucker Vacuum Truck
4. Drum Hoist Truck
5. Water Truck
6. Yard Horse Tractor
7. Fire Truck
8. Calgon Air Purification Unit
9. Canopy Structure
10. Staging Trailers
11. Fork Lift
12. Drum Crusher
13. Truck Scale
14. Pick-up Truck
15. Emergency Van
16. Compactors
17. Box Trailers
18. Tankers
19. Dump Trailers
-Equipment training
-Fire fighting training
-Use of safety gear
-Operations training
•Engineering consisted of the following:
-A base topographic map in aid in the restoration of the site to the
existing grades
-A metal detection survey to accurately locate the perimeter
boundaries of each trench
-A hydrogeological study to aid in the development of pro-
cedures to deal with groundwater movement during operations
and to designate areas of soil borrow for lining and capping of
the exhumed trench
•Monitoring Network—Including the installation of six sampling
stations for measurement of particulates and vapor emissions
caused by the exhumation operation. This consisted of the place-
ment at five locations of sampling pumps and carbon tubes for
vapor absorption and Hi-volume air samples for particulate
measurement. A series of groundwater monitoring wells and sur-
face water monitoring points had previously been established both
during and after original landfill operations.
150
-------
REMEDIAL RESPONSE
151
OPERATIONS
Removal Operations
Equipment placement—Before beginning any of the removal
operations, all equipment must be placed at area to be excavated.
Excavations will begin at the end of each trench opposite the entry
road to the trench. The extent of the area to be excavated was
determined by the markers left during the initial metal detection
survey and as verified by use of the onsite metal detector. Once
the end line of drums has been determined, the canopy structure is
moved into place. The canopy was located to overlap the area to
be excavated. Upon positioning of the canopy, all required elec-
trical connections for equipment attached to the structure were
made. Once the canopy was in place, all mobile operating equip-
ment was placed along the excavation boundaries as shown in
Fig. 2.
Cap removal and excavation—Before proceeding with the cap
removal, the perimeter of the excavation area must be protected
from surface runoff with the construction of a small runoff con-
trol berm. Clay from the cap should be used for this exercise.
The cap will then be removed to expose the uppermost lift of drums
over the area to be worked and always under the canopy. The
clay cover will be removed utilizing a rubber tired front end loader.
Soil material removed is segregated as follows: (1) Topsoil for later
use for vegetative growth, (2) Clay for lining and capping the
trenches, and (3) Miscellaneous soils, mine spoils, tree trunks, etc.,
to be placed between the upper and lower clay layer
As the cap is being removed, a hand auger is used to probe the
remaining depth to the top drum lift. This probing will assist the
operator in the depth of cut to be taken to eliminate the potential
for puncturing drums. The site metal detector will also be used to
ascertain relative depths to metal drums.
The soil removed will require testing in accordance with the
sampling and analysis section of this paper. If it is determined
to be hazardous, it must be loaded into bulk solid trailers for ship-
ment off site. During the cap removal process, a vertical excava-
tion shall be made just forward of the drum location to provide a
bench area from which to begin removal operations. The depth of
this section shall be at least equal to the base of the uppermost
drum lift. The excavation shall be made as close to the first lift of
drums as possible without contacting drums with the backhoe.
Drum removal—Before drum removal can begin, it is necessary
to remove small quantities of soil from around the surface of the
drums. This should be accomplished with the use of non-spark-
ing hand tools or a soil vacuum such as the Supersucker.
Once this soil has been removed, the drum should be visually in-
spected for deterioration. If the drum is leaking or is not in a physi-
cal condition to be lifted out of the excavation, the drum should re-
Figure 2.
Trench 24
Typical Operation Flow Diagram
main in the excavation until the liquid contents are either pumped
from the drum or the drum is overpacked.
If the drum is in a condition which would allow for hoisting from
the trench, the drum should be rolled out of its location to a point
where it can be fitted with the sling from the hoist and removed
from the trench. Once removed, it should be placed on the staging
trailer.
If the drum contains liquid and cannot be hoisted from the
trench, the contents should be pumped from the drum using drum
pumps. This material should be pumped to a clean drum.
If the drum contains solids and cannot be lifted from the trench,
the drum should be overpacked and hoisted from the trench. If
the contents of the drum have leaked to the surrounding soil, the
liquid should be vacuumed with the vacuum truck. The soil which
has subsequently become hazardous should be either vacuumed or
shoveled from the excavation and placed in the bulk solids trailer
for removal off site.
As the operation proceeds from drum to drum the soil between
drums should be mechanically shoveled from the trench to the daily
soil staging area for testing. Drums which have leaked their con-
tents and are now empty should be removed from the excavation
and sent to the drum crusher.
Contaminated Soil and Sludge Removal
Soil which is excavated from the trenches can be classified as
hazardous or non-hazardous depending upon the degree of chem-
ical contamination. All soil which is removed from between drums,
from the sidewalk or from the base of the excavation during a
day's operation should be placed in the soil staging area adjacent
to the excavation. This area should be bermed to prevent infiltra-
tion of surface water runoff. During rainstorms, the soil should be
covered with a PVC liner.
At the end of each workday, the laboratory will sample the pile
for analysis and classify the material as hazardous or non-hazard-
ous. If the material is classified as hazardous, it should be loaded
into a bulk solids dump trailer for disposal off site. If the labora-
tory classifies the material as non-hazardous, it can be used as
backfill for the trenches. If this soil is not immediately usable, it
should be stockpiled with other similar materials in the three stock-
pile classifications previously mentioned.
Sludge and bulk solids which were originally disposed as waste
material must be disposed off site. This material, when encounter-
ed, shall be removed by a backhoe and placed in a bulk solids
dump trailer. If the viscosity of the material is low, it may be
vacuumed with the Supersucker or placed in fiber pack drums if
found to be incinerable.
Free liquid removal—The presence of leachate in the trenches
is highly likely and will present numerous operational difficulties.
When encountered, this liquid should be tested and vacuumed with
the Supersucker or pumped directly to a bulk liquids tanker parked
at the side of the trench. If vacuumed with the Supersucker, the
liquid will require transfer to bulk liquid tankers.
Backfill Operations
Clay baseliner—After removal of all waste has taken place in a
minimum of 30 ft of the trench and the walls and base have been
determined to be nonhazardous, the lining of the base of the ex-
cavation should take place. Inspection of the base and sidewalls
should take place to detect the presence of sand lenses.
Since this strata represents a possible path of migration, it should
be sealed with clay during the baseliner operation. Clay from the
nonhazardous clay stockpile should be loaded into the' trench util-
izing the front end loader. The clay should be placed in 12-24 in.
lifts and compacted using a plate compactor. Three feet of clay
shall be placed along the base of the excavation in this manner.
Where necessary, the clay liner shall be dished up the sidewalls to
intersect sand layers.
Lime neutralization liner—The volume between the 3 ft base-
liner and the 4 ft cap will be filled with mine spoils, clay soils,
-------
152
REMEDIAL RESPONSE
and soils other than clay. The amount of this material required
will vary from trench to trench, depending upon the volume of
material (drums, soil, sludges) which are removed for off site dis-
posal.
Before placing mine spoils into the trench, the Laboratory Man-
ager shall be consulted to determine the quantity of lime to be
placed below the mine spoils to neutralize the effect of any leach-
ing. This lime should be spread uniformly on the base.
Mine spoils—The gob pile should be utilized to fill the remainder
of the excavation to an elevation just below the base of the cap.
The mine spoils should be loaded into the on-site dump truck with
the front-end loader. The dump truck will then transport the ma-
terial to the trench for unloading into the excavation. The loader
or plate compactor shall be utilized to compact this layer before
placement of the top cap.
Clay cap—After the mine spoils have ben placed into the trench,
the installation of the clay cap should begin. Construction of this
layer will consist of 12-24 in. lifts of clay which should be com-
pacted with the front end loader or plate compactor. The total
thickness of this cap will be 4 ft upon completion.
The top elevation of the cap will match the existing elevation
of the trench before exhumation was begun. The baseline topo-
graphic map should be utilized to determine appropriate eleva-
tions. The cap should be crowned on a 5 % slope to enhance rapid
runoff of precipitation. Only after the cap has been compacted
should the canopy be moved to its next position along the trench.
For a depiction of the final trench cross section, see Fig. 3.
Top soil and seeding—Upon completion of the exhumation and
backfill for an entire trench, the area of the trench shall be cov-
ered with 6 in. of topsoil previously stockpiled and seeded with
perennial ryegrass. Seeding shall be mechanical.
Staging Operations
Low bed staging—During exhumation of a trench, two modified
low bed trailers will be used to stage drums until the laboratory
has completed its analysis and determined the appropriate bulk-
ing requirements or treatment/disposal option. The trailers will be
placed along the sides of the excavation and receive drums which
are hoisted from the trench.
Drummed liquids which are removed from the trenches should
be placed on a staging trailer and affixed with a consecutively num-
bered drum tag. The material handlers will then open the drums to
allow sample taking. Drummed solids which are removed from the
trenches should be placed on the staging trailer, affixed with a
drum label, and opened for sampling.
F1NAL_6RADE_ .SLOPE
!' COMPACTED CLAY CAP
MINE SPOILS
IMF
3' COMPACTED CLAY
(NOT TO SCALE)
Figure 3.
Trench 24
Final Cross Section
Bulk soil/sludge staging—During exhumation of drums, sub-
stantial quantities of soil and sludges contained between and
around the drums will be excavated from the trench. The exca-
vated soil will be staged in an area alongside the trench. This area
will be bermed and lined with twenty mil PVC. The size of the
area should be sufficient to store at least three days of excavated
soils. During rainfall events, this area should be covered with
twenty mil PVC liner to prevent the infiltration of water.
Sludges and bulk solid wastes which were originally landfilled
should be excavated from the trench and placed directly into the
bulk solids dump trailer. Soils which were backfilled during oper-
ations should be removed from the trenches and placed in a pile
with that day's excavated soil in the soil staging area.
At the end of each day, a sample will be taken by the labor-
atory for analysis. If the soil is determined to be hazardous, it
will be placed into the bulk solids dump trailer for disposal off-
site or the pile will be segregated at the direction of the Labora-
tory Manager and retested.
Soils which are determined non-hazardous will be utilized as
backfill to the trenches. This soil will remain in the staging area
until utilized as backfill or stockpiled further away from the exca-
vation.
Bulk liquids staging
Once a drum of liquid has been identified as being compatible
with the contents of a bulk tanker, the contents shall be pumped
directly into the tanker for shipment off site. A drum pump shall
be inserted into the bung hole and the contents pumped complete-
ly into the tanker. The empty drum should be removed from the
staging trailer and transferred to the drum crusher. This process
will continue until the tanker has been filled at which time it will
be prepared for shipment offsite.
When liquid leachate is encountered in the trenches, the liquid
shall be vacuumed with the Supersucker vacuum truck or pumped
directly into the bulk tanker designated for aqueous waste. To
minimize handling, it is preferable to pump directly from the trench
to the tanker.
Liquid which is being pumped from Well G-107 as part of the
effort to control leachate migration should be pumped directly in-
to the appropriate liquid tanker.
Drummed liquids staging—Drums of liquids which cannot be
immediately bulked into tankers must be routed as follows:
•If the drum can be bulked with future drums which are iden-
tified on the grid sheets, it should be placed into the drum storage
building provided the drum will be bulked within 90 days of its re-
moval.
•If the drum cannot be bulked with any future drums, it should be
loaded onto the appropriate box trailer for a disposal/treatment
offsite.
Drummed solids staging—Drums of solid material should be
loaded onto box trailers depending upon the disposal facility
chosen.
Empty drum staging—Drums which have been crushed after
liquid removal should be placed into the staged bulk solids trailer
for disposal as bulk.
Drum Crushing Procedures
As drums are emptied of either solids, sludges, or liquids, they
are to be stored on a dump truck until a full load is achieved.
Drums are then hauled to the drum crusher.
At the drum crusher station, the drums are to be checked for the
presence of minor quantities of liquids. If liquids are encountered)
the drum should be emptied into a container and transferred to
the bulk liquids trailer which was used to receive the majority of
liquid in the drum. If greater than 4 in. of sludge are present at the
base of the drum, this material shall be shoveled and transferred
to the bulk solids trailer. Drums will then be crushed, two at a
time, and the crushed drums reloaded onto the same dump truck
-------
REMEDIAL RESPONSE
153
for transfer to the bulk solids trailer.
SAMPLING AND ANALYSIS
Sampling
Once the drums are placed on the staging trailers, they should be
opened and sampled with either a sampling thief for liquids or a
coring tool for solids: The samples will be placed in bottles and
transported to the laboratory for analysis.
Before transport of any bulk liquid or solid materials, the lab-
oratory will receive a sample for analysis and designation of the
appropriate handling method.
Analysis
Analysis of all samples will take place in the site laboratory which
is housed in a 70 ft x 12 ft trailer. The laboratory contains the
following major equipment for analysis:
•Gas Chromatograph/Mass Spectrometer
•Atomic Absorption Unit
•Karl Fisher Titrator
•Calorimeter
•TOC/TC/TIC Analyzer
•pH/Ion Meters
•Distilled Water
•Centrifuge
•G. C. FID Unit
•Viscometer
The laboratory analysis will direct the operational crews as to the
destination of all materials exhumed from the trenches. It will also
determine the appropriate treatment/disposal option for each
waste. In addition to the sampling and analysis of each drum of
waste, the laboratory is responsible for analysis of vapor and par-
ticulate samples taken at the site perimeter.
At the end of each day, soil which was excavated from the
trenches will be sampled by the laboratory for classification as
hazardous or non-hazardous. The comparison of the soil analysis
will be made with the USEPA Multimedia Environmental Goals
For Land Ecological Criteria.
SELECTION OF TREATMENT/DISPOSAL OPTIONS
After sampling and analysis, but prior to any repackaging or
bulking (as appropriate), it will be necessary to determine the most
appropriate method of treatment and/or disposal of the exhumed
wastes. In making this determination, SCA will consider:
•Protection of human health and the environment
•The availability and appropriateness of permitted treatment/dis-
posal facilities
•Costs
SCA will evaluate the options available by waste type of deter-
mine the advantages and disadvantages of each. SCA will then
select the preferred option and make appropriate arrangements.
The treatment/disposal techniques which are, at this time, the
most appropriate include:
•Incineration—Controlled combustion to thermally destroy waste
materials, converting the waste into harmless gases and inert
solids. Parameters used to evaluate a waste type for this tech-
nique include: BTU value, organic chlorine, organic sulfur, heavy
metals, total solids, ash, etc.
•Aqueous Treatment/Detoxification—Includes oxidation, reduc-
tion, coagulation, distillation, precipitation, and neutralization.
Parameters used to evaluate a waste type for this technique in-
clude: water content, flash point, pH, acidity, alkalinity, TOC,
cyanide, sulfide, metals, etc.
•Secured landfill—For wastes that can neither be chemically
treated nor thermally destroyed. Parameters used to evaluate a
waste type for this technique are similar to those listed above.
LOADING AND TRANSPORTATION
Treatment/Disposal Facility Arrangements
Upon completion of the laboratory analysis and a designation by
the laboratory of the appropriate disposal facility, the Operations
Manager will direct the loading of drums onto the box trailer.
Since most of the drums shipped off site will be solids destined for
a secure landfill drums can be loaded onto the box trailer as soon
as they have been analyzed, unless they contain solid materials
which cannot be disposed of at a particular secure landfill. The
drums will be loaded onto the trailers from the low beds with the
use of the fork lift and hand operated two-wheelers. All box trail-
ers should contain a layer of absorbent on their base to absorb any
leaking drums.
Upon filling of a bulk liquids tanker, a composite analysis by the
laboratory will be undertaken. Upon completion of the analysis,
the laboratory will designate the appropriate treatment facility to
receive the waste.
Upon filling of a bulk solids dump trailer, the laboratory will
sample, analyze, and designate the appropriate disposal facility to
receive the waste.
Transportation Contractor Arrangements
Upon completion of loading of either a box, dump, or liquid
trailer and upon designation of the disposal/treatment facility,
the following steps will be undertaken:
•The Laboratory Manager will notify the Transportation Dispatch-
er of the destination of the waste containing trailer. He will also
forward the appropriate manifest documents to the Dispatcher
with the laboratory data filled in.
•The Dispatcher will call the transportation contractor to schedule
a pickup of the load and will complete the remaining informa-
tion on the manifest documents.
Manifest Preparations
The Dispatcher will finalize all information on the manifest doc-
uments for transfer to the truck driver upon pickup.
Vehicle Inspection and Preparation
Before a trailer is allowed to leave the exhumation area, the
Operations Supervisor must visually inspect the entire vehicle in-
cluding but not limited to the following items:
•Correct liner installation (if installed)
•Secured cover tarpaulin
•Locked lift gate
•Proper placarding
•Proper tractor to trailer hitch
•Cleanliness
•Excess waste levels
•Tire conditions
Any corrections should be made in the exhumation area.
Before a bulk liquids tanker is allowed to leave the exhuma-
tion area, the Operations Supervisor must visually inspect the entire
vehicle including but not limited to the following items:
•Closed valve positions
•Secured hatches
•Excess liquid levels
•Proper placarding
•Proper tractor to trailer hitch
•Cleanliness
•Proper venting
•Tire conditions
Any corrections should be made in the exhumation area.
Before a box trailer is allowed to leave the exhumation area, the
Operations Supervisor must visually inspect the entire vehicle in-
cluding but not limited to the following:
•Correct layer of absorbent
-------
154
REMEDIAL RESPONSE
•Quantities of bagged absorbent
•Leaking drums
•Proper packing inside van
•Locked rear doors
•Proper trailer to tractor hitch
•Proper placarding
•Cleanliness
•Condition of tires
Any corrections should be made in the exhumation area.
Loaded vehicles will then be hauled to the vehicle wash area,
using the on-site yard horse. Cleaning of vehicles should be accom-
plished by scraping, or brushing where possible, to minimize the
generation of contaminated wash water. When absolutely neces-
sary, water washing should be accomplished using a high pressure
low volume nozzle.
The yard horse should bring the loaded trailer to the Zone I area
where the transfer of trailers will be accomplished. The transpor-
tation contractor will attach to the loaded trailer and the yard horse
will attach to the empty trailer for placement in the exhumation
staging area.
Upon entering the facility, all trailers and tractors should be
weighed empty for comparison with loaded weights. After trailer
transfer, the transportation contractor shall bring his loaded ve-
hicle to the truck scale for weighing. The Dispatcher will determine
the shipment weight which will be recorded on the manifest.
Once a vehicle is prepared to leave the site, the Operations Man-
ager will call the designated disposal facility for notification of
shipment and estimated time of arrival. All drivers will be given a
specified route to travel with instructions not to vary from this
route. They also will be given a list of emergency numbers and a
copy of the site contingency plan.
RECORDKEEPING AND DOCUMENTATION
Once a drum is removed from the excavation, it should be
affixed with a drum tag containing a preprinted number. The pur-
pose of this tag is to provide a unique drum identification num-
ber to each drum which is exhumed.
The drum/bulk data form is the principal recordkeeping tool to
be used during exhumation of drums and bulk materials from the
landfill. A copy of the form is shown in Fig. 4.
All waste which is transported off site must be accompanied by
at least one manifest depending on the state of treatment/disposal
of the shipped materials. At a minimum, all waste must have a
completed Illinois manifest.
MEDICAL MONITORING AND SURVEILLANCE
Physical Examinations are required of all individuals before
commencing work at the site. The preliminary examination will be
used as a baseline to monitor any changes in an individual's health
during the exhumation period. It will be compared with the suc-
ceeding yearly physicals given to all employees. In addition to the
yearly examination, all employees working in the exhumation area
will be tested quarterly for blood, liver and urine functions.
SAFETY PROGRAM
A comprehensive safety program has been developed to pro-
tect workers from the potential effects of materials buried in the
trenches. The equipment utilized is as follows:
•Full protective suits
•Air line respirators fed from tanks filled by a breathing air com-
pressor
•Emergency pressure demand air systems
•Protective shoes, gloves and hats
All employees which enter Zone II (the trench area) are re-
quired to be suited to various degrees as determined by the Safety
Manager.
A fire truck is parked at the excavation site at all times. The
unit is equipped to fight fires with water, foam, or dry chemicals.
Portable fire extinguishers are also placed around the trench and on
all operating equipment.
Continuous monitoring is undertaken at the trench for the
following:
•Total organic vapors
•Mercury
•Explosion potential
SPILL PREVENTION AND CONTINGENCY PLANS
A spill prevention and countermeasure plan has been devel-
oped for the site personnel and truck drivers transporting waste,
All on-site personnel have been trained in this area.
A contingency plan has been developed in conjunction with the
local county emergency services officer regarding emergencies'
which could affect residents of Wilsonville and surrounding towns.
Arrangements have been made with area hospitals in the event of
an employee requiring treatment for injuries incurred on-site. A
full time nurse and an emergency van are on standby in the event
of an emergency.
SECURITY
The control of access by individuals to various sections of the
site represents one of the most important operational considera-
tions.
DRUM/BULK DATA FORM
Sampling Date Sampled:
DrumID#: Time:
Estimated Liquid Quantity:
Grid Location:
Staging Trailer Number:.
Sampler's Name:
Drum Condition:
Physical Appearance of the Drum/Bulk
Contents:
Odor:
Color:
pH:
_%Liquid:_
Laboratory
Analytical Data:.
Date of Analysis:.
Compatibility:
Hazard:
Waste ID:
T/D:
R/D Referral:.
Approval
Lab:
Site Manager:.
Date:
Date:
Figure 4.
Drum/Bulk Data Form
-------
REMEDIAL RESPONSE
155
The entire site is divided into two major areas for purposes of
access control. The areas are designated as Zones I and II and are
shown in Fig. 5.
Zone I will include the administration building and associated
parking area along with the guard house. Activities to be under-
taken in this Zone are purely administrational. No handling or
storage of waste materials of any type should be undertaken in this
Zone. Only the transport of waste materials in containers should
occur in this Zone. Activities undertaken in this area are as
follows:
•Sign in/sign out of all individuals entering and leaving the site
by the security officer in the guard house
•Office support activities
•Visitor and employee parking
Zone II will include the remainder of the 130 acre site. In-
cluded in this area will be the drum storage building, maintenance
building, the 25 trenches to be exhumed and the mine spoils pile.
Entry to this area will be via a gate segregating Zone I from Zone
II. Activities to be undertaken in this area are as follows:
•Receipt of material and supplies
•Exhumation of all trenches
•Vehicle maintenance and operational employee support areas
•Laboratory operations
•Long-term drum storage
•Transport vehicle staging
Upon entering the site each day, all employees must present their
badges to the guard and sign the log before proceeding to their
designated areas. All employees wearing orange badges will be
allowed access to*Zone I; however, only employees wearing green
badges will be allowed access to Zone II.
Visitors are defined as all individuals who are not employed by
SCA Chemical Services on site. No admittance of visitors shall be
allowed unless pre-approved by the Site Manager, Operations
Manager, Laboratory Manager, Safety Manager, or Accounting
Manager.
Figure 5.
General Site Plan
All visitors who are approved to enter Zone II must sign a re-
lease. All drivers of waste transport vehicles or material delivery
vehicles will be allowed access to Zone II provided they comply
with the requirements of the liability release. All visitors, includ-
ing drivers, must remain in their vehicles at all times.
The use of an on-site security force will be required through-
out the duration of the project. The force size will vary through-
out the duration of the project as needs change. Twenty-four hour
coverage will be required. The guard house gate shall be closed and
locked at all times after the operational hours of the site.
CONCLUSION
The entire project is scheduled to last four years during which
approximately 85,000 drums will be exhumed. Numerous tech-
niques will be developed throughout the operation which are ex-
pected to lay the groundwork for future projects of this nature.
-------
PCB'S AT SUPERFUND SITES:
REMEDIAL ACTION EXPERIENCES
JOHN W. THORSEN
ROBERT J. SCHOENBERGER, Ph.D.
Roy F. Weston, Inc.
West Chester, Pennsylvania
ANTHONY S. BARTOLOMEO
U.S. Environmental Protection Agency
Philadelphia, Pennsylvania
INTRODUCTION
Under the Comprehensive Emergency Response, Contingency,
and Liability Act of 1980 (Superfund), three types of response ac-
tions were identified: (1) emergency actions, (2) planned removals,
and (3) remedial actions. Remedial actions have the longest time
frame, i.e., greater than six months to implement.
Lehigh Electric and Engineering Company Site in Old Forge,
Pennsylvania, was the first USEPA-lead remedial action program
under Superfund. As a result of being the first, there were many in-
teresting and new procedures that were developed and lessons that
were learned in cleaning up uncontrolled sites.
In addition, efforts to clean up this site were particularly inter-
esting because of the high level of citizen interest, input and in-
volvement in the actions needed to plan and conduct a cleanup of
the site. Further, this is the first site where the USEPA and
U.S. Army Corps of Engineers worked together as a team to devel-
op a plan and contract for the cleanup services.
Lehigh Electric and Engineering is a 4.4 acre site located in Old
Forge, Pennsylvania, adjacent to the Lackawanna River. The site
was originally a coal breaker and a steam-powered electric gen-
eration facility. Since the early 1960s, it has been used as an elec-
tric equipment storage and repair facility. There is one main build-
ing in the center of the property that was originally the electric gen-
erating station.
USEPA'S INITIAL INVOLVEMENT
In Mar. 1981, the USEPA was notified by an anonymous source
of the existence of a transformer storage yard in Old Forge, Pa.,
at which improper handling of PCB-containing equipment may
have occurred. The USEPA, along with its Technical Assistance
Team Contractor (Ecology and Environment, Inc.) and the Penn-
sylvania Department of Environmental Resources (DER) con-
ducted the initial inspection of the site from Mar. 26 to Mar. 28,
gathering soil samples and water and sediment samples from the
Lackawanna River.
The initial soil sample analyses indicated the presence of PCB at
elevated levels (i.e., 5000 ppm). The USEPA On-Scene Coordina-
tor (OSC) determined that additional sampling must be performed
at the site to assess the degree of contamination, and that the site
must be secured. On Mar. 31, 1981, USEPA Headquarters issued
$50,000 in Environmental Emergency Funds to perform these
activities.
Many activities occurred during the month of April, including:
•Issuance of an Administrative Warrant to USEPA providing for
access to the site
•Collection of 50 soil samples by USEPA in a grid pattern to de-
termine the extent of surface contamination
•Taking of aerial photography utilizing the Enviropod during site
fly-overs
•Construction of a 6 ft high chainlink fence around the perimeter
of the site to provide security
•Holding of a town meeting to inform the local residents of the
USEPA's initial findings.
An Administrative Warrant was required to allow the USEPA
to take measures necessary to evaluate and secure the site since the
property in question was privately owned. The second round of soil
sampling showed the presence of PCB in the surface soil up to
65,000 ppm. The USEPA EPIC remote sensing laboratory lo-
cated in Warrenton, Virginia, completed a photo interpretation
analysis of the infrared imagery obtained from the overflight of the
site in April. Their analysis indicated gross contamination of the
site with transformer oil. They also saw evidence of distressed veg-
etative activity at the site resulting from the spillage of oil.
By May 1981, it was determined that the site was contaminated
with PCB's; it had been secured to minimize any further contact
of the local population with the contamination, and a community
relations effort had been initiated to keep local citizens apprised
of activities there. The next step was to determine the extent of
any migration of the contamination to off-site residential areas.
A comprehensive air quality monitoring survey was performed
by the USEPA Emergency Response Team (ERT). The team was
requested by the OSC to sample air quality throughout the site and
adjacent residential areas to determine if organic vapors, such as
trichlorobenzene, were emanating from the site. The study found
no indication of harmful levels of organic vapors in the air sur-
rounding the site. There were odors present at the site. However,
it was the opinion of ERT personnel that the odors were aromatic
in nature and similar to mineral oil vapors, posing no threat to
human health.
In May, the ERT Environmental Emergency Response Mobile
Laboratory was brought to the site. The ERT conducted extensive
soil sampling and analysis to determine if airborne transport of
PCB from a rudimentary incinerator operated on-site had contam-
inated the local neighborhood. Soil samples were obtained from
17 residents' yards in the area and were analyzed on-site in the
mobile lab. Laboratory analysis indicated that no PCB contam-
ination of properties adjacent to the site had occurred.
The ERT also obtained additional soil samples from the extreme-
ly oil-saturated areas on-site. These areas were targeted from the
infrared imagery obtained during the aerial overflight. Sample
analyses indicated PCB contamination in the saturated areas to be
extremely high, i.e., up to 59,000 ppm, which probably resulted
from the indiscriminate handling of dielectric fluids.
By June of 1981, the OSC had secured and assessed the site,
determined that PCB contamination did indeed exist, and that the
contamination was not migrating off-site to the nearby residential
areas. At that point, it was decided that no further emergency ac-
tivity was needed at the site, and that remedial cleanup activity
under the Comprehensive Environmental Response, Compensa-
tion, and Liability Act of 1980 should be pursued.
REMEDIAL ACTION PLANNING ACTIVITIES
The Comprehensive Environmental Response, Compensation,
and Liability Act (CERCLA) provides for government cleanup of
156
-------
REMEDIAL RESPONSE
157
hazardous waste disposal sites which are abandoned, or where
liable responsible parties do not have the resources to perform an
adequate cleanup. It allows the government to clean up sites where
responsible parties do have the resources to do so but refuse. In
such instances, the government can sue the responsible parties to
recover the cost of the Superfund financed cleanup in order to re-
plenish the fund.
In order to receive Superfund money for cleanup, a site must be
on the National Priority List, on which sites are prioritized through
the use of hazardous site ranking system. In the summer of 1981,
the Lehigh Electric Site was rated and submitted. In the month of
Oct., USEPA published what was known as the "List of 100."
This list was actually a compilation of the 115 highest ranked haz-
ardous waste disposal sites in the nation. The Lehigh Electric Site
appeared on this list and, therefore, was eligible to receive Super-
fund money for site cleanup.
All necessary administrative requirements were satisfied, and in
Oct. 1981, Roy F. Weston, Inc., the Zone 2 A/E Contractor, was
tasked to perform a site investigation and feasibility study. These
tasks would be used to evaluate the degree and extent of PCB con-
tamination at the Lehigh Electric Site, and would provide the in-
formation on which to base a cleanup strategy and design.
At the Lehigh Electric Site, as at most uncontrolled hazardous
waste disposal sites, USEPA had to deal, not only with the clean-
up of the contaminated medium (in this case soil), but also with the
source of the PCB, the electrical equipment which was still on-
site. In order to facilitate the removal of equipment from the site
as soon as possible, USEPA decided to divide both the engineer-
ing studies and the actual cleanup into two distinct phases (Table
1). Phase I was to deal with the removal of all equipment and ma-
terial from the site surface. Phase II was to deal with removal of
the contaminated soil.
Table 1.
A/E Tasks at Lehigh Electric Site
Phase I:
1. Site Survey and Map
2. Inventory and Categorize Equipment
3. Technology, Options, Cost Evaluation
4. Environmental Impact Assessment
5. Preparation of Plans and Specifications
6. Community Relations Support
Phase II:
1. Description of Current Situation
2. Investigative Support
3. Site Investigation
4. Preliminary Remedial Alternatives
5. Site Investigation Analysis
6. Final Report—Field Investigation
7. Identify Objectives and Evaluate Criteria
8. Identify Alternatives
9. Evaluation of Alternatives
10. Conceptual Design
11. Final Report
12. Community Relations Support
During the Phase I engineering studies, Roy F. Weston, Inc.,
compiled a detailed inventory of all electrical equipment on the
site. After the inventory was completed, the USEPA Environ-
mental Response Team's Mobile Laboratory was brought to the
Lehigh Electric Site for a second time. USEPA then samples every
transformer and other piece of equipment on-site. With the aid of
the mobile laboratory, they were able to determine whether or
not the equipment contained PCB liquids (Table 2).
The classification of equipment on-site was based upon the PCB
concentration. The categories were: greater than 500 ppm of PCB;
between 50-500 ppm of PCB; and less than 50 ppm of PCB. The
analytical results allowed USEPA to establish the disposal tech-
nique for the equipment in accordance with the TSCA regula-
tions.
Upon completion of the inventory and categorizing of equip-
ment, WESTON delivered a draft design for the Phase I cleanup.
Table 2.
PCB Contamination Summary
Equipment
Transformers
Capacitors
Electrical Equipment
PCB Concentration
Units <50 50
-------
158
REMEDIAL RESPONSE
formers with PCB concentrations of greater than 500 ppm. In such
a situation, the burning fuel smoke plume would carry with it nox-
ious and possibly toxic odors and particulate matter.
The County Civil Defense Coordinator established a 0.5 mile
radius of evacuation, using the site as ground zero, in which the
population would be notified of an emergency and would proceed
to designated staging points in the event of a call for evacuation.
The plan was presented to the communities of Old Force and
Moosic Boroughs and was well received by police, firemen, and
local elected officials and citizens.
The cleanup contractor began work on July 30, 1982. Their first
task was to prepare the site. This task included clearing and grub-
bing, grading, bringing work trailers on-site, and setting of a snow
fence as the demarcation line between the "clean" and "dirty"
area. For the purpose of the Phase I site cleanup, the site was divid-
ed into two areas, clean and dirty. The dirty area was where all
the equipment was located. The clean area contained no equipment
and was to be used as a staging area for the drained equipment.
The area designations were not to be construed as identifying the
soil contamination levels in those areas.
CONCLUSIONS
The foregoing description of activities at this site portrays a ra-
tional and progressive approach to cleanup. However, often the ac-
tuality of progress and activity was considerably more chaotic as
a result of the "fast track" nature of the equipment removal, the
pressures being brought to bear on USEPA, the COE and the Con-
tractor to assure the initiation of construction in July 1982, and
the fact that all parties were at the forefront of developing ad-
ministrative and technical procedures for implementation of the
program.
A very interesting aspect of the program was USEPA's de-
sire to deviate from the normal emergency cleanup contractual
methods of "time and materials" contracts to more of a "fixed
price" contract. This was unusual because fixed price contracts
assume everything is known about the site. At Lehigh Electric
and Engineering, the plans and specifications were originally put
together without a total knowledge of the contents of the electrical
equipment at the site. Therefore, a concept of lump sum and unit
price contract with estimated quantities had to be developed and
implemented. In addition, most of the contractors with the back-
ground necessary to implement and manage programs such as this
are oil and hazardous material cleanup contractors. These contrac-
tors generally respond on an emergency basis and have historically
been paid for their services on a time and materials basis because
of the unquantified nature of the level of effort needed to dean
up any given spill situation.
These firms, for the most part, do not have a long history of
working with engineering plans and specifications directing their
cleanup. They utilize their own historical background and make de-
cisions as the spill and the response unfolds. Therefore, USEPA
and the Contractor were concerned that a specification be devel-
oped in a manner that would result in bids being submitted for the
cleanup. As a result, baseline actions were included in the specifi-
cations. The manner of accomplishing the baseline actions were
left up to the contractor. An example is the construction of the
vehicle and equipment decontamination facilities. It was specified
that an emergency shower was needed in the "dirty" area which
the decontamination facility straddled. A design for the decontam-
ination facility was provided, including specifications for slope,
foundation construction, and types of materials. The successful
construction contractor placed his emergency shower in the equip-
ment decontamination portion of this facility. Therefore, it can be
seen that baseline requirements for spill control, safety and emer-
gency response, and basic construction actions were identified, but
manner of implementation was left, as much as possible, to the
construction contractor.
This approach resulted in the submittal of twelve bids. This
was determined by the Corps of Engineers to be a satisfactory
number once a contract was awarded and work begun.
This program also showed that the USEPA-U.S. COE agree-
ment works. The COE was able to provide USEPA with consid-
erable guidance on the format of the specifications including
methods to control prices, particularly in the development of the
context of the specifications, the pricing schedule and the meas-
urement and payment section. They are also showing that the
agreement works through successful construction management at
the Lehigh Electric.
Activities that have taken place include the cleanup of the elec-
trical and other miscellaneous equipment and the soil and ground-
water investigation needed to implement a cleanup of the con-
taminated land. The next step in the Lehigh Project will be to de-
velop plans and specifications and then contract for the cleanup
of the estimated 11,000 yd3 contaminated soil. Because of PCB's
low solubility in water, there is apparently no groundwater con-
tamination at the site.
-------
EVALUATION OF REMEDIAL ACTIONS FOR GROUNDWATER
CONTAMINATION AT LOVE CANAL, NEW YORK
LYLE R. SILKA
JAMES W. MERCER, Ph.D.
GeoTrans, Inc.
Herndon, Virginia
INTRODUCTION
During the coming years, considerable effort and resources will
be committed to the investigation and clean up of groundwater
contamination caused by hazardous waste disposal, spills and other
activities. The enactment of the "Superfund" legislation (Public
Law 96-152) will precipitate the spending of large sums of money to
contain and clean up inactive sites and spills of hazardous
substances. To optimize the limited resources available for the solu-
tion of these problems, and maximize the protection of public
health and the environment, remedial actions must be adequately
selected, designed and monitored.
Groundwater modeling is an excellent planning tool that can
assist in the evaluation and selection of remedial actions. Modeling
provides quantitative analyses of site hydrology and allows predic-
tion of the effects of proposed remedial actions on the fate of sub-
surface contaminants.
Simulation of remedial actions during the planning stages can
provide decision makers with considerable insight for making deci-
sions on appropriate actions to be taken at a site. Such a tool, ap-
plied from the outset of a remedial action, can reduce trial and er-
ror, costs, and time on a project. Modeling, in addition, provides
the means to predict future consequences of planned actions.
LOVE CANAL CASE STUDY
Background
In early 1980, the USEPA wanted to assess the hydrogeology of
the Love Canal area, Niagara Falls, N.Y. As part of this ground-.
water investigation, the existing groundwater remedial measures
were evaluated.
The Love Canal site is located on the east side of the City of
Niagara Falls, N.Y. (Fig. 1). The 6.5 ha landfill at Love Canal
operated for a 25 to 30 year period, and the recent remedial actions
instituted at the site are described in Leonard, et al.'
The typical strata at the Love Canal site are shown in Fig. 2. In
very general terms, the groundwater hydrology includes, from the
surface down: (1) a shallow groundwater system that is seasonably
saturated and consists of silt fill and silty sand, (2) a bed of confin-
ing material composed of clay and till, (3) a deep groundwater
system in the Lockport Dolomite, and (4) the relatively im-
permeable Rochester Shale. In this paper, the authors are concern-
ed with the hydrology of the shallow system. The analysis of the
deep system is discussed in the original report to the USEPA2 and
will be published separately.
Shallow System Hydrogeology
The shallow system at Love Canal is located in the upper units of
silty sand and silt fill. It is bounded on the north by Bergholtz
Creek, on the west by Cayuga Creek, and on the south by the Little
Niagara River (Fig. 3).
Flow in the shallow system may be discontinuous and seasonally
saturated. Field observations showed that the direction of surface
and shallow groundwater flow (before remedial action) was from
the northeast to southwest, toward the Niagara River from about
Read Avenue (Fig. 1), an area which includes about two-thirds of
the site.'
From about Read Avenue, north, the direction of surface water
and shallow groundwater movement was toward the northwest.
This conclusion is supported by vegetation damage to the north of
the canal.5
Given these observations on groundwater movement, and noting
that much of the chemical waste was placed in the southern end of
the canal, it is not surprising that the problem of contaminated
leachate movement into sumps and storm sewers was far more
serious on the 97th Street (west) side of the landfill, and more par-
ticularly on 97th Street between Wheatfield Avenue and Frontier
Avenue (near the south end of the canal). Field data indicate that
the water-table gradient is very low and that a slight mounding of
water within the fill could have resulted in migration of con-
taminants in all directions from the fill.
Remedial Actions Instituted
In summary, remedial action at Love Canal consisted of install-
ing barrier drains, a clay cap,3 and a wastewater treatment plant.
The barrier drains are approximately 5.5 m below grade and 1.2 m
across, containing French drains consisting of perforated tile pipe,
about 0.6 m of uniformly sized gravel, and sand backfill to the sur-
face. The drains were designed to intercept shallow lateral contami-
Figure 1.
The location of the Love Canal Site in Niagara Falls, New York.3
159
-------
160 REMEDIAL RESPONSE
OBfTM (M) DESCRIPTION
0.6-0.8
1.2-1.7
2.3-2.0
3.2-3.5
6.8-6.2
10.4-12.8
30.0-46.0
n
r&o £>
-------
REMEDIAL RESPONSE
161
•The shallow system is composed of silty sand and clayey silt fill
and has been estimated to have a K equal to or greater than
10~7 m/s7. The probable range of K for the shallow system is
1(TS to 1-7m/s
,-•„
10 100 10OO 1OOOO
TIME (days)
Figure 4.
Sensitivity of Modeled Results on the French Drain Flow Rate
Per Lineal Length Over Time for Various Combinations of
Hydraulic Conductivity (K), Evapotranspiration (ET) and Recharge (R).
Curve A Represents Low ET and High R; Curve B, Medium ET and R;
and Curve C, High ET and Low R.
x
a
K
O
•» FIELD DATA
* MODELED K :
- MODELED K :
1).
Definite conclusions can be drawn from these sensitivity
analyses. An average, or bulk, hydraulic conductivity for the
shallow system is between 10 ~6 and 10 ~7 m/s with a specific yield
of 10-15%. The observed flow rates are completely bracketed by
the modeled flow rate curves using these ranges of values.
Although the above analyses treat the shallow system as a
homogeneous media with averaged properties, the system really is
heterogeneous. Inserting earth materials of various hydraulic con-
ductivities into the model may provide results that would match the
field data better than the simplified assumptions represented by
Figs. 4 and 5. In Fig. 6, for example, the results of a simulation
with two higher permeability zones (K of 10 ~4 m/s and K of 10 ~5
m/s) are shown. The shallow system in Fig. 6 has a K of 10~7 m/s,
but the addition of the higher permeability zones effectively in-
crease the flow rate in the drain to a fairly close fit to the observed
flow rates. A better match could probably be obtained by adjusting
the parameters further. However, the information gained would be
of little value since there is no way to determine the validity of these
more complex simulations.
Water Table Configuration
The effects created by the French drain on the shallow system
water table are shown in Fig. 7. The top curve represents the head
distribution in the shallow system with a K of 10~7 m/s. For com-
parison, the lower curve represents a higher permeability material
with a value of K at 10 m/s. In the areas where hydraulic conduc-
tivity is 10 m/s, the zone of influence of the drain extends about
55 m from the drain.
-------
162
REMEDIAL RESPONSE
x
3
If
O
» FIELD DATA
•» HETEROGENEOUS
10 100 1000 10000
TIME (days)
Figure 6.
Modeled Results of the French Drain Flow Rate Per Lineal Length
Over Time for a Heterogeneous Aquifer of Clay and Sand.
UJ
-i
IU
•» MODELED
•ltl/8
m/s
CAVUO.A CREEK
FRENCH DRAIN
fOO 400 COO 800
DISTANCE FROM DRAIN (m)
Figure 7.
Predicted Effects of Two Values of Hydraulic Conductivity on
the Shape of the Water Table with Installed French Drain.
In those areas where material with a K of 10 ~4 m/s occurs, the
zone of influence would extend about 518m from the drain, if con-
tinuous high permeability material extended that far. Thus, in these
zones of influence, the French drain would cause a reversal of flow
direction back toward the drain.
Seasonal Effects
Seasonal effects on the shallow system and remedial actions will
lower the water table during the summer and raise the water table
during the late fall and early spring. A hypothetical well
hydrograph for a well 176 m west of the French drain and associated
drain flow rates over time, beginning Oct. 1, 1979 and extending in-
to 1981 is shown in Fig. 8. With a simplified two-season model, the
hydrograph illustrates the water-table highs that occur in early spr-
ing and the lows in late summer. Paralleling the water-table level
changes, the drain flow rates declined over the summer of 1980 and
closely match the observed values. Drain flow rates increased again
the early spring of 1981.
Time to Achieve Steady State in Drain Discharge
Modeling results indicate that the drain should achieve a steady-
state condition in about three years, reacing 90% of the steady-
state value in about one year. Thus, the shallow system at Love
Canal was approximately at equilibrium with respect to the French
drain, ignoring the seasonal fluctuations in head at the time of this
study.
I-
<
>
UJ
-I
IU
171
b
^
M
X
3
K
D
B
A S 0 N D
1B7B
J f I
loeo
TIME
-••FIELD DATA
••MODELED K
••-MODELED K
10'f
1 Q-'
m/t
Figjire 8.
Predicted Hydrograph for a Well 76 m west of the French Drain (Fig. 8A)
and Predicted French Drain Flow Rate Per Lineal Length Over Time
for Two Values of Hydraulic Conductivities as Compared to Field Data
Available as of January 1981 (Fig. 8B)
Travel Time and Flow Velocity
The predictions presented here are- posed in terms of probable
conditions based on the best estimates for the hydrologic
parameters. If a continuing monitoring program is maintained, im-
proved estimates of hydrologic parameters can be determined.
This, in turn, can result in refinement of predictions with less
uncertainty.
Predictions of travel time and flow velocity are based on the
equation:
Ki
V ~ * (D
where v = interstitial velocity
K = estimated hydraulic conductivity
i = head gradient provided by the models
<(> = effective porosity (specific yield)
Using Equation 1, groundwater velocities for various hydrologic
parameters representing the hypothetical shallow systems and
swales or sand lenses were calculated and are shown in Table 2. As
may be seen, the various combinations of parameters produce
velocities ranging from 0.02 to 19 m/yr. These values may be used
to estimate travel times and distances for contaminants traveling
with the velocity of water.
Table 2.
Groundwater Flow Velocity for Various Parameter Combination*
(Effective Porosity = 0.15)
Hydraulic
Conductivity
m/s
1.0 x 10-"
1.0 x IO"5
1.0 x 10-6
1.0 x 10-7
Gradient
rn/m
o.ooio
0.0010
0.0012
0.0013
Interstitial Pore
Velocity
m/yr
19.00
1.90
0.23
0.02
-------
REMEDIAL RESPONSE
163
Conclusions from the Love Canal Modeling
Conclusions and predictions which have been drawn from this
modeling study for a hypothetical, average shallow system (K of
10T6 m/s) are:
•Without the remedial actions taken to date, the groundwater in
the average shallow system would have migrated up to an addi-
tional 1.8 m away from the canal in the next 5 years.
•For the average shallow system, the area of influence of the
French drain will encompass all groundwater which may have
been contaminated by subsurface migration over the past 30
years. The area of influence for material with a K of 10~6 m/s
extends out 70 m from the canal while the groundwater traveled
only about 10 m from the canal during those 30 years (assuming
a more conservative effective porosity of 10%).
•With the remedial actions installed to date, the French drain will
cause all groundwater presently existing between the canal and
drain to be discharged to the drain during the next 5 years. Out-
side the drain, groundwater will flow back to the drain from as
far as about 70 m from the canal during the next 5 years. These
predictions assume an average shallow system K of 10~6 m/s,
effective porosity of 10%, and gradient of 0.036.
Conclusions and predictions which have been made for the
highest permeability zone hypothesized (K of 10~4 m/s) are:
•Without the remedial actions, the groundwater in material with a
K of 10'4 m/s (effective porosity of 15%) would travel in the
next 5 years up to an additional 96 m to a maximum distance of
675 m from the canal. The probability of this occurring is un-
known, but most likely is low because of the discontinuous,
heterogeneous nature of the shallow system.
•In swales, sand lenses, utility trench backfill, or other higher per-
meability zones, the drain's area of influence could extend out
533 m (K of 10 ~4 m/s) while groundwater could have traveled
about 570 m during the past 30 years (assuming effective por-
osity of 15%). Therefore, the drain should cause most of the
groundwater that could have been contaminated during the past
30 years in these higher permeability zones to begin flowing back
toward the drain.
•With the remedial actions taken to date, all the groundwater in
material with K of 10~4 m/s and effective porosity of 15% be-
tween the drain and canal will be discharged to the drain. All
groundwater out to a distance of about 275 m from the canal
will be discharged to the drain during the next 5 years.
•Less than 11 years will be required for groundwater which
originated at the canal 30 years ago to be drawn back to the
French drain in material with K of 10~4 m/s and effective por-
osity of 15%. This disregards attenuation effects on contamin-
ants and assumes a maximum travel distance of 580 ft from the
canal over the past 30 years.
ALTERNATIVE TRANSPORT MECHANISM
The above analysis of the extent of groundwater contamination
at Love Canal is based on the assumption that the contaminants
were transported via subsurface flow only. An alternate mode of
travel which was not considered in this study is that of overland
flow. Additional insights into the problems at Love Canal can be
gained by considering the case at West Valley, NY. Here, burial
trenches for disposal of solid radioactive waste were excavated in
glacial till with a low hydraulic conductivity similar to that ob-
served at Love Canal, about 5 x 10~10 m/s."
Prudic and Randall" observed that above the water table, frac-
tures and root tubes extended 3 to 4.5 m below the natural land sur-
face. Preliminary simulations suggested that the hydraulic conduc-
tivity in these fractured tills is an order of magnitude greater than
the unfractured till. It is thought that the fractured till allowed
recharge to reach the trenches, and since the material surrounding
the trenches has a much lower hydraulic conductivity, water began
to fill up the trenches. As a result, water overflowed in the older
trenches briefly in 1975 until they were pumped out. Therefore,
most soil and shallow groundwater contamination away from the
West Valley site has been attributed to an intermediate transport
mechanism of overland flow.
It is difficult to assess if similar overland flow of contaminants
occurred at Love Canal. The hydrogeologic conditions at the two
sites are similar, and any interpretation of the mechanisms of con-
tamination at Love Canal should allow for this possibility. This is
especially valid considering the reported exposure of the waste in
Love Canal to surface recharge at various times during its ex-
istence.'
SUMMARY
The utilization of computer modeling to assess and predict the
effectiveness of remedial actions for groundwater contamination
cases can provide the decision maker with an important tool in
selecting the best plan. From generic and specific modeling applica-
tions, insight can be gained on operating characteristics of the
remedial method. If employed in the design phase, a modeling
study can provide reasonable estimates of the amounts of con-
taminated groundwater produced, extent of contaminant migration
over time and the required life span of remedial operations. These
data will assist the engineer in optimizing the final remedial design.
Although solute transport models may be necessary for some
contamination problems, much information can be gained by con-
sidering groundwater flow alone. These applications have been
described in detail by Nelson12 and are demonstrated by the model-
ing applications to the Love Canal area described in this paper.
Field investigations of groundwater contamination cases should
be carried out in such a manner as to facilitate a responsive, yet
scientific, appraisal of the situation. The best approach to such an
investigation is through an iterative, feed-back system linking data
collection with analysis. Through this phased approach, the collec-
tion of large amounts of data that may be largely useless will be
avoided. Resources will be maximized to produce the most accep-
table results for all parties involved. This iterative approach is
outlined as:
•The review and analysis of existing data
•The performance of preliminary modeling to gain a better
understanding of the hydrogeologic system
•The determination of further data requirements and field work,
based, in part, on preliminary modeling results
•The analysis of the additional data
•The refinement of the preliminary model and determination of
further data requirements
•Continuation of the above iteration of data collection, analysis,
and interpretation until an acceptable level of confidence is at-
tained
Following this general pattern provides a flexible approach to in-
vestigating groundwater problems. Integration of analysis and in-
terpretation with data collection can allow the investigator to
dispense with unnecessary data and, therefore, reduce cost. The use
of groundwater models, as described in this paper, can assist in
even greater cost savings by allowing the investigator to achieve a
better understanding of the groundwater system and contaminant
behavior in the subsurface.
Because of the time constraints imposed on the Love Canal
study, this type of iteration and feedback was not possible and the
groundwater investigation preceded the laboratory analysis. Even
with the lack of data and the time constraints (three months to
complete the study), it was still possible to make calculations that
gave valuable insight into the groundwater flow regime at Love
Canal. This fact is proven by data made available only recently, the
monthly French drain flow rate for the period from Aug. 1979
through Feb. 1981 are given in Table 3. The comparison of these
observed flow rates versus the predicted flow rates over the same
period (Fig. 9) provides additional confidence in the validity of the
model.
-------
164
REMEDIAL RESPONSE
E
i*
b
x
3
oc
Q
FIELD DATA •
MODELED K : 10'*m/s-
MODELED K : 1
-------
MONITORING CHLORINATED HYDROCARBONS
IN GROUNDWATER
D.A. PALOMBO
J.H. JACOBS
Engineering-Science, Inc.
Cleveland, Ohio
INTRODUCTION
Groundwater contamination by chlorinated hydrocarbons has
been found at numerous sites around this country and others. The
literature contains examples of sites where contaminated ground-
water would be expected, but just as many where contaminated
water was found at unpredictable lateral and vertical locations and
at unexpected concentrations. Trichloroethylene (TCE) and related
chemicals seem to be the most common of the groundwater con-
taminants. The literature, coupled with experience at sites across
the country, has led to several considerations in groundwater
monitoring activities at such sites and to several conclusions about
subsurface migration of some chlorinated organics.
The authors' purpose in this paper is to discuss these considera-
tions and conclusions so that they may spawn criticism or further
development. The authors believe that it is necessary to develop
and provide, to all concerned, the basic information required to ef-
ficiently conduct remedial actions at groundwater contaminated
sites. Several sites are discussed only in terms of their general
features, but are based on real sites and illustrate both the complex-
ity of natural hydrogeologic systems and the difficulties of
monitoring chlorinated hydrocarbons below the water table.
Specific information on contamination by chlorinated organic
compounds has been the subject of very few technical publications.
This is probably the result of a lack of knowledge on the subject as
well as the confidentiality of much of the available data. Addi-
tionally, the physical properties and the migration characteristics of
these compounds do not lend themselves easily to standard prin-
ciples of hydrogeological investigations. This latter consideration
has become a strong incentive toward research activities in this
area.
The following discussion, though not exhaustive, presents a sum-
mary of the available information on some chlorinated organics in
the subsurface and some of their pertinent physical properties as
they would affect subsurface migration. Although numerous
chlorinated compounds exist and probably have variable transport
characteristics, only the listed compounds in Table 1 have been
considered here.
Significant Physical-Chemical Properties
Information on the occurrence, migration and fate of particular
chlorinated organics in an unsaturated environment is relatively
rare. Some of what is available is site-specific and cannot yet be
transferred to the public. The literature does provide important
physical characteristics of the compounds of concern here, as well
as some of their characteristics in dilute aqueous solution. The pro-
perties of these compounds, which appear to be of the most impor-
tance in attempting to predict subsurface movement, are given in
Table 1.
One obvious property of the compounds is their relatively high
vapor pressure and low boiling point. Experiments conducted by
Dilling, et al.' provide information on the time required for an in-
itial 1 mg/1 aqueous solution of each listed compound to decrease
(evaporate) 50% and 90% in concentration (at 25 °C). In general,
the listed compounds decreased 50% in less than 0.5 hr and
decreased 90% in about 1.5 hr. The values are not unique because
of uncontrolled variations, but they do indicate a rapid decrease in
concentration from open containers at 25 °C.
Dilling repeated the experiments after adding certain "con-
taminants" to the initial solution. These "contaminants" were dry
bentonite, peat moss, sand, and dolomite. The experiments were
again repeated with the contaminants in closed containers. All of
these experiments were an attempt to determine which of the con-
taminants increased the rate of disappearance of the organic com-
pounds from the aqueous solution. The authors concluded that
peat moss (organics matter) was the most effective in "adsorbing"
the chlorinated compounds from the sealed container.
Although these experiments were simple, the implication exists
that the higher the organic content of the soil, the better able they
will be in adsorbing a release of chlorinated organics at least from a
dilute aqueous solution.
Table 1.
Physical Properties of Selected Chlorinated Hydrocarbons
of Low Molecular Weight
Constituent
carbon tetrachloride
chloroform
methylene chloride
trichloroethylene
1,1,1 trichloroethane
tetrachloroethylene
water
chlorobenzene
1 ,2,4-trichlorobenzene
Solubility Liquid1
in Water Density
(mg/1) (gm/cc)
Dielec-
tric1
(20 °O
Vapor
Pres.
(mmHg
@20°C)
Biodeg."
800
8,000
16,700
1,100
4,400
150-
—
500
.59
.49
.49
.45
.33
.62
.00
.10
2.238
4.806
9.08
3.4
...
...
80.37
5.708
91
192
438
74
100
19
17.5
9
D
A
D
A
B
A
-
D
28.6 1.45
•A— Significant degradation with gradual adaptation— in aerobic aqueous solution.
B — Slow to moderate biodegradative activity, concomitant with significant rate of volatilization.
D— Significant degradation with rapid adaptation.
T— Significant degradation with gradual adaptation followed by deadaptive process (toxicity).
1— from Weast".
2 — from Tabak, et a/.11.
Another important physical property of the chlorinated
hydrocarbons of concern here is their high liquid densities; all are
greater than water near 20 °C (Fig. I).2 According to the National
Research Council "if the chlorinated compounds were spilled or
released as a slug they would tend to settle in any receiving water
before dispersion, volatilization, emulsification or solubilization
takes place".3
This density factor, when considered in terms of a liquid spill on
land or a release to the subsurface (tank leakage) will assist in the
downward migration through both the unsaturated and saturated
zones. One particular case, described generally in Saines4 recog-
nized the high density of a contaminant solution and its migration
165
-------
166 REMEDIAL RESPONSE
TEMPERATURE (°F)
100 200 300
• MELTING POINT
A BOILING POINT
• CRITICAL POINT
40 80 120
TEMPERATURE
2.5
280 320
Figure 1.
Liquid Density of Selected Chloromethanes at Various Temperatures
From Yaws, 1976'
by gravity rather than by the hydraulic gradient. The schematic in
Fig. 2 illustrates a particular hydrogeologic situation where dense
leachate had moved beyond the normal groundwater discharge
point. Although some production wells may have induced the
migration across the river somewhat, the denser-than-water
leachate promoted migration down the slope.
Studies to determine the effects of organic leachate on the
permeability of clay liners' have concluded that other factors are in-
volved in affecting the rate of organic leachate movement in the
subsurface. Specifically, the ability of some organic solvents to
structurally alter the clay and make the clay more permeable has
been observed. Although none of the specific chlorinated com-
pounds of concern here were tested, the neutral non-polar com-
pound xylene was used. Of the four classes of organic fluids tested,
namely, acidic, basic, neutral polar and neutral non-polar, the
chlorinated organics are of the neutral non-polar type. All four of
the different clayey soils treated with xylene showed a very rapid in-
crease in permeability followed by a leveling at a permeability
about two orders of magnitude greater than the clay's permeability
to water. These increases are much greater than can be attributed
just to viscosity/density of xylene. They hypothesize that structural
changes had occurred in the clay.
It is not suggested here that the xylene typifies the affects of TCE
or other chlorinated solvents in the tested clays, however, it is not
unreasonable to assume that they would act in a similar fashion.
Some work by Green et a/.6 does support this contention in studies.
The experimenters tested the effects of several reagent-grade
organic fluids on three clayey soils: xylene, carbon tetrachloride
and trichloroethylene. The experiment, designed to obtain a
permeability of the particular solvent through the clay, showed that
the solvents initially exhibited a permeability about one order of
magnitude less than that of water. Those solvents with low dielec-
tric constants (including xylene, carbon tetrachloride and trichloro-
ethylene), caused significant degradation of the clay soils. After a
period of days, these solvents would break through the test col-
umns and allow a bulk transfer (or release) of the solvent through
the column. The solvents of low dielectric constants caused an ob-
vious shrinking and cracking of the clays in both consolidometer
and column tests. The authors developed an empirical relationship
between the permeability, the solvent's dielectric constant and the
packed bulk clay density before breakthrough. Results of these
tests may be significant in developing predictions of direct releases
of solvents (tank spills, leaks or other bulk loss) onto unsaturated
sediments.
UNSATURATED ZONE TRANSPORT
At several locations, soil samples obtained during the drilling of
monitoring wells or exploratory borings, were analyzed for
suspected chlorinated organics. The procedure was generally one of
hydraulically collecting an undisturbed sample below the lead
LOWER ARTESIAN
WELL
WELL
I SEAL8P 10 1W
SURPACI
WELL V
WELL 9
HO CONTAMINATION
HlfiH CONTAMINATION OF LI6HT OMANICS ANP
LWC0NTAMINATIOM Of PBN6E 6HLOWNATBP
LW CONTAMINATION Of LIGHT CWbANICfc '
HiaH CONTAMINATION Of PtHft 6HLOXIKI^
NO CONTAMINATION of LlflHT OK6ANICS ANP
LOW £ONTAMINATI0H Of OF .PCNSB CHU^lHATBP (SKiaAMICfe
Figure 2.
General Pattern of Monitoring Wells in a
Simplified Multi-Aquifer Setting
auger, minimizing its air exposure, and transporting the sample to a
laboratory. The actual procedures are detailed in recommended
sampling methods.7
Results of the laboratory testing provided some interesting con-
clusions about the migration of specific constituents in the subsur-
face. Negative results for chlorinated hydrocarbons in unsaturated
sediments (especially coarser-grade) do not allow one to conclude
that a source at the surface above the sample's location did not ex-
ist (Fig. 3). Detailed data from soil column experiments' using a
mass balance approach indicate that virtually all trichloroethylene
from a 1 mg/1 aqueous solution could be accounted for either by
volatilization or by the liquid collected after passing through a san-
dy soil. The average soil composition was 92% sand, 5.9% silt,
2.1 % clay and 0.087% organic carbon. Though the TCE was slow-
ed to approximately one-half the velocity of the standard water
tracer, it was not adsorbed on the soil particles. This non-adsorbent
tendency of TCE and perhaps other compounds on soil particles
may account for the observed negative results in the soil samples
discussed above.
These studies also involved column experiments for other
organics which indicated a much higher tendency than the TCE to
be retained in the soil. This group included such compounds as
chlorobenzene, 1 ,2-dichlorobenzene and 1 ,2,4-trichlorobenzene.
When compared to TCE, the laboratory tests indicate that these
compounds would be a lesser threat for groundwater contamina-
tion since they are better adsorbed on the soil particles. Data ob-
tained from field studies at several sites indicate that this conclusion
OR6A.KIIC FLUID
SOLID BEDROCK
GWOUNDWATBR FLOW IS PERPENDICULAR
TO THE PLANE op CROSS sectioM-
WEULS 1,3 HA.V6 |_OtV LEVEL. Op
CONTAMINATION.
Figure 3.
General Pattern of Monitoring Wells in a
Gravel-Filled Bedrock Channel
_56\LEP TO THE
SURFACE
-------
REMEDIAL RESPONSE
167
is indeed the case. But this effect probably results from both the
greater adsorptive capability and also greater tendency toward
biodegradability of these compounds.
In predicting the degree of possible attentuation, investigators'
have found that the most significant properties of the soil are its
bulk density and the soils' organic carbon content. A mathematical
model, based on the organic compound's solubility and the organic
carbon content of the soil, was used to predict the retardation of in-
dividual compounds as they moved through a soil column. The
retardation factors listed in Table 2 provide a relative indication of
how fast the particular compound travels through the sandy soil
tested. Of significance also was the volatility of the compounds
from the soil was reduced by about an order of magnitude. The im-
plications upon release of a concentrated slug on top or within
sediments are obvious in that loss through volatilization becomes
much less of an attenuation factor.
Table 2.
Retention of Organic Compounds by Soils
(Adapted from Wilson, et al.')
Compound
chloroform
1,2-dichloroethane
tetrachloroethylene
trichloroethylene
chlorobenzene
1,4-dichlorobenzene
1,2,4-trichlorobenzene
toluene
Water Solubility
(mg/1)
8000
8690
150
1100
500
79
28.6
515
Predicted Retardation
Factor*
1.2
1.2
3.2
1.6-2.0
1.9
4.0
7.0
1.8
'Retardation Factor = interstial water velocity
velocity of compound
SATURATED ZONE MIGRATION
Although the results of laboratory column experiments are of
prime importance, further studies similar to those conducted by
Roberts, et a/.10 are necessary. These studies (sponsored by the
Groundwater Research Laboratory at Ada) attempted to determine
transport properties of organics compounds within an aquifer. The
behavior of several types of organic compounds were analyzed by
means of a series of observation wells surrounding an injection
well.
In this study, Roberts et al. considered the aquifer as a quasi-
closed vessel where inflow and outflow from an aquifer-element
could be mathematically adapted to the stimulus-response concept.
This approach was used to develop relationships of contaminant
concentrations versus time in the observation wells. Based on the
shape of a semi-logarithmic plot, conclusions were made on the
relative importance of the mechanisms of dispersion, adsorption
and biodegration active within the aquifer.
Of those organics tested, chlorobenzene was most rapidly
transported with a half-time of response about 40 times less than
that of a chloride tracer. The primary retention mechanism ap-
peared to be adsorption. The dichlorobenzene isomers and 1, 2, 4
trichlorobenzene were more strongly adsorbed than chlorobenzene,
and naphthalene which is a fused benzene ring seemed to undergo
some biodegradation. Although results are presented for essentially
just benzene compounds and do not necessarily reflect the behavior
of all chlorinated organics, they generally indicate that some at-
tenuation processes are active within a confined aquifer.
Giger and Kubica' reported on groundwater contaminated by the
compound, tetrachloroethylene in wells near Dubendorf,
Switzerland. Levels of up to 1 mg/1 were determined from some
groundwater samples. The authors concluded: "this organic sol-
vent is reasonably soluble in water, has a low tendency to adsorb
compared to pesticides and petroleum hydrocarbons, and is only
very slowly degraded. Thus, it is likely to be transported with the
water flow and to be quite mobile as well as long-lived in the
groundwater environment.'"
In general, aquifer contamination by TCE and other organics
solvents has been found where groundwater samples had been
analyzed for these constituents. Owing to the widespread use of
these chemicals, it is probable that there are many more local in-
stances of contamination which will go unnoticed. Even in those
areas known to be affected, often the time of release, concentra-
tion/volume of the release, nature of the release (i.e. spill or seep)
and the source itself cannot be determined. Therefore, economic
practicality has dictated that rehabilitation efforts be directed
toward wastewater treatment, rather than exploration, monitoring,
and rehabilitation of the aquifer itself.
MONITORING CONSIDERATIONS
Effective groundwater monitoring at a site contaminated by
chlorinated hydrocarbons requires the collection of more site
details, more carefully collected water quality data and more
technical than at sites of inorganic or light hydrocarbon contamina-
tion. The advantage of having historical information on type,
quantity, rate, location of waste disposal or release cannot be
overstated. However, this is usually the most limited information
available. Generally, the monitoring program must be developed by
progressively obtaining subsurface information until the major
hydrogeologic controls on lateral and vertical migration are deter-
mined. It then becomes a matter of identifying the limits of the con-
taminants in both the vertical and lateral directions. Of course,
data on concentration levels in both the sediments and groundwater
can and should be obtained concurrently with the hydrogeologic in-
formation.
Somewhat standardized monitoring procedures for sites of con-
taminated groundwater have been developed. Detailed procedures
for obtaining groundwater and sediment samples' emphasize
specific problems and solutions for sampling volatile organic com-
pounds. The discussion here provides observations at sites of active
monitoring where standard hydrogeologic assumptions and pro-
cedures may have to be altered where chlorinated hydrocarbons are
present. The following list provides some specific considerations:
•Sites of active monitoring seem to indicate that a strong relation-
ship exists between the relative density of the chlorinated com-
pounds and their apparent migration below the water table. Ex-
amples of the generalized subsurface features and findings at
several sites are presented in Figs. 2, 3 and 4.
•The assumption that a (normal hydraulically impermeable layer)
clayey unit of 10~7 cm/sec or less permeability will function as a
barrier to downward migration of a concentrated release is not
necessarily true.
•Where concentrations of total chlorinated organic compounds
exceed several mg/1 in groundwater (or in sediments), considera-
tion should be given to bentonite's capability of providing an
impermeable barrier or seal. The fact that laboratory tests show
that some organic solvents will attack the structure of clay im-
plies that some aqueous concentration level exists, above which
the clay's ability to prevent migration becomes questionable.
For monitoring wells, it can be seen that the bentonite pellets or
slurries used for isolating the monitored interval may not be ap-
propriate where high levels of these organics exist.
•Monitoring wells should be constructed so that the vertical dis-
tribution of contaminants can be identified, above and below a
"confining" layer if chlorinated organics are present (at any
aqueous concentration level).
•Technical evaluations must consider the slope of the top of an
"impermeable base," such as solid granite, as one of the signifi-
cant methods for prediction of migration of a dense slug of con-
taminant. It is possible that if this slope is different than the
water table (or potentiometric) slope, the migration rate and di-
rection is not necessarily estimated by standard hydraulics.
•The seepage or release of chlorinated hydrocarbons onto a sandy
sediment, where organic matter in the soil is low, provides a high
probability that the chlorinated organic will reach the water table
even if the depth to the saturated zone is 30 m or more. The rate
of downward movement in the unsaturated zone will, of course,
-------
168
REMEDIAL RESPONSE
etrin
11 .souses
4* ein
TO fJe SURFACE
WCLLS 1 , Z ' WO COMIAMWATIOM
WELLS 3,5: LOW CONTAMINATION
kVCLL 4 : HI6M CONTA.M»HATIOH
Figure 4.
Generalized Appearance of Contamination Plume in a
Multiple Groundwater Flow System
be a function of the flushing action of precipitation or other sur-
face water driving forces.
•A monitoring program designed to determine the effectiveness
of a well withdrawal system must consider the effect of the
density of a concentrated slug of chlorinated organic. Studies
have shown that the pumping of a slug below the water table re-
quires a much larger pump (with no plastic or rubber internal
motor parts) in order to life the denser fluid out of the well and
into a collection vessel. Because the density of the fluid requires
this extra power, then the monitoring wells must be located
closer to the source than would otherwise be necessary. These
wells would provide measurements vertically within the affected
aquifer and reflect changes in the thickness, shape and character
of the slug.
•Because proposed standards for specific solvents in drinking water
are about 0.5 mg/1 or less, the disposal or loss of only a small
volume of solvent can cause concern. The procedures and de-
gree of accuracy must be established for laboratories conduct-
ing analyses for these organic solvents. The utility of TOX (total
organic halogen) scans, gas chromatography results or GC/MS
results should be discussed prior to testing.
•Monitoring well construction should consider possible interac-
tions among various chlorinated and non-chlorinated solvents
which may influence the behavior of complex mixtures.
DATA REQUIREMENTS FOR MONITORING
AND REMEDIAL ACTION DESIGN
Listed below are the categories of information that must be in-
cluded for the waste disposal site and the immediate surrounding
vicinity. The data must be scrutinized, especially with regard to the
vertical and lateral extent of groundwater quality degradation. A
detailed knowledge of the instruments of contaminant detection
and migration must be obtained. For instance, assume that a water
sample taken from a water supply well exhibits a small degree of
contamination. If this well is open in several producing zones, this
sample may say nothing more than a problem is indicated. The
sample should therefore, not be used to identify the vertical, lateral
or degree of contamination of a particular zone. The issue is com-
pounded by problems associated with sampling and laboratory
analysis. Clearly, it is of critical importance that interpretation of
available data be made in full knowledge of how the data have been
obtained.
•Geologic. A description of geologic structure and stratigraphy is
necessary. At the subject site, the actual areal and vertical ex-
tent of rock type and/or unconsolidated units must be known.
•Hydrologic. A description of each hydrologic unit including
areal/vertical extent, hydraulic properties and identifying whether
it is a confining unit, a permeable or producing zone, or a semi-
confining interval is required. For producing zones, it is neces-
sary to determine the lateral hydraulic gradient, recharge areas,
unconfined or confined conditions, effective velocity of water,
hydraulic relationship to other producing zones and to surface water.
•Water Use. One must determine (as much as possible) the past
and present groundwater use which has or may have had an in-
fluence on the groundwater flow beneath the site. A detailed
knowledge of producing zones (as described above) and the loca-
tion and construction of production wells are important factors,
but are of particular concern where groundwater is extensively
used.
•Water Quality. One should identify the water quality features of
each producing zone in a vertical and lateral sense. Obviously,
the differences involved in sampling from producing zones of
2 m thick and 30 m thick must be considered. As previously
noted, the data for water samples must be interpreted in terms
of the well construction, well location, sampling techniques and
analytical testing.
•Chemical/Physical Properties of Leachate. This is a very impor-
tant consideration which is not easily characterized if the history
of operations is not fully known. The reactability of the leachate
with certain local clays and any liner material must be known.
Other physical properties such as solubility, density, volatility
can be directly related to the potential mobility of the organics
in a subsurface environment. Knowledge of these properties of
the leachate is particularly important in locating and designing
off-site migratioan control systems.
One must also recognize the uncertainties involved in an in-
vestigation of this type. Information such as geologic correlation,
past hydraulic gradients, operational history all are considerations
which may have caused an area or pocket of contaminated ground-
water to exist at a seemingly impossible location. The degrees of
uncertainty are, of course, site specific, but awareness of the poten-
tial uncertainties may result in & data collection scheme that will
minimize their consequences.
REFERENCES
1. Dilling, W.L., Tefertiller, N.B., and Kallos, G.J., "Evaporation
Rates and Reactivities of Methylene Chloride, Chloroform, 1,1,1-
Trichloroethane, Trichloroethylene, Tetrachloroethylene and Other
Chlorinated Compounds in Dilute Aqueous Solutions," Envir. Sci.
Tech., 9, Sept. 1975, 833.
2. Yaws, C. T., Chem. Eng., 83(14) 1976, 81-89.
3. National Research Council, "Chloroform, Carbon Tetrachloride,
and Other Halomethanes: An Environmental Assessment," National
Academy of Sciences, Environmental Studies Board (Commission on
Natural Resources), Washington, D.C., 1978.
4. Saines, M., "Errors in Interpretation of Ground-Water Level Data,1'
Groundwater Monitoring Review, I, No . 1, Spring 1981.
5. Anderson, D.C., Brown, K.W., and Green, J., "Organic Leachate
Effects on the Permeability of Clay Liners," Proceedings National
Conference on Management of Uncontrolled Hazardous Waste
Sites, Washington, D.C., Oct. 1981, 223.
6. Green, W.J., Lee, G.F., and Jones, R.A., '-'Clay-soils Permeability
and Hazardous Waste Storage," Water Pollution Control Federa-
tion, 53, 1981, 1347.
7. Scalf, M.R., McNabb, J.F., Dunlap, W.J., Cosby, R.L., Fryberger,
J., Manual of Ground-Water Sampling Procedures, Robert S, Kerr
Environmental Research Lab., USEPA, 1981.
8. Wilson, J.T., Enfield, C.G., Dunlap, W.J., Cosby, R.L., Foster,
D.A., and Baskin, L.B., "Transport and Fate of Selected Organic
Pollutants in a Sandy Soil," J. of Envir. Quality, 10, No. 4, 1981.
9. Giger, W. and Molnar-Kubica, E., "Tetrachlorethylene in Contam-
inated Ground and Drinking Waters," Bulletin of Environmental
Contamination and Toxicology, 19, Apr. 1978.
10. Roberts, P.V., McCarty, P.L., Reinhard, M., Schreiner, J., "Or-
ganic Contaminant Behavior During Groundwater Recharge," J.
Water Pollution Control Federation, 52, 1980, 161.
11. Weast, R.C., Handbook of Chemistry and Physics, 45th ed., The
Chemical Rubber Company, Cleveland, Ohio, 1964.
12. Tobak, H.H., Quave, S.A., Mashni, C.I., and Barth, E.F., "Biode-
gradability Studies with Organic Priority Compounds," /. Water
Pollution Control Federation, 53, 1981, 1503.
13. Chiou, C.T., Peters, L.J., and Freed, V.H., "A Physical Concept of
Soil-Water Equilibria for Nonionic Organic Compounds," Science,
206, Nov. 16, 1979.
14. USEPA, Procedures Manual for Ground-Water Monitoring at Solid
Waste Disposal Sites, USEPA Office of Solid Waste, SW-611, 1977.
-------
DRUM HANDLING PRACTICES AT ABANDONED SITES
ROGER WETZEL
KATHLEEN WAGNER
JRB Associates
McLean, Virginia
ANTHONY N. TAFURI
U.S. Environmental Protection Agency
Municipal Environmental Research Laboratory
Edison, New Jersey
INTRODUCTION
Land disposal of drums containing hazardous wastes.has been a
widespread practice throughout the United States. Many problems at
abandoned waste disposal sites have been attributed to drum
disposal. The results of a 1980 survey of remedial actions at 169 sites
conducted by the USEPA indicates that about one quarter of all
abandoned sites have major drum related problems.'
Past drum disposal practices have resulted in many instances of
groundwater, air, soil and surface water contamination. At some
sites explosions and fires have resulted from reactions of incompati-
ble wastes which leaked from drums.
The cleanup of waste sites containing drums poses several hazards
to field workers and to the environment. In many instances the in-
tegrity of the drums is poor and they are prone to rupture and
leakage. The drum contents are frequently unknown and it is not
unusual to find drums containing incompatible wastes in the same pit
or trench. The hazard of cleaning up these waste drums is exacer-
bated in many instances, by the unsuitable site conditions in which
they are found. Drums containing hazardous wastes have been found
in a variety of environments including wetlands, the banks of rivers
and streams and highly industrialized, confined areas.
Because of the safety and environmental hazards involved in
handling drums and the widespread efforts currently underway to
clean up hazardous waste sites, the USEPA Oil and Hazardous
Materials Spills Branch in Edison, New Jersey, has undertaken a
research effort to evaluate drum handling practices and improve
upon the existing equipment and methods where safety and effec-
tiveness are seriously lacking.
As a means of documenting the state of the art and identifying
research needs, a manual reviewing the applicability, advantages and
disadvantages of equipment and methodologies for handling drums
was developed. The manual addressed the following activities:
•Detecting and locating drums
•Determining drum integrity
•Excavation and on-site transfer of drums
•Recontainerization and consolidation
•Storage and shipping
The selection and implementation of equipment and methods for
handling drum related problems is made on a site-specific basis. Site-
specific factors which affect the suitability of various equipment and
methods include: the number of drums present at a site, depth of
burial, accessibility of the site, integrity of the drums, site drainage,
and types of waste present.
The typical sequence of activities involved in a site cleanup opera-
tion can logically be divided into five phases:
I. Preliminary Site Assessment
II. Field Investigation and Drum Detection
III. Development of a Cleanup Strategy
IV. Site Cleanup
V. Temporary Storage and Transportation
The report focuses on equipment, methods and protocol that are
used for activities undertaken during Phase II through V. Phase I,
the Preliminary Site Assessment is generally completed by the time
the drum handling operation begins and is not considered in detail.
Nevertheless, the cleanup contractor, and Federal and state officials
will need to draw on information gathered during this phase to
reduce safety hazards and the cost and time required for subsequent
activities. Background data on the site can often provide a great deal
of insight on the types of wastes, condition of the drums and site
specific variables which will influence the selection of equipment.
As with Phase I, Phase II, Field Investigation and Drum Detec-
tion, relies upon a considerable amount of field work that has
generally been completed by the time the drum handling operation is
begun. Drum detection using aerial photography, geophysical
surveying and soil sampling are examples of activities that are con-
ducted in Phase II.
Phase III, Development of a Cleanup Strategy, provides an over-
view of the four major variables which effect the selection of equip-
ment and methods for handling drums:
•Safety precautions
•Protective and mitigative measures for environmental releases
•Site specific conditions as they influence equipment selection and
performance, and
•Costs
Phase IV considers equipment and methods for site cleanup after
drum removal in terms of their applicability, safety, advantages,
limitations and relative costs. Equipment and methods for tem-
porary storage and transportation of wastes from drum cleanup
operations are considered in Phase V.
The manual emphasizes both safety measures and measures for
preventing or mitigating environmental releases during drum
handling (Tables 1 and 2).
Although there are a variety of equipment types and methods
available for handling drums at abandoned waste sites, there are
several areas where the safety of the operation may be improved by
additional research and development. As a result of discussions
with cleanup contractors, equipment manufacturers, and USEPA
and State officials, additional research and development is recom-
mended in the following areas:
•Remote drum opening methods
•Protocols for handling lab packs and gas cylinders
•Recommendations for minimum compatibility testing
•Inventory of major equipment used for drum handling
•Further evaluation of the use of non-destructive methods for de-
termining drum integrity.
The next phase of this project will develop equipment to satisfy one
or more of these needs.
DRUM HANDLING ACTIVITIES
Detecting and Locating Drums
Detecting drums at an abandoned site involves the use of historic
and background data on the site, aerial photography, geophysical
surveying and sampling. Background data should be carefully ex-
169
-------
no
REMEDIAL RESPONSE
amined since it can minimize the cost and increase the safety of
subsequent activities. Aerial photography involves the use of
historic aerial photographs to show changes in the site over time,
such as filling in of trenches, or mounding of earth. It also includes
the use of current aerial imagery (usually color or infrared) to show
spills, seepage or changes in vegetation which may indicate the
presence of drums.
The applicability, reliability and cost effectiveness of geophysical
survey methods are highly dependent upon site specific
characteristics. Magnetometry is generally the most useful survey
tool for detecting drums. Metal detectors may be of value if drums
are close to the surface. Ground penetrating radar is extremely sen-
sitive but is easily subject to interference. Electromagnetics and
electrical resistivity are often used together to determine the boun-
daries of leachate plumes rather than the location of drums,
although electromagnetics has been used for locating drums. Fre-
quently a combination of geophysical survey methods is recom-
mended. The results of any geophysical survey must be verified by
sampling.
Determining Dram Integrity
Determining drum integrity is one of the most important and dif-
ficult aspects of the cleanup operation. Excavation and subsequent
handling of unsound drums can result in spills and reactions which
may jeopardize worker safety and public health.
The method generally used for determining drum integrity is to
make a visual inspection of the drum surface for corrosion, leaks,
swelling and missing bungs. This approach requires close contact
with the drums, but is the only effective method available for deter-
mining drum integrity. A variety of non-destructive testing
methods (i.e., eddy current, x-rays, ultrasonics) have been in-
vestigated as tools for determining integrity but all have been found
to have serious drawbacks or limitations. Use of ultrasonics or eddy
current, for example, requires that the surface of the drum be
relatively clean and free from chipping paint in order to get ac-
curate readings. Drums which have been buried are frequently
covered with soil, residue and chipping paint and cannot be safety
and easily cleaned. Also, the integrity of the under-side of the drum
cannot be determined with these methods lifting it.
Table 1.
Safety Precautions for Drum Handling Activity
POTENTIAL SAFETY
HAZARD
Unknown location and
contents of drums can
lead lo unsuspected
hazards
Process of visual in-
spections require close
contact with drums of
unknown content
Exposure to toxic/
hazardous vapors: rup-
ture of drums
SAFETY PRECAUTIONS
Locating Drums
•In conducting geophysical surveys, use hand
held instruments rather than vehicle
mounted systems where drums are close to'
the surface
•Conduct soil sampling only after the geo-
physical survey is completed to minimize the
possibility of puncturing drums
•Use non-sparking tools for sampling
•Use direct-reading, air monitoring equip-
ment to detect hot spots where contamina-
tion may pose a risk to worker safety
Determining Drum Integrity
•Any drum which is critically swollen should
not be approached; they should be isolated
using a barricade until the pressure can be
relieved remotely
•Approach drums cautiously, relying on air
monitoring equipment to indicate levels of
hazards which require withdrawal from
working area or use of additional safety
equipment
Drum Excavation and Handling
•Where buried drums are suspected, con-
duct a, geophysical survey before using any
construction equipment to minimize the
possibility of rupture
•Use the drum grappler where possible and
cost effective, to minimize close contact
with the drums
•If the grappler is not available, pump or
overpack drums with poor integrity
•Use non-sparking hand tools and non-
sparking bucket teeth on excavation equip-
ment
•Where slings, yokes or other accessories
must be used, workers should back away
from the work area after attaching the ac-
cessory and before the drum is lifted
•Critically swollen drums should not be
handled until pressure can be relieved
•Use plexiglas shields on vehicle cabs
•Use "morman bars" which fit over the
teeth of excavation buckets to prevent drum
puncture
Release of toxic,
hazardous vapors,
rupture of drums
Mixing of incompatible
wastes
Mixing of incompatible
wastes
•Use direct-reading, air monitoring equip-
ment when in close proximity to drums to
detect any hot spots
Drum Opening
•Use remote drum opening methods where
drums are unsound
•Conduct remotely operated drum opening
from behind a plexiglas shield if backhoe
mounted puncture is being used
•Isolate drum opening from other activities
to minimize a chain reaction if an ex-
plosion or reaction did occur
•Use only non-sparking hand tools if drums
are to be opened manually
•Remotely relieve the pressure of critically
swollen drums before opening
•Clean up spills promptly to minimize mix-
ing of incompatible materials
Staging Recontainerization
•Allow adequate spacing during staging of
drums to provide rapid exit in case of
emergency
•Perform on-site compatibility testing on all
drums
•Drums which are leaking or prone to leak-
age or rupture should be overpacked or the
contents transferred to a new drum before
staging
•Clean up spills promptly to avoid mixing of
incompatible wastes
•Intentional mixing of incompatible wastes
such as acids and bases should be per-
formed under controlled conditions in a re-
action tank where temperature and vapor
release can be monitored
•Monitor for hot spots using direct-reading
air monitoring equipment
Storage and Transportation
•Segregate incompatible wastes using dikes
•Maintain weekly inspection schedule
•Allow adequate aisle space between drums
to allow rapid exit of workers in case of
emergency
•Clean up spills or leaks promptly
•Ensure adherence to DOT regulations re-
garding transport of incompatible wastes
-------
REMEDIAL RESPONSE
171
Table 2.
Measures for Minimizing Environmental Releases During Drum Handling
Potential
Environmental
Problem
Groundwater
Contamination
Surface Water
Contamination
Preventive Measures
•Improve site drainage around the drum hand-
ling area and minimize run-on and run-off by
constructing a system of dikes and trenches
•Where groundwater is an important drinking
water source, it may be necessary to hydrologi-
cally isolate the work area using well-point
dewatering
•Use liners to prevent leaching of spilled ma-
terial into groundwater during drum handling,
drum opening, recontainerization and decon-
tamination
•Use sorbents or vacuum equipment to clean up
spills promptly
•Locate temporary storage area on highest
ground area available; install an impervious
liner in the storage area and a dike around the
perimeter of the area; utilize a sump pump to
promptly remove spills and rainwater from
storage area for proper handling
•Construct dikes around the drum handling
and storage areas
•Construct a holding pond downslope of the
site to contain contaminated run-off
•Use sorbents or vacuum equipment to promply
cleanup spills
•Design the dikes for temporary storage area to
contain a minimum of 10% of total waste
volume; ensure that holding capacity of storage
area is not exceeded by utilizing a sump pump
to promptly remove spills and rainwater for
proper handling
Air Pollution 'Avoid uncontrolled mixing of incompatible
wastes by 1) handling only one drum at a time
during excavation; 2) isolating drum opening
operation from staging and working areas;
3) pumping or overpacking leaking drums and
4) conducting compatibility tests on all drums
•Promptly reseal drums following sampling
•Any drum which is leaking or prone to rup-
ture or leaking, promptly overpack or transfer
the contents to a new drum
•Utilize vacuum units which are equipped with
vapor scrubbers
•Where incompatible wastes are intentionally
mixed (i.e., acids and bases for neutraliza-
tion) in a "compatibility chamber" or tank,
releases of vapors can be minimized by cover-
ing the tank with plastic liner
Fire Protection *Use non-sparking hand tools, drum opening
tools and explosion proof pumps when hand-
ling flammable, explosive or unknown waste
•Avoid uncontrolled mixing of incompatible
waste by 1) handling only one drum at a time,
2) pumping or overpacking drums with poor
integrity, 3) isolating drum opening, and 4)
conducting compatibility testing on all drums
•Use sand or foams to suppress small fires be-
fore they spread
•Avoid storage of explosives or reactive wastes
in the vicinity of buildings
•In a confined area, reduce concentration of
explosives by venting to the atmosphere
•Cover drums which are known to be water re-
active
The safety of workers in close contact with the drums must be
protected by the use of appropriate safety gear and equipment. A
variety of direct-reading, air monitoring equipment is available to
determine radioactivity, oxygen levels, levels of explosives, and
concentrations of toxic gases. This equipment generally has remote
sensing capabilities and alarms to warn field personnel of impen-
ding danger as they approach the drum.
Excavation and On-Site Transfer of Drums
Some waste disposal sites may require certain pre-excavation ac-
tivities to improve site access for heavy equipment, to drain a site to
provide a more stable work surface, or to hydrologically isolate the
site to prevent surface water and groundwater contamination.
Drum excavation is generally accomplished by using a combina-
tion of conventional excavation, lifting and loading equipment
such as backhoes, front end loaders and bobcats, but with special
equipment modifications or accessories adapted to hazardous
waste sites. The most valuable piece of equipment for handling
drums is the barrel grappler (Fig. 1), which is a modified crawler-
mounted backhoe with a rotating grapple head. The grapple attach-
ment can rotate 360° along a given plane and is hydraulically self-
adjusting in grip radius so that it can grab and lift various size con-
tainers as well as containers which are slightly dented or bent.
Other attachments include nonsparking buckets to prevent ex-
plosions, morman bars to cover the teeth of backhoes to avoid
puncturing drums, plexiglas safety shields for vehicle cabs, and
drum lifting attachments such as nylon yokes and metal hoists.
Equipment used in the excavation and staging of drums must be
suitable for digging, grabbing, lifting, loading, and manipulating
drums. Complete handling of a drum related problem usually re-
quires a combination of equipment, particularly where the grappler
is not available.
Figure 1.
Barrel Grappler, removing drums from pit excavation
(Courtesy of O.H. Materials Co., Findlay, Ohio)
When drums are found to be leaking or structurally unsound, the
drum should be overpacked or the contents transferred to a new
container in order to avoid spills or releases which could jeopardize
worker safety. Where the grappler is available, it is generally not
necessary for the worker to be in contact with drums and spills or
rupture of unsound drums may not be a threat to worker safety.
Drum Opening, Sampling and Compatibility Testing
Once a drum has been excavated, a series of activities are under-
taken which lead to final disposal or treatment of the wastes. The
first step is to segregate the drums based on their contents (liquids,
-------
172
REMEDIAL RESPONSE
solids, gases, etc.). Next the drums are opened, sampled to deter-
mine compatibility and resealed. Drum opening and sampling
should be performed in an isolated area in order to minimize the
possibility of explosion and fires if drums spill or rupture. Drum
opening tools include non-sparking hand tools, wrenches and
deheaders and remotely operated plungers, debungers (Fig. 2) and
backhoe attached spikes (Fig. 3.)
When remotely operated methods are used, further protection
should be provided by conducting the operation from behind a
plexiglas shield. Measures should be taken to contain and mitigate
spills which occur during drum opening. In general, remote drum
opening is recommended where drum integrity is poor or the wastes
are suspected of being highly toxic.
Comparability testing should consist of rapid, on-site procedures
for segregating incompatible wastes based on such factors as reac-
tivity, solubility, presence of oxidizers, water content, etc.' The
waste drums containing similar wastes are consolidated in order to
reduce the costs and time involved in cleanup and transportation.
Figure 2.
Pneumatic Bung Wrench: Attachment to drum and remote operation setup
Figure 3.
Backhoe Spike (nonsparking) puncturing drum held by grappler
of O.H. Materials Co., Findlay, Ohio)
Recontainerization or Consolidation
There are a number of options available for consolidating wastes
or for recontainerizing them, if consolidation is not possible. In
general, use of vacuum trucks is recommended where there are t
minimum of 30 to 40 drums which can be consolidated.
Industrial strength vacuum units are available for consolidating
both solids and liquids. Skid mounted vacuum units are available
for areas inaccessible to trucks or where the number of drums to be
consolidated is small. When using vacuum equipment, care must be
taken to avoid incompatible waste reactions which can result in
costly damage to the vacuum cylinder. When incompatible waste
reactions are possible, wastes should be allowed to react under
carefully controlled conditions in a "compatibility" chamber prior
to transfer to the vacuum truck.
If drums containing wastes cannot be consolidated because of in-
compatibility or where the number of drums is too small to make
consolidation economical, they can be overpacked or their contents
transferred to new drums for final disposal. This is frequently a
more costly method of final disposal since drums and overpacks are
bulky to transport. In weighing the relative costs of overpackingor
consolidating wastes using vacuum equipment, the costs of decon-
taminating the vacuum equipment after use should also be includ-
ed.
The final step in the cleanup process is decontamination. Equip-
ment is rinsed in the contaminated work area and the contaminated
rinse water is collected for treatment, if it is determined that it will
contribute to groundwater or surface water contamination. Con-
taminated soils may be combined with sludges and solids and/or
bulked, packed in drums, or treated on-site depending upon.the
volume of soils and the nature and concentration of contaminants.
Empty drums are also decontaminated in some cases depending
upon the nature of the wastes which were stored in them. This pro-
cess may be facilitated by a drum shredder. If the number of empty
drums is small, they can be crushed and overpacked. If the drums
are not considered hazardous they can be crushed and loaded in
bulk onto vans or flat-bed trucks.
Decontamination of field workers, protective clothing and
sampling equipment is an essential pan of the final cleanup pro-
cess. Procedures for decontamination can be found in several
publications including a Hazardous Waste Site Management Plan,'
and the EPA Safety Manual for Hazardous Waste Site Investiga-
tions.'
Storage and Shipping
In some instances it may be necessary to store drums on-site tem-
porarily until funds become available for final treatment/disposal
or until a suitable disposal site is found. If drums are to remain on-
site for a period of three or more months, then requirements for
RCRA permitted facilities should be followed. These requirements
include the following:
•Use of dikes or berms to enclose the storage area and to segregate
incompatible waste types
•Installation of a base or liner which is impermeable to spills
•Sizing of each storage area (containing compatible wastes) so
that it is adequate to contain at least 10% of the total waste
volume in the event of a spill
•Design of the storage area so that drums are not in contact with
rainwater or spills for more than 1 hour
•Weekly inspections
Transportation and final disposal of wastes from abandoned sites is
governed by Department of Transportation regulations as well as
by state laws, disposal facility requirements, and in the case of
highly toxic materials, by the Toxic Substance Control Act.
CONCLUSIONS
The selection and implementation of equipment for drum handl-
ing is made on a case by case basis based upon the following con-
siderations:
-------
REMEDIAL RESPONSE
173
•Safety of the field workers
•Protection from environmental releases which could jeopardize
public health
•Site-specific factors which effect equipment performance
•Costs
There are a number of precautions which should be taken to pro-
tect field personnel. Many of these measures such as use of protec-
tive clothing, medical surveillance, communications, etc. apply to
all waste investigations and cleanup operations and are not unique
to drum handling. Other safety measures, such as use of non-
sparking drum opening tools, or isolation of the drum opening
operation are unique to drum handling.
In addition, precautions must be taken to minimize en-
vironmental releases. These include both measures for preventing
spills, leaks and incompatible reactions as well as mitigative
measures for containing and controlling spills or reactions which do
occur.
Site specific variables such as accessibility of site, the number of
drums, site drainage, drum integrity, etc., effect equipment perfor-
mance and safety, and must be considered in selecting and im-
plementing drum handling methods (Table 3).
Table 3.
Major Site Specific Variables Influencing Drum Handling Activities
DRUM LOCATION
Large Number o f Drums vs
Few Drums
Expenditure for remote sensing must be
kept in perspective; If the number of
buried drums is small, use a simple
remote sens ing tool, (i.e., metal
detector), to locate drums; may noc
be worth the expense to quantify; hand-
held tools are suitable for small sites
whereas vehicIe-towed equipment is
socnetimes needed at large sites
Remote Site \/_s Accessible
or Populated Sice
Remote site may require that geographical
surveying be done manually rather than bv
vehicle; clearing and grubbing may be
needed to conduct continued surveys
High Water Table v
Low Water Table
Presence of a hLgh water cable mav prevent
use of a vehicle Co tow remote sensing
equi pment ; i £ groundwacer is sa I me ,
ground penetrating radar is I ike Iy CD be
unsuitable; interpretation of electro-
magnet LCS data can be difficult if water
table is low
DRUM INTEGRITY
Method of determining drum integrity
depends more on type of waste than the
ac tual number of drums
[f the site is in urban area, air
monitoring needs may be more
intensive to assure populat ion safety
as well as worker safety; accessibility
of lab to do quantitative analysis
should be considered prior to under-
taking quantitative sampling, espe-
cially if the site is remote
Drum integrity is likely to be poor in high
groundwater areas; monitoring for drum
integrity may be more intensive in area of
high water table depend ing upon use of
groundwater for drinking; if poor integrity
is suspected in high groundwater area it
may be necessary to improve site drainage
or to hydrologically isolate the site
before handling drums
DRUM OPENING
Use highly mobile, high production
equipment with large bucket capacities
for large numbers of drums: backhoes
grapp lers, rubber-tired loaders; use
severa 1 equipment vehicle combinations
for larger sites with over 500-1,000
drums to be removed; smaller sites
(<500 drums) will require less of a
variety of excavation vehicles; small
vehicles such as front end loaders are
economical for small sites
Remote sites may require special site
preparation, such as clearing and
grubbing for access road construction;
may dictate transport of Eewer, larger
equipment types (backhoes, grappiers,
crawler tractors, cranes) to disposal
site
For readily accessible sites, equip-
ment size and number not limited;
favors mobile rubber-tired vehicles
(Bobcats and backhoea)
[f congested, urban site, may need
smaller machinery (forklifts and
backhoes vs. cranes for I ift ing and
transfer; may also use hoists or
slings to Lift drums from congested
areas
Large number (>500 ): backhoe plunger
or remote conveyor should be considered
Sma11 number: hand tools should be
considered depending on conditions of
drums and drum contents
Congested site favors manuaI or
portable remote opening
Water-logged sices may require surface
runoff diversion with trenches and berms ,
and subsurface hydro Logic isolation with
wellpoint pumping or slurry-trench cut-off
wal Is; wet muddy sites favor crawler-
mounted vehicles vs, rubber-t ired; swamp
pads (extra-wide crawler tracks) and timber
mats may also be useful; for dry sites,
less site preparation is needed and mobile,
rubbe-tired vehicles can be used
Uppn drums in control ted area with
secondary containment d ikes or berms and
I iners
RECONTAINERIZATION
AND
CONSOLIDATION
equipment suited to size of problem,
keeping in mind possible incompacibi1ity
Few drums: more prone to use overpacks,
or smalI skid mounted units; large
numbers of drums: vacuum trucks are more
•suitable for compatible wastes
If site ia inaccessible, may need to
uae skid mounted units rather than
vacuum trucks; if site is congested,
drums should be staged, opened and
conso L idated in sma 11 groups to
prov ide adequate spacing; if site is
remote and wastes are biodegradable,
contaminated soils couH conceivably
be left on site (determine on case
by case basis); detonation of lab
packs shou Id not be considered in
populated or congested areas
Genera IIv does not affect reconcainer-
ization, unless the presence of a high
water table prevents access by vacuum
trucks
STORAGE/SHIPPING DOT regulations are more concerned with
the typ*»s of wastes rather than the
quantity; large numbers: consolidation
of compatible wastes for more economical
shipping; few drums: usually over-pack
or transfer to a new drum before shipping
Storage area should be as distant
as possible from populated areas;
rcactives and explosives should be
stored away from buildings
Tn areas with high water cable, storage
area should be placed on highest ground
pegs ible
-------
174
REMEDIAL RESPONSE
The final factor affecting the selection of equipment is costs.
Because the number of drums, their contents and integrity are
generally now known with any certainty at the outset of the cleanup
operation, costs for cleanup can only be approximated. Costs can
be minimized however, by the selection of the most cost-effective
contractor and equipment types and by close management of the
cleanup operation.
ACKNOWLEDGEMENTS
This project was sponsored by the USEPA's Municipal En-
vironmental Research Laboratory, Cincinnati, OH, Oil and Haz-
ardous Materials Spill Branch, Edison, NJ, under contract No.
68-03-3113. The content of this publication does not necessarily
reflect the views or policies of the USEPA, nor does mention of
trade names, commercial products or organizations imply endorse-
ment or recommendation by the US Government or JRB
Associates.
JRB Associates expresses their appreciation to Mr. Anthony
Tafuri and Mr. Frank Freestone, USEPA Oil and Hazardous
Materials Spills Branch for providing technical direction for this
project. Technical assistance from James Walker and Robert Pan-
ning, OH Materials, Findlay, OH; William Bloedorne, Wizard
Drum Tools, Milwaukee, WI; Gregory Heath, Peabody Clean In-
dustries, East Boston, MA; and Joseph Mayhew, Chemical Manu-
facturers Association, Washington, D.C. is also gratefully
acknowledged.
REFERENCES
1. Neely, N. et al., "Remedial Actions at Hazardous Waste Sites. Survey
and Case Studies." EPA 430-9-81-005. SCS Engineers, Covington, KY
for USEPA Municipal Environmental Research Laboratory, Cincin-
nati, OH. 1981.
2. Blackman, W.C., Jr., et al. "Enforcement and Safety Procedures for
Evaluation of Hazardous Waste Sites." Proc., National Conference on
Management of Uncontrolled Hazardous Waste Sites, Washington,
D.C., Oct., 1980, 91.
3. Chemical Manufacturers Association, "Hazardous Waste Site Man-
agement Plan." Washington, D.C. 1981.
4. USEPA. "Safety Manual for Hazardous Waste Site Investigations."
Office of Occupational Health and Safety and the National Enforce-
ment Investigation Center, 1979.
-------
GEOTECHNICAL ASPECTS OF THE DESIGN AND
CONSTRUCTION OF WASTE CONTAINMENT SYSTEMS
JEFFREY C. EVANS
Woodward-Clyde Consultants
Plymouth Meeting, Pennsylvania
HSAI-YANG FANG, Ph.D.
Geotechnical Engineering Division, Lehigh University
Bethlehem, Pennsylvania
INTRODUCTION
Disposal of hazardous and toxic wastes in the subsurface en-
vironment has resulted in the widespread application of geotech-
nology to the design and construction of waste containment sys-
tems. The adaptation of conventional passive groundwater and
surface water barriers to waste containment requires certain spe-
cial considerations. In this paper, the authors present a systematic
engineering approach to passive containment alternatives designed
to mitigate contaminant migration. Certain design and construc-
tion aspects of the application of geotechnology to these systems
are also present.
The three major passive containment components which are
covered in detail are: 1) top seals, including covers of native clay,
processed clay and polymeric (synthetic) membranes, 2) barrier
walls, including soil-bentonite slurry trench walls, cement-ben-
tonite slurry trench walls, and vibrating beam cutoff walls, and
3) bottom seals, including liners of native clay, processed clay, and
polymeric membranes. Emphasis is placed on the geotechnical
aspects of design and construction practices of waste encapsula-
tion, including advantages and limitations. The functions of each
of these components are defined. The design process is then de-
tailed with emphasis on the difference between conventional ap-
plications versus waste containment applications (for example,
slurry trench cutoff walls for waste containment must incorpor-
ate additional considerations in design as compared to their use
to control groundwater during excavation).
Many site investigations are conducted by hydrogeologists to
obtain data for assessment of contaminant migration. Subse-
quently, waste containment systems may be engineered as part of
a pollution abatement program. It is essential that the geotech-
nical engineering information required be obtained during the site
investigation phase in order to allow for a thorough, yet cost-
effective, engineering design. In this paper, the authors present
a guide to the engineering data required to enable geologists to
plan and conduct subsurface investigations which will provide geo-
technical as well as hydrogeological data for design.
TYPES OF FACILITIES
Most waste disposal facilities can be classified (Table 1) in one
of three categories: 1) past disposal sites referred to as abandoned,
Table 1.
Waste Disposal Facility Classification.
Facility
Category
I
III
Facility
Description
Past Disposal site
Active Disposal
Future Disposal
Generic
Names
abandoned
inactive
retired
midnight dump
uncontrolled site
orphaned
secure landfill
sanitary landfill
waste treatment
complex
recycling facility
Control of
Wastes Disposed
Little to none
Some
Well-controlled
inactive, retired, or uncontrolled, 2) existing active disposal sites
such as sanitary or hazardous waste secure landfills, and 3) future
disposal sites which could include waste treatment, recycling and
disposal complexes. In this paper, the authors will limit the dis-
cussion to waste containment at past disposal sites. However, much
of the geotechnology presented may be applicable to new site de-
sign and waste containment at active disposal sites. However, the
engineer's ability to control the type of wastes for which waste
containment systems must be designed and constructed to con-
tain is limited.
ENGINEERING APPROACH TO WASTE CONTAINMENT
The application of geotechnology to waste disposal requires an
orderly, systematic approach which will permit the engineer to
fully assess both the site and subsurface conditions and evaluate
the applicability of containment alternatives. The project approach
described below is for past uncontrolled disposal sites. The follow-
ing four steps are required for each project and, although they
are presented chronologically, there is a continual reassessment
of the previous steps as the project progresses and more data be-
comes available:
•Review existing information including historical site data, geo-
logic data, and groundwater data: It is necessary to obtain as
much information as possible on the types of waste disposed,
the timetable of waste disposal, and the previous disposal prac-
tices (i.e., drums, solid waste, lagoons). Air photos, which are
generally readily available, can be extremely useful. Information
regarding the subsurface conditions can be obtained from pre-
vious records and borings, from the site construction history,
and from regional geologic information. Forming a conceptual
model of the site and subsurface conditions at this time will en-
able the engineer to better plan and conduct site and subsurface
investigations.
•Assess in detail existing site conditions, which include geologic
conditions, groundwater conditions, and contamination distri-
bution: Field work will probably be required at this state. The
use of geophysical tools prior to test borings or monitoring well
installations can provide valuable insight into the subsurface
conditions' and will probably result in a more complete and cost
effective boring and sampling program.
•Quantify site conditions, including the direction, volume and
velocity of groundwater flow, the interaction of groundwater
with surface water, the distribution of contamination in the
groundwater system and the contaminant loading: The degree of
spphistication of this quantification phase may vary from a sim-
ple model to a complex computer model.
•Develop the containment/treatment program: the portion of the
program where the application of geotechnology receives the
major emphasis.
Many subsurface investigations are conducted primarily as hy-
drogeologic investigations. Geotechnical physical property test-
ing (Atterberg limit, grain size distribution, water content) may
not normally be part of the routine investigation. If remedial ac-
tion is anticipated, it may be desirable to have geotechnical en-
gineering input during the site investigation phase to avoid future
175
-------
176
BARRIERS
data gaps. Thus testing could include physical property tests as well
as engineering property tests for strength, compressibility and
permeability.
CONTAINMENT ALTERNATIVES
A waste containment system can consist of many components
(Fig. 1.). These components are classified in two general cate-
gories: (1) active and, (2) passive. Active components of a con-
tainment system are those which require ongoing energy input.
Examples of active components are pumping wells, disposal wells,
and treatment plants. Conversely, passive components of a con-
tainment system are those which do not require ongoing energy
input. Examples of passive components include drain tile collec-
tion systems, liners, covers and varrier walls. Utilizing quantified
site conditions, the geotechnical engineer should try to provide
components are required and/or feasible. It may also be necessary
to develop alternates for every component in the system.
Finally, as required in most geotechnical engineering practices,
field observations must confirm design assumptions. Construc-
tion inspection is critical for waste containment systems.
COMPONENTS OF PASSIVE REMEDIAL SYSTEMS
Important details regarding each of the major components of
passive waste containment systems and how geotechnical engi-
neering technology is required for their design and construction
are discussed below. Further discussion in this paper is limited to
passive components of waste containment systems.
Top Seals (Covers)
The primary function of a top seal is generally to control water
PUMPING WELLS
PUMPING TROUGH
PUMPING RIDGE
DISPOSAL WELLS
INJECTION WELLS
WASTE CONTAINMENT
SYSTEM
ACTIVE COMPONENTS
ACTIVATED CARBON
SPECIALTY PROCESSES
PASSIVE COMPONENTS
Figure 1.
Waste Containment Alternatives
DRAIN TILE COLLECTION
SYSTEM
FRENCH DRAINS
TOP SEALS
NATURAL CLAY CAPS
PROCESSED CLAY CAPS
POLYMERIC MEMBRANE CAPS
BARRIER WALLS
SOIL-BENTONITE CUTOFF WALLS
CEMENT BENTONITE WALLS
VIBRATING BEAM CUTOFF WALLS
BOTTOM SEALS
NATURAL CLAY LINERS
PROCESSED CLAY LINERS
POLYMERIC MEMBRANE LINERS
conclusions, recommendations, and design criteria to maximize the
use of passive components of the containment program and to
minimize the need for active components.
However, even passive components may require maintenance.
For example, it is frequently necessary to keep vegetation roots
from penetrating a clay cap. A passive system is generally there-
fore not a maintenance-free system.
In developing waste containment alternatives, one must bal-
ance the design approach for the site considering the magnitude
and extent of the contamination problem, the available active
and passive components, both the short and long term containment
effectiveness, capital and operating costs, and the eventual termi-
nation of the system. After the development of containment al-
ternatives, geotechnical engineers are often required to develop
final design criteria. During the site quantification program it is
necessary to make basic decisions as to which of the possible major
movement so as to maximize precipitation runoff and to minimize
infiltration and subsequently leachate production by the land dis-
posal facility. Geotechnical aspects of top sealing utilizing native
clays, processed clays, and polymeric membrances are discussed
in following subsections. Geotechnical input required for covers of
bituminous and portland cements, tars, liquid and emulsified as-
phalts and bituminous fabrics are not discussed. For other aspects
of cover design and construction, two manuals prepared by th*
U.S. Army Engineer Waterways Experiment Station are ref-
erenced.2'3
The USEPA states "Because clays will generally last longer than
synthetic materials, clay caps rather than synthetic caps should
usually be the materials chosen...'" However, to avoid the bathtub
effect, wherein more water enters the facility than can drain out,
"...will require the installation of a synthetic membrane cap when-
ever the bottom liner is synthetic."4 Clear guidance is therefore
-------
BARRIERS
177
provided regarding the choice of cover materials.
Native Clay Top Seals: In general, the most cost effective top
sealing can usually be completed utilizing native clay materials
from local borrow sources. In many areas, local sources are avail-
able to provide compactible clays of relatively low permeability.
If suitable clays are not available, the obvious decision would be
to use another of the top sealing alternatives.
In testing clay covers, one must consider the natural variability
in the material and the test method in order to provide an accurate
representation of the hydraulic conductivity. Permeability should
be determined in geotechnical laboratories utilizing permeability
tests conducted in a triaxial cell where the stress and gradient con-
ditions can be closely controlled. Since clay caps are designed to
minimize precipitation infiltration, it is normally not necessary to
utilize anything other than tap water as a permeant. Distilled
water is not recommended.5
The thickness of a clay top seal is generally a minimum of 18 in.
to 2 ft, based upon the following considerations. First, the top sev-
eral inches of the clay cap cannot generally be as well compacted
as the thickness remainder. Further, it may be difficult in the
long term to maintain the clay density in the top few inches due to
potential desiccation, cracking and frost action. Finally, the bot-
tom of the clay cap may become somewhat contaminated with
the subgrade material during installation. Therefore, the "effec-
tive thickness" of a nominal 2 ft cap may be on the order of 1 ft.
Native clay caps must be protected from erosion due to rain
water, cracking due to drying, differential subgrade movement,
penetration by deep tap roots of vegetation, and rutting due to
traffic. In virtually all cases, the clay cap should be covered by a
minimum of 6 to 12 in. of topsoil. Conventional hydrologic
studies to determine erosion potential are required to prevent
gullying and erosion of the clay top seal and topsoil. A typical sec-
tion illustrating the use of a native clay cap is found in Fig. 2a.
A clay top seal can generally be placed on a slope which is not
steeper than about two horizontal to one vertical. It is usually
preferable to utilize flatter slopes such as three or four hori-
zontal to one vertical. The flatter slope more readily permits
compaction and reduces the risk of local slumping. A geotech-
nical analysis of slope stability may be required to assess the prob-
ability of instability. The thickness of the clay on side slopes may
be slightly less than the thickness on top of the area. The larger
side slope gradients enhance runoff greatly reducing the time avail-
able for precipitation to percolate downward into the waste con-
tainment area.
Construction inspection of clay top seals is particularly im-
portant. If the top seal is not covered soon after the clay is placed
and compacted, precipitation can result in erosion of the clay top
seal. Control of the cover thickness can be by accurate pre- and
post-placement surveys or by probing during placement.
Processed Clay Top Seals: If natural clay is not available of
sufficient quality, quantity, or price, alternate capping materials
must be considered. Processed clay has become a common alterna-
tive to natural clay caps. Processed clay is typically a sodium mont-
morillonite from Wyoming, commonly called bentonite.
The construction of a processed clay cap .requires: 1) the appli-
cation of the bentonite at a controlled rate (e.g., 2 lb/ft2, 2) ade-
quate mixing with in-place soils into a predetermined loose lift
thickness, 3) compaction of the cap, and; 4) hydration of the ben-
tonite. A 4 to 6 in. layer of clayey soil with a low hydraulic con-
ductivity, generally around 1x10-7 cm/sec, is thereby formed.
This cap is then subsequently covered to prevent desiccation and
erosion as would be a cap of natural occurring clay.
The principal advantage of the processed clay cap is the relative-
ly low hydraulic conductivity which can be achieved. The prin-
cipal disadvantage relates to controlling a uniform application
rate. Other disadvantages may be cost, depending upon the avail-
ability of naturally occurring clays, and the fact that since the
seals are so thin there is little room for error.
The effectiveness of a processed clay top seal in reducing infil-
» tration requires design and construction control of the subgrade
I'lTYPICAL) VEGETATED FINAL COVER
2'(TYPICALI NATIVE CLAY
I i IO'Tcm /itc.
1.5'fTYPICAL) NATIVE
CLAY
K= I x I0'scm /stc.
Figure 2a.
Typical Section: Native Clay Top Seal
VEGETATED FINAL
COVER
POLYMERIC MEMBRANE SEAL
HEGRAOED FILL
BEDDING MATERIAL
DIVERSION
DITCH
Figure 2b.
Typical Section: Polymeric Membrane Topseal
materials, subgrade preparation, application rate, application uni-
formity, mixing, compaction, hydration and cover. Deficiencies in
any one of these aspects can result in reduced top seal effective-
ness.
The subgrade materials should be uniform and free of roots,
sticks, cobbles, or other miscellaneous debris which would preclude
a homogeneous blend for the specified seal thickness. Further, the
material must have water content and material characteristics
which will not impede the uniform blending of the processed clay
throughout the soil matrix. This moisture content should be near
optimum for the soil-bentonite mixture to aid in the subsequent
compaction. Compaction studies in a geotechnical laboratory are
therefore generally required to establish the moisture-density re-
lationship for the soil-bentonite mixture. The subgrade must be
sloped in such a way as to provide positive drainage characteris-
tics and preclude ponding of precipitation upon the seal. The sta-
bility of the processed clay seal must also be considered. Slopes
which are steeper than three horizontal to one vertical are not gen-
erally recommended. Steep slopes increase the risk of slumping.
The application rate is generally determined by consideration of
the desired coefficient of permeability and the available subgrade
materials. For a given coefficient of permeability, the application
rate will vary with the soil type proposed for use in the top seal.
Laboratory studies can be undertaken to investigate the relation-
ships between coefficient of permeability, application rate, soil
type, and degree of compaction. The specification can then be op-
timized with respect to these variables.
Field control of the application rate is essential if the top seal is
to perform as predicted by the engineering studies. Three methods
are generally employed to apply the powdered processed clay.
The first involves the utilization of an agricultural-type lime spread-
er. The second involves the use of a pressurized container and dis-
tributor (Fig. 3). The third method uses hand-spreading.
With either of the first two methods the application rate can be
checked by placing a relatively flat container (such as a trimmed
cardboard, or a tarpaulin) beneath the spreader as it passes. The
weight of the material deposited in the container can then be de-
termined. The application rate is then computed as a weight per
unit area. An additional check can be made by determining the
total bentonite used for the total area treated. This results in the
average application rate. Application rates should be checked
frequently if mechanical applicators are employed.
-------
178
BARRIERS
Figure 3a.
Pressurized Tank and Distributor
Figure 3b.
Close-up of Distributor
Figure 3.
Application of Processed Clay
The processed clay can also be applied by hand-spreading the
material from bags in pre-marked grid squares. The bags are
broken open and the material is raked into an "uniform" thick-
ness across the grid. Hand-spreading may be more costly, but
tighter control can generally be maintained. The probability of
zones with inadequate application rates can be reduced with hand-
spreading as compared with mechanical equipment application.
As soon as possible after the application of the processed clay
is completed, the material is thoroughly mixed with the subgrade.
Mixing should result in a uniform blend of processed clay and sub-
grade soils for a top seal of specified thickness. Adjustable rotary
tillers appear to provide a positive means to control top seal thick-
ness and homogeneity. These tillers can blend the soil and pro-
cessed clay with depth control devices, and can result in a fairly
uniform layer thickness. Agricultural disks, graders, and other
equipment have also been used but control is more difficult.
Immediately after blending, the processed clay top seal is com-
pacted to the minimum density determined during the design
studies. Compaction is generally with smooth-drummed or pneu-
matic rollers. Sheepsfoot or padfoot rollers are not permitted.
Hydration of the processed clay cap is usually from infiltrating
precipitation. Should the site hydrology characteristics show then
is a potential for contaminated water to reach the top seal, the cap
should be prehydrated with uncontaminated water.
Cover of processed clay top seals to prevent desiccation and ero-
sion is required, as with top seals of native clays. The cover thick-
ness required for processed clay seals is generally greater than that
required for native clay caps. This is because there is little margin
for disturbance without jeopardizing the seal integrity.
Polymeric Membrane Top Seals: Polymeric membranes can be
utilized as top seals for waste containment.' In the past, these
have only received limited usage due to their relatively high cost.
Polymeric membranes are available in a wide range of material
types from numerous manufacturers and distributors. Examples of
polymeric membranes include polyethylene, chlorinated polyethy-
lene, chlorosulfonated polyethlene, polyvinyl chloride, butyl rub-
ber, and neoprene. This paper is limited to discussion of the geo-
technical aspects of polymeric membrane usage. Additional in-
formation regarding the advantages and disadvantages of the avail-
able membranes can be obtained from cited references.1'8
From a geotechnical standpoint, the most important aspects of
utilizing a polymeric membrane as a top seal are the subgrade ma-
terial, subgrade preparation, slope and final cover (Fig. 26). Equal-
ly important to these geotechnical aspects is the membrane design,
placement procedures, field joining of seams, and field testing of
the membrane and field welds, all of which are beyond the scope
of this paper.
The subgrade soils must be free of materials which could punc-
ture the membrane top seal including sticks, large stones, and mis-
cellaneous debris. Further, the subgrade should be graded and
compacted to provide for runoff and to prevent liquid ponding.
The final cover is typically approximately 18 in. thick and vegetated
to minimize erosion. The cover is necessary to protect the mem-
brane from trafficking, and ultraviolet degradation. Careful con-
sideration of slide slopes is required to preclude sliding of cover ma-
terials along the membrane top seal. As precipitation infiltrates
the cover and flows along the membrane, a saturated and weak-
ened zone in the cover material may develop causing a slump of
the cover material.
Barrier Walls
The containment of contaminant migration from existing dis-
posal sites or impoundments frequently necessitates some sort of a
subsurface barrier to horizontal groundwater flow. Barrier walls
are presently typically constructed as soil-bentonite slurry trench
cutoff walls, cement-bentonite slurry trench cutoff walls, and vi-
brating beam cutoff walls. For barrier walls to be effective, they
must generally key into an impermeable stratum of natural mater-
ials beneath the site. However, this requirement is not always essen-
tial, depending upon the hydrogeologic conditions.
Soil-Bentonite Slurry Trench Cutoff Walls: The method of con-
structing soil-bentonite slurry trench cutoff walls is well docu-
mented. •10 A trench is excavated below the ground surface and
trench stability maintained utilizing a slurry of bentonite and water.
This slurry maintains trench stability in much the same way at a
drilling fluid maintains borehole stability. The bentonite-water
slurry is designed by the geotechnical engineer to have certain den-
sity, viscosity, and filtrate loss properties which allow for the for-
mation of a filter cake along the walls of the trench and which re-
sults in a computed factor of safety for trench stability greater
than 1.0.
Trench depths are generally limited to about 35 ft using conven-
tional backhoes. In order to achieve greater depths, a modified
-------
BARRIERS
179
dipper stick is required, which can usually be provided by specialty
slurry wall contractors. To go deeper than 53 ft usually requires
the utilization of a clamshell. An extended stick backhoe capable
of excavating to depth of 73 ft has also been developed.'' Clam-
shell digging is typically slower, and can increase the cost of the
barrier wall.
Once the trench is excavated, the result is a trench filled with
bentonite-water slurry. It is then necessary to backfill the trench.
The backfill normally consists of a matrix of a material with or
without natural fines, which is mixed with the bontonite-water
slurry. The backfill materials are mixed and generally controlled on
the basis of slump. The backfill is usually mixed to a consistency
of high-slump concrete, and then the trench is backfilled. Care
must be taken to achieve a uniform mixing of the backfill, and to
avoid entrapment of pockets of pure bentonite-water slurry in the
trench. A schematic of the excavating and backfill procedure is
shown in Fig. 4 while a typical section of the completed soil-
bentonite cutoff wall is shown in Fig. 5.
BACKFILL
PLACED -
HERE
Figure 4.
Schematic Section of Slurry Trench Excavation and Backfill
- 3' .1 L 3' .1
TER
SLE >
AOUIFER
'////*
xAOUICLUDE /
/AOUITARD /
'/ / / '
TYPICAL
S7
V
,. —
L -*
WATER ,
-SLURRY LEVEL TA8LE>
— BENTONITE -WATER
SLURRY
15% TYPICAL)
— -Fl LTER CAKE
// / ' / /// / /
/ AOUICLUDE/
/ - / '/AOUITARD
'•-' •••'// ////A
TYPICAL
f^
AOUI PER
— SOIL-BENTONITE
BACKFILL
1-2% BENTONITE
[TYPICAL I
— FILTER CAKE
' / / / / .
_,. / '
/ , ,
0) DURING EXCAVATION
blAFTER EXCAVATION
Figure 5.
Typical Section: Soil-Bentonite Slurry Trench Cutoff Wall
Much has been written regarding the design of soil-bentonite
cutoff walls for conventional groundwater control applica-
tions.9'10'11'12'13'14 In order to design soil-bentonite slurry walls for
waste containment, studies are required beyond those normally re-
quired for other applications. Chemical analysis of the samples of
on-site materials considered as potential backfill materials should
be considered. If on-site materials are contaminated consideration
of the use of off-site borrow areas to provide the backfill ma-
terials must be made.
Once the potential source or sources of backfill materials is
identified, waste compatibility testing is typically conducted.
Several bentonites are available which are identified as being con-
taminant resistance. Contamination resistance is a relative term and
it should be recognized that treated bentonite is not totally contam-
inant resistant to all contaminants at all concentrations. Contam-
inated soils may inhibit or reverse the hydration of the benton-
ite. Conversely, backfill mixed with contaminated soils may be sub-
ject to smaller property changes when subjected to pollutants than
backfills mixed with uncontaminated soils.9
Upon selection of a backfill source or sources, the geotechnical
design studies are typically undertaken. These studies should in-
clude, as a minimum, analysis of trench stability, backfill mix de-
sign, waste compatibility, and the site-specific subsurface con-
ditions. From these studies the project specifications can be pre-
pared. The question of compatibility between the waste and the
cutoff wall is approached in two steps. First, the nature of the
liquid waste or leachate must be characterized. A review of the
published information investigating the pore fluid effects upon clay
behavior9'15'16>17can then be made to allow a preliminary assess-
ment of the compatibility. After the preliminary compatibility
assessment is made, a laboratory testing program can be designed
and conducted to provide site-specific waste compatibility data.
Once the waste capability program has been developed, samples
of backfill need to be created and tested in the laboratory utiliz-
ing the bentonites under consideration. The most important of
these laboratory tests is the triaxial permeability test.18' 9 The test
is typically conducted ultimately utilizing the site leachate, liquid
waste or groundwater as the final permeant. The samples are first
set up in a triaxial cell, consolidated, and permeated with water.
Initial hydration of the bentonite must be with the water planned
for use during construction, and not distilled water or water from
some other source. Off-site water may be necessary if the only
available on-site water is contaminated so as to preclude adequate
hydration of the bentonite. The bentonite utilized in the testing
program must also be that planned for project usage. Permeabil-
ity versus time and volume change versus time can then be calcu-
lated from the test data.
The tests are conducted to determine the change in permeabil-
ity in response to the waste liquid. Typical results are presented in
Fig. 6. The data are evaluated in terms of pore volume displace-
PORE VOLUME DISPLACEMENT (PVO)
2.0
NOTES: I. SAMPLE PREPARED WITH SAHO AND It
8ENTONITE (AMERICAN COLLOID SALINE SEAL)
I. SAMPLE PERMEATED KITH SITE GROUNO«ATES
s m
12000 I4OOO
ELAPSED TIME (MINUTES)
PORE VOLUME DISPLACEMENT (PVD)
2.0 3.0 4.0 3.Q
= NOTES: I. SAMP
^"' IS SEHTONITE (AMERICAN COLLOID PREMIUM GEL) zfJLtZ
(» * l.O i 10 em/Me h
00 12000 I6OOG
ELAPSED TIME (MINUTES)
Figure 6.
Permeability Test Results
-------
180
BARRIERS
ment as well as time. Pore volume displacement appears to provide
an acceptable means at this time to account for "time" effects of
the waste upon the material. Generally, three to five pore volumes
are required to establish equilibrium for these particular tests, no
degradation in permeability was observed with either untreated
or treated bentonite. Bentonite suppliers often conduct perme-
ability tests free of charge to clients that do not have suitable lab-
oratory capabilities. These tests are conducted in fixed wall
permeameters as shown in Fig. 7. A complete discussion of perme-
ability testing to determine the permeant affect on fine-grained
soils/18'19
When applying cutoff wall geotechnology to waste contain-
ment, laboratory verification of the slurry and backfill prop-
erties is essential prior to the start of construction. Shown in
Fig. 8 is the effect of bentonite content and type on slurry vis-
cosity for a given mixing water. The Marsh viscosity in seconds is
Pressure C-augs
Plate Cap
Gasket
1" x 2V Glass
Hippie
Gasket
Leachate Solution
1" x H" 3eli
Seducer
Gasket
" x 6" Nipple
Sealed Soil
wall Seal - Daarp
•Volclay
Permeable Sand Ease
Gasket
Mesh
Perforated Plate
Bottcn
Figure 7.
Fixed Wall Permeameter ( from American Colloid Co.)
• SG-40
o PREMIUM GEL
a SAL IDE SEAL
BENTONITE CONTENT (S)
Figure 8.
Marsh Viscosity vs. Bentonite Content
an indirect measure of the hydrated viscosity characteristics of the
bentonite. Note that as bentonite content increases the Marsh
viscosity increases (i.e., the slurry became thicker). Also note that
the Marsh viscosity varies at a given bentonite content as the type
of bentonite varies. The bentonite for these tests was hydrated
utilizing water proposed for use during cutoff wall construction.
These examples are presented to show that the application of
cutoff wall technology to waste containment requires analysis of
the interaction between the bentonite, mixing water, backfill, and
site groundwater/leachate. The long-term performance of a soil-
bentonite cutoff wall for waste containment must be fully inves-
tigated.
Should a soil-bentonite slurry trench cutoff wall prove to be
practicable, detailed construction specifications must be written,
These specifications must include the source of the mixing water,
required hydration time or slurry properties, allowable methods of
mixing the bentonite slurry and the backfill, the bentonite-water
viscosity and density limits, approved sources of the backfill, and
allowable methods of backfill placement. An excellent treatment
of slurry wall specifications has been written by Millet and
Perez."
Close construction control is required to ensure construction
consistent with design assumptions and intent. A resident geotech-
nical engineer should document depth to key material, test the
backfill and slurry, and provide on-site technical representation
for the owner/engineer. Consideration must be given to the dis-
posal of excavated soils should they be categorized as contami-
nated. Finally, proper planning for work safety is essential should
a barrier wall be planned for containment of previously disposed
wastes.
Cement-Bentonite Slurry Trench Cutoff Walls: As an alterna-
tive to soil-bentonite cut-off walls, cement-bentonite cutoff walls
can be utilized. The trenches are excavated in a manner similar
to soil-bentonite walls utilizing a slurry to maintain trench sta-
bility. However, in contrast to a soil-bentonite wall, the slurry
consists of cement in addition to water and bentonite, and no
backfill is added. The slurry is left in the trench and allowed to
harden. A strength equivalent to stiff to very stiff clay can be ob-
tained with the cement-bentonite slurry wall after a period of a
month or so. Design considerations include the cement and ben-
tonite content and type, and their relationship to the strength and
permeability of the backfill. An example of the laboratory test
-------
BARRIERS
181
results investigating the strength of various mix designs is pre-
sented in Fig. 9.
Leachate compatibility tests must be conducted utilizing the site
pollutant as permeant. The overall permeability of a cement-
bentonite cutoff wall is generally higher than for soil-bentonite
walls. Close control again must be given to the source of the mix-
ing water.
Vibrating Beam Slurry Walls: Barriers to horizontal ground-
water flow and contaminant migration have been designed and
construction using the vibrating beam injection method (Fig. 10).
This technique utilizes a vibratory-type pile driver to cause the
penetration of a beam of specified dimensions to the design depth.
Slurry is added through injection nozzles as the beam pene-
trates the subsurface soils and as the beam is withdrawn.
The slurry utilized with the vibrating beam technique is gen-
erally either of two types, cement-bentonite, or bituminous grout.
Mix design considerations for cement-bentonite were previously
discussed in this paper. Bituminous grouts are prepared as a homo-
geneous blend of asphalt emulsion, sand, portland cement and
water. Flyash may also be included. It is reported that this bitum-
inous grout can resist strong acids and high saline content wastes.
The engineer must be aware of the detailed aspects of thin slurry
wall barriers installed by the vibrating beam technique in order to
assure an adequate waste containment design and installation.
The specification may include the slurry mix design, installation
equipment requirements, batch mixing equipment requirements,
verticality limits, injection pressure, overlap, depth, injection and
extraction rates and procedures, and wall thickness.
Control of the beam tip location cannot be guaranteed, par-
ticularly with deep penetrations. For example, the presence of
cobbles or boulders may cause a deflection of the tip. As in the
case of conventional slurry walls, compatibility testing is neces-
sary to investigate the slurry resistance to the contaminant being
contained. A principal advantage of this technique is the elim-
ination of the need to excavate potentially contaminated soils,
possibly an important safety consideration for a barrier wall
around active or retired facilities.
Bottom Seals (Liners)
In working with new facilities, or transferring existing contam-
inated materials to new impoundments, it is frequently neces-
sary to design some sort of a liner system. The major function of
a liner is to prevent leachate or waste from entering the ground-
water regime. Liners, as with covers, can consist of native clays,
processed clays, or polymeric membrane liners. It is important to
note that under the recent "Interim Final Regulations",4 the
-- PUBLISHED CURVES(AFTER RYAN. 1977)
/ 28 D«S
/ /'
1
3 -
d
st
1
2:0 tt<
1.0 llf
0.5 tif
10 20 30 40 50 60
CEMENT CONTENT IT TIME OF MIKING (PERCENT)
Figure 9.
Strength vs. Cement Content
Cement-Bentonite Slurry Trench Wall
USEPA considers a synthetic membrane liner best to "prevent"
migration of wastes; whereas a clay liner will "minimize" migra-
tion of wastes. With all materials, compatibility testing is essen-
tial to determine the liner resistance to the waste or leachate to be
contained.
The geotechnical considerations for the design and construc-
tion of liners include many of the same considerations discussed
in the previous section discussing top seals. Only additional de-
tails unique to the use of natural clay, processed clay, and poly-
meric membranes as liners are discussed in the following subsec-
tions.
Native Clay Liners: The compatibility between the natural clays
and the waste is an important design consideration for the use of
natural clays as liners. It is important to ascertain the volume
change and permeability change characteristics of the proposed
clay liner material. The bulk transport of liquid waste must be pre-
cluded. Bulk transport of liquid through clay liners could occur due
to differential settlement of the foundation materials. Tensile
stresses within the liner could also result in cracking and subse-
quent bulk transport of liquid waste through the liner. Finally,
and probably most importantly, physical-chemical stresses due to
the pore-fluid-clay interaction could cause cracking."
The determination of liner/waste compatibility requires site-spe-
cific studies. The compatibility is a function of both waste type
and concentration. Studies of clay-waste compatibility conducted
to date15'16'17'21'22 have shed considerable light upon the subject.
Despite these recent advances, laboratory tests under triaxial stress
and gradient conditions can yield site-specific data from which to
evaluate the suitability of a natural clay liner.
Processed Clay Liners: The unique geotechnical design and con-
struction considerations for the use of processed clay for liners
also relate to waste compatibility. The volume change characteris-
tics of the processed clay are especially important. Generally, the
processed clay is mixed with the subgrade material to form the im-
permeable liner. The impedance to groundwater flow is typically
primarily by the processed clays, especially when the matrix soil is
relatively free of natural fines. Hence, if the processed clay is sub-
ject to shrinkage upon exposure to the waste, large increases in
permeability can occur. Even greater flow can occur if bulk trans-
port of liquid waste occurs due to liner cracking. The hydration of
a processed clay liner with uncontaminated water prior to waste
disposal is recommended."
Polymeric Membrane Liners: As with other liner types, waste
compatibility (not a geotechnology problem) is the major design
consideration. However, the permeability of a polymeric liner can
increase with liner stretching. Thus, total and differential foun-
SLUHRY
WALL
Figure 10.
Schematic of Vibrating Beam Slurry Wall
-------
182
BARRIERS
dation settlement can impact the liner design. Close construc-
tion control is essential to the overall system performance. The
"permeability" of an installed membrane liner system is generally
a function of bulk transport through seam, joints, tears, holes and
pinholes. Additional, typically non-geotechnical, aspects of poly-
meric membrane liners can be found elsewhere.24'
SUMMARY AND CONCLUSIONS
It is concluded that the application of conventional ground and
surface water control techniques to waste containment systems re-
quires special considerations beyond conventional design practices.
Further, the application of geotechnology is essential to the ade-
quate performance of these systems.
ACKNOWLEDGEMENTS
This work was conducted under the sponsorship of Woodward-
Clyde Consultants, Plymouth Meeting, Pennsylvania, Mr. Frank
S. Waller, Managing Principal. The opinions, findings, and con-
clusions expressed in this report are those of the authors, and are
not necessarily those of the sponsor. Special thanks are due to Mr.
Frank S. Waller and Dr. Arthur H. Dvinoff for their review of the
manuscript.
REFERENCES
1. Kolmer, J.R., "Investigation of LiPari Landfill Using Geophysical
Techniques". Proc. 7th Annual Research Symposium, Land Dis-
posal: Hazardous Waste, USEPA Report No. 600/9-8l-002b, Mar.
1981,298-311.
2. Lutton, R.J. "Evaluating Cover Systems for Solid and Hazardous
Waste" USEPA Report No. SW-867, Sept. 1980, 57 p.
3. Lutton, R.J., Regan, G.L. and L.W. Jones, "Design and Construc-
tion of Covers for Solid Waste Landfills", USEPA Report No.
60012-79-165, Aug. 1979, 249 p.
4. USEPA, 40 CFR Part 260, July 26, 1982.
5. Olson, R.C. and Daniel, D.E., "Field and Laboratory Measure-
ment of the Permeability of Saturated and Partially Saturated Fine-
Grained Soils" Geotechnical Engineering Report 80-5, University of
Texas, Austin, Tx, 1979, 78 p.
6. Emrich, G.H. and Beck, W., "Top Sealing to Minimize Leachate
Generation—Status Report". Proc. 7th Annual Research Symposium,
Land Disposal: Hazardous Waste, USEPA Report No. 600/9-81-
002b, Mar. 1980,291-297.
7. Haxo, H.E., Jr., "Durability of Liner Materials for Hazardous Waste
Disposal Facilities", Proc. 7th Annual Research Symposium, Land
Disposal: Hazardous Waste. USEPA Report No. 600/9-81-002b.
Mar. 1981, 140-156.
8. Kays, W.B., Construction of Linings for Reservoirs, Tanks and
Pollution Control Facilities, John Wiley & Sons, Inc., New York,
N.Y., 1977.
9. D'Appolonia, D.J., "Soil-Bentonite Slurry Trench Cutoffs" Journal
of the Geotechnical Engineering Division, ASCE, 106, GT4, 1980,
399-417.
10. Xanthakos, P.. Slurry Walls, McGraw-Hill, New York, N.Y. 1979.
11. Case International Co., "Case Study No. 5," Case Slurry Wall Note-
book, 1981.
12. Boyes, R.G.H., Structural and Cutoff Diaphragm Walls, Halstead
Press, New York, N.Y. 1975.
13. Ryan, C.R., "Slurry Cutoff Walls, Design Parameters and Final
Properties, An Interim Report", paper presented at Technical Course,
Slurry Wall Construction, Design, Techniques, and Procedures,
Miami, Fl., Feb. 1977.
14. Millet, R.A., and Perez, J.Y., "Current USA Practice, Slurry Wall
Specifications", Journal of the Geotechnical Engineering Division,
ASCE, 107, GTS, 1981, 1041-1056.
15. Evans, J.C., Chancy, R.D., and Fang, H-Y., "Influence of Pore
Fluid on Clay Behavior", Fritz Engineering Laboratory Report No.
384.14, Environmental Geotechnology, Lehigh University, Bethlehem,
Pa., 1981, 73 p.
16. Green, W.J., Lee, G.F. and Jones, R.A., "The Permeability of Clay
Soils to Water and Organic Solvents: Implications for the Storage
of Hazardous Wastes". Report submitted to USEPA, Oct. 1980.
17. Anderson, D. and Brown, K.W., "Organic Leachate Effects on the
Permeability of Clay Liners". Proceedings, 7th Annual Research
Symposium, Land Disposal: Hazardous Waste. USEPA Report No.
600/9-8 l-002b, Mar. 1981, 119-130.
18. Zimmie, T.F., Doynow, J.S., and Wardell, J.T., "Permeability
Testing of Soils for Hazardous Waste Disposal Sites", Proc. 10th
International Conference on Soil Mechanics and Foundation En-
gineering, Stockholm, Sweden, 1981, 403-408.
19. Wilson, D.B. and Fang, H-Y., "Hazardous Liquid Effects on Soil
Permeability: Theory and Testing" Fritz Engineering Report No.
462.8, Environmental Geotechnology, Lehigh University, Bethlehem,
Pa., 1982.
20. Fang, H-Y., Chancy, R.C., Failmezger, R.A. and Evans, J.C.,
"Mechanism of Soil Cracking", Fritz Engineering Laboratory Re-
port No. 462.4, Environmental Geotechnology, Lehigh University,
Bethlehem, Pa., 1982.
21. Fong, M.A. and Haxo, H.E., Jr., "Assessment of Liner Materials
for Municipal Solid Waste Landfills" Proc. 7th Annual Research
Symposium, Land Disposal: Municipal Solid Waste. USEPA Report
No. 600/9-81-0020, 1981, 138-162.
21. Fond, M.A. and Haxo, H.E., Jr., "Assessment of Liner Materials
for Municipal Solid Waste Landfills". Proc. 7th Annual Research
Symposium, Land Disposal: Municipal Solid Waste. USEPA RE-
port No. 600/9-8l-002a, 1981, 138-162.
22. Matrecon, Inc., "Lining of Waste Impoundment and Disposal Facil-
ities", USEPA Report No. SW-870, 1980.
23. Hughes, J., "Use of Bentonite as a Soil Sealant for Leachate Control
in Sanitary Landfills", Volclay Soil Engineering Report Data 280-E,
1975.
24. Gunkel, R.C., "Membrane Liner Systems for Hazardous Waste
Landfills". Proc. 7th Annual Research Symposium, Land Disposal:
Hazardous Waste. USEPA ReportNo. 600/9-8l-002b, 1981, 131-139.
25. Pacey, J.G., Brisley, C.G., Jr., and Dooley, R.L., "Field Verifica-
tion of Liners". Proc. 7th Annual Research Symposium, Land Dis-
posal: Municipal Solid Waste. USEPA Report No. 600/9-8 l-002a,
1981,163-169.
-------
COVERINGS FOR METAL CONTAMINATED LAND
A.K. JONES, Ph.D.
R.M. BELL, Ph.D.
L.J. BARKER
A.D. BRADSHAW, F.R.S.
Environmental Advisory Unit and Botany Department
Liverpool University
Liverpool, United Kingdom
INTRODUCTION
Metal contamination of land in Britain has arisen at all stages of
metal utilization from the mining of metalliferous ores to the dis-
posal of metal laden sewage sludges. Britain was one of the world's
major non-ferrous metal ore producers from the late 17th century
until the mid-19th century.1 The major ores mined were those of
lead, zinc, copper and tin. Ore extraction and processing produced
large quantities of waste materials ranging from coarse waste rock,
low in metal content, to very finely ground tailings.
During much of the period of maximum mining activity metal
recovery techniques were poorly developed and the fine tailings
frequently contained very high concentrations of heavy metals (up
to 10% zinc, 3% lead and 400 ug/g of cadmium at some lead/zinc
mines). These fine tailings were at first disposed of directly to
water courses but 1876 legislation resulted in disposal in impound-
ments.2 However, wind transport of these materials can affect soils
for some distance downwind of abandoned mine sites,3 while depo-
sition of metal rich particulates during periods of flooding has
caused severe contamination of agricultural land." The total num-
ber of individual mine sites if high high. For example, a survey of
mines in Wales revealed a total of over 750 individual sties, mostly
small in size but some containing over 20ha of contaminated land.5
Metal smelters have, in the past, been serious sources of land
contamination through emissions from furnace stacks, dust blown
form ore stockpiles and disposal of large quantities of waste slags
• containing high concentrations of metals. The major center of non-
ferrous metal smelting in the U.K. from the 18th to the early 20th
century was the Lower Swansea Valley in South Wales. Here 400ha
of land was despoiled as a result of the activities of numerous metal
smelting and refinery works from 1717 to 1974.6 In common with
most other derelict metal smelting and refinery sites the land in-
volved is in an urban area, thus restoration and reuse is important.7
Other industries which have given rise to problems of metal con-
tamination are foundries, scrap yards and metal reclaiming works,
paint pigment works and acid manufacturing plants.
Sewage sludge, especially when derived from urban/industrial
areas, can contain high concentrations of zinc, copper, tin, nickel,
cadmium, lead and mercury.8 Sewage sludge was formerly
dumped on agricultural land known as sewage farms. These farms
frequently encompass considerable areas and can be highly con-
taminated.9 The metals tend to be very available and can become
even more so as the sewage organic matter decomposes. Many of
these sites are close to urban areas and are prime sites for redevel-
opment.
COVERING LAYERS
There is thus a wide variety of metal contaminated land through-
out Britain in both rural and urban situations. Its restoration is im-
portant for public health, pollution control, aesthetic and econom-
ic reasons. One of the most widely used restoration methods for
these sites in the U.K. is to cover the contaminated material with
"clean" imported soil in order to provide a rooting medium for
plants and to isolate the potentially hazardous material from
human and animal contact. Such coverings can be simple layers
of topsoil or suitable sub-soil placed directly over the contam-
inated materials, or some intermediate layer may be incorporated
between the surface covering and the toxic substrate. These layers
have been included in order to inhibit root growth into the con-
taminated soil and also as a result of concern over the possible fail-
ure of restoration schemes through the effects of upward migra-
tion of metal ions in capillary water during dry periods.10'"'12'13
These intermediate materials, frequently termed barrier layers,
can be divided into three main types:
•Impervious seals—
which are intended to prevent solute movement between contam-
inated material and topsoil by virtue of their impermeability.
Included are heavy clay soils and artificial materials such as butyl
rubber and high density polyethylene. These materials have been
used on domestic refuse sites and sites containing organic contam-
inants.
•Porous capillary breaks—
which are intended to prevent the upward movement of water by
breaking the hydraulic conductivity between the topsoil and con-
taminated materials. Examples of materials used on metal con-
taminated sites include building rubble and limestone gravel.
•Chemical barriers—
which are intended to chemically precipitate metals in solution at
the waste/amendment interface. Ground limestone, pulverized
fuel ash and Leblanc waste* have been suggested as suitable ma-
terials.
•Waste material from the old Leblanc process for sodium carbonate manufacture.
It has a pH of 12 and consists largely of calcium sulphide and calcium hydroxide.
RESEARCH PROGRAM
In order to investigate the long term effectiveness of surface
covers on metal contaminated land the U.K. Dept. of the Environ-
ment have funded a research program whose prime aims are to
assess:
•Under what circumstances and to what extent contamination of
covering layers would occur under field conditions
•The extent of metal uptake into vegetation growing on covered,
metal contaminated materials.
The research program consists of three main areas of work con-
sisting of:
•Laboratory experiments
•Field experiments
•Survey of reclaimed sites
Laboratory Experiments
Experiments incorporating a range of waste materials and barrier
layers in artificial soil profiles have been constructed inside 300mm
diameter PVC tubes. Details of the construction of these profiles
were given in a previous paper.14 A section through one of the
columns is shown in Fig. 1.
183
-------
184 BARRIERS
The barrier layers used were estuarine clay, gravelly glacial clay,
sea won sand, land won sand, pulverized fuel ash, building rubble
Table 1.
Restored sites included in survey.
• 30OFJI) -
300mm
300mm
4OOmn>
Topsoil
Barrier
Layer
Contaminated
Waste
Perennial Rvegra
" Vhite Clover
-P V C Pipe
-
4t4
4 :: 4 •
•*• * -r +•» + -t- f <-. +
- .-. *« *• — — ^-
"T" ~' **• "• " "^ ~
-' Sand '.-.-— ' -• J
rr :
•+JI
,P V C Pipe
..Water Tsblo
"Ccpillory Hatting
-Polythene Lined
Trench
-Nylon Meih
Figure 1.
Section through soil column.
and ground limestone. Soil and vegetation samples from these pro-
files are collected at regular intervals and total metal concentra-
tions determined.
To date, under extreme drought conditions, one of the barrier
layer treatments, clay, has shown elevated metal concentrations
in the supported vegetation. The reasons for this are being inves-
tigated in a further series of experiments.
Field Experiments
A series of experimental plots has been constructed on metal
contaminated material at a former munitions works. The treat-
ments used duplicate those used in the laboratory experiments, but,
under field conditions. The plots are described in a previous pa-
per."
Uptake of metals into vegetables grown on the plots has been
minimal. Other experiments include an investigation of metal up-
take by vegetables and fruits grown on top of covering layers on an
old sewage farm."
Survey of Reclaimed Sites
A survey of sites in Britain where metal contaminated materials
have been covered and vegetated is being carried out. Investigations
at these sites consists of the collection of surface soil and vegeta-
tion samples, and extraction of cores through the covering layers
into the original ground surface. Total metal concentrations in
these samples are determined, analysis depending on the nature of
the contamination at the site.
The sites which have been investigated are described in Table I.
The metals of biological significance most commonly encountered
as contaminants are lead, zinc, copper (reflecting their industrial
importance) and cadmium (also widely used but its presence on
older contaminated sites is usually due to its strong association with
zinc).
Site
No.
1
2
3
4
5
6
r
a
9
10
11
12
13
14
15
16
of site Contaminants Restored
Lead/zinc Pb.Zn.Cd
mine
1855-1963
Lead/zinc Pb.Zn.Cd
•ine
1622-1921
Lead/zinc Pb,Zn,Cd
nine
1620-19O1
Lead/zinc Pb.Zn.Cd
mine
1892-1920
1948-1956
Lead/zinc Pb.Zn.Cd
nine
1710-1914
Lead/zinc Pb.Zn.Cd
mine
1850-1920
Lead Shelter Fb.Cu
1805-1913
Copper Snclter Pb.Zn.Cd.Cu
1810-1924
Copper/Arsenic Pb,Zn,Cu,As
Smelter
1866-1905
Steel Vorks Zn
1839-1975
Chromate Vorks Cr
1880-1968
Chromate Works Pb.Zn.Cd.
line smelter, Cr,As
Le b 1 an c Vorks .
Sulphuric acid
vorks
Leblanc works, Pb.Zn.Cu.As
sulphuric acid
plant
1847-1920
Sewage Sludge Zn
disposal areas
Domestic refuse Pb,Zo,Cd,
disposal area Hg,As
1895-1950
Domestic refuse Pb,Zn,Cd,Cu
disposal area
1962-1974
1977 20O-400mn coarse
sliale quarry
waste
1972 Varying depths of
waste rock and
sandy topsoil
1981 150-iOOmm stony
subsoil over
150->350mm wast*
rock
1977 150-250nai coarse
shale quarry
waste
1975 150-900mm of
burnt *nd unburnt
colliery spoil
1975 150-900mp of
including topsoil.
burnt and unburnt
colliery spoil
1971 100-200mm clay
topsoil over
30-30Omm clay
aubsoil
1973 150mm clay loaa
topsoil over
300mm limestone
gravel
1964 150-300mm acid
sandy-clay subsoil
1980 150-2 50mm clay
subsoil
1975 250-2000mm
coarse sandy soil
1975 Varying depths
of sand and top
soil over 350mm
Leblanc vast*
1975 100-200mm silty
clay over 150-
300mm railway
ash
19-68 Varying depths
1979 °f topsoil over
sand
1978 125mm topsoil and
25mm sand over
150mm flint gravel
1975 550-80Omm clay
gravel subsoil
Sheep
grazing
Sheep
grazing
Sheep
gracing
Sheep
grazing
Revegetation
R« vegetation
Public open
space
School
playing
field
Public
open apace
Public
open space
Public
open apace
Public open
space and
sports fioldi
Public
open
apace
Sheep
gracing
Public
open
space and
sports
fields
Allotment
gardens
The dates of restoration schemes are seen to be largely post 1970,
although many of the sites had been disused and derelict for many
years. This reflects the increasing public and official awareness of
the necessity for restoration of these contaminated sites and dere-
lict land in general during the 1960s. Local authorities were given
powers to acquire derelict land in 1963, and from 1966 they have
been eligible for generous grants to defray the costs of restoration.
The actual level of grant varies from 50% to 100% depending on
the area within which a site is situated."
The choice of cover used has been largely based on local avail-
ability of materials since transport costs represent a' major ele-
ment in the total cost of a restoration scheme utilizing imported
surface covers. In some cases decisions have also been based on
preliminary experiments and/or scientific advice.
The covering systems used have varied from simple covers of
subsoil to more complicated schemes involving intermediate barrier
layers between the waste and the final surface cover. The materials
used have encompassed the complete physical range from very
-------
BARRIERS
185
coarse waste rock through coarse gravel and sandy subsoils to
heavy clays. Chemically the materials have ranged from highly
alkaline chemical waste to acid sandy subsoil.
At some of the sites, covering of the contaminated material
has been incomplete due to insufficient material being imported. At
site 7 part of the area was left uncovered in order to preserve a
series of smelter fume condensation flues of archaeological inter-
est. As a result, materials containing very high metal concentra-
tions were still present at the surface. Metal concentrations in the
exposed contaminated materials compared with the restored areas
of the sites are given in Table 2. This situation would clearly not
be accepted at an intensively used site.
Table 2.
Sites with Incomplete Cover.
Metal concentrations in soils from covered and non-covered areas
Metal concentrations ug/g dry wt.
Site
number Pb Zn Cd Cu
1 covered 251 1140 4.5 —
non-covered 4460 20800 55.8 —
7 covered 160 183 0.5 43.7
non-covered 5280 512 0.6 386
9 covered 548 869 6.7 232
non-covered 1460 2350 54.0 67.2
"normal"
unmineralized 2-200 10-300 0.1-2.0 2-100
soil
At several sites, very coarse materials such as waste mine rock,
quarry waste, and colliery spoil have been used as cover materials.
Zinc concentrations are shown in Fig. 2a for a typical profile
through such a material (placed over tailings containing high con-
centrations of lead and zinc sulphides) in revegetation trials at a
derelict lead/zinc mine (site 5). There is no evidence of any in-
crease in zinc concentrations in the cover materials beyond 50mm
above the waste/amendment interface. The slightly elevated con-
centrations in the lowest sample of cover material was probably
a result of physical incorporation of waste into the cover during
restoration and also the difficulty of obtaining "clean" samples of
this very coarse material in close proximity to the contaminated
material. The very high concentrations of zinc in surface samples
were due to contamination from adjacent exposed areas of mine
waste.
a Site 5—Zinc profile
b Site 1—Zinc profile
100-
500-
£*»
«
Q700-
aoo-
..,,. ,
— *(. — I ' I100-
J fl».
T *m
t •g.jjj.
°«B-
• : '^'v. '•:•-..- •:•:'•"••.'.:.:::: 'S',';- ••;:[
, 1
I
I
I
I
I
•••.. :•• .•••':- • '..••:• l • ' •:'•": f- .'•*
' WO 600 1200 ' 'UOO ' UOO
Zn Mg/g
S Quarry orerburden
Lead/zinc mine vaate
1200 ' 1600' 2000 36800 37200
Zn tig/g
1 Unburnt colliery spoil
J Lead/zinc mine vaste
dence of any such movement since its sulphide, and its weathering
product, zinc sulphate, are more soluble than the corresponding
lead salts, and its concentration relative to that of cadmium is very
great although their chemistry is similar.
Burnt colliery spoil and alluvial sandy loam topsoil were also
used in these revegetation trials. The metal profiles in these ma-
terials were very similar to those in unburnt colliery spoil. As a
result of this trial and a similar series at site 6 a number of lead/
zinc mines have been covered with porous materials similar to
colliery spoils. Fig. 2b shows zinc concentrations in a typical pro-
file through coarse quarry waste at site 1.
At site 9, an acid, coarse, sandy subsoil was used to cover the
site of a former copper/arsenic smelter. Copper concentrations in
a typical profile from this site are shown in Fig. 3a. There was no
evidence of any movement of copper into the cover. The same is
true for lead, zinc, cadmium and arsenic. Copper concentrations
in a profile from the same site but in an area where vegetation
showed signs of toxicity are given in Fig. 3b. This condition was
simply due to an insufficient depth of cover above the toxic ma-
terial. Similar effects due to shallow cover layers have been ob-
served at other sites.
0-
100-
-q
-
,200-
0 200 400 600
AOO 800 1200 3200 3600
[_] Sandy subsoil
|C:jCopper/arsenic smelter vaste
Figures.
Site 9 Copper profiles
At site 11, a similar sandy subsoil was used to cover the site of a
chromate smelter. Chromium concentrations in a typical profile
from this site are shown in Fig. 4a. Some mixing of chromate
waste into the cover had occurred during restoration resulting in
the elevated concentrations in the lower horizons. Chromium con-
centrations in an area where a high water table was present with-
in the cover are shown" in Fig. 4b. This situation resulted in the
contamination of the sandy subsoil with soluble chromium salts.
The water table actually emerged at the surface in some areas of the
site resulting in complete death of vegetation due to high chrom-
ium concentrations in the plant rooting zone.
1
100- •* 1 1
I •*• f
•** IBB ¥t** *»*•* 1 \*/«%*H *»"»"»*•*«*»
0 10 100
Or |ig/g
Sandy subsoil
V Sandy subsoil
'•'.'• Chromate vaste
100-
1 1 20°-
., .."..i -^yo-
i • ' ' '^ • "^ COO
1000 10000 a
500-
600-
and chromate vaste 7QQ.
^3
rtf i i
.^
-S
T^-"
I
3
< Wafor
I
) 10
c
100 BOO 10COO
Figure 4.
Site 11—Chromium profiles
Figure 2.
The other metals determined in samples from this site (lead and
cadmium) showed very similar vertical distributions with no evi-
dence of contamination of the amendment by movement of metals
from below. Of the three metals, zinc was most likely to show evi-
At site 12, a layer of Leblanc waste has been incorporated into
the covering system as a chemical barrier between topsoil and chro-
mate waste. The position of the barrier varied depending upon the
requirements of landscaping. Chromium concentrations in a typical
-------
186
BARRIERS
0-
100 •
5 300-
a
Q 1.00-
soo-
600'
» >>
y/1 ' 1 100-
K^^ 5 300-
''•'^'\v'\''\'\'\'\'\'\\!\^'\^'\ 50°"
I
1
-:, i
•;\\\\x. \\\\\\\\'
•ry^XVN \\\\\\\l
*V\\^x\\\N
"'X\\\\\\\\\^\i
> V BO 1000 10000 100000 0 10 100 1000
Cr m/g Cr ug/g
vironmental Advisory Unit (Director, Dr. G.D. Parry) and Botany
Dept. for technical assistance and useful discussion, and the offi-
cers of numerous local authorities for access to sites.
D
Figure 5.
Site 12—Chromium profiles
core from this site are shown in Fig. 5a. There was evidence of a
slight increase in chromium concentrations in the lower horizons
of the barrier. Chromium concentrations in an area of the site
where vegetation shows signs of toxicity are shown in Fig. 5b. They
were elevated throughout the cover profile. This core was taken
in an area at the base of a steep slope where water would be
expected to emerge during wet weather.
Heavy clay soils have been used as cover material at three sites.
These sites showed evidence of mixing of contaminated waste with
the lowest cover horizons, but there was no evidence of any other
contamination of the cover soils.
CONCLUSIONS
The information derived from this project has shown that the
measures used for restoration of metal contaminated land have
been successful. In general, where covering has been carried out in
accordance with recommendations there is no evidence that re-
gression due to reappearance of toxicity will occur. Toxicity prob-
lems still exist at some sites due to areas of contaminated ma-
terial remaining uncovered.
At some sites the provision of an insufficient depth of cover has
resulted in toxic concentrations of metals within plant rooting
zones causing adverse effects and in some cases, death of vege-
tation. These investigations at reclaimed sites have produced no
evidence of upward migration of metals into covering layers
through transport of metals in solution in capillary water under
field conditions.
At some sites elevated metal concentrations in surface covers
have occurred through physical incorporation of contaminated
materials into the amendments during restoration work. At two
sites where chromium is present in a soluble anionic form, con-
tamination of the surface covers with resulting toxicity to vege-
tation has occurred where water enters waste mounds, becomes
contaminated, and re-emerges through the covering layers at lower
levels. At none of the sites was there any evidence of similar move-
ment of cationic metal ions through soil covers.
ACKNOWLEDGEMENTS
This research was carried out under contract ICRCL 3/78 for
the U.K. Dept. of the Environment. The views expressed are those
of the authors and do not necessarily represent those of the De-
partment. The authors wish to thank their colleagues in the En-
REFERENCES
1. Barnes, J.W., "The first metal workings and their geological setting."
in Subterranean Britain ed. Crawford, H., London, John Baker, 1979.
2. H.M. Govt., "Rivers Pollution Prevention Act" H.M.S.O. London,
1876.
3. Davies, B.E. and White, H.M., "Environmental Pollution by Wind
Blown Lead Mine Waste: a case study in Wales, U.K.," Sci. Tot.
Env., 20, 1981,57-74.
4. Alloway, B.J. and Davies, B.E., "Trace element content of soils af-
fected by base metal mining in Wales," Geoderma, 5, 1971,197-208.
5. Liverpool University/Welsh Office, "Survey of Abandoned Metalli-
ferous Mines in Wales," Unpublished report, 1978.
6. Hilton, K. J. (Ed.), The Lower Swansea Valley Project, London, Long-
mans, 1967.
7. Lavender, S. J., New Land for Old, Bristol, Adam Hilger, 1981.
8. Dept. of the Environment/National Water Council, "Report of the
Sub-Committee on the Disposal of Sewage Sludge to Land," Stand-
ing Technical Committee Report No. 20, London, National Water
Council, 1981.
9. Pike, E.R., Graham, L.C. and Fogden, M.W., "An Appraisal of
Toxic Metal Residue in the Soils of a Disused Sewage Farm." /. Assoc.
of Public Analysts, 13, 1975, 19-33.
10. Ames, S., "Migration of acid substances in Sullivan tailings—a col-
umn study." In: Reclamation of Lands Disturbed by Mining. Proc.
3rd. Annual British Columbia Mine Reclamation Symposium, Vernon,
British Columbia, March 7-9, 1979.
11. Craze, B., "Investigations into the' revegetation problems at Cap-
tain's Flat mining area." J. Soil Conserv. Service of New South Wales,
33, 1977, 190-199.
12. Gemmell, R.P., "Revegetation of derelict land polluted by a chromate
smelter. Part 2: Techniques of revegetation of chromate smelter
waste". Environ. Pollut., 6, 1974, 31-37.
13. Johnson, M.S. and Bradshaw, A.D., "Ecological principles for the
restoration of disturbed and degraded land," Applied Biology, 4,
1979, 141-200.
14. Jones, A.K., Johnson, M.S. & Bell, R.M., "The Control of Metal
Dispersal in Reclaimed Land," Mining Magazine, 144,1981,249-257.
15. Jones, A.K., Johnson, M.S., Bell, R.M. and Bradshaw, A.D., "Bio-
logical aspects of the treatment of heavy metal contaminated land for
housing development schemes, Paper C3 in Reclamation of Contam-
inated Land, Proc. Soc. of Chemical Industry Conference. East-
bourne, Sussex, U.K. 1979.
17. Lepp, N.W. and Harris, M.R. "A strategy for evaluation of soil cov-
ering techniques to reduce trace metal uptake by soft fruits and veg-
etables," Paper C7 in Reclamation of Contaminated Land, Proc. Soc.
of Chemical Industry Conference, Eastbourne, Sussex, U.K. 1979.
18. Chisholm, M. and Howells, J., "Derelict Land in Great Britain",
in Dealing with Dereliction—the redevelopment of the Lower Swan-
sea Valley, ed. Bromley, R.D.F. and Humphrys, G., 1979, 3-19.
-------
SELECTION, INSTALLATION, AND POST-CLOSURE
MONITORING OF A LOW PERMEABILITY COVER OVER A
HAZARDOUS WASTE DISPOSAL FACILITY
MARKJ.DOWIAK
ROBERT A. LUCAS
ANDRZEJ NAZAR
DANIEL THRELFALL
PEC Division, NUS Corporation
Gaithersburg, Maryland
INTRODUCTION
A closure plan, recently designed and implemented for a haz-
ardous waste disposal facility, included the installation of a low-
permeability cover over the wastes. This facility, located in western
Pennsylvania, consisted of a lime neutralization plant and a 26-acre
sludge impoundment. The facility started operations prior to the
existence of strict hazardous waste disposal regulations and was not
constructed with a liner system to protect groundwater.
As part of the closure requirements of the Pennsylvania De-
partment of Environmental Resources (PADER), a low-perme-
ability cover as well as a leachate collection and treatment sys-
tem had to be installed. In addition, the facility owner is respon-
sible for monitoring aquifers beneath the site to determine changes
in water quality and the effectiveness of the liner. Presented here
are the hydrogeologic conditions, the considerations that went into
cover selection and design, the cover installation procedure and
cost, and the groundwater monitoring plan.
SITE CONDITIONS
The site is located in a sparsely populated area of farms and
woodland. During the active life of the disposal facility, it received
mostly acidic wastes from the regional steel industry for lime treat-
ment and disposal. Prior to use as a treatment and disposal
facility, the site was mined for the Pittsburgh Coal. Deep mining
occurred first, starting just after 1910. Haulageways and rooms
were left open beneath the surface from where the coal was re-
trieved. This mining method left coal in place to support the over-
lying rock. The amount left was sometimes up to 50% of what was
originally there.
During several periods between 1930 and 1950, surface mining
was performed to remove the remaining pillars of coal. Large
equipment was used to excavate in successive cuts along the con-
tour of the hillside, working deeper with each pass to reach the
coal. Overburden was cast behind and downslope from each new
cut.
After the last cut was made, a long open pit was left, with a
highwall of in-place rock on the north side and mine spoil over-
burden piles on the south. No laws existing at that time required
reclamation of the land.
This property was then purchased to utilize the open cut as a dis-
posal lagoon. A treatment plan, consisting of several holding tanks
and mixers, was constructed at one end of the abandoned strip cut.
Once wastes were neutralized, they were discharged as a slurry.
They flowed by gravity into the strip cut. The waste would then
partially dewater and form a semi-solid sludge. The operation
continued from 1953 until 1981, when the amount of wastes ap-
proached the capacity of the cut.
At that time, PADER gave standards for closure of this facility.
The major concern was groundwater contamination.
To minimize leachate generation a low-permeability cover was
required. In addition, a network of leachate collection drains along
with a leachate treatment plant were required.
GEOLOGIC SETTING
The site is located in the unglaciated portion (Kanawha Sec-
tion) of the Appalachian Plateau Physiographic Province. The
regional structure generally consists of subparallel anticlines and
synclines trending northeast and plunging to the southwest.
Rocks exposed at the site are assigned to the Monongahela and
Conemaugh Formations of the Pennsylvanian age. The base of the
Pittsburgh Coal is the boundary between these formations. The
rocks immediately underlying the base of sludge belong to the Con-
emaugh Formation. The lower portion of the Monongahela
Formation is also present, but in much of the area these rocks
have been removed by strip mining.
The rocks of the Monongahela Formation consist primarily of
sandstone, shale, and limestone. The thickness of this unit varies
from 50 to over 80 ft at this site.
The Conemaugh Formation consists of a very complex series of
delta deposits. Most facies in the formation are lenticular, and the
geology can vary significantly in the horizontal direction.
The study area has substantial mining history, because of the oc-
currence of the Pittsburgh Coal seam. The hill abutting the waste
disposal site on the north was deep mined by the Pittsburgh Coal
Company in the 1920s and 1930s. Mining was by the room and
pillar technique, a common method at that time. Entries to the
mine were at the area presently forming the northeast corner of the
easternmost sludge pond. Surface mining took place in this area in
the 1930s and early 1950s.
Mine maps show that deep mining operations extend through
portions of the south-facing highwall. Mining also occurred
through the center of the highwall where a depression extends to
the north. Sludge filled this depression and formed a fingerlike
configuration extending northward from the main sludge area.
The heavy metals in the waste sludge leach very little, most like-
ly because the lime treated sludge has a pH of 8 to 10. Leachate
testing has indicated that significant levels of NH3N, NO3-N,
chloride and phenol are dissolved from the sludge.
Extensive sampling and analysis has been performed on the num-
erous seeps around the site. The seeps located upgradient from the
site do not indicate any effect on groundwater quality from the
disposal facility. Their water quality represents typical acid mine
drainage water with low pH and high concentrations of Fe, Mn,
and sulfate.
Seeps located along the south perimeter of the site downgrad-
ient indicate high concentrations of NH3-N, chloride, phenol, and
heavy metals (Al, Fe, Ni, Mn, Zn) where the overall quality has
deteriorated with respect to pH. Mine waste leaching was sus-
pended to contribute to low pH and increased metal concentra-
tions.
During facility operation, seepage originated from upslope run-
off, direct rainfall, and free water from sludge deposition. The
major portion of the water flowed southward and passed through
the impoundment dikes. These waters continued their generally
southward and downslope flow until reaching the perimeter of the
187
-------
188
BARRIERS
spoil areas, where considerable surface seeps appeared. These seeps
then combined with two main water courses draining site.
The general geologic conditions at the site were determined by
surface reconnaissance, mine maps, and investigation of highwall
exposure in a strip cut on the opposite side of the south-facing
highwall which formed the northern wall of the sludge pond.
SUBSURFACE INVESTIGATION
The objective of the subsurface investigation was to define hy-
drogeologic conditions and determine the existence or probability
of groundwater contamination. The hydrogeology was studied via
10 monitoring wells (Fig. 1). Six shallow monitoring wells and four
deep monitoring wells have been installed. The wells monitor
shallow and deep groundwater flow systems. Test borings were
drilled using the rotary method with air or water. Roller bit or core
samplers were used to obtain rock samples. After coring, each
borehole was reamed with a 7.875 in roller bit, and 4 in. diameter
PVC monitoring wells were installed. In-situ permeability tests
were performed. Falling head or pressure tests were conducted, de-
pending on the rock type being evaluated. A generalized geologic
cross section is shown in Fig. 2.
I
LEGEND
A SURFACE WATER MONITORING POINTS
O SHALLOW GROUNDWATER MONITORING POINTS
• DEEP GROUNOWATER MONITORING POINTS
=0 DIRECTION OF SHALLOW GROUNDWATER FLOW
•^^ DIRECTION OF DEEP GROUNDWATER FLOW
I I SLUDGE IMPOUNDMENT
LOCATION OF SAMPLING POINTS
Figure 1.
Location of Sampling Points
Test borings through the Conemaugh Formation indicated that
the geology of this formation can be divided into four units.
Unit I—0 to 100 ft below base of sludge. The uppermost unit is
geologically the most complex. It consists primarily of claystone
and siltstone, but contains numerous lenses of sandstone, shale,
and limestone. At the northwest end of the site, claystones are
very abundant in the lower part of this unit, but these often grade
southeastward into sandstones. The sandstones are somewhat
better developed in the southeast portion of the area. A continuous
limestone layer was discerned in the uppermost part of this unit.
This lenticular limestone bed is elongate in a northwesterly direc-
tion. The limestones are separated by siltstone and claystone.
TEST BORINO
*«TEH TMLE
PICZOHCTRIC SUKFUE
Figure 2.
Generalized Geologic Cross-Section
The uppermost unit ranges from 85 to 100 ft thick where it has
not been breached by erosion. The dip of this unit is northeast-
ward.
Unit II—100 to approximately 130 ft below the base of sludge.
The rocks underlying the uppermost unit consist mostly of the
sandstone. Medium to medium-coarse grained argillaceous sand-
stone with abundant bituminous partings occur in the southeastern
portion of the area. In the northwest portion of the area, it is a silty
sandstone. It is 20 to 30 ft thick in the southeast and approximate-
ly 40 ft thick in the northwest.
Unit III—130 to 140 ft (approximately) below the base of sludge.
Beneath the sandstone is an interlaminated unit of sandstone with
black siltstone. It is 3 to 11 ft thick, with the greatest thickness
occurring in the west.
Unit IV—140 to 215 ft (approximately) below the base of sludge.
The interlaminated unit overlies the thick sandstone layer. It is
approximately 75 ft thick, with some siltstone beds in the lower
part. A claystone marks the base of the sandstone layer.
During this investigation, three principal water-bearing zones
were identified. These are:
•A shallow groundwater flow system in the mine spoil, rock strata,
and abandoned deep mines above the Pittsburgh Coal underclay
and at the base of the sludge elevation
•An aquifer in the limestone layers beneath the Pittsburgh Coal,
approximately 5 to 15 ft below the base of sludge
•A deeper aquifer in a sandstone unit of the Conemaugh Forma-
tion, approximately 140 to 160 ft below the base of the sludge.
The first aquifer is represented by the shallow groundwater flow
system. This flow system occurs in bedrocks and mine spoil above
the Pittsburgh Coal underclay and also in the abandoned deep
mine. Water that infiltrates through this rock strata or mine spoil
becomes perched on the Pittsburgh Coal underclay. Perched
groundwater on the Pittsburgh Coal underclay is recharged from
the north areas of the waste disposal site. This water flows be-
neath the existing sludge or at the interface between the sludge and
mine spoil. Several seeps at the south perimeter of the site indicate
this is the discharge area for this groundwater flow system.
The second principal aquifer occurs in the limestones from 5 to
15 ft below the base of sludge. These continuous and discontin-
uous limestone layers, although thin, appear to be good aquifers
because of their high permeability. The limestone units represent
a semi-confined aquifer. The flow is largely controlled by local
topography but is generally southward. A swampy area southeast
of the sludge impoundment may be the groundwater discharge
area. This aquifer is separated from the underlying, deeper aquifer
-------
BARRIERS
189
by a continuous layer of the aquitardes and aquicludes. These lay-
ers consist mainly of claystone and siltstone.
Six shallow monitoring wells have been installed into this
aquifer. Three are upgradient and three are downgradient wells.
The downgradient wells have consistently higher chloride, NH3,
and NO3-N. This ground water is found in close proximity to the
cropline of the unit and travels a short distance before it is dis-
charged.
The third aquifer at the site is a sandstone from 140 to 160 ft
below the base of sludge. The thickness of this sandstone aquifer
varies from 20 to 40 ft. This unit is particularly permeable at its
top. However, it grades into a silty sandstone to the northwest; as
it does so, it becomes less permeable. The most permeable zone
occurs near its base. The ground water flow direction is westward.
This waterbearing stratum is a confined aquifer. Four monitoring
wells have been emplaced into this aquifer.
The following observation was made during the drilling one of
the monitoring wells. Water in the limestone aquifer was sealed off
with casing and bentonite. As the rest of the hole was drilled, care-
ful observations were made to detect water. The hole remained
totally dry until a depth of 160 ft. The rocks below 160 ft, however,
were saturated with water. This water is under pressure, as indi-
cated by the fact that water level in the hole stands at 107 ft. The
rock unit below the limestone aquifer can be attributed with being
the confining layer which separates these two aquifers. The sand-
stone aquifer does not indicate any effect by leachate.
Since installation of the monitoring wells, water table measure-
ments and samples have been periodically taken.
EVALUATION AND SELECTION OF COVER MATERIAL
The objective *of the cover placement was to reduce or elimi-
nate rainfall percolation through the sludge and consequently pre-
vent leachate generation and contamination of groundwater.
PADER required that the cover have a hydraulic conductivity of
less than 1 x 10"7 cm/sec.
Cover options considered were: compacted local soil or mine
soil, local soil with bentonite additives, bentonite or clay import
or a synthetic membrane.
Compacted local soil or mine spoil. This option was considered
as a result of the abundance of this material available due to the
local strip mining. There were several problems which prevented its
use. Much of this material was coarse. The amount of fines was
relatively small, which would leave a significant amount of void
space. Compaction could not be expected to overcome this prob-
lem because high densities could not be attained over the semi-
solid sludge surface. In addition, some of this material consisted of
carbonaceous shale fragments, which would produce an acidic
leachate as rainfall infiltrated through the cover. This leachate
would then flow through the sludge, releasing even more contam-
inants as a result of its increased apidity. The extraction of non-
carbonaceous fines in suitable quantity from the local material for
compaction over the sludge was considered technically and eco-
nomically infeasible.
Bentonite additives to local soil. The high swelling characteristics
of bentonite clay make it an effective soil sealer. The application
rate is a function of the physical gradation and void ratio of the
soil. The coarse nature of the local soil, with resultant high void
ratio, required that high application rates be made to achieve low
permeability. These application rates made this option economical-
ly infeasible.
Bentonite or clay import. A bentonite or clay import was con-
sidered. Bentonite could have been hauled in by rail from Wyom-
ing; however, a supply adequate to cover 26 acres was relatively
expensive. Other clays closer to the site did not possess the superior
qualities of bentonite and would not have performed suitably as a
cover.
Synthetic membrane. The use of a synthetic membrane was
found to be the most economical cover option. The only other
practical option from a strictly technical viewpoint was the ben-
tonite import, but it was too costly in this particular instance.
Under different circumstances, the synthetic membrane may not
have been the option ultimately selected. If local soil had more
fines, it may have been suitable for bentonite additives. Further-
more, if leachate collection and treatment was not part of the site
closure and if the cover was the sole or primary protective feature
to maintain groundwater quality, then questions relating to the life
of the synthetic membrane may have resulted in selection of a ben-
tonite cover.
FINAL COVER INSTALLATION
The final cover is a composite layer and consists of a local bor-
row material grade coarse, two layers of geotextile fabric, PVC
membrane, and a vegetation and soil cover layer. Storm water run-
off and seepage drainage systems are incorporated on and within
the cover layer. A typical cross-section of the final cover is shown
in Fig. 3.
DRAINAGE SWALE
VARIES - 3' to 10'
2' COVER (SPOIL/ REDOOG/
ORGANIC MIXTURE )
ANCHOR TRENCH
4' PERF. DRAIN TUBING
2' GRADE COURSEIMin I
Figure 3.
Detail of Low Permeability Cover
The cover surface was designed to slope at a minimum of 2% to
achieve drainage of storm water runoff. Since the 30-acre sludge
impoundment was developed in a contour strip mine, its shape is
irregular and has a variably curved perimeter. Width of the im-
poundment ranged from 200 to 700 ft. This uncommon config-
uration necessitated a cover fill of varying thickness and grade to
achieve conveyance and drainage of storm water runoff.
Material for the grade coarse fill was borrowed from local mine
spoil and mine refuse piles to the south of the site. The mine spoil
material is weathered rock overburden and soils from strip min-
ing of the local coal seam in the 1920s and 1930s. The mine spoil
was typically soft to hard shale and sandstone rock fragments from
0.5 to 24 in with little fine-grain particles. Mine refuse is the re-
ject material from preparation of coal removed from deep min-
ing of the same local coal seam in the 1920s. Mine refuse is typical-
ly more carbonaceous and acidic because of the higher percentage
of coal and pyritic minerals. Under these conditions the mine re-
fuse commonly oxidizes to form a hard, shale-like semi-fused mass,
commonly called reddog. Reddog was present throughout the mine
refuse deposits. Reddog is typically hard shale and sandstone frag-
ments from 0.5 to 24 in with trace fine-grained particles.
The grade course fill ranged in thickness from 2 to 11 ft. Ap-
proximately 316,000 yd' of material was required. Self-loading
scraper pans of 31 to 44 yd3 capacity (Caterpillar 637 and 657)
and 200 and 300 hp bulldozers (Caterpillar D7 and D8) were used
to excavate and place the borrow material.
Prior to placement of the grade course fill on the sludge im-
poundment, a number of test fill ramps were constructed on the
sludge surface to evaluate the bearing capacity of the sludge ma-
terial. The semi-solid character of the sludge deposits raised the
question of the feasibility of placing a cover fill with the heavy
equipment.
In-situ sludge water contents were found to be 150 to 240% with
a unit weight of 80 to 90 lb/ft3. The sludge is predominantly a
silt-size material with a void ratio of approximately 8.0 and a com-
-------
190
BARRIERS
pression index from approximately 2 to 2.5. Natural soils with com-
pression indices greater than about 0.4 are considered to have high
compressibility, and few natural soils have compression indices
greater than about 1.0. Therefore, the sludge was known to have a
high compressibility and thought to possess a low bearing capacity.
The fact that the sludge surface would support a person in-
dicated that the sludge possessed some shear strength and bear-
ing capacity. The ground pressure of a 200 Ib individual standing
on the impoundment surface is estimated to be approximately 0.5
to 1.0 psi. This is considerably lower than the average ground pres-
sure of 10 psi for a 300 hp bulldozer and up to 20 psi for a loaded
scraper pan.
Since the sludge properties were found to vary with depth and
area! extent, a field test fill was judged to be the most effective
means to evaluate sludge bearing capacity for the covering opera-
tions.
The test fill was performed to determine the bearing strength of
the sludge and the minimum cover fill thickness needed to sup-
port the proposed range of earthmoving equipment. A 4 to 5 ft lay-
er of borrow material was placed on the impoundment by end
dumping from the retaining dikes and pushing the material onto
the sludge with the smallest available dozer. Initial results showed
that the sludge would support 5 ft of cover material alone which
indicated a minimum bearing capacity of approximately 4 psi.
The dozer then slowly extended the test fill out into the sludge to
create a ramp approximately 25 ft wide. Since the sludge was
demonstrated to support the dozer and 5 ft of cover material, the
minimum suggested bearing capacity was increased to about 15 psi.
This procedure was continued with a series of test fill ramps
across the thickest parts of the sludge. Equipment loadings were
increased in increments up to a fully-loaded 44 yd' scraper pan
without any failure in sludge shearing strength.
Once the supportability of the sludge impoundment was ade-
quately demonstrated to the owner, engineer, contractor, and
regulatory agency officials, the installation of the final cover was
able to proceed.
The complete cover was installed on the sludge impoundment
over a four month period, from July to Nov. 1981.
The grade course fill was placed on the sludge impoundment
at the required final slopes for drainage of storm water and seep-
age. Since the grade course material included hard, angular, rock
fragments, a protective, puncture resistant support layer was placed
on the finished surface of the grade course to protect the PVC
membrane. This support layer was a non-woven polypropylene
fabric manufactured by DuPont (TYPAR 3401). The fabric was
placed in 20 ft wide sections and overlapped a min of 6 in.
A 20 mil PVC membrane manufactured by B.F. Goodrich was
placed directly on top of the geotextile support fabric. The mem-
brane was delivered to the site in 100 ft wide by 200 or 300 ft long
folded sections. An average field crew of 18 laborers and a fore-
man installed 1,140,000 ft' of membrane in 3.5 weeks.
Two technical representatives of the liner manufacturer were on-
site full time to supervise liner installation. This supervision and
final approval of installation was written into the project specifica-
tions and was considered a critical item in construction quality
control.
The membrane was solvent bonded using an average 4 in overlap
on each 100 ft section. Solvent was applied with a brush to the
contact sides of each membrane piece. The seam was formed by
pressing both solvent coated pieces together with a hand roller.
Curing of each seam was completed in approximately 24 hr, de-
pending on ambient air temperature.
Care was taken to avoid membrane installation at temperatures
less than 40 °F, since this would have an adverse effect on seam in-
tegrity.
The membrane was keyed along the impoundment perimeter us-
ing an average 1 ft deep, soil-backfilled, anchor trench. A per-
forated drainage pipe with a coarse aggregate backfill was placed
continuously along the toe of the membrane to drain and convey
water flowing on or near the membrane surface.
Another layer of geotextile fabric, identical to the fabric beneath
the membrane, was placed directly on the membrane surface. This
layer protects the membrane from puncture by angular rock frag-
ments in the soil cover layer and additional dynamic stresses created
by heavy equipment traffic.
Since the soil cover layer was required to support vegetation, a
common borrow alone as used in the grade coarse fill was not ac-
ceptable. Soil cover layer specifications required a maximum 4 in.
particle size, significant fine-grained soil fractions, and an organ-
ic soil supplement at an application rate sufficient to improve
physical fertility for maintenance of good vegetative growth. Lime
and fertilizer applications were also specified for seedbed prepara-
tion. The soil cover layer averaged 2 ft in thickness. Approximate-
ly 84,000 yd3 of material was required.
Because the volume of soil cover was significantly less than the
grade course fill, alternate borrow areas were investigated for a
suitable material. Significant quantities of residual clay-silt soils
and topsoil were found in undisturbed areas beyond the edge of the
mine wastes. Organic silty soils were borrowed from the local
stream beds as part of the foundation excavations for the sedimen-
tation basin embankments.
The available amount of these soils was not sufficient to meet
the required volume for the soil cover layer. It was decided to
blend the organic fine-grained soils with mine spoil material less
than 4 in in size to provide the final cover layer.
The mix ratio was approximately 2 parts mine spoil to 1 part
organic-silt soil. Mixing was performed by the earthmoving equip-
ment during material placement on the fabric-covered membrane.
Care was taken not to operate equipment directly on the covered
membrane. Cover materials were placed in lifts, with the initial 1 ft
lift being mine spoil, and subsequent 6 in. lifts being mine spoil and
organic silts. The placed cover layer was disced and sacrificed to
complete mixing prior to hydroseeding of the lime, fertilizer, and
soil-legume seed mix.
COSTS
The total cost for placement of the impermeable cover over 26
acres was approximately $2,030,000 (1981 costs). This converts to
$78,077/acre or approximately $1.80/ft2.
This cost included the following items:
•Mobilization and site preparation
•Clearing and grubbing
•Sedimentation basins
•Excavation and fill
•Storm water drainage channels'
•Underdrain
•PVC membrane and geotextile fabrice
•Vegetation and mulch
•Supervision, field engineering, and general labor
POST CLOSURE MONITORING
It is too early to determine the effect, if any, of the low-perme-
ability cover on groundwater quality. Post-closure monitoring will
include sampling of 11 surface water locations (seepage points
and streams) and 10 monitoring wells. Samples are being analyzed
for:
PH
Free cyanide
Ammonia Nitrogen
Nitrate Nitrogen
Nitrite Nitrogen
Phenols
Specific conductance
Total dissolved solids
Hexavalent Chromium
Mercury
Zinc
Chloride
Fluoride
Aluminum
Cadmium
Calcium
Total Iron
Manganese
Lead
Nickel
Sodium
Potassium
Sampling occurs on a monthly basis and is routinely reported to
PADER. An on-going evaluation is underway to determine the
long-term effectiveness of the cover.
-------
POLLUTION MIGRATION CUT-OFF USING
SLURRY TRENCH CONSTRUCTION
PHILIP A. SPOONER
ROGER S. WETZEL
JRB Associates
McLean, Virginia
WALTER E. GRUBE, JR. Ph.D.
U.S. Environmental Protection Agency
Municipal Environmental Research Laboratory
Solid and Hazardous Waste Research Division
Cincinnati, Ohio
INTRODUCTION
Over the last two decades, cut-off walls emplaced by slurry
trenching have been installed at many solid and hazardous waste
facilities. This paper presents the interim results of a project for
USEPA's Office of Research and Development to develop a tech-
nical handbook on the use of slurry trenching techniques to con-
trol pollution migration.
Slurry trenching is a method by which a continuous trench is
excavated (by backhoe or clam-shell grab) under a slurry of
bentonite and jvater. This slurry, or more correctly, colloidal sus-
pension, supports the trench walls and allows excavation to great
depths with no other means of support. Once a trench has been
excavated under slurry to the required depth and length, it is back-
filled to form a continuous, nearly impermeable barrier, or cut-off
wall, to groundwater flow.
In some cases, the slurry is a mixture of water, cement, and
bentonite which hardens in place to form the final barrier. In other
instances, this trenching technique is used to place soil-benionite
cut-off walls, where the backfill is composed of the excavated
soil with small amounts (1 to 4%) of bentonite added.
In some instances, borrowed soil materials, such as additional
fines, or gravel may have to be added to meet the requirements of
lower permeability or higher strength. Another variation of this
technique, seldom used for pollution migration control, is the use
of pre-cast or cast-in-place concrete diaphragm walls, which are
installed when both pollution migration cut-off and structural sup-
port are required.
History
The use of bentonite slurries in trench excavation is an out-
growth of their use as drilling muds in petroleum exploration. In
the late 1920s it was found that bentonite slurries aided drilling by
lubricating the bit, removing the rock cuttings, and sealing the sides
of the hole.1'2 This use is reflected in the current bentonite classif-
ication system developed by the American Petroleum Institute
(API). Under the API system, bentonites are rated by the number
of barrels of drilling mud (of a given viscosity) that are obtain-
able from one ton to bentonite.
In Europe during the 1930s, it was found that borings and
trenches can be kept open using these slurries and that the tech-
nique is well suited for use in restricted work areas and excavations
in unstable materials.3 The use of slurry walls for pollution cut-off
began during the early 1970s and to date, many have been installed
in the U.S." Some typical wall installations for pollution control
are listed in Table 1. Nonetheless, there are less than a dozen firms
experienced with the construction technique, and only a few of
those appear qualified to install slurry walls for pollution migra-
tion cut-off.
SLURRY MATERIALS AND FUNCTION
The material used in the trenching slurry is a commercially avail-
able product called bentonite, sodium bentonite, or Wyoming ben-
tonite. Bentonite is considered by many to be a clay but is, in fact, a
rock composed mostly of the clay mineral montmorillonite, with
smaller amounts of other clays and various metallic oxides. Mont-
morillonite crystals contain negative charges which attract and
adsorb cations. In most bentonites, the montmorillonite is satur-
ated with calcium ions: however, the products which are used for
slurries and drilling muds are the Wyoming bentonites which con-
tain more of the sodium saturated variety. This sodium benton-
ite is generally highly colloidal and plastic, and has the ability to
swell many times its volume and to form thixothropic gels in
water.2
Many commercially available bentonite products are highly mod-
ified by the addition of polymers, peptizers or other chemicals.
This practice has been adopted in order to increase the API yeild
value of lower grade bentonite deposits, and in the case of poly-
mer extended products, to increase piping resistance in liners.
The remainder of this discussion assumes the use of "unaltered"
high grade (i.e., high sodium montmorillonite content) bentonite.
The slurry used in trenching is composed of clean water mixed
with from 5% to itfo bentonite by weight.5 Clean water is used to
Table 1.
Example Slurry Walls Installed for Pollution Control.
LOCATION
Indiana
Illinois
Louisiana
Pennsylvania
New York
New Hampshire
Wisconsin
New Jersey
Saskatchewan
Florida
Ohio
WALL
COMPOSITION
Soil-
Ben tonl te
Soll-
Bentonlte
Soll-
Bentonlte
Cement-
Bentonlte
Cement-
Bentonlte
Soll-
Bentonlte
Soll-
Bentonlte
Soll-
Bentonlte
Soll-
Bentonlte
Soll-
Bentonlte
Both Soil-
& Cement
Bentonite
Cement-
Bentonlte
SIZE* ~tOST POLLUTANT
1000 ft2 (S/ft2) CONTAINED
500
13.5
30
I
77.3
20li
SB.*
10
53.*
58
25
10
3
I,
2.5
2
5.5
5
2.8
Inorganics
Ash Disposal
Methane Gas
Ethylene
Dlchlorlde
PCB Oils
Jet Fuels
Mixed
Organlcs
Solid Waste
Landfill
Leachate
Phenols
Uranium HI II
Tal 1 Ings
Methane Gas
Landfill
Leachate
Acid Mine
Drainage
Radloact 1 ve
On-going
Completed
Completed
Completed
Completed
On-going
Completed
Completed
Completed
Completed
Completed
Completed
*These walls range from 2 to k ft In width 6 15 to 100 ft In depth.
191
-------
192
BARRIERS
avoid chemical interactions which could hinder the performance of
the slurry. When introduced into the trench, the slurry should be
fully hydrated with an "apparent viscosity" of from 15 to 20 cp
and a density of from 64 to 94 lb/ftj 6>? (1030 to 1500 kg/mj).
As stated earlier, the primary role of the slurry is to keep the
trench open during excavation. A number of mechanisms have
been suggested for trench support, including hydrostatic force ex-
erted by the slurry, formation of a filter cake on the trench sides,
and compaction of the earth on either side of the trench during ex-
cavation.8 Of these three, the first two are the most important and
are closely related.8'9
In order for the force exerted by the weight of the slurry to sup-
port the trench, the force must be exerted on the soil grains and
not just on the soil pore fluids. The formation of a low permeabil-
ity filter cake on the trench sides, caused by water loss from the
slurry into the ground, eventually seals off the soil pores, allowing
the hydrostatic force of the slurry to be transferred more directly
to the sides. Filter cake formation a function of clay type, clay con-
centration in the slurry, and formation and pressure. Permeabilities
of filter cakes have been measured as low as 2.3 x 10" ® cm/sec.5'9
Additional trench support can be expected by the plastering effect
of the filter cake holding the soil grains together, and by gelation
of the slurry in the soil pores into which it infiltrates.1 The contri-
bution of the filter cake to the final cut-off wall is examined later in
this paper.
BACKFILL MATERIALS AND FUNCTION
As noted earlier, there are three major types of slurry trench
backfill: soil-bentonite, cement-bentonite, and concrete. The
choice of backfill is dependent on specific site characteristics and
the requirements of the completed cut-off wall. Each of these is
discussed below.
Soil-Bentonite Cut-Off Walls
Soil-bentonite (SB) cut-off walls are often backfilled using the
soil materials excavated from the trench. These soils, or soils with
more favorable properties borrowed from another source, are
mixed with small amounts of slurry from the trench until they are
a homogenous paste. The recommended consistency for the back-
fill is a mixture that will stand on a 10:1 slope (as .measured by
slump cone) with a water content of 25% to 35%. The minimum
amount of bentonite in the backfill mixture will range from 0.5%
to 2% depending on the initial water content of the excavated
soil.1
The permeability of the backfill is very much dependent on the
gradation of the soil used, as well as on the bentonite content and
initial backfill water content. The lowest permeabilities are ob-
tainable when the backfill is well graded and contains an apprec-
iable amount of (i.e., 10-20% or more) of fines (particles passing a
200 mesh sieve) and especially plastic fines.' The extent to which
o 10
10 '• «r7 10
SB Backfill Permeability, cm, sec
Figure 1.
Permeability of Soil-Bentonite Backfill
Related to Fines Content
(After D'Appolonta and Rvan 19791
10'4
10-
10-
u
.="
u>
•B ,0-6
I
I""
10-"
10-
Well Graded
Coarse Gradations
(30-70% + 20 Sieve!
w/10 to 25% NP Fines
Poorly Graded
Silly Sand w/
30 to 60% NP Fines
Clayey Silty Sand
w/30 to 50% Fines
12345
% Bentonite by Dry Weight of SB Backfill
Figure 2.
Relationship Between Permeability and
Quantity of Bentonite Added to SB Backfill
lAfter O'Appolonia and Ryan 19791
fines and plastic fines can affect backfill permeability is shown in
Fig. 1 while the contribution of added bentonite to the permeabil-
ity of the backfill is shown in Fig. 2.
In cases where the excavated sojls are not well suited for backfill,
borrowed soils may have to be added to meet the design criteria.
For example, if low compressibility is called for in the design, and
the trench materials lack coarser particles, additional sand and
gravel may be needed. If low permeability is of greater concern
30
70
60
SO
40
30 -
20
10
0
• Plastic Fines
O Non-Plastic Fines
I-D Compression
Stress Increment
0.5 to 2.0 kg/cm2
.02
.04
.06
08
.10 .12
• Plastic Fines
O Non-Plastic Fines
. /»
Isotroplc Compression
Stress Increment
0.5 to 0.2 kg/ cm2
02 .04 06 08
Compression Ratio. Cc/1 + Cg
Figure 3.
Compressibility of SB Backfill
Related to Fines Content of Mix
(After D'Appolonia. 19801
.10
12
-------
BARRIERS
193
than low compressibility, addition of borrowed coarser materials
would be unnecessary. If fines are lacking in the trench soils, and
low permeability backfil ( £ lxlO~7 cm/sec) is called for, borrowed
material could be required. Bentonite can make up only a portion
of the plastic fines in the backfill mixture due to its high compress-
ibility. In brief, gradation of the backfill is a trade off between
permeability and compressibility as illustrated in Figs. 1 and 3.
The permeability of the final wall is a function of both the filter
cake permeability and the backfill permeability. Because the back-
fill is less permeable than the surrounding soil it is possible that
the downgradient filter cake is forced into the soil matrix under
the hydraulic gradient across the wall. This gradient would act to
force the upgradient filter cake into the backfill, but given proper
backfill gradation (i.e., sufficient fines) it would remain intact.s
The relative contribution of the backfill and one filter cake to the
final wall permeability is illustrated in Fig. 4. This figure also shows
that the contribution of the filter cake is greater for more perme-
able backfill and vice versa. It also illustrates that, even with very
permeable backfill, the maximum permeability of a SB cut-off wall
will be on the order of lxlO~6 cm/sec due to the contribution of
the filter cake.
Cement-Bentonite Cut-Off Walls
Cement-bentonite (CB) cut-off walls are of two major types:
(1) self hardening slurries, which both support the trench and are
10-
=- 10-
ID
o
10'" -
k Wall Permeability
kc Cake Permeability
k(, Backfill Permeability
tb Backfill Thickness
tc Cake Thickness
/
^
t£^
^
Wall Thickn
tb
(- * -)
\kb Kc /
i...
—
ess = 80 cm
10-9/sec
15 x 1
-------
194
BARRIERS
D
STAGE 1
)
STAGE 2
1 (j |— — Guide Trench
Excavate panel 1 and fix stop end tubes.
' c n 3 a
Ditto panel 3 white concreting panel 1
and withdrawing slop end tubes.
) ' C * ) 3 C
STAGE 3
Ditto panel 3 while excavating panel 2.
) ' ( 2 ) 3 (
STAGE 4
Concreting panel 2.
standardized or applied throughout the industry. In general, the
tests include:
Figure 6.
Typical Sequence Used in Construction of
Diaphragm Walls (After Nash, 1974)
CB walls with respect to permeability and chemical resistance
provided that the joints are sound. Since they are seldom used for
pollution migration control, they are not considered further here.
COMPATIBILITY
The presence of organic or inorganic compounds in the ground-
water can have a detrimental effect on the bentonite slurry used
during wall construction as well as the ability of the finished wall
to contain pollutants. These chemicals can affect the physical/
chemical properties of the bentonite and the backfill material, and
can lead to failure of the wall either during construction or during'
its operational lifetime. Thus, before a slurry wall is considered as
an appropriate remedial response, the effects of the leachate on
the bentonite slurry and the finished wall must be determined. This
testing procedure will provide the information that is necessary to
properly select the grade of bentonite and the type of backfill ma-
terial that should be used in the slurry wall construction.
Chemicals can affect the physical and chemical properties of the
bentonite and backfill material, leading to:
•Flocculation of the slurry
•Reduction of the bentonite's swelling capacity
•Structural damage to the bentonite or backfill material
In general, flocculation of the bentonite slurry can be caused by
chemicals, such as electrolytes and metals, that reduce the natural
repulsive forces between the hydrated clay particles. From a prac-
tical standpoint, the exact mechanisms are unimportant. However,
if the slurry in the trench flocculates, due to chemicals in the make-
up water or the soil, it can lose its ability to support the trench,
leading to collapse." This problem can be avoided by adequate
site investigation and testing.
Numerous organic and inorganic compounds can, through a
variety of mechanisms, cause bentonite clay particles to shrink or
swell. All of these mechanisms affect the quantity of water con-
tained within the interspacial layers of the clay structure. Many
chemicals can reduce the double layer of partially bound water sur-
rounding the hydrated bentonite, thus reducing the effective size of
the clay particles.l4'15 This effect can increase the permeability of
the backfill. Proper testing and backfill composition, however, can
prevent reduced swelling from having a detrimental effect on the
final cut-off.14
Strong organic and inorganic acids and bases can dissolve or
alter the bentonite or soil in the backfill material, leading to large
permeability increases." Aluminum and silica, two of the major
components of bentonite, are readily dissolved by strong acids or
bases, respectively. For example, when four types of clay were ex-
posed to acetic acid, significant soil piping occurred due to the soil
components being dissolved by the acid." This condition leads to a
significant increase in permeability in all of the clay types. Destruc-
tion of the proposed backfill mixture would be revealed during
testing and could eliminate certain backfills from consideration.
Compatibility testing is done using actual site leachate or ground-
water and the proposed slurry and backfill. The tests are not all
•Apparent slurry viscosity
•Filter press test
•Permeability
Viscosity testing indicates the ability or inability of the slurry to
form a thixotropic gel in the presence of the contaminants. The
filter press test gives a quick indication of the ability of the slurry
to form a filter cake in a contaminated trench. Together, these
tests can tell the design engineer whether problems with trench
support might occur.
Permeameter tests of the proposed backfill mixture, using the
site leachate as the permeant, are essential in determining the ex-
pected cut-off effectiveness. Following wall installation, perme-
ability testing of actual backfill can be used to see'if the design
permeability was achieved. Care must be taken in collection of the
leachate to ensure that the samples used in the tests are represen-
tative of the site conditions.
SLURRY WALL APPLICATIONS
Slurry walls have been, and are being, applied to a variety of
waste site situations. The composition, dimensions, and place-
ment of the walls are, of course, designed on a site specific basis.
As noted above, wall composition is dependent on a number of
factors including site lay-out, site chemistry and strength require-
ments.
The dimensions of a slurry wall are essential design parameters,
especially for depth and length. Wall width is largely a function
of the excavation equipment used, and generally ranges from 0.6
to 1.6 m.14 The depth of the wall is most dependent on the be-
havior of the contaminants and the site stratigraphy. In most situa-
tions, it is necessary to key the wall into a low permeability
underlying strata (aquiclude) such as bedrock or a clay layer. In
some cases, as with floating contaminants such as petroleum pro-
ducts, a so-called "hanging wall" can be used. Hanging walls are
not keyed into an aquiclude but are placed only deep enough to in-
tercept the top several meters of the water table, thereby allowing
capture of floating contaminants.'7
The length and shape (placement) of a slurry wall are largely de-
pendent on the location of the pollution source and any resultant
contamination plume, as well as the intended role of the wall. In
general, a wall can be placed surrounding a source, or upgrad-
ient or downgradient from the source with respect to groundwater
flow.
Upgradient placement can divert groundwater flow around a
pollution source, thereby protecting that groundwater from con-
tamination and reducing the amount and rate of leachate genera-
tion (Fig. 7). Care must be taken in the design to ensure that
groundwater does not flow around the wall ends or rise to overtop
the wall.
Groundwater Flow
Slurry Wall
Figure 7.
Upgradient Wall Placement
-------
BARRIERS
195
Groundwater Flow
Extraction Wells
• Slurry Wall
Figure 8.
Downgradient Wall Placement
Groundwater Flow
Slurry Wall
Extraction Wells
Figure 9.
Circumferential Wall Placement
Downgradient placement, shown in fig. 8, can be used in situa-
tions where the amount of flow into a source is very low. In this
practice, the slurry wall serves to capture the contamination plume
so that it can be collected (by wells or drains) and treated. Care
must be taken to ensure that contaminated water does not have a
destructive effect on the wall (olr is kept from contact with it), and
that the contamination does not over top the wall.
Circumferential wall placement is employed in situations where
most complete containment is desired. As shown in Fig. 9, the
wall surrounds the source and prevents migration into and out of
the site. Again, care must be taken to prevent wall destruction by
thorough compatibility testing and/or by preventing contaminant-
wall contact.
Slurry walls are rarely, however, the sole type of engineered re-
medial measure. Most often, additional measures such as extrac-
tion wells, leachate collectors, and surface seals are used in con-
junction with slurry walls as part of the total remedial effort.
Site Characterization
In order to design and install a successful slurry wall, exten-
sive site characterization, especially with respect to subsurface con-
ditions, is necessary. Site investigations for construction pro-
jects are frequently accomplished in two phases, a preliminary
phase, and a design phase. A preliminary investigation will re-
veal many of the physical constraints that would affect slurry
trenching operations. Among the most important considerations
are:
•Topographic restrictions
•Buried or overhead obstructions
•Equipment access
•Sufficient work area
These factors, and their interrelationships must be identified and
considered very early in the design phase. Additional, general in-
formation will be collected on the subsurface conditions, which
will allow the design engineer to select the most likely construction
method(s).
The in-depth or design phase investigation will be more involved
and more narrowly focused, centering on developing design param-
eters. For slurry walls, the principal focus will be on complete geo-
logic and geohydrologic and, in some cases, geophysical charac-
terization of the site. This is accomplished through a series of soil
borings, initially general in nature and becoming more detailed as
variability and anomalies are assessed. Among the most important
parameters produced by this investigation are:
•Soil distribution and nature
•Aquifer geometry and gradients
•Groundwater and leachate chemistry
The accuracy and completeness of the design phase investigations
are of prime importance to the success of a slurry wall project and
should never be under estimated. Thorough site characterization
can allow problems to be anticipated and planned for before they
become detrimental to the project.
SLURRY WALL CONSTRUCTION
The first phase of slurry wall construction is to prepare the site
for work. This will vary considerably in nature and extent from
site to site, but will involve such tasks as leveling the trench line,
locating and/or moving overhead and underground obstructions,
identifying storage areas and haul routes, and constructing im-
poundments for slurry hydration and storage. In cases where
groundwater is very near the surface, a construction platform is
often built along the trench so that the slurry can be kept well above
the water table. Extra care in site preparation can help maintain
construction schedules by identifying problems and planning for
their resolution.
Excavation of the trench is accomplished using either a hydraulic
backhoe or a cable or Kelly Bar mounted clam-shell grab. Back-
hoes are usually used for depths less than 16m; however, in re-
cent years, modified or "extended stick" backhoes have reached
73 ft. Below these depths, clam-shell grabs have successfully in-
stalled walls over 94m.17 If a linear wall is being installed, back-
filling can begin as soon as excavation has reached the point be-
yond the calculated backfill slope (Fig. 10). For a continuous, cir-
cumferential wall, however, the entire trench is often excavated
before backfilling is begun.
Slurry is usually prepared on the site as close as practical to the
trench. The bentonite is weighed and mixed with water in a high
shear mixer. From the mixer, it is pumped to a pond to hydrate
fully under circulation. The hydration time is a function of benton-
ite quality, mixer shear, and pond circulation, and van vary con-
siderably. From the hydration pond, the slurry is usually pumped
to a storage pond before introduction into the trench. This proced-
ure allows new slurry to be mixed and hydrated and provides a re-
serve supply of slurry in case the slurry level in the trench should
drop suddenly when an unexpected previous zone is encountered.
All slurry trenching operations are accompanied by a simple mo-
bile laboratory for testing new bentonite shipments, and for test-
ing the slurry at various times during construction. The slurry is
generally tested for viscosity (Marsh), density (mud balance), sand
content, and pH. In many cases, filter loss and gel strength are
also measures. The testing of the slurry is usually done after mix-
ing, as it is introduced into the trench, and again as backfill is
placed (a sample is taken from the trench bottom near the leading
edge of the backfill). If the slurry in the trench is found to be too
dense, additional new slurry is added to bring it back within specif-
ications.
Backfill for a SB wall is usually mixed alongside the trench us-
ing a bulldozer. If additional soil or dry bentonite is needed it is
spread evenly over the trench spoils and tracked and bladed with
the bulldozer. During the mixing, slurry from the trench is added
to the backfill mixture to form a homogenous, plastic paste. When
the backfill has reached the specific composition and consistency,
it is ready to be placed in the trench.
-------
1%
BARRIERS
(After Rvan. 19801
Figure 10.
Typical SB Wall Construction
Backfill placement is begun using a clam-shell grab to lower the
backfill to the trench bottom. This continues until a slope of back-
fill extends to the surface (Fig. 10). Thereafter, the backfill is
pushed into the trench with a bulldozer and allowed to flow down
the slope. If excavation and backfilling are well coordinated, the
level of slurry in the trench should not fluctuate drastically.
As with any construction operation, slurry trenching requires
that certain quality control procedures be followed. These that cer-
tain quality control procedures be followed. These procedures may
differ from job to job, but are aimed at ensuring trench con-
tinuity (width and depth), backfill composition and placement, and
slurry quality.''
Trench continuity is extremely important to the finished wall be-
cause unexcavated material or improper key-in can greatly affect
final permeability. This is mainly checked by observing the ex-
cavation operation and the material being brought to the surface.
In shallow excavations, it is possible to sound the trench using a rod
to "feel" for unexcavated material.
Quality control of backfill mixing and placement is second in
importance to excavation but is essential nonetheless. Mixing is
checked mostly by observation but samples are taken for labor-
atory analysis at various times during the job to determine whether
they meet specifications. The control of backfill placement is
focused on the initial backfill that is often placed with a clam-
shell. Care must be taken to ensure that no pockets of slurry are
trapped in the backfill.
Control of the slurry properties often receives the most atten-
tion but is not of extreme importance to the final wall. Control
procedures, outlined above, center on ensuring that the slurry is
of sufficient density to adequately support the trench, but not so
dense that it is not easily displaced by the backfill. Evidence in-
dicates that as long as the slurry is at least 240 kg/m3 less dense
than the backfill, no problems should occur.5
The cost of a completed slurry wall will vary a great deal de-
pending on depth, ease of excavation, and materials involved. The
range of costs of SB and CB walls for different depths and media
is shown in Table 2.
CONCLUSIONS
The use of slurry walls for controlling the migration of pollution
from waste sites has become a popular alternative. The evidence
gathered to date indicates that a properly designed slurry wall, in-
stalled by a competent contractor, can be effective in long-term
pollution control.
Unfortunately, most of the slurry wall installations to date have
been in the private sector, from which little monitoring data are
available. Until such data are assembled and critiqued, it remains
to be seen just how effective, and how long term this remedial
measure is in controlling the spread of contaminated groundwater.
Table 2.
Approximate Slurry Wall Costs as a Function of Medium and Depth. (In 1979 Dollars per Square Foot) (After Ressi di Cerva).
SOFT TO MEDIUM SOIL
N < 40
HARD SOIL
N 40-200
OCCASIONAL BOULDERS
SOFT TO MEDIUM ROCK
N > 200 SANDSTONE, SHALE
BOL'LDCR STRATA
HARD ROCK
GRANITE, GNEISS, SCHIST*
SOIL-DENTONITE
WALL
DEPTH
< 30
FEET
2-4
4-7
4-8
6-12
15-25
DEPTH
30-75
FEET
4-8
5-10
5-8
10-20
15-25
—
DEPTH
75-120
FEET
8-10
10-20
8-25
20-SO
50-80
___
CEMENT-BENTONITE
WALL
DEPTH
< 60
FEET
15-20
25-30
20-30
50-60
30-40
95-140
DEPTH
60-150
FEET
20-30
30-40
30-40
6C-85
40-95
140-175
DEPTH
> 150
FEET
30-75
40-95
40-85
85-175
95-210
'175-235
••NOMTNAL PKNFTRATrnN ONLY
POR STANnAnn RF TNFORrrnrNT IN SIUKFV WALLS Ann S8.no PFR so. FT.
FOP CON^TPIirTTON TN MRR/VN FNV1PONVFNT Aim ?S7 TO W nf r-rit"-
-------
BARRIERS
197
REFERENCES
1. Nash, J.K.T.L. "Slurry Trench Walls, Pile Walls, Trench Bracing,"
Sixth European Conference on Soil Mechanics and Foundation En-
gineering, Vienna, March 1976.
2. Grim, R.E., "Clay Minerology," McGraw-Hill Book Co., New York,
N.Y., 1968.
3. U.S. Army Corps of Engineers. Foundation Report. Design, Con-
struction, and Performance of the Impervious Cut-Off at W.G. Hux-
table Pumping Plant, Marianna, Arkansas. Volume I. April, 1978.
4. Ryan, C.R., "Slurry Trench Cut-Offs to Halt Flow of Oil-Polluted
Groundwater." Presented at: American Society of Mechanical Engi-
neers, Energy and Technology Conference and Exhibition, New Or-
leans, LA., Feb. 1980.
5. D'Appolonia, D.J., "Soil-Bentonite Slurry Trench Cut-Offs,"
Journal of the Geotechnical Engineering Division, ASCE 106, No.
GT4, April 1980.
6. LaRusso, R.S., "Wanapum Development Slurry Trench and Grouted
Cut-Off," Grouts and Drilling Muds in Engineering Practice, But-
terworth's, London, 1963.
7. Ryan, C.R., "Technical Specifications, Soil-Bentonite Slurry Trench
Cut-Off Wall," GEO-CON Inc., undated.
8. Veder, C., "Excavation of Trenches in the Presence of Bentonite
Suspensions for the Construction of.Impermeable and Load-bearing
Diaphragms," Grouts and Drilling Muds in Engineering Practice,
Butterworth's, London, 1963.
9. Nash, J.K.T.L., and Jones, O.K., "The Support of Trenches Using
Fluid Mud," Grouts and Drilling Muds in Engineering Practice, But-
terworth's, London, 1963.
10. Jefferis, S.A., "Bentonite-Cement Slurries for Hydraulic Cut-Offs,"
Proc. of the Trench International Conference on Soil Mechanics and
Foundation Engineering, I, Rotterdam, June 1981.
11. Ryan, C.R., "Slurry Cut-Off Walls; Methods and Applications,"
Presented at GEO-TEC '80, Chicago, IL, Mar. 1980.
12. Nash, K.L., "Diaphragm Wall Construction Techniques," Journal
of the Construction Division, ASCE, 100, No. C04, Dec. 1974.
13. Xanthakos, P.P., "Slurry Walls," McGraw-Hill Book Co., New
York, N.Y. 1979.
14. D'Appolonia, D.J., and Ryan, C.R., "Solil-Bentonite Slurry Trench
Cut-Off Walls," Presented at Geotechnical Exhibition and Tech-
nical Conference, Chicago, IL, Mar. 1979.
15. Anderson, D., and Brown, K.W., "Organic Leachate Effects on
the Permeability of Clay Liners," Land Disposal: Hazardous
Wastes, Proc. of the Seventh Annual Research Symposium, USEPA,
MERL, Cincinnati, OH, 1981.
16. Anderson, D., Brown, K.W., and Green, J., "Effects of Organic
Fluids on the Permeability of Clay Soil Liners," Land Disposal:
Hazardous Wastes, Proc. of the Eighth Annual Research Symposium,
USEPA, MERL, Cincinnati, OH, 1982.
17. Case Slurry Division, Case International Co., "Case Slurry Wall
Construction Notebook," Roselle, IL, undated., 139p.
18. Ressi di Cervia, A.L., "Economic Consideration in Slurry Wall
Applications," Proceedings of a Symposium on Slurry Walls for
Underground Transportation Facilities, Cambridge, MA, Aug. 1979.
-------
GELATINOUS SOIL BARRIER FOR REDUCING
CONTAMINANT EMISSIONS AT WASTE DISPOSAL SITES
B.E. OPITZ
W.J. MARTIN
D.R. SHERWOOD
Battelle Pacific Northwest Laboratory
Richland, Washington
INTRODUCTION
The disposal of hazardous wastes, such as sanitary landfill
leachate, chemical waste, or radioactive waste, is an increasing pro-
blem. Current practices call for waste disposal in earthen pits or
repositories. However, these waste contaminants may eventually
migrate so that potentially hazardous elements could be
transported to the biosphere. Thus, reducing the transport of
chemicals from waste disposal sites poses an immediate challenge.
Under sponsorship of the Department of Energy's Uranium Mill
Tailings Remedial Action Program (UMTRAP), Pacific Northwest
Laboratories (PNL) has investigated the use of engineered barriers
for use as liners and covers for waste containment. A comparison
of clay-soil mixtures and a gelatinous soil additive was performed
to evaluate their use as a cover in reducing the escape of radioactive
radon gas from abandoned uranium mill tailings piles. In addition,
permeability measurements were performed to examine and com-
pare the ability of gelatinous materials to reduce leachate move-
ment and the migration of contaminants such as trace or heavy
metals. The results of these studies led to the development of a low
permeable, multilayer earthen barrier for effectively reducing con-
taminant emissions from waste disposal sites.
COVER SCHEME
Every potential application of the gelatinous-soil material
depends on site-specific conditions, such as the type of waste,
remedial action requirements, and climatic conditions. However,
using thinner but better engineered cover schemes (Fig. 1) can
reduce the material costs that would be required to meet or exceed
the expectations of currently proposed waste control strategies
(i.e., 3 m or more thick).
The system consists of a capillary barrier and topsoil or over-
burden layer overlying the gel-soil mix barrier. The capillary barrier
is comprised of a layer of crushed rock (minimum conditions) and
FINAL GROUND SURFACE
r°-?vORAyil CAPILLARY BARRIER v"c)io?rj
a5^^:KSX^^^^^^Ki^&^'^y
Figure 1.
Multilayer Cover Scheme
is used to isolate the moist gelatinous-soil barrier layer from a drier
topsoil or overburden layer. By isolating the two layers, "wicking"
from a moist layer to a dry layer is prevented, helping to maintain
the moisture content of the gelatinous-soil layer. In areas with high
rainfall, the capillary barrier may not be needed if the moisture
content of the overburden layer is high enough to prevent wicking.
The topsoil or overburden layer is added: (1) for revegative pur-
poses, and (2) to prevent freezing and subsequent thawing of the
gel-soil barrier layer. With selective revegative efforts, control of
the infiltration of moisture through the topsoil layer will be enhanc-
ed by evapotranspiration. Depending on the location of the waste
site, the topsoil layer would be thick enough to ensure that the
gelatinous-soil barrier is below the frost line of the area. This would
eliminate the possibility of cracking due to freezing and thawing of
the moist barrier layer.
Several factors control the movement of gases and/or liquids
through porous geologic media: total porosity, air-filled porosity,
moisture retention, bulk density, permeability (hydraulic conduc-
tivity) and gelatinous precipitants. For example, the total porosity
of the gel cover ranges from 20 to 30%, as compared to 40% for
clay. This reduction decreases the total amount of void space
within the confining layer of the cover material available for gas or
liquid movement. Similarly, the air-filled porosity of the gelatinous
soil layer is near zero, limiting the space available for gas or liquid
transport. In addition, the hygroscopic constituents within the gel
cover layer retain moisture even at very high capillary pressures,
thus assuring low gas transport even under relatively dry condi-
tions.
Bulk densities greater than 2.1 g/cm3 have been achieved in the
field using standard road construction equipment to compact the
gel layer. These density values increase structural stability and
longevity. Gel-soil layer columns tested in the laboratory had
permeabilities on the order of 10~8 cm/sec with distilled water, thus
allowing even lower flow rates than recommended by USEPA (less
than 5 x 10 ~7 cm/sec).
RADON GAS ATTENUATION STUDY
During the normal operation of a uranium mine thousands of
tons of uranium containing ore are processed daily. The extraction
process, which is specific for the removal of uranium, does not alter
the concentrations of the radioactive uranium daughter products
which have come to equilibrium over many thousands of years.
During the life of the mill, the concentration of these daughter pro-
ducts is constantly increasing as more and more of the spent feed
material containing the members of uranium's decay chain are be-
ing expelled to the mill tailings disposal site.
Eventually, the concentration of the daughter products in the
mill tailings is many times greater than the normal concentrations
found in nature. As a result, the uranium mill tailings themselves
could remain radioactive indefinitely due to the long lived uranium
decay products such as 230Th (half-life of 80,000 yr) and ^Ra (half-
life of 1,620 yr). Neither one of these isotopes is significantly
198
-------
BARRIERS
199
removed during the uranium extraction process; in fact, greater
than 90% of the radium content originally present in the ore goes to
the uranium mill tailings. The 238U decay chain with the correspon-
ding half lives and emissions down to stable 206Pb is shown in Fig.
2.
ALPHA
«.S x II9
BETA BETA ALPHA
YR 24.1 DAY 1.1 WIN 2.5xlOSYR
ALPHA ALPHA ALPHA ALPHA
8.0 x 10* YR 1620 YR 3.8 DAY 3.05 MIN
BETA
26.I MIN
BETA
19.7 MIN
ALPHA
l.t x IO-« SEC
BETA
22 YR
BETA
S.O DAY
ALPHA
HO DAYS
Figure 2.
Uranium Decay Chain
Radium in trace amounts is found in all soils and water. Since
there is such a high residual 226Ra content in the tailings, on the
order of several hundred times the average background concentra-
tion, one of the major consequences associated with uranium mill
tailings is the tailings pile acting as a source for the evolution of
radioactive 222Rn gas.
Radon poses a serious environmental and health impact to
populations acting as a source of radiation exposure. The im-
mediate ratioactive decay product of radon, 238Po, is a solid which
can adhere to soil or aerosol particles. Upon inhalation these par-
ticles can remain lodged in the lungs and may be linked to
pulmonary disorders or lung and respiratory cancer.
Realizing the health impacts of radiation exposure from uranium
mill tailings piles, interim performance guidelines for management
of uranium tailings were established by NRC in May of 1977. One
of the performance objectives for post reclamation of a tailings
disposal site was to reduce the radon flux from the tailings to about
twice the flux in the surrounding environment. The current stan-
dard proposed by the USEPA will allow a tailings radon flux of not
higher than 20 pCi/mVsec. Therefore, the reduction and control of
radon gas emission from a uranium mill tailings disposal site is an
area of major concern in the mining and milling section of the
nuclear industry.
Currently, general practice calls for earthen or other cover
materials to control radon gas emissions. Much of the emphasis has
been on earthen covers of the order of three or mof e meters thick to
reduce emissions below the proposed 20 pCi/mVsec standard set
by the USEPA. Many of these proposed strategies for containing
gases (i.e., synthetic covers, deep lake disposal, or thick earthen
covers up to 10 m thick) have been either too expensive or difficult
to install or lacking in long term stability.
The thinner covers offer the potential for significant savings in
remedial actions with the possibility of improved long term radon
control. However, the development and use of thinner engineered
covers requires a better and more complete analysis capability.
Field tests of multilayer earthen covers for radon control have
been conducted for the past 2 years at Grand Junction, Colorado.
The initial tests conducted in 1980 used relatively thin (less than 20
cm) compacted layers of clay mixed with gravel to limit radon
escape from tailings piles. Results indicated that compacted
clay/gravel layers have been relatively effective in reducing radon
fluxes (greater than 98% effective) but during the first year flux
levels at all monitoring points were observed to exceed the USEPA
limit then of 2 pCi m~2 sec~'.
The 1981 field test was designed to test multilayer earthen
systems that could reduce surface radon fluxes to less than 2 pCi
m~2 sec"1. The materials used were bentonite clay/gravel, benton-
ite/lime/gravel, and alum/lime/gravel mixes which were separately
incorporated into the multilayer scheme shown in Fig. 1.
Initial Radon Flux Measurements
Radon flux measurements were made using the aluminum tent
system described by Hartley et al.10 A pressure balanced system cir-
culates air over the tailings surface and the radon is swept into an
activated charcoal collector.
Tests were run to determine the effectiveness of activated char-
coal as a radon trap at ambient temperatures. The results of those
tests showed that using 400 g of an activated charcoal at a flow rate
of 2 1/min through a 4.75 cm diameter convolated steel tube, the
activated charcoal was greater than 99.9% efficient in collecting
radon at temperatures up to 50 °C. Therefore, a radon measure-
ment system was designed around a 2 1/min flow rate with 400 g of
activated charcoal at ambient temperature (approximately 27 °C).
The field radon measurements consisted of measuring the radon
flux both before and after the multilayer barrier system was applied
to the test area and consisted of the following steps:
•The tent was placed in the predetermined test spot and sealed to
the surface by packing tailings or overburden around the edge of
the lip of the tent dependent on whether p're- or post-application
measurements were being made.
•The flow rate was adjusted to 2 1/min and the radon flux meas-
urement was taken for 4 hr from the time the flow starts through
the activated charcoal radon trap.
•After the measurement was taken, the activated charcoal was
placed in 2.5 cm x 15 cm diameter petri dishes and subsequently
counted using an intrinsic germanium diode or Nal detector and
a multichannel analyzer. The count rate from the 609 Kev peak
of Bi-214 was used in calculating the radon flux (J). The equation
used was:
J (pCinT2 s"1)
where
E
A
C
•y
( 104 \
(TTTK)
(1)
counting efficiency of detector, count/disintegration
cross sectional area of tent (cm2)
net count rate of Bi-214 peak
radon decay constant, 2.09 x 10~6 (sec~')
exposure time
12 = time from initial exposure to start of count
13 = time from initial exposure to end of count
Material Preparation
Several clay/gravel and gelatinous soil/gravel mixes were tested
in 1981. The materials were loaded into a portable Calenco model
30 TP 14H pugmill rated at 300 t/hr which proportioned and mixed
the radon barrier materials. The composition of the radon barriers
are given in Table 1. Depending on the barrier materials being mix-
ed, i.e., bentonite or the gelatinous mix, either water or aluminum
sulfate (Alum) solution was required.
Cover Application
Upon completion of the barrier materials mixing and stockpiling,
a 11 m3 paddle-wheel scraper self-loaded the material and
transported it to the test plot. The scraper applied the material in
-------
200
BARRIERS
Table 1.
Gelatinous and Clay Mixtures Used as Radon Control Layers
at the 1981 Grand Junction Field Test
INFLUENT PORT
Plot
15 cm thick
Gelatinous
40 cm thick
30 cm thick
Material
1.9 cm minus roadbase
topsoil
lime
alum
1.9 cm minus roadbase
bentonite (CS-50)*
lime
1.9 cm minus roadbase
bentonite (CS-50)®
Wt
74.4
13.1
4.3
8.1
86
12
2
88
12
(dry)
• American Colloid, Skokic, IL.
layers ranging from 7 cm to 15 cm. Following application of each
lift a Bomag model BW 210 PD vibratory tamping foot compactor
made several passes to blend the added moisture and compact the
barrier materials. A Bomag model BW 220A smooth drum com-
pactor completed the compaction process.
The final compacted barrier thicknesses were: 15 cm for the
gelatinous material, 30 cm for the bentonite/gravel mix and 40 cm
for the bentonite/lime/gravel mix. Average compaction densities
for the barriers were 1.96 g/cm3 for the gelatinous mix, 1.96 g/cm3
for the bentonite/gravel mix and 1.92 g/cm3 for the bentonite/
lime/gravel mix.
The final phase of the multilayer earthen cover included the
capillary barrier and overburden application.
Post Cover Radon Measurements
Radon flux measurements after application of the barrier layers
were made in the same manner as the initial measurements except
the aluminum tests were sealed to the measurement site with the
surrounding overburden material in place of tailings.
Results of the radon flux measurements show a substantial
reduction in radon emanation. Values of 96% to greater than 99%
reduction of radon gas from the tailings were achieved by the
multilayer clay seal and the inorganic gelatinous soil barriers. The
difference in effectiveness of the two types of barriers, clay based
or gelatinous based, can be seen in the layer thickness required to
achieve the results. The gelatinous soil barrier layers thickness was
15 cm thick while the bentonite and bentonite/lime based layers
tested were 30 cm and 40 cm thick, respectively. For relatively iden-
tical radon flux reductions the gelatinous soil barrier required 50 to
60% less material dependent on which clay barrier is used for the
comparison. Preapplication radon flux values and a number of
post application radon flux values are shown in Table 2.
HYDRAULIC CONDUCTIVITY STUDIES
The hydraulic conductivity (K) is qualitatively defined as the
ability of soil to transmit water. Column experiments used in deter-
mining hydraulic conductivity followed the ASTM method for
determining constant head permeability with some minor modifica-
tions (Fig. 3). The flow of solution was from the bottom to the top
to ensure saturated flow through the sediment.
These column experiments can be used in predicting the long-
term effect of sediment-tailings solution interaction. The rate of
flow is then monitored and hydraulic conductivity calculated using
Eq.2:
k (cm/sec) =
AQ(cm3) xL(cm)
(2)
where
AQ =
L =
AT =
A =
H =
AT (days) x A(cm)2 x H(cm) x 86,400 sec/day
change in effluent
length of the permeameter cell
change in time
cross-sectional area of the permeameter cell
amount of head
REGULATOR
VALVES
FLOW DIRECTION
CAS PRESSURE SOURCE
PERMEAMETER
CELL
SEDIMENTS
Figure 3.
Illustration of the Apparatus Used for Determining Hydraulic Conductivity
Three permeameter cells were packed with the gelatinous soil
materials. The characteristics of each cell are listed in Table 3. The
permeameter cells were then contacted with distilled water to ac-
quire a constant hydraulic conductivity and to saturate the
gelatinous material. Once saturated, they were transferred from
one apparatus (Fig. 3) to a similar apparatus which contains tailings
Table 2.
Summary of Radon Flux Measurements Made at Grand Junction
Before and After Cover Application (PNL Quarterly Report 9/82)
pCi m-2 s-1
Multilayer
Barrier
15 cm
Gelatinous
30cm
Bentonite
40cm
Bentonite/lime
Before After Application
July 1981 Oct/Dec Feb 1982 July 1982 Average
Avg ± SD 1981 Avg ± SD Avg ± SD % Flux Re-
Avg ± SD ductloni
367 ± 219 10 ± 11 12 ± 8 6 ± 5 96
175 ± 70 11 ± 8 38 ± 18 7 ± 4 92
240 ± 179 II ± 7 19 ± 8 8 ± 6 93
Table 3.
Characteristics Associated with the Permeameter Cells
Dry Bulk
Cdl Deiully
No. (g/cm')
A 1.96
F 1.92
G 1.88
Porally
0.26
0.28
0.29
Pore
Vol (ml)
25.08
18.72
19.7J
Length
(cm)
4.97
3.0
3.0
Dlimtlcr
(cm)
4.97
5.37
5.37
-------
BARRIERS
201
solution. The chemical composition of the tailings solution is
outlined in Table 4. Triplicate samples of tailings solutions were
analyzed by Inductively Coupled Plasma (ICP) and Atomic Ab-
sorption Spectroscopy for major cations, Ion Chromatography
(1C) for major anions and Intrinsic Germanium Detector for ra-
dionuclides.
The acidic tailings solution contains a large amount of total
dissolved solids at a low pH (~1.9). There was a sharp decline in
hydraulic conductivity in all three permeameter cells after contact
with the tailings solution (Fig. 4). With the buffering of the acidic
Table 4.
Tailing Solution Characteristics (mg/l except as noted)
Parameters
Al
As(a)
Ba
Ca %
CdU
Co a
Cr(a)
Cu
Fe
K
Mq
Mn
Mo(a)
Na %
Pb a
Se(a)
Si
5?.)
NHjW
F
Cl
N03
S04
U-238(c)
Ra-226(c)
Pb-210(c)
Th-230(c)
Tailing Solution
721 + 8
0.54 + 0.05
0.07 TO. 01
557 "6
0.12 + 0.01
2.45 ~0.2
1.29 + 0.2
21.1 TO. 6
1974 + 20
245 ~ 15
1369 ~ 13
1946 + 18
0.17 + 0.01
1073 T 22
1.62 + 0.1
or
292 + 3
11 ~ 0.1
2.99 + 0.27
311 + 11
210 + 6
1064 ~ 7.1
98 + 0.7
22107 f 323
4466 + 95
1784 + 48
8665 ~ 397
136705 + 3001
Detection
Limit
0.1
0.05
0.05
1.0
0.01
0.05
0.05
0.05
0.1
0.2
0.5
0.1
0.05
0.5
0.02
0.15
0.1
0.1
0.05
5.0
0.1
0.3
0.3
0.3
50
50
15
400
(a) Analyses performed by Atomic Absorption
Graphite Furnace
(b) Analyses performed by Nesslerization and
Distillation
(c) Values in pCi/liter
DL = below detectable limits
tailings solution from approximately 1.9 to 9.2 and the reduction in
total dissolved solids from 34.77 to 4.05 g/1, presents evidence that
the reduction in hydraulic conductivity is due to the blocking of
pore spaces by precipitants.
The reactions of importance involving the ability of a soil to
neutralize a given solution is based on the buffering capacity of the
sediment. Buffering capacity of a sediment is defined as the
amount of solution which can be neutralized to a certain pH per
gram of sediment. The calcium carbonate content of a soil gives an
indication of the volume of solution which the sediment can
neutralize. Calcium carbonate buffers the pH by dissolving the
calcium carbonate in acid and forming carbon dioxide and water.
The continued dissolution of calcium carbonate is driven by the
mass action precipitation of calcium sulfate, the evolution of car-
bon dioxide gas, and the increasing hydrogen ion concentration as
additional tailngs solution migrates through the sediments. The
reactions involved in this neutralization can be written simply in the
following equations:
CaC03(s) +
Ca2+ + S0=
C0(g)
CaC03(s) + 2H + S04
CaS04-2H20(s) +
(3)
(4)
(5)
The slight excess of Ca(OH)2 in the gelatinous material, used to en-
sure a complete reaction with the A12(SO4)3 forming the gelatinous
substance. The combination of the aluminum sulfate [AL^SO^j]
and the calcium hydroxide [Ca(OH)2] produce some insoluble gela-
tionous precipitates, aluminum hydroxide [A1(OH)2] and calcium
sulfate (CaSO4) as illustrated in Eq. 6:
A12(S04)3-14H20 + Ca(OH)2*f 2A1(OH)3 + 3CaSCy2H20 + 8H20 (g)
thus enhancing the buffering capacity of the barrier. The following
reaction occurs when contacted with the acidic tailings solutions:
20H"
Ca2+ + 20H~ + 2H+
Ca(OH)
(7)
(8)
(9)
The dissolution of calcium hydroxide aids in the neutralization by
removing H+ ions from solution causing an increase in pH. Since
the solubility of many of the tailings solution constituents are pH
dependent, the change in hydrogen ion concentration causes
precipitation reactions to occur resulting in a reduction of total
dissolved solids content in the column effluent solutions.
K / \CELL '
DISTILLED WATER
CELL A TAILING SOLUTION
, I I I , I I I I I I I I I i I . I , I
PORE VOLUMES (ml)
Figure 4.
Hydraulic Conductivity Versus Pore Volume of the Gelatinous Material
CONCLUSIONS
Based on the results of the cover study the gelatinous soil barrier
and bentonite mixes reduced overall radon fluxes by greater than
90% over initial radon flux values. The gelatinous soil mix tested in
the Grand Junction, Colorado field project was 15 cm in depth or
approximately 50% to 60% less barrier material than the bentonite
clay soil barriers tested. From an economic standpoint, this means
less total material would be required to cover a waste site and could
result in a substantial reduction in remedial action costs.
Hydraulic conductivity measurements on the gelatinous soil mix-
ture show dramatic reductions in flow rates upon contact with the
highly acidic uranium mill tailings liquor. Initial hydraulic conduc-
tivity values of 5 x 10 ~8 cm/s or less were achieved in laboratory
experiments with distilled water. Upon contact with the acidic tail-
ings solution flow was reduced by more than two orders of
magnitude in flow reductions on 0.5 to 4 pore volumes of sediment
solution interaction resulted in hydraulic conductivities on the
order of 5 x 10~10 cm/s.
Some of the proposed strategies for controlling gas and leachate
migration at waste disposal sites have been very expensive, difficult
-------
202
BARRIERS
to install or lack long term stability. However, the use of the
gelatinous-soil mixture as a cover or liner can reduce or eliminate
some of these common problems associated with waste con-
taminants:
•Inaccessible materials. The main component in the multilayer
scheme is comprised of 80 to 85 % standard road base normally
used in the construction of highways. The size distribution of
the roadbase material varies somewhat from state to state but it
is a material-thai is readily available throughout the country.
•Swelling and shrinking. As mentioned earlier, high bulk densi-
ties, low total porosities, and low permeabilities can be achieved
whereas clays are extremely difficult to work with and require
careful control of moisture content. As it swells, it will reduce
the total porosity of the layer. When the moisture content is lower
the clay shrinks and the porosity increases. The gel material being
very hydrogsopic entrains moisture into its pores and tends to
maintain a consistent moisture content. This consistency elimin-
ates any cyclic shrinking or swelling and gives a layer with very
predictable properties.
•Ease of handling and application. In many cases, clay-soil ma-
terials must be mixed and applied at some moisture content less
than optimum due to nonuniform wetting of the mix. Also, mix-
ing clays at too high an initial moisure content causes the clays to
adhere to the equipment being used. The gel material can be mixed
at the optimum moisture content initially because of a more uni-
form moisture distribution. The gel material will not adhere to
the machinery being used. Since the main component of the ma-
terial is standard road base, large-scale application is achieved
using standard road construction equipment.
REFERENCES
1. Black, C.A., et al., "Methods of Soil Analysis, Part I." American
Society of Agronomy, Inc., Madison, WI, 1965, 210-215.
2. Borrowman, S.R., and Brooks, P.T., "Radium Removal from
Uranium Ores and Mill Tailings." USBM Report of. Investigation
8099, 1975.
3. ERDA, Summary Report—Phase 1> Study of Inactive Uranium Mill
Sites and Tailings Piles. Energy Research and Development Admin-
istration, Washington, D.C., 1974.
4. Federal Water Pollution Control Administration, Disportion and Con-
trol of Uranium Mill Tailings Piles in the Colorado River Basin.
U.S. Department HEW, Denver, CO, 1966.
5. Ford, Bacon and Davis, Phase II Title I Engineering Assessment of
Inactive Uranium Mill Tailings— Vitro Site. Salt Lake City, UT,
1976, 2-13.
6. Freeze, R.A., and Cherry, J .A., Ground Water. Prentice-Hall, Inc.,
Englewood Cliffs, N.J., 1979.
7. Gee, G.W., et al., "Interaction of Uranium Mill Tailings Leachate
with Soils and Clay Liners," NUREG/CR-1494, PNL-3381, Battelle,
Pacific Northwest Laboratory, Richland, WA, 1980.
8. Meites, L., ed., Handbook of Analytical Chemistry. McGraw-Hill
Book Company, New York, N.Y., 1982.
9. Handbook of Chemistry and Physics, 61st Edition, Chemical Rub-
ber Co., Cleveland, OH.
10. Hartley, J.N., etal., "Asphalt Emulsion Sealing of Uranium Mill Tail-
ings." Annual Report. DOE/UMT0201, PNL-3752, Pacific North-
west Laboratory, Richland, WA, 1981.
11. Koehmstedt, P.L., et al., "Use of Asphalt Emulsion Sealants to
Contain Radon and Radium in Uranium Tailings," BNWL-2190,
Pacific Northwest Laboratory, Richland, WA, 1977.
12. Nelson, R.W., Gee, G.W., and Oster, C.A., "Radon Control by
Multilayer Earth Barriers—Modeling of Moisture and Density Ef-
fects on Radon Diffusion from Uranium Mill Tailings." Presented
in the proceedings of the Third Symposium, Fort Collins, CO, 1980.
13. Tanner, A.B., "Radon Migration in the Ground. A Supplementary
Review," Proc. of the Natural Radiation Environment III, 1, 5-56.
-------
TREATMENT OF HIGH STRENGTH LEACHATE
FROM INDUSTRIAL LANDFILLS
R.C. AHLERT
P. CORBO
C.S. SLATER
Department of Chemical & Biochemical Engineering
Rutgers University
New Brunswick, New Jersey
INTRODUCTION
Landfill Leachates
Definition
Landfill leachate is grossly polluted liquid, produced as rain, sur-
face water and groundwater, remove various components from li-
quid and solid wastes. Depending on the wastes deposited at a land-
fill site, leachate may contain various synthetic and biogenic organic
species, bacteria, viruses and toxic chemicals, including heavy metals
and known or suspected carcinogens.12 The water in the aqueous
phase of leachate originates from groundwater infiltration into a
landfill, surface water or precipitation seepage through the site, or
may be inherently associated with wastes and decomposition pro-
cesses.6
Leachate takes the character of a high strength waterwater similar
to a concentrated industrial waste stream. Leachates generated from
industrial and sanitary landfills diplay COD, TOC, SS, TDS and tur-
bidity that exceed measurements of these common pollution
parameters for municipal wastewater (Table 1). In addition to these
high general contaminant values, levels of individual organic and in-
organic contaminants are frequently very high.10
Hazardous Waste Disposal Sites
The USEPA has indicated there are over 32,000 sites nationwide
that contain significant amounts of hazardous waste. As many as
2,000 of these landfills present imminent public health hazards.'
Many of these hazardous waste disposal sites are located in the
northeastern sector of the United States. In fact, nine of the 25 most
dangerous are in New Jersey.13
The major concern with landfill leachate is surface or ground-
water, with attendant public contact and possible impacts on sources
of potable water. Recently, considerable attention has beenjpaid to
Table 1.
Leachate Characteristics
Parameter
pH
COD
TOC
TDS
Turbidity
Color
Conductiv-
ity,
re mhos/cm
Na
Fe
Mg
Ca
Leachates in
this study**
5.5-6.0
23,000-29,000
8,000-11,000
16,000-25,000
75-150 NTU
Brown
Leachates*
(Nat'l Sample)
3.0-7.9
25-41,000
11-8,700
1,450-15,700
Variable
Domestic Sewage*
(med. conc'n)
13,000-18,000
1x10'
IxlO4
IxlOMxlO'
lxlO!
800-2,000
5-1,350
90-678,000
25-453
164-2,500
8.0
500
200
500
Brown
700
150
0.1
30
50
•Taken from reports by Mikucki et a/*, Shuckrow et al." and Clark el a/.'
••All values in mg/1 unless otherwise noted.
actual and potential impacts on humans. Beck' points out that until
the Love Canal incident, in which industrial toxic wastes literally
drove people from their homes, the general public was not greatly
concerned about landfill leachate and did not seek studies to monitor
or treat it.
Detrimental Effects
There is the potential for severe effects of landfill leachates due
to migration to municipal drinking water systems or to private
wells. Although experts disagree on the toxic and/or carcinogenic
effects of synthetic organic chemicals (SOC), they have concluded,
based on animal tests, that organics must be kept below maximum
safe levels in drinking water.7'
Many SOC are chlorinated hydrocarbons or aromatics; in addi-
tion, many pesticides have given positive laboratory results and are
possible carcinogenic substances. SOCs include: vinyl chloride, car-
bon tetrachloride, trichloroethylene, PCBs, benzene and trihalo-
methanes. Inorganic compounds pose a health threat, also. These
include arsenic, barium, cadmium, chromium, lead, mercury,
selenium and silver.11
Renovation
There are processes available to treat leachates generated by in-
dustrial and chemical landfills. Most of the literature on treatment
processes is based on industrial waste and sanitary landfill leachate
treatment. Basically, hazardous waste treatment aims at ultimate
disposal but may only accomplish one or more of the following:
•Detoxification
•Concentration of hazardous constituents in a reduced volume
•Fixation of the waste to inhibit leaching
The treatment approach chosen depends on waste characteristics
and external factors that influence the degree of treatment
necessary. The intent of this overall study is to evaluate and imple-
ment techniques of hazardous waste treatment.
In this paper, the authors describe examples of two very different
and successful means of leachate renovation. These techniques are
reverse osmosis and aerobic biodegradation; both requiring some
pre-treatment.
PRETREATMENT
The high strength complex industrial landfill leachate used in this
research study is the aqueous phase of bulk oil/water leachate mix-
tures. This material was provided by the USEPA Oil and Hazard-
ous Materials Spill Branch. A Gravity Separator was used to
remove much of the bulk oil, leaving some oil dispersed in the
aqueous phase. Some bulk oil remained, however, floated on the
surface of leachate samples. Leachates designated EPA-01, 02 and
03 were obtained at different points and times at the same landfill.
Samples were stored at temperatures below 10 °C, to retard changes
in composition during storage.
As received, samples of aqueous phase contain oil-in-water and
water-in-oil-in-water emulsions and colloidal suspensions, as well
203
-------
204
TREATMENT
as oil adsorbed on organic solids and clay particles that are difficult
to settle. These dispersed oil phases are thermodynamically stable.
Further, they interfere with any attempt to remove or convert
dissolved inorganic salts and organic species. Activated carbon par-
ticles are wetted and macropores are sealed, reducing effective sur-
face and adsorbtive capacity to nil. Ultrafiltration and reverse
osmosis membranes are fouled and blinded very rapidly. Fluxes
quickly drop to zero. Cell suspensions are flocced and settled. The
cell surface area available for substrate transfer is reduced by
clumlping and masked by adsorbed oil. Respiration and growth are
severely retarded. Hence, pretreatment is necessary to eliminate
turbidity due to dispersed oil.
A suitable scheme was developed and adopted as a standard for
all samples and process research investigations. The pH of the
aqueous phase of leachate samples was raised to 12 by addition of
lime. Doses in excess of 6 gm/1 may have been required. Slurries
were centrifuged and/or settled for 10-20 hours and decanted. The
supernatant flue was recarbonated to a pH between 7 and 8. Car-
bon dioxide was used to avoid air stripping and minimize
anomolous volitilization. The recarbonation slurry was settled and
decanted, yielding a second sludge. A final adjustment to a pH of 7
was made with concentrated sulfuric acid. A third settling step was
usually required, to remove a small amount of calcium sulfate
sludge.
Iron, magnesium and manganese were removed quantitatively by
this pretreatment procedure. Turbidity reduction varied with settl-
ing time, but was never less than 95%. Dispersed oil was eliminated
quantitatively, also. However, the oil was only a very small part of
the initial TOC, i.e., 3-5%. After pretreatment, dispersed oil ceas-
ed to have any observable effect on subsequent treatment studies.
REVERSE OSMOSIS
General Description
Reverse osmosis is the separation of one or more solutes from
solution by means of pressure exerted on the solution to force sol-
vent through a semipermeable membrane. Most present work in
reverse osmosis utilizes cellulose acetate membranes that sorb water
preferentialay to ionic species and dissolved matter. The treated
water fraction is known as permeate; the reject species stream is
termed the retentate. 0
Membrane pores of 5 to 20 A in diameter are formed by the
structure of the cellulose acetate matrix. Molecules that do not
orient in a water-like crystalline structure are rejected by the mem-
brane. In addition, the membrane pores exclude molecules with
molecular weights greater than 200 by a simple sieve process,
although this is not the mechanism of rejection. Because of
preferential sorption of water and rejection of ions and most
organic molecules, water of extremely high quality can be produc-
ed.
Normal osmosis occurs when water passes from a less concen-
trated solution to a more concentrated solution through a
semipermeable membrane. Potential energy exists between the two
solutions across the semipermeable membrane. Water flows, due to
this energy difference, from the less concentrated to the more con-
centrated solution until the system is in equilibrium. Pressure on
the concentrated solution will stop transport of water across the
semipermeable membrane, when the applied pressure equals the
apparent osmotic pressure between the two solutions. The latter is a
measure of the potential energy difference between the solutions.
As additional pressure is applied to the more concentrated solu-
tion, water will begin to flow from the concentrated solution to the
less concentrated solution. The rate of transport is a function of
pressure applied to the concentrated solution, the difference in the
absolute osmotic pressure of each solution, and the membrane area
pressurized. The absolute osmotic pressure is the potential energy
between the solution and pure water.
Membrane Performance Parameters
Effectiveness of reverse osmosis in treatment of high strength in-
dustrial wastewater is determined by several process and solution
parameters. Organic removal and membrane fouling are important
features of membrane performance.
Parameters used to measure the efficiency of membrane separa-
tion include ionic strength, conductivity, total dissolved solids,
chemical oxygen demand, and total organic carbon. These
parameters indirectly reflect inorganic ions, dissolved solids and
organic matter removed. The efficiency of operation can be defined
as the percent rejection of these quantities, by comparison of raw
waste to permeate.
Membrane Fouling
In membrane separation of high strength wastewater, the quanti-
ty and quality of permeate may deteriorate due to membrane foul-
ing.2'4 Eykamp defines fouling as a condition in which a membrane
undergoes plugging or coating, by some element in the waste
stream being treated, such that flux is reduced and a buildup on the
membrane surface commences.
Foulants can be classified as sparingly soluble inorganic com-
pounds, colloidal or particular matter and dissolved organics.
Ultimately, if fouling is allowed to continue, effective membrane
service life is diminished rapidly. Therefore, pretreatment processes
must be utilized in high strength complex wastewater treatment.
Equipment
In this study, the authors employed a tubular cellulose acetate
reverse osmosis membrane, manufactured by Abroc, In c. (Wilm-
ington, Mass.). The membrane was 5 ft long and 0.5 in. in
diameter, with a net surface area of 0.65 ft2. The reverse osmosis
system could operate in steady-state (complete recycle) and
unsteady-state modes. Most data were collected in the steady-state
mode; permeate and retentate are recirculated back to the feed con-
tainer. This procedure allowed sampling at all points to quantify
operational effectiveness, but did not diminish the quantity of ex-
perimental wastewater; the retentate stream and feed were iden-
tical. The system was operational basically in an unsteady-state
mode. Permeate and retentate were-separated and collected; the
retentate was concentrated in rejected species. The system was
equipped with a 55 gal, water-cooled feed tank and several staging
tanks, including a pretreatment system.
Procedure and Results
In Experiment 3 EPA-02 leachate pretreated with the standard
lime pretreatment process was used. This experiment was perform-
ed at steady-state, with recirculation of permeate and retentate to
the feed tank. This simiulates operation in the unsteady mode when
the concentration of the retentate does not change significantly in
response to changes in raw waste concentration. Operation was
continued for approximately one week to evaluate flux and
separatiaon efficiency. Separatiaon effectiveness was observed as
TDS and conductivity reductions over the period of the run. TOC
and COD assays were performed to evaluate removal of dissolved
organics. Experiment 4 was a continuation of the previous test. The
system was switched to unsteady operation. Changes in system
parameters were observed and compared to steady-state perfor-
mance.
In Experiment 3, 40 gal of leachate with the following
characteristics were used: 17830 micro-mhu/cm conductivity, 4.9
pH, 16,400 mg/1 TDS, 27,120 mg/1 COD and 8,120 mg/1 TOC.
Process operating parameters throughout 142 hr of steady-state
operation were: pressure, 400 psig, temperature, 21.6°C, and flow
rate, 1 gal/min. Salt rejection averaged 95.%, reducing raw feed
conductivity to 890 micro-mhu/cm. Permeate ionic strengtlj
decreased slightly after start-up, but leveled off quickly (Fig. 1).
Total dissolved solids removed was greater than ion rejection,
displaying 97.9% reduction. Dissolved solids were reduced from
16,400 mg/1 to an average of 340 mg/1.
The permeate was colorless, clear and waterlike, with a slight
odor of the parent waste. Feed solids, after evaporation, appeared
brown, whereas permeate solids showed a trace of yellowish brown.
Permeate dissolved solids decreased only slightly, in time. The feed
-------
TREATMENT
205
'Permeate -Conductivity vs. Time in Experiment R03
SOU
g 100
3
S
> 303
i*
U
-I
g
—I—
2.5
1.6 2 2.S 3
Operating Time (hr.)
3.S
Figure 1.
Permeate Conductivity vs. Time in Experiment RO3
pH of 4.9 decreased to a pH of 3.3. Thus, the ionic species that
passed the cellulose acetate membrane are highly acidic in
character. The flux did not change appreciably during 142 hr of
continuous operation. Permeate flux averaged 4.4 gal/day/ft2 with
a recovery of 0.20%.
Pretreated leachate feed COD, with an average value of 26,400
mg/1, was reduced by 68%. A permeate COD of 8,570 mg/1 was
produced by reverse osmosis. Approximately 3 % of the TDS and
one third of the organic matter in the leachate feed can permeate
the membrane. TOC, COD and TDS are interactive. However,
solutes that pass the membrane are more organic in character than
solutes in the bulk of the leachate. TOC results support these obser-
vations; 50% of the TOC was removed from a feed with an average
TOC of 8,480 mg/1, yielding a permeate TOC of 3,440 mg/1. Slight
reductions in permeate COD and TOC may be due to a non-fouling
gel layer formed on the interior membrane surface, augmenting
organic separation.
In Experiment 4, a continuation of Experiment 3, the unsteady-
state mode was used for an additional week of operation. An elec-
trical outage midway through the run caused the system to be in-
operative for several hours, leading to slight loss in pressure.
Overall performance in this study suggested that, as the concentra-
tion of the reject species in the feed increased, concentration in the
permeate increased, also. This result follows the basic theory of
membrane concentration gradients. Ion rejection decreased from
95%, at the start of the unsteady operation, to 91% after 116 hr.
Both feed and permeate conductivities increased, reaching 27,620
and 2,490 micro-mhu/cm, respectively. TDS followed conductivi-
ty, with solids rejection decreasing from 98 to 96% over the span of
operation. The dissolved solids concentration of the feed increased
40%, to 22,280 mg/1, while the permeate concentration was close
to 1,000 mg/1 in the final hour. Considering the volume of
permeate, this was not significant. Forty percent of the permeate
was obtained from feed. Flux rate decreased only slightly to 3.5
gal/day/ft2.
Summary
Reverse osmosis proved capable of processing pretreated
leachate, EPA-02, reducing TDS of 16,400 mg/1 by 98% and ionic
strength by 95%. Reverse osmosis separated this waste into a clear,
colorless permeate, with some odor. Also, it reduced the COD and
TOC of the leachate by over one-half; rejections were 68 and 59%,
respectively. Although solids removal was high, much of the solids
permeating the membrane were organic in character.
Degenerative fouling of the membrane did not pose a problem.
Over 250 hr of continuous, steady-state and unsteady-state leachate
operation did not disclose a fouling problem. This was also the case
with high strength simple salt and binary salt and carbon solute
mixtures.
Reverse osmosis process effectiveness with high strength
wastewaters is affected by several operating and solution variables.
The most important of these is pressure. An increase in pressure in-
creases permeate flux rate and the separation efficiencies for some
solutes. An increase in temperature improves flux and rejection
slightly, not significantly. The effect of wastewater pH on mem-
brane performance is.slight. In a binary salt and carbon system, the
removal efficiency improves slightly with increasing pH.
AEROBIC BIODEGRADATION
General
The study of bacterial metabolism of synthetic organic chemicals
can use one of several approaches. Investigations of pure com-
pound utilization by selected microbial strains are common. The ef-
fects of a single SOC on mixed microbial populations, such as ac-
tivated sludge systems, are of great interest. Such experiments are
essential to avoidance of upsets in wastewater treatment plants.
The extreme complexity of potential substrates in landfill leachate
mandates the approach in the present case. It is not to be expected
that one or a limited number of microbial species have or can ac-
quire the enzyme systems requisite to fully utilize this diversity of
carbon sources.
Aerobic biodegradation studies have concentrated on the use of
ssewage organisms. Secondary sludge is taken from an activated
sludge plant with substantial industrial inflows and cultured in the
laboratory to prepare a master batch of seed for comparative
respiration and growth experiments. The medium is due to
Krishnar and Gaudy (1977) and includes:
Glucose
(NH4)2S04
MgSo4'7H2O
FeCl3'6H20
lg/1
500 mg/1
100 mg/1
500 mg/1
pH is buffered with 1M K2HPO4 solution. The master batch is
grown to a nominal concentration of 10 g/1 of total solids. It is then
separated into 20 or more seed aliquots for the 100-200 ml
Erlenmeyer flasks used in shake aeration studies.
Control flasks combine sludge seed, glucose medium and distill-
ed water to volume. The desired sludge solids level is 2.5 g/1. Test
flasks receive the seed aliquot, glucose and/or leachate carbon, and
nutrient medium. Replicate control and test flask sets are sacrificed
at preselected times after inoculation. The progress of the experi-
ment is monitored by one or more of the following variables:
•Respiration Rate—specimens to Gibson or Warburg respirometer
•Volatile Suspended Solids, Total Suspended Solids, Total Solids,
Protein, Optical Density—surrogates for cell mass
•TOC— total organic carbon in system—suspended solids, ad-
sorbed carbon and medium
•Glucose—preferred substrate
Various experimental series were designed to test the utility of
sludge acclimation, cometabolism and supplementary glucose addi-
tion. Measurements were intended to compare respiratory assimila-
tion and cell mass or growth rate, as well as test for cellular adsorp-
tion or cellular uptake and storage without metabolism.
Experiment 120281
This experiment was intended to compare respiration and
organic carbon utilization, with and without leachate carbon at a
concentration approximating the initial glucose carbon. Glucose
concentrations were measured for samples taken during the initial
12 hr, from both test and control flasks. Optical density, DIC and
respiration rates were monitored for almost 4 days.
-------
206 TREATMENT
Experimental Scheme
Control
1.5 g/1 glucose
2.5 g/1 sludge solids
nutrient medium
Test
1.5 g/1 glucose
2.5 g/1 sludge solids
2.0 g/1 leachate DOC
nutrient medium
Glucose Uptake
Glucose uptake in control flasks was rapid and complete in the
12 hr observation period. Glucose was taken up by the flasks with
leachate added after substantial delay. Measurement after 24 hr
found glucose had disappeared completely from test flasks.
Respiration
Respiration rates of control flasks samples exceeded those of test
flask specimens, during glucose uptake by the former, at 0 and 5
hr. By 12 hr, test flask rates were seven-fold greater than control
flasks, although control flasks respiration did not display an effect
of substrate depletion (Table 2). Respiration in the test flask re-
mained higher for the rest of the experimental period. In addition,
respirometer measuresments were linear and very reproducible. Ox-
ygen uptake data taken at 36 hr is plotted in Fig. 2. Average respira-
tion rates for the entire experiment are described in Fig. 3.
0, UI'IAH VS IIHI in E>P[«|HtHl
see
406
100
3' hrs.
" - control 1 Slope-O.d
control 2 slope-0.65
— test 1 slope-B.18
test 2 sloiwS.36
IS 26 ES 30 35 46 -15 58 55 GO 65
1IHE (MINIIUS)
Figure 2.
02 Uptake vs. Time in Experiment 120281
Table 2.
Degradation of Leachate
fiCSPIRflllON RAIE VS. 1IW IN tmBIMtll! ]?0?B1
Volatile Solids Glucose In Solution Respiration Rite
Time (mg/l) ( g / 1 ) (pl/min)
(hr) Control Test Control Test Control Test
0 1310
1
2
4
5
6
8
10
12 2621
1310 1.36 1.30 1.60
..
1.08
- 0.84
-
-- 0.65
- 0.98
-- 0.04
1755 0.00
.26
.47
3.92
.23
.43
.20 --
.13 1.33
0.77
-
„
_
1.53
—
—
—
9.73
Biodegradation
Samples were taken from test and control flasks at 12 and 24 hr
intervals, over a period of 84 hr. Volatile solids (determined by op-
tical density) increased in both control and test flasks, but increased
much more for the test flasks (Table 3). Optical density appears to
give better estimates of cell mass than direct volatile solids
measurements, probably because there is negligible interference of
dissolved leachate solids with the measurement.
Soluble TOC showed an overall decrease of 72% between 24 hr
to 84 hr. From time zero, a decrease of 85% was observed. The
TOC of the control remained constant at zero, except at 48 and 84
hr. This result is due possible to lysis of cells and release of cell mat-
ter into solution. This corresponds to a decrease in volatile solids
between hours 36 and 84; refer to Table 3, also.
Comment
A lag in glucose uptake accompanied the presence of leachate
carbon. Degradation of leachate appeared to take place in this case,
however. This is suggested by increased respiration rates of test
over control flasks, a large increase in cell mass in the test flasks,
and a large decrease in soluble TOC in the test flasks. It is not clear
whether the TOC removed from solution was oxidized fully or used
in cell growth. It was not determined whether leachate carbon is ad-
sorbed onto cell surfaces, or taken up and dissolved in cellular fatty
structures.
1.2
0.8
e.e ,
e.-t
s e.a
control
test
18 2e 30 49 56 60 70 80 90 100
llHf (minim
Figure 3.
Respiration Rate vs. Time in Experiment 120281
Table 3.
Biodegradation of Leachate—Experiment 120281
Time
(nra)
0
12
24
36
48
60
84
Volatile Solids"
mg/l
Control Test
1310
2620
2330
2860
2790
2540
2480
1310
1760
1950
4000
4440
5720
4700
Soluble TOC
mg/l
Control Test
*
•
0
0
10
0
210
2600t
•
1120
510
410
310
310
Respiration Rile
/il/mln/ng vw
Control TeH
0.60
0.10
0.07
0.03
0.02
0.03
0.03
0.23
1.10
1.23
0.41
0.12
0.11
0.07
•not measured
tapproximate value
"approximated by optical density measurements
-------
TREATMENT
207
Experiment 022882
Experimental Scheme
Control Flasks
400 mg/1 DOC as glucose
2000 mg/1 DOC as sucrose
suspended solids
~-2500 mg/1
Test Flasks
400 mg/1 DOC as glucose
2000 mg/1 DOC as leachate
suspended solids
~2500 mg/1
The growth and respiration rates of an activated sludge culture
containing leachate and glucose, as carbon sources, were compared
with the rates of the same culture using glucose and sucrose, with
sucrose contributing DOC equivalent to that of the leachate. The
experimental period was 84 hours, nutrient medium was added to
test and control flasks.
Glucose Depletion Study
Glucose uptake was measured from hours 0 through 12, at 2 hr
intervals. The test flasks showed uptake rates similar to previous
experiments. However, the control flasks showed a high level of
glucose present throughout the 12 hours, probably due to
hydroloysis of sucrose to its monomers fructose and glucose.
Respiration was measured after 0, 6 and 12 hr; the overall rate of
the test flasks was slightly lower than the control flasks. Refer to
Fig. 4 and Table 4.
IE
14
12
110
I 8 -
3
I 6 .
o
A .
3 .
ie 20 30 40 se 60 70 80 90 100
lime (hrs.)
Figure 4.
Respiration Rates vs Time in Experiment 022882
Table 4.
Biodegradation of Leachate—Data for Experiment 022882
Suspended
Solids
Time (mg/1)
(lire)
Control Test
Vol. Suspended
Solids (mg/1)
DOC (mg/1)
Control Test Control
0
12
24
36
48
60
72
84
2500
3760
4990
5510
5040
4500
4710
4870
2500
3570
3530
5540
4930
3930
5110
4350
3260
4620
4610
4021
4130
3850
4270
2110
2330
4060
3100
2590
3110
3540
2400
1040
400
140
140
135
150
ISO
>/l) Respiration
/J/min/mg/vss
Test
2400
1470
1210
1050
890
740
560
130
Control
—
0.328
0.320
0.182
0.124
0.106
0.104
0.116
Test
...
0.683
1.284
0.377
0.564
0.312
0.160
0.131
Biodegradation
After the glucose depletion study, respiration was measured at 12
hr intervals. The overall rates for the test flasks were higher than
for the control flasks (Fig. 4). Since the test and control flasks con-
tained equal amounts of DOC, as leachate and sucrose, respective-
ly, this may indicate that leachate is preferred over sucrose as an
energy source, by the activated sludge, rather than as a carbon
growth source. This is further suggested by the higher values of
suspended and volatile suspended solids in the control flasks
relative to the test flasks. DOC reduction in test flasks, over the 84
hr period, averaged 82%, a value similar to data from previous ex-
periments. The DOC reduction for the control flasks averaged 93%
(Fig. 5).
DOC vs Ti™ In £«perlm«nt 022882
3000
2500 .
2000 .
1500 .
1000 .
500
10 20 30
50 60
Time (IH-S )
70
—T~
90
Figure 5.
DOC vs Time in Experiment 022822
Apparently, leachate serves as both an energy source and a car-
bon growth source for the activated sludge seed. Respiration rates
are higher and growth rates are lower than rates observed with
sucrose, a readily available carbon source.
CONCLUSIONS
Pretreated industrial landfill leachate can be renovated by reverse
osmosis, without degenerative membrane fouling. Greater than
95% of the inorganic species and 50% of the organic species pre-
sent in the leachate can be removed.
Subsequent to acclimation in the presence of glucose, sewage
organisms degrade leachate under aerobic conditions. Respiration
rates are enhanced substantially in the presence of leachate.
DISCLAIMER
Reference to commercial products or processes in this paper does
not constitute endorsement by Rutgers University of by the U.S.
Environmental Protection Agency.
ACKNOWLEDGEMENTS
This work is supported, in part, by cooperative agreement CR
807805010 between Rutgers University and the USEPA.
-------
208
TREATMENT
REFERENCES
1. Beck, E. "The Love Canal Tragedy," USEPA J., 5, Jan. 1979, 16-18.
2. Brunelle, M.T., "Colloidal Fouling of Reverse-Osmosis Membrance,"
Desalination, 32, 1980, 127-135.
3. Clark, J.W., Viessman, W., Jr., Hammer, M.J. Water Supply and
Pollution Control, 34th Ed., Harper and Row Publishers, New York,
N.Y.. 1977.
4. Eykamp, W. AICHE Symposium Series, 74, No. 172, 1976/7, p. 234.
5. Krishnan, P . and Gaudy, A.F., J r. "Response of Activated Sludge
to Quantitative Shock Loading Under a Variety of Operational Con-
ditions." Proceedings of the 30th Industrial Waste Conference, Pur-
due, Ann Arbor Science, Ann Arbor, Mi., 1977, 637-644.
6. Mikucki, W.J., Smith, E.D., Fileccia, R., Bandy, J., Gerdes, G.,
Kloster, S., Schanche, G., Benson, L.J., Staub, M.J. and Kamiya,
M.A. Characteristics, Control and Treatment of Leachate at Mili-
tary Installations, U.S. Army Corps of Engineers, Const. Eng. Re-
search Lab., Interim Report N-97, Feb., 1981.
7. Pendygraft, G.W., Schleger, F.E. and Huston, M.J. "Organics in
Drinking Water: A Health Perspective," Journal AWWA, 1979,
118-126, March.
8. Pendygraft, G.W., Schleger, F.E. and Huston, M.J. "Organics in
Drinking Water: A Health Perspective," Journal AWWA, Mar. 1978,
118-126.
9. Pojasek, R.B. "Disposing of Hazardous Chemical Wastes, Environ-
mental Science and Technology, 13, 1979, 810-814.
10. Shuckrow, A.J., Pajak, A.P. and Touhill, C.J. "Management of
Hazardous Waste Leachate," EPA-530/SW-S71, 1980.
11. Sorg, T. J. Treatment Technology to Meet the Interim Primary Drink-
ing Water Regulations for Inorganics, Journal AWWA, 1978,105-107.
12. Staats, E.B., "Waste Disposal Practices—A Threat to Health and
the Nation's Water Supply," Report to the Congress of the United
States by the Comptroller General, CED-78-120, U.S. GAO, June,
1978.
13. USEPA. Superfund Site List, Hazardous Materials Control Monthly,
2, Oct/Nov., 1981, 8.
-------
TREATMENT OF TNT AND RDX CONTAMINATED
SOILS BY COMPOSTING
R.C. DOYLE, Ph.D.
J.D. ISBISTER, Ph.D.
Atlantic Research Corporation
Alexandria, Virginia
INTRODUCTION
The manufacture and handling of TNT and RDX, major ex-
plosives used in conventional weaponry by the United States, has
resulted in contamination of soils and sediments at facilities carrying
out these operations. Because of the recalcitrant nature of these com-
pounds, the contamination of soil and sediment around these
facilities has increased over a period of years.
A number of investigators have studied biodegradation of RDX
under aerobic and anaerobic conditions. Sikka et al.' reported that
aerobic biodegradation of RDX occurs very slowly only in the
presence of sediments. Under anaerobic conditions, they found that
RDX was degraded in few in the presence of yeast extract. McCor-
mick et al.1 evaluated the biodegradability of 14C-RDX and observed
that degradation occurred only under anaerobic conditions. Rapid
disappearance of 14C-RDX in static cultures was observed although
little of the initial label was trapped as 14C-labeled gas. Most of the
label remained int he aqueous fraction. Several byproducts of
anaerobic degradation of RDX have been tentatively identified as
formaldehyde, nitrosotriazine, hydrazine, dimethyl-hydrazine and
methanol.2.
In the environment, the nitro groups of TNT are reduced in a step
wise fashion to amino groups. The intermediates in this reduction
scheme may condense to form dimers, and subsequent reduction and
condensation reactions can lead to the formation of complex polar
polymers.3 Major transformation,products have been identified as
monoaminodinitrotoluenes and tetranitroazoxytoluene. No evidence
for ring cleavage in soil or fresh water environments has been
found.3'4
The fate of TNT in soil amended with nutrients was studied by
Osmon and Andrews.5 The amendments did not alter the reductive
transformations observed to occur in unamended soils. A
preliminary study on composting of TNT by these investigators in-
dicated that composting resulted in a rapid decrease in TNT concen-
tration. The usual transformation products were not detected as pro-
ducts of TNT composting. In this paper, the authors describe the
results of laboratory and greenhouse scale studies on composting for
cleanup of TNT and RDX contaminated lagoon sediment.
MATERIALS AND METHODS
All composts were composed of 45% chopped alfalfa hay
(Medicago sativa), 45% Purina Sweetena horse feed and 10% soil
[air dried, sieved (2 mm) Lakeland sand (typic quartzipsamments),
thermic coated)] contaminated with either TNT or RDX. Addi-
tionally, a small quantity of active uncontaminated compost was us-
ed to seed each compost. The laboratory bench-scale composts were
50 g (dry weight) with sufficient distilled water added to bring the
moisture level up to 60% (wet weight basis). Test composts were
spiked with either 1.0 jtCi of I4C-TNT or 0.72 pCi of 14C-RDX. Both
radiolabeled compounds were uniformly ring labeled with radiocar-
bon purities of 96 and 97% for 14C-TNT and 14C-RDX, respectively.
Control composts contained no 14C tracers. The final concentration
of explosive on both test and control composts was 10,000 /ig/g (1 %
by weight).
Test composts that were not immediately sacrificed were placed in
the aeration and gas trapping apparatus illustrated in Fig. 1. Each
compost was incubated at 55°C and continuously aerated with
humidified, warm, CO2-free air. Off-gases from each compost were
Thermocouple and
Thermistor Meter
Air
NaOH
Trap
NaOH
Trap
-Perforated
Tubing
Figure 1.
Laboratory Compost Aeration Apparatus
209
-------
210
TREATMENT
individually scrubbed through concentrated H2SO4 and 5 N NaOH,
and then dried and passed through activated carbon to remove any
volatile I4C compounds emitted from the composts. The NaOH traps
were replaced weekly or more often to avoid CO2 saturation of the
trap. The H2SO4 traps were replaced after three weeks. Subsamples
of the liquid traps were assayed for 14C-activity by liquid scintillation
counting. At the conclusion of the experiment, the carbon traps were
crushed to a fine powder and subsamples were combusted (850 °Q
for 30 min. A 1:5 (v/v) mixture of CuO and AL,O, was used as a
catalyst. The "CO2 released was collected in Carbosorb and
quantified by liquid scintillation counting.
Control composts were incubated and aerated as described for
the test composts, but the off-gases were not scrubbed. Weekly air
samples were removed from the control composts and analyzed by
gas chromatography for CO2, CH4, N2, and O2 (Varian 3700 gas
chromatograph; 6 ft CTR column from Alltech, thermal conduc-
tivity detector).
After 0, 3 and 6 weeks of composting, both control and test com-
posts were sacrificed for analysis. At each time interval, three
replicate test composts for each explosive were sacrificed. Each
compost was extracted three times with 160 ml of appropriate sol-
vent^). The I4C-RDX composts were extracted with acetone. The
14C-TNT composts were extracted once with methanol/benzene
(40:120) and then re-extracted twice with benzene. All extractions
were carried out at 37 °C. Extracts for each individual compost
were combined and subsamples were counted by liquid scintillation
counting to quantify the total extractable 14C.
A portion of each extract was concentrated by rotary evapora-
tion and then analyzed by thin layer chromatography (TLC). The
TLC development was in a saturated atmosphere using silica gel 60
F-254 TLC plates. The developing solvents used were cyclohex-
anone for RDX and petroleum ethenethyl acetate:hexanes
(160:80:25) and benzene:hexanes:pentane:acetone (50:40:10:3) for
•O
O 2-Amlno DNT
O *' AmtflO ONT
l.«-Dlamlno NT
0 TMT if*
TETKA
O
solvent
system 1
O TNT
O TETHA
8
2 • Ammo DNT and 4-Am I no ONT
• • 2,6-Olamlno NT
2, 6- Olamlno NTT
• Origin
TNT - 2.4,6 - trinitrotoluene
TETRA - 2, a. 2', 6' - i«tr«nltro - 4,4' - noxytoluene
2-Ammo DNT - a- amino-4.6 -dinltrotoiuenc
4-Amino DNT 4-ammo - 2. 6 - dlmtrotoJuene
2.6- Dlimtno NT 2.0- dlimmo -4 • nitrotoluene
Figure 2.
Two Dimensional Thin Layer Chromatograph of a Six-Week
TNT Compost Extract Including Metabolite Standards
TNT— 2.4.6- tnrmroiolucne; TETRA—2,6.216'-[eiranitro-4.4'-azoxvtoluene;
- - A.mino DM — 2-amino-4.fr-dinitrotoluenc; 4-Ammo DNT— 4-,iimne-2,6-dinitrololuene;
2.*»-Did mi no N'T—2,6-diamino-4-m[rotoluenc
TNT. The TNT TLC plates were developed in two directions. The
separation of major TNT transformation products on a two dimen-
sional TLC plate is illustrated in Fig. 2. Radiolabeled compounds
on the TLC were located by autoradiography and were identified
by comparing Rf values with authentic standards. Each 14 labeled
spot was scraped from the TLC plates arid counted by liquid scin-
tillation counting to determine its activity. Following solvent ex-
traction, the compost solids were freeze dried, weighed, and
ground to a powder. Subsamples of the powder were combusted as
previously described to determine residual 14C activity.
One RDX control and one TNT control compost were sacrificed
at time 0 and duplicate samples were sacrificed for the 3 and 6 week
composts. These samples were assayed for percent moisture (80°C
for 24 hours), pH (approximately 1:9 solid to water ratio), total
Kjeldahl nitrogen content (6) and total carbon content. The total
carbon was determined by combustion with the CO2 being trapped
by NaOH solution, which was then titrated to the phenolphthalein-
end point.
All liquid scintillation counting was performed on a Beckman
LS-7500. Samples were counted in the 300-655 window with
automatic quench correction to adjust the window appropriately
for varying levels of quench. Samples were corrected for quench
from a standard quench curve using Beckman's H number as a
measure of quench.
Greenhouse scale composts were 10 kg (dry weight) in size. The
composting chambers were 38 x 56 x 46 cm wooden boxes sealed
with varnish. Each chamber had a removable lid, which did not
form an air tight seal when in place and was insulated on all sides to
reduce heat loss from the comp
-------
TREATMENT
211
was reduced from an average of 93.5% at time zero to 46.6 and
16.6% respectively after 3 and 6 weeks of composting, respectively
(Table 1). Despite this substantial loss of TNT, only insignificant
quantities of 14C were trapped as 14CO2 (Table 2) suggesting that
cleavage of the TNT ring did not occur to any appreciable extent.
Other volatile losses of 14C from the TNT composts (in H2SO4 and
carbon traps) were insignificant. Transformation of TNT through
the reduction of nitro groups to amines, as reported in earlier
work,3'4 did not appear to be a primary mechanism of TNT loss
during composting. No 14C-labeled amino derivatives of TNT were
detected in the compost extract after 3 weeks of composting. After
6 weeks of composting only small quantities of 2-amino-2,4-di-
nitrotoluene and 4-amino-2,6-dinitrotoluene were found in the ex-
tract of one of three replicate composts. Negligible quantities of
2,6-diamino-4-nitrotoluene may have been present in the 6 week
composts but the presence of polar materials in the solvent extract
interfered with the identification and quantification of this com-
pound.
The reduction of TNT levels in the compost was correlated to a
reduction in the solvent extractable 14C (R = 0.9997) and was paral-
leled by a significant increase in residual 14C activity (R = 0.9835).
The composting process apparently altered TNT to form products
which are insoluble in benzene and/or are very strongly sorbed to
the compost materials.
Analysis of the control composts indicated that the C/N ratio
and percent moisture did not limit microbial activity (Table 3). Air
V!80
£
I70
|60
ft
ow
o.
140
U
30
TNT
o Replicate A
"Replicate B
7 14 21
Length of Composting (Days)
28
Figure 3.
Temperature Profiles for TNT Greenhouse Composts
Table 2.
"C Recovered from Uniformly Ring-Labeled 14C-TNT Composted 0, 3
and 6 Weeks in the Laboratory
% Recovery of 14C
Length of
Compost-
Weeks
0
3
6
MCOj
0.0 bt
0.2 b
0.5 a
H2S04
trap
0.0 a
0.0 a
0.0 a
Carbon
trap
0.0 a
0.0 a
0.0 a
Solvent
extract
93.5 a
47.8 b
19.3 c
Resi-
dual 14C
1.7 c
37.8 b
66.5 a
Tota
95.2
85.8
86.3
tValues in a column not followed by the same letter are significantly different at the 5 % level of pro-
bability according to the Student-Newman-Kuel multiple range test.
Table 3.
The pH, Moisture Content and C/N of the Laboratory
Control TNT Composts
Length of
Composting
0 wks
3 wks
3 wks
6 wks
6 wks
pH
5.9
8.1
6.0
8.0
4.7
Moisture
(%)
60.0
58.8
59.7
61.9
56.3
C/N
15.6
14.9
17.7
13.4
20.6
analyses of the compost atmospheres showed that the compost re-
mained aerobic throughout the experiment.
Analysis of the extracts from the greenhouse scale composts con-
firmed the rapid disappearance of TNT observed in the laboratory
scale composts. TNT levels were reduced by more than 99%, below
the detection limit of 16.9 jtg/g. within three weeks (Table 4). The
greenhouse composts maintained temperatures between 55 and
75 °C with no external heat applied (Fig. 3). The elevated
temperatures of the greenhouse compost evidently increased the ef-
ficiency of TNT transformation over that observed in the
laboratory study.
RDX
Laboratory studies demonstrated that RDX is rapidly metabol-
ized in compost. The 14C recovered as RDX accounted for 112.3,
68.9 and 21.6% of the 14C initially added to the composts after 0, 3
and 6 weeks of composting, respectively (Table 5). Thin layer
Table 4.
TNT Concentrations in Greenhouse Compost Material
Table 1.
14C Recovery from TNT Laboratory Compost Extracts
% of total 14C
Length of Repli-
Compost- cate
tag (wks) TNT At
0 A 89.8
3D SB S
O oo.o —
C 101.8
3 A 44.5
B 48.9
C 46.5
6 A
B 37.0
C 12.9 0.6
At — 2-anuno-2l4-dinitrotoluene
B— 4-amino-2,6-dinitrotoluene
B C D E
__^
—
j 5
1.0
0.9
0.1 1.0 0.6*
1.1 2.0
1.1 0.8 0.9
C — this was not a discrete spot on any chromatograph but an area that would contain
2,6-diamino-4-nitrotoluene if it were present
E— other unidentified He-compounds
t— no 14C-activity detected
•—present in two spots
l«l l/ig/g;
Sample 0 Week 3 Week 4 Week
Box 1
(control) *:17 t <17 <17
Box 4 19,678 <17 <17
Box 5 20,404 <17 *17
tDetection limit for quantification of TNT from compost was 19.6 fig/g
Table 5.
14C Recovered from "C-RDX Laboratory
Composts at 0, 3 and 6 Weeks.
% Recovery of 14C
Length of MCO2 H2SO4 Carbon Solvent Resi-
Compost- trap trap extract t dual I4C Total
ing
Owks 0.0 c t O.Ob 0.0 112.3 a 6.1 b 118.4
3 wks 19.6 b 0.3 a b 0.0 68.9 b 13.5 a 102.3
6 wks 55.8 a 0.7 a 0.0 21.6 c 16.1 a 94.2
t'4C-RDX was the only detectable 14C component of the solvent extract
t Values in a column not followed by the same letter are significantly different at the 5% level of
probability according to the Student-Newman-Kuel multiple range test.
-------
212
TREATMENT
chromatographic analysis of the compost extracts indicated that
RDX was the only I4C containing compound present in the ex-
tracts. No byproducts of RDX degradation were detected.
10 2O 30
Length of Composting (days)
4O
Figure 4.
Cumulative Percent '«C Recovered as I4CO2 From "*C-Labeled
RDX in Compost
Significant quantities of 14C were volatilized from the composts
as I4CO2. The cumulative average recovery of 14CO2 with length of
composting is presented in Fig. 4. The rapid loss of 14C-RDX with
the concurrent evolution of I4CO2 suggests that the RDX molecule
is very quickly metabolized with a substantial portion of the carbon
being released as CO2. The breakdown products of RDX appear to
be assimilated into the microbial biomass as quickly as they are pro-
duced. Initial losses of 14C as CO2 are therefore presumed to reflect
the rate of RDX breakdown. However, as composting proceeded,
increased amounts of I4C would be incorporated into the biomass
resulting in additional I4CO2 losses through secondary metabolism.
On this basis, the 14CO2 recoveries can be used to estimate the
breakdown of RDX with length of composting. Recovery of
14CO2 during the first four days of incubation showed that RDX
degradation began shortly after compost initiation. The rate of
RDX breakdown increased during the first two weeks of com-
posting before reaching a maximum rate which was maintained for
an additional week. During the last three weeks of the experiment,
the rate of 14CO2 evolution declined. Depletion of RDX readily
available for microbial attack, shifts in the microbial populations
or decreased activities of select group(s) of compost organisms
could be responsible for this decline. 14CO2 resulting from secon-
dary metabolism probably constituted a significant portion of the
I4CO2 recovery during the last two weeks of the study. Recovery of
residual-14CO2 showed a statistically significant increase during the
first three weeks of composting as a fraction of the carbon from
RDX metabolism was incorporated into the microbial biomass.
There was no substantial change in residual 14C between the three
and six week composts presumably due to secondary metabolish of
the 14C in the biomass with resultant 14CO2 evolution.
Analyses of the controls indicated that the composts were not
limited by aeration, water or C/N ratio (Table 6).
Table 6.
Moisture Content, pH and C/N of Laboratory RDX Control Composts
Sample
0 week
3 week
3 week
6 week
6 week
PH
5.9
8.3
4.8
8.5
8.5
% Moisture
60.0
66.3
53.1
64.5
70.1
C/N
14.8
13.4
19.0
11.4
12.7
Sao
I70
!<50
050
°40
30
"Replicate A
• Replicate B
7 14 21 28
Lengl/i ol Composting (days)
35
42
Figure 5.
RDX Greenhouse Temperature Profiles
Breakdown of RDX in the greenhouse scale composts was initial-
ly more rapid than that observed in the laboratory (Table 7). After
three weeks of comosting, 55 % of the RDX was degraded in the
greenhouse composts compared to a 31% reduction in the
laboratory composts. As with TNT, the temperature of the
greenhouse composts was significantly higher than that of the
laboratory composts (Fig. 5) and the increased rate of breakdown
could have resulted from the elevated temperatures. The six week
old composts from both greenhouse and laboratory experiments
had approximately the same level of residual RDX (22% laboratory
and 24% greenhouse). The increased temperatures and higher
levels of microbial activity of the greenhouse composts were not
more effective than the laboratory composts in reducing RDX over
an extended period of composting.
Table 7.
RDX Concentration in Greenhouse Compost Material
ug/g in Compost
Compost
Control
RDX
RDX
T0 Week
ND t
9,240
9,414
T3Week
ND
3,284
5,093
T6 Week
ND
3,142
1,277
tDetection limit for quantification of RDX from compost was 794.7 jig/g
CONCLUSIONS
TNT rapidly disappeared in both the 50 g laboratory and the 10
kg greenhouse scale composts. The rate of disappearance was much
greater in greenhouse scale composts. It is assumed that the increas-
ed rates of disappearance are related to the higher temperatures and
greater microbial activity of the larger scale composts. Recovery of
I4CO2 from the laboratory composts was negligible suggesting that
the benzene ring in TNT is not cleaved during composting. The
recovery of very small quantities of 14CO-TNT reduction products
in relation to the large disappearance of TNT suggests that the
biotransformation pathway involving reduction of the nitro groups
and subsequent polymerization3'4 is not a primary route of TNT
loss from the compost. It is hypothesized that a mechanism of TNT
transformation unique to composting is responsible for incor-
porating the TNT ring structure into solvent insoluble residue in the
compost.
Composting is an effective method for degrading RDX. A large
portion of the ring carbon (14C-labeled) was evolved as I4CO2 dur-
ing six weeks of composting. The pattern of 14CO2 evolution in-
dicated that RDX metabolism begins very quickly after composting
is initiated and that the rate of degradation increases during the
first two weeks of composting. 14C-labeled byproducts of RDX
breakdown were not detected in acetone extracts of the composts at
-------
TREATMENT
213
any of the sampling times. Probably the RDX molecule is largely
destroyed in the compost. After 6 weeks of composting 22 to 24%
of the parent RDX molecule remained in both laboratory and
greenhouse scale composts. No explanation for the relatively high
levels of RDX remaining in both the laboratory scale and
greenhouse scale composts after six weeks is apparent from the
data. A build up of a substance(s) that inhibits or blocks RDX
metabolism is possible. The consistency of the levels of residual
RDX could be explained if this inhibitory material is synthesized
either directly or indirectly as the result of RDX breakdown. Other
possible phenomena that could have contributed to a high level of
residual RDX include shifts in the compost microbial population as
the compost ages, a general decrease in the microbial activity of the
compost, and a bonding or adsorption of RDX to the compost
materials thus reducing RDX susceptability to microbial attack.
REFERENCES
1. Sikka, H.C., Banerjie, S., Pack, E.J. and Appleton, H.J., "Environ-
mental Fate of RDX and TNT," Syracuse Research Corp., Contract
DAMD17-77-C7026, 1980.
2. McCormick, N.G., Cornell, J.H. and Kaplan, A.M., "Biodegrada-
tion of Hexahydro-l,3,5-trinitro-l,3,5-triazine," Applied and Environ-
mental Microbiology, 42, Nov. 1981.
3. Hoffsommer, J.C., Kaplan, L.A., Glover, D.J., Kubose, D.A., Dick-
enson, G., Kayser, E.G., Groves, C.L. and Sitzman, M.E., "Biode-
gradability of TNT: A Three Year Pilot Study," Naval Surface
Weapons Center, White Oak, Md., NSWC/WOL TR77-136, 1978.
4. Spanggord, R.J., Stanford Research Institute, Personal Communica-
tion, 1980.
5. Osmon, J.L. and Andrews, C.C., "The Biodegradation of TNT in
Enhanced Soil and Compost Systems," Army Armament R&D Com-
mand, ARLCD-TR-77032. NTIS, ADE-400 073, 1978.
6. Bremner, J.M., "Total Nitrogen," p. 1171-1175, in C.A. Black et al.
(eds.) Methods of Soil Analysis—Part 2. Chemical and Microbio-
logical Properties, American Society of Agronomy, Madison, Wi.,
1965.
-------
HAZARDOUS WASTE INCINERATION:
CURRENT/FUTURE PROFILE
C.C. LEE, Ph.D.
U.S. Environmental Protection Agency
Cincinnati, Ohio
EDWIN L. KEITZ
GREGORY A. VOGEL
The MITRE Corporation
McLean, Virginia
INTRODUCTION
The control of hazardous waste is one of USEPA's highest
priorities for the 1980s. EPA estimates that 57 million tons of
organic hazardous wastes are generated annually in the United
States.' Approximately 70 percent of this waste could be disposed
of by using,thermal destruction technologies. USEPA regards in-
cineration as a principal technology candidate for destroying hazar-
dous waste. Since the Congress enacted the Resource Conservation
and Recovery Act of 1976 (RCRA), incineration has been includ-
ed among those hazardous waste disposal technologies that are
regulated by the Agency.
The objective of the authors is to discuss: (1) the extent to which
incinerators are currently being used in industry for the disposal of
hazardous waste, (2) the basic characteristics of such incinerators,
and (3) innovative technologies which are currently under investiga-
tion for possible future applications. Since incineration can serve as
an alternate to land disposal for many hazardous wastes, this paper
will assist those interested in uncontrolled waste site cleanup.
In 1980 EPA promulgated regulations requiring every facility
which is treating, storing or disposing of hazardous waste to file
Part A of the RCRA permit application form2 under OMB permit
No. 158-ROOXX. The data submitted on these forms was stored in
a computer information system entitled "Hazardous Waste Data
Management System" (MWDMS) which is operated in each of the
10 EPA Regions.
The Incineration Research Branch of USEPA's Industrial En-
vironmental Research Laboratory in Cincinnati has expanded
HWDMS and is developing the Hazardous Waste Control Tech-
nology Data Base (HWCTDB) to manage detailed incineration
engineering data, trial burn data and related information.
The information presented in this paper is based on part of the
data assembled for the HWCTDB project. The topics to be discuss-
ed include conventional incinerators, innovative technologies, a
profile of existing incinerator facilities and a profile of incinerator
manufacturers.
CONVENTIONAL INCINERATORS
Liquid Injection Incinerators
Liquid injection incinerators are currently the most commonly
used type of incinerator for hazardous waste disposal. A wide
variety of units is marketed today, with the two major types being
horizontally- and vertically-fired units. A less common unit is the
tangentially-fired vortex combustor.
As the name implies, the use of the liquid injection incinerator is
confined to hazardous liquids, slurries and sludges with a viscosity
value of 10,000 SSU or less. The reason for this limitation is that a
liquid waste must be converted to a gas prior to combustion. An
ideal size droplet is 40 microns or less, and is attainable mechanical-
ly using rotary cup or pressure atomization, or via gas-fluid nozzles
and high pressure air or steam.
The key to efficient destruction of liquid hazardous wastes lies in
minimizing unevaporated droplets and unreacted vapors. Typical
combustion chamber residence time and temperature ranges are 0.5
to 2 sec and 1300T to 3000 °F, respectively. Liquid injection in-
cinerators are variable dimensionally, and have feed rates up to
5,600 Ib/hr.
The combustion chamber is a refractory-lined cylinder. Burners
are normally located in the chamber so that the flames do not im-
pinge on the refractory walls. The combustion chamber can be ac-
tively cooled by process air prior to its entry into the combustion
zone, thus preheating the air to between 300 °F and 700 °F.
Rotary Kilns
Rotary kiln incinerators are generally refractory-lined cylindrical
shells mounted at a slight incline from the horizontal plane. The
speed of rotation may be used to control the residence time and
mixing with combustion air. They are generally used by industry,
the military and municipalities to degrade solid and liquid com-
bustible wastes, but combustible gases may also be oxidized.
Recently, rotary kiln incinerators have been used to successfully
dispose of obsolete chemical warfare agents and munitions.
Two types of rotary kilns are currently being manufactured in
the United States: (1) cocurrent, with the burner at the front end
with the waste feed, and (2) countercurrent, with the burner at the
back end. Optimal length to diameter (L/D) ratios range from 2 to
10, and rotational speeds of 1 to 5 ft/min at the kiln periphery are
common, depending on the nature of the waste. Residence times
vary from a few seconds for a highly combustible gas, to a few
hours for a low combustible solid waste. A typical feed capacity
range is 1300 Ib/hr to 4500 Ib/hr for solids, and 630 Ib/hr to 2250
Ib/hr for liquids at temperatures ranging from 1475 °F to 2900T.
Fluidized Beds
Fluidized bed incinerators are vessels containing a bed of inert
granular material, usually sand, which is kept at temperatures in a
range from 850 °F to 1550°F.
Fluidizing air is passed through a distributor plate below the bed
and agitates the heated granular material. Hazardous waste
material and auxiliary fuel are injected radially in proportionately
small amounts and mixed with the bed material which transfers
heat to the waste. The waste in turn combusts and returns energy to
the bed.
The reactor vessel is commonly about 20 to 25 ft in diameter and
30 ft high. Bed depths are typically 3 ft while at rest and 6 ft during
operation. Variations in the depth affect both residence time and
pressure drop, resulting in a compromised depth which optimizes
residence time and excess air to ensure complete combustion.
Multiple Hearths
A typical multiple hearth furnace includes a refractory-lined steel
shell, a central shaft that rotates, and a series of rabble arms with
teeth for each hearth. Sludge and/or granulated solid combustible
waste is fed through the furnace roof by a screw feeder or belt and
flapgate. The rotating air-cooled central shaft with air-cooled rab-
ble arms and teeth distributes the waste material across the top
hearth to drop holes. The waste falls to the next hearth and then the
next and so on until discharged as ash at the bottom. The rabble
arms also agitate the waste as it moves across the hearth to con-
tinually expose fresh surfaces to the hot gases.
214
-------
INNOVATIVE TECHNOLOGIES
Molten Salt
Molten salt destruction3 is a method of combusting organic
material while, at the same time, scrubbing in-situ, any objec-
tionable by-products of the combustion and thus preventing their
emission in the effluent gas stream. This process of simultaneous
combustion and scrubbing is accomplished by injecting the
material to be burned with air, or oxygen enriched air, under the
surface of a pool of molten sodium carbonate. The melt is main-
tained at approximately 1650 °F; at this temperature the organic
hydrocarbons are immediately oxidized to carbon dioxide and
water. Halogens in the waste form halide salts while phosphorus,
sulfur, arsenic or silicon (from glass or ash in the waste) form ox-
ygenated salts such as sodium phosphate, sulfate, arsenate or
silicate. These products are retained in the melt as inorganic salts
rather than released to the atmosphere as volatile gases.
In 1981, under contract to USEPA, Rockwell International
tested hexachlorobenzene and chlordane in their pilot-scale molten
salt reactor. The results show that the destruction and removal effi-
ciency of their molten salt reactor exceeds 99.99% for these
chemicals.
High Temperature Fluid Wall
The high temperature fluid wall (HTFW) process is a high
temperature process for quickly reducing organic wastes to their
elemental state." A cross-section of a typical high-temperature
fluid-wall reactor is shown in Fig. 1. The reduction is carried out in
a reactor which consists of a tubular core of porous refractory
material capable of emitting sufficient radiant energy to activate
reactants fed into the tubular space. The core material is designed
to be of uniform porosity to allow the permeation of a radiation-
,transparent gas through the core wall into the interior. The core is
completely jacketed and insulated in a fluid-tight pressure vessel.
Electrodes located in the annular space between jacket and core
provide the energy required to heat the core to radiant temperatures
around 4000 °F.
During operation, the waste material to be pyrolyzed is finely
ground to a 20 mesh and introduced into the top of the reactor. As
the material falls through the tubular space it is exposed through
radiative coupling to power densities of over 1200 watts/in2 The
finely divided reactants are heated through the direct impingement
of electromagnetic radiation.
Through a cooperative agreement between USEPA and the State
of California, the developer of HTFW, Thalgard Inc., will test
their 3 in. diameter reactor in late 1982 in California.
Wet Air Oxidation
Wet air oxidation (WAD) is the aqueous phase oxidation of
dissolved or suspended organic substances at elevated temperatures
and pressures.3 Water, which makes up the bulk of the aqueous
phase, aids in catalyses to the oxidation reactions proceed at
relatively low temperatures, 350 °F to 650°F, and at the same time
the water moderates the oxidation rates removing excess heat by
evaporation. Water also provides an excellent heat transfer medium
enabling the WAD process to be thermally selfTSustaining with
relatively low organic feed concentrations.
The oxygen required for the reactions is provided by an oxygen-
containing gas, usually air, bubbled through the liquid phase in a
reactor used to contain the process; this then is the derivation of the
commonly used term "wet air oxidation." The process pressure is
maintained at a level high enough to prevent excessive evaporation
of the liquid phase, i.e. 200 to 3,000 psi.
Through a cooperative agreement between USEPA and the State
of California, Casmalia Resource Management Inc. in California is
in the process of installing a reactor in the Casmalia industrial
dump site to detoxify wastes in the site. The manufacturer of the
reactor, Zimpro Inc., will test other compounds of interest in the
reactor in addition to the wastes at the site.
ULTIMATE DISPOSAL
I
215
Expansion Bellows
Power Feedttirough
Cooling Manifold
Power Clamp
Porous Core
Radiometer Port
HeatShMd
Insulator
Cooling Jacket
Blanket Gas InM
(Typical)
Figure 1.
Cross-section of a typical high-temperature fluid-wall reactor
Plasma Reactor
Plasn is have been referred to as the fourth state of matter since
they do not always behave as a solid, liquid or gas. A plasma may
be defined as consisting of charged and neutral particles, having an
overall charge of approximately zero, all exhibiting collective
behavior.
The most common method of plasma generation is electrical
discharge through a gas.' A typical plasma reaction vessel
schematic is shown in Fig. 2. The plasma, when applied to waste
disposal, can best be understood by thinking of it as an energy con-
version and transfer device. A low pressure gas is used as a medium
through which an electrical current is passed. The type of gas used
is relatively unimportant in creating the discharge, but will
ultimately affect the products formed. In passing through the. gas,
electrical energy is converted to thermal energy by absorption by
the gas molecules, which are activated into ionized atomic states,
losing electrons in the process. Arc temperatures up to 50,000°C
can be achieved along the centerline recirculation vortex. Ultra-
violet radiation is emitted when molecules or atoms relax from the
highly activated states to lower energy levels. Waste materials are
atomized, ionized, and finally destroyed as they interact with the
decaying plasma species. The products which result are simple
because their activated states are atomic.
-------
216
ULTIMATE DISPOSAL
Figure 2.
Plasma reaction vessel schematic
USEPA and the State of New York are currently in negotiations
to form a cooperative agreement to build a mobile plasma reactor
with a capacity of 50 gal/hr. The objective is to destroy sludges
from the Love Canal treatment facilities.
PROFILE OF EXISTING INCINERATION FACILITIES
The existing facilities data discussed in this paper were assembled
in support of the Regulatory Impact Analysis Program for hazar-
dous waste incineration. The approach to the data assembly began
with preparation of a list of all known facilities which might have
one or more operational hazardous waste incinerators. As of the
Nov. 30, 1981 cutoff date established for this list, 612 such facilities
were identified. The HWDMS contained 566 of these and 46 were
identified from other sources. However, at that time, it was known
that some of the Part A applications had not yet been entered into
HWDMS. Based on later information obtained from HWDMS in
July 1982 (estimated 100% complete), it was calculated that the list
of 612 facilities was approximately 90% complete.
Initial telephone contacts with many of these facilities revealed
that a significant number did not have an operational hazardous
waste incinerator. Of the 612 facilities, a total of 537 facility
spokesmen indicated whether or not their facility had an opera-
tional hazardous waste incinerator and provided varying amounts
of additional information. The summary findings discussed here
are based on the information verified by these 537 facilities.
The hazardous waste incineration status of these 537 facilities
divided by USEPA regions is shown in Table 1. A total of 284
operational HW incinerators were identified at 219 facilities. Thus
only 40.8% of the facilities contacted verified having an opera-
tional HW incinerator. Inspection of the list of facilities in the
HWDMS data base in July 1982 showed that there were 128
facilities not previously contacted during the telephone campaign.
If it is assumed that these facilities have the same verification rate
as those contacted earlier, then an additional 52 facilities should
Opera-
tional
Facil.
10
22
23
46
29
62
8
5
14
0
219
Opera-
tional
Incin.
12
28
30
59
31
95
8
5
16
0
284
No.
Opera-
tional
HWI
34
46
48
52
44
28
11
9
14
0
' 286
Under Stilus
Conslr Unknown
3
4
5
6t
3
4t
3
1
3
Of
32f
1
1
0
0
0
0
0
1
0
0
3
Sample
Size
48
73
76
102
76
93
22
16
31
0
537
have 68 operational HW incinerators. Therefore, there are approx-
imately 350 operational HW incinerators at 270 facilities in the
United States.
The number of operational HW incinerators by type is shown in
Table 2. Of the 264 incinerators whose type was specified, 208
(79%) are capable of burning liquids by injection. Twenty-nine
units (11 %) are capable of burning bulk wastes (solids or liquids).
The remaining types are mostly special purpose units such as steel
drum reconditioning burners or military ammunition disposal
units.
Table 1.
Status of Hazardous Waste Incineration Facilities
in Each EPA Region*
Region
I
II
III
IV
V
VI
VII
VII
IX
X
Total
•Information obtained from an estimated 81% of the HWI facility population.
t3 facilities have both an operational unit and a unit under construction; 2 in Region IV and 1 in
Region VI.
Table 2.
Type and Number of Operational HW Incinerators*
Type
Liquid Injection
Hearth with Liquid Injection
Fume with Liquid Injection
Rotary Kiln with Liquid Injection
Combination System f
Rotary Kiln (Solids Only)
Hearth (Solids Only)
Ammunition and Explosives
Drum Burner
OtherJ
Total Specified
Type not Specified
Total
•Information obtained from an estimated 81% of the HWI facility population.
tlncludes interconnected multiple units (e.g., rotary kiln in scries with liquid injection unit).
{Includes such items as fluidized bed incinerators.
The design capacities of operational HW incinerators are shown
in Table 3. Design capacities were reported for 180 incinerators
burning liquids and 44 incinerators burning solids. The median
design capacity of incinerators burning liquids is 150 gal/hr with
most units (86%) not exceeding 1000 gal/hr. Incinerators burning
solids tend to have similar capacities with the median being approx-
imately 650 Ib/hr (equivalent to 78 gal of water).
The temperature and gaseous residence time for operational HW
incinerators is shown in Table 4. Combustion temperatures were
reported for 173 incinerators. Gaseous residence times were
reported for 104 incinerators. The median combustion temperature
was approximately 1800°F, and median gaseous residence time was
slightly under 2 sec.
Number
136
33
24
10
5
1
23
12
7
13
264
20
284
% of Total
Specified
51
12
9
4
2
1
9
5
3
i
100
-------
ULTIMATE DISPOSAL
217
Table 3.
Design Capacity of Operational Hazardous Waste Incinerators*
Incinerators Burning Liquids
Capacity
(gal/hr)
0- 50
51- 100
101- 200
201- 300
301- 500
1000- 2,000
2001-10,000
Total
Specified
Unspecified
Total
No.
48
28
22
22
23
17
8
180
28
208
% of
Total
Specified
27
16
12
7
13
9
4
100
Capacity
(Ib/hr)
0- 100
101- 300
301- 500
501- 1,000
2,001- 5,000
5,001-10,000
10,001-20,000
Total
Specified
Unspecified
Total
No.
4
5
12
6
7
1
2
44
17
61
Incinerators Burning Solids
% of
Total
Specified
9
11
27
14
16
2
5
100
•Information obtained from an estimated 81% of the HWI facility population.
Table 4.
Temperature and Gaseous Residence Time for
Operational Hazardous Waste Incinerators*
Maximum Temperature
<1600°F 1600°F 1901°F>2200°F Not
Total
*
-------
218
ULTIMATE DISPOSAL
capacities up to 170 million Btu/hr. Typical rotary kiln and liquid
injection incinerators have approximately the same capacity.
Although the largest incinerator listed in Table 7 has a capacity of
150 million Btu/hr, some manufacturers have received requests to
bid on facilities as large as 300 million Btu/hr.
Table 6.
Number of Hazardous Waste Incinerators Sold in the United States
Type of
Incinerator
Liquid Injection
Fixed Hearth
Rotary Kiln
Fluidized Bed
Multiple Chamber Hearth
Pulse Hearth
Rotary Hearth
Salt Bath
Induction Heating
Reciprocating Grate
Infrared Heating
Open Drum
Total
•Includes five units in construction
tlncludes one oscillating kiln
tOne unit is in construction
0. Of
Ifg
0
}
2
7
9
I
2
1
HW
Incin
Sold
219
59
42* t
9
7
2
2*
0
0
1
It
0
342
%of
Total
64.0
17.3
12.3
2.6
2.0
0.6
0.6
...
—
0.3
0.3
...
100.0
Incinerator
Type
Liquid Injection
Hearth
Rotary Kiln
Fluidized Bed
Range
(Ib/hr)
30-24,500
25- 2,500
1200- 2,080
and the waste characteristics. The most important operating condi-
tions are the combustion zone temperature, combustion gas
residence time at elevated temperature, and excess air usage.
Typical operating conditions are summarized in Table 8. Hazar-
dous waste incinerators may be designed to operate outside the
ranges of these typical values. These data are obtained from a
relatively small sampling of incinerator manufacturers and may not
be indicative of the entire industry. Incinerator manufacturers
often determine operating conditions from trial burns or a
customer's waste.
COMPARISON OF OPERATIONAL DATA
WITH MANUFACTURERS' DATA
A comparison of the number of HW incinerators reported by
manufacturers and existing HW facilities is made in Table 9. A
total of 284 operational HW incinerators were identified at 219
facilities. If a projection of these facility figures is made to account
for the estimated 128 facilities from which data were not obtained,
the total operational HW incinerator population would be approx-
imately 350 at 270 facilities. This figure agrees very well with the
335 operational units reported by manufacturers. In contrast, ex-
isting facilities reported 32 units under construction which is much
higher than the 7 reported by the manufacturers.
A comparison of the types of operational HW incinerators
reported by manufacturers and HW facilities is made in Table 10.
The manufacturers' data and the projected total existing popula-
tion agree extremely well for the liquid injection and hearth type in-
cinerators. However, the rotary kiln population reported by
manufacturers is more than double the number reported by
Table 7.
Design Capacities of Hazardous Waste Incinerator Types
(From Manufacturers' Data)
Typical Value
(106 Btu/hr)
8
4.9
10.3
45.5
Mass Capacity
Statistical
Value and
Population
Median
Average
Average
-
43
48
2
1
Typical Value
(Ib/hr)
1,600
810
1,600
31,000
Range
(106 Btu/hr)
0.125-130
3- 9
1-150
8.5- 67
Thermal Capacity
Statistical
Value and
Population
Median
Average
Median
Average
50
4
34
5
In addition to the refractory lined combustion chamber, hazar-
dous waste incinerators may be equipped with automated loading
and ash removal systems, energy recovery equipment and air pollu-
tion control equipment. The most prevalent energy recovery equip-
ment is a firetube or watertube boiler generating steam. Air pollu-
tion control equipment is located downstream of the combustion
chamber and energy recovery equipment. It may consist of one or
more of the following components:
•Quench chamber to cool cumbustion gases
•Particulate collection device
Venturi scrubber
Baghouse
Electrostatic precipitator
Cyclone
Ionizing wet scrubber
•Gas absorbing device
Packed tower scrubber
Plate or tray scrubber
Spray tower scrubber
Ionizing wet scrubber
Most hazardous waste incinerator manufacturers buy energy
recovery and air pollution control equipment from vendors rather
than manufacture the equipment.
Incinerator manufacturers design hazardous waste units to
operate at specific conditions depending on the type of incinerator
Table 8.
Ranges of Incinerator Operating Conditions
(From Manufacturers' Data)
Incinerator
Type
Liquid Injection
Rotary Kiln
Afterburner
Hearth
Primary Chamber
Secondary
Chamber
Fluidized Bed
Combustn
Zone Temp
(°F)
1800-3000
1200-2300
2000-2500
1200-1800
1400-2200
1400-2000
Combustion
Gas Residence
Time
(sec)
0.3-2.0
2 hr (solids)
1.0-3.0
—
1.5-2.5
1.0-5.0
Excess Air
(% stoi-
chiometric)
120-250
50-250
120-200
30-200
200-400
100-150
facilities. Reasons for the discrepancies may include: some of the
units sold since 1969 may no longer be in use or may now burn non-
hazardous wastes; some manufacturers might not have been given
enough information to know whether the customer's wastes are
hazardous; or, some manufacturers may not know if the
customer's wastes are regulated under RCRA or the Toxic
Substances Control Act. All of these may cause high or low
estimates of the number of incinerators.
-------
ULTIMATE DISPOSAL
219
Table 9.
Comparison of Number of HW Incinerators Reported by
Manufacturers and HWI Facilities
Reported by HWI
Facilities Contacted
Operational
Incinerators
Units Under
Construction
Total Reported
Actual No.
Reported
284
32
316
Proj. for
Total Popula.
350
40
390
Rep'd by
Mfr's
335
7
342
Table 10.
Comparison of the Types of Operational HW Incinerators
Reported by Manufacturers and HWI Facilities
Reported by HWI
Facilities Contacted
Liquid Injection
Hearths
Rotary Kiln
Fluidized Bed
Others
Types Not
Specified
Total Opera-
tional
Actual No.
Rep'd
160*
56t
13tt
4
31**
20
284
Proj. for
Total Popula.
213
75
17
5
42
0
352
Rep'd
Mfr
219
70
37
9
tt
tt
335
by
'Includes fume/liquid units
tlncludes units both with and without liquid injection
{Includes 2 rotary kilns in combination units
"Includes 3 combination units not having a rotary kiln
tfThis category not obtained from manufacturers
CONCLUSIONS
In this paper the authors have presented a brief discussion of
hazardous waste incineration, including: (1) the technologies, (2)
existing facilities, and (3) incinerator manufacturing. A wide varie-
ty of incinerators were shown to be in operation throughout the na-
tion.
In addition, new processes offer promise for improvement in the
capability for more complete hazardous waste destruction and ex-
tension of this capability to cover a wider range of wastes. It is
becoming increasingly undesirable to select untreated landfilling as
the method for disposal of hazardous wastes, Many other options
are preferable.
REFERENCES
1. USEPA, "Engineering Handbook for Hazardous Waste Incineration,"
Publication No. SW-889, September 1981.
2. USEPA, "Part A of Hazardous Waste Application Requirements:
122.13 and Form 3," 45FR33543, May 19, 1980.
3. Yosim, S.J., "Disposal of Hazardous Wastes by Molten Salt Combus-
tion," Proceedings of Annual Meeting of AIChE, 1979.
4. Matovich, E., "Management of Toxic Wastes with the Thagard High
Temperature Fluid Wall Reactor," Proceedings AIChE Meeting, April
21, 1981.
5. Randall, T., "Wet Oxidation of Toxic and Hazardous Compounds,"
Technical Bulletin 1-610, Zimpro, Inc., Rothschild, WI, 1981.
6. Barton, T.G., "Plasma Destruction of Polychlorinated Biphenyls,"
Royal Military College, Kingston, Ontario, 1981.
-------
THE USE OF GROUT CHEMISTRY AND TECHNOLOGY
IN THE CONTAINMENT OF HAZARDOUS WASTES
PHILIP G. MALONE, Ph.D.
NORMAN R. FRANCINGUES
JOHN A. BOA, JR.
U.S. Army Corps of Engineers
Waterways Experiment Station
Vicksburg, Mississippi
INTRODUCTION
Uncontrolled hazardous waste sites include a wide variety of waste
disposal locations where toxic wastes from manufacturing or mining
activities have been discarded or dumped in a manner that poses a
threat to human health and the environment. The potential
pollutants vary from highly toxic organics to inorganic weathering
products from mine spoils. In the superfund list of 160 problem sites,
134 sites involve significant pollution from toxic organics and 26 in-
volved inorganic industrial wastes (such as battery or plating wastes),
mine tailings or mine drainage or slag. The wastes at the disposal sites
can be in tanks, lagoons, drummed storage or waste piles. In most
cases, soil and groundwater contamination from leakage or inten-
tional discharge is already evident. Remedial actions at the sites
usually take the form of removing waste to a safer location and/or
developing a strategy for on-site containment by installing barriers to
control waste movement or by immobilizing toxic constituents in
place.
In this paper, the authors discuss grouting technology and
grouting equipment in remedial actions other than diversion of un-
contaminated groundwater. While (grouts have long been used to
stabilize soils for foundations or to produce barriers to control
groundwater movement in tunnelling or excavation, there are other
aspects of grout chemistry that can be used to directly restrict
chemical transport from waste materials.
GROUTS AND GROUTING METHODS
Grouting involves the injection of liquid solution or suspensions
into granular materials to fill voids and cement the particles together
so as to restrict fluid flow in the media or improve its strength or
bearing capacity.1-2 Two classes of grouting materials are generally
recognized: (1) particulate grouts and, (2) chemical grouts (Table 1).
The particulate grouts are generally Portland cement and blends of
clay and Portland cement. The chemical grouts are solutions or
emulsions of various types that "gel" or polymerize after injection.
Table 1 includes data on some of the types of chemical grouts that
are currently available and are potentially of use.5 These materials
have very little in common except that each can be made into a low
viscosity, injectable liquid that will "gell" or polymerize to form a
solid or highly viscous fluid within a particulate medium.
Typically, grouts are injected into sediments using a perforated
pipe with expandable packers to restrict the movement of grout in the
boring. The pressures used during injection are relatively low, being
not more than 1 psi for each foot of overburden. Excessive pressures
can cause the porous medium to be fractured or wedged open.
Natural planes of weakness or porous zones in an alluvial unit will
accept more grout. By using expandable packers in a boring, grout
can be restricted to specific horizons and each horizon can be grouted
to capacity. A typical grouting setup includes a mixer, an agitator; a
pressure pump and an injection system (Fig. 1).
Grouts are typically injected by pumping the grout into sections of
a prepared boring. In alluvial material, it is necessary to inject an an-
nular layer of grout into the boring to hold it open for introduction
of more grout at each horizon. The annular grout is fractured by
pressure during subsequent grout injection. Second phase injection
grouts the porous units further out from the borehole.
Table 1.
Examples of Grouts Potentially Useful in Waste Immobilization'1-6)
RCTURN SECTION OF CIRCULATION LINE
Group
Particulate
Chemical
Chemical
Chemical
Chemical
Type
Clay-Cement
Particulate Clay
Sodium
Silicate
Cationic
Asphalt
Emulsion
Resorcinol
Polyurethane
Gelling System
Solidification of
Portland cement.
Swelling and gela-
tion of expanding
clays.
Polymerization to form
silica gel on mixing
with salts or acids.
Emulsion is broken by
adding hydrated lime
slurry.
Polymer is formed
with formaldehyde.
Reaction is catalysed
with hydrochloric
acid.
Polymer is formed
by combination of
polyisocyanates with
polyols, polyethers,
glycols or castor oil.
Reaction is catalysed
with tertiary amines
and tin salts.
Special Characteristics
Clay can be selected to
bind specific metals.
Clay can be selected
for particular chemical
properties.
Reaction is known to
bind heavy metals.
System is highly
alkaline.
System is highly al-
kaline, material re-
mains pliable after
gelation; coating of
mineral grains occurs.
System is acidic at
gelation. pH 1.5-2.5.
System will poly-
merize at a wide
range of pH's.
Figure 1.
Example of a typical grouting setup for a recirculating system."-2*
220
-------
ULTIMATE DISPOSAL
221
The grouts are typically mixed by the batch method using a
highshear mixer and the prepared grout is pumped into an agitator
that keeps the material mixed during injection. Set time or gelling
time is regulated by mixing in gelling accelerators or retardants. A
positive pressure pump is used to move the grout into the bore hole.
Pressures are regulated using the pump in conjunction with a system
of valves and bypass lines.
Where chemical grouts are used, specialized equipment is added to
meter the components of the grout and stationary inline mixers can
be used for blending. Often, care must be taken to use non-
corroding, non-reactive materials in pumps and piping.
APPLICATION OF GROUTING AT WASTE SITES
Grouting technology can potentially be used in a variety of ways
to provide better waste containment at a disposal site (Fig. 2).
Shallow grouting can be employed to improve the cover over a
waste site by employing closely-spaced, shallow grout injections to
develop a nearly impervious layer in the existing cover. Closely-
spaced, shallow grouting could also be employed to prevent the
dispersal of contaminated soils (or mine spills) at a disposal site and
to reduce infiltration and transport of toxic contaminant to the
water table. The viscosity and particle size of a grouting material
generally restricts the type of sediment or alluvium that a grout can
penetrate. For chemical grouts, the lower the viscosity, the finer the
grain size of the sediment that can be penetrated.
TOP SEALING OR STABILIZATION
OF CONTAMINATED SOIL
BOTTOM SEALING
WASTE REMOVAL AND SOLIDIFICATION
Figure 2.
Techniques for employing grouts in hazardous waste containment.
In particulate grouts, the grain size of the grout is the limiting
factor. Particulate grouts are generally restricted to materials that
have an average grain size of medium sand or greater (Fig. 3). The
organic polymer grouts can be used in materials as fine as a medium
silt. Organic polymer grouts also have more closely regulated set-
ting times; a characteristic that is very necessary when silts are being
injected.
Bottom sealing involves injecting grout into granular material
below the waste. This approach has been suggested by Tolman.7
The major difficulty in employing grout in this way is assuring that
GRAVEL
FINE
SAND
COARSE
MED.
FINE
CLAY-SOIL
COARSE
SILT
SILT (NONPLASTIC)
I
Mil
III
1L
Illl
III
IRESI
Hill
II 1 1
ISILIC/
MM 1
Mill
1L
II
Illl
1
CHF
NS
\TES
1
O
8ENTONITE (CLAY)
Illl
PORTLAND CEMEI\
Illl
III
1
T
I/IE
-LIGN
II
OR
PO
N
II
GA
_Y
N
t/II
C
:R£
).0 1.0 0.1 0.01 0.0(
GRAIN SIZE, mm
Figure 3.
Grouts typically used in different soil types."'
a continuous seal is developed. Grout is normally injected as bulbs
of material and care would have to be taken in planning injection
holes to assure that the bulbs of grout merge to form a continuous
bottom seal.
An optional strategy would involve grouting back up into the
waste and producing a solidified waste mass. This approach would
require much more grout and would involve injecting into wastes
that frequently are fine-grained (clay-sized) materials.8 This techni-
que involves some uncertainty with regard to the completeness of
the treatment. If the grout can restrict contaminant movement
because it alters the pH or bonds the toxic materials chemically,
even partial grouting may reduce the rate of escape of con-
taminants.
Complete waste solidification could be assured if the waste were
pumped or dug out and mixed into the grout and pumped into a
second storage site to gel or set (Fig. 2). This approach of combin-
ing the waste directly with a grout has been used successfully with
radioactive waste where the material was immobilized and injected
into geologic strata with very low permeability.9
The four basic approaches to using grout chemistry and tech-
nology in remedial Action can be combined a variety of ways to im-
prove waste containment. In most cases, the same material and
equipment could be employed. It would be possible to bottom seal,
top seal and solidify surrounding contaminated soils using very
similar mixers, pumps and injection equipment.
Waste/Grout Interactions
Most grout formulations used in foundation and construction
applications are tolerant systems that set under adverse conditions;
but chemical waste may contain materials that can prevent a set or
degrade the grout. The usefulness of the grouts in waste contain-
ment can be improved by selecting grouts or adding materials that
bind or sorb specific contaminants.
Cement and silicate grouts are alkaline systems and the high pH's
produced in a waste generally reduce the mobility of toxic metals.
Clay grout with specific absorption characteristics can be employed
to retain toxic metals.
As a group, organic pollutants may be the least easily retained
wastes.'" The use of specific sorbents for organics may be helpful in
improving containment.
Problems related to compatibility and chemisorptive properties
are illustrated in Table 2. Fillers can be added to many grouts to in-
crease absorption. For example, clays, diatomaceous earth, and fly
ash are common fillers in cement grouts. Organophilic materials as
fillers in grout may enhance adsorption and allow the grouts to be
used as chemical barriers to toxic organics.
-------
222
ULTIMATE DISPOSAL
Table 2.
Compatibility and Cbemisorptive Properties
for Selected Types of Grout
Grout Type
Clay or
Cement/
Clay
Sodium
Silicate
Cationic
Asphalt
Emulsion
Resorcinol
Polyurethane
Materials known to cause
set retardation or
degradation
High concentrations of
metal salts or some or-
ganics prevent setting.
Smectite clays are ad-
versely affected by many
organics.
Presence of acids or con-
centrated salts can cause
early gelation.
Organic solvents especially
hydrocarbons can soften
asphalts.
Presence of strong bases
can retard set.
Ammonia or ammonium
hydroxide inhibit gel
formation. Strong bases
cause swelling.
Materials known to
be chemisorbed
Metals may be se-
lectively sorb on some
clays.
Silica gels trap
divalent and trivalent
metals.
Foam is used as a
sorbent for benzene,
kerosene, n-butylalde-
hyde, and phenol.
EXAMPLES OF GROUTING IN WASTE CONTAINMENT
Madawaska Uranium Mine, Bancroft, Ontario
The mine used an acid leaching process to remove uranium from
ore. The tailings were placed in a disposal area surrounded by a tail-
ings dam. The mine water contained approximately 35
picocuries/liter of radium-226.''
There was particular concern that mine water was seeping
through the mine tailings and into a recreational lake near the mill
site. Canadian regulations require that all discharge water con-
tainless than 3 picocuries/liter.
Grout mixtures of illite, smectite, and bravaisite clays (a mixed
clay) and cement were prepared and pumped into a triple line of
bore holes placed between the lake and the tailings pond. Even
before the project was complete, the amount of dissolved radium in
water collected below the barrier showed a definite decline. The
clay/cement slurries adsorbed the radium from the seepage water,
creating a chemical barrier for the radium, but water still moved
through the barrier. Approximately 2,600 tons of grout were placed
in the grout curtain. It has been estimated that the amount of clay
involved is sufficient to adsorb the radium escaping from the pond
for thousands of years.
Runit Island, Eniwetok Atoll Test Site
In 1978, an effort was undertaken by the U.S. Army Corps of
Engineers to decontaminate areas on several islands associated with
the hydrogen bomb testing in the 1950s. Several areas of con-
taminated coral sand were to be contained. A grout batch plant
(Fig. 4) was fabricated on the site and the waste was mixed directly
into a Portland cement grout and pumped into a large blast crater
on the island. The batch plant separated the coral sand fines in a
screening operation and mixed them with Portland cement and an
attapulgite-based suspending agent and sea water.
The grout was pumped approximately 300 ft and laid down in
bottom of a blast crater (Fig. 5). The grout was designed to set up
under salt water and contain a minimum amount of cement with
the maximum amount of waste. Formulation for lowest cement
content grout proposed consisted of:
Component
Portland Type I
Eniwetok fines
Attapulgite
Friction Reducer
Salt water
Total
Weight Ob)
264
1827
47
2.8
810
2950
Laboratory batches of grout showed an unconfmed compressive
strength of 135 psi after sevendays.12
An 18 in. thick concrete cap was placed over the hardened
grout/waste mixture. The solidification process has been con-
sidered successful in containing the radioactive constituents in the
contaminated sand and the grout had satisfactory engineering pro-
perties.
CONCLUSIONS
Grout chemistry and grout technology can be useful in the treat-
ment and containment of some types of toxic wastes. Grouts can be
applied as barriers to water or waste movement or mixed directly
with the waste. Because a wide variety of grouts have been
developed and the advanced technology is available for their for-
mulation, mixing and/or injection, they will find increased use
where in-situ containment of waste is necessary. Examples of suc-
cessful applications to large scale waste problems are already
available. The most obvious potential application involves sites
where widespread low-level contamination" (mine tailings or con-
taminated soil) is the major problem.
Figure 4.
Batch plant used for the preparation of grout/waste mixtures
ai Runil Island, Eniwetok Atoll.
Figure 5.
Grout pipeline and grout placement system at Cactus Crater,
Runit Island, Eniwetok Atoll.
-------
ULTIMATE DISPOSAL
223
REFERENCES
1. U.S. Dept. of Army, "Grouting Methods and Equipment." TM 5-
818-6, U.S. Government Printing Office, Washington, D.C., 1970.
86 p.
2. Water Resources Commission. "Grouting Manual," 2nd Edition.
Water Resources Commission, New South Wales, Australia, 1977.
3. U.S. Dept. of Army, "Chemical Grouting. Engineering Manual,"
EM 1110-2-3504, Office of the Chief of Engineers, Washington,
D.C., 1973. 82 pp.
4. Gebhart, L.R. "Experimental Cationic Asphalt Emulsion Grouting,"
Journal of Soil Mechanics and Foundation Division, ASCE, 98, No.
SM7, July, 1972.
5. Shroff, A.V., and Shoh, D.L., "Resorcinolic Grout for Injecting
Sandy Foundations," Journal of the Geotechnical Engineering Di-
vision, ASCE, 106, No. GT10, Oct., 1980.
6. Vinson, T.S., and Mitchell, J.K., "Polyurethane Foamed Plastics in
Soil Grouting," Journal of Soil Mechanics and Foundation Division,
ASCE, 98, No. SMI, Jan., 1972.
7. Tolman, A.L., et al. "Guidance Manual for Minimizing Pollution
from Waste Disposal Sites," U.S. Environmental Protection Agency,
Publ. No. EPA-600/2-78-142, Cincinnati, OH, 1978, 267 p.
8. Bartos, M.J., Jr., and Palermo, M.R., "Physical and Engineering
Properties of Hazardous Industrial Wastes and Sludges," U.S. En-
vironmental Protection Agency, Publ. No. EPA-600/2-77-139, Cin-
cinnati, OH, 1977, 89 p.
9. Weeven, H.O., Moore, J.G., and McDaniel, E.W., "Waste Disposal
by Shale Fracturing at ORNL," pp. 257-260 in McCarthy, G.J., "Sci-
entific Basis of Nuclear Waste Management," Col. 1, Plenum Press,
New York, 1979, 562 p.
10. Anderson, D., Brown, J.W., and Green, J., "Effects of Organic
Fluids on the Permeability of Clay Soil Liners," pp. 179-190, U.S.
Environmental Protection Agency, Publ. No. EPA-600/9-82-002,
Cincinnati, OH, 1982, 549 p.
11. Henderson, J.K., "Guide to Alluvial, Rock and Chemical Grout-
ing," J.K. Henderson, Inc., Buffalo, NY, 1980, Revised Edition,
121 p.
12. U.S. Army Engineers, "Final Report of Evaluation of Portland Ce-
ment Grout Mixtures for Use in Disposal of Contaminant-P.O.D.
Eniwetok." USAE Waterways Experiment Station, Vicksburg, MS,
1977, 6 p.
-------
CRITERIA FOR COMMERCIAL DISPOSAL
OF HAZARDOUS WASTE
MILO G. WUSLICH
Frontier Chemical Waste Process, Inc.
Niagara Falls, New York
INTRODUCTION
Specific techniques and needs of off-site hazardous waste treat-
ment firms are not well known. Even for chemists with some
background in waste water chemistry, the day to day goals,
specifications, and problems associated with the operation of a
commercial treatment plant, are often foreign.
The burden of managing a generator's hazardous wastes usually
falls on the engineering department of that generator. Therefore,
an engineer's knowledge of what commercial firms look for in their
daily operations is critical for cost effective collection and segrega-
tion of wastes.
Many times a simple policy instituted at the point of generation
could mean the difference between getting paid $15.00-$30.00 for a
drum of waste or having to pay nearly $100.00 per drum in disposal
charges. This often happens when a company is generating flam-
mable organic wastes (e.g., xylene, toluene, various thinners) and
chlorinated solvents (e.g., 1,1,1-trichloroethane, TCE, other
degreasing solvents). If these wastes are mixed in the same drum
when collected, the disposal option is usually limited to high
temperature destructive incineration. The cost for this technique in
drummed qualities is in the $100.00 per drum range.
If on the other hand these wastes were kept separate, that is if the
chlorinated materials were collected in different drums from the
flammables, the disposal costs would be dramatically different.
Unmixed chlorinated solvents with a high enough recoverablej'ield
are often purchased from generators by reclaimers. Flammable
solvents with a low halogen content can be used for their thermal
value and usually incur a low disposal charge.
This same kind of information on how an outside TSDF (treat-
ment, storage and disposal facility) operates is also essential for a
consulting engineer who finds himself facing a mountain of
thousands of drums, two 500,000 gal lagoons, and one million gal
of tankage on an abandoned site. If a consulting engineer were try-
ing to prepare a meaningful bid package in the abo"e situation, but
did not know what criteria would be needed by the TSDFs, chances
are the figures quoted in the TSDF's proposal would be inflated to
cover those gaps in the information or to satisfy unnecessary re-
quirements.
Some of the most common techniques used by TSDFs operating
in the eastern U.S. today include: 1) chemical secure landfill, 2)
deep well injection, 3) encapsulation, 4) incineration, 5) recovery,
and 6) chemical treatment. Immediately following each type of
TSDF disposal process description will be a summary of some of
the most critical criteria used by that TSDF to evaluate waste
streams. These criteria will be the basis for the TSDF's decision on
acceptance or rejection of a stream and decisions on pricing. Much
of this information is applicable to both current generators of
wastes and also abandoned site cleanup categorizations.
In the last section of this paper, the author will offer suggestions
for better bid package preparation.
TECHNIQUES AND CRITERIA OF TSDF OPERATIONS
1. Chemical Secure Landfill
Some landfills constructed in New York State have generally had
to truck in and compact the clay needed to achieve permeability
ratings of *1 x 10~7 cm/sec. A 10 ft bottom layer of clay is com-
pacted first. Then a 30 mil hypalon membrane is .placed over the
bottom layer. Next 2 ft of compacted clay covers the membrane.
Drums of waste are placed vertically on this surface which is sloped
to various leachate collection points. Bulk dump trailer loads are
also accepted. Stand pipes are placed at the collection points to
eventually pump leachate to the surface. Vent pipes are also install-
ed to permit gases to escape. Clay walls are arranged within the
landfill to form cells and subcells to prevent co-mingling of incom-
patible wastes. The drums are covered daily and eventually the
landfill takes a domed shape. It is covered again with layers of clay,
polyethylene, sand, more polyethylene, clay and seeded cover.
Monitoring wells are also installed. A grid system is used as the
landfill is being filled to record the locations of each drum.
The secure landfill in Alabama is designed to meet similar re-
quirements. However, the natural clay deposits at that site facilitate
construction. In many areas the first aquifer lies beneath 700 ft of
clay. The walls of the landfill are cut vertically into the clay. The
drums are laid on their sides and covered with clay and absorbent
material. Bulk sludges as well as liquids are accepted. Leachate col-
lection and subcell segregation are similar to New York landfills.
Some of the criteria to keep in mind that secure landfills use are:
•NY—No free liquid by state regulations.
•NY— <5 % (approximately 1 in) free air space in drums.
•NY, AL—No reactives. At this writing this rule is in a state of
flux for both locations. Currently NY permits <1% CN and
Alabama permits <5% CN.
•NY—Bulk sludges must be solid enough for a man to walk on in
order to be acceptable. If drums, material must not flow if the
drum is tipped over.
•AL—Liquids are accepted in bulk, however, they are actually
mixed with absorbent material and landfilled as solids.
•NY—No water solubles are acceptable, e.g. nitrates.
•NY—PCB solids only. The amount of PCB residue must be ex-
pressed in Kilograms and the amount of filler material also ex-
pressed in Kilograms must be provided.
•AL—PCB liquids <500 ppm accepted. This will also be land-
filled after being solidified.
•NY—Flash point of solids must be <80%.
•AL—Flammable solids are acceptable if the drums are banded
together and then banded to pallets.
•NY—An EP Toxicity test on the leachate is usually required be-
fore a solid is approved for acceptance.
2. Deep Well Injection
Wastes are injected under pressure to depths of 2,800 ft into
porous rock (usually sandstone). The porous rock rests on im-
permeable granite layers. Before injection, the wastes are blended
into acidic lagoons and then filtered.
Criteria for deep well injection are:
•Wastes must meet a relatively tight suspended and settleable sol-
ids specification to prevent clogging of the sandstone layer.
•Streams usually preferred are inorganic with low organic con-
tamination. The receiving lagoons are acidic so as to minimize
precipitation of solids, therefore precluding wastes which would
generate toxic fumes, layer, or cause odor problems upon acidi-
fication (e.g. sulfides, mercaptans, etc.).
224
-------
ULTIMATE DISPOSAL
225
3. Encapsulation
This technique employs a cement-like matrix to encapsulate the
wastes into a non-leaching solids. Some companies are able to use
the solids as fill or building materials. Stabilization works best on
materials like pickle liquors and other inorganic waste streams.
Most often small organic contamination adversely affects the
solid's solubility.
4. High Temperature Incineration
This method thermally degrades wastes which are of greater than
average toxicity, usually very stable, and not readily treated using
less expensive technology. Most incinerators blend wastes with sup-
plemental fuel and inject them into the burn area. Temperatures of
approximately 2,000 °F and dwell times over 2 sec provide enough
energy to break the aromatic ring compounds and split halogens
from carbon. There is usually a chamber to further burn the waste
gases before cooling and passing through a caustic scrubber, then
up the stack. The criteria for incineration are:
•The solids must be able to pass through the nozzle readily. Each
incinerator has its own specification, however, they usually can
accommodate solids which can pass through a 60 to 80 mesh
screen.
•The thermal value has a considerable effect on the disposal
price. Obviously an aqueous stream which has to be slowly in-
jected, using up a great deal of supplemental fuel, will be more
expensive to dispose of.
•The amount of halogens and alkali metals in the waste stream al-
so has a significant bearing on the price since these contaminants
deplete the caustic scrubber solution and attack the refractory
linings of the equipment.
pH LOWERED TO
2.5 WITH ACID
DISSOLVED IRON
•The ash content or ash on ingition also is a factor which incin-
erators have to consider. The ash disposal cost to the incinerator
and is passed along to the generator, usually as an incremental
cost per gallon of waste.
•Some cement companies and more recently steel companies are
paying for organic wastes with good thermal value. They have
strict specifications for halogen, metals, and sulfur content, how-
ever. Generally a waste is a candidate for this method if the halo-
gen content is <2%. The metals and sulfur must not jeopardize
the burners air emission permits, or cause undue damage to the
refractory lining here also. The thermal value should exceed
15,000 Btu/lb.
5. Recovery
Although the amount of reclamation in the U.S. is increasing the
country has a long way to go to approach European successes.
Various US regional industrial waste exchanges are having increas-
ingly good results. Checking these should become a standard part
of procurement procedures.
The most common area of reclamation is chlorinated solvent
recovery. Many companies have stills which can reclaim 1,1,1-tri-
chloroethane, trichloroethylene, methylene chloride, and per-
chloroethylene. Some reclaimers have refrigerated columns and can
reclaim Freons as well. Other larger reclaimers have fractional
distillation columns and work on a larger bulk scale.
The single most important criterion for recovery of these solvents
is that the different species must be kept separate when collecting
the wastes. Most of these solvents form azeotropes when mixed.
This means that they do not boil off at their respective boiling
points when being distilled. Instead they both come over at a new
boiling point. This makes its resale virtually impossible. Instead of
a valuable waste product, the generator now has an expensive
liability.
pH RAISED TO 11.5 SLURRY TO
WITH HYDRATED LIME AIR AGITATION
TANKS
PRETREATMENT OF
HEX-CHROME WITH
S02
OPTIONAL
OXIDATION
PRETREATMENT
WITH HYPO
PRETREATMENT
AIRSTRIPPING
DISCHARGE
SOLIDS TO SECURE
CHEMICAL LANDFILL
OPTIONAL
Figure 1. OPTIONAL
Chemical treatment flow chart
-------
226
ULTIMATE DISPOSAL
The same holds true for metals. Some reclamation of metals (e.g.
Cr, Cu) is economically feasible today if they are kept separate, and
not in a mixed metal hydroxide sludge.
6. Chemical Treatment
This technology (Fig. 1) usually employs ferro-lime treatment of
aqueous streams with small to relatively large concentrations of
organics and metals. First, iron is added to the waste. Normally fer-
ric chloride is used, but many times spent pickle liquors are
employed as a substitute. Also ferrous sulfate crystals can be dis-
solved and used. The addition of an acidic iron solution serves: to
adjust the pH (usually to approximately 2), as a filter aid later on in
the system; and as a reducing agent. Next the pH is adjusted back
up to 11-12 with NaOH or lime. NaOH has the advantage of not
generating much sludge; however, it tends to produce more soluble
salts. Lime does improve the filterability of the sludge but generates
more of it. This treatment step forms insoluble metal hydroxides
which start to precipitate. At several points in this initial pH adjust-
ment sequence, or after filtration, oxidizing agents such as
chlorine, hypochlorite, or hydrogen peroxide may be introduced to
attack the organic constituents of the stream.
The heavy metal hydroxide slurry is now ready to be dewatered.
Methods employed include centrifugation, rotary vacuum filtra-
tion, and plate and frame filtration. The latter seems to be the most
forgiving in terms of ability to filter a less homogeneous feed. The
filter cake, approaching 50% solids by weight, is laden with heavy
metals and a considerable amount of organics which the iron seems
to attract. This cake is disposed of in chemical secure landfills.
After another pH adjustment downward to neutral, the superna-
tant is now ready for secondary treatment. Most companies use
carbon adsorption to further remove organics. Some companies
utilize biological treatment before or after this step. The aim with
activated carbon, whether it be powdered or granular, is to remove
organic materials, long-chained aliphatics and ring compounds.
These chemicals are generally the more toxic organics and include
halogenated compounds, pesticides, herbicides, and other potential
carcinogenic or bio-accumulative compounds.
The final effluent is discharged into a waterway or to a municipal
wastewater treatment plant which in turn holds the permit to
discharge into a waterway.
Other waste streams cannot be directly introduced into this ferro-
lime system. Wastes like cyanides for example must first be
pretreated using alkaline chlorination before the reaction products
are put back into the main waste feed. Sulfides also are first broken
down in this fashion. Ammonia containing streams must first be air
stripped. Hexavalent chromes must be reduced to trivalent.
Some of the criteria used by chemical treatment plants in
evaluating wastes are:
•For both drummed wastes or in bulk there should be a descrip-
tion of layering, solids, and how the material became a waste.
Layering generally refers to different liquid phases, for example
20% top oil layer, 10% middle emulsified oil layer, and 70%
bottom aqueous layer. The term solids to a TSDF does not just
mean hard, dry material. It includes that of course, but also is
taken to mean sludges, and non-pumpable viscous material like
some resins.
Many times an item can only be identified by a trade name or
its manufacturer will not divulge its proprietary composition.
This is why it is often very helpful for a TSDF to know how that
item was used. What was the process that changed the commod-
ity into a waste and what contaminants could have been picked
up in that process? An experienced staff at the TSDF has prob-
ably seen 90% of the total variety of waste streams generated
within a 400 mile radius. The waste generation information will
often help the TSDF's laboratory to determine a starting point.
For example, an aluminum anodizing operation usually signals
the presence of hexavalent chrome. Any strong caustic stream is
always suspected of containing cyanides. If a person tells the
TSDF that the waste came from a cadmium plating bath, the
suspicion is nearly always confirmed. If, however, the caustic
stream was called aluminum De-Smut, it would be highly un-
likely to contain cyanide.
•Is the waste aqueous or organic? If it is aqueous, is it acidic or
alkaline/neutral? If acidic, TSDFs need to know the type of acid.
Sulfuric acid is generally handled easily. Corrosion problems oc-
cur at medium concentrations, however. Hydrochloric acid is
also relatively simple to treat, however, corrosion problems are
significantly greater. Nitrating acids have the potential for re-
leasing NOx fumes and cannot be handled by many TSDFs. Hy-
drofluoric acid is difficult to handle because its corrosivity is
similar to HC1 with respect to stainless steel, but it also attacks
the glass fibers in many fiberglass liners or reinforced tanks.
Therefore the TSDF must be extremely careful to greatly dilute
these acids before treatment. The following components are also
important information for a TSDF when evaluating acids: pH,
free and total acidity, hexavalent chrome concentration, total
organic carbon (TOC), chemical oxygen demand (COD), metal
concentration, and phenol concentration.
If the waste is alkaline or neutral, evaluations will be based
on: pH, free and total alkalinity, TOC/COD, amount of solids,
and the presence and concentrations of ammonia, cyanide, met-
als, sulfides, and phenols.
The above factors are critical in both the acidic and alkaline/
neutral categories because the TSDF will be doing treatability
studies to answer questions such as:
•What is the initial organic concentration?
•How much acid is required to lower the pH?
•Are there any fumes liberated upon acidification?
•What metals are present?
•How much alkaline solution is required to raise the pH?
•What kind of temperature rise is associated with the acid and
lime steps?
•Are the reactions controllable?
•How well does the metal hydroxide slurry filter?
•What is the solids concentration?
•How well are the organics removed after ferro-lime treatment?
•How much carbon is consumed per gallon of waste?
•What is the final organic concentration after carbon treatment?
•How much further oxidation is required with hypochlorite,
chlorine, peroxices, permanganates?
•Will the waste require, and is it amenable to, biological de-
gradation?
•How much hypochlorite and caustic will be needed to com-
pletely destroy any cyanides?
If the waste is organic, it is usually broken down into flam-
mable or halogenated categories.
•Flammable organic criteria are: what types of compounds,
thermal value per pound or gallon, halogen content, alkali met-
als concentration, metals such as lead and chrome which would
cause air emission concern, water content, both free and en-
trained, specific gravity, suspended and settleable solids, size of
solids and presence and concentration of PCBs.
•Chlorinated organics criteria are: specific gravity, % recover-
able yield, species, whether two or more species mixed, oil conp
tent, flammables present, PCB concentration, (see Fig. 2).
SUGGESTIONS ON IMPROVING ABANDONED
SITE BID PACKAGE PREPARATION
Now that the basics for treating wastes by the most common
types of TSDFs have been outlined, some direction can be given to
the approach taken when preparing abandoned site categorizations
and requests for proposals (RFP).
-------
ULTIMATE DISPOSAL
227
Drums or Bulk
Layering, Solids, How Generated
Aqueous Organic
Acidic Alkaline/Neutral Flammable Halogenated
Type pH Type Specific Gravity
pH Free & Total BTU Value % Recoverable
Alkalinity Yield
Free & Total TOC/COD % Halogens Species Type
Hex Chrome Solids Alkali Metals Species Mixed
TOC/COD Ammonia Lead, Chrome Oil Content
Metals Cyanide Sulfur, Water Flammables
Phenol Metals Specific Gravity PCBs
Sulfides Suspended Solids
Phenols Settleable Solids
Particle Size
PCBs
Figure 2.
Criteria for evaluating common waste streams
This is an appeal for practicality. The overall concern at aban-
doned sites should certainly be safety. However, due to the emo-
tional impact of the term hazardous waste in the post-Love Canal
era, governmental agencies at all levels have required increasing
levels of overkill. Overkill when it comes to site characterizations
by engineering firms, which becomes magnified in the directions
those firms give to TSDFs in the RFP. Consequently an RFP may
become so specific, down to the last detail of how a site should be
cleaned up, that the RFP effectively precludes techniques that
might be just as safe and many times more practical and cost sav-
ing.
It is not the purpose of this discussion to change the direction of
government procedures. Instead, this discussion is directed at the
consulting engineers who categorize and prepare RFPs for aban-
doned sites, and corporations' normal waste disposal needs. You
may be constrained to fulfill the same governmental requirements
as in the past, probably more in the future. But, if you try to incor-
porate the following suggestions in your work, the net result will be
reduced cleanup time and lower overall cost to everyone. Hopefully
this will have the added benefit of stretching superfund dollars,
enabling more sites to be cleaned up sooner.
First of all, much of the chemical analyses provided is extensive,
elegant, and unnecessary. A complete metals scan does not con-
tribute to the information needed by a TSDF. Most of these metals
will be removed during treatment anyway. Likewise a complete
priority pollutant scan for every so many drums is very expensive
and also superfluous. Assuming the TSDF is using proper oxidizing
agents and carbon adsorption, the organics in question will be
reduced so that the TSDF's discharge is acceptable. If the disposal
method is high temperature incineration, the toxic organics will
also be degraded properly. Moreover most laboratories can only
perform 10 scans at a time, with a 4 to 6 week lag in providing
results.
Physical information is often overlooked. It is important to note
the presence of layers in tanks or drums and estimate the amount of
layered material within each vessel. For example, information like
the following is virtually worthless: out of 8,000 drums, 3,000 were
found to contain a top organic layer or a composite sample of these
drums showed metals and priority pollutant concentrations as
follows; etc.
Much more valuable information would look like this: out of
8,000 drums, 3,000 contained a top organic layer which averaged
20% of the drum volume. Of these, 2,100 appeared to be a medium
viscosity petroleum hydrocarbon with a flash point of ;>200°F. Of
these 2,100 drums, 75 were found to contain organo-chlorine levels
of >2% but <10% and 530 had >10% chlorine with the balance
below 2%. The 900 remaining layered drums showed a flash point
of <100°F and had an odor of xylene. Chlorine levels were all below
2%. In all 3,000 organic-layered drums PCBs were found to be less
than 50 ppm using 10 drum composites, etc.
Likewise physical information should be provided about the
solids content of the drums. The condition of the containers and
the need for overpacks is important. How many drums are open
top (lid and ring) and how many are bung type is useful. If the
material is in tanks, are there solids and layering in them? An
estimate of how much solids and their pumpability is needed. What
kind and sizes of fittings or access openings are available to a
tanker is good information. A real bonus would be the availability
of samples for bulk materials or even drum composites so that
TSDF's could perform their own treatability studies. If the site
needs considerable excavation work, a two stage bid should be
prepared. The first, to contract experts in this phase with adequate
supervision for safety considerations. Then the wastes would be
more closely examined and a more meaningful RFP could be
engineered for the second cleanup phase.
Some practical financial considerations should also be con-
sidered. A lump sum bid requirement for the entire cleanup or even
on individual phases is the least cost saving way of all. With so
many unknowns at an abandoned site, a TSDF which is forced to
respond with one price for the job will most certainly build in
enough cushion to guarantee a profit. It will probably be
somewhere in the area of the most pessimistic costing for all of the
wastes thought to exist at the site. If the agency overseeing the ad-
ministration of the contract insists on this type of lump sum
response, a means of adjustment should be provided in case of ex-
tenuating circumstances. For example, some RFPs provide that if
an error of ± 10% in the estimated volume of any given line item
results in >± 5% total bid cost variation, then adjustments will be
made. These can be in the favor of either the agency or the contrac-
tor.
The best type of proposal as far as both fairness and cost effec-
tiveness is concerned would be a combination of a unit price and a
cost-plus. The cost-plus aspect of the job would have to be strictly
supervised to be sure. However, this would give the agency via the
consulting engineering overseer even better control of the costs and
disposal methods to be employed. Flexible bid options are also a
good idea. A contractor should be able to quote as directed, but
also include a second price if he could do the work a different way.
For example, he might say, "our price is x, but since we will be us-
ing high temperature incineration for item y, our price will be z due
to the elimination of 5 priority pollutant composite scans."
Certain calculated risks have to be considered when dealing with
the practical aspects of bid requirements. For example, if after
categorizing the wastes on site it can be safely assumed that
radioactives and explosives are not present, requirements that
radiation-bomb proof buildings be constructed for these materials'
storage are unnecessary. A phrase in the RFP that could save con-
siderable time and expense would ask that these be built if these
types of items were(discovered.
Likewise, on most abandoned sites much of the ground is
permeated with a number of different wastes. It seems most un-
necessary to require that extensive decontaminated areas, docks for
quality control, washing areas for trucks, etc., be constructed only
to have them all excavated and landfilled upon site cleanup comple-
tion.
Smaller items but ones that are often overlooked are: much more
time should be provided for the TSDF to quote; technical contacts
and phone numbers need to be provided; workers who are required
to wear the complete self-contained breathing apparatus and moon
suits in the summer might last 15 minutes at a time. If they were
able to carry air packs or other emergency escape devices with self-
contained devices in stationed areas, efficiency would be greatly im-
proved without sacrificing safety; keep in mind the costs of various
laboratory analyses. Sometimes requiring a drum by drum PCB
analysis or small 5 drum composite sample analyses of a large
number of drums will result in the laboratory and regular disposal
cost far exceeding the cost of treating the drums as if they all con-
tained PCBs.
Again, by being aware of the treatment methods available today
and following some practical suggestions on approaching the pro-
blems associated with hazardous waste evaluation, engineers can
significantly reduce costs to the taxpayer and industry alike. This
will contribute to a cleaner environment by directly reducing the
costs of cleanup of the hundreds of sites targeted around the coun-
try. It will stretch federal and state superfunds further to do more.
-------
ABOVE GROUND STORAGE OF HAZARDOUS WASTE
CHRISTOPHER J. LOUGH
MARK A. GILBERTSON
Pope-Reid Associates, Inc.
St. Paul, Minnesota
STEPHEN D. RINER
Minnesota Waste Management Board
Crystal, Minnesota
INTRODUCTION
A long-term monitorable and retrievable storage facility for hazar-
dous wastes was investigated as an alternative to the land disposal of
hazardous wastes in Minnesota. The proposed conceptual design and
subsequent analysis was conducted for the Minnesota Waste
Management Board as part of Minnesota's Comprehensive Haz-
ardous Waste Management Plan.
During the statewide meetings regarding the management of Min-
nesota's hazardous waste, public sentiment was strongly in support
brSbove-ground storage as an alternative to a hazardous waste land
disposal facility. Prompted by public interest and a lack of definitive
information on this concept, an in-depth analysis of the technical,
economic, and institutional feasibility of above-ground storage was
undertaken.
Monitorable and retrievable storage describes a secure system for
the long-term management of hazardous wastes in anticipation of the
development of new economical recycling, treatment or destruction
technologies. Wastes stored above-ground remain visible where a
leak can be detected and contained before release to the environment
occurs. An above-ground storage facility has applicability to wastes
generated in the state as well as those recovered from uncontrolled
hazardous waste sites in the upper midwest. Wastes accepted by such
a facility would likely be those which:
•There does not currently exist practicable alternatives to land-
fillings
•Are considered too toxic, persistent, and leachable for land dis-
posal
•In the case of wastes removed from an uncontrolled site, are
stored while the responsible parties are identified and litigated
The study identified the types and quantities of Minnesota's wastes
which cannot be reasonably recycled, treated, or destroyed and
analyzed the regulatory requirements, design and operating
parameters, environmental aspects, costs, and economic feasibility of
above-ground storage.
WASTE INVENTORY
The quantity of Minnesota's hazardous waste available for long-
term storage was estimated by completing the following tasks:
•Identifying the types and quantities of hazardous waste generated
each year
•Specifying the likely management technologies for each waste
stream category
•Determining those waste streams for which the only management
technology is secure landfilling
•Identifying those treatment technologies which produce a haz-
ardous residual
The waste categories and quantities are presented in Table 1, and
are based on management plans reported by waste generators to the
Minnesota Pollution Control Agency and the Metropolitan Counties
through December, 1981. As indicated in the table, approximately
52,000 tons/yr (TPY) of hazardous wastes were reported to be
generated in Minnesota at the time this study was undertaken. Subse-
quent plan data received by the regulatory agencies and estimates
made by the Waste Management Board since the study's completion
indicate that the actual waste generation amount is higher. The
relative mix of waste types did not change significantly, however.
Waste streams were then assigned probable management
technologies such as oil and solvent recovery, incineration,
neutralization, precipitation, and biological treatment processes that
exist within the State. Technologies that generated a hazardous
residual (e.g., sludge, ash, etc.), which would normally be landfilled,
were also identified. Estimates of waste and residual quantities (and
their physical characteristics) for each selected management
technology were prepared. Wastes which could neither be treated nor
destroyed would require a secure landfill, and consequently became
candidates for above-ground storage.
An annual amount of 6650 tons, or approximately 13% of Min-
nesota's hazardous waste stream, are available for above-ground
storage. This total quantity represents 4450 TPY which cannot be
practically treated plus 2200 TPY of residual sludges generated by
various waste treatment technologies. Nearly 90% (5900 TPY) of the
wastes available for long-term storage are sludges/solids that require
containerization. The remaining 10% of the wastes are liquids that
would be stored in above-ground tanks.
Table 1.
Summary of Minnesota's Reported Hazardous Wastes11'
Waste Quantity
(tons/yr)
Waste Category
1. Solvents
2. Halogenated Hydrocarbons
3. Cyanides
4. Oils
a. Petroleum oil
b. Organic, synthetic, and other mixed unknown
5. Other Organic Wastes
a. Paints/inks
b. Organophosphates
c. Mixed, unknown pesticides
d. Phenol
e. Polymerics
f. Polynuclear aromatics
g. Other, mixed, and unknown, general
h. Photochemicals
6. Acids
7. Aklalis
8. Metals and Toxic Non-Metals
9. React!ves
a. Oxidants
b. Ammonium salts
c. Other and mixed, unknown
6,884
1,550
163
15,795
2,214
1,258
20
21
51
590
196
878
4,249
4,168
7,536
4,368
1,488
145
339
TOTAL 51,913
REGULATORY BACKGROUND
Federal and State hazardous waste regulations, as well as ap-
propriate fire and building code requirements, were reviewed and
applied to the proposed facility design. In addition to requirements
assuring container and tank integrity, waste compatibility, and
monitoring procedures, proposed State hazardous wastes regula-
tions2 require secondary containment more stringent than existing
Federal rules.3 Container and tank storage areas holding free li-
quids must be designed and operated with the following major pro-
visions:
228
-------
ULTIMATE DISPOSAL
229
•An impervious base that contains all leaks, spills, and accumu-
lated precipitation
•A drainage system to remove standing liquids, unless the con-
tainers or tanks are protected from contact with accumulated
liquids
•A containment system which is able to contain 10% of the vol-
ume of the containers or tanks (or volume of largest container/
tank, whichever is greater).
Fire protection standards regarding the construction and opera-
tion of liquid storage facilities were evaluated from the National
Fire Protection Association Code.4 The code specified minimum
distances between adjacent tanks, elements of the containment
dikes, and fire control system requirements that were useful in
developing the facility design.
Building code standards' for the proposed container storage
buildings specified additional fire protection measures. Using Type
I fire resistive construction, the allowable floor area for buildings*
one story in height containing various classes of flammable and
hazardous materials is 15,000 sq ft. Buildings of this type include
steel, iron, concrete or masonry structural elements that provide
four hours of fire protection to exterior bearing and non-bearing
walls, and three hours to interior bearing walls and structural
frame. Container buildings were conceptually designed in accor-
dance with these and other applicable requirements, such as set-
back, access, aisle spacing, and air exchange provisions.
•Each portion of a building separated by one or more area separation walls
may be considered a separate building provided the area separation walls
meet building code requirements. Area separation walls shall be not less
than four-hour fire-resistive construction in Type 1.
CONCEPTUAL DESIGN AND ENGINEERING
Based on the quantities of waste allocated to above-ground
storage, the facility was designed to store 22,000 drums in a con-
tainer building and 185,000 gal in bulk-liquid tanks each year. By
TANK
0
i
.-4
0
i
1—
._«•*. -
e)
tJ
m
.*•_
•D 1
3 *
i '
i
-------
230
ULTIMATE DISPOSAL
•A packaged wastewater treatment system will be installed. Di-
lute, decontamination wastewaters from spills on roads, receiv-
ing areas, or in buildings will be conveyed via the sewer system to
the wastewater treatment system.
•A waste supply, storage, and distribution system will be pro-
vided, and a sewer system will be constructed. Provision will be
made to limit contaminated water or waste liquids entering the
sewer system.
•A fire detection system, consisting of fire, sprinkler, and stand-
pipe alarms, control panels, and remote annunciators will be in-
stalled and monitored in the control room.
•A fire pump, hydrants, hoses, and other miscellaneous fire
equipment such as nozzles, couplings, and portable extinguishers
for all lift trucks and maintenance vehicles will be provided.
•Groundwater monitoring wells will be installed on a basis of one
well per 20 acres of area, with a minimum number of 4 wells.
•All roads will be asphalt paved and curbed with 6 in. concrete to
contain any spills or leaks.
•A microcomputer, printer, and storage discs will be purchased for
inventory control. Functions to be provided include: waste
identification, location, quantity, type, generator, storage date,
and special considerations or handling procedures.
Container Storage Design Assumptions
•Container storage buildings will be constructed as two units over
a two year period, each unit has a capacity for one year of drums.
•Each completed container building will have a capacity for two
years of drum generation (approximately 44,000 drums).
•Storage buildings will be further divided into a ten week operat-
ing bay due to limitation of allowable floor space of 15,000 sq ft.
•Drums will be placed on steel storage racks, four high, with the
maximum number of drums in a double row section to be 64.
•The container building will be constructed with fire resistive,
preformed insulated concrete panels and steel roof joists.
•Each building will contain an automatic sprinkler system and
heating and ventilation components.
•Each ten week operating area will be separated by a fire wall and
automatic overhead fire doors on each aisle.
•Monitoring instrumentation to detect harmful ambient air levels
of a variety of organic constituents will be installed.
DRUM STORAGE ELEVATION
TEN WEEK STORAGE PLAN
ONE YEAR STORAGE PLAN
Figure 2.
Container Storage Plan
-------
ULTIMATE DISPOSAL
231
•Each building will contain floor drains and sumps to collect and
store temporarily any liquid spills that cannot be immediately
contained and removed with absorbent materials and redrummed.
•A central drum receiving building will be constructed near the
container storage buildings. This building, which will have an
available floor space for five days of inventory, will serve as the
truck receiving, record-keeping, testing, and pick-up area for
drums entering the storage buildings.
•The redrumming area will be a separate, well ventilated portion
of the building, complete with spill containment diking, drains,
monitoring instrumentation, and an automatic fire suppression
system.
•The drum receiving area will also contain a redrumming area and
equipment used to recontainerize leaking containers in overpack
drums.
Tank Storage Design Assumptions
•Each individual tank will be sized to a capacity approximately
equal to five years generation of a specific waste type.
•The materials specified in the construction of each tank were se-
lected based on their compatibility with the individual waste types.
•All tanks will be insulated and vented to condensers that recycle
the condensate to the tank.
•All tanks will include continuous level monitoring instrumen-
tation.
•A concrete dike will be constructed around each tank to contain
the contents of the full tank, plus an allowable freeboard of at
least four inches, and to keep individual waste streams from com-
ing in contact with each other in the event of a tank failure.
•The dike network and tank foundations will be constructed with
concrete and sealed to form an impervious barrier to waste spills.
•Service roads will be constructed between the diked tank area to
provide easy access for fire control, tank maintenance, and
spill/precipitation removal.
The design of this facility was based on the assumption that the
State's waste stream would remain relatively constant and few, if
any, stored wastes would be removed for processing during the
operating life of the facility. These assumptions were made in order
to determine the maximum amount of storage space needed for the
State. If, during the lifetime of the facility, some wastes are remov-
ed for processing or disposal, the resulting storage space would be
available for incoming waste. Coupled with a projected decrease in
the amount of incoming waste due to the advent of new treatment
technologies, the design capacity of the facility could be reduced.
Predictions of future treatment capabilities and their affects on
Minnesota's waste stream are, of course, difficult to make. Any
potential reduction in the design capacity, however, would likely be
offset by the acceptance of materials from uncontrolled hazardous
waste sites.
OPERATIONS
From an operating standpoint, the most important consideration
is the condition in which the drummed wastes are received. The
condition of the container will determine the associated risks and
costs of the material during its storage. To help assure container in-
tegrity, while minimizing repacking costs, containers will be in-
spected in the receiving area for proper labeling, content and condi-
tion. Containers failing the facility standards during the initial
check-in will be denied long-term storage and returned to the
generator.
In addition, a program will be instituted to monitor all container
failures or related problems necessitating drum repacking. Data
collected in this program will be periodically analyzed to determine
the probable reasons for failure. Facility receiving requirements,
handling procedures, and inspection schedules will be modified
based on program results in an effort to reduce the number of con-
tainer failures. With a concerted effort to reduce the likelihood of
container deterioration and failure, drum repacking rates would be
minimized. Other important operating conditions and assump-
tions, as well as some resulting cost components, are presented
below:
•Container repacking requirements were based on an assumed re-
placement rate of 1 % of the accumulated inventory each year
(excluding repacks).
•Damaged or leaking containers will be repacked in 80 gal over-
pack drums, absorbent added, and the drum returned to a con-
tainer building.
•Costs for repacking include contracted labor, overpack drum,
absorbent and miscellaneous items (protective clothing, spill con-
tainment materials, decontamination chemicals, etc.). Total re-
packing costs were estimated at $150/drum.
•Utilities and fuel cost estimates were based on natural gas re-
quirements for heating 'all buildings, electricity for building
ventilation, tank heating, pumps, office, lift-trucks, and lights,
and fuel requirements for maintenance vehicles and standby
electric generator.
•Laboratory analysis of incoming wastes will consist only of such
basic analysis as pH, ignitability, corrosivity, etc. More de-
tailed analyses, when required, will be conducted under a con-
tract established with a private analytical laboratory.
•Sufficient water storage and supply will be maintained for
sprinkler and fire hydrant demands for a duration of two hours.
•Disposal costs of the inventory at facility closure were esti-
mated by allocating an average cost of $200/ton for transporta-
tion and disposal of the waste at a secure landfill in Illinois.
ENVIRONMENTAL ASPECTS
The proposed above-ground storage facility would significantly
reduce the environmental impact of a land disposal facility by pro-
viding monitorable and retrievable storage of hazardous wastes
within controlled environmental conditions. Wastes in the pro-
posed storage facility can be inspected easily and problems repaired
quickly before they develop into significant environmental hazards.
The sources of potential environmental impact from the above-
ground facility are container failures, spills during mateials
transfer, fires or explosions, and volatile emissions. Specific
designs and operating procedures were developed to mitigate or
minimize these potential hazards. Leaks or spills resulting from
container or tank failures, for instance, will be contained in either
the storage buildings (each with concrete walls, floors, and floor
drains) or in the tank basin which contains an impermeable base
and diking. Both the container building and tank basin are con-
structed to contain the material until it can be removed.
Spills incurred during materials transfer, either in the drum
receiving buiding where repacking is conducted, or in the liquid
unloading structure, will be contained and removed. The drum
repacking area will be constructed with spill containment walls and
drains. The liquid unloading structure will contain an underground
holding tank for emergency by-pass, an impermeable foundation,
curbing, and emergency pumping capabilities. All site roads will be
paved and curbed for spill containment.
The provisions to prevent and control the spread of fires at the
site are extensive. Briefly, all container buildings will be sprinkled
and possess fire partitions. The drum receiving building and liquid
unloading structure will also be sprinkled. A site fire detection and
control system will be present. In addition, potentially reactive
materials will be segregated within facility buildings and tank areas.
Emissions of organic vapors will be present from the tanks and
container buildings. These emissions, however, will be minimized
through the use of insulation of the tanks and buildings to control
temperature fluctuations, the use of internal floating roofs and
condensers on specific tanks, and the placement of ambient air
monitors in the buildings and other site locations to monitor emis-
sion rates.
Other measures taken to assure the environmental integrity of
the site and safety of the public include:
•Placement of groundwater monitoring wells
•Security fencing and personnel to prevent unauthorized entrance
•Monitoring instrumentation, communication, and process control
located in central control room
-------
232
ULTIMATE DISPOSAL
•Auxiliary power supply
•Rigorous facility standards, inspection schedules and personnel
training
•On-site wastewater storage and treatment
•Drum repacking capabilities to mitigate or prevent leaks
•On-site laboratory analyses
•Drum failure analyses and preventative program
COSTS AND ECONOMIC FEASIBILITY
Facility investment costs were prepared and are summarized in
Table 3. These costs would be incurred in the development of the
site prior to actual facility operation. Significant component costs
in this develoment stage are the bulk-liquid tanks (5-year storage
Table 3.
Initial Facility Investment Costs
Investment Cost
1. Land
2. Fencing & gates
3. Security station
4. Facility hdqtrs.
5. Parking area
6. Scale
7. Signs
8. Drum receiving bldg
9. Communication system
10. Bldg Tire detection
11. Water distribution
system
12. Auxiliary power supply
13. Groundwater moni-
toring wells
14. Inventory control
computer
IS. Wastewater treatment
system
16. Office equip/furn
17. Laboratory equip
18. Site clearing
19. Site vegetation
20. Outdoor lights
21. Gas distribution
22. Visual screening
23. Site roads, curbing
& paving
24. Tank truck unload
25. Tank basin & dike
construction
$300,000
78,000
1,500
180,000
18,000
41,000
500
80,000
tem 20,000
9,000
220,000
ply 40,000
8,000
14,000
•nt
66 000
16,000
25,000
73,000
23,000
24,000
*>l f¥Yl
Zl ,IAJU
4,000
243,000
11,000
95,000
26. Tank Costs— 5 yr
storage capacity
•paints & inks
•Organophosphates &
pesticides
•Other mixed organics
•Reactives
•Reactives
•Reactives
•VOC removal
•Delivery pumps
•Emergency pump
•Retrievable pump
Tank Subtotol
(includes fees)
27. Facility vehicles
•Lift trucks
•Maintenance trucks
28. Air monitoring instru-
mentation
29. Container storage
bldg (2 yrs)
30. Rack storage constr
31. Subtotal (1 )
32. Engineering/con-
tractor's fee (15%)
33. Subtotal (2)
34. Contingency (15%)
35. Subtotal (3)
36. Tank Subtotal
(fees included)
37. Total
443,000
130,000
707,000
146,000
44,000
48,000
72,000
67,000
15,000
24,000
$1,696,000
90,000
25,000
32,000
4,536,000
440,000
$6,734,000
$1,010,100
$7,744,100
1,161,000
$8,905,700
1,696,000
110,602,000
capacity) and the construction of a container storage building
(2-year storage capacity). Investment costs would also be incurred
during the life of the faculty. On a yearly basis, with the exception
of the first and last operating years, the facility will require the con-
struction of an additional container building unit, storage racks,
and air monitoring instruments. After five years, costs will also be
incurred for new lift trucks, maintenance vehicles, tank basin, and
bulk-liquid storage tanks. All facility cost estimates are in second
quarter 1982 dollars.
Annual facility operating and maintenance costs were estimated
and appear in Table 4. These costs assumed that the facility would
be owned and operated by the state. Other ownership options were
considered in the study, but are not included here. A state owned
and operated facility would be exempted from property taxes, and
would assume all site liabilities during operating and post-closure
periods.
An annual revenue requirement of $7.2 million was determined
for the state owned and operated facility. No return on investment
was assumed, and a 9% capital recovery rate was applied. Costs for
final disposal of the inventory were also included. The price for
hazardous waste storage at the facility (under state ownership and
operation) was nearly $1100/ton. Other ownership options that in-
clude private investors and operators increase the price to approx-
imately $1300/ton.
CONCLUSIONS
In summary, the above-ground storage facility would present
hazards similar to those from manufacturing operations where
hazardous materials are handled. Within the facility, workers
would bear some risk of exposure to hazardous wastes. The risk to
the environment and to persons outside the facility from an acci-
dent, although real, would be small since the facility would contain
safeguards against foreseeable accidents, and operating procedures
would be designed to minimize the impact from any accidents.
There are no technological barriers to storing hazardous wastes in
any form. It was possible to conceptualize a facility meeting all cur-
rent hazardous waste regulations and building codes which would
serve the intended purpose.
The high storage costs were due to significant capital investment
(mainly buildings) necessary for the storage facilty. Changing
design or operating assumptions to account for removal of wastes
during the operating life of facility, variations in ultimate process-
ing cost (or credit), or increasing the proportion of wastes stored in
bulk did not lower the cost of storage to a level comparable to other
waste management methods.
Year
982
983
984
985
986
987
988
989
990
99 1
992
0
1
2
3
4
5
6
7
8
9
1
Total:
Investment
$10,602,000
„
3,298,000
3,298,000
3,298,000
5,166,000
3,298,000
3,298,000
3,298,000
3,298,000
0
$38,854,000
Drum
Repacking
Table 4.
Annual Operating and Maintenance Costs
Total
Labor
Utilities
&Fuel
Lab Sup
& Contr
Wastewtr
Trtmt
Off Oper
Supplies
$ 33,000
66,000
98,000
130,000
162,000
193,000
224,000
255,000
285,000
316,000
$522,000
522,000
522,000
522,000
522,000
522,000
522,000
522,000
522,000
522,000
$139,000
198,000
258,000
317,000
382,000
442,000
501,000
561,000
620,000
620.000
$25,000
25,000
25,000
25,000
25,000
25,000
25,000
25,000
25,000
25,000
$15,000
15,000
15,000
15,000
15,000
15,000
15,000
15,000
15,000
15,000
$5,000
5,000
5,000
5,000
5,000
5,000
5,000
5,000
5,000
5,000
Main!
Supplies
Total Ann
O&M
$106,000 $ 845,000
139,000 970,000
172,000
205,000
257,000
290,000
323,000
356,000
389,000
389,000
,095,000
,219,000
,368,000
,492,000
,615,000
,739,000
,861,000
,892,000
$14,096,000
REFERENCES
1. Overall reported amounts from disclosure documents submitted to the
Minnesota Pollution Control Agency and license applications submit-
ted to the Metropolitan counties by Minnesota industries through De-
cember 1981.
2. Minnesota Pollution Control Agency, Proposed Hazardous Waste
Regulations, Minnesota State Register, 6MCAR 4.9101-4.9560, June
7, 1982.
3. Federal Register, 7. January 12, 1981, 46, No. 215, November 6, 1981.
4. Flammable and Combustible Liquids Code, NFPA 30, National Fire
Protection Association, Inc., Boston, Mass., 1981.
5. Uniform Building Code, 1982 Edition.
-------
UNCONTROLLED HAZARDOUS WASTE SITE CONTROL
TECHNOLOGY EVALUATION PROGRAM
RONALD HILL
NORBERT SCHOMAKER
IRA WILDER
U.S. Environmental Protection Agency
Solid and Hazardous Waste Research Division
Cincinnati, Ohio
INTRODUCTION
In anticipation of the passage of the Comprehensive Environ-
mental Response, Compensation, and Liability Act of 1980
(CERCLA or Superfund),1 the Office of Research and Develop-
ment, of the USEPA began a program in 1980 to support the
Agency's activities concerned with uncontrolled hazardous waste
sites. In the area of environmental engineering and technology, the
Agency looked to the ongoing and established program in the Solid
and Hazardous Waste Research Division of the Municipal En-
vironmental Research Laboratory. This Division had a base of ex-
pertise that could quickly relate to the uncontrolled hazardous
waste problem.
The Oil and Hazardous Materials Spills Branch (OHMSB),
located at Edison, New Jersey, had been actively pursuing re-
search on the identification, containment, control, removal, and
ultimate disposal of hazardous spills since 1971. These activities
could be directly related to the removal aspects of Superfund. The
Disposal Branch, located in Cincinnati, Ohio, has been actively
pursuing research in the area of waste disposal control to the land
since 1965. These activities could be directly related to the remedial
action aspects of Superfund.
Since CERCLA only provided for a five year program for the
uncontrolled hazardous waste site problem, time was not available
to establish a fundamental research and development program.
The approach taken by the Agency for the Office of Research and
Development was one of technical support to the Office of Emer-
gency and Remedial Response. Technologies that had been devel-
oped under the Clean Water and Solid Waste programs were
adapted to the uncontrolled hazardous waste site situations. In
addition, construction techniques, e.g., slurry trench cutoff walls,
injection grouting, and chemical stabilization, that had been used
for other purposes, were evaluated to determine their applications
to uncontrolled sites. Also there were very limited data available
on the cost and effectiveness of various remedial techniques. The
task of collecting and analyzing the available data was initiated.
Once CERCLA became law, the Office of Research and Devel-
opment developed, in consort with the Office of Emergency and
Remedial Response, a five year support strategy which is updated
each year. The strategy outlined a program with peak funding in
the early years to meet the immediate needs of the program; the
latter years concentrates on technical assistance.
The environmental engineering and control technology research
program divides activities along the lines of CERCLA, i.e., re-
moval and remedial actions. In the following sections, details of
each program are presented.
REMOVAL TECHNOLOGY PROGRAM
Approach
The overall goal of the Oil & Hazardous Materials Spills Branch
is to provide scientific and engineering expertise in the area of re-
moval (emergency response) activities. Specifically, the OHMSB
evaluates and demonstrates new or improved equipment, devices,
systems, and data analysis techniques for the prevention, iden-
tification, containment, control, removal and ultimate disposal of
hazardous substances released to the environment. This includes
the cleanup and recovery of hazardous substances from accidental
releases as well as from uncontrolled hazardous waste sites, and is
consistent with the definition of "removal" in Section 101 of the
Superfund legislation. Additionally, the OHMSB demonstrates the
equipment and systems to actively encourage the commercial use
of cost-effective, advanced technologies during cleanup operations.
Once an item is completed and has undergone various field test-
ing, the plans, specifications, and other information are made
available publicly for the purpose of encouraging commercializa-
tion of the new technology. Numerous systems, including a mobile
water treatment unit and a mobile laboratory, have been com-
pleted and are now available commercially.2
The OHMSB also provides input into regulation development,
enforcement, and technical support needs of the USEPA's Pro-
gram Offices and the ten Regional Offices. Regulation develop-
ment is being assisted through technical background investigation,
such as updating the list of Superfund designated hazardous sub-
stances and attendant "reportable quantities." Regulation support
is also provided by evaluating new cleanup techniques that will be
incorporated in the regulations by reference. In this area, the
OHMSB provides user manuals for emergency response, includ-
ing sampling and analysis, monitoring techniques, technology eval-
uation, and guidance for on-scene response personnel. In addition
to specific projects addressing decontamination of personnel and
equipment, and specialized protective clothing for personnel, all
projects in this program area place special emphasis on personnel
health and safety.
Major Outputs to Date
Mobile Incineration System
The OHMSB recently completed construction of a mobile in-
cineration system3 designed for field use to destroy hazardous
organic substances collected from cleanup operations at spills and
at uncontrolled hazardous waste sites. The system is designed to
the USEPA's PCB destruction specifications (under the Toxic Sub-
stances Control Act) to provide start-of-the-art thermal detoxifica-
tion of long-lived, refractory organic compounds. Hazardous sub-
stances that could be incinerated include compounds containing
chlorine and phosphorus (i.e., PCB's, kepone, dioxins, and organ-
ophosphate pesticides) which may be present in sludges or in
soils. In order to systematically evaluate and demonstrate the
equipment, a trial burn is currently underway.
Ultrasonic Submerged Pollutant Detector
Using existing ultrasonic reflectometry technology, a detector—
a sophisticated "fish-finder" for locating insoluble hazardous sink-
ers (chemicals that sink instead of float or are soluble) at the bot-
tom of waterbodies—has been developed. The detector measures
233
-------
234
RESEARCH AND DEVELOPMENT
variations in acoustic return echoes, and can be used to unique-
ly identify the acoustic "signature" of a sunken pollutant. Dur-
ing its development, the device was used to profile a spill of approx-
imately 350,000 gal of toxic ethylene dichloride into Lake Fergu-
son near Greenville, Mississippi.4 The device performed exception-
ally well and located pools of pollutant ranging in depth from less
than '/2 in. to 20 in.
Hazardous Material Spill Case History Computer System
A computerized data base system,' which provides a centralized
information bank of past hazardous substances incident response
experiences has been developed and is currently being evaluated.
This system will serve to aid on-scene personnel in deciding what
treatment or technique to use, what degree of cleanup to employ,
and what priorities to initiate for cleanup in relation to environ-
mental fate and effects.
The system is based upon a standardized after-action data re-
port form which is to be filled out by On-Scene Coordinators,
their advisors, or a trained interviewer at the conclusion of a haz-
ardous material incident. The report is in a format and arranged
in such a way that the experiences can be subsequently retrieved
for use by others who may be facing the same or similar situa-
tions. The computerized data base is continually updated with
after-action report forms.
Field Test Kit for Measuring Redox Potentials of Waste Chemicals
A field test kit, for measuring oxidation-reduction (redox) po-
tentials of organic and aqueous waste chemicals, has been devel-
oped and evaluated. Using the test kit, measurements can be made
of the redox potential by using a portable, battery-operated instru-
ment containing electrode probes and electrolyte solutions. The en-
tire procedure for obtaining redox measurements requires only a
few minutes and can be performed by inexperienced operators. The
redox kit was developed as a screening procedure for segregating
drums at uncontrolled hazardous waste sites to preclude the
danger of an explosion due to a reaction between oxidizing and
reducing agents. Field evaluations of the redox kit were success-
fully performed during Jan. and Nov. 1981.
Major Future Outputs
Mobile Soils Washer
A mobile treatment system has been designed for on-site ex-
traction of a broad range of hazardous materials from excavated
soils.' The system is expected to be an economical alternative to
the current approach of excavation, hauling off-site to a landfill,
and replacing the excavated soil. The system can be used to extract
contaminants from soils—"artificially leaching" the soil using
water—and enabling operators to leave the treated soil on-site.
The extracted hazardous materials are separated from the washing
fluid using physical/chemical treatment procedures. The cleaned
washing fluid is recirculated, and the separated and concentrated
hazardous materials are disposed of by appropriate means. The
system is currently undergoing shakedown tests and is expected to
be available for field demonstration during FY-83.
Mobile Carbon Regenerator
Water contaminated with hazardous substances has been suc-
cessfully cleaned using water decontamination equipment such as
the USEPA's mobile physical/chemical treatment system.7 This
system, which utilizes granular activated carbon to concentrate
dilute dissolved organic contaminants, can be made more cost-
effective with on-site regeneration of the spent carbon as opposed
to transporting the carbon for off-site regeneration or placing it
in a secure landfill.
In order to provide a safe and effective method for handling
contaminated carbon, the OHMSB has developed a mobile unit
for detoxifying/regenerating the carbon at the cleanup site.1 The
system has recently undergone initial shakedown and preliminary
testing, and is expected to be ready for field demonstration and
evaluation during FY-83.
Mobile In-Situ Containment/Treatment System
The OHMSB has developed an innovative, mobile system for
treating contaminated soils in place at reduced costs, in terms of
dollars per pound of contaminant removed.' The technique em-
ploys flushing with additives and detoxification by chemical re-
action. In-situ containment is accomplished by the mobile unit
through direct injection of grouting material into the soil around
the contaminated area in order to isolate the released chemicals.
The chemicals are then treated in place by water flushing with
additives, or by other methods such as oxidation/reduction, neu-
tralization, or precipitation. The collected chemically contaminated
wash solution can be processed through a mobile water treat-
ment unit where contaminants are removed. The mobile in-situ
containment/treatment system is currently undergoing shakedown
tests and will be available for field evaluation during mid FY-83.
Manuals
The OHMSB is currently preparing documents for release to the
user community during FY-83/84. Each of these user-oriented
field manuals is being prepared in close cooperation and coordi-
nation with representatives of private organizations who would po-
tentially use the manuals. These manuals are the following:
•Environmental Emergency Control Handbook for First Re-
sponders which will cover specific environmental-related practices
to assist first-on-scene personnel, such as firefighters, in their de-
cision-making process during the first critical minutes of a haz-
ardous substance spill or release incident, where fire is not in-
volved.
•Manual on Physical and Chemical Countermeasures which will
provide general recommendations for using physical and chemical
countermeasures to mitigate frequently occurring hazardous sub-
stance releases in subsurface soils and in large, relatively quiescent
waterbodies such as lakes, ponds, canals, and slowly moving riv-
ers. These recommendations will take the form of a matrix of
countermeasures versus release types and will be applicable to the
cleanup of spills as well as uncontrolled hazardous waste sites.
•Spill Prevention Manual, which will provide a matrix for various
classes and groups of chemicals and spill-prevention techniques
for these chemicals. This matrix will be developed primarily
through communication with trade associations (such as the
Chemical Manufacturers Association and others) and organiza-
tions engaged in producing, storing, and transferring hazardous
substances. The manual will also be developed into a training
course/workshop.
REMEDIAL TECHNOLOGY PROGRAM
Approach
The overall goal of the Disposal Branch is to assess and val-
idate new or improved remedial action technologies or schemes to
minimize pollutant release from uncontrolled hazardous waste dis-
posal sites. More specifically, the remedial activity includes site
survey and assessment studies, bench and pilot studies, field ver-
ification studies, cost-effectiveness and model studies. These stud-
ies are being performed to validate control technologies as they
relate to surface water control, groundwater control, plume man-
agement, chemical immobilization, and excavation and reburial.
The activities are consistent with the definition of remedial ac-
tion in Section 101 of Superfund (CERCLA).
The Disposal Branch has pursued activities for new landfill
design in the research areas of pollutant identification, pol-
lutant generation, pollutant transport, pollutant control, and
economics. These activities have direct relationships to the remedial
action schemes for uncontrolled landfill sites. These research areas
include bench, pilot, and field studies accompanied by the predic-
tive modeling work. This research activity has produced eight tech-
nical resource documents10"17 which reflect best engineering judg-
ment for the design of waste disposal facilities as relate to land-
fills, land treatment, and surface impoundments.
-------
RESEARCH AND DEVELOPMENT
235
The Disposal Branch also assists the Office of Emergency
and Remedial Response and several of the Regional Offices in
the areas of regulation development, technical and enforce-
ment support, and assistance in the development of the National
Contingency Plan." This support typically includes the develop-
ment of technical documents describing the design and construc-
tion of a variety of remedial action schemes which could be util-
ized as control measures at uncontrolled landfill sites.
Major Outputs to Date
Handbook—Remedial Action at Waste Disposal Sites
The Disposal Branch, in consort with the Office of Emergency
and Remedial Response, recently published the subject technology
transfer document." With this information the reader can then
develop a preliminary remedial action plan and cost estimate. The
objectives of the Handbook are twofold: (1) to provide the reader
with a generalized understanding of the pollutant pathways in-
volved in waste disposal sites, the remedial actions as they apply
to each pathway, and the process of selecting the appropriate re-
medial actions; and (2) to provide detailed information on specific
remedial actions including applications, state-of-the-art, design,
construction, and/or operating considerations, advantages, dis-
advantages and costs.
Remedial Actions at Hazardous Waste Sites: Survey and Case
Studies
A survey20 of 169 waste disposal sites was performed to identify
what type of remedial action was implemented. Technologies em-
. ployed at these sites included: (1) containment, (2) removal of
wastes for incineration or secure burial, (3) institution of surface
water controls,' and (4) institution of groundwater controls. A
major deficiency of this study was that only 9 of the 169 sites were
able to be investigated in detail. The other 160 sites were given
only a cursory survey investigation. Remedial measures usually
consisted of containment of contaminants or waste removal. The
survey determined that a lack of sufficient funds and/or selection
of improper technologies was responsible for remedial actions hav-
ing been applied effectively at only a few of the uncontrolled waste
disposal sites. This survey is currently being updated.
Guidance Manual for Minimizing Pollution from Waste Dis-
posal Sites
The purpose of this manual21 is to provide guidance in the selec-
tion of available engineering technology to reduce or eliminate
leachate generation at hazardous waste disposal sites. The manual
emphasizes remedial measures for use during or after closure of the
sites. Some of the measures are passive, that is, they require little
or no maintenance once emplaced. Others are active and require a
continuing input of manpower or electricity.
Block Displacement Technique of Waste Isolation
A field demonstration of a technique to construct a clay isola-
tion barrier around hazardous waste sites was recently completed.
The block displacement technique is accomplished through a phase
sequence of drilling, fracturing, and bentonite slurry injection
around the bottom and sides of a waste disposal site with the resul-
tant upheaval of the waste site to form a block, isolated by an
impermeable bentonite barrier. USEPA was unable to demonstrate
the full isolation at the study site, especially in the vertical plane,
because of certain site specific anomalies such as the presence ,of
tree roots below the local groundwater level and the presence of a
variable iron-cemented strata immediately overlying the hori-
zontal plane of bentonite injection across the bottom. However,
there was evidence that the bentonite slurry did_penetrate the hor-
izontal fracturing plane.
Guidance Manual for Slurry Trench Design and Installation
A Guidance Manual for slurry trench cut-off wall design, con-
struction, and performance evaluation is near completion. It pro-
vides recommendations on a variety of scientific and technical
parameters relevant to using this approach to isolate hazardous
chemicals in near-surface groundwater regimes. The accomplish-
ment of this effort required extensive information gathering and
integration of technical data gathered from a diverse array of ex-
perience and authorities.
Conclusive recommendations reported within the Guidance
Manual were determined by investigating areas of influence. These
areas included historical perspective, present methodology, chem-
ical compatibility tests, shortcomings of common backfill mater-
ials, positive recommendations for resistant backfill materials,
detailed site soil and geologic characteristics which impact cut-off
wall success, vegetation, checklists of design construction, per-
formance factors, and documenting the quality and performance
of completed construction.
Major Future Outputs
Physical and Hydrogeologic Models for Hazardous Waste Sites
Remedial action alternatives for uncontrolled hazardous waste
sites must be described in terms of attenuation of mitigation of
existing or eminent public health/environmental problems, cap-
ital costs, O&M costs, design life, and risk of failure. This task will
develop two complimentary levels of modeling.
Level 1 will be a relatively detailed modeling level for spe-
cific site engineering, and will consider detailed site factors, con-
taminant migration, detailed models of technologies, interrelation-
ship among technologies, costs, design lifes, and risk of failure.
Level 2 will be simplified desktop procedures for use primarily
by state and federal personnel for problem assessment, preliminary
screening of the cost and effectiveness of remedial actions, and
rapid review of remedial action plans. Level 2 will be based on sen-
sitivity and factor analysis of key site and technology charac-
teristics using the detailed Level 1 models. Both levels will describe
the effectiveness of remedial actions considering site factors and
characterization of the control technologies.
Cost Analysis of the Effects of Human Safety and Degree of
Hazard as They Affect Remedial Actions at Uncontrolled
Hazardous Waste Sites
This study will seek to determine the factors which contribute to
the increased costs and ascertain the magnitude of additional costs
associated with various components of remedial action unit opera-
tions. Primary source for data will be the private contractors and
project officials having knowledge of the specific elements of pro-
ject costs, and the manner in which these vary as a result of prox-
imity to hazardous waste materials. Estimates of the additional
costs incurred will be indicated in terms of percentages of ordinary
or unusual costs, and in absolute terms where appropriate. The
information produced will be valuable to program offices and
others in evaluating costs of remedial actions.
Update of Survey Information on Completed and Ongoing
Remedial Action Efforts
Since the time the study "Remedial Actions at Hazardous
Waste Sites: Survey and Case Studies"20 was made more recent
information on remedial actions has become available. There is
an obvious need to bring together and analyze the current up-
to-date information, including effectiveness and cost. This in-
formation will serve as a foundation for future decision making
with regard to presently uninitiated remedial action efforts.
Reliability of Available Technologies When Considered with Cost-
Effectiveness
This study has been initiated to develop procedures for con-
ducting the cost-effectiveness analysis at uncontrolled hazardous
waste sites. Various remedial action options are available for any
uncontrolled hazardous waste site. In broad categorical areas they
are: alternative measures, active measures, and passive measures.
The "alternative measures" category is meant to include such
actions as moving the affected population away from the site or
providing alternative water supplies. The "active measures" would
-------
236
RESEARCH AND DEVELOPMENT
include treatment scenarios applied directly to the site such as
excavation and reburial in a secure site, waste stabilization, neu-
tralization, treatment and/or elimination of problem. "Passive
measures" would include entombment or isolation techniques us-
ing slurry walls, grout curtains, capping, bottom sealing, etc.
Collection of Data on Compatibility of Grouts with Hazardous
Wastes
Available data from the literature and industrial sources will be
collected on the compatibility of various types of suspension and
chemical grouts with various classes of hazardous wastes and leach-
ates. In addition, information will be collected on the procedures
available to test the durability of grouts in the presence of haz-
ardous wastes and leachates.
In Situ Treatment Techniques Applicable to Large Quantities of
Hazardous Waste Contaminated Soils
The project addresses detoxification of large quantities of haz-
ardous waste contaminated soils located at Superfund or other un-
controlled hazardous waste sites. In situ chemical and/or biolog-
ical treatment methods presently available will be identified and
evaluated. The feasibility and effectiveness of these methods will
be assessed based on waste, soil type, site conditions and economic
considerations. The output, a technical handbook, will include per-
tinent information concerning soil sorption and chemical and soil
interaction influences on waste degradation. The remedial action
identified as the most promising of the evaluated methods may
be applied in a follow-up study (Phase II) on a full scale basis
at a Superfund site.
Field Evaluation of Drum Encapsulation Techniques
A process for encapsulating drums containing hazardous waste
is being demonstrated. Efforts will include evaluating the over-
pack/cover weld, resistance to physical stress (drop test, punc-
ture resistance, etc.), and equipment performance. Mobility of
the process and equipment is important and evaluations will devel-
op criteria for mounting the hardware on a flat bed tractor/trailer
vehicle.
Development of Methods and Pilot Test for In Situ Hazardous
Waste Stabilization by Injection Grouting
This project will provide pilot scale tests to predict applicability
to specific sites with specific waste compounds and expand the
state-of-the-art to hazardous waste in situ stabilization. The
pressure injection of grout to a variety of waste types will be in-
vestigated to develop a matrix of grout types to waste types with
appropriate grout pressures and tube spacings included.
Manuals
The Disposal Branch is currently preparing documents for re-
lease to the user community during FY-83 and FY-84. Each of these
user-oriented field manuals is being prepared in close cooperation
with representatives of private organizations, and State and Fed-
eral agencies who would potentially use them. These manuals in-
clude the following:
•User Guide for Evaluating Remedial Action Technologies: The
objective of this task is to produce a Remedial Action Technical
Resource Document describing how the technologies and methods
for evaluating proposed RCRA new hazardous waste disposal
sites can be applied to site-specific remedial response activities
for uncontrolled hazardous waste sites. The Remedial Action
Document will be based on the state-of-the-art technical and
cost information in eight TRDs10"17 for design and evaluation of
new hazardous waste disposal sites under RCRA. That informa-
tion will be reviewed for relevance to remedial response at un-
controlled hazardous waste disposal sites, and will be edited and
refocussed to address the needs of personnel involved in re-
sponse and remedial action planning under CERCLA.
•Guidance Manual for Fixation/Solidification of Wastes in Sur-
face Impoundments: This project will define the limits of applic-
ability of fixation/solidification techniques to remedial actions at
uncontrolled sites when considered in perspective with alterna-
tive or competing options. Information will be drawn from avail-
able fixation/solidification technology. The major problem with
using these techniques at uncontrolled hazardous waste sites is
that the composition of the waste is often unknown.
•Guidance Manual for Cover Design and Installation: A specifica-
tion manual for the selection, design, and installation of covers or
surface caps for uncontrolled hazardous waste sites is being pre-
pared. Much of the information developed from two existing
EPA reports10'22 will be incorporated into this manual that will
be specific to the problems of uncontrolled sites.
REFERENCES
1. U.S. Congress, Public Law 96-510, "Comprehensive Environmental
Response, Compensation, and Liability Act of 1980," Washington,
D.C.1980.
2. Bennett, G.F., Feates, F.S. and Wilder, I., "Hazardous Materials
Spills Handbook," McGraw-Hill Book Company, New York, NY,
1982, pp 9-24 through 9-39.
3. Brugger, J.E., Yezzi, J.J., Jr., Wilder, I., Freestone, F.J., Miller,
R.A., and Pfrommer, C., Jr., "The EPA-ORD Mobile Incinera-
tion System: Present Status," Proc. of the 1982 Hazardous Ma-
terials Spills Conference, Milwaukee, WI, Apr. 1982, 116-126.
4. New York Times Newspaper, "Missing Chemicals Found in a Lake in
Mississippi," New York, NY, Sept. 28, 1981.
5. Meyer, R.A. and Stone, W.L., "Development of a Hazardous
Substance Incident Data Base for Response Personnel," Proc. of
the 1982 Hazardous Materials Spills Conference, Milwaukee, WI,
Apr. 1982,381-387.
6. Scholz, R. and Milanowski, J., "Mobile System for Extracting Spilled
Hazardous Materials from Excavated Soils," Proc. of the 1982 Haz-
ardous Materials Spills Conference, Milwaukee, WI, Apr. 1982, 111-
115.
7. Gupta, M.K., "Development of a Mobile Treatment System for
Handling Spilled Hazardous Materials," EPA-600/2-76-109, USPEA,
Cincinnati, Oh., 1976.
8. Griwatz, G.H. and Brugger, J.E., "Activated Carbon Regeneration
Mobile Field-use System," Proc. of the 1978 Hazardous Materials
Spills Conference, Miami Beach, FL, Apr. 1978, 350-355.
9. Huibregtse, K.R. and Kastman, K.H., "Development of a System to
Protect Groundwater Threatened by Hazardous Spills on Land,"
EPA-600-2/81-085, USEPA, Cincinnati, Oh, 1981.
10. USEPA, "Evaluating Cover Systems for Solid and Hazardous Waste"
Report SW-867, Washington, D.C., Sept. 1980.
11. USEPA, "Hydrologic Simulation on Solid Waste Disposal Sites,"
Report SW-868, Washington, D.C. Sept. 1980.
12. USEPA, "Landfill and Surface Impoundment Performance Evalua-
tion," Report SW-869, Washington, D.C., Sept. 1980.
13. USEPA, "Lining of Waste Impoundment and Disposal Facilities,"
Report SW-870, Washington, D.C., Sept. 1980.
14. USEPA, "Management of Hazardous Waste Leachate," Report SW-
871, Washington, D.C., Sept. 1980.
15. USEPA, "Guide to the Disposal of Chemically Stabilized and Solid-
ified Wastes," Report SW-872, Washington, D.C., Sept. 1980.
16. USEPA, "Closure of Hazardous Waste Surface Impoundments,"
Report SW-873, Washington, D.C., Sept. 1980.
17. USEPA, "Hazardous Waste Land Treatment," Report SW-874,
Washington, D.C., Sept. 1980.
18. 40 CFR "National Oil and Hazardous Substances Contingency Plan,"
Federal Register/47, No. 137, Friday, July 16, 1982.
19. USEPA, "Handbook-Remedial Action at Waste Disposal Sites,"
report EPA-625/6-82-006, Washington, D.C., 1982.
20. USEPA, "Remedial Actions at Hazardous Waste Sites: Survey and
Case Studies," report EPA 430/9-81-051, Washington, D.C. 1982.
21. USEPA, "Guidelines Manual for Minimizing Pollution from Waste
Disposal Sites," report EPA 600/2-78-142, Cincinnati, Oh, 1978.
22. USEPA, "Design and Construction of Covers for Solid Waste Land-
fills, report EPA-600/2-79-165, Cincinnati, Oh, 1979.
-------
APPLICATIONS OF SOLUBLE SILICATES AND DERIVATIVE
MATERIALS IN THE MANAGEMENT OF
HAZARDOUS WASTES
ROBERT W. SPENCER
RICHARD H. REIFSNYDER
JAMES C. FALCONE, JR., Ph.D.
The PQ Corporation
Lafayette Hill, Pennsylvania
INTRODUCTION
Soluble silicates are currently being used in chemical based
methods of waste disposal, but the authors feel they could be utilized
much more. This report is a review by the authors of the literature
pertaining to waste treatment for disposal, with emphasis on
solidification of liquids and sludges. The chemical methods currently
used, using both silicate-based, and those that use silicates indirectly,
are described. Some theoretical papers with implications for this use
are reviewed, and possible new ways of using silicates are described:
•Solidifying solvents for transportation
•Grouting landfills to reduce permeability and divert or block sub-
surface flows
•Modifications of other processes to improve leaching and physical
properties
SOLUBLE SILICATES
The soluble silicates useful for waste management are a class of
materials made by reacting NAjCO3 and quartz sand, SiO2, in oil- or
gas-fired open hearth regenerative furnaces.' This reaction;
nSiO
CO
(1)
where the "ratio", (n) varies between 1 %) silica poly-
mer "sol" crosslinking occurs and a gel network is formed.2'3 The
rate of gel formation is maximal near neutrality and increased silica
concentrations generally increase the rate of gelation and gel
strength.
Other matrix components may have a significant effect on
mechanical strength. These gelatin properties account for their
usefulness in soil stabilization and for diversion or blocking of
subterranean fluid flows.
Reaction with Multivalent Ions
The reaction of soluble silicates with metal ions under ambient
conditions are generally thought to lead to mixtures of silica gel and
hydroxylated metal ion gels. Specific interactions with the silica are
expected to be primarily due to cation exchange, i.e.,
SSi-OH + Me
H
0)
In more dilute solutions, the activities of several metal ions in the
presence of higher ratio soluble silicates were suppressed to a great
degree than expected based on solution pH value alone.4 This may be
attributed to a stronger affinity of the metal ion for the surface of the
more condensed silicate species.
In more practical studies Connors5'6 and Gowman' have shown
that the metal ion leaching from solidification methods employing
soluble silicates is significantly reduced compared to hydroxide
precipitation.
Ponomareva et al.s suggest that the removal of cations from
solution by calcium silicate slags proceeds by the dissolution of the
calcium silicate, which provides silicate ions, followed by the
precipitation of heavy metal silicates:
^ (CaSi03)m.n + nCa** + nSio3-- (4)
nSi03-- + m Men+ ^± Mem (SiO3)n (5)
However, the dissolution of the slag is fairly slow and limits the rate
of removal of the cations. The authors have shown that the in-
troduction of silicate ions into the system as sodium silicate results
in much faster reaction with the heavy metals, and decreases the
amount of slag needed.
The use of Ca(OH)2 and other alkalies (including sodium silicate)
for the removal of heavy metal cations, in the absence of slag,
results in the formation of fine precipitates with poor settling
characteristics and dewatering properties. However, the continued
use of slag and sodium silicates gives a sludge with good dewatering
properties, because the slag serves as a support for the epitaxial
deposition of precipitates.
237
-------
238
RESEARCH AND DEVELOPMENT
Environmental Acceptability
Since silicates are made from sand and alkali, they are universally
judged to be ecologically acceptable. Recently the Task Force on
ecological effects of non-phosphate detergent builders' concluded
that sodium silicate is neither toxic nor otherwise harmful to the
fresh water environment. Evidence was presented as part of this
report that the presence of silicates in waste water being treated led
to an inceased collection of metal ions in the waste sludge. High pH
values were identified as a possible problem; however, silicates used
in waste treatment would presumably be neutralized and harmless
to the environment.
In summary, it appears that the chemical properties of these
ecologically safe soluble silicates make them interesting raw
materials for the modification for matrix leachabiiity
characteristics through reduction in matrix permeability, diversion
of fluid flows and enhanced metal ion affinity.
GENERAL CHEMICAL METHODS
OF WASTE TREATMENT
The term solidification implies methods that fix or encapsulate
wastes in solid matrix end products. Fixation processes bind the
wastes by chemical or physical means, using solidification agents.
Encapsulation methods physically surround the wastes with the
agent. These methods have been used in the U.S. for radioactive
wastes, but until the RCRA and other environmental protection
acts were passed, most non-radioactive wastes have been disposed
of by the less acceptable methods such as lagooning, sewering,
landfilling in an unfixed state, bleeding into stream, etc. Under
pressure from the recent environmental protection laws, many
more wastes are being solidified.
Pojasek10'" discusses methods for converting hazardous wastes
to non-hazardous substances. He describes five basic technologies,
one of which can be adapted to handle any specific waste:
•Silicate and cement based
•Lime-based
•Thermoplastic-based
•Organic polymer-based
•Encapsulation techniques
Table 1, taken from Pojasek's work,10 gives the advantages and
disadvantages of these solidification processes. The cement- and
lime-based methods are suitable for toxic inorganics and stack gas
scrubber sludges, but are generally not suitable for toxic anions and
organics. The other three methods are suitable for toxic inorganics,
but are not suitable for strong oxidizers and organics.
Maugh12 has also reviewed the hazardous waste problem, and
discusses various general methods of coping with it. He emphasizes
that the non-nuclear is much greater than the nuclear waste pro-
blem. As of 1979, about 5,000 tonnes of nuclear wastes had been
accumulated since the beginning of the nuclear era; this is four
orders of magnitude smaller than the volume of non-nuclear toxic
waste generated in one year.
Table 1.
Technical Comparison of Solidification Processes
Proc»i$ Advantages
Cement-based 1. Additives are available at a reasonable price.
2. Cement mixing and handling techniques are
well developed.
3. Processing equipment is readily available.
4. Processing is reasonably tolerant of chemical
variations in sludges.
5. The strength and permeability of the end-
product can be varied by controlling the
amount of cement added.
Lime-based 1. The additiv.es are generally very inexpen-
sive and widely available.
2. Equipment required for processing is simple
to operate and widely available.
3. Chemistry of pozzolanic reactions is well
known.
Thermoplastic 1. Contaminant migration rates are generally
lower than for most other techniques.
2. End-product is fairly resistant to most
aqueous solutions.
3. Thermoplastic materials adhere well to
incorporated materials.
Organic polymer
Encapsulation
1.
2.
Only small quantities of additives are
usually required to cause the mixture to set.
Techniques can be applied to either wet or
dry sludges.
3. End-product has a low density as compared
to other fixation techniques.
1. Very soluble contaminants are totally iso-
lated from the environment.
2. Usually no secondary container is required,
because the coating materials are strong and
chemically inert.
Disadvantages
1. Low-strength cement-waste mixtures are often
vulnerable to acidic leaching solutions. Extreme
conditions can result in decomposition of the
fixed material and accelerated leaching of the
contaminants.
2. Pretreatment, more-expensive cement types, or
costly additives may be necessary for stabiliza-
tion of wastes containing impurities that affect
the setting and curing of cement.
3. Cement and other additives add considerably to
weight and bulk of waste.
1. Lime and other additives add to weight and
bul k of waste.
2. Stabilized sludges are vulnerable to acidic
solutions and to curing and setting problems
associated with inorganic contaminants in the
waste.
1. Expensive equipment and skilled labor are
generally required.
2. Sludges containing contaminants that volatilize
at low temperatures must be processed care-
fully. •*'
3. Thermoplastic materials are flammable.
4. Wet sludges must beiried before they can be
mixed with the thermoplastic material.
1. Contaminants are trapped in only a loose resin-
matrix end-product.
2. Catalysts used in the urea-formaldehyde
process are strongly acidic. Most metals are
extremely soluble at low pH and can escape
in water not trapped in the mass during the
polymerization process.
3. Some organic polymers are biodegradable.
4. End-product is generally placed in a container
before disposal.
1. Materials used are often expensive.
2. Techniques generally require specialized equip-
ment and heat treatment to form the jackets.
3. The sludge has to be dried before the process
can be applied.
4. Certain jacket materials are flammable.
E\ccrpicd b> special permission from Clwmical Enginttnng, Augusl 13. copyright 1979. by McGraw-Hill, Inc.. New York. NY 10020.
-------
RESEARCH AND DEVELOPMENT
239
The approaches which Maugh suggests, in order of decreasing
preference, are these: eliminating the production of wastes, re-
cycling to recover values, chemical or biological degradation, in-
cineration, deep well injection, and finally solidification and land
filling. While not a preferred technique, an increasing amount of
toxic waste is being solidified, and at least 41 different companies
have developed proprietary processes for this step.13 Maugh divides
these into cement-based', pozzolanic or lime-based, thermoplastic,
and organic binders. Advantages and disadvantages for the various
techniques, as outlined by Maugh, agree with Table 1.
SOLUBLE SILICATE BASED
SOLIDIFICATION METHODS
Methods Using Soluble Silicates Directly
Chemfix: The earliest patents for a silicate-based solidification
system are those issued to Conners, of Chemfix, Inc., at that time a
Pittsburgh, PA Company.14'15 The patents describe the use of an
alkaline metal silicate and a setting agent containing polyvalent
metal ions (particularly those which provide free calcium ions) to
solidify waste sludges. They claim that wastes consolidated with
soluble silicate and cement give leachates with low heavy metals
concentrations. In fact, when leachates from raw, unconsolidated
wastes, containing relatively high toxic metals levels were per-
colated through beds of waste treated by the Chemfix method, the
metal concentrations in the original leachate are greatly reduced.
Reductions from 85 to 100% are found. The same is true for
cyanide, ammonia and nitrates, and chemical oxygen demand. The
treated waste is therefore recommended as a liner for landfills, and as
a cover for layers of unconsolidated waste.
The Chemfix process got considerable publicity in the 1970s16"24.
In addition Conners prepared a number of papers, information
packages, and data sheets describing the process and giving both
laboratory and field test results.6'7'25"28
Leaching tests were reported with H2SO4 and HC1 solutions at
pH values of 1.0 without significant solubility of the metals. (Con-
centrations of Cr, Cu, Fe, Mn, Ni, and Zn remained below 0.1
mg/1.) It is said that for many Chemfix sludges which contain high
levels of toxic heavy metals, the leachates easily meet USPHS drink-
ing water standards.
The Chemfix process increases the volume of the waste by
5-10%. Treated wastes can be used for landfill and can support
vegetation.
A recent British Patent Application29 describes a trailer-mounted
apparatus for mixing a setting agent and silicate with waste sludges
for solidification. An example cites 500 kg Portland cement and
0.225 m3 of sodium silicate for 5.5 m3 of waste sludge. This appears
to be very similar to the Chemfix method.
Hittman Nuclear: This company is engaged in the fixation and
transporting of low level nuclear wastes. The Hittman method30 is
to convert the waste slurries to a rock-like mass, using Portland ce-
ment as the main solidifying agent. A certain amount of anhydrous
sodium metasilicate is added to the slurry to give a harder product
and a quicker set time.
Elaborate conveying and feeding equipment is used to mix the
slurried wastes, cement and silicate in the proper proportions. The
treated wastes are dropped into disposable containers, sealed and
transported to secure landfills. The Hittman organization is plann-
ing to market their method and services in the non-nuclear hazar-
dous waste field.
This approach differs in two ways from the Chemfix method:
•Cement is the main ingredient, rather than the setting agent for
the silicate
•A solid anhydrous metasilicate is used instead of a high ratio
liquid
Hayes Method: This patent31 specifically addresses the fixation
of cesium. Limited success is obtained using clays, shales or the
Chemfix approach. Hayes claims that the use of sodium silicate, a
silicate hardening agent and finely divided shale results in decreased
mobility of the cesium. In this process the function of the shale is to
ion-exchange with the radioactive cesium. When silicate is present
in the solidifying mass, leaching of the cesium is decreased by a fac-
tor of 10.
Methods Using Soluble Silicates Indirectly
Soliroc Process: This process32 is based on the premise that heavy
metal hydroxides, precipitated in the pH range of 9-11, will dissolve
significantly when exposed to leachates at low pH values. The in-
ventors claim further that other silicate-based methods simply en-
capsulate the wastes; if the solidified waste products are crushed, as
in hauling, dumping or compacting, the resulting higher surface
areas would allow significantly increased leaching of the wastes.
In this approach, the heavy metals are dissolved in acid, and are
reacted at low pH with a "siliceous reagent". The mixture is then
adjusted to a higher pH value causing the silicates to polymerize
and gel. Lime and a setting agent such as cement are used for final
solidification.
The siliceous reagant (also called low molecular weight or low
polymerization silicic acid) is prepared by adding alkali or alkaline
earth silicates or aluminum silicate to an excess of mineral acid.
Residues or waste streams are suggested for both the silica source
and the acid starting materials. For instance, siliceous wastes such
as blast furnace slag or flue dusts are suggested for the silica source,
and waste plating solutions as the acid source.
The dissolved heavy metal ions react with the low molecular
weight silicic acids, and as the pH is raised to about 7 with an
alkaline material, the liquid starts to gel. Final setting is ac-
complished by the addition of more lime and cement. Hardening
takes about three days.
Low leachabilty and high mechanical strength are claimed, but
no information on these are given in the patent. Rousseaux and
Craig33 have evaluated the process. A blend of four wastes was
solidified by this process. The source of the silica was not specified.
Results for five runs, made under different conditions, are sum-
marized below:
Metal
Cd
Cr
Cu
Ni
Pb
Zn
Cone's
in Waste (mg/1)
2890
10390
7100
2640
1710
7650
Leachate Conc'n
for five runs (mg/1)
0.01
0.01
0.05
0.01
0.008
0.41
-5.6
-0.5
-0.64
-1.30
-0.017
-34.8
Compressive strengths ranged from 8 to 156 psi.
Ontario Liquid Waste Disposal, Ltd.: A patent describes the
following process34:
•Reacting the liquid waste with an acid ferrous solution at a pH
range 2-5
•Neutralizing the reaction mix with an alkali (NaOh or lime) to
raise the pH to the range 9-12, which yields an inert precipitate
•Adding a silicic compound to form large complex silicate mole-
cules
•Adding an alkaline earth material to raise the pH to the range
12-13
The silicic compounds for the third step can be sodium silicate,
but also cement, fly ash, general ash, siliceous slag, clay, silt, sand,
and siliceous stone or soil. The alkaline earth compounds for the
final step include lime, MgO, BaO, Ca(OH)2, and cement. Harden-
ing takes place over several days.
The compressive strengths and leaching properties of samples
cited in this patent were as good as the Soliroc samples. This treat-
ment suppressed the offensive odors of putrified wastes.
The Research-Cottrell approach35 is to blend fly ash with the
sludge which has been dewatered to about 69-70% solids. The pro-
portion of fly ash must be greater than 80% of the total to achieve a
dry, transportable product. This process is most suitable for low
-------
240
RESEARCH AND DEVELOPMENT
sulfur coal containing 2% S or less. For high sulfur coals, they
recommend the addition of an extra reactant, such as lime,
Portland cement, sodium silicate or other proprietary substances.
The Stablex Corp. patent" describes a process of treating hazar-
dous wastes with an aluminum silicate or aluminosilicate, and with
Portland cement, to form a slurry which will set into a rock-like
crystal matrix having low permeability and high ultimate strength.
The aluminosilicate is generally fly ash. The use of alkali
metasilicate to shorten set time is mentioned in the patent in one ex-
ample and in one claim. The process appears to be sorption of the
toxic wastes by the aluminosilicate, followed by the immobilization
of the loaded fly ash in a cement matrix.
The process has been used since 1974 in England, and has also
been used in Japan." Waste treatment units are currently being
built in the U.S." Stablex claims that a wide range of wastes, com-
prising about 70% of existing hazardous and toxic wastes, can be
handled.
GENERAL ASPECTS OF SILICATE
AS A MODIFIER
Heacock and Riches" compare three processes for the solidifica-
tion of radioactive wastes. These are Portland cement, sodium
silicate-modified Portland cement, and urea-formaldehyde. Unfor-
tunately the details of the processes are not given in the paper, and
neither the ratio nor the concentration of the silicate is specified. In
a table of ten solidification properties for the three processes, the
silicate-modified cement method equaled or exceeded the com-
peting processes in each case; in five comparisons, physical proper-
ties, shipping efficiency, residual water, and tolerance for chemical
composition of basic wastes and regenerant wastes, it was listed as
"best". (Shipping efficiency is related to the volume expansion
caused by the solidification step.)
In annual operating cost comparisons, the silicate-modified ce-
ment method showed a definite advantage: costs for the silicate-
cement approach were only 70-80% of those for the other two
methods. These results combined with the work of Ponomavera
suggested that soluble silicates could be beneficial additives to other
cement based processes, e.g., the Stabatrol Corp., Terra-Tite4
method and the Pec-Engineering, Petrifix4 method.
A SUMMARY OF RECENT LABORATORY STUDIES
A preliminary investigation of individual hazardous wastes was
conducted in PQ Corporation laboratories to determine the effec-
tiveness of sodium silicates in treating materials with a wide range
of physical and chemical properties. Wastes selected are listed in
Table 2; each was chosen because it represented a different disposal
problem. A variety of silicate based treatment processes were
employed to solidify, fix or encapsulate these materials and the
resulting waste/sludge mixtures were evaluated using standard
analytical procedures. The following is a brief summary of the
work done.
Sodium silicate consolidation and fixation of waste fly ash was
evaluated using two different types of ash (type A&B). While the
ashes chosen displayed comparable physical properties, the ashes
were quite different chemically (Table 2). Two silicate based techni-
ques were used with equal success in treating these materials, and a
lime-based solidification method was used as the standard treat-
ment. These techniques are outlined in Table 3 and the results for
physical strength and teachability are outlined in Tables 4 and 5.
Fly ashes consolidated using these silicate based methods displayed
better strengths and water resistance and reduced leaching
characteristics than the lime-based standard.
The benefits of treating municipal and plating wastes with
sodium silicates were assessed for two sludges containing a mixture
of organics and toxic heavy metals (analysis in Table 6).
These materials were consolidated using both a gel solidification
and a silicate-cement process (Table 7). Leachability of the
materials was evaluated using the EP batch leaching test and results
compared to those for the untreated wastes are shown in Table 8.
Both silicate based treatments resulted in a solid material able to
meet the EPA limits for heavy metal leaching. It has been proposed
Table 2.
Wastes Selected for Sodium Silicate Treatments
Waste
Fly Ash A
Fly Ash B
Plating
Sludge
Municipal
Sludge
Refinery
Waste
Liquid
Organic
Solvents
Criteria for
Hazardous Status
Corrosivity (low pH)
Respiratory irritant
(particle size)
EP toxicity (heavy metals)
Respiratory irritant
(particle size)
EP toxicity (heavy metals)
EP toxicity (heavy metals
and organics)
Corrosivity (low pH)
Ignitability (low flash pt)
Disposal
Problems
Material must be neutral-
ized and consolidated for
landfilling
Toxic metals must be fixed
for safe landfilling
Material must be consoli-
dated for landfilling
Toxic metals must be fixed
for safe landfilling
Solidification of sludge to
improve landfilling
Toxic metals must be fixed
for safe landfilling
Solidification of sludge to
improve landfilling
Waste must be neutralized
for landfilling
Solidification to improve
landfilling and contain
organic material
Flammability must be re-
duced for landfilling
Liquid material must be
solidified for transport
and landfilling
Table 3.
Fly Ash Solidification Techniques
Direct Solidification
Fly ash is wetted out with diluted sodium silicate solution.
Dilution factor and amount of solution added to ash were calculated
to deliver 4% SiO2 based on weight of material.
Material may be air set or oven set at 75 °C.
Gelled Solidification
Fly ash is slurried in 14% sodium silicate solution (4% SiO2 based on
weight of ash) and then gelled with 12% H2SO4.
Lime Treatment (Standard)
Fly ash slurried in H2O. 10% lime added based on weight of ash. Ma-
terial is air set.
Table 4.
Compression Results for Consolidated Fly Ashes
Fly Ash
A
B
Fresh*
(IHr)
64.3
58.4
Oven Dried* Water
(75 °C) Soaked*
(4 Mrs)
128.4 75.5
134.4 87.0
•Results reported in LBS/1N1
that this reduction in heavy metal leaching is due to the formation
of less soluble metal silicates and an improved containment matrix,
both directly related to the addition of sodium silicate.
Compatibility and effectiveness of silicate based treatment pro-
cesses with wastes containing high levels of organic and petroleum
by-products were determined during our investigation of refinery
sludge. The waste was composed of waste-produced from an acid-
oil treatment process that had become saturated with water. This
sludge had a low pH value, low heavy metal concentrations and
high sulfur content. Treatment was directed at achieving a solid
material with a minimum compressive strength of 1 ton/ft to allow
proper landfilling. The waste could be solidified by a number of
-------
RESEARCH AND DEVELOPMENT
241
Table 5.
Leaching Results for Fly Ashes A and B
All Concentrations In mg/l.
RESULTS FROM EP LEACHING TEST
FLY ASH A
HEAVY
METAL
CO
Cd
Zn
Sn
Cu
Ni
Cr
Pb
FRESH
0.31
0.86
36.0
0.57
2.0
1.6
0.14
3.2
GELLED
0.08
0.19
11.0
0.48
0.13
0.28
0.09
1.2
%
CHANGE
- 74
- 78
- 69
- 23
- 94
- 83
- 36
-167
FRESH
0.41
0.73
15.0
0.77
1.8
2.2
0.39
0.51
LIME
0.13
0.27
7.0
0.83
0.48
0.84
0.08
0.44
%
CHANGE
- 68
- 63
- 53
+ 8
- 73
- 62
- 79
- 14
FLY ASH B
HEAVY
METAL
Co
Cd
Zn
Sn
Cu
Ni
Cr
Pb
FRESH
0.12
64.0
8600
0.94
37.0
0.4
0.13
170
GELLED
0.07
24.0
4200
0.77
14.0
0.23
0.06
74
%
CHANGE
- 46.8
- 62.5
- 51.2
- 18.1
- 62.2
- 42.5
- 53.8
- 56.5
FRESH
0.12
64.0
8600
0.96
37.0
0.4
0.13
170
LIME
0.12
26.0
4100
1.1
20.0
0.28
0.26
350
%
CHANGE
0.0
- 59.3
- 52.3
+ 14.8
- 45.9
- 30.0
+100
+106
Table 6.
Analysis for Mixed Municipal and Plating Waste
% Solids
PH
COD
Arsenic
Cadmium
Copper
Chromium
Lead
Mercury
Nickel
Selenium
Silver
Acetone
Chloroform
Mixed
Municipal
28.1
9.1
217,000
23
0.9
8.9
353
823
0.06
564
0.07
0.1
Organics in Sludge
56.9
5.5
All metal concentrations reported in mg/l.
Plating
38.2
9.5
N/A
N/A
3.2
10.4
56.0
1048.0
N/A
23.4
20.1
20.1
N/A
N/A
Table 7.
Treatment Formulations for Mixed Municipal
and Plating Wastes
Gelled Solidification
14 pphw N" sodium silicate added to waste. Material gelled with 1.5-
2.0 pphw H2SO4.
Cement/Silicate Solidification
14 pphw N< sodium silicate and 10 pphw Portland cement Type I added
to waste; set time—1 week.
pphw—parts per hundred parts of waste
sodium silicate-cement-fly ash combinations, with strengths ex-
ceeding the required 1 ton/ft2 value. Costs for several of these
treatments were lower than the cement-lime-clay standard treat-
ment.
The final phase of these initial investigations concentrated on
solidifying a variety of organic solvents (Table 9) with a gelling pro-
cess to reduce ignitability and flash point and produce a material
that would be safe and easy to transport. Three distinct techniques
utilizing sodium silicates were developed (Table 10).
The physical properties and characteristics of solvents solidified
by silicate gellation are compared to standard absorbent treatment
and reported in Table 11. All solvents investigated were successfully
-------
242
RESEARCH AND DEVELOPMENT
Table 8.
Heavy Metal Leaching Results for
Mixed Municipal and Plating Wastes
•All concentrations in mg/l.
PLATING WASTE
HEAVY
METAL
Co
Cd
Zn
Sn
Cu
Ni
Cr
V
Pb
FRESH
WASTE
1.1
0.33
20.0
4.0
2.6
44.0
2.5
0.32
8.6
GELLED
WASTE
0.15
0.05
6.2
1.8
0.14
1.4
0.42
0.47
2.7
%
CHANGE
-86
-76
-69
-55
-95
-97
-83
+ 47
-69
HEAVY
METAL
Co
Cd
Zn
Sn
Cu
Ni
Cr
V
Pb
C£M£NT
FIXED
WASTE
0.47
0.12
4.1
4.0
0.48
1.9
1.4
1.5
12
GELLED
WASTE
0.15
0.05
6.2
1.8
0.14
1.4
0.42
0.47
2.7
%
CHANGE
- 68
- 58
+ 51
- 55
- 71
- 26
- 70
- 69
- 78
MIXED MUNICIPAL WASTE
HEAVY
METZVL
Co
Cd
Zn
Sn
Mi
Cr
V
Pb
FRESH
WASTE
0.19
0.20
30.0
0.78
4.7
0.13
0.07
4.5
GELLED
WASTE
0.04
0.06
8.3
0.27
1.7
0.15
0.04
2.5
%
CHANGE
-79
-70
-72
-65
-64
+15
-42
-44
HEAVY
METAL
CO
Cd
Zn
Sn
Ni
Cr
V
Pb
CEMENT
FIXED
WASTE
0.09
0.10
22.0
0.13
2.2
0.19
0.04
3.2
GELLED
WASTE
0.04
0.06
8.3
0.17
1.7
0.15
0.04
2.5
%
CHANGE
- 44.4
- 60.0
- 37.7
+130.0
- 77.3
- 78.9
0.0
- 15.5
Table 9.
Solvent/Solvent and Solvent/Silicate Compatibility
Water
£ = 78.5
Glycerol
£ = 42.5
Ethylene
Glycol
£ = 37.7
Methanol
£ = 32.6
Acetone
£ = 20.7
Pet. Ether
£ = 10.0
Toluene
£ = 2.4
M
C
G
M
C
G
M
C
G
M
I
P
M
I
P
N
I
S
N
1
S
M
C
G
M
C
G
M
I
P
M
I
P
N
I
S
N
I
S
M
C
G
M
I
P
M
I
P
N
I
S
N
1
S
M
I
P
M
I
P
M
I
P
M
I
P
M
I
P
M
I
P
M
1
P
M
I
S
M
I
S-
M
I
S
M—Solvents Mix
N—Solvents Don't Mix
C— SoKent Compatible wiih Silicale
1—Solvent Incompatible with Silicate
G—Gel Formation
P—Precipitation
S—Pha.ve Separaiion
Table 10.
Approaches to Solvent Gelling
Direct Gelling
For solvents compatible with water and silicates. Solvents characterized
by high dielectric constants and miscible with water. Gel forms di-
rectly in solvent by condensation of silica to form gel network.
Examples: Water, Ethylene Glycol, Formamide
Indirect Gelling
For solvents compatible with water but not with silicate solutions. Sol-
vents miscible with water and have medium range dielectric constants.
Silica in solution must be protected from dehydration by solvent.
Acidify silicate solution and initiate gelling before adding to solvent.
Examples: Acetone, Alcohols
Encapsulation
For solvents incompatible with water and silicate solutions. Not
miscible with water and low dielectric constant. Solvent must be en-
capsulated in a gelling silicate through mechanical procedures.
Examples: Toluene, Benzene, Pet. Ether
Table 11.
Comparison of Solvent Solidification Using
Shale Absorbent and Silica Gel
Gel
Amount solidification
material added (based on
weight of solvent). 30%
Volume increase
Ignitability
Would not ignite
or support
combustion.
Absorbent
110%
40%
Material ignited and
burned until solvent
exhausted.
-------
RESEARCH AND DEVELOPMENT
243
solidified and contained using one or more of these techniques and
in each case the ignitability and the ability to support combustion
was reduced.
In summary, these results show that silicate addition reduces
leaching characteristics of heavy metals from wastes vs untreated
wastes and the standard treatment used. Further benefits observed
were the improved mechanical properties of the cement based
method used. Containment properties of gelled silicates extend to
certain solvents in additionto inorganic wastes.
FUTURE WORK
The results discussed in the previous section show that silicates
can be used to solidify a variety of materials, and that silicate-
bonded wastes appear to have advantages in fixing heavy metal
ions, or in suppressing the teachability of those ions. In addition, a
wide range of organic solvents can be solidified by silicate based
methods and solidified organics have much lower flammability
characteristics than the liquid form.
The key objective of future work will be to optimize the use of
silicates in three areas; permeabilty, leachability and mechanical
strength.
•Permeability of silicate-bonded wastes and silicate grouts vs.
wastes bonded by other methods
•Resistance to leaching of silicate-bonded heavy metals vs hydrox-
ides, over a range of pH values
•Mechanical strengths of silicate-based consolidated wastes
REFERENCES
1. Vail, J.G., "Soluble Silicates" 1 and 2, ACS Monograph Series #116,
Reinhold Publishing Corp., New York, 1951.
2. Merrill, R.D. and Spencer, R.W., J. Phys. and Coll. Chem., 54, 1950,
806
3. Greenberg, S.A. and Sinclair, D., J. Phys. Chem., 59, 1959, 435.
4. Falcone, J.S., Jr., in J.S. Falcone, Jr., ed., "Soluble Silicates" ACS
Symposium Series #194, ACS Washington, D.C., 1982, 133.
5. Conners, J.R., Industrial Water Engineering, July/August 1977.
6. Conners, J.R., "Ultimate Disposal of Liquid Wastes by Chemical
Fixation", 29th Annual Purdue Industrial Waste Conference, Purdue
University, West Lafayette, Ind., May 7, 1974, 906.
7. Gowman, L:P., "Chemical Stability of Metal Silicates vs. Metal Hy-
droxides in Ground Water Conditions", Proc. National Conference
on Complete WateReuse, L.K. Cecil, ed., AlChe, N.Y. 1976, 783.
8. Panomareva, L.V., Dushina, A.P. and Aleskovskii, V.B..Z/J. Prikl.
Khim., 18, 1975, 2142.
9. Final Report on Inorganic Builders to the Great Lakes Science Ad-
visory Board of the International Joint Commission, July 1982, J.
Shapiro (chairman).
10. Pojasek, R.B., Chem. Engineering, 86, Aug. 13, 1979, 141.
11. Pojasek, R.B., Envir. Sci. & Tech., 13, 1979, 810.
12. Maugh, T.H. II, Science, 204, 1979, 819, 930, 1188, 1295.
13. Pojasek, R.B., ed., "Toxic and Hazardous Waste Disposal", Vols I
and II, Ann Arbor Science Publishers, Inc., Ann Arbor, MI, 1979.
14. Conners, J.R., USP #3,837,872, (9/24/74).
15. Conners, J.R., USP #3,841,102, (10/15/74).
16. Anon, Chemical Engineering, 78, Nov. 1, 1971, 29.
17. Anon, Products Finishing, Jan. 1972.
18. Anon, Chemical Week, 121, Jan. 26, 1972, 41.
19. Anon, Business Week, June 30, 1973, 32F.
20. Josephson, J., Env. Sci. & Tech., 9, 1975, 622.
21. Weismantel, G.E., Chem. Eng., 82, Oct. 13, 1975, 76.
22. Anon, Clippings from the Cleveland Press, the Plain Dealer, De-
troit News, and Lorain, Ohio Journal.
23. Anon, Iron Age, undated.
24. Conners, J.R., "A Critical Comparison: Ultimate Liquid Waste Dis-
posal Methods", Plant Engineering, Oct. 19, 1972.
25. Anon, Chemfix Data Package I.
26. Conners, J.R. and Gowman, L.P., "Chemical Fixation of Activated
Sludge and End Use Applications", in Proc. National Conference on
Complete WateReuse, L.K. Cecil, ed., AlChe, N.Y. 1976, 740.
27. Anon, Chemfix Data Sheet Series, Technical Data Information:
#101 Chemistry, 11/72; #102 Leaching, 11/72; #103 Current Indus-
trial Applications, 11/72; #105 Leaching Tests for Solids Developed
from Treatment of Industrial Wastes.
28. Anon, Ultimate Disposal of Liquid Wastes by Chemical Fixation—A
New Concept. Chemfix Information Package.
29. King, P.W.H. and King, A.S.H., Brit. Pat. appl. 2,062,604,
(5/28/81).
30. Phillips, J.W., "Applying Techniques for Solidification and Trans-
portation of Radioactive Wastes to Hazardous Wastes" presented at
the Proc. of National Conference on Management of Uncontrolled
Hazardous Waste Sites, Washington, D.C., Oct. 1981, 206.
31. Hayes, J.F., U.S. Patent #4,173,546 (11/6/79).
32. Societe Internationale de Publicite et D'Agences Commerciales, Brit.
Pat. 1,518,024(2/19/78).
33. Rousseaux, J.M. and Craig, A.B., "Stabilization of Heavy Metal
Wastes by the Soliroc Process", USEPA, Off. Res. Dev., [Rep.] EPA
1981, EPA-600/2-81-028.
34. Krofchak, D., Can Pat 1,024,277, (1/10/78).
35. Stevens, N.J., in R.B. Pojasek, ed., "Toxic and Hazardous Waste
Disposal", Vol I, Chapter 7, Ann Arbor Sci., Ann Arbor, Michigan,
1979, 119.
36. Chappell, C.E., U.S. Pat #4,116,705 (9/26/78).
37. Schofield, J.T., in R.B. Pojasek, ed., "Toxic and Hazardous Waste
Disposal", Vol I, Chapter 7, Ann Arbor Sci., Ann Arbor, Michigan,
1979, 297.
38. Anon, Chem Week, 128, Feb. 7, 1979, 41.
39. Heacock H.W. and J.W. Riches, "Waste Solidification Cement or
Urea Formaldehyde", ASME Winter Meeting, New York, NY, Nov.
1974.
-------
DEVELOPMENT AND DEMONSTRATION OF SYSTEMS TO
RETROFIT EXISTING LIQUID SURFACE IMPOUNDMENT
FACILITIES WITH SYNTHETIC MEMBRANE
JOHN W. COOPER
DAVID W.SCHULTZ
Department of Geosciences
Southwest Research Institute
San Antonio, Texas
INTRODUCTION
Surface impoundment facilities such as pits, ponds, and lagoons
are used extensively throughout the United States to store, treat
and dispose of hazardous wastes. These impoundments usually are
designed to contain fluids, utilizing native materials or liners and in
the past, many were constructed with little concern for preventing
seepage from the facility. In fact, many such facilities were de-
signed to seep water into the ground as a means of disposal. Un-
fortunately, many of these facilities are presently causing ground-
water contamination problems.
With the establishment of regulations passed under authority
of the Resource Conservation and Recovery Act (RCRA) legisla-
tion, retrofitting in-service disposal and treatment impounds may
be required to prevent continued seepage. Retrofitting with im-
permeable liners may accomplish this objective. Liners may also
be required to protect against wind and water erosion, another
objective of RCRA regulation requirements.
Retrofitting an in-service impoundment with a flexible mem-
brane liner generally implies the emplacement of liner material on
the bottom and sides of a facility to mitigate leakage without re-
moving any in situ fluid (hazardous material) in the impoundment.
Such action may be desirable as a temporary measure, while a
new impoundment is constructed, or as a permanent step to up-
grade an untrustworthy facility. As RCRA regulations are im-
plemented and groundwater studies confirm contamination from
seeping impoundments, information about retrofitting techniques
will be needed to assist owners and operators in upgrading ex-
isting facilities.
The main purpose of a membrane liner system in a fluid im-
poundment is to prevent loss of liquid to the groundwater, but
a liner can also prevent bank erosion due to wave and wind ac-
tion. Flexible liner membranes are manufactured using a variety of
polymeric materials such as polyethylene and polyvinyl chloride.
During construction of new impoundments, the liner materials are
installed over compacted subgrade. Sheets of membrane liners are
placed over the subgrade and connected together using a variety of
seaming methods. The most common methods are heat sealing and
solvent bonding. The sheet is attached at the top of the perimeter
using an anchoring system, usually an earthen anchor trench. The
finished liner serves as a relatively impermeable continuous bar-
rier covering the earthen impoundment.
The implementation and enforcement of RCRA regulations con-
cerning hazardous waste impoundment will create a need for own-
ers/operators to cease the contamination of groundwater due to
seepage. In many cases, it may be desirable or even necessary, to
make design modifications while the facility is in service, while in
others it may be necessary to temporarily reduce seepage while a
facility is closed or a new one is constructed. Retrofitting of mem-
brane liner systems offers a potential way for owners of such haz-
ardous waste fluid impoundments to reduce or stop seepage.
Unfortunately, retrofitting existing impoundments has not been
routinely performed. Many questions need to be answered before
this approach to sealing an existing impoundment can be imple-
mented. Information concerning the size of facilities which can be
retrofitted, equipment needed, the most efficient retrofitting sys-
tem, labor requirements and costs needs to be developed.
In this paper, the authors report on an ongoing project that
Southwest Research Institute is conducting for the USEPA, for
the development and demonstration of systems to retrofit exist-
ing liquid surface impoundment facilities with synthetic membrane
liners.
PROJECT OBJECTIVES
The primary objective of this project is to determine the feasi-
bility of retrofitting an existing surface impoundment with an effec-
tive membrane liner system without taking the facility out of
service. Anticipated project results are the following:
•Identification of one or more technically feasible retrofit liner
installation techniques
•Demonstration of the feasibility of each system on a pilot scale
•Preparation of a full scale demonstration plan if the systems prove
feasible at pilot scale
The results of the project will determine the feasibility of retro-
fitting existing surface impoundments with a membrane liner sys-
tem. Conceptual approaches to retrofitting different types of im-
poundments will be developed and evaluated. The most promising
approaches will be attempted at a pilot scale to verify or deny their
feasibility. From these activities, the important procedural and
physical parameters to consider when retrofitting will be estab-
lished. Limitations of each approach will be defined, e.g., which
approach is best suited for a particular impoundment geometry.
A plan will be developed for full scale implementation of those
retrofitting systems shown to be feasible at pilot scale. Potential
sites will be identified where the retrofit approaches could be im-
plemented.
The most significant benefit which could be expected as a result
of successful completion of this project will be the development
of a promising technique(s) which could be used to halt the release
of hazardous fluids into the environment. Preventing or stopping
the contamination of surface and/or groundwater may be impor-
tant to protect drinking water supplies. If a surface impound-
ment containing hazardous wastes is found to be seeping contam-
inated fluids, retrofitting the facility with an effective membrane
liner system should provide mitigation. This project should provide
information needed about retrofitting in order to ascertain the feas-
ibility of full scale utilization.
WORK ACCOMPLISHED
Load Analyses
A theoretical stress/strain analysis has been performed for the
various retrofit membrane materials. Mechanical properties and
coefficient-of-friction and failure data for different types of liner
244
-------
material were obtained and used during the analyses. A load analy-
sis program has been written to calculate the stresses on liner ma-
terials resulting from installation using the pull-thru concept for
retrofitting. The program is set up to run on a HP9325A desktop
computer. Required input data are:
•Length of the first sheet to be pulled into the pond
•The footage of liner covering the end slopes of the pond
•The footage of liner covering the side slopes of the pond
•Coefficients of friction values for soil and water
•The tensile strength of the liner material at elongation
•The thickness and density of the liner
•The length and width of the pond
This program will calculate these stresses for any impoundment
geometry and liner material. Calculations for four generic liner ma-
terials and three different sizes of impoundments are shown in
Table 1. The pond geometry varies as does the size of the panels
to be seamed together as the material is pulled into the water.
These calculations provide a very gross estimate of the forces to
be expected. Critical factors yet to be answered are: (1) can the
maximum force required to pull the liner be distributed evenly
along the edge of the liner; (2) what are the coefficient of fric-
tion values for each material and how can they be reduced; and
(3) what effect will weighting the draw bar to keep the liner on the
bottom have on the stress distribution to the liner? The values
shown in Table 1 are likely more accurate for the pull-over retro-
fit concept than the pull-through concept. Note that the calcu-
lated tensile values for all cases do not exceed the tensile at elonga-
tion values of the various materials.
Considerations in Retrofitting Liners
A study has been made of considerations in retrofitting liners.
Two general approaches to retrofitting ponds have been analyzed.
These are pulling the liner along the contour of the pond and pull-
RESEARCH AND DEVELOPMENT 245
ing the liner on top of the fluid, with subsequent sinking of the liner
to conform to the contour. Each approach has positive and nega-
tive features.
Pull-Through Concept
Assembly
The liner will be assembled by one of three methods:
•All of the membrane panels are laid out, seamed and then fan
folded.
•The panels are laid out, seamed and then fan folded.
•The panels are laid out and seamed one at a time.
Pulling
The leading edge of the assembled panel will be wrapped around
a heavy catenary cable or chain or clamped between sections of
heavy steel plates to distribute the pulling load along the edge of the
panel. A battery of cable pullers or winches would be mounted at
the opposite end of the impoundment to pull the panel into and
across the basin.
Bottom Contour and Side Slope Following
While new impoundments with membrane liners generally have
smooth flat bottoms to facilitate proper placement of the liner, ex-
isting unlined impoundments will most likely not have smooth flat
bottoms or straight side slopes since that would not have been a de-
sign requirement at the time they were built. Thus, even with the
use of rollers at the bottom of the side slopes and in the corners, it
will be difficult to prevent bridging by the liner as it is pulled across
depressions in the bottom or irregularities along the sides.
Large pockets of liquid will then be trapped beneath the liners.
The implications of this entrapped liquid are discussed in a later
section. However, a drag caused by the digging in of the leading
edge of the liner, or by the buildup of mud or sludge balls along
Table 1.
Calculated Tensile and Pulling Forces for Four Liner Materials*
Liner
Type
Thickness
(In)
Pond
Size
(ac)
Geometry
(ft)
Panel
Size
(ft)
Liner
Density
(gm/cm^)
Liner
Tensile
(pal)
Slide
Slope
(ft)
Liner
Weight
(Ibs)
Max
Tensile
(psl)
Max
Force
(Ibs)
CSPE
CSPE
CSPE
CSPE
CSPE
CSPE
HOPE
HOPE
HOPE
HOPE
HOPE
HOPE
PVC
PVC
PVC
PVC
PVC
PVC
CPE-S
CPE-S
CPE-S
CPE-S
CPE-S
CPE-S
.036
.036
.036
.036
.036
.036
.080
.080
.080
.080
.080
.080
.030
.030
.030
.030
.030
.030
.030
.030
.030
.030
.030
.030
1
1
5
5
10
10
1
1
5
5
10
10
1
1
5
5
10
10
1
1
5
5
10
10
200 x 200
100 x 400
447 x 447
300 x 667
1414 x 1414
500 x 4000
200 x 200
100 x 400
447 x 447
300 x 667
1414 x 1414
500 x 4000
200 x 200
100 x 400
447 x 447
^00 x 667
14f4 x 1414
500 x 4000
200 x 200
100 x 400
447 x 447
300 x 667
1414 x 1414
500 x 4000
200
200
200
200
200
200
33
33
33
33
33
33
200
200
200
200
200
200
200
200
200
200
200
200
1.3
1.3
1.3
1.3
1.3
1.3
.94
.94
.94
.94
.94
.94
1.3
1.3
1.3
1.3
1.3
1.3
1.2
1.2
1.2
1.2
1.2
1.2
1500
1500
1500
1500
1500
1500
2800
2800
2800
2800
2800
2800
2300
2300
2300
2300
2300
2300
1433
1433
1433
1433
1433
1433
20
'20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
9752
9752
48711
48782
487428
487574
15699
15699
78271
78384
783216
783453
8126
8126
40592
40652
406189
406312
7501
7501
37470
37525
374944
375058
102
203
132
153
145
294
37
76
40
58
46
157
102
155
132
153
145
294
94
143
122
141
134
272
8776
8776
25560
19816
88694
63544
7018
7262
17375
16703
62520
75324
7314
5578
21300
16513
73912
52954
6751
5149
19662
15243
68226
48880
Coefficient of friction of soil -.90; Water - .01 for all cases.
-------
246
RESEARCH AND DEVELOPMENT
this edge, can greatly increase the required pulling force or cause
the cable or load distributing panel to fail.
Mitigating actions to reduce bridging and entrapment of liquids
will be accomplished by the frequent use of rollers to keep the liner
on the bottom of depressions or irregularities and by the careful de-
sign of the leading edge of the load distributing panel to prevent
digging in. A slight rise in the weighted leading edge will cause the
liner to dig through the sludge but not into the harder underlying
soil. If the impoundment is filled with filamentous sludges (as
might be the case for some sewage wastes) it may not be possible
to place the liner by dragging because of balling of the sludge.
Tearing
Existing impoundments are likely to have protrusions that will
cause excessive stress concentrations in the liner and failure by tear-
ing. Irregularities in the bottom, rocks, pipes and other debris,
known or more likely unknown and unseen, can cause ripping.
Because the ripping is likely to occur beneath the surface of what
is perhaps a murky liquid the tear may not be noticed or detect-
able.
Mitigating actions may be to search the impoundment with a
drag wire survey to locate and/or remove the larger protrusions,
by reducing friction forces so that the localized tension forces at
the protrusion are less, and thus the near buoyant liner might ride
over it without tearing, or by using a tougher fiber reinforced liner
that is less likely to tear on small protrusions.
Float and Sink Concept (Pump-Over)
Assembly
The liner is assembled using one of the methods described for
dragging, however, the end that is pulled across the impoundment
is doubled back over a stiff hose, or sections of rods, for a suffic-
ient distance so that when the liner is deployed, the end will fall
several feet short of the "shoreline" when fully deployed. Clamps
around the liner and hose or rods are then installed to permit pull-
ing the liner without excessive strain.
Placement
The liner will be pulled across the impoundment in a fashion sim-
ilar to the pull-through concept, however, in this method of place-
ment, the leading edge of the liner and the pulling lines will be kept
as light as possible so that the liner would remain on or near the
surface during the deployment period.
Sinking of Liner
The deployed liner will be lightly held in position during the
sinking phase. If the liner has a density greater than the liquid it
will slowly sink. As it sinks, the liquid underneath the liner would
escape and flow over the top at the open water margin between the
folded end of the liner and the far shore line.
If the liner is made of buoyant material, e.g., HOPE, ballast
such as coarse sand would be used to sink the liner. The ballast
would be placed on the liner starting at the initial deployment end.
Only a small amount of ballast should be .required. Only 2.51b/
100 ft2 should be required to sink 80 mil HOPE. Similar to the
dense liner, the displaced water will move toward the gap at the far
end as the liner is slowly covered with ballast. When the liner has
completely sunk, the doubled back flap is opened up and pulled to-
ward the far end to close the gap. One end of the gap will be left
down until the very end to let any remaining liquid escape. For the
pump-over concept, fluid would be pumped onto the liner rather
than allowing it to flow over the leading edge.
Advantages
Placement of the panel will cause minimal stress to the liner since
only the dry land friction should be of any magnitude and, as dis-
cussed earlier, this can be mitigated. There should be negligible
tearing since the liner is not dragged across potential tearing pro-
trusions. Holing of the liner should be limited to localized punc-
tures by very sharp protrusion. Since the liner sinks under min-
imum tension, it should settle into most bottom depressions and
side irregularities.
Sinking and Liquid Recirculation
Even with the provision of a folded back flap at one end to per-
mit the liquid beneath the liner to recirculate to the top, proper
sinking of the liner may be difficult to achieve. The liner will prob-
ably not sink uniformly by itself but will have to be contrained
around the perimeter as it sinks. Otherwise, the liquid underneath
may attempt to recirculate to the surface at locations other than
planned with the liner settling to the bottom possibly out of
position. This could result in excessive folds in the liner or lack of
sufficient coverage around the side slopes.
Also, since the tension in the liner during placement will be less
than for dragging, extra care will have to be taken to reduce dam-
age or misplacement due to wind or wave action. Pumping of fluid
would allow placement of the water at the end of the facility oppo-
site from the leading edge. This may facilitate controlled sinking of
the liner.
Entrapped Liquids and Sludges
Whether by pulling or sinking, some liquid or sludge will be
trapped between the liner and the bottom of the impoundment.
The specifics of the type of waste and the underlying soil condi-
tions will determine what actions should be taken to accommo-
date these entrapments.
Non-Gas Generating Entrapment^
If the entrapped liquid or the underlying soil does not generate
gases, it may be acceptable to leave the liquid and not attempt to re-
move it. If the underlying soil is permeable enough, and the water
table low enough, the entrapped liquid will percolate into the soil.
If the underlying soil or entrapped sludge is not permeable, the
bubble of liquid will remain. There is then the danger of ripping or
perforation of the liner due to the repeated working of the fabric
over the bubble under wave induced or other hydraulic motions.
Methods to remove entrapped bubbles of liquid may include the
following:
•One Way Valves could be sidely distributed over the liner. In the
placement mode, they will assist in the rapid and uniform sink-
ing of the liner and would permit the entrapped liquids to escape.
Detrimental aspects of their use will include clogging by the
sludge, complexity of installation and susceptibility to damage,
and the likelihood of leakage, albeit a small amount, due to grit
lodging on the valve faces.
•The impoundment could be lined with an array of drain' pipes
with the inlets covered with porous sand, gravel and/or a geo-
technical filter membrane. Entrapped liquids would then be
pumped out. The placement of the array in an existing impound-
ment may be difficult and complex. There will also be suscep-
tibility to drain blockage by sludge.
•By the use of either ballast or weighted rollers, entrapped liquids
could be rolled out, Starting in the middle or deepest point of im-
poundment, ballast is placed on the liner to establish an outward
moving radial front of the displaced entrapped liquid. The en-
trapped liquid is then allowed to escape to the surface around the
perimeter. Rollers operating in a sequence of outward moving rad-
ial rolls can similarly work the entrapped liquids toward the edges
and then the surface.
Gas Generating Entrapments
Many waste liquids, sludges and underlying soils may generate
gases. Bubbles of gas under the liner can raise the liner to the sur-
face or at least separate it from the bottom and make it suscep-
tible to wear and failure from hydraulic motions or induced
stresses. In some cases, if the waste material and underlying soil
are permeable enough, the entrapped gas will escape by itself.
However, if the gas cannot escape, then mitigating actions will
have to be taken using the methods for removing entrapped liq-
-------
RESEARCH & DEVELOPMENT
247
uids or other techniques:
•If clogging of the values can be prevented and if leakage is ac-
ceptable, one way values can be effective in the removal of en-
trapped gas. The escaping gas may also unclog the valve.
•Similarly, if clogging can be prevented, drain fields can be used
to remove gases.
•Rolling out or ballasting will not be an acceptable method for gas
removal unless only a very small quantity of gas is generated for
a short period after the retrofit of the liner.
•Non-clogging geotechnical filter membranes that can block sludge
or sediment while passing liquids and gases can be installed under
the retrofitted liner. They will then conduct the entrapped liquids
and gases to the surface or to either one way valves or a drainage
field where the gas can escape. Uncertainties with using permeable
membranes include the long term blocking of the pores of the ma-
terial and compression of the pores due to the overlying load of
the waste material.
Sludge and Contaminated Soil Removal
The float and sink method will trap all, and the dragging method
much, of the solids in the sludge under the retrofitted liner. This
sludge and any contaminated soil under the impoundment can
leach undesirable trapped liquids or generate gas for many years.
In some cases, the leachate may be able to be pumped out from
peripheral wells. However, if this practice is determined to be unac-
ceptable it may be necessary to completely remove the sludge and
contaminated soil.
The size of many impoundments and other constraints will pre-
clude draining the impoundment and temporarily storing the waste
material in another impoundment. Thus, the solid material will
have to be retrieved in situ and prevented from settling before the
retrofitted liner can be placed. Two methods of dredging and liner
installation might be used to achieve this objective. Both methods
require that the waste material in the impoundment has the prop-
erty that suspended solids will settle out under gravitational forces
so that there is a clear liquid overlying a sludge blanket. Except
for some colloidal suspensions, most aqueous wastes have these
properties. Oily wastes may have oil or foam on the surface but
there is usually a clear liquid underneath.
Dragging—As the liner is pulled along the bottom and into
position, a dredge will remove the sludge and contaminated soil
immediately ahead of the liner. The dredged spoil is initially pump-
ed and discharged at a far corner of the impoundment. There the
suspended sediment settles to the bottom and the clear liquid recir-
culates back into the impoundment. It may be necessary to install a
floating silt curtain around the discharge to confine the dredged
material until it has settled out. When approximately half the liner
is dragged into position, the discharge will be moved so that the
spoil is deposited at a far corner of the lined part of the impound-
ment. Care will have to be taken to widely spread the spoil over
the liner since it will still be moving into position and thus, it will
be desirable not to create abnormal stresses in the liner by weighting
it down with sediment. If there is a very large mass of heavy set-
tled sludge, this method of installation may not be usable.
Placement—In this method, part of the impoundment is sec-
tioned off by a floating curtain similar to those used in controlling
oil spills.- The bottom edge of the curtain is weighted and lies on the
bottom of the impoundment as a seal. The sediment from the sec-
tions is dredged and deposited far behind the curtain. The clear
supernatent is permitted to recirculate over the top of the curtain.
After dredging, the liner is placed using the methods discussed
earlier. The curtain is then moved to enclose the next adjacent
section. After completion of dredging of this section, the liner is
placed with an overlap over the previously placed liner. With suffic-
ient overlap, acceptable small leakage rates can be achieved. Sim-
ilarly to the procedure for the pull-through method of installation,
when the impoundment is half lined the spoil discharge point is
moved to a remote position over the retrofitted liner. Since in the
placement method of installation the liner is not moved over the
bottom, the mass of the settled sediment at a particular place is
not as critical using this method. While small road transportable
cutter head or dust pan dredges can be used to move the sludge,
consideration should be given to pneumatic dredges that have been
developed for dredging contaminated sediments. This type of
dredge entrains considerably less liquid and the discharged spoil
tends to settle out with a considerably smaller surface plume.
Folds—The topography of impoundments in which retrofitted
liners will be placed will most likely to quite irregular. In install-
ing the liner, it will be very difficult, if not impossible, to prevent
folding of the liner. Where the liner is above the liquid, as around
the sides of the impoundments, folds can be cut out and the liner
patched. However, when folds occur in the liner beneath the sur-
face of the liquid waste, it may not be possible to remove the folds
using these methods because (1) the fold may not be detectable
if covered by sludge or a murky waste and (2) it may not be pos-
sible to patch the liner underwater. If the fold is left unattended,
the liquid confined in the fold may drain out by itself or may be
withdrawn using the methods for removing liquids and gases dis-
cussed earlier. However, the base of the fold may seal the fold thus
trapping the liquid in the fold. Similar to the entrapped bubble,
the material in the fold will work under various hydraulic actions
which may utlimately cause the liner to wear or rip. If the fold ma-
terial just wears enough so the entrapped liquid can escape, the fold
may collapse and two overlapping halves may form a seal. If the
fold material tears, the two halves of the fold may move apart and
an appreciable leak path may be created. Thus, a liner material that
wears and leaks may be preferable to one that does not wear but
tears. Folds in the subsurface part of the retrofitted liner may have
to be accepted. Perhaps in areas where folds are likely to be en-
countered, a second layer will be installed to reduce leakage
through failed folds.
Double Perforated Liner With Permeable
Geotechnical Filter Membrane
For the many reasons discussed in this presentation, the prob-
ability of having leakage after the installation of a retrofitted liner
is quite high. Proper design, selection of materials and placement
can result in an installation with greatly reduced leakage but there is
still likely to be some. If this less than perfect performance is
acceptable then the utilization of two layers of liners that are inten-
tionally perforated may lead to a superior installation system. As-
pects of a two liner systems are:
Installation—If the perforated liner is installed using the place-
ment method it will sink more rapidly than a non-perforated liner
and will require less control around the perimeter. Any residual
trapped liquid will be displaced as the liner sinks to the bottom.
Liquid trapped in folds will also readily leak out, permitting the
fold to collapse. In the installation of buoyant liners, porous ballast
such as sand or gravel would not block the perforations. If the
second layer of the perforated liner is laid at right angles to the
first likelihood of folds lying on top of each other and of being par-
allel is reduced.
Perforations—Perforations in the liner would not be uniformal-
ly placed but would be located in a pseudo-random pattern. That
is, there might be one hole randomly placed in each square foot.
With this random placement there would not be a row of holes in
one liner aligning over a row of holes in the other liner.
Leakage—If the lower layer is lying firmly on the bottom and the
facing surfaces of the two liners are clean, then the double perfor-
ations would make an effective one-way valve or hydraulic diode.
Either due to negative buoyance or added ballast the two layers
will be pressed together blocking any leakage. If there is any
trapped liquid or gas at a higher pressure than the overlying liquid,
the upper layer will lift, permitting the gas or liquid to escape
through the perforations in the upper layer. Once the fluid has es-
caped, the pressure is reduced and the upper layer will collapse re-
sealing the bottom layer. Since, with many randomly placed holes,
some holes in one layer will align up with holes in the other layer
there will be some leakage. However, the area of the perforations
will be a very small part of the total area so leakage would be very
-------
248
RESEARCH AND DEVELOPMENT
small for a double layer system. A 0.375 in. hole in every square
foot of each liner would permit the upward flow of all trapped
liquids, yet would have a porosity of less than 10~3. A double layer
would then have a porosity of less than 10"6. When combined with
existing porosity of the impoundment bottom there should be very
little leakage.
Clogging—The efficiency of a perforated double layer retrofltter
liner system is dependent upon keeping the perforations open and
the facing surfaces of the liner clean. Blockage of the perforations
will not permit the trapped fluids to escape and grit caught between
the two layers might create a leakage path. If a liquid and gas
permeable geotechnical filter membrane is installed under the per-
forated liners, sludge and grit underneath would not move up and
block the perforations or contaminate the facing surfaces of the lin-
ers. A very effective retrofitted liner system would then be
achieved.
Gas—Gas trapped beneath the liner system or subsequently gen-
erated in the underlying sludge or soil will be able to escape through
the filter membrane to a perforation and then to the surface. If
there is a heavy or impermeable waste fill in the impoundment,
the gas will move laterally until it can escape through an unblocked
perforation or until it reaches the surface around the perimeter of
the impoundment.
Laboratory Scale Studies
A 10 by 20 ft scale model impoundment has been constructed
and is being used to make qualitative studies or retrofitting liners.
The model was constructed outside on the ground, lined with 10 mil
black polyethylene sheeting for water retention and then filled with
an inch or so of sand to simulate an unlined impoundment. T,he
sand was also banked up the sides and stabilized with cement.
At the time of writing only one series of experiments has been
undertaken: the floating and sinking by pump-over of a buoyant
membrane liner. A 4 mil polyethylene liner was pulled over and
floated on the impoundment, overlapping both the sides and ends.
Water was pumped from one end and released at the opposite end
at the same time as sand ballast was being put on the liner to sink
it. The following observations were then made:
•The liner would not sink to the bottom even with the ballast and
water added. Archimedean buoyance forces and tension in the
liner from friction forces around the perimeter kept the liner
afloat.
•The liner was moved back several inches from the pumping end of
the impoundment and a fold worked back toward the ballasted
end. When the fold was adjacent to the ballast the liner would
then sink to the bottom.
•With no longitudinal tension the ballasted section of the liner will
fold under the remaining buoyant liner. Light tension will keep
the size of this fold manageable.
•An advantage of floating and sinking is that the liner is not
dragged over the bottom. However, the liner will move a little
along the sides of the impoundment as it is sunk: Provision will
have to be made to accommodate this movement without over-
stressing the liner.
Pilot Scale Demonstration
A pilot scale demonstration facility has been designed to test the
retrofit system on a larger scale. The facility is presently being con-
structed at Southwest Research Institute. It will allow field testing
of retrofit concepts under controlled conditions and in the absence
of hazardous fluids. This is necessary in the early stages of system
development to avoid risk of injury to field crews.
FUTURE WORK
Following further laboratory scale studies retrofitting concepts
will be selected for demonstration in the pilot scale facility. The
pilot scale demonstration will be conducted and then evaluated. A
retrofit system field test plan for an actual hazardous waste site
will then be developed.
ACKNOWLEDGEMENT
The work reported on in this presentation has been accomplished
by Southwest Research Institute under Grant No. CR 809719 from
the Municipal Environmental Research Laboratory of the USEPA.
-------
A BLOCK DISPLACEMENT TECHNIQUE TO ISOLATE
UNCONTROLLED HAZARDOUS WASTE SITES
THOMAS P. BRUNSING, Ph.D.
Foster-Miller, Inc.
Waltham, Massachusetts
WALTER E. GRUBE, JR., Ph.D.
U.S. Environmental Protection Agency
Solid and Hazardous Waste Research Division
Cincinnati, Ohio
INTRODUCTION
Foster-Miller, Inc. has developed a technique for complete in
situ isolation of uncontrolled hazardous waste sites. This technique
is intended to emplace a seal at low cost around the sides as well as
underneath contaminated ground. Under contract to the USEPA,
demonstration of the Block Displacement Method (BDM), was
conducted over a four month period adjacent to the Whitehouse
Oil Pits site in Whitehouse, Florida.
The purpose of the demonstration project was to add this techni-
que to the list of available construction technologies applicable to
chemical waste remedial action by:
•Demonstrating the BDM in the geologic conditions of an exist-
ing chemical waste site
•Developing BDM specifications for "A User's Guide for Evaluat-
ing Remedial Action Technologies'"
•Evaluating the applicability of the BDM to existing site geologic/
hydrologic conditions
•Establishing BDM implementation procedures for remedial ac-
tion designers and contractors
DESCRIPTION OF THE BDM
The BDM is a patented2 technique for vertically lifting a large
mass of earth. The technique produces a fixed underground
physical barrier placed around and beneath an earth mass. The
barrier is formed by pumping slurry composed of local soil, ben-
tonite and water, into a series of notched injection holes. The
resulting barrier completely encapsulates the earth mass or block.
A typical BDM barrier is shown in Fig. 1.
The BDM is of particular value in strata where unweathered
bedrock or other impermeable continuum is not sufficiently near
the surface for a perimeter barrier alone to act as an isolator. The
barrier material should be compatible with soil, groundwater, and
leachate conditions.
A perimeter separation is first constructed and then surcharged
to ensure a favorable horizontal stress field in the formation (Fig.
2). Surcharge is additional pressure transmitted to the fluid slurry
in the perimeter separation by raising the slurry fluid level above
ground level.
Construction of the perimeter separation can proceed using one
of several techniques including thin slurry wall, vibrating beam or
drill notch and blast techniques. The thin perimeter separation
must be constructed on a slight angle, , off vertical tapering in-
ward toward the block center. Upward displacement, d, of the
block resulting from injection along the bottom barrier will then
increase the perimeter separation, W0, to the desired barrier
thickness, w, according to:
W = d sin 0 + W0
(D
A bottom barrier is formed when lenticular separations extending
from horizontal notches at the base of injection holes coalesce into
a larger separation beneath the block. Continued pumping of
slurry under low pressure produces a large uplift force against the
bottom of the block and results in vertical displacement propor-
tional to the volume of slurry pumped.
Construction of the bottom barrier proceeds in four phases: (1)
formation of notches at the base of the injection holes, (2) initial
bottom separation at the notched holes, (3) propagation of the
GROUNDWATER
LEVEL_'
J.
4 ' < PERIMETER
-i ) BARRIER
POSITIVE SEAL THROUGH
INJECTED BENTONITE
MIXTURE
BOTTOM BARRIER
Figure 1.
BDM Barrier in Place
PERIMETER
SURCHARGE
(WHEN
REQUIRED)
PERIMETER
SEPARATION
ill!
INJECTION
^ HOLES X
UPLIFT
/ PRESSURE \
MM
FRACTURED BEDROCK
Figure 2.
Creating Separation to Induce Displacement
249
-------
250
RESEARCH AND DEVELOPMENT
local separations at each injection point coalescing into a single
larger bottom separation, and (4) generation of a complete bottom
barrier by controlled vertical displacement of the earth mass using
low pressure slurry injection into the horizontal separation. Each
of these phases is carried out through controlled monitoring of the
slurry pressure, slurry flow rate, total volume injected, and slurry
composition.
The notching operation requires a high pressure rotating jet at
the base of the injection hole. The jetting slurry must be composed
in a manner which optimizes notch erosion, removes cuttings, and
minimizes leak off into the soil. The initiation of bottom separation
(3) requires a slurry pressure, P0, at the separation defined by:
P0 = Prgh + AP (2)
where
Pr is the average earth mass density
g is the gravitational constant
AP is the pressure in excess of the overburden
h is the depth of the bottom separation
AP increases with the increasing slurry viscosity and decreasing
notch radius and in general depends on soil characteristics and the
speed of operation. The bottom separation is initiated when slurry
pressure exceeds P0 inducing flow.
Separation coalescence is brought about by adding slurry volume
and by gradually increasing the viscosity of the slurry. Slurry
pressure required to propagate the horizontal separation will
decrease during this phase due to the increased area over which it is
acting. Increasing the viscosity of the slurry serves to limit flow in
preferential directions.
Vertical displacement utilizes the maximum capacity of the pum-
ping equipment, along with a high solids slurry that will form the
final barrier. The pressure required to continuously increase the
barrier thickness by lifting the buoyed block eventually decreases.
The slurry pressure, in excess of the overburden, AP, measured at
the bottom of the injection hole approaches that required to
balance the resistance of the perimeter surcharge (q) and to over-
come fluid drag in the perimeter (AP2) and in the bottom separa-
tion (AP,). The final pressure relationships are:
AP = AP, + AP2 + q (3)
HOLE Tl HOLE T2 HOLE T3
„ _ STANDARD PENETRATION
u RESISTANCE (BUMS)
GEOLOGIC KEY
SAND
HARDPAN
INTERBEDDED
HARDPAN AND
SILTY SAND
Figure 3.
BDM Demonstration Geology
FRACTURE
6 in.
DRILL HOLE
NOTCH
32 NOTCHED LATERAL
SEPARATION HOLES
+q = (pr-pm)gh
(4)
*
*
-------
RESEARCH AND DEVELOPMENT
251
3). A circular block 23 ft deep and 60 ft in diameter was to be
displaced in excess of 1 ft vertically.
Two types of holes, 32 uncased perimeter holes and 7 cases injec-
tion holes were drilled in preparation for the block displacement
process. The perimeter separation was to be constructed by line
drilling 6 in. diameter perimeter holes with a taper angle, , of 14°.
These holes were to be notched and blasted with line charged ex-
plosives to induce fracturing between the holes.
Following trial blasts on three test holes spaced 3 ft and 6 ft apart
respectively, a final perimeter pattern of 6 in. diameter holes on 6
ft spacing was drilled (Fig. 4). These holes were each notched from
top to bottom with a IS in. span notching tool (Fig. 5), following
tests with three earlier notching tool designs.
Figure 5.
Perimeter Notching Tool
Seven injection holes were drilled and cased with 25 ft long 6 in.
diameter PVC tubing which was cemented in place at the base. The
casing extended to a depth of 23 ft leaving 2 ft of casing exposed
above ground. Horizontal notches approximately 2 ft in diameter
were cut on all seven injection holes using a slurry jet notching tool
(Fig. 6).
In parallel with bottom notching, an 18 in. diameter, 5 ft high
concrete forming tube was placed over each drilled and notched
perimeter hole and filled with high density slurry (Fig. 7). Upon
culmination of injection hole bottom notching all 32 perimeter
holes were line loaded with prima cord explosive (50 grains/ft) and
blasted simultaneously. Connecting fractures were observed at the
surface between adjacent perimeter holes.
Figure 7.
BDM Bottom Notching Operation with 32 Perimeter
Surcharge Tubes in Place
Figure 6.
Bottom Notching Tool
Figure 8.
Upward Displacement of Block
BDM system analysis indicated horizontal separation should in-
itiate from the bottom notches at an injection hole pressure of ap-
proximately 20 psi. Initial attempts at injection required pressures
in excess of 40 psi to induce slurry flow indicating that renotching
of each injection hole followed by immediate slurry injection was
required.
Bottom notched separation propagation was finally realized by
increasing notch diameter to approximately 4 ft using a slurry jet in
air medium to cut the formation. Each notching operation was
followed by immediate slurry injection to propagate a local separa-
tion.
Renotching and local propagation was carried out in all six in-
termediate injection holes. This procedure was then applied to the
central injection hole followed by steady state slurry pumping.
-------
252
RESEARCH AND DEVELOPMENT
Separation coalescence between injection holes was observed
after approximately 500 gal of slurry were pumped into the central
injection hole. Pressure response was documented at intermediate
injection hole well heads followed by observance of slurry flow up
the perimeter holes at multiple points around the block. Perimeter
flow was characterized by small flows from approximately 10 of the
concrete forming tubes. During the course of block displacement
several of these tubes were replaced with 10 to 20 ft long plugs to
impede excessive flows at the perimeter.
Following separation coalescence, block displacement was
observed over most of the block surface. Block displacement then
proceeded over a two week period pumping approximately 2 ydVhr
alternately pumping to each injection hole. The resulting upward
displacement of the block is shown in Fig. 8.
Demonstration Results
Displacement of the block was monitored by recording the
change in elevation of 16 fixed rods (Fig. 9). The numbers in paren-
60 ft
r—A
^ D _^f* *
<~~Z
(2.9)
Jar
^
^7p
,(4.9) A (Ml
2»(12.2)
5.1)
(2.9)
17]
(3.8)
J.9)
7.8)
.(13V
(3.1)
25p
(6.4)
15p
(0)
(4.4)
22p
(3.3) *
V
6 ft
KEY (DIMENSIONS IN INCHES)
lt • NEAREST HOLE #
(6-1L.ACCUMULATED LIFT
A-— LOCATION BOTTOM
*-* SLURRY SAMPLES
[11.5] THICKNESS OF
SLURRY
CORRECTED FOR
ROD RELOCATION
Figure 9.
Recorded Block Lift and Verified Seal Thickness
theses in the diagram are the total recorded lift in inches at the cor-
responding positions on the block. Elevations were read through a
surveyor's level located 40 ft outside the block perimeter. Addi-
tional survey points beyond the block boundary were monitored
during a portion of the lift operation to verify that no lift was
occurring outside of the perimeter. Survey points 0 through 6
correspond to rods located adjacent to (approximately 3 ft from)
injection holes. The remaining 9 survey points 3p through 30p
correspond to rods located just inside the perimeter. These survey
INJECTION HOLES
-BOTTOM. SLURRY BABBIER
SECTION *-*
5 In.
Figure 10.
Final Block Displacement
point numbers correspond to the closest perimeter drilled hole
numbered progressively from 1 to 32 in a counterclockwise direc-
tion.
Approximately 10,000 gal of bentonite slurry were injected dur-
ing the block lift operation. Data from surface topographic surveys
combined with accumulated daily lift records was combined to give
two profiles of the final block position (Fig. 10). In total the block
was displaced upward 12 in. at its highest point and tilted approx-
imately 1 ° from horizontal. A crescent shaped portion of the block
outside of injection hole #3 was sheared free of the lifting block and
did not move appreciably. The entire remaining portion of the
perimeter lagged behind the inner portion of the block lifting from
3 to 6 in.
The crescent shaped shear zone, tilting of the block and
perimeter displacement lag were all attributed to an incomplete
fracturing and freeing of the block around the perimeter. Follow-
ing completion of the block lifting operation, 4 to 6 ft deep trenches
were excavated through the perimeter and shear zone to try to
determine causes of resistance to perimeter displacement.
Tap roots as large as 12 in. in diameter and typically on 6 ft
centers were found extending from approximately 5 ft below sur-
face to the upper boundary of the hardpan layer. Tap roots in-
tersecting the perimeter separation may have acted as dowels to
restrict block movement at the perimeter. In addition, perimeter
fractures filled with gelled slurry extended down only a few feet
before becoming unmapabale hairline fractures.
Thin walled tube soil samples were taken 4 weeks after discon-
tinuation of slurry pumping to determine the integrity of the bot-
tom seal. Obtaining undisturbed soil samples of the soft slurry
material bounded by hardpan on top and unconsolidated sand
beneath was a difficult task. In eight attempts three acceptable
samples were obtained at the locations shown in Fig. 9. Sample #8
(Fig. 11), indicated a well defined boundary of separation, be-
tween injected clay slurry and the overlapping native sand. Traces
of slurry migrated into fractures in the underlaying sand.
Final evaluation of data obtained from the field operation will be
conducted following additional core sampling and geophysical
surveying to assess the complete integrity of the bottom seal.
-------
RESEARCH AND DEVELOPMENT
253
BENTONITE
•SLURRY
BARRIER
BENTONITE
•SLURRY
BARRIER
Figure 11.
Core Sample #8 Showing Slurry Barrier Placed in
Native Sand at 23 ft Depth
CONCLUSIONS
Final conclusions and recommendations are pending completion
of data evaluation and verification testing still to be conducted.
Preliminary results indicate the following:
•The bottom seal was demonstrated and is apparently continuous
under the lifted portion of the block.
•Creating the perimeter separation using the drill, notch, and
blast technique was unsuccessful possibly owing to tap roots, in-
sufficient explosive charge size, or excessive local plastic de-
formation of the soil.
•BDM is a viable method for isolating contaminated ground in
unconsolidated stratified soils when used with a proven perimeter
barrier construction methodology. Additional information re-
garding BDM applicability will be contained in the project3 final
report to be presented to USEPA by the end of 1982.
•Sufficient data were obtained to write specifications for use of
BDM for bottom barrier construction. These specifications have
been included in the "User's Guidelines for Remedial Action
Technologies".1 The guidelines include slurry designs, general
engineering information and suitability of bottom barrier use
with applicable complementary perimeter barrier construction.
•BDM implementation procedures should include feasibility analy-
sis, a thorough geologic investigation of the site, and a pilot
operation to verify site suitability prior to proceeding with a full
scale operation. Details on implementation procedures will be
contained in the project final report.3
ACKNOWLEDGMENTS
Acknowledgment is given to Wendy J. Davis-Hoover, the
USEPA technical project officer during site selection and prepara-
tion, for coordinating program operation with Federal, State and
local government agencies.
REFERENCES
1. "User Guide for Evaluating Remedial Action Technologies", USEPA,
To be published.
2. Cleary, J.M., "A Method for Displacing Large Blocks of Earth",
U.S. Patent No. 4,230,368, February 1979.
3. Foster-Miller, Inc., Block Displacement Technique, Final Report, to
be issued to the Solid and Hazardous Waste Research Division,
MERL, Office of Research and Development, USEPA, Cincinnati,
Ohio under JRB Contract 60-03-3113, Task 37-1.
-------
CALLAHAN UNCONTROLLED HAZARDOUS WASTE SITE
DURING EXTREME COLD WEATHER CONDITIONS
WILLIAM E. RLTTHALER
Ecology & Environment, Inc.
Kansas City, Kansas
INTRODUCTION
The Ellisville sites are located in St. Louis County, Missouri, ap-
proximately 0.25 mi west-northwest of Ellisville (Fig. 1). To date,
investigative efforts have disclosed three locations in this vicinity
where waste materials have been deposited: Bliss, Callahan and
Rosalie properties. Investigations in the Ellisville vicinity are a
result of suspected past activities of a waste oil hauler and disposer
of industrial wastes, who is located approximately 0.25 mi east of
the Callahan site.
In this paper, the author will discuss only activities pertaining to
the Callahan site and subsequent emergency removal activities a»
accomplished by the Missouri Department of Natural Resources
(MDNR), the Region VII USEPA, Ecology and Environment,
Inc., (E&E), and Emergency Environmental Services, (EES).
BACKGROUND
Discovery
The Callahan site was discovered, along with other area sites,
through investigative efforts and information supplied to the
Missouri Department of Natural Resources. The actual site loca-
tion is at 162 Strecker Road, St. Louis County, Missouri, on a
12-acre parcel of land owned by Jean Ellen Callahan, the wife of
the alleged disposer.
Initial Investigation
A preliminary investigation was conducted of the Callahan prop-
erty near Ellisville, Missouri on Sept. 16 and 17, 1980. The purpose
ST. LOUIS
AND VICINITY
Figure 1.
Location of the Callahan Waste Site
ARROW POINTS TO LOCATION OF SITE
254
-------
CASE HISTORIES
255
of the investigation was to determine if the property had been used
for the disposal of hazardous waste as alleged by at least two infor-
mants.
The investigation was conducted on the surface of the property
only; no excavation was done. During the course of the investiga-
tion, 38, 55-gal drums were found either protruding from a filled
ravine or below the filled area. A composite sample was prepared
from the contents of three of the drums. Analysis of that composite
sample revealed five chemicals listed in "Missouri's Hazardous
Waste Law, Rules and Regulations" as well as seven other organic
chemicals not ubiquitous to the environment.
Positive responses from a metal detector were recorded over the
entire area of fill. Probing the soil with a metal rod indicated buried
metallic objects.
Although it was determined that there were drums of hazardous
waste partially buried on the Callahan property, it had not been
determined how extensive the hazard was, nor had it been deter-
mined to what extent the existing hazardous waste had impacted
the groundwater in this area.
Geology
The geologic setting of the Callahan property is the Caulks Creek
Watershed. Drummed wastes were piled in a steep gully in the head
waters of Caulks Creek, a tributary of the Missouri River, and were
thinly covered with dirt. The soil in the area is a clayey silt
(modified loess) overlying the Burlington limestone. The Burl-
ington limestone is heavily fractured, and is used extensively for
domestic water supplies in the area. Approximately 30 households
in the area draw drinking water in the Caulks Creek Watershed.
Contamination at the Site
The Callahan site was used for the disposal of containerized li-
quid and solid wastes. Based on limited analysis prior to the initial
removal actions taken by the Missouri Department of Natural
Resources, the following contaminants were determined to be
leaching into the water of Caulks Creek: nitrobenzene, napthalene,
anthracene, BHC, aldrin, DDE, phenols and various phthalates.
State Activities on Site
Based on Mr. Callahan's testimony that approximately 80-100
drums of material had been disposed of at the site and preliminary
sample analysis data gathered, the Missouri Department of Natural
Resources initiated removal actions on Dec. 14, 1981. All removal
actions accomplished at the Callahan site were funded under the
Comprehensive Environmental Response Compensation and
Liability Act, more commonly referred to as the "Superfund."
Mr. Keith Schardien was appointed as the State On-Scene Co-
ordinator (OSC), and Environmental Emergency Services (EES) of
Chesterfield, Missouri was chosen as the cleanup contractor to per-
form the removal operations of the buried drums.
On the site itself, the restricted zones and the hot line or red zone
areas were posted utilizing red banner tapes for the hot line area
and yellow banner tapes for the restricted access areas. The re-
stricted zone, the periphery of which was marked by the yellow
banner, was considered to be a buffer zone surrounding the red or
hot zone. In this area personnel were required to be wearing, at a
minimum, an ultra twin respirator with combination cartridges,
splash or acid suit, rubber safety boots and rubber gloves. The
joints of the boots and gloves had to be taped with duct tape in
order to prevent contamination by a liquid running into them.
The State of Missouri's scope of work called for a test trench to
be dug diagonally across the site from northeast to southwest and
another smaller trench at the south face of the fill area where visual
evidence on the surface indicated the greatest concentrations of
buried drums. As the drums were excavated, the following pro-
cedures would be followed to assure that proper custody and
documentation would be maintained:
•A control number was assigned to and permanently marked on
each drum as it was excavated. Any drum found to be leaking
would be placed into an overpack drum.
•The description of each drum type, condition and any visible
markings on the drums would be recorded on an individual drum
sheet. Photographs would be taken of each drum with visible
markings.
•Two full depth representative samples would be obtained by con-
tractor personnel under state supervision of each drum. Full
chain-of-custody was to be observed on all field data and sample
collection using the bound volume National Enforcement In-
vestigations Center (NEIC) system. All on-site sample collection
and analysis would be accomplished under the direction and sup-
ervision of a professional chemist.
•On-site analysis would consist of the following checks or tests:
(1) radiation checks with a general survey meter and (2) head-
space readings at the bung hole of the drum with an organic vapor
analyzer.
By Dec. 17, 1981, 187 drums had been excavated from the site
and were stored on the surface.
It was obvious that completion of the removal operation would
not be possible because funds originally allocated were exhausted
by the end of the first week's operation. In addition to the lack of
funds several hundred drums were visible in the area excavated. At
this point a formal request was made by the State of Missouri to the
USEPA for assistance.
Mr. William J. Keffer, Chief of the Region VII USEPA
Surveillance and Analysis/Environmental Services Division, was
appointed federal On-Scene-Coordinator (OSC) and requested to
evaluate the site.
On Dec. 18, 1981, Mr. Keffer and the writer evaluated the site.
They were joined by Dr. Joseph P. Lafornara and Mr. Rodney
Turpin of the Environmental Response Team (ERT).
On-site inspections (between Dec. 18, 1981 and Dec. 23, 1981)
conducted by USEPA, Region VII OSC, TAT, ERT technical staff
and state OSC documented the following imminent hazards to
public health and the environment:
•Leachate discharges containing several hundred mg/1 were en-
tering the Caulks Creek Watershed which enters the Missouri
River 2 mi upstream of the St. Louis County, Hazard Bend water
intake.
•The fractured Burlington limestone underlying the area is used
extensively for domestic water supplies in the area. Approxi-
mately 30 households in the area draw drinking water from the
groundwater in the Caulks Creek Watershed.
•Volatile organic material were being released to the air at the
site. One thousand people live within a one mile radius of the
site; the closest residences are within 100 yards and clusters of
single family dwellings within 0.25 of the site.
•Drums of toxic organic materials were exposed at the face of
the fill and others were stacked on the surface.
Based on this information the OSC recommended to the Region
VII Superfund coordinator and Henry D. Van Cleave, Acting
Director of the Emergency Response Division, Washington, D.C.,
that immediate removal funds be authorized with the proposed'ac-
tions to:
•Continue removal operations with EES
•Provide a guard on-site to prevent casual trespass
•Post the site with warning signs
•Have all on-site media contacts and releases be managed by
MDNR—State OSC as part of the community relations package
for the remedial actions for all Ellisville sites
•Drain the small farm pond immediately upgradient of the dis-
posal area because it was significantly adding to the leachate and
groundwater contamination problems. The pond was to be
drained and sealed with Bentonite Slurry.
•Construct surface runoff diversion ditches around the periphery
of the disposal area and a small temporary lined containment
pond to intercept leachate
•Using a backhoe supported by manual labor, remove and recon-
tainerize when necessary, sample, sort and secure on-site all
-------
256
CASE HISTORIES
drummed material in an aisle arrangement, segregating potentially
incompatible materials. On-site sorting to be based on laboratory
tests for visual characteristics, flashpoint, photoionizer readings
and simple incompatibility testing.
•Two full depth representative sample sets are to be collected by
contractor personnel under state supervision using the drum num-
bering system presently used and maintaining full chain-of-
custody. One sample set is to be delivered to USEPA with full
documentation of all field tests by a professional chemist.
•Composite soil samples are to be collected at the direction of
the OSC and analyzed to determine disposal needs. If disposal of
contaminated soil is necessary based on evaluation of contam-
inants MDNR will arrange for bulk disposal at Bob's Home
Service in Wright City, the state's only approved hazardous
waste site.
•Drums to be sorted and secured on the Callahan property and
placed on a 6 in. gravel pad enclosed by a 6 ft chain link fence
pending full identification and arrangement for disposal under
the remedial portion of the effort.
FEDERAL REMOVAL ACTIONS
Legal
On Dec. 30, 1981, USEPA Region VII received approval for up
to $210,000 to accomplish the immediate removal operations at the
Callahan site. Contacts with potentially responsible parties were
made by Region VII Office of Regional Counsel (CNSL) and verbal
notice of 48 hr given as required. None of the potentially responsi-
ble parties responded positively and the landowner, refused to give
USEPA clear access to take the necessary removal actions. On Dec.
31, 1981, a search warrant was obtained by the OSC in the US
District Court Eastern District of Missouri (81MISC202). All ac-
tions on-site during the removal were carried out under the authori-
ty of the search warrant.
On Jan. 4, 1982, Mr. Keffer signed the contract with En-
vironmental Emergency Services and discussions were held at this
time to include site safety procedures and cleanup priorities.
The owner of the property was also served with the search war-
rant and the purpose of the on-site cleanup was explained.
CLEANUP
Removal Operations
The first week on-site was dedicated almost exclusively to site
preparation and the acquisition of the necessary equipment and
supplies to carry out the balance of the emergency removal action.
A crew of approximately 20 persons was assembled; EES, the
cleanup contractor (11 workers), MDNR (2 to 3), EPA (1 to 4), and
the Region VII Ecology and Environment, FIT and TAT teams (6
people nearly full time).
Excavation
Excavation of the buried drums at the site was accomplished
utilizing a track mounted hoe. A series of hand signals were worked
out by the operator. These hand signals enabled the operator to
continue removal of the buried drums from the fill area while not
actually being able to see them at times.
By carrying out the excavation portion of the removal operation
in this manner the overall efficiency and safety of this phase of
operation was greatly enhanced; 1,238 drums were removed from
the fill area. The entire excavation was accomplished without any
incident, which could have directly affected the health and/or safe-
ty of personnel directly associated with this phase of operation.
Drum Data Collection
The data collection procedure used at the Callahan site is unique
and was developed by the on-site team to provide a workable
system meeting the following criteria:
•Provide adequate records to support responsible party actions by
USEPA legal staff
•Provide a permanent marking/identification system for all drums
removed and stored on-site
•Permit transmittal of information from the excavation and
sampling areas to the command post by radio because of ex-
treme weather conditions
•Permit on-site containment of spilled and leaking drums in a
time frame commensurate with the emergency nature of the re-
moval actions
The system developed for use at the Callahan site is described as
follows:
Data Sheets
To keep track of the drums a single sheet was prepared for each
drum containing the following information:
•Person who tagged the drum assigning it a unique and sequential
drum number stamped onto a metal tag
•External conditions and markings on each drum
•Photograph number and photographer's name
•Quantity and appearance of drum contents
•Field test results
•Screening category color code assigned to drum based on screen-
ing analysis and field tests
•Storage category color code assigned to drum based on screening
analysis and field tests
A second set of notes were called the field progress record which
was a series of sheets with sequential drum numbers for all drums
removed from the site. These sheets contained the following infor-
mation:
•Drum number
•Date and time samples were collected from the drum
•Summary of all items on drum sheets
•Disposition of the drum—bulk disposal to Bob's Home Service
or on-site row storage
•Date drum entered secure storage
Immediately after a drum was removed from the fill area by the
backhoe, it was placed on the ground in an adjacent vacant area.
The State OSC would then examine the drum carefully, scrubbing
it when necessary to uncover markings. After photographing the
markings and tagging the drum with its unique and sequentially
numbered tag, this information was transmitted by radio to the
command post where the information was recorded. The team
would proceed to the next drum following the same procedure.
The drum was then transported to the sampling area where a
visual description of the waste was recorded along with an HNU
organic vapor measurement, radiation and hydrocyanide (HCN)
test results. This information was also transmitted by radio to the
command post where it was also recorded by specific drum number
on the respective drum sheet.
Sampling, Waste Characterization and Color Code
All of the sample jars were prelabeled in the laboratory before
being sent to the site. This minimized the recording responsibility
by the sampling and staging crew whose manual dexterity was
severely hampered by several layers of rubber gloves. The empty
labeled jars were transported from the laboratory to the site and
stored in a locked, metal storage shed. To preserve the integrity of
the sample jars after they were prelabeled, the cases were sealed
with filament tape to be opened only at the drum sampling staging
area. When stored on-site, the empty jars were locked in the metal
shed.
The cases of sample jars were opened by the sampling crew at the
drum sampling area. The sampling crew was responsible for perfor-
ming the following field tests as the drums were being opened:
•Check for radiation with radiation meter
•Measure the volatiles with an HNU photoionizer unit calibrated
to benzene
-------
CASE HISTORIES
257
•Check for hydrocyanic acid (HCN) with a Bendix air monitor-
ing pump
•Give a visual description of the contents and amount of waste in
a drum
These results and description were immediately transmitted from
the drum sampling crew to the OSC in the command trailer. The
OSC then recorded the data on the respective drum data sheet.
Three (3) samples (sample A, B, and C) were simultaneously taken
of each drum by the drum sampling crew.
After the sample jars were filled, the drum number, time and
date were recorded on the sample labels and then the sample jars
were sealed. A layer of aluminum foil was placed over the mouth of
the sample jar before the lid was screwed on. When the three (3)
cases of sample jars (representing 11 drums with a set of blank jars
as a control) had been reached, they were transported to the redline
and placed in a plastic bag. The sets of samples and the respective
drum data sheets were then transported to the field lab, located at
EES, via government vehicle.
A Region VII TAT member accepted the drum samples at the
laboratory and segregated them according to their A, B, or C
designation. Analytical tests were run on the C set. These tests in-
cluded a water reactivity test, pH determination, oxidizer test, and
an open torch flammability test and a Seta flash closed cup test set
at 140 °F.
After the above tests had been run, the EES chemist prescribed
the appropriate compatibility drum color code. (See Table
1—Compatibility Chart and Table 2—Color Code). The drum col-
or code was then relayed to the OSC via telephone or radio. The
OSC could then have the individual drum placed in the appropriate
compatibility category section. The overpacked drum was then
marked with itsscolor code utilizing a grease pencil and placed into
its proper secure storage area.
Table 1.
Compatibility Field Tests
Test
l.pH*
Category
Caustic (NF)
Caustic (F)
Acid (NF)
Acid (F)
Oxidizer (F)
Oxidizer (NF)
2. Water Reactive
3. Oxidization/Reduction
4. Radioactives
5. Volatile vapor/gases
6. Flammability
*pH is the level at which the release of cyanide, sulfide and sulfide gases pose a threat.
(F) Flammable
(NF) Non-Flammable
Table 2.
Waste Group Color Codes and Corresponding Waste Properties
Color Code
white
red
red/orange
blue
yellow
yellow/green
orange
green
blue/green
blue/white
red/white
Waste Properties
pH>7, non-flammable
pH >7, torch test properties
pH >7, Seta flash positive
pH <•!, non-flammable
pH <7, torch test positive
pH <7, Seta flash positive
water reactive
oxidizer, torch test positive
oxidizer, Seta flash positive
oxidizer, non-flammable
radioactive
The procedure utilized for the lab screening analysis was based
on the "Compatibility Field Testing Procedures for Unidentified
Hazardous Wastes,"1 developed by Turpin, Lafornara and Allen,
Environmental Response Team, USEPA, Edison, NJ.
Chain-of-Custody
Mr. Michael demons was responsible for chain-of-custody of
the samples after they reached the screening laboratory. The "B"
drum samples were handed to MDNR along with a copy of
MDNR's chain-of-custody form for every drum sample, soil sam-
ple and blank jar sample. Due to the large quantity of the "A" and
"C" drum, soil and blank jar samples, a type-written chain-of-
custody form was devised. The samples were aligned in numerical
order according to drum number and then placed 11 to a case with
a blank jar sample. Each case was then marked in indelible ink:
•Callahan
•"A" or "C" respectively
•The span of drum samples contained with the case, i.e. 13-23
Each case was then sealed with filament tape and an USEPA
custody seal placed over the taped joint of each cardboard case. AH
of the "A" and "C" samples were signed over to William Keffer,
OSC of the Callahan site on Feb. 11, 1982.
Personnel Safety and Decontamination
The guidelines used to determine the appropriate levels of per-
sonnel protection are outlined in two manuals developed by E&E
for USEPA.
•Personnel Protection Manual2
•Hazardous Waste Site Investigation Training Manual3
In addition to these guidelines and with assistance from ERT's
safety officer, safety guidelines were developed and furnished to
the cleanup contractor and other on-site personnel prior to initia-
tion of the federal removal actions.
These procedures required all personnel inside the posted red
zone to be wearing a minimum of a sealed plastic splash suit and an
air purifying respirator with combination cartridges for organic
vapors and particulates; the masks were changed daily. Neoprene
boot and neoprene gloves were also worn with all joints taped.
Those personnel in the sample collection area who were routinely
leaning over the drums were required to be on supplied air.
All contaminated clothing and used cartridges were placed in the
bulk disposal bins and sent to Bob's Home Service for disposal.
Monitoring of the site area and personnel was accomplished by dai-
ly use of passive carbon vapor monitors supplemented by carbon
tubes and Bendix pumps when weather permitted. In addition,
HNU photoionizers were used in the sampling and excavation area,
as well as at a 20-station network around the site, to determine
organic vapor levels.
Disposal
As a result of the removal activities, 1,238 drums were removed
from the fill area. Of the drums removed, 592 were determined to
be empty or non-flammable solids with no other hazardous
characteristics. The contents of these drums were placed in 15 and
30 yd3 containers and disposed of at Bob's Home Service. All
transportation of drums from the site for disposal was made under
Missouri Manifest #02121005001 as Hazardous Waste N.O.S. with
the state OSC serving as the generator. This procedure resulted in a
very substantial cost saving for the federal action and also minimiz-
ed the on-site storage space ultimately required.
The remaining 613 drums were all overpacked and were placed in
secure row storage according to chemical compatibility category, as
determined by the screening test.
Costs
A summary of the costs incurred during the Federal Removal Ac-
tions is given in Table 3.
-------
258
CASE HISTORIES
Table 3. Both the MSA 401 self contained breathing apparatus and MSA
Summary of Job Costs ultratwin respirators on occasion experience operational dif-
Callahan Site—Ellisville, Missouri ficulties. The low pressure alarms on the regulator units of the
U.S. Environmental Protection Agency MSA ^j must be adequately lubricated to insure that the warning
Region VII, Kansas City, Kansas bdl gQes Qff ^ 5QQ psi ^ individuai should always be aware of his
Management $34,824.50 tank operating pressure when working in extremely cold
Labor 68,123.63 temperatures.
Subsistence & per diem 2,576.00 The hand-keyed microphones used would freeze. Communica-
Overpacks 52,140.00 tion was also difficult while wearing either an air-purifying
Disposal 4,105.00 respirator or self-contained breathing apparatus.
Equipment 34,549.68 Decontamination procedures, normally considered to be quite
Analysis (on-site) 10,900.90 routine, must be modified. For example, when attempting to rinse
Analysis (off-site—contract) 6,325.00 an individual and washdown boots and equipment in winter
Safety Equipment & Materials 28,119.55 weather conditions, copious amounts of alcohol or methanol must
Security Service 5,057.67 be added to the wash and rinse to prevent freezing. This solution
Fencing 4,317.10 win also freeze posing a slipping hazard for the workers.
Road & Barrel Pad Material 4,942.65
Fuel & Lubricants 1,289.41 CONCLUSIONS
Miscellaneous Expenses 1,411.31 Jn retrospect) a great deal was iearned about conducting an
Total $258,542.92 emergency removal action of this nature under extreme and often
Cost per drum excavated $1205.00 harsh weather conditions. These emergency actions were ac-
Cost per drum removed 214.56 complished successfully in a cost effective manner despite the
adversity of the weather and operational difficulties encountered.
Unusual Working Cooditions Each difficulty was overcome without serious incident.
A number of unusual working conditions were encountered ACKNOWLEDGEMENTS
while conducting the removal operations at the Callahan site. The „, iU . , t . . . ,., „..,,. „ „ ™ . c r
weather proved to be the main adversary of the site operations. k ™e aut.1?,or wlshAes ° ^knowledge Mr .William Keffer, Chief of
Conditions were indeed variable and ranged from temperatures of th* Surveillance Analysis/Environmental Services Division who
- 50 °F wind chill, deluge rain storms and blinding snow which ac- acte,d as the On-&ene Coordinator for the Region VII, USEPA. As
cumulated to more than 24 in. such- he was a dnvmf f°rce and in«piration to most of us who
Operations which under normal circumstances are considered to snpent the dcurat.lon of the f™**0™rt.e' A'S° Enviro"menta
be routine, suddenly became unusually difficult, time consuming Emergency Serv.ce personnel, the Missouri Department of Natural
and frustrating. When the temperature is below freezing, the field Resources personnel, and Ecology & Environment, Inc., all of the
person suddenly is faced with developing new working techniques. author ? C°»ea8ues who participated on- and off-site. The technical
For example, instead of using a glass rod of spoon, chisels and rei"?*™ *" .""T, aCtl°n *" ^ *
hammers are required to obtain samples of certain materials. The and Wllliam Kittnaler.
rubber gloves required to maintain adequate personnel protection REFERENCES
become stiff, making it extremely difficult to perform even routine
tasks such as writing or opening sample bottles or jars. Rubber '• Turpin, R., Lafornara, J.P. and Allen, H., "Compatibility Field Test-
boots normally issued for hazardous waste site operations are in- in8 Procedure for Unidentified Hazardous Wastes." USEPA, Edison,
adequate to protect individual team members from frost bite or ex- N"''
treme discomfort. Special insulated safety boots become a necessi- 2- USEPA, Personal Protection and Safety, Course 165.2, 1980.
ty. 3. USEPA, Hazardous Waste Site Investigation Training. 1980.
-------
OPERATING EXPERIENCES IN THE CONTAINMENT
AND PURIFICATION OF GROUND WATER AT
THE ROCKY MOUNTAIN ARSENAL
DONALD G. HAGER
Rubel and Hager, Inc.
Tucson, Arizona
CARL G. LOVEN
Rocky Mountain Arsenal
Commerce City, Colorado
INTRODUCTION
The Rocky Mountain Arsenal (RMA) was established by the
United States Department of the Army in 1942 on approximately
20,000 acres northeast of Denver. Its mission has been to manufac-
ture chemicals needed by the military branches of the U.S. Govern-
ment. In more recent years the facilities have been used to destroy
and detoxify obsolete chemical weapons and contaminated hard-
ware as well.
Since 1946, some of the facilities have been leased to a private
contractor who has manufactured various types of insecticides and
herbicides for commercial agricultural applications. In 1974, a por-
tion of the RMA site was taken over by Stapleton Airport (Fig. 1)
and additional land acquisition by the airport is expected in the
near future. As a result of these manufacturing and demilitariza-
tion activities both soil and water have become contaminated with
chemicals which are both inorganic and organic in nature. Many of
the materials that were processed were either toxic and/or
suspected carcinogens.
In 1974, it was discovered that several of the chemicals were be-
ing transported off site by ground and surface waters. To address
this and other potential problems, a Program of Installation
Restoration was established at RMA.
Figure 1.
Groundwater Contours and Bedrock Highs
259
-------
260 CASE HISTORIES
Although many purification techniques were evaluated, adsorp-
tion on activated carbon was the most cost effective. Dynamic flow
granular activated carbon column testing indicated that diisomethyl
phosphonate (DIMP) was the first of the contaminants to appear in
the effluent of the adsorption test columns. Based upon these
laboratory results, several dewatering and reinjection wells were in-
stalled in the area shown in Fig. 2 and a full scale interim granular
carbon system was constructed in July 1978.
Calgon Adsorption Service equipment was used for the interim
purification facilities; it consisted of two multi-media filters follow-
ed by a fixed granular carbon adsorber. This system operated from
July 1978 to June 1981 using a maximum effluent DIMP concentra-
tion of 500 jig/I as the principal operating parameter for carbon
performance and periodic replacement. The system design features
and salient performance data for the period of July 1978 through
June 1981 are shown in Table 2.
Table 2.
RMA North Boundary
Interim Granular Carbon Treatment
(1/7/78 through 6/30/81)
Figure 2.
North Boundary Plan View
Adsorption Mode
Carbon Bed Volume ft3
Flow Rate (gpm)
Contact Time (minutes)
Type of Carbon
DIMP Concentration
Influent
Effluent
Carbon Use Rate
Projected
Actual
Single stage fixed bed
666
70-100
50-300
Calgon Service Carbon
1200/ig/l average
non-detectable
0.33 Ibs per 1000 gallons
l.OOlbper 1000 gallons
CONTAMINATION CONTROL ACTIVITIES
Since 1974, extensive hydrogeologic and decontamination studies
have been conducted under the direction of RMA, Waterways Ex-
perimental Station (WES) of the Corps of Engineers and the
Mobility Equipment Research and Development Command
(MERADCOM). These studies (Fig. 1) have identified specific
areas which are sources of contamination and have defined the pat-
tern of flow of groundwater at the RMA site.
RMA North Boundary Interim Facility
The initial contamination control project at RMA was initiated
at the North Boundary. Several organic contaminants were
detected in the bog water (Table 1). An extensive program was in-
itiated to examine this water as well as nearby water from Wells
PW2 and PW3 (Fig. 2).
Table 1.
RMA North Boundary
Groundwater Contaminants (Wells PW2 and PVV3)
Compound Concentration 0*g/D
Aldrin
Dieldrin
Dichloropentadiene
Diisomethyl phosphonate
Endrin
Dibromochloropropane
p Chlorophenylmethyl
sulfide
solfoxide
sulfone
PW2
<2.0
4.5
1000
530
8.6
7.6
68.3
53.3
40.5
PW3
< 2.0
<: 2.0
82
2800
<2.0
<10.0
<10.0
<10.0
The successful three year operating period demonstrated that the
granular carbon adsorption process would reliably remove those
organic contaminants found in RMA North Boundary ground-
water. The system operated virtually unattended except for
periodic monitoring, backwashing and carbon replacement pro-
cedures.
The carbon exhaustion rate for the demonstration period was
one pound of carbon per 1000 gal of water treated. This compares
with a projected exhaustion rate of 0.33 Ib C/1000 gal based upon
the pilot studies previously discussed. Part of this difference may be
attributable to air pockets in the granular carbon beds which form-
ed when water drained from the adsorbers during periodic shut-
downs. The installation of an anti-siphon valve corrected this early
problem.
Based upon these encouraging results, a permanent boundary
water containment and purification system was designed and in-
stalled.
RMA North Boundary Expanded
Permanent Facility
The permanent facility for groundwater containment at the
North Boundary included a 6800 ft long clay slurry wall which was
installed 200 ft inside the RMA perimeter. Dewatering and reinjec-
tion wells along the clay barrier remove water for treatment at a
central facility and deliver the purified water back to the ground
flow.
The dewatering system is divided into three separate water collec-
tions systems which have waters of varying organic composition. It
was anticipated that three adsorption process trains would provide
added treatment flexibility than that afforded by treating a mixed
influent (Fig. 3).
-------
CASE HISTORIES
261
r
?,., >.
*r*
1
x
hrl
JL
hr
JL
hr
JL
i
1 0.
r?
PLJII
1
?.
H
1
JL
hr
JL
hr
J_^
hH
JL
i
l ^
FIL.T6J1*
IAKTE.P-
D£tJA,T6JZ.iN^ SUMP
U1E.L-L/6
WilNJECTIOM
PIUTEJZ4.
BtO
Figure 3.
RMA North Boundary
Permanent Granular Carbon Facility
Economic analysis of capital costs associated with on-site reac-
tivation of granular carbon compared with projected carbon usage
rates indicated that on-site carbon reactivation was not justified at
that time. The granular carbon replacement cost was nevertheless
projected to be the principal operating cost. In order to optimize
carbon utilization, various countercurrent adsorption systems were
evaluated. Commercially proven countercurrent adsorption
systems are either two stage downflow fixed beds in series or
up flow pulsed beds.
A pulsed bed adsorption system supplied by Westvaco was
selected for the permanent RMA North Boundary treatment facili-
ty. The system is comprised of three adsorbers in parallel, cartridge
filters ahead of and following the adsorbers, storage vessels for
both fresh and exhausted carbon and two pressure transfer vessels
to move carbon between the equipment in a measured slurry
volume. A process diagram for the permanent adsorption system is
shown in Fig. 3 while the design features are given in Table 3.
Table 3.
RMA North Boundary
Permanent Granular Carbon Treatment
Adsorption Mode Three pulsed beds operating in parallel
Carbon Bed Volume (each vessel) 1000 ft'
70ftJ
WV-G 12 x 40
Average 250/Maximum 350
Average 84/Minimum 30
Volume of Carbon Pulse
Type of Carbon
Design Flow Rate (gpm)
Contact Time (min.)
Influent Organic
Contaminants
D1MP* /ig/1
DCPD" pg/l
DBCP*** (ig/1
•Diisomethylphosphonate "Dicyclopentadiene
Process
Stream A
700-1200
10-2000
1-5
Process
Stream B
100-500
500-2400
1-5
Process
Stream C
10-100
24
1-5
Effluent
Objectives
500
24
0.2
*Dibromochloropropane
-------
COST EFFECTIVE MANAGEMENT OF AN ABANDONED
HAZARDOUS WASTE SITE BY A
STAGED CLEANUP APPROACH
KENNETH F. WHITTAKER, Ph.D.
Camp Dresser & McKee, Inc.
Boston, Massachusetts
ROBERT GOLTZ,
U.S. Environmental Protection Agency, Region II
New York, New York
SITE DESCRIPTION
The Bridgeport Rental and Oil Services (BROS) site is a former
oil processing and reclamation facility in southern New Jersey (Fig.
1). The site has an overall area of approximately 26 acres, contains
an 11.5 acre unlined lagoon and over 88 tanks and storage vessels,
and is bounded on three sides by two fresh water ponds, a peach or-
chard and marshland (Cedar Swamp). A small creek (Timber
Creek) running through the swamp passes in the near vicinity of the
lagoon and eventually discharges to the Delaware River about 3
miles from the site.
The lagoon has an average depth of 10 to 15 ft but depths as
great as 60 ft in places have been reported. A thick layer of heavy
oil floats on the surface. Large quantities of construction debris,
trash, and several large tank trucks are partially submerged in the
lagoon and are all coated with oil. Large numbers of floating drums
are also present.
The storage vessels on site range in capacity from a few thousand
gallons to tanks with volumes of over 300,000 gal. The volume of
material in each tank is not consistent. Initial site surveys have
shown the majority of the tanks are either empty of contain only
small quantities of bottom sludges. Two of the seven major tanks
(i.e., greater than 300,000 gal) contain large quantities of liquid
material.
The site poses a threat to both surface and groundwater in the
area. Private drinking wells in the area have been contaminated
with varying amounts of volatile organic compounds. Previous
breaching of the dike surrounding the lagoon has led to discharge
of material to Timber Creek and the deforestation of 8-10 acres of
land adjacent to the lagoon. There is visual evidence of seepage of
oily materials into the surrounding fresh water ponds.
HISTORY
Since 1969, there have been no known discharges from the
lagoon but the general condition of the site has deteriorated. The
level of the lagoon has continued to rise far above the surrounding
groundwater table, suggesting the bottom sludges may have sealed
off, to a limited extent, the lagoon from the groundwater.
From 1975 to 1980, various remedial efforts, such as skimming
and pumping of the oil, booming, etc. were proposed or attempted
but the size of the lagoon, the large amounts of floating debris, and
the viscosity of the oil prevented their successful application.
Due to high lagoon levels, a new dike was constructed in 1981 us-
ing funds from Section 311 (K) of the Clean Water Act. The dike
was designed to contain the lagoon liquid for 4 to 5 years.
In Sept. 1981, Camp Dresser & McKee (CDM), as USEPA's
Zone II contractor, was directed to begin a comprehensive site
study leading to the definition of the most cost-effective means for
remedial site cleanup. The study was to consist of a review and
compilation of existing information, sampling of tanks and lagoon,
and the carrying out of a hydrogeological investigation. However,
site conditions inthe spring of 1982 dictated that a more direct ap-
proach to the problem be undertaken.
\)ffl~4
Mli-^^L
1
s 'V7SS^
Figure 1.
Schematic of Site
JUSTIFICATION FOR A STAGED
CLEAN-UP APPROACH
By June 1982, the lagoon surface had risen to approximately 8 in.
below the top of the dike, the combined result of higher than
average rainfalls and unforeseen limitations in subsurface percola-
tion and evaporation. An emergency action was initiated, and
USEPA's mobile activated carbon unit was employed to lower
lagoon levels. Operations were terminated in late July when ap-
proximately 2 ft of liquid, approximately 5 million gal, had been
removed.
Also in June, CDM was directed to develop feasible treatment
methods for surficial site clean-up, that is, lowering the level of the
lagoon by 10 ft and emptying the large capacity storage tanks.
(Lowering to levels below this depth was dismissed without further
study due to the possibilities of bottom deposit fracturing and
subsequent increases in the rate of groundwater contamination.)
This work was to be accomplished in two phases. Technically and
economically feasible methods of waste treatment or removal were
262
-------
CASE HISTORIES
263
to be specified. Following this, technical specifications were to be
prepared for incorporation into U.S. Army Corps of Engineers
standard specifications for preparation of bid documents.
The objectives of this task were threefold:
•Immediate environmental hazards which would occur as a result
of lagoon overtopping or storage tank rupture would be allevi-
ated.
•Operating problems, regulatory requirements and treatment sys-
tem reliability relative to final site remediation would be identi-
fied.
•Lowering of the lagoon levels would expose a large amount of
the currently submerged area, thus, allowing the investigation
of the lagoon bottom characteristics needed to define final site
remediation options.
For cost effectiveness, the surficial clean-up system was to be
designed, to the greatest extent possible, to be equally applicable to
the long-term site solution.
SITE SAMPLING AND WASTE CHARACTERIZATION
The first step of CDM's surficial site cleanup study was to obtain
representative samples of the lagoon aqueous phase, tanks, and
floating oil. A private contractor was called in to perform these ser-
vices. The work was carried out in three phases. First, the structural
integrity of each major storage tank was assessed to evaluate the
potential for rupture and also to identify what tanks could be used
for either off-site waste transportation staging areas or as com-
ponents of an on-site treatment system. Secondly, storage tank pro-
files were prepared detailing the volume, depth and chemical
characteristics of each phase, or in the case of empty tanks, con-
taminants remaining as determined from wipe-testing. Thirdly, a
substantial lagoon profiling was undertaken. Samples of surface
oil, aqueous materials two feet below the oil surface, and aqueous
materials at mid-depth in the lagoon were obtained during four
traverses of the lagoon. The samples from each depth were com-
posited for each traverse and submitted to chemical analysis for
priority pollutants. Lagoon depth readings were taken at each
sampling point to confirm previous data.
Significant operational problems were encountered during
lagoon sampling. Fatigue was a major problem, due to the combin-
ed effects of high temperature and the protective clothing and
respirators worn by all personnel. Also, the large amount of
floating debris and oil viscosity greatly decreased the rate at which
the floating sampling station could be pulled across the lagoon. The
end result was that only three limited traverses could be made in the
alloted one-week sample effort. A fourth "traverse" was ac-
complished by suspending sampling personnel over the lagoon sur-
face (approximately 40 ft from shore) by means of a "cherry-
picker". A diagram of the sampling points is given in Fig. 2.
Legend
• Sample Station
I
| Sample Gridline
Not to Scale
Figure 2.
Sampling Grid
Chemical Characterization of Tank Contents and Lagoon Oil
Five of the major storage tanks were essentially empty. Analysis
of wipe test samples on the empty tanks showed a large number of
generally uncharacterized hydrocarbons, various benzene, and
phenolic species, and a number of polyaromatic hydrocarbons,
such as phenanthene and naphthalene. The concentration of these
materials was generally less than 200 mg per wipe. Relatively high
concentrations (1-15 jig per wipe) of phthalate species including bis
(2-ethyl hexyl) phthalate, di-n-butyl phthalate and di-n-octyl
phthalate, were found in all tanks.
Three tanks contained significant quantities of liquid or sludge.
A composite sample of each phase (i.e. sludge, water and oil) was
taken for analysis. A general summary of the amounts of
chlorinated hydrocarbons identified during this testing, along with
those from the sludge samples in the "empty" tanks, is given in
Table 1. Also included in this table are the results from a com-
posited sample of floating lagoon oil.
The greatest threat posed by the site resides in the lagoon surface
oil with its very high concentration of solvents and PCBs. The need
for preventing the escape of this material into the environment (by
preventing overtopping of the lagoon) is readily apparent especially
when one realizes that approximately 5 million gal of this material
float on the lagoon.
Characterization of Aqueous Lagoon Contents
In general, the aqueous phase of the BROS lagoon can be
characterized as a lightly yellow colored liquid containing a
moderate amount of suspended solids. Samples of the wastewater
were noted for their strong, "motor-oil" like odor. Although there
was no visual evidence of a substantial degree of emulsified oil,
some samples were observed to contain a small number of oil
droplets and generally uncharacterized settleable materials.
The aqueous waste has a TOC ranging from 180 to 220 mg/1 and
a COD of 720 mg/1. The 5-day BOD is quite low in relation to these
other parameters, amounting to only about 90 mg/1. Total
suspended solids were approximately 60 mg/1. The oil and grease
content of the waste amounted to approximately 80 mg/1.
Based on the above analysis, in combination with treatability
studies conducted in CDM's laboratory, the waste can be said to be
composed of 33% readily volatile species and 33% large molecular
weight oily type materials. The remaining third of the measured
organic matter was generally uncharacterized, containing a wide
diversity of organic species.
Only dilute concentrations of the metal species, well below
established discharge limits, were found in the waste. Only two
metals appear to be of potential concern in meeting discharge
criteria, namely lead and zinc, and these exist at concentrations bet-
ween 0.5 and 5 mg/1.
Organic chemical characterization was provided in four general
categories, acid compounds, base/neutral compounds, volatiles
and tentatively identified compounds. Significant interferences oc-
curred in the analysis of acidic and base neutral compounds. (The
data developed showed remarkable similarity to the type of data ex-
pected from the analysis of undiluted crude oil.) The end result of
these interferences was that low concentrations of acid or base
neutral compounds and pesticides were not determinable to a high
level of reliability. Within this context, any species indicated as not
detected had to be viewed as being present at a concentration of 50
jig/1 or less (in contrast to some tank samples, PCBs were not
found in this water). The value for species which were definitely
found to be present could, however, be considered to have greater
validity. Significant interferences were not encountered in the
analysis of the volatile species, and thus the data would be con-
sidered as highly reliable.
With the exception of diverse phenolic species, there appeared to
be no substantial concentrations of acidic priority pollutant com-
pounds definitely present in the waste. The number of base neutral
compounds was also quite limited, being effectively limited to
naphthalene. Both of these species were present in quantities
significantly less than 1 mg/1.
-------
264
CASE HISTORIES
Table 1.
General Physical and Chemical Characterization of Tank Contents
•ND means noi delected
Tank No.
1
2
On-Site
Designation
C-2
C-3
RP-6
RP-7
RP-11
C-6
T-13
CC-6
Estimated
Capacity
(gal)
280,000
300,000
300,000
300,000
300,000
520,000
2,000
760,000
Estimated
Sampled
Phase
Sludge
Aqueous
Sludge
Sludge
Oil,
Aqueous
Sludge
Oil
Aqueous
Sludge
Aqueous
Depth
(Vol) of
Sampled
Phase
(400 gal)
(6400 gal)
(1400 gal)
2 in
2 in
310,00 gal
90,000 gal
13,000 gal
(850)
(500)
Status
Unsuitable for storage
Corrosion evident, un-
suitable for storage
Unsuitable for storage
Possibly usable for
storage; floor must be
tested; surrounding dike
inadequate
Unsuitable for storage
Appears struct, sound;
most stable of all tanks
Too small for use as stor-
vessel
Questionable integrity in
places
Chlorinated
Hydrocarbons (ug/ml)
Solvents PCB
ND* ND
ND ND
18 420
140 260
640,000
Lagoon
Oil
Relatively high concentrations of a variety of volatile priority
pollutants were measured. Of particular concern were benzene,
trans 1-2 dichloroethene, methylene chloride and toluene, all of
which had average concentration at or near 1 mg/1. These concen-
trations also illustrate the hazard which will be associated with
treatment of this wastewater, and the care which will be necessary
in controlling volatile emissions if required.
Information on the non-priority pollutants identified in the
waste is given in a listing of tenatively identified, compounds in a
combined waste sample presented in Table 2. A number of large
molecular species were shown to be present at or near the 100jig/l
level. Of particular significance in this listing, however, are the
large measured concentrations of additional volatile species. Very
high levels of acetone (39 mg/1), methyl ethyl ketone (9.6 mg/1) and
methyl isobutyl ketone (4.3 mg/1) were observed.
POTENTIAL TREATMENT OPTIONS
Tank Contents
Due to unforeseen delays in obtaining the chemical analysis of
the storage tank contents and wipe test samples, a detailed review
of feasible waste treatment disposal options for these materials
could not be prepared in time for manuscript submittal. Brief
review of the data indicates, however, the aqueous materials could
be treated in the proposed on-site treatment plant.
The high concentrations (greater than 500 mg/1) of various
solvents and polychlorinated biphenyls '(PBCs) in the organic
phases mandate that these materials be shipped to an approved
hazardous waste incinerator. Note that in all cases except one, the
organic materials in the tanks comprise a relatively small volume,
thus shipping costs can be limited.
Lagoon Water
Determination of the lagoon water treatment system required
three decisions: (1) how far to lower the elevation of the liquid
level, (2) how and where to treat the water, and (3) how to manage
any residues from on-site treatment.
A 10 ft reduction in elevation means removing 67% more water
than a 5 ft reduction; it would achieve a new lagoon area approx-
imately half of that achieved by a 5 ft reduction. In combination
with cleanup of stranded materials on the newly exposed lagoon
7 330
490 960
52 130
530 980
27 1540
ND ND
ND ND
ND ND
49 1230
Table 2.
Tentatively Identified Compounds in Composited Sample
of BROS Lagoon Wastewater
Est. Conceit-
Compound trallon
Name Fraction (ug/l)
2-ethyl-l-hexanol Acid-Base Neutral 100
Heptanoic Acid Acid-Base Neutral 100
2-ethyl-hexanoic acid Acid-Base Neutral 110
Octanoic acid Acid-Base Neutral 100
Decanoic acid Acid-Base Neutral 110
6-methyl-tridecane Acid-Base Neutral 100
Heneicosane Acid-Base Neutral 150
3,5,2,4 trimethyl tetracotane Acid-Base Neutral 130
Hepta decane Acid-Base Neutral 500
7-hexaleicosane Acid-Base Neutral 130
Acetone Volatile 39,060
Methyl-ethyl ketone Volatile 9,600
Methyl-isobutyl ketone Volatile 4,300
Xylenes Volatile 960
Options Related to Lagoon Surface Elevations
Past soundings of the lagoon's depth at various locations were
used to estimate the physical effects of lowering the lagoon by
various amounts, as follows:
Reduction Lagoon
Surface Elevation
(ft)
New Lagoon
Surface Area
Volume
Removed*
(gal)
0 500.000
5 320.000
10 165.000
•Neglecting rainfall during Ihe removal period.
0
15 x 10*
25 X 10*
bottom and re-diking this smaller lagoon will collect less total
precipitation than a larger lagoon. In addition, a 10 ft decrease in
surface level would eliminate the hydrostatic driving force for
fluids to move from the lagoon into the groundwater. Therefore, a
drop of the lagoon elevation by 10 ft to an approximate elevation
of 3 ft above sea level was recommended.
-------
CASE HISTORIES
265
Flow rates required to achieve this elevation drop over various
time periods were computed. Rainfall, based on the mean annual
precipitation and conservatively assuming no evaporation, is in-
cluded:
Time to Lower Lagoon
by 10 ft.
(months)
Pumping Rate
(gpm)*
2 316
4 200
12 74
For concluding surficial cleanup in an expeditious manner, CDM
recommended a flow rate of 200 gal/min. An interim treatment
system to handle this flow could be used on an intermittent basis in
the future to maintain the lagoon at an appropriate level. Opera-
tion for six to ten weeks per year would achieve this goal.
Lagoon Water Treatment and Disposal
A wide variety of cleanup options were reviewed for their ap-
plicability to the BROS site. It is not possible to discuss here each
reviewed treatment process. A summary of CDM's review is given
in Table 3. However, since the treatment plant includes both floc-
culation/sedimentation and activated carbon adsorption, a
separate discussion of each of these recommended treatment pro-
cesses for the full-scale plant is given below.
A series of jar tests were conducted on the raw waste. Different
ferric chloride doses, polymer (anionic) doses, and pH combina-
tions were tried. The best removal of suspended solids and color
was noted under the conditions of: (1) final pH of 6.0, (2) FeCl3 .
H2O dosage of 200 mg/1, and (3) 4 mg/1 of anionic polymer. The
supernatant was clear yellow after filtering and the final TOC of
Table 3.
Non-Feasible Treatment Processes Reviewed for BROS Site Wastewater
Candidate
Process
Treatability
Studies Conducted
Feasibility Status and Reasons for
Inclusion or Rejection
Cost Estimates for 200 gpm Unit
Air Stripping
Laboratory-scale
air sparging studies
Large proportion of the waste (approx.
35%) amenable to removal by air stripping.
However, NJDEP regulatory requirements
preclude significant VOC emissions.
< $15,000
Biological Treatment
Bacterial acclimation to
waste, toxicity oxygen
uptake rates
Not considered feasible due to exceedingly
slow rates of bio-oxidation, anticipated
start-up delays, and substantial volatile
emissions
Not performed due to lack of
technical feasibility
Reverse Osmosis
None
Due to extensive pre-treatment required,
influent quality to unit would have to be
superior to plant discharge limit; thus
unit would be extraneous
$170,000 (with sand filter)
Ultrafiltration
Pilot-scale studies by
vendor (CDM present)
Maximum organic removal approximately 50%
TOC with highest reject membrane.
Effluent not suitable for discharge-
concentrated stream would require disposal.
Not performed due to poor performance
noted.
Hydrophobic Resin
Beads
Parallel test and control
adsorption columns
No significant amount of organic material
removed
Not performed due to lack of technical
feasibility
Ultraviolet/Ozone
Oxidation
Pilot-scale studies by
vendor
Demonstrated effectiveness to reduce waste
TOC concentration down to acceptable dis-
charge levels. However, cost levels too
high.
$1.6 million for contractors and
ozone generator; $.8 million for
chemicals, power, etc.
Chemical Fixation
Pilot-scale studies by
vendor
Chem-fix process gives solid acceptable for
disposal under EP extraction test. However,
no decrease in volume of material treated
and limitations in on-site storage space
of material.
$2 - $3 million
Ion Exchange
None
Organic chemicals in raw waste have high
probability of fouling resin. Disposal
proven for concentrated regenerant
solutions. Metals can be removed in
treatment process as part of other
disposal techniques.
System not considered necessary
no pricing attempted.
Off-Site Disposal
(to treatment plant)
By Truck
By Pipeline
Preliminary treatability Temporary "quick-fix solution" no process
studies conducted by OuPont's equipment would remain for application
Chambers works to 1on9-term clean-up. High cost.
None
The nearest treatment plants are
approximately two miles away. Thus,
pumping and piping costs would be high.
Treatment efficiencies are questionable
and VOC emissions would be a problem.
$750,000 water treatment
$750,000 waste transportation
Not priced due to lack of
technical feasibility
-------
266
CASE HISTORIES
the filtrate was 119 mg/1, a substantial reduction from the filtered
(no flocculation) waste concentration of 175 mg/1.
It appeared highly unlikely, based on CDM's laboratory results,
that this flocculation alone would be sufficient to meet discharge
requirements. However, some additional process such as carbon
adsorption or oxidation is necessary to remove soluble organic
materials. One advantage of the use of adsorption technology
would be that this process could be beneficial in removing dissolved
metal species, especially where powdered activated carbon would
be used in combination with a ferric hydroxide coagulation step.
Activates carbon can be applied to wastewater in two ways. The
most common method of contact is through passage of the waste
through a stationary bed of granular carbon. An alternate method
of carbon contact is to add powdered carbon. Powdered carbon,
due to the fact that most of the available surface area is on or near
the external surface of the particle, results in quicker adsorption
equilibrium since materials to be adsorbed only have to travel
relatively short distances.
For evaluating columnar carbon adsorption for treatment of
BROS wastewaters, CDM was fortunate in having the results of the
on-site emergency response cleanup action which used USEPA's
mobile physical-chemical treatment system. Based on these data,
granular activated carbon adsorption, appeared unsuitable for
treatment of the wastewater. Rapid exhaustion of the 16,000 Ib of
carbon in the unit was observed. Furthermore, actual achievement
of an effluent concentration limit of 50 mg/1 TOC was observed in
only a few isolated instances.
Hydraulic surface loading rates in the physical-chemical treat-
ment system were approximately 6 gal/min/ft2 as opposed to a
"normal" operating range of 2 gal/min/ft2. Such a high loading
rate could have been responsible, to a certain degree, for the
relatively rapid rise of TOC concentration in the effluent.
Independent of surface loading considerations, however, is the
fact that, even at the very start of the run, effluent from the carbon
columns was quite poor (greater than 40-50 mg/1 TOC). Since the
empty bed detention time in the columns was considerably greater
than standard operating conditions of 15 min (through the use of
three columns in series), this essentially immediate appearance of
high effluent TOC concentration was indicative of the poor adsor-
bability of many of the compounds in the waste.
CDM suspected that the mediocre performance of the carbon
treatment unit was related primarily to the high concentrations of
oily materials in the wastewater, materials which are poorly dif-
fusable and which can block the pores for penetration of other ad-
sorbents.
Partial confirmation of the role of the oily materials in not-
penetrating and/or blocking pores was obtained when it was noted
that greater adsorption efficiency and better effluent quality was
observed as more regenerated carbon was used in the columns.
Since regenerated carbon generally has less surface area but larger
pores than virgin carbon, the observed performance seemed to con-
firm CDM's supposition that carbon adsorption efficiency was
primarily related to the actual availability of adsorbent surface area
to the adsorbable species, rather than to total surface areas.
CDM, therefore, concluded that the poor performance of on-site
carbon beds was probably due, in part, to the poor penetration of
chemical species into micropores. As a consequence, any attempt to
expand the area available for chemical species adsorption, either by
increasing the diameter of the micropores or by maximizing exter-
nal (i.e. non-intra-pore) surface area would increase performance.
Furthermore, it was surmised that long contact times would in-
crease carbon performance by allowing greater amounts of the pore
penetration to occur and, to prevent pore blockage any removal of
floating oils should be undertaken.
To investigate 'and/or confirm these suspicions, a series of
laboratory scale treatabilty studies were carried out. Calgon Cor-
poration was sent a sample of the wastewater and requested to per-
form adsorption isotherm at an acidic and neutral pH. These
studies were conducted in a standard manner using powdered ac-
tivated carbon, raw wastewater and a contact time of 17 hr. The
powdered carbon was quite successful in treating the raw waste,
reducing TOC levels to less than 50 mg/1 TOC at carbon concentra-
tions of 500 mg/1 or greater.
CDM conducted its own combined isotherm/jar test evaluation
with powdered activated carbon and raw wastewater. Significantly
greater amounts of carbon were required (in 2 hr) to reach the same
TOC concentrations as had been observed in the Calgon data in 17
hr. CDM believes that better performance would have been observ-
ed in both cases had suspended materials and oils been removed
before carbon treatment.
In addition, CDM conducted a small scale carbon adsorption
column test at a hydraulic loading rate of 2 gal/min/ft2, and an
empty bed detention time of approximaely 15 min. Coagulated
wastewater was used for this test. Only two effluent samples from
the laboratory column study, taken 2 and 4 hr after commencement
of the study, were analyzed. Both of these samples showed TOC
values in the range of 20 mg/1. This relatively poor effluent from a
new bed of carbon under normal loading, while within permit
limits, suggests that breakthrough to 50 mg/1 could be expected on
the full scale.
In summary, therefore, granular activated carbon does not ap-
pear well suited to treatment of the lagoon water to a level of 50
mg/1 TOC. Powdered activated carbon does, however, show great
potential applicability, presuming sufficient times of contact and
dosages (these two factors being inversely related) are allowed. Use
of powdered carbon was strongly recommended for treatment of
this waste for two primary reasons:
•The technology is applicable to the removal of a diverse variety of
chemical components.
•By adjustment of carbon dose (or contact time) considerable
flexibility of operation is insured, an important consideration
for the expected variability in BROS lagoon wastewater charact-
eristics.
Once candidate options were dismissed on the basis of their
technical feasibility or cost effectiveness, two potentially feasible
options remained. These were off-site disposal to DuPont's
Chambers Works Plant or on-site treatment using flocculation and
powdered activated carbon. Of these two, on-site treatment was
considered the more desirable primarily because it could be easily
incorporated into a long-term final site remediation.
The estimated costs of each alternative were roughly equivalent
at $1.3 million. An additional $600,000 in operating expenses was
predicted for the powdered activated carbon system.
Preliminary cost estimations showed that use of a granular ac-
tivated carbon system could result in a savings of approximately
$400,000 to $500,000 (primarily contributed by reductions in
capital expenditure and reduced carbon usage). However, based on
laboratory and full-scale studies previously described, CDM did
not feel that granular carbon would be adequate to meet the 50
mg/1 effluent discharge limit to Timber Creek specified by NJDEP.
However, if the discharge limit were relaxed to 100 mg/1 TOC, a
granular activated carbon unit could reliably meet that standard.
This was demonstrated by on-site application of USEPA's mobile
physical-chemical treatment system under "worst case" condi-
tions, that is, without pretreatment for the removal of oily
materials.
NJDEP and USEPA are still reviewing the report. Until review is
completed, CDM is not carrying out any design work for operation
beyond the pretreatment step for oil removal.
RECOMMENDED ACTIONS
For a 50 mg/1 TOC discharge limit, CDM recommends construc-
tion of the powdered activated carbon system shown schematically
in Fig. 3. The system is designed to provide "standard" detention
times (i.e. 5-10 min rapid mix, 30 min flocculation and 1 hr of
sedimentation) at a flow rate of 200 gal/min. A powdered carbon
contact period of 6 hr was specified based on the results of isotherm
testing (carbon dosages could be increased on an as-needed basis).
Note that all process tanks will be covered to reduce volatile emis-
sions. Final effluent will be sand-filtered to remove carbon granules
-------
CASE HISTORIES
267
DUAL MEDIA
PRESSURE
FILT6R
SEDIMENTATION
TANK
(15,000 SAL'S)
STORE.
CAKJB IN
EXIST.
TANK
Figure 3.
Recommended Treatment Process Flow Diagram
which could contribute to the TOC. CDM's decision to recommend
this process train was based upon the following considerations:
•The system will be effective in treating most of the identifiable
components in the lagoon liquid under a variety of flow, com-
position, and concentration conditions.
•Because the treatment will use only physical/chemical processes,
very little time will be required to bring the plant up to a full op-
erating condition.
•The treatment plant should be of great benefit in treating addi-
tional lagoon waters related to overall cleanup.
•The possibility exists for many treatment system components to
be used at other sites once work has been completed at BROS.
•The proposed system is the least cost option that provides con-
fidence in meeting discharge requirements.
To the maximum extent possible, CDM intends to specify package
units for the treatment processes. Use of such pre-fabricated
systems will achieve the following goals:
•Short construction time
•Low capital cost
•Ability to design the system in a short time (e.g. structural en-
gineering effort to design concrete tanks will be avoided).
The packages to be specified will be expected to hold up for several
years of continuous service, a period of time which USEPA
estimates will be required for eventual site remediation.
Specification of a Granular Activated Carbon system would
reduce capital outlay by approximately $250,000 by obviating the
need for all powdered activated carbon storage and contact
tankage. Furthermore, due to greatly decreased sludge handling re-
quirements (i.e. essentially limited to ferric hydroxide/oil sludges
from pretreatment) CDM believes the thickener, sludge drying beds
could be replaced by an automatic batch feed centrifuge.
SUMMARY
Upon direction by USEPA, CDM has specified and will design, a
treatment system for waters drawn from a waste oil storage lagoon.
Treatment processes were recommended on the basis of laboratory
and full-scale treatability tests. CDM believes that the most cost ef-
fective treatment solution is the application of a combined floccula-
tion/sedimentation oil removal, activated carbon adsorption treat-
ment system. Pending review of this report, technical specifications
for the appropriate treatment system will be prepared for formula-
tion of site construction bidding documents under the combined
direction of USEPA and the U.S. Army Corps of Engineers.
-------
PICELLO FARM, COVENTRY, RHODE ISLAND:
A SUPERFUND & STATE FUND CLEANUP CASE HISTORY
BARRY W.MULLER
ALAN R. BRODD
JOHN LEO
Rhode Island Department of Environmental Management
Providence, Rhode Island
HISTORY
In the fall of 1977, an explosion and fire alerted area resi-
dents and officials to the presence of a chemical dump site at the
Warren V. Picillo Farm (Fig. 1). The Rhode Island Attorney
General sued to enjoin Picillo from further disposal and to re-
move all hazardous wastes for proper disposal. The State at-
tempted to secure legitimate disposal outlets for these wastes.
When Picillo failed to comply with the court order, the Attorney
General became involved in a lengthy court action which con-
tinues to this day.
Figure 1.
Location of Picillo Farm, Coventry, Rhode Island.
CLEANUP BEGINS
In late 1978, the Rhode Island General Assembly passed an emer-
gency appropriation to begin cleanup activities at the site. The
Rhode Island Department of Environmental Management
(RIDEM) retained the services of a consultant to conduct a hydro-
geological assessment of groundwater contamination,-assess the
extent of wastes buried onsite and develop remedial options for
resolution of the problem (Phase 1).
The remedial action options evaluated included:
•Encapsulation Of The Site—This option considered the place-
ment of an impermeable cover over and wall around the site to
bedrock. It was rejected for the reasons that: (1) a significant
sources of chemicals would remain in a liquid state as they were
contained in deteriorating barrels, (2) the bedrock depth varied
between 25 and 35 ft which would be costly insofar as liner place-
ment was concerned, and (3) the bedrock beneath the site is highly
fractured, i.e., too permeable for a secure bases.
•Interceptor Trenches—This option considered the placement of
trenches down gradient from groundwater and leachate flow. It
was rejected for the same reasons, i.e., deep bedrock, irregular
bedrock surface and fractured bedrock.
•No Action—This alternative was examined since a swamp directly
northwest of the site was found to contain significant quantities of
leachate and acted as a "filter" or point of volatilization of many
organics. It was rejected-as the swamp could not be proven to be
an adequate treatment mechanism for all wastes. Further, no
control over this "mechanism" could be exerted since the disposal
site itself contains a significant source and quantity of chemical
wastes. In a purely social sense, the option of no action was ill-
advised and posed significant consequences.
•Drum and Chemical Removal—This alternative was considered to
be the only viable one as it would insure the source of contam-
inants would be removed. Further dispersion of contaminants in
the groundwater could be monitored.
Following receipt of the consultant's recommendations, it was
decided to excavate and dispose of all barrelled wastes and con-
taminated soil. In the spring of 1980, the RIDEM authorized ex-
cavation of the contents of the northeast trench, which was
thought to be one of the smaller trenches (Fig. 2).
Figure 2.
Location of Disposal Trenches.
268
-------
CASE HISTORIES
269
Ground penetrating radar and metal detection surveys had been
utilized to determine the extent of the trenches.1 Unfortunately,
these techniques could not predict either density of barrels or
depth of the trench. When the excavation was completed, the
depth from grade was 35 ft and the trench contained 2,300 barrels
of wastes as compared to an estimate of 270.
It appeared from the condition of the barrels that this had been
one of the oldest trenches since most barrels were leaking as they
were removed. Many barrels in the upper layers had been crushed,
generating large pools of leachate.
Several important lessons were learned from this first excava-
tion:
•Barrels were allowed to leak during removal causing additional
contamination of the soil. This soil which is contaminated with
PCS' still poses a significant disposal problem.
•The State let a dump sum agreement contract for a specific num-
ber of barrels thought to be contained in the trench. When it
was discovered that the amount of waste far exceeded the fore-
cast, delays were encountered until additional funding could be
procured. The lump sum agreement form was altered to a time
and materials format.
•It was decided to combine the contents of many barrels of sim-
ilar flammable wastes in the tanker prior to thorough final analy-
sis. The result was 4,000 gal of wastes contaminated with PCB
levels in excess of 1,500 mg/1. This waste was ultimately dis-
posed of in the spring of 1981.
CLEANUP CONTINUES
Following completion of disposal of the northeast trench wastes
in the fall of 1980, RIDEM began planning for excavation of the
northwest trench (Fig. 2). It was known from the metal detec-
tion survey and ground penetrating radar studies that the trench
extended for 250 ft and at its widest was 50 ft. From previous
experience in the northeast trench, the revised estimates for barrel
content ranged from 8,200 to 22,400.
A Scope Of Work and Request For Proposal for excavation and
disposal was distributed in Nov. 1980 (Phase 2). It specifically re-
quested a response to the technical issues of drum removal, en-
countering leaking barrels, contaminated soil, site layout, person-
nel required, personnel protection, site decontamination zones and
safety requirements.
Further requirements to be addressed included drum staging
following subsurface removal, at which time leaking drums would
be repackaged and records initiated. Other areas to be addressed
were drum waste classification, sample analysis, on-site treatment,
disposal, recordkeeping and reporting and surface water infiltra-
tion.
All interested bidders were directed to respond not later than
mid-Nov. However, in early Dec. the USEPA announced that they
had money available to fund the excavation portion of the pro-
ject if the State would assume responsibility for disposal of the
waste.
The State agreed to this option and transferred all bids to
USEPA for evaluation. The State examined its bids for disposal
capability and selected a contractor to perform disposal activities.
USEPA chose another contractor to conduct excavation activ-
ity. In early Feb. 1981, USEPA began site preparation by con-
structing several diked storage and staging areas placed so as to
effect an orderly material flow from the trench excavation area to
pumping/staging area to final storage prior to disposal (Fig. 3).
USEPA constructed several areas in which a polyethylene liner
was placed beneath the soil surface to prevent infiltration of chem-
ical wastes from leaking or spilled drums. By the end of Feb., pro-
ject mobilization was begun.
However, USEPA advised the State in mid-March that antici-
pated funding was not available and it would have to terminate its
contract. The State decided to assume the excavation portion and
dispose of as much waste as possible within its own funding con-
straints. RIDEM began preparation for the excavation require-
ments and mobilized all remaining equipment necessary on site. In
mid-Apr, actual barrel removal began.
The excavation of barrels from the northwest trench proceeded
until the end of June, 1981 by which time 4,400 barrels of wastes
had been removed and stored on site. At that point over $800,000
had been expended, necessitating a curtailment in disposal op-
tions.
USEPA obtained emergency funding of $250,000 to dispose of
the hazardous liquid wastes and 2,600 barrels of solids material.
By Nov. 1981 all barrels of liquid wastes had been disposed of with
available funds.
CLEANUP—PHASE 3
Throughout the summer and fall of 1981, RIDEM negotiated
the complex requirements of a cooperative agreement with USEPA
to obtain Superfund monies to continue the cleanup effort. Recog-
nizing the long lead time required for funds to be awarded,
RIDEM developed the scope of work and request for proposal
(RFP) in the fall of 1981. The RFP was distributed to prospec-
tive bidders in early Dec. 1981 and a bidders conference followed
to answer technical questions.
The Scope of Work and Request for Proposal was similar to that
issued for the 1981 work, and addressed the same issues: regard-
ing site preparation, layout, safety and work effort.
c
rui
c
PCB
Soli
C.ll
i
•11 t,
C.ll
B
B*
DDT
Eoll
C.ll
C
}
c.n
K
r
PH 5
PF
^
C.ll 5
C.ll «
Tr.il!
Ub
Pik
C.ll
PuBpiog A 1
B»pllL
:] 0 1
1
/£^r~
'•quip.
•t.gluB
[ fu«i
3
CD»=
Pit 2
ffij..
Cornr)
I960
!«•»-
J
L
A
\
Figure 3.
Operational Layout of Northwest Trench Excavation.
To enable quicker turn around time on unknown waste analy-
sis, an on-site laboratory would be required. Additional ground
penetrating radar work was required in order to both identify and
confirm the south and west trench location as well as to insure that
no other trenches of buried barrels were located on the site.
Negotiations between the State and USEPA continued when an
acceptable Superfund grant proposal was submitted to USEPA in
Dec. 1981. However, despite previous consultants' reports indicat-
ing that drum removal was the most viable cost effective tech-
nique to use at this site, USEPA required the state to conduct an
additional cost effectiveness study. Its intent was to provide infor-
mation necessary for the development of engineering design and
contract documents for the excavation and disposal of hazardous
waste buried at the site.
The State retained a consulting firm to provide the study and
identify cost centers of the proposed excavation project.
In the meantime, several companies submitted bids for the clean-
up of the wastes from the west and south trenches. It was as-
sumed that given the overall size of the trenches as many as
8,500 barrels might be buried in these remaining trenches.
All bids were reviewed for technical approach and analyzed for
specific cost requirements: mobilization, administration, site prep-
aration excavation, staging and storage and bulking of wastes,
analytical capabilities, transportation and disposal. Interviews with
the top three companies were conducted to evaluate key personnel
capabilities and responses to technical questions.
Final selection of the contractor was made in Mar. of 1982 and
contract negotiations were completed in Apr.
The contractor began its work effort at the end of Apr. and by
the end of May, 3,300 drums had been excavated from the site.
-------
270
CASE HISTORIES
This in fact proved to be the extent of buried drums in these re-
maining trenches. Exploratory excavations, to confirm ground
penetrating radar were conducted, but no more drums were en-
countered.
Throughout June and until mid-July 1982 final analysis, accum-
ulation (bulking) and disposal of the excavated waste material was
continued.
BARREL HANDLING AND ANALYSIS
In both the latter two phases, the procedures for barrel re-
moval and staging were discussed with contractor and subcon-
tractor personnel and thoroughly understood prior to movement.
As a barrel was unearthed, it was transferred to a pit staging
area and examined for identifying markings or labels. Any barrels
containing such markings were held aside as potential evidence
drums.
If the barrel was leaking, it was placed in an overpack drum.
Along with non-leaking drums, the contents were examined for
physical state. Based on this preliminary examination, the barrels
were transported to a liquid storage area, solid storage area or
sludge storage area.
Leaking drums, which had been placed inside overpacks in the
pit area, were retained in overpacks until the contents could be
ascertained. The overpack was moved into a storage area. Before
sealing the drum, the pumping crew sampled the contents for trans-
fer to the laboratory.
Once analysis was completed, the drum contents were trans-
ferred to compatibility chambers. After the contents of the drum
were transferred, the empty drum was inspected for residual ma-
terials content. If a significant amount of solids remained in the
drum, it was transferred to the solids holding area. If not, it was
placed in the empty drum pile.
Acid drums were repumped into polyethylene lined drums for
storage. A sample of each acid waste was taken by the pump-
ing crew, pH ascertained and transferred to the laboratory.
Storage
All non-leaking liquid barrels extracted from the trench were
directly transferred to the liquid storage area. A three man
sampling crew opened each drum, sampled the drum and closed or
secured the drum from the weather.
Each drum was sampled with a hollow 3/4 in diameter glass tube
to allow a composite sample of the barrel. To avoid cross contam-
ination of the samples, a separate tube was used for each barrel.
The characteristics of each sample were noted on a separate
card. A number was assigned to each barrel and sample.
As each drum containing solid material was extracted from the
trench, it was transferred to a secure solids storage area. Samples
were extracted from several drums with a disposable scoop and de-
livered to the laboratory for analysis.
Non-pumpable liquids or sludges from the repumping area were
staged in a sludge holding area when extracted from the trench. A
glass tube was used for sampling if possible; a disposable scoop was
used in cases of material too viscous for the tubing. Samples were
collected, numbered in the same manner as the liquids and held
pending collection for analysis.
During the sampling procedure, all wastes (solid, sludge and
liquid) were examined and physical characteristics were noted (e.g.,
odor, fuming, color, etc).
Laboratory Analysis
Upon delivery of liquid samples to the laboratory, samples were
first examined for the previously noted physical characteristics
(odor, fuming) to insure the protection of the personnel involved.
All liquid and sludge examples were checked for pH. Those with a
pH less than 3 were considered to be acid and greater than 12,
basic. Following this test, a flammability analysis using a portable
flash point tester was conducted. By hazardous waste regulation,
if such test shows the material has a flash point of less than 140°F,
it is considered to be flammable. Following this test, water reactiv-
ity was examined by dropping a minute quantity of the sample into
distilled water.
Specific gravity was examined and samples were isolated into
three groups: (1) those with specific gravity less than 0.9, (2) those
with specific gravity ranging from 0.9 to 1.1, and (3) those with spe-
cific gravity greater than 1.1.
Following the specific gravity analysis, all samples with pH
greater than 9 were checked for the presence of cyanide. Acids
were examined for reduction/oxidation potential which measures
the ability of an acid to be a reducer (water reactive) or an oxidizer
(which when mixed with organics could cause an explosion). As
an example, perchloric acid, a strong oxidizing acid, if mixed with
grease or oil forms an explosive mixture. Physical characteristics
such as fuming or sublimation were noted on the work sheets for
those samples.
Upon receipt of solids samples in the field laboratory, samples
were examined for extraordinary physical characteristics, e.g.,
fuming, odor, etc. Following this, distilled water was utilized to dis-
solve the solid, if possible, and pH was checked. Obviously, if
water reactivity was a problem, it was noted in this phase. A por-
tion of the mixture was then examined for cyanide and flamma-
bility/combustibility capability.
Based upon the analysis for both solids and liquids, drums were
segregated into preliminary compatible groups or transferred to the
compatibility chambers (e.g., flammables, acid, caustic, etc.). This
minimized redundant drum movement until disposal was immi-
nent.
Disposal Analysis
Once the preliminary laboratory investigation was completed,
samples of compatible wastes were combined in the laboratory to
provide for disposal analysis. It was the intention to bulk as many
compatible liquids as possible for disposal purposes to enable the
most cost-effective means of transportation and disposal.
Once a significant group of compatible samples was developed
(representing 80 to 100 barrels), a PCB analysis was conducted on
subgroups (generally five drums). This proved to be the most time
consuming, costly, but necessary analytical task.
PCB contamination poses difficult and expensive disposal op-
tions whenever concentrations exceed 50 mg/1. When a five barrel
composite showed significant PCB contamination (> SOmg/1),
each uncombined sample would be examined individually for con-
tamination. As can be imagined, time was a time consuming task
which posed significant problems in turnaround time for composite
analyses.
However, once a compatible group of samples had all PCB bar-
rels identified and removed from the composite, a final disposal
analysis on the remaining barrels was conducted. Combined with
all preliminary tests, the disposal analysis consisted of those tests
shown in Table 1. In comparing the analytical requirements of a
variety of disposal facilities, these tests (Table 1) represent the
analyses necessary prior to acceptance of unknown materials for
disposal.
AIR/SAFETY HAZARD MONITORING
As a result of improper storage, an unauthorized waste dis-
posal facility can be a source of significant odors. In the Picillo
case, an ongoing air pollution incident was created as a result of
barrel excavation. Leaking or open barrels, pooled leachate,
pumping and transfer operations and the storage of wastes await-
ing disposal all contribute to the creation of objectionable odors.
The generation of odors creates problems from both a worker
and community standpoint. While worker exposure was limited as
far as possible with protective/respiratory equipment, community
complaints occurred throughout the excavation and disposal opera-
tions.
Monitoring Program
During Phase II an on-site pollutant monitoring program was de-
signed and implemented to provide an assessment of the health
-------
CASE HISTORIES
271
Table 1.
Analytical Requirements For Disposal.
1. Flammability
2. pH
3. Specific gravity
4. PCB analysis
5. Thermal content (BTU/16)
6. Physical'state at 70 °F
7. Phases (Layering in liquids)
8. Solids (%)
9. Hydrocarbon composition
10. Pesticide analysis
11. Sulfur content
12. Phenols
13. Oil and grease (%)
14. Water (%)
15. Viscosity
16. Organochlorine percentage
17. Metals analysis
a. Liquids were analyzed for soluble metals.
b. Solids were extracted according to the EPA Toxicant Extraction
Procedure (24 hr) which shows leachable metals.
c. Both liquid and solids were checked for concentrations of the
following metals:
Arsenic Mercury
Barium Nickel
Cadmium Selenium
Chromium Silver
Copper Zinc
Lead
18. Both free and total cyanide content were checked.
19. Solids were checked for solubility in water, sulfuric acid and dimethyl
sulfoxide.
hazards to personnel participating in the cleanup operations.2'3 Its
purpose was to provide an assurance that personnel protection and
fire and/or explosion protection were sufficient and that the
"clean" areas on site were indeed "clean."
The focus of this program was on-site operations. First, higher
concentrations of contaminants would be encountered which
would pose a greater threat as well as facilitating more reliable
measurements. Second, the logistics of on-site sampling were easier
to consider with respect to placement of monitors and the effects of
local meteorology and topography on air contaminant movement.
The monitoring program was designed in three phases to pro-
vide:
•Identification of the chemical hazards over a range of conditions
representative of site operations
•Assessment of the site operations, layout and. personnel pro-
tection practices to define hazard levels and specific needs for
continuing pollutant monitoring
•A continuing pollutant monitoring program focusing on person-
nel hazards
Initially a three day monitoring program was conducted in which
critical operating personnel utilizing personal monitors and several
fixed locations were examined. Samples from these personnel and
area monitors were analyzed to identify and quantify exposures.
Once identification of all exposure components was made, air
sampling using portable sample pumps and adsorbent composite
cartridges was performed. Thirty-two highly volatile contaminants
were identified. Of these, 13 chemicals were considered to have
been present in significant quantities at one or more locations on
site. The maximum concentration of any of these contaminants
never exceeded 8% of the Threshold Limit Value (TLV).
The number and variety of volatile, flammable and/or explosive
industrial solvents and petroleum products identified by the analy-
ses underscored concern for the danger of fire or explosion. Al-
though the concentrations detected did not warrant great concern,
several factors suggested that continuous explosion monitoring was
necessary:
•Ignition sources are necessarily present during site excavation
operations (metal grapplers scraping drum surface can create
sparks)
•Depth of trench, combined with calm winds, and pools of vola-
tile, flammable liquids could possibly result in conditions condu-
cive to ignition
The program concluded that, in terms of individual TLVs, there
was no significant risk to human health in the operational zones
for personnel adhering to the required safety precautions. The
concern that non-protected personnel outside the work area in the
"clean" zones would be exposed to excessive levels was not sup-
ported since measurements of contaminants in the decontamina-
tion and clean zones indicated that in no instance were TLVs ap-
pro'ached.
Nevertheless, odor complaints from area residents increased
throughout the excavation phase and into the summer months.
Presumably the complaints were based upon increasing quantities
of exposed chemicals which tended to volatilize more readily in hot
weather.
Phase III—Air Monitoring
It was determined from the results of the personnel on site mon-
itoring effort conducted in Phase II to expand the effort in Phase
HI to be as representative of on site ambient conditions as possible.
In Phase III, the excavation contractor assumed responsibilities
for on-site monitoring to determine safe working conditions and
whether adequate respiration protection was being maintained.
An effort was made to quantify population exposures by posit-
ioning an air monitoring station near private residences. Air
sampling for this effort was tailored to meet requirements for
sampling three classes of problem compounds and elements:
•Volatile organic compounds (VOCs)
•Semi-volatile organic compounds (primarily pesticides and PCBs)
•Metals (including complexed mercury)
To accomplish this, three media were used to collect the com-
pounds or elements of interest:
•Glass tubes packed with Tenax-GC sorbent were used to collect
vapor-phase VOCs.
•Hi-Vol samplers were set up with particulate filters to capture
metals and PCBs and pesticides associated with particulates.
•Backing up the particulate filters in the throat of the Hi-Vol were
two polyurethane foam (PUF) plugs used to collect vapor phase
pesticides and PCBs.
Fig. 4 is a map of the area showing locatons of the seven semi-
permanent sites. Stations Nos. 1,2, 4, and 5 are based around the
perimeter of the site, Station No. 3 is the onsite station, Station
Nos. 6 and 7 were sited in the predominantly downwind (northerly)
area, and Station No. 8 (not shown) was set at one of three loca-
tions dependent on the morning wind direction for the sampling
date.
During the monitoring the concentrations of metals, PCBs and
pesticides did not vary markedly from background. After one
month of data collection these samples were discontinued.
Discussion
The air monitoring programs of Phase II and III provided data
for both community exposure and worker safety. While odor com-
plaints persisted throughout the excavation/disposal phases, the
rural, remote location of the site provided for dispersion of con-
taminants.
In no instance did the concentrations of contaminants exceed
threshold limit values. However, no synergistic effects of the com-
bined pollutants were examined.
However, the continuous air monitoring data collected was an
average of ambient concentrations, and not the maximum exposure
levels. It is possible, therefore, that higher levels existed for short
periods.
-------
272
CASE HISTORIES
Figure 4.
Air monitoring locations for the Picillo hazardous
waste site, Coventry, Rhode Island.
GROUND AND SURFACE WATER CONTAMINATION
Since site abatement efforts began in 1978, monitoring of both
surface and groundwater has taken place. As a matter of policy it
was determined to remove the source of contamination prior to
proceeding to address the water contamination problem. Current-
ly quarterly monitoring is taking place at the well network in-
stalled at the site.
It is RIDEM's intent to conduct an engineering feasibility study
to determine if groundwater treatment can be successfully and cost
effectively applied to the contaminated plume.
OPERATIONAL PROBLEMS
Analytical Backlog
In Phase II, one of the most important areas associated with the
excavation and disposal of hazardous waste from this site involved
the time required to collect, screen and analyze the samples, to
allow for the disposal of the hazardous wastes at licensed haz-
ardous waste treatment or disposal facilities. Typically this process
took more than two weeks to complete, and backlogs for PCB
analysis increased that time up to one month. Various schemes
and methods were tried in order to reduce this turnaround time:
additional personnel, overtime, various sample compositing
schemes and providing the disposal companies with samples for
their own analysis.
In Phase III, it was determined that an on-site laboratory would
be a requirement for the project. In terms of overall work effort
this significantly decreased the time frame between sample collec-
tion analysis and disposal.
However, certain requirements should be realized at the outset:
•Laboratory personnel should be familiar with the state's (clients)
requirements and be capable of providing analysis which will
allow for disposal of the waste. Since the State is the generator of
the waste, disposal responsibility falls clearly on its shoulders.
•The sampling analytical requirements should be thoroughly
developed prior to bidding the project so the contractors will pro-
vide for all the required equipment.
Funding Logistics
A second problem area experienced by the RIDEM in regard to
the effectiveness and timeliness of the cleanup effort at the Picillo
site involved funding for the project. Funding has come from four
sources to date and the availability of monies to continue the work
effort on site had been sporadic until Superfund money was avail-
able. This presents obvious planning and scheduling problems in
regard to long range plans for a phased removal operation.
Equally important is the question of cost efficiency which is
raised when the on-site work cannot continue in a steady unin-
terrupted manner due to budget cuts and funding uncertainties.
Considerable sums of money are required to mobilize a contractor
at a remote location such as the Picillo site; earth moving equip-
ment and pumping equipment fixed costs continue whether the
equipment is operating or idle. These costs eventually are borne
by the RIDEM either directly as downtime or indirectly in in-
creased contractor rates.
Contractual Agreement
Another concern of the agency/department responsible for the
cost effective cleanup of a hazardous waste site relates to con-
tract methods and procedures. As is customary in construction en-
gineering practice, elaborate plans and specifications are pro-
duced spelling out in minute detail exactly what is to be accom-
plished and which standard methods are to be employed. The con-
tractor is given little or no latitude as to construction methods or
work schedule. Progress is easily measured and completion of the
project is easily verified.
A hazardous waste cleanup project presents a different chal-
lenge. Standard methods are not in widespread use and while some
standard laboratory procedures and safety standards are in ex-
istence, work plans must be developed for each site activity on a
case-by-case basis and what hazardous materials are encountered.
Each contractor brings his own perspective and company policy
and procedures in regard to working conditions and safety issues.
The contractor's on-site work effort can be considered to be more
in the realm of professional services as opposed to construction
contract services. Thus, contract documents should be developed
as a result of requests for proposals rather than bids. This creates
basic problems of contract administration and cost control. A more
specific scope of work would increase overall project and cost con-
trol but reduce flexibility and stifle contractor creativity in develop-
ing and implementing improved work methods and procedures.
Safety
Safety issues are of primary concern as they relate to the exca-
vation, sampling, storage, transportation and ultimate disposal of
hazardous waste materials. If an unlimited supply of money were
made available, ultimate safety procedures could be instituted.
Every drum could be handled remotely by means of mechanical
devices and robots. Continuous automated monitoring for explo-
sivity, oxygen content, organic vapor content, and other known
contaminants could be performed utilizing remote samplers and*gas
chromatography. Any automatic alarm system could be utilized
and a cessation of the work effort and site evacuation could take
place each time an alarm condition arose. But the pace would be
agonizingly slow and experience to date has shown that the vast
majority of the wastes at the site do not fall into the extremely haz-
ardous categories of explosive, shock sensitive or extremely toxic
which would require such careful handling and become excessively
costly. The materials found are primarily: highly flammable waste
solvents, acids, pesticides and PCBs.
Operating within the confines of a limited budget, safety must
be addressed and every reasonable effort must be made to afford
safe working conditions. However, in the field of hazardous waste
cleanup a risk free environment is not possible. Even if ultimate
safety measures were affordable and put into practice, risks would
be reduced but not eliminated. The safety related practices and
procedures put into effect at the Picillo site included: (1) remote
-------
CASE HISTORIES
273
handling of all drums in the excavation area; (2) constant mon-
itoring of the excavation trench area for explosivity and oxygen
content of the work area utilizing various portable meters, in-
cluding explosimeter, oxygen meter and portable organic vapor an-
alyzer, periodic monitoring of the contaminated zone, decontam-
ination zone, command post and off-site areas for airborne con-
taminants; and (3) utilization of self contained breathing apparatus
for all personnel involved in excavation and material handling.
Every person in the work area was equipped with chemically re-
sistant coveralls, rubber gloves and boots, hard hats, goggles or
other eye protection as well as a cartridge or canister type filter
respirator.
In addition to this, tests have been completed attempting to de-
fine the effectiveness of personnel protection using chemical dosi-
meters which are available to measure exposure to a number of par-
ticular chemicals as well as total organic vapor exposure. All per-
sonnel were advised to adhere to strict decontamination and were
included in a medical monitoring program.
Contaminated Soil Disposal
An additional problem area involves the sampling logistics and
costs associated with the disposal of contaminated soil generated
primarily during the excavation of the northeast trench and to a
lesser extent the excavation Of the northwest, west and south
trenches. Continuous leachate pumping unit greatly reduced the
amount of soil contaminated during the northwest, west and south
trench excavation. Much of the contaminated soil was disposed of
during Phase II and III since it was highly contaminated.
Currently RIDEM is attempting to degrade 1700 yd3 of soil con-
taminated with phenols. An ongoing landfarm/leachate program
utilizing microbiological treatment to degrade the soil is in pro-
gress. -Results of the program are unavailable to date.
Considerable quantities (approximately 3300 yd3) of soil contam-
inated with organic solvents and PCBs remain on site awaiting
final disposition through disposal, on-site fixation or treatment.
Estimates for off site transportation and disposal are approxi-
mately $500,000. USEPA is funding a research project which is
attempting to destroy the PCB content utilizing sodium polye-
thylene glycolate."
Citizen Involvement
Citizen participation, which was originally perceived as a prob-
lem, developed into an asset. The involvement of concerned cit-
izens groups in the planning process and the dissemination of in-
formation as to the progress toward project goals is an absolute
necessity when dealing with the sensitive and often emotional mat-
ters associated with hazardous wastes.
An informal community relations plan was developed including
local input into the planning process, briefings at citizen group
meetings, scheduled weekly access to the site for all interested com-
munity and press representatives as well as periodic briefings for
local (town) officials and press releases. The concerned citizens
group developed into a valuable resource with regard to develop-
ing site historical information and providing accurate lists of local
affected population in addition to communicating information re-
garding progress and site conditions to the affected population.
CONCLUSION
The cleanup of unauthorized hazardous waste sites presents var-
ied operational and environmental problems. In this paper, the
authors have attempted to provide insight into the chronology,
operational techniques, air monitoring aspects, water monitoring
and problem areas associated with the abatement of such a site
situated in Rhode Island.
Field and off-site disposal analyses proved to be a problem in
terms of turnaround time and procedural techniques. Particular
attention must be paid to the analytical requirements so that they:
(1) keep pace with excavation efforts, and (2) provide the required
disposal analyses in a timely fashion.
One of the primary operational constraints involves reducing
the risk of injury and adverse health effects to personnel work-
ing on site and the surrounding population. A key aspect of this
risk reduction strategy involved monitoring the atmosphere to
assess the degree of hazard in the following areas: explosion,
oxygen deficiency and exposure to contaminants. Much of this risk
can be reduced by the use of self-contained breathing apparatus.
But at some distance away from the contaminated zone, respira-
tory protection must be removed and the air considered to be clean.
Unanswered questions at this point include the possibility of long
term effects posed to those workers not utilizing respiratory pro-
tection (in the clean zone) as well as the effects on the local pop-
ulation.
Funding of a project of this scope must account for the signifi-
cant unknown factors associated with hazardous waste cleanup,
i.e., What is the amount and type of material to be excavated
and disposed of? Every attempt must be made to closely monitor
contractual obligations. Thorough negotiations and understanding
prior to commencement of a project will reduce problems later on.
No hazardous waste disposal activity can be risk free. There-
fore, safety precautions are paramount. Yet budget limitations are
real and require consideration of the trade-off between risk immu-
nization and performance efficiencies.
The generation of contaminated soil poses a significant disposal
problem. Successful efforts were made to limit this problem in
Phase II and III in terms of the smaller amount of generated
contaminated soil. However, at some point, a disposal or treatment
option must be chosen to deal with this problem.
Finally, and most importantly, all activities should be closely
coordinated with any citizens organization or affected group that
deals with the issues of hazardous wastes. An informed citizenry
can aid procurement of funding and maintain the necessary inter-
est in the issue that influences governmental decisions. Working
closely with such organizations promotes cleanup effects, main-
tains good press relations and may accelerate completion of the
job.
REFERENCES
1. "Hazardous Waste Investigation: Picillo Property, Coventry, Rhode
Island" The MITRE Corporation, Metreck Division, Bedford, MA,
Apr. 1980.
2. "Final Report: Pollutant At The Picillo Dumpsite Phases I, II and
III" S&D Engineering Services, Inc. East Brunswick, NY, June 1981.
3. "Environmental Monitoring Services At the Picillo Site, Coventry,
RI. Air Quality Monitoring Task," Final Report GCA Corporation
Technology Division, Bedford, MA, Aug. 1982.
4. Unpublished Research, Franklin Research Institute.
-------
A COORDINATED CLEANUP OF THE OLD HARDIN COUNTY
BRICKYARD, WEST POINT, KENTUCKY
FRED B. STROUD
U.S. Environmental Protection Agency
Atlanta, Georgia
BARRY G. BURRUS
Kentucky Department for Environmental Protection
Frankfort, Kentucky
JOHN M. GILBERT
U.S. Environmental Protection Agency
Cincinnati, Ohio
INTRODUCTION
During Feb. and Mar. of 1982, a Superfund immediate re-
moval operation occurred at an uncontrolled hazardous waste site
at West Point, Kentucky, known as the Old Hardin County Brick-
yard. The purpose of the authors is to describe in this paper the
events and activities of the operation, especially highlighting the
coordination among federal, state, and local officials involved in
the response.
The authors believe that this response was very successful and
efficient, and that it sets a good example of cooperation, coor-
dination, and communications among response personnel. By ex-
amining the response operation, we will identify those factors
which we believe are keys to a successful and well-coordinated
Superfund response action.
SITE HISTORY AND BACKGROUND
The Brickyard Site is located in northern Hardin County, Ken-
tucky, approximately 0.25 miles from the city limits of West Point,
Kentucky (Figs. 1,2 and 3). The facility was operated since the
early 1900s to manufacture bricks; however, operations ceased in
the late 1960s.
In Mar. 1977, an illegal chemical waste disposal operation at
the Brickyard Site was reported to the Kentucky Department for
Environmental Protection (KDEP). Subsequent investigations re-
vealed that approximately 3,000-55-gal drums had been placed at
the site by Mr. Donald Distler and Kentucky Liquid Recycling,
Inc., who had leased the property from the owner. The wastes were
various liquids, sludges, and solids characterized as corrosive,
toxic, highly volatile and flammable.
Following initial investigations of the site, the KDEP began en-
forcement proceedings against the operator, Mr. Distler. On Jan.
19, 1979, Mr. Distler and Kentucky Liquid Recycling, Inc. were
issued an order to abate and alleviate by the KDEP. A follow-
up order and opinion was issued on May 24, 1979. Personnel from
the USEPA, Region IV, visited the site in early 1979 during re-
sponse operations at the nearby waste site known as the "Valley
of the Drums"; however, since Kentucky enforcement activities
were underway, USEPA took no further action at the Brickyard.
During Jan. and Feb. 1980, Mr. Distler relocated many of the
drums on-site and removed several hundred non-hazardous sludge
drums to a nearby landfill. This operation soon ceased, however,
and on Apr. 5, 1980, a fire involving approximately ten of the
drums occurred at the site. The fire was extinguished by the City
of West Point Fire Department.
On June 30, 1980 a Settlement Order between KDEP and Don-
ald Distler was signed. This order set a 150 day period (expiring
Nov. 30, 1980) for final disposition of all materials on-site and site
clean-up. Mr. Distler took no action, and this 150 day period ex-
pired with hazardous conditions remaining at the site.
The State continued to pursue an enforcement solution to clean-
up, and on Feb. 24, 1981, KDEP filed a complaint in Franklin
County Circuit Court against Donald Distler and Kentucky Liquid
TO LOUISVILLE
US HHY. 31W
ILLINOIS CENTRAL
GULF RAILROAD
BR1CKVARD SITE
Figure 1.
Location Map of Brickyard Site, West Point, Ky.
Figure 2.
Overview of Brickyard Site, West Point, Ky.
274
-------
CASE HISTORIES
275
Figure 3.
Old Hardin County Brickyard, West Point, Kentucky
Recycling, Inc. No response or action occurred at the site as a re-
sult of the complaint.
As a final enforcement effort, KDEP filed a Motion for Sum-
mary Judgement in Franklin Circuit Court on Sept. 24, 1981, in
order to bring the case to final resolution. At that time, KDEP ex-
pected that the case would be tied up in the courts for an extended
period of time. (In fact, the Motion for Summary Judgement has
still not been heard as of this writing.)
Environmental Health Concerns
In late 1981, with no enforcement solution in sight, conditions
at the Brickyard continued to worsen. Many of the over 2,000
drums were leaking due to deterioration and vandalism at the site.
The effects of the release of chemicals into the environment was
readily apparent in the vicinity of the drums. Spillage had killed
grass, trees and birds. In addition, irritating odors were immed-
iately apparent within 0.25 mile of the drums. Material leaking
from the drums had entered the ditch adjacent to the railroad
tracks and had entered a tributary to the Salt River.
The Illinois Central Gulf railroad tracks run directly through the
site. Complaints of dizziness due to odors had been received from
railroad personnel when rail cars had been stopped for 15-20 min,
while waiting for other trains to pass.
The entire population (2,000) of West Point was within a one-
mile radius. A computer air model indicated that a four to six mile
radius would have to be evacuated if a fire occurred. Since the
drums were located in a valley, the air model predicted the fumes
would linger in the area intensifying the problem.
There was also concern that the spillage from the drums had en-
tered the groundwater. The water supply for Fort Knox Military
Base came from a groundwater well field located 0.5 mile from the
site. Also, the nearby town of West Point used groundwater as the
major drinking water supply.
Because of these environmental health concerns, KDEP and
EPA agreed to pursue immediate removal action under Supre-
fund.
Request for Immediate Removal
On Jan. 4, 1982, USEPA Region IV received the State of Ken-
tucky's written request for an immediate removal action at the
"Brickyard". A copy of the letter was delivered to the USEPA
attorney assigned to the site. Demand letters were prepared for
the four parties that might be held responsible for the clean-up.
The OSC also requested the USEPA Environmental Response
Team (ERT) to conduct an air monitoring survey on the site to:
(1) provide documentation of air hazards, (2) determine material
present on site and concentration of material, (3) establish levels of
protection needed to conduct emergency cleanup, and (4) estab-
lish an air monitoring program to be conducted during clean-up
activities.
On Jan. 20-21, 1982, the ERT conducted a total vapor/gas air
monitoring survey. Based on results of the initial air survey, an air
monitoring scheme was developed and the levels of protection
proposed for the immediate removal action.
In the meantime, the USEPA On-Scene Coordinator (OSC) pre-
pared the required ten-point document necessary to effectuate an
immediate removal action. An estimated cost of $300,000 was pro-
jected for the removal of all liquids, sludges, contaminated soil,
and drums at the site.
In response to the USEPA demand letters, the four potential
responsible parties decided that they would rather see USEPA do
the cleanup. On Feb. 16, 1982, Mr. Distler, the primary respon-
sible party, began serving a two year prison sentence for another
illegal hazardous waste incident that occurred in Louisville in 1977.
The request for immediate removal action was submitted to
USEPA Headquarters by the OSC. On Feb. 22, 1982, approval to
commence Superfund immediate removal action at the Brickyard
was received.
RESPONSE PLANNING
Upon approval of the immediate removal request, additional re-
sponse planning commenced. Two important activities that
occurred were selection of contractors and community relations.
Selection of Contractors
Two main contractors were selected for the removal action.
One of the contractors (Resource Recycling Technology) was se-
lected on the basis of analytical support and the capability to dis-
pose of the liquids at the site. The other contractor (CMC, Inc.)
was hired to provide heavy equipment for handling of the drums,
liquids, and solids. In addition, it was planned for the OSC to ar-
range the contract for the disposal of any solid hazardous wastes
generated at the site. These wastes would be disposed of at an ap-
proved hazardous waste facility. Both contractors were signed to
"Superfund cleanup contracts" prior to commencement of re-
moval activities on March 1,1982.
Media and Community Relations
The media and community relations operations were coordi-
nated through the KDEP Office of Communications and Com-
munity Affairs (OCCA) and USEPA Region IV Superfund Public
Relations personnel. Upon approval of Superfund Money, the
OCCA was contacted to provide support in the community rela-
tions area. OCCA contacted by telephone the Mayor of West
Point, the County Judge, and the State Representative to inform
them of the upcoming operations. A press conference was sched-
uled for the morning of the first day of activities at the site, and
OCCA made all necessary arrangements, including the appearance
of the highest ranking State Environmental official, Secretary
Swigart of KDEP. Finally, a press release was prepared for review
by USEPA Public Relations at the Regional Response Team meet-
ing prior to the beginning of operations.
THE REGIONAL RESPONSE TEAM MEETING
Of all the many factors that contributed to the successful re-
sponse at the Brickyard Site, perhaps the most important was the
Regional Response Team Meeting that occurred on the day before
the actual on-site operations began. This meeting brought together
for the first time all of the main organizations and personnel to be
involved in the response.
The Response Team Meeting was held on the afternoon of Feb.
28, 1982. Present at this meeting were the following personnel and
organizations:
•USEPA On-Scene Coordinator (Stroud)
•USEPA Environmental Response Team Member (Gilbert)
•KY Superfund Coordinator (Burrus)
•KY Disaster and Emergency Services Personnel (DES)
-------
276
CASE HISTORIES
•Prime Clean-up Contractor (CMC, Inc.)
•Prime Sampling/Analysis/Lab Contractor (Resource Recycling
Technology)
•KY Office of Communications and Community Affairs
•USEP A Superfund Public Relations
The USEPA OSC conducted the meeting. At the start, a brief
background of the site was covered and a basic scope of the work
to be accomplished was presented. This gave everyone a common
base and frame of reference for the site activity. There was also
the opportunity for each group involved to get to know others on
the team and get a feel for who was going to be on the team and
what they would do. This was very important prior to commence-
ment of activities and helped establish a working environment of
trust and coordination from the beginning.
After the general introduction at the response team meeting, spe-
cific planning occurred on three basic concerns: safety, security,
and public relations. The Site Safety Plan was reviewed and further
developed. Responsibility for implementation of the Safety Plan
was assigned to the USEPA Environmental Response Team Mem-
ber (Gilbert). Site Security details were also developed and as-
signed to the KY DES with the responsibility to use local law en-
forcement personnel. Finally, the draft of the press release was re-
viewed by all, corrected, and returned to the KY OCCA contact
at the meeting for release. When all these items, together with some
final details for the first day, were completed, an excellent plan had
been developed and shared. This meeting allowed the Regional Re-
sponse Team to "hit the ground running" when operations com-
menced on the next morning.
KY Superfund Coordinator
KY Disaster C Emergency
Services
US EPA
On-Scene Coordinator
—— US EPA Environmental
Response Teem Advisor
^^— Technical Assistance
Team Members (TAT)
KY Office of Communications-
( Community Affairs
Local Government
Contacts
I I
Health Police Fire
Dept. Dept
STATE SUPPORT GROUPS
US EPA Public Ralatlons
Crime Clean-up
Contractor
(CMC. Inc.)
Subcontractors
I
Prime Sampling/Analyses
Contractor (Resource
Recycling Technology)
Subcontractors
KV Envlrorwnantal Response Teem
KV Division of Waste Management
Fire Marshall's Office
KV Dept. of Health
FEDERAL SUPPORT GROUPS
US EPA Environmental Response Te
Center for Disease Control
US Army Explosive Ordnance
Detachment
Figure 4.
Response Organization for Old Hardin County Brickyard
RESPONSE ORGANIZATION
A key factor to a well coordinated Superfund Response is the
proper organization and inclusion of the people necessary to do the
job (Fig. 4). During the planning phase, the OSC began assembling
these necessary individuals and groups. At the Regional Response
Team Meeting, the basic organization crystallized so that we were
ready to begin the next day. Other people and organizations were
utilized during the response, and by having the basic organiza-
tion established at the Response Team Meeting, these new people
were able to quickly assimilate into the appropriate organization
position.
One of the more important aspects of this response organiza-
tion was the good integration of the State and Local personnel in-
to the overall response organization. While USEPA did, of course,
maintain the lead for the overall project, the State was fully in-
volved in the response activities and coordination. This State and
local involvement, encouraged by USEPA, was an important fac-
tor in the success of the response and enhanced effective com-
munications during the response.
Figure 5.
Activities View of Brickyard Entrance
Figure 6.
Drum Grappler and Compatibility Chamber
RESPONSE ACTIVITIES
The immediate removal action at the Brickyard began on Mar.
1, 1982, and was completed on Mar. 27, 1982. The following
sections describe some of the main activities that occurred during
the response.
Initial On-Site Activities
The contractors arrived at the site on the morning of Mar. 1,
1982, and began staging their equipment and personnel by 7:00
a.m. The hotline was established and security precautions were put
into effect. To support field activities, a command post, decontam-
ination trailer, mobile laboratory, and USEPA Region IV's van
for air monitoring were positioned according to the plan (Fig. 5).
Other support items for field activities included electricity, tele-
phones, copy machine, "port-a-lets", typewriter, and a telecopier.
A press conference was also held at the site on the morning of
Mar. 1, 1982. The OSC, Ms. Jackie Swigart, Secretary, Natural
Resources and Environmental Protection Cabinet, and West Point
Mayor Gene Smith outlined the removal action.
Personnel Safety
Based on the initial air monitoring survey, personnel working
on site were required to wear USEPA's Level C protection which
included an approved air purifying mask with an organic vapor
cartridge. Tyvek/Saran hooded disposable coveralls were required
for heavy equipment operators, monitoring and surveillance activi-
ties, and observers. A safety boundary, or "hotline," was estab-
lished and the Level C protective equipment had to be removed
prior to exiting the "hot area".
When sampling of the drums began, Resource Recycling's per-
sonnel were using a bung wrench to open drums. The third drum
they attempted to sample had a pressure build-up, and the person
opening the drum was sprayed with liquid. Fortunately, he was
protected by his protective gear and was not hurt. In order to
make the drum sampling operation safer, the remaining drums
were opened utilizing a punch on the bucket of a backhoe. When
punching the drums, the operator of the backhoe was in an en-
closed cab with a fresh air supply. No personnel were allowed in
-------
CASE HISTORIES
277
the vicinity of the drums when they were being punched. This pro-
cedure worked well and there were no further incidents.
In order to maximize worker safety, a drum grappler was util-
ized to relocate the drums. The grappler provided a safe working
environment for the operator (enclosed cab with fresh air), and no
personnel were required to handle the drums. The grappler was also
used to empty the drums containing liquid wastes into the com-
patibility chamber (Fig. 6).
Sampling Program
The contractor chosen to sample the drums and determine com-
patibility developed the sampling program for conducting this por-
tion of the activities. The analytical scheme developed by Re-
source Recycling Technology that was utilized to segregate the
drummed material into categories for disposal is shown in Fig. 7.
PH
r7 pH*
ID
WATER ADDITION
SOLUBLE INSOLUBLE
FLOATS SINKS (MORE-DENSE)
(LESS DENSE)
7 pH>7 5% HCI
BASE SOL INS<
BASE
5%
NaOH
' (ORGANIC BASE) SOL
BEILSTEIN
TEST
3L PCS NEC
CHLORINATED FLAM
HC HC
(SV)
1NSOL
SWIRL
WATER
CHECK FP
WATER SOL
FLAM LIQ (SV)
CODES USED FOR MARKING DRUMS:
(ORGANIC
ACID)
FLAM HC
(SV)
ACID (pH = )
BASE CN FREE (pH = )
BASE DANGER CN
WATER (pH = )
WATER SOL SV
ORGANIC BASE
7. FLAMMABLE LIQUID (5V)
8. ORGANIC ACID
9. CHLORINATED HC
10. SOLID (MARK ON PART OF DRUM)
11. OILS
NOTE: OILS WILL BE SEGREGATED AND CHECKED FOR PCB'S INDEPENDENTLY. EACH
SOLVENT COMPOSITE (100 DRUM LOTS) WILL BE CHECKED FOR PCB'S.
Figure 7.
Analytical Scheme for Sampling/Categorization
Air Monitoring
Air monitoring surveys were conducted utilizing the Century Or-
ganic Vapor Analyzer (OVA) twice a day in order to determine
what impact cleanup operations had on the ambient air conditions.
In addition to the gross OVA readings, air samples were collected
utilizing personnel sampling pumps and Tenax collection tubes
(Fig. 8). The samples were collected over a two to three hour
period. The tubes were desorbed and analyzed utilizing the OVA.
These samples provided a time weighted average concentration and
an estimate of how many compounds were in the air. All air mon-
itoring was conducted to justify the level of protection being worn
by on-site personnel or to indicate that a change in the level of pro-
tection was warranted.
The air sampling indicated that during the time that liquids were
being poured into the compatibility chamber and the solids were
being dumped that the concentration of the organics in the air went
up substantially in the areas of dumping. However, the ambient
conditions on-site were not affected by the dumping operations.
Figure 8.
Collection of Air Samples
Disposal Options
The most cost-effective and environmentally sound method of
disposal was to be used for each type of waste that was generated
at the site. Once the drums had been staged into solids and liquids,
they were sampled and marked according to the category code and
description (Fig. 7). This eventually led to eight different disposal
options.
Everything that was considered contaminated water was trans-
ported to the analytical contractor's waste water treatment plant
for disposal. Liquids that were classified reclaimable solvents were
dumped into the compatibility chamber provided by the contrac-
tor, and a suitable reclaimer subcontracted to remove these
liquids. Other organic liquids of no economic use were incinerated
at an approved facility within the State of Kentucky. Empty drums
and contaminated soil on the site that were deemed nonhazardous
by the State were disposed at a local landfill.
The remaining sludge and solids were stabilized with fly ash and
sand to produce a "mixed solids" waste pile, and these mixed sol-
ids were disposed at an approved hazardous waste disposal facility.
Two drums of lab packs and three drums of oil containing PCB's
were redrumed on-site and properly disposed by a subcontractor at
approved hazardous waste facilities.
COMMUNITY RELATIONS DURING RESPONSE
The community relations were excellent during the response. The
Mayor and Police Chief of West Point were kept informed fre-
quently of the progress being made. Since the police were pro-
viding nightly-security at the site, there was involvement of the
local community in site activities which helped promote good com-
munity relations.
During the second week of activities, a new press release was
developed by KY OCCA, checked with on-scene personnel, and is-
sued as a status report to the press. Additional media inquiries were
answered promptly and effectively by the OSC.
At the end of the response, the Mayor of West Point was so
pleased with the cleanup, that he contacted several media and press
personnel and issued very positive statements about the operations.
This provided a good, positive final touch to the community rela-
tions aspects of the response at the Brickyard.
PROBLEMS ENCOUNTERED
Several problems were encountered during the response. Due to
continuing effective coordination among members of the Response
Team, all these problems were successfully resolved. The problems
are described in the following sections.
-------
271
CASE HISTORIES
ID order to operate the mobile laboratory and the command
port, temporary power service was required. The local power com-
pany was contacted, but they were not accustomed to dealing with
emergency hook-ups, and the service was delayed. The Kentucky
Department of Emergency Services (DBS) contacted the power
company and asked them to expedite the power hook-up due to the
emergency. Once DES contacted the power company, the power
was installed the following day.
During the liquids dumping operation, foggy conditions caused
a heavy vapor build-up on the evening of Mar. 11, 1982. This
condition began after the close of business that day. Around 10:00
p.m. the OSC received a call from the West Point Police report-
ing that many town residents were complaining about the vapors
emitting from the Brickyard.
Two Response Team members returned to the site to conduct an
air survey. By the time they arrived, the police chief reported that
the odors were not as bad as before. The air surveyed indicated
higher than normal organic concentrations around the compatibil-
ity chamber, the sludge pile, and an area of empty drums. To re-
duce the vapors emitting from these areas, they were covered with
plastic.
Since the Police Chief was aware of the activities being con-
ducted at the "Brickyard," he was ready to handle problems like
this. When people started complaining, he advised them to stay in-
doors and shut all windows. He also contacted the OSC to de-
termine the extent of the problem. This cooperation was a result of
the coordinated planning process prior to work being initiated.
&ptorive and Radioactive Materials
When the contractors encountered drums that were marked ex-
plosive and radioactive, experts were called to help assist defin-
ing the problems. An Explosive Ordnance Disposal Team (EOD)
from the U.S. Army at Fort Knox was brought in to examine
the material marked explosive. After examining the material the
EOD determined that the material did not represent an explosion
problem and the material was disposed of as sludge.
Two drums, were encountered that had markings indicating the
possibility of radioactive material. A complete radioactive survey
was conducted over the site by the Kentucky Department of Health
Radiological Survey Team and the Jefferson County Department
of Health. Neither group found any radioactive material.
Disposal
The major problem encountered during the entire response oper-
ation concerned the disposal of the mixed solids and sludge gen-
erated at the site. The planned strategy on h^nrfljng solids and
sludge was to composite it in a large pile on the concrete pad at the
site and mix with fly ash (Fig. 9). Because of the nature of the
wastes at the site, the plans were to declare this waste as hazardous
and ship it in bulk to a hazardous secure landfill. Several con-
tacts were made and finally one disposal contractor appeared to be
the most cost effective while providing the quickest response; there-
fore, arrangements were made to begin shipment of this mixed sol-
id waste, totaling about 400 tons, to the facility.
Major problems always seem to happen at 4:30 p.m. on Fridays,
and that is exactly when this one developed, at the end of the
second week of operation! That is when word was received that the
State Agency responsible for regulating this out-of-state facility
had questions about accepting this waste and needed more infor-
mation. Because of the time of day, clarification of the problem
could not be obtained; it was Monday before the responsible State
Agency official could be contacted.
During that weekend, USEPA began ^m"'"'"g other disposal
options, including disposal in Kentucky by classifying the waste as
non-hazardous. Testing of samples from the sludge pile had in-
deed shown that the waste was not hazardous by any of the four
Figure 9.
Mixed Solids Pile
RCRA characteristics (i.e., ignitability, corrosivity, reactivity, or
EP toxicity) and did not contain PCBs. Kentucky State person-
nel, on the other hand, held the opinion that the waste should be
declared hazardous under the mixture rule since 12 compounds
from those RCRA listed wastes numbers F001, F002, F003 and
POOS were detected at significant levels in the samples of the mixed
solids. Thus, on-site personnel reached an impasse as to the haz-
ardous/non-hazardous waste issue on the mixed solids, and it was
necessary to get further clarification from EPA personnel.
At USEPA's request, the Kentucky coordinator submitted a let-
ter on the following Tuesday giyng justification for classifying the
waste as hazardous. This letter was forwarded immediately to
USEPA Region IV RCRA personnel and later checked out with
USEPA Headquarters in Washington, D.C.
•- • ^K t »
>*••
Figure 10.
Response Completed at Drum Storage Area
In the meantime, all other operations had been completed ex-
cept the mixed solids disposal, so operations at the site were com-
pleted and all personnel left the site. The prime cleanup contrac-
tor was on stand-by as a decision was awaited on the disposal of the
mixed solids.
Two days later, USEPA Region IV RCRA personnel ruled that
the waste would be classified as hazardous since it contained sludge
that had been in contact with hazardous liquids that had been
shipped under USEPA hazard numbers F003-F005; therefore, be-
cause of the mixture rule, the entire 400 tons of mixed solids was
classified as a hazardous waste. This ruling hastened the negotia-
tions with the potential disposal contractor and the other State
Agency. An additional clarification letter on the waste was sent by
the Kentucky Coordinator to the other State Agency requesting dis-
posal as a hazardous waste.
-------
CASE HISTORIES
279
As several days passed without hearing further word from the
other State Agency, a new disposal alternative was presented by an-
other disposal contractor for a facility in a different state. On
further checking by the USEPA OSC, this new alternative ap-
peared more cost-effective and offered rapid response; therefore,
in conjunction with Kentucky personnel, the decision was made to
go with the new disposal option. A contract was signed the follow-
ing day, and removal operations for the mixed solids began. In
three days, the operation and total response was finished (Fig. 10).
Two important lessons can be learned from this problem and its
solution. First, even though a major difference of opinion occurred
between on-scene USEPA and State personnel over whether the
waste was hazardous, coordination had been developed success-
fully enough by that time so that the problem was effectively re-
solved and completed. The second lesson is that it is never too late
to implement a better solution!
REMEDIAL INVESTIGATION
With the completion of the immediate removal response, atten-
tion was focused on potential remedial action at the Brickyard
Site. There were several local reports of buried drums of waste at or
near the site; however, follow-up magnetometer inspections and in-
terviews have shown no presence of buried wastes. The primary
remedial concern, therefore, has been potential groundwater con-
tamination.
During Mar. 22-26, 1982, The Field Investigation Team (FIT)
of USEPA Region IV conducted resistivity surveys of the Brick-
yard Site in order to approximate the horizontal and vertical ex-
tent of any groundwater contamination.' Results of this testing in-
dicate that there is some contamination of groundwater below the
site, but that the contaminated plume has not moved off-site.
As a result of this initial testing, a program has been initiated to
install 10 groundwater monitoring wells at the site to better define
the groundwater conditions. This work is in progress and should
be completed in the Spring, 1983.
CONCLUSIONS
At the completion of this response, those who had been asso-
ciated with it realized that it was one of the most effective they
had seen. It was natural, then, to ask the question—"Why did
this response go so well?" The general answer is that this response
was coordinated well. In other words, people, equipment, and ac-
tivities were managed in such a way that they were effectively in-
tegrated to get the desired results.
Some of the key factors that we believe led to this successful and
well-coordinated Superfund Response are the following:
•Effective communications and working relationships between the
primary USEPA person (OSC) and the primary state person
(Superfund Coordinator) prior to and during the response
•Good response planning prior to the response, especially in the
areas of contractor selection and community relations
•The Regional Response Team meetings
•The integration of the State and Local team members into ac-
tive involvement and coordination of the response
•Utilization of proper equipment and techniques for handling
and disposing of hazardous wastes
These factors, together with a spirit of cooperation and commit-
ment, led to the successful cleanup of the Brickyard.
REFERENCE
1. Harman, H.D., Jr. and Hitchcock, S., "Cost Effective Prelim-
inary Leachate Monitoring at an Uncontrolled Hazardous Waste Site",
Proc. U.S. EPA 3rd National Conference on Management of Uncon-
trolled Hazardous Waste Sites, Nov. 29-Dec. 1, 1982, Washington,
D.C. Hazardous Materials Control Research Institute, Silver Spring,
MD.
-------
SILRESIM: A HAZARDOUS WASTE CASE STUDY
JOHN D. TEWHEY, Ph.D.
JOSH E. SEVEE
RICHARD L. FORTIN
Perkins Jordan, Inc.
Reading, Massachusetts
INTRODUCTION
The Silresim Chemical Corporation's chemical waste recla-
mation facility in Lowell, Massachusetts was abandoned in Jan.
1978; approximately one million gallons of hazardous material
were left behind in drum and bulk storage. The five-acre recla-
mation facility, established in 1971, had been accepting approx-
imately three million gallons of oil wastes, solvents, chemical pro-
cess wastes, plating wastes, heavy metal containing sludges and
other materials yearly. The facility was designed and licensed for
the ultimate disposal or recycle of these chemical wastes. Site in-
vestigations conducted in 1977 revealed license violations; the li-
cense was revoked when Silresim declared bankruptcy later that
year.'
The Commonwealth of Massachusetts, Department of Environ-
mental Quality Engineering (DEQE) initiated efforts to clean up
the site and by Sept. 1981 all stored materials had been removed.
In Oct. 1981, Perkins Jordan began a two-part study to char-
acterize the nature and extent of soils and groundwater contam-
ination caused by the hazardous materials and to recommend ac-
tions to remedy the contamination problem.
HYDROGEOLOGIC INVESTIGATION
The site is located at the edge of an industrial area south of
Lowell's central business district (Fig. 1). The Lowell Connector,
Boston and Maine Railroad tracks, River Meadow Brook, and
several residential areas are all close to the site. To assess the ex-
tent of surface and subsurface contamination, investigations of the
surficial soil, surface water, subsurface soil, and groundwater were
conducted at both on-site and off-site locations.
Twelve backhoe dug test pits—eight located in on-site, high-
use areas and four in adjacent surface runoff areas—were sampled
for shallow soils analysis (Fig. 2). Five borings were installed for
deep soils exploration; one was in the center of the site and four
at various locatons around the perimeter of the site (Fig. 2).
A deep monitoring well was installed at each boring location
and shallow wells were positioned at four borings. The monitoring
wells served two purposes. The physical characteristics of the
groundwater regimes were determined by means of in-situ perme-
ability measurements and groundwater level measurements. The
level and extent of chemical contamination in groundwater was
determined by means of gas chromatography/mass spectroscopy
(GC/MS) analyses of organic constituents.
Surficial soil samples were collected in or near the closer resi-
dential areas to assess the extent of airborne contaminant migra-
tion. Surface and groundwater samples were obtained from sur-
rounding areas including the River Meadow Brook. A subsurface
metal detector survey was also performed at the site to identify any
underground storage tanks.
RESULTS
The visual effects of contaminants in surface runoff were evi-
dent at the site itself: vegetation was dead or nonexistent and the
soil was discolored and emitted an odor. Laboratory analyses of
the surface soil, shallow subsurface soil, and groundwater samples
collected during the study indicated that approximately 6,000 gal of
volatile organic compounds were in the soil and groundwater be-
neath the Silresim site.
The zone of maximum soil contamination (1,000,000 ppb of
volatile organic substances) is limited to on-site high use areas in
the central portion of the site and along the northern perimeter.
Maximum soils contamination levels were found at 10 ft or less be-
low the surface (Fig. 3). The zone of maximum groundwater con-
tamination levels occurs at depths of 20 ft or less (Fig. 4). The
rate of horizontal groundwater flow beneath the Silresim site is
approximately 16 ft/yr. The direction of groundwater flow is to-
ward the north.
Figure 1.
The five-acre Silresim Site is located approximately 500 ft east of River
Meadow Brook and the Lowell Connector in Lowell, Massachusetts.
The contours represent approximate groundwater elevations in the vicinity
of the site. Scale bar is approximately 1000 ft.
280
-------
CASE HISTORIES 281
Volatile Organic Level
In ppm
>50
30-50
Q 10-30
n 1-10
Figure 2.
The heavy dashed line indicates the borders of the five-acre Silresim Site. Explorations are indicated by squares (test pits) and circles (borings with monitoring
wells). Contours represent the distribution of volatile organic substances in surface soils as determined by a photoionization meter survey. The 50 ft sampling
grid is superimposed on the figure.
SLRESM CHEMICAL CORPORATION STTE
Level of Contamination
(ppb)
> 1,000,000
100,000-1,000,000
10,000-100,000
1,000-10,000
Figures.
Contours represent the vertical distribution of volatile organic materials in soils at the Silresim Site. A 10 ft thick silt stratum is located at a depth of 15 ft
Scale bar is 100 ft.
-------
282
CASE HISTORIES
SLRESM CHBJCAL CORPORATION STTE
Level of Contamination
(ppb)
> 100,000
10,000-100,000
1,000-10,000
100-1,000
D 1-100
Figure 4.
Contours represent the vertical distribution of volatile organic materials in groundwater at the Silresim Site. Scale bar is 100 ft.
Level of Contamination
Mg/l (ppb)
• > 100.000
• 10.000-100.000
• 1.000-10.000
• 100-1.000
• 1-100
Figures.
The contaminant plume that presently exists in groundwater at the site (left) was determined by means of the hydrogeologic study. The migration of the c
taminant plume in groundwater as a function of time (right) was determined by means of predictive modeling of the groundwater regime. The illustration on
the right depicts the estimated extent of the plume in the year 2006 if no remedial action is taken. Three potential receptors are identified in the figure.
-------
CASE HISTORIES
283
CROSS SECTION
PLAN VIEW
CROSS SECTION
CLAY CAP
CLAY CAP AND SLURRY WALL
CROSS SECTION
PLAN VIEW
CROSS SECTION
GROUNDWATEH REMOVAL
SOIL EXCAVATION
Figure 6.
Each of the four remedial actions is shown at 70% level of implementa-
tion (i.e., 70% of subsurface contaminants are contained or removed).
Cross-section views depict levels of contamination in soils as depicted in
Fig. 3. Contours in plan view represent contamination levels in soils that
exist approximately 5 ft below the surface.
The horizontal contamination plots for soils and groundwater
indicate that subsurface contamination exists beyond the bound-
aries of the site to the north, in the direction of surface runoff
and groundwater flow. Groundwater samples obtained from up-
gradient monitoring wells located south of the site show little or
no evidence of contamination.
Most of the chemical contamination occurs in the shallow soils
and groundwater: about 8% of the waste volume is dissolved in
the groundwater while 92% is held in the soils. Contamination
levels in soil and groundwater decrease as distance from the site in-
creases.
Dilution of chemical contaminants in groundwater occurs by
means of dispersion through groundwater flow and molecular
diffusion. Unless the contaminants are removed from the subsur-
face, or their movement is inhibited, they will continue to migrate
away from the site, either in the form of air emissions from the soil
or through groundwater flow.
LONG-TERM EFFECTS AND REMEDIAL ACTION
How will contamination affect the areas surrounding the site 10,
15, or 50 years from now? The second phase of Perkins Jordan's
study was directed toward determining these long term effects and
recommending methods of inhibiting contaminant migration.
Mathematical models were used to predict future migration
through air emissions and groundwater flow and discharge (Fig.
5).2'3' The models were applied to three potential receptors of the
contaminants: 1) the Lowell Iron and Steel plant, the closest occu-
pied structure to the site in the path of groundwater flow, 2) the
Robinson Street area, the only residential area within the antici-
pated bounds of the contaminated groundwater flow, and 3) the
River Meadow Brook, which is the ultimate discharge area for
groundwater flow from the Silresim site.
Five remedial conditions were considered: l)"no action", 2) clay
capping, 3) clay cap and slurry wall, 4) groundwater removal and
treatment, and 5) soil excavation and removal (Fig. 6). The ground-
water and air transport models were used to evaluate each alterna-
tive remedial action in terms of the contamination levels that could
be expected to reach the three receptors over the next century.
If no corrective action is taken at the site, the first evidence
of contaminated groundwater (1 ug/ml) is estimated to reach the
three receptors in five, 20, and 27 years, respectively. Maximum
air contaminant levels (8 ug/ml) are estimated to arrive in 25, 75,
and 90 years. These levels neither exceed OSHA air quality
standards nor USEPA drinking water criteria. In addition, there
are no known users of the groundwater along the expected path-
way of the contamination plume. Based on these factors, no remed-
ial action is considered to be warranted to specifically reduce con-
tamination via groundwater flow.
However, air emissions emanating from the contaminated soils
and groundwater are a concern. The ability of the four remedial
actions to reduce air emissions was evaluated based on the con-
centration levels and amount of exposure expected at each of the
three designated receptors. The cost of construction and/or opera-
tion of the four alternative remedial actions at three levels of im-
plementation (20, 70, and 90%) and the duration of contaminant
levels at the site were also evaluated in order to assess the effec-
tiveness of the remedial schemes.
The clay cap and slurry wall, groundwater removal and treat-
ment, and soil excavation and removal options each involve ex-
cavation which will aggravate air emissions. Clay capping alone
was the most cost-effective, positive control option evaluated and
was recommended for implementation in order to reduce the ef-
fects of contamination at the Silresim site. Air and groundwater
monitoring were recommended in order to verify the findings ob-
tained from air emissions and groundwater modeling.
REFERENCES
1. MITRE Corp., "Hazardous Waste Cleanup: Silresin Site in Lowell,
Massachusetts", Report MTR 79W00204, 1979, 39 p.
2. Shen, T.T., "Estimating Hazardous Air Emissions from Disposal
Sites", Pollution Engineering, 13, Aug. 1981, 31-34.
3. Freeze, R.A. and Cherry, J.A., Groundwater, Prentice Hall, Engle-
wood Cliffs, N. J., 1979, 604 p.
4. Konikow, L.F. and Bredehoeft, J.D., "Computer Model of Two-
Dimensional Solute Transport and Dispersion in Groundwater",
Techniques of Water Resources Investigations of the U.S.G.-S., Book
7, Chapter C2, 1978, 39 p.
-------
CLEANUP AND CONTAINMENT OF PCBs—A SUCCESS STORY
BRIAN D. BRACKEN
HILARY M. THEISEN
Brown and Caldwell
Walnut Creek, California
INTRODUCTION
The General Electric Company owns a 24-acre site in Oakland,
California that serves as a location for an apparatus service shop
and a short-term storage facility for various company products.
From 1924 through 1975 the site was used for the manufacture of
transformers, motors, and switchgear and for some maintenance
and repair work.
Polychlorinated biphenyls (PCBs) were used as a dielectric in
the manufacture of transformers at the site from 1940 to 1968.
Pyranol-filled (a blend of approximately 50% trichlorobenzene and
50% Aroclor 1260, a PCB) transformers were serviced at the
Oakland facility until 1975.
During the manufacturing process, Pyranol was pumped to
tanks, tested, stored, reprocessed by filtering, and transported by
vehicle and pipes. Accidental leaks and spills occurred resulting
in PCB accumulation in surface and subsurface soils, and detec-
table levels of PCBs were found in shallow soils in 1979.
Shortly thereafter the State of California requested General
Electric to submit a plan of correction. General Electric retained
Brown and Caldwell to conduct a preliminary site investigation to
determine the extent of the problem. General Electric closely co-
ordinated initial project activity with the regional USEPA office,
the California Department of Health Services (DOHS), and the
Regional Water Quality Control Board (RWQCB). Eventually,
the DOHS was designated lead agency in the project, simplifying
subsequent coordination for General Electric.
The preliminary site investigation was expanded into another
phase, and a conceptual design for the immediate correction of the
problem was developed by Brown and Caldwell. After the ex-
panded preliminary site investigation was completed, the RWQCB,
in conjunction with the DOHS and USEPA, issued a cleanup and
abatement order requiring: (1) a Phase II study providing historical
and current information on plant operations, geohydrologic infor-
mation on the site, and additional data on the extent of site con-
tamination, and (2) development and implementation of an immed-
iate correction plan for control and removal of subsurface oil and
prevention and containment of storm water runoff.
The Phase II study was completed in June 1981 concurrently
with plans and specifications for construction of "an immediate
correction project." The project was bid in Aug. 1981. Con-
struction began the same month and was essentially complete in
Dec. 1981. The immediate correction program, now operating
successfully, includes an underground oil and groundwater collec-
tion system, an extraction sump, a treatment system for removal of
PCBs, and an extensive site sealing and drainage system.
The oily groundwater treatment system was started up in Dec.
and began discharging cleaned effluent to the local utility district
sanitary sewer system in Jan. 1982. Long-term monitoring of the
treatment system, the surface sealing system, and the shallow
groundwater system has continued since startup.
In this paper, the authors describe the methodology employed
in meeting the requirements of the regulatory agencies and the Gen-
eral Electric Company; specific aspects of the project approach
and correction plan that may have application to other sites are also
discussed. Unique components of the corrective plan are described
that contributed to the project receiving first place in the 1982 Cal-
ifornia Competition of the Consulting Engineers Association of
California, an honor award from the American Consulting Engi-
neers Council, and a certificate of merit from the California Water
Pollution Control Association.
Significant features of the project are: (1) the extent and na-
ture of PCB concentration at tlje site were determined, (2) PCBs
were found in subsurface oil and absorbed to soil particles but
were not found in the shallow groundwater, (3) the potential of
PCB migration in shallow soils to deeper confined aquifers was
determined to be insignificant, (4) PCBs are being effectively re-
moved from oil groundwater extracted from the site, (5) PCB
loadings leaving the site through storm water runoff are being kept
below acceptable limits, and (6) the direction of shallow ground-
water movement has been effectively altered to reduce the poten-
tial for PCB migration off site.
SITE INVESTIGATION
The site investigation or problem definition activity for the Oak-
land Site Project was conducted in three phases: Phase I, Phase
Ib, and Phase II. The Phase I investigation was performed in
response to an order dated Nov. 29, 1979, from the California
DOHS. Work involved a brief review of existing site characteristics
and plant operations, a field investigation including soil borings
and monitor well construction to obtain .data on PCB distrib-
ution in the soil and the quality and movement of shallow ground-
water, and development and evaluation of alternative correction
programs.
The Phase Ib investigation included collection of soil boring and
surfac" grab samples in an unsampled area of the site, construc-
tion 01 additional monitor wells, including a multicased well
through a subsurface free oil zone, and sampling and analysis of
PCBs and volatile organics.
The Phase II site investigation was performed in response to a
Cleanup and Abatement Order issued by the Regional Water Qual-
ity Control Board and concurrently ordered by the DOHS and
USEPA. This work expanded upon the previous investigations by
detailing site history and operations, describing the existing site
drainage pattern, providing additional information on PCB dis-
tribution and water quality through more soil borings and monitor
well construction and completely characterizing the shallow
groundwater system.
The multiphase site investigation was comprehensive. After the
final phase was completed, the State of California acknowl-
edged to General Electric in writing that the Phase II Definition
Study adequately addressed the issues raised in the Cleanup and
Abatement Order and at subsequent meetings, and fulfilled the in-
tent and requirements of the order. Highlights of the site investi-
gations are presented below.
284
-------
CASE HISTORIES
285
Soil Sampling and Drilling Techniques
Soil samples were collected using a California split spoon drive
sampler through a hollow-stem, continuous-flight auger. For the
soil borings and shallow monitor wells, the first three samples at
0, 5, and 10 ft below grade were initially analyzed. If these samples
had detectable levels of PCBs (more than 0.1 ppm), additional
samples were analyzed until a depth of nondetectable PCBs was
found. If samples from 0 to 10 ft had nondetectable PCBs, no
additional samples were analyzed. For deep monitor wells drilled
adjacent to shallow wells, samples at 25, 30, and 35 ft were in-
itially analyzed and the same procedure described above was used
to determine the need for any additional analyses.
During the investigation, it was necessary to construct three mul-
ticased wells. These wells were drilled through PCB-contaminated
soils. To prevent downward movement of contaminants while
drilling and to collect representative soil samples below a shallow
zone of high PCB concentration in soil, it was necessary to seal off
the contaminated zone before constructing the monitor well. To
meet these requirements, the multicased wells consisted of a series
of concentric steel casings of decreasing diameter. A drillable grout
plug was placed at the bottom of each casing and each casing was
cleaned and sealed to prevent contamination of the next lower cas-
ing. After installing each casing, a cement-bentonite grout seal was
emplaced between the well casing and borehole.
Extent of PCB Contamination
PCB concentration in soils at the site ranges from nondetec-
table to 11,000 ppm, with high concentrations centered in the area
where pyranol was stored. PCB concentrations in soils attenuate
to nondetectable levels at a depth of 10 ft over most of the site with
some exceptions attenuating at 25 ft in the tank farm area and one
boring having measurable PCB concentrations to a depth of 35 ft in
the old PCB storage area. Detectable concentrations of PCBs were
not found in grand water samples.
Monitor Well Construction
The multiphase investigation involved drilling 89 soil borings
and constructing 76 monitoring wells. Over 800 soil samples and
70 fluid samples were collected and analyzed for PCBs. Over 300
water level measurements and 30 permeability tests were made to
help characterize the ground water system.
Free Oil
Free oil contaminated with PCBs was found in several mon-
itor wells, screened in thin layers of sand strata located at a max-
imum depth of 32 ft below grade under an oil tank storage area
(tank farm). The free oil measured in the monitor wells was a
maximum of 8 in thick.
Groundwater Mound
Groundwater flow in the shallow strata of the site, defined as
the upper 60 ft, tended to be away in all directions from a ground-
water mound in the vicinity of the tank farm. The source of the
groundwater mound was identified as the diked area of the tank
farm.
Hydraulic Continuity
A detailed flow net evaluation was made of the shallow ground-
water system; the potential for vertical migration of PCBs in
shallow soils on the site to deeper confined aquifers was determined
to be negligible.
DESIGN
The design involved preparation of bidding documents, plans,
and specifications. In addition, pilot tests were conducted on a
small scale oil-water separator for verification of performance ex-
pectations. Permits and approvals were arranged with various regu-
latory agencies, and as-built drawings were prepared after construc-
tion. The facilities designed included:
•A surface sealing system of bentonite and soil overlaid with a
permeable gravel layer
•Surface runoff control facilities including curb and gutters,
catch basins, drainage piping, drainage channels, and a mon-
itoring and sampling station
•A three-trench French Drain system with central collection sump
and mechanical extraction system
•A groundwater treatment system and storage facilities for treated
groundwater, sediment, and PCB-contaminated oil
•Modification of the existing tank farm (tanks previously used for
oil storage and mixing) for approved temporary bulk storage of
PCB fluids and drummed storage of sediments.
Groundwater Collection and Extraction System
A plan and section of the French Drain collection system and
extraction sump is shown in Fig. 1. The system was located in the
central area of the subsurface free oil area to facilitate oil collec-
tion. Three levels of 6 in diameter perforated pipe were designed
for each French Drain arm to provide flexibility in collecting oil
and groundwater.
TRENCH (TYPICAL)
6" PERFORATED LINE
Figure 1.
Plan and Section of French Drain and Extraction Sump
Oily-Water Treatment System
A schematic of the oily-water treatment system is shown in
Fig. 2. Oil collected via the surface skimmer is pumped directly to a
storage tank. Oily groundwater is pumped to an oil-water sep-
arator consisting of a rectangular fabricated steel box approx-
imately 20 ft long, 6 ft wide, and 6 ft high containing a series of
vertical and horizontal coalescing plates that are hydrophobic and
oilophilic. The plates cause coalescence of the fine oil droplets into
larger globules which more readily float to the surface for easy
collection via a static skimming pipe.
Solids settle in the separator, dropping into sludge hoppers for
collection by periodic pumping. Clean effluent passes under a baf-
fle and over a weir and flows by gravity through a monitor-
ing station and then into the sanitary sewer collection system of the
local utility district. Polymer may be added to the oil water sep-
arator influent to enhance coagulation and separation of solids.
The effluent can be pumped to a storage tank for testing before
discharge to the sewer.
-------
286
CASE HISTORIES
• SANITARV
SEWER
Figure 2.
Oily Groundwater treatment system
Surface Sealing and Drainage System
The Oakland plant site and the surface sealing and drainage
system are shown in Fig. 3.
The design provided two types of soil sealing systems: (1) a
bentonite-soil mixture covered with permeable rock, and (2)
asphalt-concrete paving and base rock coated with a surface seal-
ant. The bentonite (a natural low-permeability clay) soil mixture
was applied using 4 in. of imported soil and 4 Ib of bentonite/ft2
to achieve a final permeability not greater than 1 x 10~7 cm/sec
when compacted to 80% at optimum moisture content. Six inches
of crushed drain rock was spread over the bentonite-soil mixture.
This surface sealing system was used over portions of the site con-
taining high concentrations of PCBs, where vehicular traffic would
be prohibited, and where no facility expansion plans existed.
Part of the site was sealed with asphalt-concrete paving over an
aggregate base. A seal coat was applied at a rate of 0.10 gal/yd! to
both new and existing paving.
The site investigation showed that the western portion of the
property had so little PCBs that sealing appeared unnecessary.
Therefore, with concurrence from the regulatory agencies, this sec-
tion was left in its natural state except for some grading to provide
a holding basin for accumulation of runoff during severe storms.
The design contains three separate, structured drainage systems:
(1) roof runoff, (2) paved area runoff, and (3) bentonite seal area
runoff. The discharge of all three structured systems is combined
and passed through a monitoring station before discharge to a
storm sewer.
Having three systems provides the flexibility of system isolation
if monitoring results indicate abnormally high PCB levels in the
combined discharge. Part of the storm water collected in the un-
paved area runs off the site overland in a northwesterly direction,
while remaining storm water collected in this area drains to the
monitoring station.
Figures.
Surface Sealing and Draining System at Oakland Plant Site
PERMITS
One agency approval and three agency permits were required
for construction and operation of the Oakland site project. Ap-
proval was necessary from the Alameda County Flood Control and
Water Conservation District (District) for discharge of storm water
collected on site to the storm sewer system. An approved connec-
tion point, flow calculations, and a drainage system design draw-
ing with plans for a short-term holding basin were needed to satisfy
requirements of the District.
A building permit was required from the City of Oakland to con-
struct a structure over the tank farm. Securing this permit did not
permit any difficult problems.
An Authority to Construct and a Permit to Operate was re-
quired from the Bay Area Air Quality Management District. This
penriit was necessary due to the potential for volatile organic emis-
sions from the oil-water separator escaping to the atmosphere.
The most complicated permit required was a wastewater dis-
charge permit from the East Bay Municipal Utility District
(EBMUD). Under the provisions of this permit, treated ground-
water from the oil-water separator could be discharged to the
EBMUD sanitary sewer system. Permit provisions require flow
monitoring of the treated groundwater discharged to the sewer and
periodic (every 8 weeks) analysis of total suspended solids, filtered
chemical oxygen demand, oil and grease, and PCBs from a 16 hr
composited samples. The PCB discharge requirement is for a max-
imum of 2.3 ug/1 and an average of 1.4>ig/l.
The flow measurements are used to compute a sewage disposal
service charge. Quarterly reports are required documenting flow
readings and quantities of oil and solids stored on site or off-
hauled.
Provisions of the Resource Recovery and Conservation Act and
the Toxic Substances Control Act were explored for permit require-
ments applicable to the Oakland facility. No specific requirements
were found and it was therefore concluded, with concurrence of
the State of California, that no additional permits were required.
CONTRACTOR PROCUREMENT AND CONSTRUCTION
Prospective bidders for the Oakland site project were solicited by
issuing a request for qualifications which contained a brief des-
cription of the project. A pre-bid conference was held at the project
site to familiarize prospective bidders with the proposed project.
Subjects covered included a description of the project, contract
requirements, safety with respect to PCBs, and the contractor
selection process. Major items of concern expressed by the prospec-
tive bidders were containment of extracted groundwater and safety
provisions in working with PCBs. Key safety provisions discussed
were equipment cleaning, personal protective equipment, and con-
trol of excavated materials.
Six bids were received for the Oakland site project. The bids were
evaluated based on: (1) submitted project activity schedule, (2)
qualifications of key project personnel, (3) submitted safety pro-
gram, (4) extent of subcontracting planned, (5) any special con-
ditions of the bid, and (6) bid price. Underground Construction
Company of San Leandro, California, was selected as the suc-
cessful bidder for the project.
Construction
Construction work on the Oakland site began in Aug. and ended
in Dec. 1981. A particularly important feature of the construction
phase that facilitated maintaining a tight construction schedule was
the assignment of a full-time project manager to the site by General
Electric. This individual had personal knowledge of the site history
and ongoing operations and had complete authority to execute con-
tract change orders and make field decisions for the company.
Prepurchased Equipment
The oil-water separator and five pumps were prepurchased by
General Electric and furnished to the contractor. While attractive
in concept, this approach did cause problems. Equipment deliv-
-------
CASE HISTORIES
287
eries were delayed. When the equipment finally arrived, parts and
installation instructions were occasionally missing, and in some
cases, specifications had not been followed. Numerous hours were
spent by Brown and Caldwell and General Electric expediting pre-
purchased equipment deliveries. As a result, the contractor was de-
layed and the project schedule suffered.
A total of 49 change orders were executed during the project.
All but three of these change orders increased the contract price.
Construction Problems
Sheet Piling. The contractor drove sheet piling to prepare for
excavation of the extraction sump and French Drain trenches. The
sheet piling was to be driven 20 to 30 ft to prevent soil cave-in dur-
ing excavation and to afford some degree of groundwater isola-
tion. Unfortunately, neither a pneumatic pile driver nor a diesel
pile driver would drive the piles through the site soils. Piles shifted
and the ends buckled. Use of newer and thicker piles may have re-
sulted in more success. Eventually, a 24 in-diameter auger was used
to drill several deep holes in the extraction sump, permitting the
piles to be driven.
The French Drain lateral trenches were excavated to the re-
quired depth in sections by backhoe with sheet piling added after,
rather than during excavation. Very little groundwater was en-
countered in the lateral trenches or the extraction sump during con-
struction. Excavation of the extraction sump and lateral trenches
required three times longer than the contractor expected, which
proved very costly and resulted in the entire project falling behind
schedule.
Monitoring Wells. The 76 monitoring wells installed on site dur-
ing the site investigation phase all had to have well-casing exten-
sions installed so the site grading crew could see them. Even then,
approximately one third of the wells were damaged by impact with
earthmoving equipment and had to be repaired. Fortunately, none
of the multicased wells were damaged. In retrospect, it would have
been better to have extended each monitoring well a few feet above
the ground surface during initial construction, brightly colored the
well casings, and surrounded them with temporary fencing.
Buried Pipes. Since operations had been occurring at the Oak-
land site for a long time and as-built drawings of the facility were
not available, location of underground piping and utility lines
proved to be a problem. The recollection of General Electric per-
sonnel and numerous "potholing" expeditions were necessary to
locate buried pipes and lines. This situation caused considerable de-
lay in construction and many change orders but prompted careful
preparation of as-built drawings for the newly completed construc-
tion.
Wipe Tests. The contract specifications required that all con-
struction equipment leaving the site must be inspected and cleared
by the resident engineer. This provision was designed to ensure
that PCB-contaminated soil material did not leave the site and
contaminate another location. A wipe test was used; it consisted
of wiping a 1 ft2 area with a swab coated with acetone and analyz-
ing the swab material for PCBs. Equipment was cleared to leave
the site if the wipe test analysis had a PCB concentration of less
than 100jig/ft2.
Cleaning equipment to this specification required dirt scraping,
water washing, and detergent steam cleaning. The process proved
to be very time consuming and expensive for the contractor, par-
ticularly with rented equipment.
OPERATION AND MAINTENANCE
Operation and maintenance activities for the Oakland site pro-
ject involved basically the extraction system, the oil-water separa-
tor, and the material storage systems. An operation and main-
tenance manual .was prepared covering key aspects of these sys-
tems, including relevant construction features, operational theory,
and equipment maintenance.
Training sessions were conducted by Brown and Caldwell oper-
ator specialists to provide instruction to General Electric person-
nel on proper operation of the treatment plant. Startup assis-
tance, including process adjustment, was provided during the first
few weeks of plant operation with hands-on training for General
Electric personnel. Long-term operating assistance is being pro-
vided by Brown and Caldwell specialists for the first year of oper-
ation. This assistance involves solving process-related problems,
making system modifications, and assisting in monitoring perform-
ance.
Operating problems and solutions resulting from the first few
months of operation are described below.
Groundwater Quantity Extracted. It was predicted that as much
as 16,000 gal/day of groundwater might be extracted from the
French Drain system. Operating history has shown close to 1,000
gal/day of groundwater is being extracted. Because of the low flow
rate and General Electric's desire to man the treatment plant dur-
ing the same time period each day, automatic operation of the
groundwater treatment system was discontinued. The system is
now operated in batch mode.
Oil Skimmer Operation. The oil skimmer in the extraction sump
is designed to efficiently operate only if sufficient oil is present
to completely wet the skimmer membrane. Without sufficient oil,
water leaks into the skimmer, tripping an al.arm. To overcome
this problem, the skimmer is removed when the thickness of the oil
layer in the extraction sump is less than 1 in. When the oil layer
becomes thick enough, the skimmer is dropped to the water sur-
face briefly for oil removal to bulk storage.
Solids Buildup. Solids have accumulated in the oil-water separa-
tor downstream of the horizontal and vertical oil coalescing plates.
These solids are believed to have entered the system during start-
up and they have been removed. If continued buildup occurs, it
may necessitate adding foam filter pads to the oil-water separator
between the vertical and horizontal banks of coalescing plates.
Flowmetering System. The ultrasonic flowmetering system con-
sisting of a flume, ultrasonic transducer, and display panel mal-
functioned repeatedly during the early months of operation. The
problem was finally traced to a hairline crack in the transducer
that presumably occurred during installation. After transducer re-
placement, the system is performing reasonably well.
LINE OF EQUAL WATER
LEVEL ELEVATION IN FEET
-- SHALLOW GROUNDWATER
FLOW DIRECTION
Figure 4.
Water Level Contour Map in September 1981
-------
288
CASE HISTORIES
Ultrasonic Level Probe Fouling. During startup of the plant,
the ultrasonic level probes in the oil-water separator, which acti-
vates the effluent pump, fouled and the unit flooded the building
with extracted groundwater. The fouling substance consisted of a
sticky, viscous material suspected of being packing around the co-
alescing plates. After probe cleaning, the problem has not recurred.
Steam Condensate Sidestream. A steam condensate sidestream
from the General Electric apparatus service shop with high con-
centrations of solids, chemical oxygen demand, oil, and arsenic
was initially added to the oil-water separator influent flow. This
action proved to be problematic as the sidestream contained sur-
factants which reduced the coalescing of oil droplets in the oil-
water separator. Eventually, the sidestream was split off and
handled separately.
Figures.
Water Level Contour Map in August 1982
LONG-TERM MONITORING
In response to a request by the RWQCB, a long-term monitoring
plan was developed to monitor: (1) groundwater, (2) recovered,
treated groundwater, (3) recovered sediment and oil, and (4) storm
water runoff.
Groundwater
Monitoring of the groundwater is done to identify any PCBs in
groundwater and to determine changes in the groundwater levels.
Monitoring consists of water level measurements in all 76 on-site
monitor wells and sampling and chemical analyses from 19 selected
wells. Analyses are to be made for: water levels, PCBs, and oil and
grease.
A water level contour map prepared immediately before con-
struction of the French Drain is shown in Fig. 4; the water level
contour map 8 months after startup of the French Drain and
groundwater extraction system is given, for contrast, in Fig. 5.
Note the reversal of groundwater flow that has occurred and the in-
fluence the French Drain extraction system has had on the shallow
groundwater system water levels.
Recovered, Treated Groundwater
The recovered, treated groundwater (effluent from the oil-water
separator) is monitored to evaluate the unit performance and to
confirm that the discharge to the sewer system is meeting the re-
quirements of the receiving utility. Determined are flow rate,
PCBs, total identifiable chlorinated hydrocarbons, oil and grease,
and total suspended solids. Average values for weekly measure-
ment of these parameters during the first 8 months of plant opera-
tion are shown in Table 1.
Recovered Sediment and Oil
Solids and oil recovered by the oil-water separator are sampled
and analyzed to determine the PCB concentration of these wastes.
Sediment and oil PCB concentrations measured on one occasion
during the first 6 months were 27 ppm for dry sediment and 870
mg/1 for oil.
Storm Water Runoff
Storm water runoff is monitored to detect any PCBs, permit-
ting an evaluation of the effectiveness of the bentonite and asphalt
sealing systems. For the period Nov. 1981 through Apr. 1982,
PCBs in the runoff averaged 2.4pg/l.
Table 1.
Recovered, Treated Groundwater Monitoring Results
Parameter
Flow, gpd
PCBsa, ppb
Oil and grease , ppm
Total suspended
soilds, ppm
Average value
Influent
-
6.5
9.2
5.8
Effluent
1,100
0.1
7.1
5.8
aActually total identifiable chlorinated
hydrocarbons.
Note: For first 8 months of operation.
-------
IMPLEMENTATION OF REMEDIAL ACTIONS AT ABANDONED
HAZARDOUS WASTE DISPOSAL SITES
DALE S. DUFFALA
PAUL B. MAC ROBERTS
Black & Veatch
Kansas City, Missouri
INTRODUCTION
In this paper, the authors describe current efforts to study, de-
sign, and implement remedial actions at abandoned hazardous
waste disposal sites in USEPA Regions VI through X. They also
discuss the legislative background and enabling authorities to pro-
vide such remedial activities, the provisions of Black & Veatch's
contract with USEPA, issues related to the objectives of Super-
fund, and investigations and feasibility studies at an abandoned
site.
LEGISLATIVE BACKGROUND
There are three major laws that cover remedial activities at aban-
doned hazardous waste disposal sites. These laws include the Clean
Water Act (CWA), the Resource Conservation and Recovery Act
(RCRA), and the Comprehensive Environmental Response Com-
pensation and Liability Act (also known as Superfund).
Section 311 of the Clean Water Act required the preparation
of the National Contingency Plan to provide for emergency re-
sponse to discharges of oil or hazardous substances. RCRA pro-
vides the overall legislation governing the management of haz-
ardous wastes including standards for generators, transporters,
and owners/operators of treatment, storage, or disposal facilities.
Regulations promulgated under RCRA and the Consolidated
Permits Program provide facility design guidance, outline closure
and financial requirements, and set general operating standards.
The most important legislation governing abandoned sites is
Superfund, which provides $1.6 billion for cleanup. Superfund also
established a Postclosure Liability Trust Fund and required revis-
ion of the National Contingency Plan while establishing liability
provisions for cleanup costs. Superfund provided the basis for a
USEPA contract to provide design services for remedial actions at
abandoned sites.
CONTRACT BACKGROUND
Under the provisions of Superfund, last year USEPA pro-
cured engineering services to provide remedial actions at aban-
doned disposal sites. The country was divided into three zones:
Zone 1 is Regions I, II, and IV, Zone 2 is Regions II and V,
and Zone 3 is Regions VI through X. Black & Veatch was selected
to perform engineering services in Zone 3. The contract is a cost
plus fixed fee work order type contract and calls for provision of
certain services at abandoned disposal sites where the USEPA
takes the lead to clean them up.
The scope of work calls for five separate areas of service to be
provided including initial field investigation, feasibility study, de-
sign, construction management, and general technical support. The
objective of the initial site investigation is to establish the degree of
contamination, sources, possible remedies, and impacts to public
health and environmentally sensitive areas.
Field Investigation
Typical tasks associated with the field investigation include:
•Determination of the amount and composition of hazardous sub-
stances
•Determination of the extent of contamination
•Groundwater and air quality monitoring and modeling
•Collection of additional data required by subsequent activities
•Identification of environmental pathways
•Characterization of potential exposure
•Preliminary data processing and storage
•Preparation of reports and participation in project review
meetings
FEASIBILITY STUDY
The feasibility study incorporates a review of problem history
and the results of field investigation to analyze and compare re-
medial options. The comparison is based on technical feasibility,
expected duration, cost effectiveness, economic reliability, O&M
considerations, and public acceptance. Typical tasks included in
the feasibility study are:
•Processing and review of data collected during the remedial in-
vestigation and definition of additional data requirements
•Analysis of complete data base
•Definition and analysis of discrete problems impacting the devel-
opment of remedial options
•Definition of alternative remedial measures, phasing options,
and consequences
•Performance of treatability, compatibility, and pilot scale studies
•Performance of cost effectiveness and risk analyses
•Performance of preliminary environmental assessments and
economic impact analyses
•Recommendation and approval of preferred options
•Preparation of conceptual design of the selected options
•Identification of discrete work packages for an integrated design-
construction approach
•Development of preliminary cost estimates, cash flow require-
ments, and schedules for subsequent activities
•Performance of detailed environmental assessment and economic
impact analysis
•Preparation of reports and participation in project review meet-
ings
Engineering Design
The engineering design phase defines the selected remedial ac-
tion in sufficient detail to permit direct implementation. De-
pending on site-specific conditions, a wide range of potential de-
sign solutions are available, including:
•Construction plans for berms, dikes, ditches, dams, and other
runoff diversion structures
•Underground and surface pipelines for short distance pumping
and transfer of wastes on site
•Lagoon training, chemical fixation, neutralization
•Provision for incineration of wastes on site
•Ground water interceptor well networks
289
-------
290
CASE HISTORIES
•Ground curtain, curtain wall, and sheet piling cutoff wall em-
placement
•Dike stabilization
•Temporary and permanent holding tank or reactor tank construc-
tion
•Temporary and permanent biological and physical/chemical treat-
ment plants
•Construction of alternative water supply pipelines
•Specifications for off-site disposition of wastes, including thermal
destruction, chemical treatment, and land disposal
Typical tasks involved in engineering design include:
•Acquisition of additional data required
•Procurement of applicable federal and local permits and rights-
of-way
•Monitoring availability of suitable disposal sites
•Preparation of engineering designs and specifications, including
post-closure, resource replacement, and relocation plans
•Integration of containment and cleanup feasibility considerations
•Preparation of cost estimates and schedules
•Preparation of bid documents for construction and cleanup and
technical support in reviewing and evaluating bids
•Preparation of procedures and forms
•Preparation of operation and maintenance manuals
•Development of operator training programs when needed
•Preparation of reports and participation in project review meet-
ings
Construction Phase
Construction phase services typically include the following:
•Answer questions during the bidding period by the Contracting
Officer regarding meaning or interpretation of the drawings and
specifications and prepare the required changes
•Assist the Contracting Officer in analyzing and evaluating pro-
posals and bids
•Check and recommend approval action on shop drawings, ma-
terial samples and proposals submitted by the construction Con-
tractor
•Be available for consultation and advice during the planning,
construction, and final inspection of the project
•Prepare any additional details or detailed drawings as required for
clarification of construction
•Revise the original drawings to show construction as actually
accomplished
•Visit the construction site to review and record the progress of
the work and to inspect all aspects of the work for conformance
with the plans and specifications
•Prepare plans and specifications for any changes to the contract
that may be required
•Check the validity of all invoices for payment submitted by the
construction Contractor
•Review claims submitted by the construction Contractor and
make recommendations for their settlement to the Government
Project Manager
ISSUES RELATED TO SUPERFUND
The implementation of remedial actions is a complex process.
There are several issues that relate to the success of Superfund
programs including:
•Site specific variables
•"How clean is clean?"
•Communications
Site specific variables can influence the ultimate choice of remed-
ial options for implementation. Some of these variables include:
•Site geotechnical and hydrological characteristics
•Extent and magnitude of contamination
•Local political situation
•Proximity to off-site disposal facilities
•Proximity to experienced contractors
The issue of "how clean is clean?" is related to cost since the in-
cremental cost of achieving "complete" cleanup can be prohibi-
tive. It is necessary during the feasibility study to determine which
remedial alternative results in an acceptable reduction of risk and
also an acceptable level of contaminant reduction. This is de-
pendent on the characteristics of both site and wastes and repre-
sents the most difficult engineering decisions in the entire cleanup
process.
Communications between state officials, regional USEPA, and
USEPA Washington are critical because overall program manage-
ment comes from Washington, whereas each Region assigns a
project monitor to each specific site. It is essential that the con-
cerns of state officials are adequately addressed during cleanup
activities.
CASE HISTORY
One of the sites under this contract is the MOTCO disposal
site located near Texas City, Texas. This site is an 11-acre tract of
land that contains approximately 7 acres of abandoned lagoons.
The lagoons were originally used for storage in a waste styrene tar
reclamation operation. For several years the facility was also used
as a dumping ground for various waste products, including vinyl
chloride bottoms, acids, metallic catalysts, and elemental lead and
mercury. The waste materials are contained in unlined lagoons
that vary in depth from approximately 12 to 30 ft. Several attempts
were made to recycle the waste materials for commercial pur-
poses, but all of them failed. It was determined by the USEPA
and the U.S. Coast Guard that the site was posing a serious threat
to navigable waters and, thus, an action under Section 311 of the
Clean Water Act was taken. The immediate threat was eliminated
by the construction of dikes and fencing. However, the extent of
subsurface contamination was not investigated.
Black & Veatch conducted initial and supplemental field inves-
tigations to determine the nature and extent of contamination. A
site-specific safety plan was prepared and samples were taken from
air, surface water, surface soils, ground water, subsurface soils,
and the waste pits. Analytical data cannot be released due to pend-
ing litigation, but off site contamination has been discovered.
In addition to reporting the results to USEPA, Black & Veatch
developed some preliminary order of magnitude costs for eight po-
tential remedial action alternatives. The next level of investiga-
tion will be a feasibility study to determine the recommended
remedial action. This study will include a detailed cost analysis
that will replace the order of magnitude cost figures derived prior
to beginning the feasibility study.
Work is also progressing at other sites throughout Zone 3 includ-
ing:
•Ellisville, Missouri—Preparation of bid package in drum removal
and feasibility study for cleanup of three sites
•Council Bluffs, Iowa—Preparation of Remedial Action Master
Plan (RAMP) and feasibility study for site cleanup
•Commerce City, Colorado—RAMP preparation
•Tucson, Arizona—RAMP preparation and monitoring well in-
stallation
•Globe, Arizona—RAMP preparation
•Tacoma, Washington—Remedial investigation and feasibility
study in contaminated municipal well field. Remedial investiga-
tion in South Tacoma Swamp
•Denver, Colorado—Feasibility study for radium sites. RAMP for
abandoned chemical site
After feasibility studies are completed, the Corps of Engineers
will manage design and construction (implementation) contracts
while USEPA will approve the appropriate remedial alternative.
It is evident that through the Superfund program, the country
has taken significant steps toward protecting public health and wel-
fare. The cleaning up of abandoned hazardous waste disposal sites
is a well directed program which will be recognized as producing
benefits for exceeding expenditures.
-------
PROPOSED CLEANUP OF THE GILSON ROAD HAZARDOUS
WASTE DISPOSAL SITE, NASHUA, NEW HAMPSHIRE
FREDERICK J. MC GARRY
BRUCE L. LAMARRE
R.F. Weston, Inc.
Concord, New Hampshire
INTRODUCTION
The Gilson Road Hazardous Waste Dump site is located in
Nashua, New Hampshire, near the New Hampshire/Massachusetts
State line. The 6 acre site is located in a residential area of the
city with two large mobile home parks directly adjacent to it (Fig.
1).
The site was developed as a sand borrow pit by the owners, with
substantial quantities of sand having been removed over several
years of operation. The sand excavation at many locations ex-
tended into the ground water table underlying the site. At some
point in the late 1960s, the owner decided to discontinue the sand
mining operation and began the operation of an unapproved and
illegal Refuse dump with the apparent intent of filling the pit left by
the sand mining.
Household refuse and demolition materials were the initial com-
ponents of the wastes dumped at the site. Eventually, chemical
sludges and aqueous chemical wastes were also dumped there. The
domestic trash and demolition material were usually buried while
the sludges and liquid chemicals were disposed of in several ways:
mixed with the trash and buried, placed in steel drums and either /
buried or stored on the surface, or dumped into a makeshift leach-
ing field and allowed to percolate into the ground.
It was estimated that the site was used for the disposal of haz-
ardous wastes for approximately five years. During the last eight
months of operation at the site, Jan. to Oct. of 1979, over 800,000
gal of aqueous wastes were known to have been disposed of there.
With the dump having been used for at least five years for aqueous
waste disposal, the total volume of liquid waste dumped there is
likely substantial.
Cleanup activities began in 1980, as soon as legal access to the
site could be obtained. During May and June of that year, 1,314
drums, which were accessible, were removed by a contractor and
disposed of at approved secure landfills in the States of New York
and Ohio.
Immediately following removal of the drums, a ground water ex-
ploration and monitoring program was initiated to determine the
magnitude and extent of the contamination. In July of 1981, an en-
gineering report prepared by another consultant stated that there
were high concentrations of heavy metals, volatile organics, and
extractable organics in the ground water under the site. The metals
and much of the extractable organics could be traced to the domes-
tic refuse while the volatile organics likely originated from the
makeshift leaching field used for the disposal of most of the
aqueous wastes.
The concentration of organics, measured as total organic car-
bon, was found to be over 4,000 mg/1. Total metals concen-
trations, in the most contaminated portion of the plume, were in
excess of 1,000 mg/1. Some of the major organic contaminants
included methylene chloride, methyl ethyl ketone, toluene, ben-
zene, chloroform, tetrahydrofuran, and acetone. The metals found
at the site included iron, manganese, nickel, zinc, barium and
arsenic.
The ground water plume of contaminants extended over an area
in excess of 20 acres and was moving at a rate estimated at 0.8 to
1.6 ft/day. The plume was moving toward the Nashua River and
contaminants from the upper portion of the plume had already en-
tered Lyle Reed Brook, a small stream tributary to the Nashua.
The primary remedial measure recommended by the initial con-
sultant was the encapsulation of a 12.5 acre site with a slurry trench
cut-off wall and an impermeable surface cap. For that portion of
the ground water plume outside the containment wall, a treatment
system would be installed which would intercept and treat the
groundwater. The treatment process would consist of air stripping,
biological oxidation, and activated carbon adsorption (Fig. 2).
Chemical precipitation of the metal contaminants was considered
to be an optional treatment step.
Following cleanup of the acquifer outside the slurry wall, all
further activities would be terminated and the site would be left
with the slurry wall to permanently contain the 12.5 acres of con-
taminants.
Figure 1.
Location Map
WELL
FIELD
VAPOR
RECOVERY
WITH
ACTIVATED
CARBON
t
AIR
STRIPPER
r
i
i
pH
ADJUSTMENT
WITH
SODIUM
HYDROXIDE
(OPTIONAL)
1
BIOLOGICAL
TREATMENT
WITH
R 8. C
CARBON
ADSORPTION
-»
LEACHING TRENCHES
Figure 2.
Proposed Treatment System in Initial Report
FURTHER INVESTIGATIONS
Shortly after the completion of the report in 1981, the State of
New Hampshire received a grant from the USEPA, funded under
291
-------
292
CASE HISTORIES
the Comprehensive Environmental Response, Compensation,
and Liability Act of 1980, for the cleanup of the site. A condition
of the grant called for the preparation of an interim report deal-
ing with the feasibility of groundwater treatment.
WESTON was retained by the State of New Hampshire to pre-
pare this report. Among the concerns raised by the State and
USEPA were: the effects on the environment of talcing no action
at the site, the long-term integrity of the slurry wall, and the cost-
effectiveness of the proposed treatment system.
No Action Alternative
WESTON's investigation found that taking no action at the site
would have impacts both locally and some distance from the site.
A small brook borders the site (Fig. 3) and, as previously men-
tioned, it was found that upper portion of the plume had migrated
to the brook and low levels of contamination were being detected
downstream.
Several municipalities in Massachusetts draw their drinking
water directly from the Merrimack River and provide service to a
combined population of over 100,000. Through direct surface
water flow and groundwater seepage into the Nashua River, a no
action alternative would result in exceeding suggested USEPA
Water Quality Criteria for a lifetime cancer risk of 1 in 1,000,000
for methylene chloride and chloroform and exceeding 50% of
the levels for trichloroethylene, benzene, and 1,2-dichloroethane
in those downstream communities. Based upon soils flushing ex-
periments and the current rate of natural groundwater flow
through the site, it was likely that the cumulative effects of the
concentration of organics at the downstream water supplies would
exceed USEPA Water Quality Criteria for a minimum of 20 years.
The effects on the aquatic communities within the Nashua and
Merrimack Rivers were anticipated to be significant during low
flow periods. Long-term effects are unknown since there is little if
any research into the effects of the organic contaminants on spe-
cies found in the receiving waters.
Locally, adjacent to the site, substantial volatilization could be
expected as the Lyle Reed Brook flowed to the Nashua River. Al-
though the predicted airborne concentration of the organic vapors
would not be toxic, they would, cumulatively, exceed the recom-
mended long-term exposure limits for a cancer risk of 1 in
1,000,000 by a factor of 100 for residents of the mobile home park
located some 30m from the brook. Expected odor levels adjacent
to the brook, as well as in the mobile home park, were expected
to be high, in some instances more than 400 times the odor thres-
hold.
Containment
A second major concern was the long-term integrity of the slurry
wall and the potential seepage of contaminants through the wall or
into the underlying bedrock. The wall in the previous consultants'
recommendations, would be used to permanently contain the con-
taminants within the 12.5 acre site. It was found through a re-
view of the literature, discussions with contractors, and laboratory
work by a hydrogeological consultant to the state, that the integ-
rity of the wall would be suspect because of the contaminants
contained within the plume. Bentonite has a tendency to shrink
Figure 3.
Location Map
-------
CASE HISTORIES
293
when subject to some of the chemical contained in the plume,
thereby increasing the permeability of the wall. The laboratory re-
sults indicated an increase in permeability from 10-7 to 10-!
cm/sec when the bentonite was exposed to the contaminants con-
tained within the ground water.
In addition to the concern with the integrity of the wall, there
was also concern with seepage of contaminants into the underly-
ing bedrock. Even with an ideally constructed slurry wall and cap
containment system, there will be flow into the site through the
cap and underflow of the wall by way of the fractured bedrock.
The combined volume of flow from these sources was estimated at
5,000 gal/day. This uncontaminated flow would displace an equal
volume of contained groundwater which would exit the site
through the bedrock. This would then be expected in time to enter
the Nashua River. Even at a rate of only 5,000 gal/day, the com-
bined concentration of the contaminants would equal the USEPA
Water Quality Criteria for a cancer incidence of 10 -'.
Because of these concerns, long-term use of the slurry wall for
containment of the contaminated groundwater was not recom-
mended. However, a slurry wall and surface cap system could be
used quite effectively as a "clean water exclusion system." With
such a concept, the wall and the cap would be constructed not to
contain the contaminants but to exclude uncontaminated ground
water from entering the contained site. Such a system would be
used in conjunction with a groundwater treatment process.
Wells, located within the slurry wall, would pump contam-
inated water, treat it, and return it for recharge within the wall.
This system would significantly reduce the volume of contam-
inated groundwater leaving the site since there would be induced
flow into the site from the bedrock because of the pumping from
within the wall. It was recommended to utilize this system to sur-
round a 20-acre area of highly contaminated groundwater. A
schematic of the recommended system is shown in Fig. 4.
Investigation of Treatment Alternatives
Various treatment alternatives were investigated to determine if
the recommendations contained in the initial report were cost-
effective. Because of the three distinct components of contam-
ination in the groundwater, the alternatives for treatment had to
address each of the contaminant groups. Heavy metals removal
was the first process investigated.
The chemical analyses of the groundwater had indicated high
concentrations of iron and manganese, averaged 350 mg/1 and
80 mg/1 respectively. Although not toxic or known to be carceno-
genic in these concentrations, these compounds could severely
affect the operation of a treatment processes. It had been found
both in the laboratory and by personnel working at the site, that
the introduction of air into the contaminated groundwater would
cause immediate precipitation of these metals. Failure to remove
them would likely result in plugging or fouling of any process where
air was introduced.
Four alternatives were reviewed which could be effective in re-
moving the heavy metals component of the plume. These con-
sisted of: chemical precipitation, ion exchange, electrodialysis,
and reverse osmosis. Because of the high content of organics in
the groundwater, only chemical precipitation with lime was deter-
mined to be a reliable, cost-effective alternative. In the labora-
tory, this process removed virtually all of the iron and manganese
and provided a secondary benefit by removing about 65 % of the
other heavy metals.
From the standpoint of both health and environmental effects,
the volatile organics fraction of the contaminants is the most sig-
nificant component of the groundwater plume. Some of the
analyses of the groundwater taken from wells on the site had
shown concentrations of volatile organics of over 1,800 mg/1.
Groundwater from one well, located directly adjacent to the form-
er "leaching field", had a volatile organic concentration of 11,200
mg/1. This high concentration of low molecular weight organic
compounds has a significant effect on the selection of treatment
alternatives.
^ Carbon adsorption, which is effective for the removal of high
molecular weight insoluble compounds, had produced poor results
in treating the groundwater from this site. Breakthrough of the
volatile organic fraction occurred after operating only a short
period of time. Because of this situation, activated carbon was
not considered a biable treatment alternative. After reviewing six
other treatment alternatives, the two most effective processes ap-
peared to be steam stripping and biological treatment.
^ Air stripping of the volatile organics, recommended in the prev-
ious engineering report, was found to be only partially effec-
tive in removing the contaminants. With an initial TOC of the
groundwater of 4,000 mg/1, the best result in bench scale tests pro-
duced a stripping column effluent of 3,400 mg/1.
^- Steam stripping, in which the groundwater was preheated to 85-
90°C, produced significantly better results. This process takes ad-
vantage of the high volatility of most of the compounds at ele-
vated temperatures. The results from the bench scale test indi-
cated that the TOC of the waste stream could be reduced from
4,000 mg/1 to 1,400 mg/1. Nearly all of a selected number of rep-
resentative volatile compounds were completely removed. As might
be expected, this process was only moderately effective in remov-
ing the non-volatile fraction of the organic compounds.
Biological treatment (activated sludge or rotating biological con-
tactors) was very effective in removing both volatile and extrac-
PRECIPITATION
| | SANDS a BEDROCK
{'.','. '-~ | FRACTUftSD BEDROCK
I I a.
I ARGE GROUND WATER FI 0*
LAY WALL a CAP
Figure 4.
Proposed Containment/Treatment System
Figure 5.
Proposed Groundwater Treatment System,
Gilson Road Hazardous Waste Dump Site
-------
294
CASE HISTORIES
table organics, eliminating nearly 90% of the total organic car-
bon. However, because of the need to add substantial quantities
of nutrients and the relatively high capital cost of the process,
compared with steam stripping, biological treatment alone was
determined to be too costly for this application.
Biological treatment in the form of an aerated lagoon was found
to be cost-effective in removing the remaining extractable organ-
ics following the steam stripping process.
Because of the reduced levels of organics entering this process
and the resolubilizing of some of the nutrients as the biological
sludge decomposes within the lagoons, lower levels of nutrient
addition would be expected than with other forms of biological
treatment.
The recommended treatment process would consist of chemical
precipitation of the inorganics followed by pressure filtration to
remove any remaining metallic floe. The waste stream would then
be preheated to 85-90 °C and then passed through a stripping col-
umn. Air would be passed counter-currently through the column
and the off gasses would be burned in a fume incinerator.
The effluent from the stripping column would then be pumped
to an aerated lagoon for removal of the remaining extractable
organics. Nutrients in the form of ammonia and phosphoric acid
would be added to the lagoon to insure adequate biological treat-
ment. The lagoon effluent would then be pumped to recharge
trenches located within the bentonite slurry wall. A diagram of the
treatment process is shown in Fig. 5.
In addition to establishing the most cost-effective treatment pro-
cess, it was also necessary to determine the design treatment flow
rate and the length of operation of the treatment system.
Soil flushing experiments indicated that 90% of the contam-
inants could be removed from the soil after two full clean water
flushes. Seepage of the remaining contaminants, when diluted with
the surrounding groundwater and down-stream surface waters, was
not expected to cause any health problems. The 20 acre slurry wall
would hold approximately 137 million gal of contaminated ground-
water. The estimated clean-waster flow into the site from the bed-
rock and through the slurry wall was estimated at 23,000 gal/day.
Capital and operating costs and salvage values were developed
for five different flow rates of the recommended treatment pro-
cess. These rates ranged from 25 to 400 gal/min. After preparing
a plot of the capital, operating, and salvage costs of the system
(Fig. 6), it was found that the optimum treatment rate was ap-
proximately 300 gal/min, dependent upon rate of return used in
the analysis. At this rate, it is expected to take 1.74 years to provide
the two full flushes of the site necessary to remove 90% of the con-
taminants retained by the soil. The estimated capital costs would
be $4,920,000 and the annual operating costs are anticipated to be
$1,380,000.
CURRENT STATUS
In late 1981, a groundwater recovery and discharge system was
installed in an effort to "freeze" the position of a major portion
of the plume. This system appears to have been effective. A con-
tractor began installation of the surface cap and slurry wall in
Aug. of this year and is expected to complete the wall by Dec.
A pilot plant will be constructed to accurately determine spe-
cific design parameters and fine tune the operation to minimize
O&M costs. This experimental work will be followed by the de-
sign and construction of the treatment facility which should be on-
line in early 1984.
COST EFFECTIVE ANALYSIS CURVES
(7-S/8* INTEREST)
SILSON ROAD HAZARDOUS WASTE OUMP SITE
Figure 6.
Cost Effective Analysis Curves
ACKNOWLEDGEMENTS
The authors wish to acknowledge the participation in this project
of Dr. Enos Stover of Environmental Engineering Consultants of
Stillwater, Oklahoma. They also wish to thank John Ayers and
the staff of Goldberg-Zoino and Associates of Cambridge, Massa-
chusetts for the cooperation which they provided, on this project.
The assistance and support of Mr. Michael P. Donahue and Mr.
Thomas Roy of the New Hampshire Water Supply and Pollution
Control Commission was greatly appreciated.
REFERENCES
1. GHR Engineering Corporation, "Hazardous Waste Site Investigation
Sylvester Site, Gilson Road, Nashua, New Hampshire," Final Report
prepared for the New Hampshire Water Supply and Pollution Control
Commission and the Region 1, USEPA Laboratory, July 1981.
2. R.F. Weston, Inc., "Final Report— Sylvester Hazardous Waste Dump
Site Containment and Cleanup Assessment" prepared for the New
Hampshire Water Supply and Pollution Control Commission, Jan.
3. Roy F. Weston, Inc., "Supplemental Study to Final Report on Syl-
vester Hazardous Waste Dump Site Containment and Cleanup Assess-
Water Supply and
-------
REMEDIAL ACTIVITIES AT FLORIDA'S
UNCONTROLLED HAZARDOUS WASTE SITES
VERNON B. MYERS, Ph.D.
DAN DI DOMENICO
BRENT HARTSFIELD
Florida Department of Environmental Regulation
Tallahassee, Florida
INTRODUCTION
Florida's water resources are not limited to its expansive 1200 mile
coast line and over 7000 freshwater lakes. It is also blessed with an
abundant supply of groundwater. The Floridan Aquifer underlies the
majority of peninsular Florida and supplies most of the potable
water needs of north and central Florida. The Floridan Aquifer is at
the surface in many areas of central Florida. In coastal areas where
the Floridan Aquifer is saline, surface aquifers are used for potable
drinking water. The majority of the state has very pervious soils
which allow for rapid transport of contaminants to the shallow sur-
face aquifers; hence the groundwater is extremely vulnerable to con-
tamination.
Toxic materials are finding their way into groundwater from
various sources: hazardous waste dumps, waste lagoons and pits,
sanitary landfills, accidental spills, mining operations, and storm
water runoff. A recent assessment funded by the USEPA1 showed
over 70% of Florida industries using unlined impoundments for
waste disposal. This is particularly alarming when 98% of the 6000
impoundments were located over useable drinking water aquifers.
Florida's prominent position on EPA's Interim Priority List of un-
controlled hazardous waste sites is due, in part, to the susceptibility
of the state's groundwater to contamination.
SUPERFUND SITES
Florida recognizes that an integrated effort is needed to protect the
state's ground and surface waters, fauna and flora, and inhabitants
from exposure to hazardous wastes. In 1981, approximately 200
potentially dangerous uncontrolled hazardous waste sites were iden-
tified in the state. Twenty-seven of the most dangerous sites were
ranked based on their relative hazard potential using USEPA's
model. Sixteen of these sites were listed among the initial 115 USEPA
interim sites eligible for funding under Superfund2. Twenty-four ad-
ditional sites were ranked in 1982 and nine of these are being con-
sidered for addition to the priority list.
Florida's 16 sites (Table 1) are located throughout the entire state
which emphasized both the extent of the problem and vulnerability
of the environment. Very few areas of the state can be considered un-
susceptible to potential hazardous waste contamination.
Principal contamination sources at Florida's uncontrolled hazar-
dous waste sites (Table 1) include: municipal and industrial waste
disposal facilities, fuel and solvent spills, battery casings, and in-
dustrial wastes from drum recycling, galvanizing, wood treating,
electroplating, solvent extraction, battery salvaging, agricultural
chemical manufacturing, polyester resin manufacturing, and
solderless terminal manufacturing. Contaminants from these opera-
tions include both heavy metals and synthetic organic compounds.
These toxic contaminants have been found in groundwater and/or
surface waters and soils at these sites.
Activities at these hazardous waste sites fall into five main
categories: enforcement by USEPA, Florida Department of En-
vironmental Regulation (DER), or local programs, monitoring
studies by responsible parties or DER, USEPA funded remedial ac-
tion master plans or studies, USEPA/DER cooperative agreements
under CERCLA, and cleanup by responsible parties. Enforcement is
taking place where viable responsible parties have been identified.
USEPA guidance3 currently specifies that before CERCLA
cooperative agreement activities can proceed, the enforcement
avenues must be pursued. Monitoring studies are underway where
the extent of the site contamination is not known.
In the remainder of this paper, the author will discuss joint
USEPA/DER CERCLA efforts. As of Sept. 1982, two sites have
USEPA funded remedial action master plans (RAMPs) in prepara-
tion, three sites have USEPA/DER cooperative agreements for
remedial investigations, and one site has an USEPA zone contractor
study ongoing.
REMEDIAL ACTION MASTER PLANS
The USEPA has authorized the preparation of RAMPs for the
Rollings worth and American Creosote sites. The purpose of a
RAMP is to provide a general planning document for a hazardous
waste site. A RAMP can be used to initiate work by a USEPA Zone
Contractor, be the basis for a cooperative agreement with USEPA,
or provide a plan of action for the site to assist with negotiations
with responsible parties by the State and USEPA enforcement pro-
grams. The RAMPs will summarize available information for these
sites and scope out remedial activities following the procedures
established by the National Contingency Plan.
Hollingsworth
The Hollingsworth Solderless Terminal Company, located in
Broward County, Florida, was a manufacturer of electrical
solderless terminals. As part of the manufacturing process, the
parts were heat-treated in molten salt baths, degreased in trichloro-
ethylene, and electroplated. Over the years, Hollingsworth dis-
posed of its wastes in several ways. Machine parts were routinely
cleaned on site and the solvents used were washed into the ground.
Oil, grease and other unknown materials were routinely dumped in-
to the septic system. Waste trichloroethylene was dumped routinely
into a shallow drain field. The final mode of waste disposal was via
an unauthorized injection well into the Biscayne Aquifer. The well
was initially used to provide water for the plant cooling system. Ap-
proximately three to five years ago the use of the well was modified
to inject solvent wastes into the ground.
Several Broward County public water supply wells and a major
City of Ft. Lauderdale wellfield are all located within five km of the
Hollingsworth site. One Broward County well is only about 120 m
from the site. It has been estimated that more than 10,000 people
are served by groundwater in this area.
The unauthorized disposal practices were not discovered until
Hollingsworth contacted the Broward County Environmental
Quality Control Board to report that oil was bubbling up out of the
ground in the area surrounding the disposal well. Subsequent
groundwater monitoring has shown the presence of trichloroethy-
lene, and cis and trans 1,2-dichloroethane at the site. High levels of
copper, nickel and tin were also found. Monitoring is underway to
determine if the two public water supply wells are contaminated.
295
-------
296
CASE HISTORIES
Table 1.
Florida CERCLA Sites
Site Location* Ranking?
Biscayne Aquifer Dade 2nd
58th Street Landfill
Miami International
Airport
Miami Drum Services
Picketville Road Landfill Duval 3rd
Reeves Southeast
Galvanizing Hillsborough 3rd
American Creosote Escambia 4th
Taylor Road Landfill Hillsborough 4th
Davie Landfill Broward 5th
Pioneer Sand Escarabia 5th
Timberlake Battery Hillsborough 5th
Whitehouse Oil Pits Ouval 5th
Cole-nan-Evans Duval 6th
Hollingsworth Broward 6th
Alpha Chemical Polk 7th
Zellwood Groundwater Orange 7th
Gold Coast Oil Dade 8th
Sapp Battery Jackson 8th
Tower Chemical Lake 8th
Notes:
1 County
2 EPA ranking groups of 10
Contamination Source
Waste disposal facility
Fuel and solvent spills
Drum recycling wastes
Industrial wastes
Wastes from galvanizing
operation
Wastes from wood treating
faci lity
Waste disposal facility
Waste disposal facility
Electroplating waste disposal
landfill
Battery casings
Waste oil and acid sludge
pits
Wastes from wood treating
facility
Wastes from solderless
terminal manufacturer
Wastes from polyester resin
manufacturer
Wastes from agricultural
chemical or drum recycling
operations
Spills from drum storage and
solvent extraction operation
Wastes from battery salvage
operation
Wastes and spills from agri-
cultural chemical manufacturer
Action
EPA zone contract study
DER Enforcement/ground-
water Monitoring
Monitoring study
EPA Cooperative Agree-
ment for surface clean-
up and groundwater study
Monitoring study
County Enforcement/
groundwater monitoring
EPA RAMP/DER Enforcement
EPA Enforcement
Monitoring study
DER Enforcement
DER Negotiation
EPA Cooperative Agree-
ment for Phase I inves-
tigation
DER Enforcement/
groundwater monitoring
EPA RAMP/County Enforce-
ment
Monitoring study
Monitoring study
Responsible party
cleanup
EPA Cooperative Agree-
ment for Pahse I invest-
gation
Judgement/EPA Enforce-
ment
American Creosote
The American Creosote site comprises approximately 6 ha.
American Creosote Works is an inactive wood treatment facility
located in a densely populated residential/commercial area of ur-
ban Pensacola. Florida. The facility had been treating wood from
1901 to 1982 and discharged contaminated wastewater into two
unlined percolation ponds. The facility utilized the "Rurping pro-
cess" for treating wood with pentachlorophenol and creosote The
condensate water from this process was discharged into the unlined
ponds. This wastewater contained entrained oils, preservatives, and
-------
CASE HISTORIES
297
wood carbohydrates. Bottom sediment sludge from the treatment
process was also discharged into the ponds.
Sampling of soils and waters at the American Creosote facility
has revealed the presence of numerous substances including pen-
tachlorophenol, oils and greases, phenol, 2,4-dimethyl phenol,
4-raethyl phenol, fluorene, pyrene, fluoranthene, phenanthrene,
and numerous other compounds. These contaminants have per-
colated through the highly permeable soil into the shallow ground-
water table. The groundwater beneath the site is moving in a
southerly direction. Contaminants from the American Creosote
facility have been found in the groundwater 100 m south of the
facilty. Pensacola Bay is approximately 500 m south of the facility,
in the direct path of the leachate plume.
USEPA ZONE CONTRACTOR STUDY
The USEPA funded a zone contractor study of the Biscayne
Aquifer site in Apr. 1982. The purpose of this study was to identify
the extent of contamination of the Biscayne Aquifer in northwest
Dade County by examining existing data and to propose further
monitoring if warranted. The Biscayne Aquifer site is Florida's
highest priority site for Superfund assistance, since this aquifer is
the sole source of fresh water utilized for water supply in Dade
County. It is a highly permeable, wedge-shaped unconfined surface
aquifer composed of limestone and sandstone which underlies the
entire county. It has an approximate thickness of 30 to 50 m along
Biscayne Bay, and is less than 3 m thick along the western edge of
Dade County. The Biscayne Aquifer is considered the most produc-
tive shallow non-artesian aquifer in Florida and one of the most
permeable aquifers in the world. Average transmissitivity of the
aquifer is about 60 million I/day per meter, the storage is about
0.20, and the permeability is approximately 2.5 million 1/m2.4
Several major wellfields are in the general vicinity of potential
contamination sources in northwest Dade County. These wellfields
include: Miami-Springs (80 mgd), Preston (70 mgd), and Medley
(47 mgd). In addition to these public water supply wells, over 400
private wells are present in this area. The identified sources of con-
tamination to the aquifer in this area of Dade County include: the
Northwest 58th Street Landfill, the Miami International Airport
site, and the Miami Drum Services site.
Northwest 58th Street Landfill
The Northwest 58th Street Landfill consists of a 260 ha area. The
landfill began operation in 1952 as an open dump. Through July
1982 the estimated disposal rate for garbage and trash was about
90,000 tons/month and for liquid wastes was about 1 million
I/month. Resistivity investigations by the U.S. Geological Survey
and Technos, Inc. have determined that, leachate from the landfill
has infiltrated the Biscayne Aquifer beneath and adjacent to the
landfill site.5 The infiltration is in the form of a groundwater plume
moving in an easterly direction along with the natural groundwater
gradient. The close proximity of wellfields to the landfill has
stimulated much concern regarding contamination of these drink-
ing water sources. The landfill is less than 5 km upgradient of the
Miami Springs and Preston wellfields a 2 km from the Medley
wellfield.
This site is not permitted as a sanitary landfill by DER. Accor-
ding to preliminary close-out plans for the landfill, it is classified as
an open dump and is operating in violation of a 1979 consent order
between DER and Dade County. Final close-out plans for this
landfill are now being prepared and this facility is expected to be
closed by the end of 1982.
Miami International Airport
Miami International Airport is located less than 1 km south of
the Miami Springs wellfield. Industrial operations associated with a
typical commercial airport have resulted in hydrocarbon con-
tamination of surface and ground waters over the past years. Since
1966 approximately 15 spills and leaks have been recorded.
The total discharge of hydrocarbon materials is estimated to be
7.5 million liters. This includes the loss of an estimated 5.5 million
Table 2.
Summary of Potential Contaminants by Site
for the Biscayne Aquifer
Site
NW 58th
St Land-
fill
Miami
Drum
Site
Miami
Int'l
Airport
Vol
Org. Heavy Pestl-
Comp Metals PCBs cides
Base/
Neutral Oil &
Phenols Cyanide Exfrac Grease
Priority
Pol'ts
B
A
A
B
B
A
A
A
A—Presence detected in groundwater sampling
B—No groundwater analytical data available but the nature of activities at the site indicates that the
contaminant may be present
C—No groundwater analytical data available and the nature of activities at the site indicates that
it is unlikely that the contaminant is present
liters of Varsol (a petroleum-based cleaning solvent) discovered at
the Eastern Airlines maintenance base. Recovery operations have
removed less than 10% of the estimated spills. No testing has been
done to measure the dissolved fraction of hydrocarbons in water
around the airport. The proximity of the airport to the water sup-
ply wells and several major canals is reason for concern.
Miami Drum
The Miami Drum Services site is approximately 0.5 ha in area.
The site is an inactive drum recycling facility located in a pre-
dominantly industrial area. About .5,000 drums of various
chemical waste materials including corrosives, solvents, phenols,
and toxic metals were observed on the site while the company was
operating. Drums were washed with a caustic cleaning solution,
and the waste solution was disposed of on-site.
Recent electromagnetic studies indicate a groundwater contami-
nant plume associated with the site. Preliminary results from five
shallow monitoring wells indicate that the plume contaminants may
include PCBs, trichloroethylene, dichloroethene, and other halo-
genated organic compounds, as well as heavy metals, oil, and
grease. The plume is moving toward the Miami Springs wellfield
and toward the Medley wellfield which is less than 400 m away.
Biscayne Aquifer Study Results
The first phase of the zone contractor study' concluded that the
toxic contaminants of particular concern include volatile organic
compounds, heavy metals, phenols, cyanides, pesticides, and PCBs
(Table 2). Although no indications were found of contaminant
plumes containing high levels of toxic materials, there is a general
lack of groundwater monitoring data for the area. Further sampl-
ing may indicate the presence of a high-level contaminant plume at
any of these sites. A unified, planned, and intensive sampling effort
is needed to determine the magnitude of groundwater contamina-
tion in order to protect Dade County's drinking water.
COOPERATIVE AGREEMENTS
CERCLA allows USEPA and states to enter into cooperative
agreements for taking remedial actions at hazardous waste sites on
the Priority List. Cooperative agreements allow cost sharing of ap-
propriate remedial activities which are consistent with the National
Contingency Plan. The agreement assigns specific responsibilities
to the state and USEPA for each phase of a site's response actions.
Three phases have been identified for response actions. Phase one
involves investigation and feasibility activities, phase two includes
final design of appropriate remedial alternatives, and phase three
allows for actual implementation of the selected remedial alter-
natives.
-------
298
CASE HISTORIES
Miami Drum
The USEPA and DER have entered into a cooperative agreement
for surface cleanup and a groundwater study at the Miami Drum
Services site. In 1981 Dade County obtained the Miami Drum Ser-
vices site through eminent domain proceedings to use the property
for construction of a maintenance yard for the county's rapid rail
transit system. The county contracted in Dec. 1981 for excavation
and off-site disposal of contaminated soils. The "how clean is
clean" issue had to be addressed. The amount of contaminated
material to be removed was based on analyses of extensive soil-core
borings at the site. Initially, a quadrant of 25 3-m cores spaced over
the site was used to identify contaminated soils. The soil cores were
analyzed every 0.5 m for suspected contaminants. The determina-
tion of the amount of soil removal was based on an assessment of
the soil contaminant data and EP toxicity tests.
In addition to removing obviously contaminated soils based on
bulk analyses, soils displaying EP toxicity parameters concentra-
tions in excess of 10 times the state of Florida "minimum criteria"
for groundwater were excavated. Excavation removed the first 2 m
of soil and up to 3 m of the contaminated portion of the underlying
limestone. Approximately 8,000 m3 of material were removed.
Contaminated soils were transported to and disposed of in the
secure USEPA approved hazardous substance disposal facility in
Emelle, Alabama. A portion of this material was below the water
table; therefore, treatment of the contaminated water encountered
during excavation was performed prior to on-site disposal. The
total cost of this remedial activity was $1,600,000.
A $50,000 remedial investigation and feasibility study of
remedial alternatives for cleanup of groundwater contamination is
planned for the site. Preliminary evidence suggests that con-
taminants have entered the groundwater underlying the site. A
groundwater monitoring study will be implemented in order to
determine the extent of contamination and to identify alternatives
for dealing with the sursurface contamination.
Sapp Battery
The USEPA and DER have entered into a cooperative agreement
for $236,000 for a remedial investigation and feasibility study for
the Sapp Battery Salvage site. The Sapp Battery Salvage site is a 12
ha site located in Jackson County, Florida. Prior to the closing of
the facility in 1980, Sapp Battery Salvage was engaged in recovering
lead from spent case batteries. During the period of operation,
wastewater containing lead, zinc and sulfuric acid was discharged
into settling pits which overflowed to an unlined pond. Spent bat-
tery casings were disposed of in several on-site fill areas. The pond,
along with runoff from the Sapp property, discharges through two
culverts into a natural creek system. Dead and stressed vegetation,
as well as strong sulfurious odors, were noted along the drainage
route from the site. Significant levels of metals including lead,
magnesium and zinc have been found in sediments several miles
downstream from the site.
The USEPA conducted a partial cleanup of the site in Aug. 1980.
The efforts were geared toward raising the extremely low pH of the
pond and berming certain areas to prevent contaminated runoff.
Despite these efforts, runoff continued to move off-site. Since the
shallow groundwater is closely connected to surface waters in this
area, there is a strong likelihood of groundwater contamination.
In order to permanently solve the environmental problems
associated with the Sapp Battery Salvage site, a phased program
has been implemented. Phase one will consist of a remedial in-
vestigation and feasibility study. The remedial investigation will
consist of groundwater and soil testing to define the extent of con-
tamination. The feasibility study will identify the most cost effec-
tive long-term solution.
Whitehouse Oil Pits
The USEPA and DER have entered into a cooperative agreement
for $265,000 for a remedial investigation and feasibility study at the
Whitehouse Oil Pits site. The site is located in the community of
Whitehouse which is west of Jacksonville in Duval County,
Florida. The oil pits were owned and operated between 1958 and
1968 by Allied Petro-Products, Inc., which used a sulfuric acid pro-
cess to recycle used petroleum products. Waste from this operation
was dumped into the pits and included acid sludges as well as waste
oil containing PCBs. The pits were abandoned in 1969 when the
company went bankrupt.
The Whitehouse site was given national attention when it was
identified on an ABC television presentation.' On several occasions
the pit levees have ruptured and spilled contaminants onto adjacent
private property and into a creek. Chemical contaminants included
heavy metals, PCBs and other halogenated organic compounds.
Since the water table is generally within 1.5 m of the land surface,
the oil pits pose a threat to the shallow aquifers used for water sup-
ply purposes.
The City of Jacksonville, DER, and USEPA have attempted to
control pollution from the site by reinforcing the pit dikes,
dewatering the pits, and capping with clay and soil. Recent efforts
have included: liming to increase pH; rerouting of a perimeter
drainage ditch; lining another ditch with clay to prevent leachate
migration; and surface revegetation. Recent site inspections in-
dicate that leachate continues to seep from the covered pits despite
these efforts.
Additional efforts will be needed to achieve a stable site condi-
tion and prevent further migration of contaminants. An initial
remedial measure and an investigation-feasibility study are plann-
ed. The initial remedial measure will be an effort to reduce the
leaching of contaminants into surface water. This will reduce the
dangers to public health and reduce further release of contaminants
into the environment during the course of the remedial
investigation-feasibility study. Additional work will consist of field
investigations including groundwater monitoring. The feasibility
study is designed to identify the most cost-effective long term solu-
tion.
CONCLUSIONS
Florida has recognized the need for an integrated effort to pro-
tect the state's environment and inhabitants from the exposure to
hazardous wastes. Superfund remedial activities initiated by DER
and USEPA during the past year should lead to the cleanup of at
least five of the 16 Florida priority sites. Ongoing enforcement ac-
tivities will result in remedial actions at other priority sites in 1983.
In order to meet the growing demands of hazardous waste remedial
activities in the state, increased state funding is needed for staff,
equipment, and matching funds for USEPA-funded cleanups.' The
mitigation of the adverse effects associated with uncontrolled
hazardous waste sites will require a multimillion dollar annual ex-
penditure.
REFERENCES
1. Florida Department of Environmental Regulation. "Florida Surface
Impoundment Assessment Final Report." Report, January 1980.
2. U.S. Environmental Protection Agency. "EPA announces first 114
top-priority superfund sites." Environmental News, Oct. 23, 1981.
3. USEPA, "Guidance: Cooperative Agreements and contracts with
states under the Comprehensive Environmental Response Compensa-
tion and Liability Act of 1980." Memo, Mar. 1982.
4. Klein, Hoy, and Schroeder. "Biscayne Aquifer of Dade and Broward
Counties." USGS report No. 17, 1958.
5. CH2M HILL, Inc. "Biscayne Aquifer/Dade County: Phase I—Com-
pilation and evaluation of data for the protection of the Biscayne
Aquifer and environment in North Dade County, Florida." Report,
Sept. 1982.
•6. CH2M HILL, Inc. "Biscayne Aquifer/Dade County: Evaluation of
the cleanup activities already undertaken at the Miami Drum Ser-
vices hazardous waste site, Dade County, Florida " Report Sept.
1982.
7. ABC, Inc. "The Killing Ground." TV documentary, Mar. 1979.
8. Florida Governor's Hazardous Waste Policy Advisory Council.
"Hazardous Waste: A Management Perspective." Report, Dec. 1981.
-------
SAFETY AND AIR MONITORING CONSIDERATIONS AT THE
CLEAN UP OF A HAZARDOUS WASTE SITE
D.A. BUECKER
M.L. BRADFORD
Ecology and Environment, Inc.
San Francisco, California
INTRODUCTION
Effective management of a hazardous waste site cleanup re-
quires realistic integration of safety concerns into all operating
procedures. Decisions must continuously be made that not only
minimize the risk to workers and the public, but also promote oper-
ational expediency. Experience is often the best teacher in mak-
ing these decisions.
In this paper, selected decision-making factors and elements
essential to the management of a safety and air monitoring pro-
gram are critically reviewed with respect to the practical experience
gained at the General Disposal Co. clean up, a major removal ac-
tion of a drummed waste facility. Positive features of the safety
program are addressed, as are logistical and organizational diffi-
culties that arose during the clean up. It is hoped .that this review
will help clarify realistic options available to decision-makers when
developing future safety programs.
In the paper, the authors specifically address site background,
organizational management, levels of personnel protection, decon-
tamination procedures, health surveillance, evacuation procedures,
operations safety and air monitoring. Selected hazard assessment
criteria are presented, as are other decision-making criteria used to
justify program elements.
BACKGROUND
On July 10, 1981, a fire erupted at the General Disposal Co.
facility in Santa Fe Springs, Ca., burning nearly half of the esti-
mated 18,000 drums of assorted paint wastes and unidentified
chemicals stored on the one acre site. A series of explosions pre-
vented control of the three alarm blaze until the next morning and
the site continued to smolder for nearly two weeks after the inci-
dent (Fig. 1).
Numerous local, state and federal agencies responded to the
scene, supported by a private clean up contractor. Based on con-
cerns over the continued threat posed by the site, USEPA Reg-
ion IX submitted a request to USEPA-Headquarters and was
granted funding (or emergency removal action under the Com-
prehensive Environmental Response, Compensation and Liability
Act of 1980 (CERCLA or "Superfund").
A unique situation developed after two months of intensive
cleanup effort, when a generator of products found at General Dis-
posal assumed financial responsibility for completion of the clean
up. This led to the termination of USEPA's clean up contrac-
tor in early Oct. and replacement by a second contractor selected
by the generator.
The familiarity of the generator and their contractor with the
majority of products on the site and the USEPA experience on
site allowed for significantly modified operations and a greatly re-
duced work crew. The modifications effectively allowed the second
contractor to implement less stringent personal protection require-
ments. The relationship between site safety monitors and clean
up contractor personnel also varied significantly over the two
phases of clean up operations. The sequence of significant events
as they occurred at General Disposal is shown in Table 1.
Table 1.
Sequence of Significant Events at General Disposal
Date Event
7/10/81 First at General Disposal
7/11-23/81 Site stabilized; fire watch maintained
7/24/81 Superfund implemented; cleanup efforts begin
8/14-27/81 Work shutdown due to lack of further funding
approval
Cleanup begins after shutdown
Work again shutdown
First contractor demobilization
8/28/81
9/10/81
9/28-10/8/81
10/19/81
1/6/82
Second contractor stages equipment and begins
work
Site work completed
The brief summary provided below is intended to highlight some
of the technical operations which occurred over the duration of
cleanup efforts. A site map showing cleanup operations is found
in Fig. 2.
Summary of Technical Operations
•Runoff and sediment from firefighting efforts was collected from
residential streets and returned to the site. Drums and debris
scattered offsite by explosions were also consolidated onsite.
•Site security was provided by the L.A. County Sheriff's De-
partment. Temporary fencing was constructed around the site. A
fire watch was maintained by the Santa Fe Springs Fire Depart-
ment until the site ceased to smolder.
299
-------
300
PERSONNEL SAFETY
Figure 2
Site Sketch of General Disposal on August 14, 1981
•Staging areas were cleared by removing non-chemical debris, in-
cluding old trucks, trailers and dumpsters.
•Drums containing liquid were moved to a staging area and opened
with hydraulic drum handling equipment mounted on a hydraulic
excavator. The drums were numbered and sampled with one
sample being characterized for compatibility and one being sup-
plied to USEPA site monitors for documentation and storage.
After the drum contents were characterized by an on-site lab-
oratory, liquid products were batched in tanks for disposal.
Basic characterization distinguished between flammable/combus-
tible organics and neutral products and in some cases specific
compounds. Any drums found to contain product requiring spe-
cial handling (e.g., PCS, acids, bases or oxidizers) were isolated
for further analysis and overpacking. This process was employed
by the first contractor; the second contractor's familiarity with the
material at General Disposal allowed them to simplify compati-
bility testing procedures. Only materials that were unfamiliar to
site personnel were subjected to detailed analysis.
•Drums with solid materials or those containing charred residues
were ripped open by a backhoe and dumped into a sludging pit.
Sand was mixed with the solid residues and loaded into lined
trucks for disposal.
•USEPA site monitors maintained a photographic log of each
drum to accompany written descriptions. Samples were docu-
mented, packaged and stored in a secured shed for possible
future enforcement actions. These samples were later transferred
to the custody of the California Department of Health Services.
•Empty drums were crushed with a bulldozer (first contractor) or
hydraulic drum crusher (second contractor), placed in lined
dumpsters or trucks and disposed of as solid waste. All drums
and liquids were disposed of at California Class I disposal sites.
•Following removal of all drums, including the overpacks, 6 in.
of topsoil was removed from the entire site. Soil samples were
collected by the California Department of Health Services prior
to the installation of a 6 in. clay cap on the site. Subsequent an-
alysis of the soil samples indicated the presence of heavy metals
and polychlorinated biphenyls which may preclude any future
development of the property.
•Costs and site activities were monitored by the USCG Strike
Team and Ecology & Environment, Inc. USEPA expenditures
between July 24 and Oct. 8, 1981 reached approximately $1.3
million; the generator spent approximately $1 million to complete
the cleanup by Jan. 6,1982.
ORGANIZATIONAL CONSIDERATIONS
A well-defined organizational hierarchy is probably the single
most important factor for instilling a strong safety ethic into oper-
ations during the cleanup of a hazardous waste site. Recogniz-
ing the need for such an organization at General Disposal, USEPA
and the other agencies represented at the site delegated explicit'
safety responsibilities to the Federal On-Scene-Coordinator (OSC)
and, in addition, created the rather unique role of the Operations/
Safety Advisor (OSA).
In accordance with the National Oil and Hazardous Substances
Pollution Contingency Plan, the OSC is to direct and coordinate
all Federal pollution control efforts at the scene of a discharge or
potential discharge. With respect to safety, the OSC at General
Disposal also had specific responsibility for:
•Assuring that appropriate personnel protective equipment was
available and properly used by USEPA, USCG and contractor
personnel in accordance with the site safety plan
•Assuring that personnel were aware of the provisions of the safety
plan, were instructed in the work practices necessary to ensure
safety and were familiar with the planned procedures for dealing
with emergencies
•Assuring that personnel were apprised of the potential hazards
associated with site conditions and operations
•Supervising the monitoring of safety performance by all personnel
to ensure that required work practices were employed
•Correcting any work practices or conditions that would result
in injury to personnel or exposure to hazardous conditions
To assist the OSC in carrying out those duties, the OSA served
as the daily contact for operational advice and safety monitoring
between the cleanup contractors and the OSC. The OSA did not
directly supervise or manage contractor personnel, but rather pro-
vided advice and oversight. The OSA had responsibility to:
•Identify tasks and activities to be performed by the contractor
with respect to the daily work order issued by the OSC
•Provide advice to the drum removal contractor and the OSC on
operational and logistical options
•Conduct or request to be conducted site monitoring of personnel
hazards to determine the degree of hazard present
•Recommend the proper and necessary clothing and equipment to
ensure the safety of operating personnel
•Evaluate chemical hazard information and recommend to the
FOSC and cleanup contractor necessary modifications to work
and safety plans
•Monitor the safety performance of all personnel to ensure that
the required practices are employed
Although the OSA did not have direct line authority, his vis-
ible interaction with the cleanup contractor and ready access to
the OSC encouraged cooperation and adoption of recommended
safety procedures (Fig. 3). The practical advantages of an inti-
mate advisory role are numerous: (1) the OSA position may be
filled by a technical consultant rather than an empowered gov-
ernment agent, (2) the OSA is generally more technically oriented
than the OSC, who must deal more in policy-related matters, and
(3) the OSA can provide succinct recommendations on significant
technical issues and refer policy issues to the FOSC.
Additional safety support was provided at General Disposal by
representatives of the USCG-Pacific Strike Team. An informal
-------
PERSONNEL SAFETY
301
FEDERAL ON-SCENE COORDINATOR IEPAI
N
\
\
N
\
OPERATIONS/SAFETY ADVISOR SA^f™JlO/lITOI"NG
SUPfOHT STAFF
/
/
PRIME CONTRACTOR-OPERATIONS SUPERVISOR
SUBCONTRACTOR-OPERATIONS SUPERVISOR
Figure 3
Organizational Structure at General Disposal (July through October 1981)
communications relationship was implemented where safety viola-
tions observed by USCG personnel were referred through the OSA
for discussion with the contractor and/or OSC. Supplemental
support such as that provided by the USCG greatly enhanced the
overall effectiveness of the safety program.
During the period when USEPA was funding the cleanup ef-
fort, the OSA role remained dominant and influential. However,
after responsible parties assumed funding, the OSA's influence
gradually deteriorated. During the latter phase recommendations
could be made, but without a management incentive the contrac-
tor was not compelled to adopt them. The more detached presence
of the OSC during the final 2-3 months of the cleanup also con-
tributed to the decreased effectiveness of the advisory role.
For future remedial actions, the management organization as
applied at General Disposal is only recommended for Superfund
cleanups where there is a strong Federal or state agency presence
so that the OSA can function between the funding source and the
cleanup contractor. The role of the OSA as discussed above essen-
tially ceases to exist when a third party is funding cleanup activ-
ities. In that event, a safety supervisor with direct line authority
may be necessary.
SAFETY PLAN ELEMENTS
Obviously, personnel and public safety is of paramount concern
whenever remedial actions are undertaken at a hazardous waste
site. However, integration of a realistic safety program into a major
cleanup effort requires constant trade-offs between operational
efficiency and expediency and the protection of worker and ulti-
mately public health. In general, the best safety program can only
be expected to minimize, and not eliminate, the risks inherent in a
site cleanup.
The development of a comprehensive safety program is typically
the end result of a democratic process involving numerous inter-
ested parties. Pragmatically, however, a single individual or agency
must initiate and present a flexible outline upon which refine-
ments can be made. In the case of General Disposal, safety plan
development involved input from USEPA Region IX, USEPA En-
vironmental Response Team, USCG-Pacific Strike Team, Los An-
geles County Sheriff's Department, City of Santa Fe Springs Fire
Department, California Department of Health Services, the prime
and subcontractors, and E & E. E & E provided the funda-
mental format upon which each agency, at various times through-
out the cleanup, contributed or refined selected elements.
The safety program is also one of the first components of the
cleanup that needs to be definitized as soon as the operational
scope evolves. In planned remedial actions, this is not too diffi-
cult. However, in emergency actions of the magnitude of General
Disposal, circumstances often dictate an evolutionary approach to
safety planning. Long before a definitive document can be pro-
duced and distributed, safe operating procedures should be verbal-
ly discussed with the contractor as well as local agencies that may
be involved in an evacuation. A hard copy, however, should
immediately follow, which tends to legitimize the safety proto-
cols and is a valuable reference for the numerous management
personnel on duty rotations.
A safety plan should be specific yet broad enough to accommo-
date the dynamic and unanticipated events that will certainly
occur. A safety plan for a drummed waste removal action should
contain clarification of the following fundamental elements:
•Applicability—to all site personnel, contractors, government
agencies, government agents and visitors (assumption of liabil-
ity should be discussed)
•Responsibility—definition of roles, organizational hierarchy,
site supervision, government liaison, safety requirements, con-
tractor hierarchy (accommodate rotating supervisory person-
nel, flexible definitions, integrate safety responsibilities into man-
agement roles)
•Site description—configuration, size and type of contamination;
services, materials and equipment within contamination zone,
contamination reduction zone, standby zone, clean zone (con-
sider space constraints, hazard assessment, utility and transpor-
tation access, weather variables, security and emergency re-
sponse)
•Personnel/Equipment decontamination—contaminant specific,
sequence of stations, manpower support (consider reuse and stor-
age of protective gear, wet/dry decontamination, laundry, on-
site/offsite cleaning, discharge of contamination wash water or
materials, frequency of use)
•Levels of protection—description of protective clothing and res-
piratory gear for drum handlers and samplers, equipment oper-
ators, ancillary personnel (consider exposure potential, job func-
tion, work station by zone, level of site activity; allow for case-
specific modifications and input from air monitoring results)
•Operations safety—daily safety meetings, buddy system, health
monitoring and first aid capability, weather conditions, fire sup-
pression, safety observers, incident log, heavy equipment, etc.
•Air Monitoring—on-site, perimeter, and off-site; mobile and sta-
tionary, real-time or time-weighted (address objectives, agents,
background influence, onsite operations, episodes, worst case
scenarios, instrument selectivity and sensitivity, monitoring loca-
tions, frequency, duration, recordkeeping)
•Emergency evacuation—on-site personnel and public (consider
worst case and likely events, signal techniques, action levels, fire
department notification, rescue techniques, first aid and emer-
gency equipment, communications, emergency transportation and
routes, hospital/poison control centers, use of mobile laboratory
for rapid chemical identification).
DECISION CRITERIA
For anyone who is routinely in the position of making de-
cisions that influence the health and safety of other people, the
concept of acceptable risk is not new. Such decisions require
constant juggling of dozens of factors, few of which are suffic-
iently quantified or clarified in a timely fashion for the decision-
maker. The consequence of lacking complete knowledge, partic-
ularly when addressing work practices at a hazardous waste site,
is a very conservative approach to worker and public safety.
In few cases is this dichotomy between conservative and exped-
ient approaches more apparent than in the cleanup of a hazardous
waste site comprised of thousands of charred, bulging, and crypti-
cally-labelled drums. Uncertainty over the wastes present, coupled
with the potential for an accident, demands a conservative ap-
proach, as adopted in the first phase of the General Disposal
cleanup.
At General Disposal two distinctly different approaches to saf-
ety were implemented based on the confidence the two funding
sources, USEPA and the generator, had in knowing what the bar-
rels contained. The urgency placed on the cleanup and the level of
effort expended are two other factors that contributed to the dif-
-------
302
PERSONNEL SAFETY
fcrence in the scope and sophistication of the safety programs.
LEVELS OF PROTECTION
The selection of appropriate respiratory protective equipment
and protective clothing can be accomplished after practical con-
sideration of a series of interrelated hazard assessment factors
broadly categorized under two groups—toxicity and exposure po-
tential. Although semi-quantified decision trees are often applied
to this selection process, the authors prefer a more flexible format
consisting of two lists of considerations subjectively prioritized
on the basis of specific site conditions, manpower constraints or
logistical concerns (Table 2). Subjective weighting of the factors by
a safety professional will generally yield a conservative selection of
safety gear or procedures because of the strong likelihood that all
toxicity or exposure factors are not accurately known for all sub-
stances.
Table 2.
Decision-making Considerations for Selection of
Levels of Personal Protection
Toxicity-Related Factors
•chemical agents
•concentrations (background,
episodic)
•dose-response relationship
•physiologic/synergistic
consequences
•TLVs, ceiling limits, STELs
•odor thresholds
•percutaneous characteristics
Exposure Potential Factors
•job function
•proximity to zones of contamination
(work station)
•accident/major release potential
•level of site activity
•physical properties of agents
•frequency of exposure
•route of exposure
•atmospheric dispersion characteristics
At General Disposal, the basic set of protective equipment was
selected on the basis of two dominant criteria, job function and
proximity to zones of contamination. Because the contents of the
drums could not be confirmed and the potential for a major acci-
dental release remained high, the highest level of respiratory pro-
tection (positive-pressure self-contained breathing apparatus) was
selected. Similar rationale was followed for selection of the acid
resistant, PVC, full body coverall. These universally-applied levels
of protection, however, represented a conservative compromise
over what was originally considered adequate.
Initial efforts to apply more exposure criteria and reduce the
clothing and respiratory requirements for selected non-contact per-
sonnel resulted in confusion for the safety monitors and disgruntle-
ment among the workforce. Primarily to avoid misunderstandings
and encourage a spirit of cooperation all personnel in the contam-
ination zone were subsequently required to wear the same basic
outfit regardless of job function or work station (Fig. 4).
Case-specific exceptions to the basic protection levels were made
by the OSA through application of the subjective exposure po-
tential factors in Table 2. For example, at the end of a work day
when all activity had ceased, organic vapor emissions on-site had
dropped to background levels and winds were picking up. a quick
foray onto the site to retrieve some forgotten equipment could
be done with an air purifying respirator and a lightweight dis-
posable coverall.
PERSONNEL DECONTAMINATION PROCEDURES
Large scale remedial efforts generally require a different ap-
proach to decontamination than those applied at short duration
investigative activities such as initial site entries. Procedures must
still be representative of the hazard potential presented by the on-
site materials, however few long-term remedial actions are able to
accommodate personnel decontamination stations of buckets,
shuffle pits, brushes and soapy water.
General Disposal presented several obstacles to such procedures
because of the large number of workers onsite (up to 20), fre-
quent movement on and offsite (every 25 min), the unavailabil-
ity of a sewer discharge or holding pond and the severely restricted
space.
Fortunately.there was no evidence of extremely hazardous ma-
terials onsite that would require extraordinary decontamination
precautions. Consequently, a "dry" decontamination procedure
was applied consisting of four stations: outer protective clothing
removal, shower trailer, dressing trailer and standby zone (Fig. 5).
As teams moved offsite, SCBAs and reusable clothing were
"dropped" at the first PDS station, after which the worker immed-
iately moved to either the standby zone or through the shower
and dressing trailers to the lunch/break area. A pipe rack was used
to hang garments labelled with the workers' names. Most work-
ers found it advantageous to tape outer gloves and neoprene boots
to the PVC coverall for quick donning.
Two sets of cotton coveralls were issued to each worker. At the
end of the day the coveralls were bagged and sent for laundering.
; ON-SITE
( DRUM HANDLERS/SAMPLERS
I EQUIPMENT OPERATORS
I SUPERVISORS
STATION NO 1
• REMOVE SCBA
• REMOVE AND HANG UP OR DISCARD
OUTER PROTECTIVE CLOTHING.
BOOTS. AND CLOVES
• PDC ASSIST ANT PRESENT
STANDBY ARIA
• SHADE CANOPY
• TAKES AMD WNCHIS
• LIQUID REFRESHMENTS
• NO SMOKING OH (ATINO
ALLOWED
STATION NO. 1
• SHOWER TRAILER. SERVING AS
PATHWAY TO STATION NO. 3
DURING DAY
• SHOWER REQUIRED AT END
Of SHIFT FOR ALL SITE
PERSONNEL
STATION NO. 3
• HANDWASHING AND DRESSING
TRAILER
• CLEANUP REQUIRED BEFORE
BREAKS AND LUNCH
OFF-SITE BREAK AREA
• SHADE CANOPY
• USED FOR CONTRACTOR
LUNCH AREA
Figure 4
Figure 5
Personal Decontamination Sequence
-------
PERSONNEL SAFETY
303
The most obvious health concern with a "dry" contamination
procedure is that workers inevitably touch outer surfaces of their
garments during donning and disrobing despite training on tech-
niques to minimize contact. This procedure would not be recom-
mended if extremely hazardous materials were suspected or if cloth-
ing became chronically and heavily contaminated.
HEALTH SURVEILLANCE
Preparation for contingencies at a remedial cleanup depends
upon a realistic evaluation of worst case scenarios and likely events
based on the proposed level of effort and the environmental con-
ditions at the site. Clearly, contingency procedures must antic-
ipate physical and chemical injuries, as well as heat stress inci-
dents when impermeable protective clothing is worn. Distinct emer-
gency and injury prevention networks must be developed for each
of these possibilities, including the development of emergency
transportation systems, identification of medical centers, notifica-
tion protocols, first aid/CPR and protocols for identifying specific
chemical exposures.
The following comments are based on the experiences at General
Disposal. Physical injuries in the form of sprained ankles or backs,
puncture wounds, or broken bones are nearly unavoidable despite
safety awareness and training. The preponderance of jagged metal,
uneven terrain, construction debris, heavy lifting and encumbered
movement due to protective equipment increases the likelihood of
physical injuries.
Significant chemical exposures are probably less likely to occur
due to the conservative precautions already taken in the form of
respiratory and dermal protection. For chemical injuries, on-site
first aid is largely limited to the use of eyewashes, deluge show-
ers and oxygen inhalators. In anticipating a chemical injury, em-
phasis should be placed on location and maintenance of first aid
equipment and developing protocols for its use. An additional
network should be developed for identifying the chemical agent(s)
to which the worker(s) may have been exposed.
At General Disposal real-time air monitoring was available, as
was a mobile laboratory for agent-specific determinations. Safety
protocols required collection of soil, liquid or air samples within
the general vicinity of any exposure. This information would then
be relayed to the attending physician shortly after arrival of the in-
jured person at the hospital.
In the one incident where a chemical injury did occur at General
Disposal, the chemical identification protocols were satisfactorily
activated. However, the chemical results proved questionable. By
the time the individual had been attended to and samples collected,
several minutes had passed. Vapors had subsequently dispersed and
the quality of air had changed, leading to uncertainty regarding
the specific chemicals involved and their concentrations at the time
and place of the exposure. At best this protocol would yield only
an approximation and any results obtained should be relayed to
the physician as only a possible exposure.
Heat stress incidents may not be a health concern at many re-
medial cleanups. However, almost anytime impermeable protec-
tive clothing is worn, significant planning into preventive measures
and emergency response is essential. Practical field techniques for
heat stress management can vary greatly in sophistication. De-
termination of pulse and sweat rates, inner body temperatures, and
application of one of the many heat stress indices may be deemed
appropriate in some cases, while modification of traditional work
schedules may suffice in others.
At the peak of activity at General Disposal, the heat stress man-
agement program was designed around the use of work-rest cycles.
The work day was also shifted toward the cooler morning hours
after the possibility of nighttime work was eliminated as too haz-
ardous due to lighting problems and disturbance to the public.
Generally, teams worked for the duration of-a 45 ft3 air cyclinder
(about 25 min) and then rested for approximately the same period.
Drinking water and dilute electrolyte replacement fluids were pro-
vided during rest cycles. As ambient temperatures increased, longer
periods were devoted to testing and work cycles were shortened.
Despite seemingly liberal rest periods in the shaded standby area
(some approaching 45 min) several mild cases of heat stress—heat
cramps and exhaustion—occurred. As temperatures hovered over
100 °F on some afternoons, the activity of the field crews was
severely reduced, although equipment operators maintained a near-
ly normal pace.
Lack of adequate acclimatization played an important role in the
heat stress incidents. The occurrence of at least two major shut-
downs and the rotation of new unacclimatized personnel into the
work force contributed significantly to the occurrence of heat stress
incidents. Failure to recognize the need for reacclimatization after
the lengthy shutdowns and the lack of a personnel tracking log to
identify first-time workers were probably the greatest deficiencies
in the heat stress management program. Unfortunately, the ur-
gency of the cleanup effort tended to overshadow these needs. At
future sites requiring impermeable protective clothing, it is felt that
a program based on work-rest cycles is practical and adequate
but that a required acclimatization schedule (3-5 days for fit males)
be implemented.
EVALUATION CONSIDERATIONS
The uncontrolled release of chemical vapors or a conflagra-
tion that threatens onsite personnel or the public is clearly a worst
case situation that must be anticipated. At a drummed waste facil-
ity the most likely cause for evacuation is fire and/or explosion
from a spark or chemical reaction, although some scenarios may
also include uncontrolled releases of volatile vapors (ruptured or
leaking barrels can generally be covered or contained before the in-
cident escalates to evacuation status).
Typically, sirens and air horns have been used to warn workers
of a pending disaster. However, at many sites they may have lit-
tle practical merit. This is particularly true at sites where distance,
noise of heavy machinery, respiratory protection equipment and
protective clothing combine to muffle an emergency signal. Dur-
ing peak operating periods at General Disposal the noise levels were
significant. To alleviate this problem a 20 ft observation tower was
erected at the southern end of the site with radio-communica-
tion both to on-site and off-site personnel. From the scaffold, the
observer not only functioned to warn of pending incidents, but also
kept track of individual workers and noted safety infractions. At
least one evacuation drill should be practiced during the first full
week of operation to test the efficiency of the evacuation proto-
cols.
Decisions need to be made on the amount of on-site disaster
mitigation that should take place before an evacuation is ordered.
It was felt at General Disposal that limited firefighting could be
handled by on-site personnel smothering flames or chemical re-
actions with sand or lime. Quantities of both were readily avail-
able. Similar procedures were to be followed should volatile vapor
concentrations reach certain onsite or perimeter action levels. Fire
extinguishers were to be used only for personal protection or equip-
ment fires. The Santa Fe Springs Fire Department approved of the
two step emergency notification protocol and was prepared to
respond to any major emergencies.
An independent community evacuation plan prepared by the
local fire or law enforcement agencies is an integral part of the
safety plan. On-site personnel will not have the manpower or the
authority to order or carry out a community evacuation. The plan
should contain: 1) a description of responsibilities and command,
2) procedures for notifying residents (drive by, air, door-to-door),
restricting traffic and security protocols, 3) evacuation destination
and routes, 4) communication frequencies and reporting proced-
ures, 5) coordination with ambulance and hospital services, and 6)
maps, telephone numbers, and names. Ideally the plan should be
prepared and disseminated prior to any on-site activity, which re-
quires early interaction with the respective local agencies.
OPERATIONS SAFETY
Throughout the daily operation of a remedial action, countless
decisions are made that influence the movement of men, materials
-------
304
PERSONNEL SAFETY
and equipment. Often, many of these are unanticipated or inade-
quately addressed during safety plan preparation. Operations
safety is the one area where most tradeoffs are made between oper-
ational expediency and minimization of risk to the worker and
public.
Each site is unique with regard to operations. Discussed below
are selected operational considerations that impacted safety at Gen-
eral Disposal. Some of the site operations are illustrated in Fig. 6.
Figure 6
•The use of a drum opening bunker constructed with sandbags or
concrete dividers was originally planned, but later discarded be-
cause of the logistical difficulties and increased handling involved
in moving an estimated 8,000 barrels to and from the bunker. The
bunker technique would have essentially isolated each barrel as it
was opened, thereby reducing the explosion potential, but would
have required more handling and contact. An alternative ap-
proach was taken where the barrels were staged in rows of two or
three with 12-18 in. between each barrel. It was felt that the spac-
ing would minimize the domino effect if one barrel exploded
and/or caught fire. This alternative allowed for rapid remote
opening of the drums and eliminated the need to relocate the
barrels.
•All heavy equipment, including rental backhoes and bulldozers,
were fitted with 0.75 in. plexiglass screens to protect the operators.
Inadvertent or deliberate barrel crushing often spewed contents
toward the operators. On numerous occasions the screens saved
them from possible injury.
•To minimize air emissions, crushed drums were removed from the
site daily and batching tanks were covered with polyethylene dur-
ing shutdown periods.
•Airline respirators were used only by heavy equipment operators
adapted to 220 ft1 air cylinders mounted on the equipment. At-
tempts to use traditional airline equipment for workcrews during
the second phase of the cleanup proved to be inefficient and
dangerous.
•Daily meetings were held to brief workers on work assignments
and discuss safety infractions observed the previous day. After
each successive day of working without an accident or injury
there was a tendency to become less cautious and more liberal
in bending the rules. These "tailgate" meetings reinforced safe
work practices and were critical to maintaining the high level of
safety consciousness necessary at the site.
•A variation of the buddy system for on-site work tacitly evolved.
Because many tasks occurred simultaneously (e.g., drum open-
ing, drum sampling, drum documentation, sludging and bulk-
ing), it was not always feasible for workers to work side by side.
Except for drum handlers who always worked in groups of two or
more, workers with less physical drum contact (e.g., air moni-
toring, drum documentation and supervisors) were often not
paired with a buddy. This procedure was justified because of the
relatively small size of the site and the large number of workers.
The overhead observer also acted as an onsite "buddy"
AIR MONITORING
The scope and adequacy of a comprehensive air monitoring pro-
gram is without question one of the most important and contro-
versial elements of the overall safety and health program. Two in-
terests must be addressed: the health and safety of on-site personnel
and the health and safety of the surrounding public. Without this
dual perspective a seemingly adequate on-site air monitoring pro-
gram can seriously fail to address the qualitative and quantita-
tive parameters needed to evaluate the hazard to the unprotected
public. Distinct programs must be implemented for each.
At General Disposal, the task of procuring and maintaining the
instruments and conducting the air monitoring was left largely to
the prime cleanup contractor. The contractor judiciously focused
on air monitoring for personal safety as outlined in the safety plan.
As the cleanup progressed it became necessary to quantify the
offsite emissions to address the concerns of the local health of-
ficials, even though on-site levels indicated a minimal hazard.
Eventually, a perimeter monitoring program was implemented dur-
ing active work periods.
On-site organic vapor emissions (within 2 ft of a source) were
commonly around 10 ppm with rare instantaneous peaks of up to
100 ppm in the drum crushing area or sludging pit. Draeger colon-
metric detector tube analysis indicated that acetone was a common
and predominant species. Perimeter concentrations were generally
less than 5 ppm total organic vapor with peaks (3-5 times daily)
of up to 30 ppm. The final air monitoring scheme is shown in Table
3.
Two sets of arbitrary action levels were established for on-site
and offsite locations. On-site concentrations of 30 ppm total or-
ganic vapor (flame ionization or catalytic combustion detectors)
in the breathing zone would indicate the need to verify the source
and modify operations as necessary. Public evacuation criteria and
strategy were developed by the California Department of Health
Services (DHS) and were based on the sustained presence (a few
minutes) of selected volatile organics at the perimeter fenceline.
The action levels selected by DHS were the threshold limit
values (TLVs) of four compounds alleged to be on-site (benzene,
hexane, methyl butyl ketone, carbon tetrachloride). Sustained
readings of 5-50 ppm total volatile organics at the perimeter on
Table 3.
Air Monitoring Strategy
General Disposal Co. (August 1981)
(refer to Figure 2 for locations)
Location
•drum opening/sludging/
staging area
•mobile monitor around
site and perimeter
•perimeter-north and
south
•drum crushing area/
bulking lank
Purpose/Procedure Instrumentation
evacuation criteria;
action level 30 ppm
total organic vapor;
stationary w/
audible alarm
evacuation criteria/
air characterization;
hourly spot checks
adjusted to changes
in operations
evacuation criteria;
continuous monitor-
ing; adjusted to
wind shifts
air characterization;
agent-specific; lime-
weighted average;
stationary; as
requested
•combustible gas monitor (Oil-
Tech Model 1238 or OXJA)
•combustible gas monitor
(BacharachTLV Sniffer)
•organic vapor detector (Cen-
tury System OVA or HNu
Photoionization Detector)
•colorimetric detector tube*
(Draeger Detection System)
•organic vapor detector (Cen-
tury Systems OVA w/recorder
or HNu Photoionization De-
tector w/recorder)
•sampling train with penonal
sampling pump, charcoal and
Tenax collection media
-------
PERSONNEL SAFETY
305
either the flame ionization or photoionization detectors would
trigger the use of colorimetric detector tubes to verify the presence
of one of the criteria compounds.
Several comments are worthwhile on how the action levels were
determined. The use of conservatively staggered on-site and peri-
meter action levels, in which there is corrective intervention if the
on-site level is exceeded, greatly reduces the chance that conditions
on-site will deteriorate to a point where the perimeter action level
is breached. For example, when on-site levels approached 30 ppm,
the ambient air concentrations at the perimeter still remained far
below the 5-50 ppm action levels for the specific compounds.
This scheme, however, has several drawbacks as applied at General
Disposal. First, the use of the TLV as the criteria for setting an
action level is not appropriate when the consequences are as serious
as a residential neighborhood evacuation. In fact, they clearly rep-
resent a most conservative criteria, being based on an eight-hour
time-weighted average. Furthermore, unless a major fire occurs in
which there would be no question of an evacuation because of the
proximity of residents, vapor emission episodes are likely to be
short and controllable. Despite the presence of more susceptible
elderly or young children in the downwind population, an inter-
mediate action level approaching the short-term-exposure limits
(American Conference of Governmental Industrial Hygienists)
may be more realistic for fenceline evacuation criteria.
Another drawback is that perimeter action levels were based on
specific substances—substances that were suspected yet not proven
to be on-site in significant quantities. Considering the instrument
and operator capability at the site, substance-specific determina-
tions were difficult. Colorimetric detector tube verification of spe-
cific agents would be time-consuming, possibly futile, and subject
to all interference and sensitivity variables of colorimetric detector
tube methodology. An alternative approach could have been to use
a total organic vapor reading instrument with a backup of a port-
able gas chromatograph for species-specific determination. Stand-
ards run previously could be used for baseline comparison. With-
in 10-15 min a chromatogram could be complete and appropriate
action taken (either on-site mitigation or evacuation).
Some of the fundamental decision-making factors considered in
planning an air monitoring program are summarized in Table 4.
CONCLUSIONS
The unique evolution of cleanup efforts at the General Disposal
waste storage facility in Santa Fe Springs, California, illustrates
the breadth of occupational and public safety factors and decisions
that apply to many site cleanups. General Disposal was a valuable
lesson in the differences between textbook approaches to safety
management at waste sites and the functional realities of actually
implementing a comprehensive program. Critical factors which af-
fected implementation of a safety program included the emergency
nature of the cleanup, non-traditional work schedules, unexpected
shutdowns of site work and discontinuity in cleanup contractors.
The relationship between site monitors and cleanup personnel var-
ied significantly over two phases of operations involving differ-
ent prime contractors with different sources of funding.
General Disposal's proximity to residential neighborhoods and
small businesses also created significant concerns relating to public
health and safety. Effective site management therefore included a
broad range of health concerns pertaining to both contractor per-
sonnel and the public at large.
There are several general axioms that grew from the General Dis-
posal experience:
•The willingness to accept risk determines the latitude the safety-
decision-makers will have in designing operating procedures. Gen-
Table 4.
Fundamental Monitoring Applications and Considerations
Application
Evacuation Criteria
Modification of Levels
of Personnel Protection
Air Characterization
and Documentation
Sample Characterization
Consideration
•Occupational and public health focus
•real-time monitoring
•variable action levels
•on-site and perimeter locations
•volatile organics and agent-specific
•stationary, downwind
•occupational health focus
•real-time and TWA monitoring
•arbitrary action levels based on exposure and
toxicity factors
•on-site, worst-case locations
•volatile organics and agent-specific
•personnel and area monitoring
•informational or planning focus
•real-time or TWA monitoring
•episode related, correlate with operation
•on-site, perimeter, and off-site (background)
•volatile organics and agent-specific
•area monitoring
•emergency assistance focus
•real-time monitoring
•quantify to extent possible
•on-site, incident location
•agent-specific
•grab sample
erally, the less risk the responsible parties are willing to take, the
more expensive and resource intensive the cleanup becomes. It is
the role of the safety professional to optimize both concerns.
•Establishment of a safety supervisor with line management au-
thority or a safety advisor with access to supervisory personnel is
essential to a strong, consistent safety program.
•Allocation of off-site/perimeter air monitoring responsibility to
a qualified public agency or contractor independent of the clean-
up contractor is preferred because of conflict of interest and tra-
ditional occupational focus.
•Fundamental safety procedures and air monitoring protocols
should be developed prior to initiation of work, even in emergency
cleanups. The urgency of a cleanup should not be allowed to
supersede basic program elements. Equipment and instrumenta-
tion must be in place with qualified operators before operations
begin.
•Levels of personnel protection should be established equally upon
known or anticipated toxicity and exposure potential factors as
well as realistic accident scenarios. However, speculative worst
case accident scenarios should be avoided when translating them
into operations safety.
•All agencies having interest in the design of the safety program
should be solicited for their input before operations begin, al-
though a single individual or agency will have to draft initial pro-
cedures and absorb the burden of implementing modifications.
•Formal understandings should be developed between site mana-
gers and cleanup contractors regarding the efficient and frequent
transfer of essential data that could influence site or public safety.
This agreement should include air monitoring results, accident/in-
jury data, laboratory operating procedures, waste characteriza-
tion protocols, sample results and performance of protective
equipment.
-------
USES AND LIMITATIONS OF ENVIRONMENTAL MONITORING
EQUIPMENT FOR ASSESSING WORKER SAFETY IN THE FIELD
INVESTIGATIONS OF ABANDONED AND UNCONTROLLED
HAZARDOUS WASTE SITES
CHRISTINE L. MC ENERY
Fred C. Hart Associates, Inc.
Denver, Colorado
INTRODUCTION
Hazardous waste site field investigations pose a risk from haz-
ards such as potentially flammable or explosive wastes, the pres-
ence of radioactive materials, and direct or indirect exposure to
chemical wastes that might necessitate special protective clothing
and equipment for workers. In order to fully assess the hazard
and minimize risk, a comprehensive background data search must
be completed prior to any on-site work.
At some sites, information concerning the types and quantities
of waste present may be obtained from Federal, State and local
officials, manufacturers, and others who live or work in close prox-
imity to the site. However, in many cases adequate data are not
available concerning the waste types, toxicity, or past disposal
practices, especially at abandoned or uncontrolled sites.
At the sites where background information is unavailable or in-
adequate, it may be necessary to characterize the hazards prior to
the field investigation through the use of environmental monitor-
ing equipment such as combustible gas indicators, oxygen meters,
gas detector tubes, radiation detectors, photoionization detectors
and flame ionization detectors. The initial site entry characteri-
zation with these instruments will determine the nature of the
waste present and its hazard potential. Obtaining reliable data from
the environmental monitoring equipment is the most important
aspect of the site characterization with respect to determining the
proper level of personnel protection required at that site.
In order to prescribe the proper level of protection which will
maintain a balance between health, safety, and efficiency, the
monitoring equipment used in site characterization must be re-
liable and applicable to the varied and unique conditions at each
and every site. The lowest level of protective equipment and cloth-
ing necessary to ensure worker health and safety is usually the
most desirable level since the efficiency and effectiveness of the
workers are often greatly reduced when higher levels of personnel
protection are used. Some of the monitoring equipment presently
being used for site characterization was designed more specifically
for other intended uses in the industrial work place and is not
ideally suited for use at a typical hazardous waste site where com-
plete background information is lacking.
Other authors have previously addressed the subject of using en-
vironmental monitoring equipment to assess worker safety, how-
ever, their discussions have been limited to the theoretical versus
the practical application and effectiveness of using this equipment
to characterize hazardous waste sites. The purpose of this paper
is to detail the uses and limitations of the environmental mon-
itoring equipment currently available for use by field investiga-
tion teams and to demonstrate the data gaps which result from
using equipment in the field in a capacity for which it was not spe-
cifically designed.
In early 1980, contractor Field Investigation Teams (FIT) were
established 10 provide nationwide support services to each of the
ten regional offices of the USEPA. As a member of the Fred C.
Hart Associates Field Investigation Team for Region VIII of the
USEPA, the author has gained valuable experience from partic-
ipating in numerous field investigations of abandoned and uncon-
trolled hazardous waste sites. The multi-disciplinary team con-
tributes scientific and management services to identify, investi-
gate and remedy existing uncontrolled and/or abandoned haz-
ardous waste sites. Over 200 investigations have been completed
ranging from site inspections and sample collection to detailed
studies concerning the nature and extent of surface water, ground-
water, air, and soil contamination at a particular site. The exper-
ience gained in the field enables the author to address from a prac-
tical and field-oriented point of view the benefits, drawbacks, and
improvements needed in the equipment currently used to charac-
terize sites and assess worker safety.
In this paper, the author focuses on the uses and limitations of
environmental monitoring equipment to assess the hazards and de-
termine the levels of contaminants present onsite. However, the
actual selection of specific clothing or equipment to protect the
workers from these various levels of contaminants is beyond the
scope of this paper. See references 1 through 6 for more com-
plete information on selecting protective equipment.
Because of the limited availability of monitoring equipment de-
signed specifically for hazardous waste field work, it is hoped that
the information contained in this paper will encourage modifica-
tions to the existing instruments or development of new instru-
ments which will provide more specific and dependable data.
INITIAL SITE ENTRY CHARACTERIZATION
Toxic materials can enter the body in three ways: 1) gastroin-
testinal tract; 2) skin; and 3) lungs. Of these three modes of entry,
the human respiratory system presents the quickest and most direct
avenue of entry because of its intimate association with the cir-
culatory system and the constant need to oxygenate tissue cells.
Therefore, the most significant hazards posed to a field investiga-
tion team at a site, where background information is insufficient,
are related to the ambient air quality. These hazards include oxy-
gen-deficient air, combustible gas, airborne radioparticulates,
and gaseous and/or particulate contaminants. Of these four haz-
ards, the determination of the type and concentration of contam-
inants in the air is the most difficult due to the numerous varieties
of contaminants which may be present and the limitations of the
instruments used to detect their presence.
The initial site entry characterization is performed by first de-
termining the oxygen content of the air with an oxygen indica-
tor, then the explosive potential of the astmosphere is measured
with a combustible gas indicator, and then radiation levels onsite
are evaluated using one or more of the radiation' detection sur-
vey instruments. If there is sufficient oxygen in the atmosphere,
no explosive gases are detected, and radiation levels are concur-
rent with background readings, then the characterization will con-
tinue. Otherwise, work may be halted at the site due to the extreme
and unsafe conditions encountered. The final step in the character-
ization is to determine the type and concentration of contam-
306
-------
PERSONNEL SAFETY
307
inants present in the ambient air using a combination of gas de-
tector units and organic vapor detector instruments.
According to USEPA guidance documents, the initial site entry
characterization is performed with the work party wearing SCBA's
unless field conditions dictate otherwise. Problems have been en-
countered with this blanket rule. First of all at many sites, there
may be a strong political pressure from the client to maintain a
low-key profile while performing investigative field work to avoid
attention, causing undue stress to area inhabitants. This is a diffi-
cult situation to be confronted with and a blanket rule is not al-
ways applicable. Therefore, it is imperative that a field investiga-
tion team be allowed to choose the proper level of personnel
protection.
ENVIRONMENTAL MONITORING EQUIPMENT
There are many instruments and procedures available for meas-
uring concentrations of airborne substances. There is no single,
universal instrument for all such measurements, and there prob-
ably never will be. In fact, the trend seems to be toward devel-
opment of a greater number of specialized instruments.
Instruments and procedures can be classified as follows: 1) those
that give a direct reading; 2) those that remove the substance from
a measured volume of air for later analysis; and 3) those that col-
lect and retain a measured volume of air for later analysis. The
choice of the instrument and procedure to be used depends on
many factors: portability, ease of operation, sensitivity, accuracy,
reliability, availability, the type of information desired and per-
sonal experience.
Grab, or instantaneous direct reading tests, require only a few
seconds to a few minutes to attain results. They indicate fluctua-
tions in concentration of airborne substances and are useful in de-
termining maximum and minimum concentration.
A continuous test or collected sample requires from several min-
utes to an entire day to collect. Such tests give information on the
average concentration of the airborne substance.
There is a need for both grab and continuous methods, as both
give useful information. However, direct reading instruments are
most applicable for the initial site entry characterization because
the information required is gained instantly and can be monitored
for changes throughout the site. However, no matter what instru-
ment is selected, only qualified individuals with complete famil-
iarity with the unit and its operation should be permitted to use the
instrument in order to assure accuracy of the results.
Oxygen Indicator
The normal content of oxygen in the air is 20.9% by vol-
ume. A reading of less than 19.5% is to be considered an oxygen-
deficient environment and all further work must be performed in
air-supplied respirators. Oxygen concentrations below the range of
16-19.5% will not support combustion and are considered unsafe
for human exposure because of harmful effects on bodily func-
tions, mental processes, and coordination. Levels higher than 25 %
indicate oxygen-enriched conditions and all further work is halted
due to fire hazard.
The oxygen indicator measures atmospheric concentrations of
oxygen over a range of 0-25%. Oxygen diffusing through the face
of the galvanic cell undergoes redox reactions which generate a
current proportional to the oxygen partial pressure. The current
is converted to a proportional voltage displayed on the indicator as
a percent of oxygen.
The typical applications for which the instrument was designed
include checking utility manholes, sewers and mines. The read-
ings obtained with the oxygen indicator can be influenced by
temperature, relative humidity, and altitude as well as high con-
centrations of interfering chemicals. High concentrations of strong
oxidants, such as fluorine, chlorine, and ozone will lead to erron-
eously high oxygen readings.
The oxygen indicator is simple to use, compact and practical
for most all field investigations. The model used by the FIT has
the sensing cell in a separate plastic holder at the end of a 10 ft
cable. This extension facilitates the testing of hard to reach areas
before entering. The only notable problem that has been encoun-
tered while using the meter in the field is inconsistency and fluc-
tuation in the readings obtained while dangling the sensor cell from
the body of the meter.
Combustible Gas Indicator
The combustible gas indicator is used to determine the explo-
sive potential of an atmosphere. Although there are many units
available commercially, most test for the concentration of flam-
mable gases and vapors by measuring the heat produced by the
combustion of a test sample. A wide variety of materials are com-
bustible, and a combustible gas indicator does not express their
concentrations directly, but rather as a percentage of the lower
explosive limit (LEL). The LEL is defined as the lowest con-
centration of flammable vapors or gases by volume in air which
will explode, ignite, or burn when there is an ignition source.
The meter must also be capable of determining if flammable
gases or vapors are present in concentrations above the upper ex-
plosive limit (UEL). The UEL is defined as the highest concen-
tration of flammable vapors or gases by volume in air which will
explode, ignite, or burn when there is an ignition source. Although
mixtures of flammable gases or vapors will not explode if they
are present in concentrations above the UEL, these mixtures can
easily be diluted into their explosive range.
While the operation of the combustible gas indicator is relative-
ly straightforward, it is important that the operator be skilled in
interpreting the meter reading correctly. Many physical and chem-
ical properties of the sample and other factors affect instrument
responses. These include oxygen level, relative humidity, the physi-
cal nature of the sample (solid, liquid, or gas/vapor), the presence
of interfering substances, and temperature. Most combustible gas
indicators will not detect an atmosphere which is explosive due to
dusts or mists. They will not record accurately in oxygen-defic-
ient atmospheres and can generate explosions in oxygen-enriched
atmospheres that contain combustible vapors.
In order for the combustile gas indicator to properly func-
tion, the atmosphere must be normal and not one that is oxy-
gen deficient or .oxygen-enriched. Therefore, the oxygen indicator
must be used in conjunction with the combustible gas indicator to
ensure valid results. Any reading on the meter above 10% of the
LEL requires evacuation to reevaluate the work plan. A reading
above 20% of the LEL is immediately dangerous to life and
health, and all investigative work must stop until the hazard is re-
duced.
Leaded gasoline, silanes, silicones, silicates and other com-
pounds containing silicon in the tested atmosphere may seriously
impair the response of the meter. Even minute traces of these ma-
terials can rapidly poison the filament and destroy the sensitivity
of the instrument. This may prove to be a problem at a site where
an explosive gas such as methane is present in conjunction with
possible petroleum product vapors. Two alternative approaches to
testing in the presence of leaded gasoline vapors are to use an in-
hibitor filter or to use a specialized model which was designed for
service where leaded gasoline vapors are present. Another limita-
tion is that petroleum vapors and combustible gases cannot be
differentiated unless a charcoal filter is employed.
The combustible gas indicator is simple to operate, easily por-
table and generally reliable. No major problems have been en-
countered while using it in the field to determine the explosive
potential of the atmosphere.
Radiation Survey Meters
It is imperative to be able to confirm the presence of absence
of radiation levels above background level near the site. If radia-
tion is detected above background readings, then the investiga-
tion is not completed until the advice of a health physicist is ob-
tained.
The radiation survey instrument used by the FIT during an initial
site entry characterization is a personnel radiation monitor, a large
-------
308
PERSONNEL SAFETY
area detector probe used with a portable pulse count ratemeter
and a sensitive, wide range survey meter. The latter two instru-
ments should be used only by persons who have been trained in
the proper interpretation of its readings. These instruments are in-
tended solely for the detection of ionizing radiation.
In addition to the two survey instruments, a personnel radiation
monitor is worn by field personnel during on-site work. The
pocket-sized radiation monitor gives an audible warning when a
gamma radiation field is encountered. It can be easily worn in the
chest pocket of a coverall. This is an effective device for limiting
personnel exposure to radiation both during an initial site entry
and during field activities when onsite conditions can change.
These instruments are used by the FIT to detect radiation lev-
els during an initial site entry characterization. This enables the
safety officer or radiation specialist to discriminate between pro-
ceedings or halting planned field work to ensure worker safety. A
trained radiation specialist is valuable for interpreting the read-
ings, assessing the unique conditions at each site which contribute
to those readings, and deciding if the readings indicate a safe level
for work party exposure.
Gas Detector Tubes
There are many commercially available gas detector tubes and
associated pumps. In general, the tubes operate on the specific re-
action of a chemical compound with a colorimetric reagent. The
detector tube and pump are the two major components of the sys-
tem. Detector tubes are normally species specific. Some manufac-
turers produce tubes for groups of gases such as for hydrocar-
bons in general. Pumps used for drawing air through the tubes
come in two basic forms: bellows pump and piston-type (syringe).
The pumps are manufactured under strict specifications so as to
draw only a specified volume of gas.
Each detector tube has specific detector chemicals which react
with the specifically designated gas being sampled. Each tube re-
quires a set volume of gas to be drawn through by the pump.
Once the proper volume of gas has been drawn, the tube can be
examined. A chemical reaction between the sample gas and the de-
tector chemical is represented by a color change. The length of
the color is proportioned to the concentration.
Gas detector tubes are a prime example of an instrument de-
signed specifically for use in the industrial workplace which is be-
ing used for hazardous waste site field investigations with limited
success. The problems encountered stem from the sharp con-
trasts between the overall conditions found at a hazardous waste
site and the industrial workplace. In the industrial workplace,
there is usually a thorough knowledge of the process, related equip-
ment, raw materials, end-products and by-products which can pos-
sibly create an exposure hazard. As a result the air monitoring
performed at the industrial workplace can be site-specific at a
known location for a known chemical species. The situation at an
abandoned or uncontrolled hazardous waste site will often be close
to the opposite due to a combination of inappropriate waste man-
agement/disposal practices and inadequate background informa-
tion. In addition, there are many substances that interfere with
the color reaction and give false-positive readings. Although these
interferences are usually well documented, the unknown nature of
hazardous waste sites usually precludes absolute knowledge of po-
tentially interfering substances. While the tubes are accurate under
controlled laboratory conditions, the variable and inconsistent con-
ditions in the field contribute to unreliable and limited applic-
ability.
The FIT has been fortunate to have access to additional pieces of
air characterization equipment besides the gas detector tubes. The
effectiveness of an initial site entry characterization would be mark-
edly diminished if the gas detector tubes were used alone. Their
ability to detect ambient air contamination is weak, even at sites
where specific chemicals are suspected or known to be present.
Their applicability for characterizing drummed material is also
weak and limited by the hundreds of tubes which have to be used if
the contents are unknown. Therefore, it is obvious that in many
cases the gas detector tubes cannot provide substantive data to
assess worker safety.
Organic Vapor Detectors
The two types of instruments used by the FIT to monitor for
organic vapors are a portable vapor analyzer with a photoioniza-
tion detector and a portable organic vapor analyzer with flame
ionization detector. Each of these instruments employs a differ-
ent technique to detect and measure the concentrations of trace
gases in the atmosphere.
The photoionization analyzer employs the principal of photo-
ionization for detection. A wide variety of organic compounds
and some inorganic compounds can be measured with this tech-
nique.
The flame ionization detector is designed to detect and measure
trace quantities of organic materials in air by utilizing the prin-
cipal of hydrogen flame ionization. The instrument measures or-
ganic vapor concentration by producing a response to an unknown
sample, which can be related to a gas of known composition to
which the instrument has previously been calibrated. During nor-
mal survey mode operation a continuous sample is drawn into the
probe and transmitted to the detector chamber by an internal
pumping system.
In areas where mixtures of organic vapors are present, it often
becomes necessary to determine the relative concentration of the
components and/or to make quantitative analysis of specific com-
pounds. The flame ionization detector has an available gas chrom-
atographic (GC) column attachment useful for separating mixtures
of gaseous compounds. When the instrument is adapted for GC the
instrument takes a preset volume of ambient air and passes it
through a column containing a material with which the organic
components in the air interact in a non-destructive manner. As the
sample passes through the column the components of the sample
are separated (in time) by their respective different interactions
with the column material. By recording the detector response ver-
sus time on a strip chart recorder a complex mixture can be an-
alyzed.
The instrument was designed for use as a portable survey in-
strument for determining organic air pollutants and for monitor-
ing potentially contaminated areas. In hazardous waste site inves-
tigations, its ability to detect contaminants has varied from site
to site. At an inactive municipal/industrial landfill it was very
effective for detecting compounds in concentrations varying from 1
to 1,000/ppm in the survey mode. In addition, some of these com-
pounds were identified with the instrument in the GC mode. The
instrument was ineffective at an investigation at a petroleum waste
disposal site, where it was unable to pick up the lower molecular
weight compounds associated with petroleum products. This could
have been related to the weather conditions during the survey. Am-
bient temperature was about 40 °F, which is not conducive to
volatilization.
The complementary nature of the organic vapor detector and
the photoionization detector used together in the field, increases
the confidence of the data gathered. However, their effectiveness
for ambient air characterization has been found to be inconsistent
and unpredictable at different sites depending upon the nature of
contamination present.
CASE STUDIES
In the following section two hazardous waste site investigations
where the environmental monitoring equipment was used in the
initial site entry characterization with polarized results are dis-
cussed.
The first site is located in a suburb of Denver, Colorado in a
moderately industrialized area. Between the late 1950s and 1965 a
chemical company operated a pesticide formulation facility on this
site. In 1965, the facility was destroyed by fire and the demolition
debris from the building was disposed of on a vacant lot directly
adjacent to the original site.
-------
PERSONNEL SAFETY
309
Concern was expressed in 1979, when a worker installing a bur-'
ied sewer line at the boundary of the vacant lot complained to
the State health department of noxious odors. Subsequent
sampling by Federal, State and local agencies identified the pres-
ence of a number of pesticides in the soil (Table 1). In Mar. 1982,
the Region VIII office of the USEPA directed the Fred C. Hart
Associates, Inc. FIT to assess the nature, magnitude, and extent of
contamination to the soil, groundwater and surface water on and
originating from the site.
Table 1.
Results from Soil Samples Collected at a Former Pesticide Facility in 1979.
Pesticides (mg/1)
Aldrln
L-BHC
DDT
D1eldr1n
Endrin
Heptachlor
Lindane
Methoxychlor
Sample Number
1
0.15
420
0.8
3
0.4
0.45
ND
1.6
2
0.5
240
8
5
0.7
2.2
ND
6.7
3
1.3
700
60
3
4
1.3
ND
23.5
4
ND
35
0.13
0.4
ND
ND
ND
0.3
5
ND
375
0.4
0.4
0.05
ND
ND
0.4
6
35
700
294
88
447
103
38
116
7
11
ND
152
45
224
85
45
87
8
0.1
ND
0.4
0.5
0.1
ND
ND
0.4
9
ND
ND
ND
ND
ND
ND
ND
ND
10
ND
ND
ND
0.?
0.1
0.1
ND
0.1
ND - Not detected
Prior to commencing field activities, an exhaustive background
information search was performed. Valuable information was ob-
tained and a study plan was developed, including a sampling plan.
An initial site entry characterization was performed to assess ex-
isting hazards. It was necessary to re-evaluate the existing hazards
because over a year had passed since previous fieldwork was com-
pleted and conditions at the site could have changed.
The characterization was not performed in SCBAs because the
highest levels of contaminants previously detected were low enough
to be able to safely wear an air purifying respirator with an organic
vapor/particulate cartridge. Two members of the FIT dressed in
steel toed rubber boots, Tyvek disposal coveralls, surgical gloves,
full face respirators and hard hats. They carried an oxygen in-
dicator, a combustible gas indicator, the radiation survey meters
and the organic vapor detectors. Gas detector tubes were not util-
ized because there are no detector tubes manufactured which will
identify the suspected pesticides.
From the beginning of the site characterization it was under-
stood by the FIT that detection of any pesticides with either of the
organic vapor detectors was highly unlikely due to their low vola-
tility and ionization potential. A manufacturer's representative was
consulted for. each instrument, concerning its applicability for de-
tecting the suspected pesticides. The consultation confirmed the
FIT'S idea concerning the instrument's limitations. However, the
site characterization with the organic vapor detectors was carried
out to monitor for any other contaminants which might be detected
with the instruments.
A perimeter survey was performed initially to assess the site haz-
ards and to indicate localized hot spots for more detailed sampling.
The oxygen content was 20.5%, no explosive gases were detected,
the radiation levels were consistent with area background levels and
no organic vapors were detected above background levels.
The characterization continued and moved onto the site. The
work party began walking with the monitoring equipment in a grid
pattern equally spaced approximately every 15 ft to maximize areal
coverage of the site. Similar readings were obtained throughout
the entire site; none were above the background levels recorded
earlier. The next step was to focus the monitoring equipment di-
rectly around the obvious rubble disposal areas. The piles of rubble
visible above ground were scrutinized with both of the organic
vapor detectors. No vapors above background levels were de-
tected. The rubble piles were then stirred with a long handled stain-
less steel spoon to allow volatilization of any compounds present
which might elicit a response on the instruments, but again there
were no responses above background readings.
Although the hazards associated with the volatilization of the
pesticides were low, there was concern about the exposure risk
associated with inhalation of pesticide contaminated soil particles
in the air, especially during soil sample collection. This is a case
where the results of the initial site entry characterization could have
misled the FIT to believe no exposure potential existed if the back-
ground information was not available. However in this case, be-
cause of the extensive background data, the results of the site char-
acterization reflected the instruments limitations in detecting pesti-
cides versus the site being free of contaminants.
The second_sjte_js located approximately 15 mi southeast of
downtown Denver in a predominantly rural area that is within 3 mi
of a rapidly growing and populated residential area. The site is
1,600 acres in area and contains both an active and inactive land-
fill. From 1966 to 1980 the owner/operator of the landfill ac-
cepted municipal refuse and chemical waste. The solid wastes were
dumped and periodically compacted and covered with earth. The
liquid wastes were poured into unlined trenches and eventually
were filled with trash and covered with earth. In 1980, a new oper-
ator assumed management responsibilities of the landfill as a waste
storer, treater and disposer. The facilities were expanded and im-
proved, three solar evaporation ponds and a drum burial cell were
constructed and segregation of municipal refuse from industrial
waste began.
Odors, arising from the improper management of the landfill in •
its earlier years, became a major issue of concern for residents and
health officials alike. Residents of the area, even as far away as 5
mi, complained strongly and repeatedly to local officials of the ob-
noxious odors associated with the landfill. The number and magni-
tude of complaints prompted the USEPA to direct the Fred C.
Hart Associates, Inc. FIT to assess the extent to which vapors were
migrating and to determine what effects these vapors were hav-
ing on the air quality both on and off site.
A study plan was developed to satisfy the following objec-
tives: (1) to identify the origin of organic vapors at the landfill,
(2) determine the extent to which these vapors were migrating from
the landfill and then, based on the survey results, (3) develop an
ambient air monitoring program for qualitative and quantitative
analysis.
The use of the organic vapor detector was integral to the success
of the investigation from the initial site entry characterization to
the development of a more intensive air monitoring program. The
survey was conducted in three stages. The first investigation con-
sisted of an organic vapor detector perimeter survey of the land-
fill to determine the background levels and ambient air levels of
total organic hydrocarbons possibly traversing the landfill boun-
dary. A second survey consisted of monitoring the ambient air on
the landfill in the vicinity of the active landfill waste disposal
operations such as brine ponds, solar evaporation ponds, sludge
landfarming operations and potential clay liner breaches.
Additionally groundwater monitoring wells on the landfill were
uncapped and subject to organic vapor scans. The instrument
was operated in the GC mode for a number of the monitoring
wells. The instrument was also employed for night surveillance one
mile north of the landfill. Finally, the organic vapor detector was
employed to produce a gas chromatograph strip chart of trade or-
ganic vapors after two puncture holes were induced in the clay cap
at the southwest section of the landfill. The gas chromatograph
verified the presence of vinyl chloride, trichloroethane and trich-
loroethylene.
Under the prevailing meteorological wind regime during the sur-
vey period, no significant values of total hydrocarbons were de-
tected outside the landfill boundary. Background levels of total or-
ganic hydrocarbons were on the order of 5 ppm with a maximum
reading of 7 ppm. Elevated instrument readings of 200-300 ppm
were detected in one of the wells during the survey operation, and
a 350 ppm reading was obtained from the clay cap puncture holes.
Because of the successful organic vapor detector survey results,
the FIT proposed an intense program of ambient air sampling for
the disposal facility.
-------
310
PERSONNEL SAFETY
CONCLUSIONS
In the new and rapidly-growing field of waste management, spe-
cialized monitoring equipment for effective field investigations is
developing slowly. The best available technology for assessing haz-
ards at hazardous waste sites continues to be the equipment devel-
oped for use in the industrial workplace.
Of the major pieces of equipment discussed, three of them:
gas detector tubes, the photoionization detector and the organic
vapor detector should be re-evaluated for their adequacy in assess-
ing the contaminants which may exist at a hazardous waste site.
All three do not detect certain substances. There are not enough
cross checks between the instruments to confirm or substantiate
the results of each individual piece of equipment. However, the im-
portance of this equipment cannot be overemphasized since it is
used to detect what contaminants are present in order to employ
the proper protective clothing and equipment which will ensure
health and safety of the work party.
The author has not suggested the development of a universal in-
strument for performance of a full hazard assessment, but has
rather discussed problems encountered in the field of the existing
equipment. A greater understanding of the problems encountered
with this equipment may encourage modifications which will im-
prove the reliability of the results.
REFERENCES
1. American National Standards Institute, American National Standard,
Practices for Respiratory Protection. ANSI 288.2 New York, N.Y.,
1980.
2. Ecology and Environment, Inc., Course Guide: Hazardous Waste Site
Investigation Training. Prepared for USEPA under Contract No. 68-
01-6056. Arlington, Va., 1982.
3. Mackison, F.W. and Stricoff, R.S., ed., N1OSH/OSHA Pocket Guide
to Chemical Hazards. National Institute for Occupational Safety and
Health and Occupational Safety and Health Administration, U.S. Gov-
ernment Printing Office, Washington, D.G., 1980.
4. Pritchard, John A., A Guide to Industrial Respiratory Protection.
Interagency Agreement Nos. IA-74-23, IA-75-25, and IA-76-9. U.S.
Department of Health, Education and Welfare, Public Health Service,
Center for Disease Control, National Institute for Occupational Saf-
ety and Health, Cincinnati, OH. NIOSH Publication 76-189, 1976.
5. Sax, N.I., Dangerous Properties of Industrial Materials [5th ed.] Van
Nostrand Reinhold, N.Y., 1979.
6. U.S. Department of Labor, OSHA Safety and Health Standards.
29CFR 1910 (revised), Washington, D.C., 1979.
-------
WORKER SAFETY AND DEGREE-OF-HAZARD
CONSIDERATIONS ON REMEDIAL ACTION COSTS
J. LIPPITT
J. WALSH
A. DIPUCCIO
SCS Engineers
Covington, Kentucky
M. SCOTT
SCS Engineers
Reston, Virginia
INTRODUCTION
The passage of the Comprehensive Environmental Response,
Compensation, and Liability Act (CERCLA) in Dec. 1980 estab-
lished the legal and economic basis for a concerted effort by the
federal government to address the environmental and health prob-
lems associated with uncontrolled hazardous waste sites. Among
the requirements of this "Superfund" legislation is a provision in
Item 2 of Section 105 to develop methods for evaluating the rela-
tive costs of remedial actions.
Responsibility for developing the procedures to evaluate remed-
ial action costs was placed with the USEPA in the Office of
Emergency and Remedial Response (OERR) and the Office of En-
forcement (OE). The Office of Research and Development (ORD)
was, in turn, requested to provide support in developing these
procedures. SCS Engineers was then contacted te review, update,
compile, and integrate existing data on remedial act/on costs at
hazardous waste sites.
In this paper, the authors review two SCS projects undertaken
to determine remedial action costs. The basic approach to devel-
oping costs was to identify the remedial action unit operations
which may be combined to provide a remedial action plan. Com-
ponent cost items were identified which contribute to each unit
operation. By determining the cost associated with the component
cost items, the total unit operation cost can be developed. A com-
pleted remedial action plan for a site can be evaluated by sum-
ming the costs of each of the remedial action unit operations in-
volved.
The first project, "Costs of Remedial Actions at Uncontrolled
Hazardous Waste Sites", compiled existing data in terms of scope,
location, and time-frame and provided a methodology for com-
puting the costs of remedial action unit operations. The sources
of available cost information did not incorporate the increased
costs associated with health and safety considerations. Therefore,
a methodology was needed to identify and compute the increased
costs encountered on hazardous waste sites due to health and safety
concerns.
The second project, "Impact of Health and Safety Considera-
tions on Remedial Action Costs", was designed to incorporate the
incremental costs associated with the degree-of-hazard conditions
at uncontrolled hazardous waste sites. Incremental cost factors,
representing a percentage increase in base construction costs at four
different degree-of-hazard levels, would be computed for each unit
operation. The increased cost for a remedial action unit opera-
tion would be determined by multiplying the base costs (com-
puted using procedures outlined in the first project) by the incre-
mental cost factor associated with the degree-of-hazard conditions
at the site during remedial actions.
A related project, "Selecting Among Alternative Remedial Ac-
tions for Uncontrolled Hazardous Waste Sites", was also under-
taken by SCS Engineers to expand on the use of procedures out-
lined on the first project. However, information concerning this
third project is not included in this paper.
DETERMINING COSTS OF REMEDIAL ACTIONS
Remedial Action Unit Operations
Twenty-five remedial action unit operations have been identified
in the course of the two projects. These were derived primarily
from previous USEPA reports.''2'3'4 These remedial action unit
operations are identified in Table 1 and grouped into four general
categories: (1) surface water controls, (2) ground water controls,
(3) migration controls, and (4) waste controls
Cost Variations
Three major factors impacting costs were evaluated in the first
project. These factors were: (1) geographic location, (2) facility
type, and (3) facility size. The 1980 costs for each unit operation
were computed at three separate cost levefs reflecting geographic
locations of: (1) upper U.S. average, (2) lower U.S. average, and
(3) specific costs encountered at Newark, N.J. Facility types inves-
tigated included each of landfills and surface impoundments. Re-
medial action unit operation costs were then examined at each of
five separate scales of operation for landfills and surface impound-
ments. Unit costs appropriate to various remedial actions (e.g.,
dollars/square meter of wall face for a grout curtain) are given in
Table 2. Costs per hectare for direct comparisons among "compet-
ing technologies" are shown in Table 3. These costs reflect the var-
iations due to regional location.
The impact of geographic location, facility type, and facility size
varies for different unit operations. An example of unit operation
costs of a bentonite slurry trench at five hypothetical landfills and
demonstrates the marked impacts of geographic location are shown
in Fig. 1. In contrast, the effect of facility size with unit operation
costs for a well point system at five different sizes of surface im-
poundments is illustrated in Fig. 2.
Table 1.
Remedial Action Unit Operations
Surface Water Controls:
1. Surface Sealing with Synthetic Membranes
2. Surface Sealing with Clay
3. Surface Sealing with Asphalt
4. Surface Sealing with Fly Ash
5. Revegetation
6. Contour Grading
7. Surface Water Diversion Structures
8. Basins and Ponds
9. Dikes
Ground Water Controls:
10. Well Point System
11. Deep Well System
12. Drain System
13. Injection System
14. Bentonite Slurry Trench
15. Grout Curtain
16. Sheet Piling Cutoff
17. Grout Bottom Sealing
Gas Migration Controls:
18. Passive Trench Vents
19. Passive Trench Barriers
20. Active Gas Extraction Wells
Waste Controls:
21. Chemical Fixation
22. Chemical Injection
23. Excavation and Reburial
24. Leachate Recirculation
25. Treatment of Contaminated Water
311
-------
312
PERSONNEL SAFETY
Table 2.
Remedial Action Costs—Appropriate Units*
REMEDIAL ACTION UNIT
! Surface Sealing with Synthetic Membrane
Surface Sealing with Clay
Surface Sealing with Asphalt
Surface Sealing with Fly Ash
Revegetation
Contour Grading
Surface Water Diversion Structures
Basins and Ponds
Dikes
Well Point System
Deep Well System
Drain System
Injection System
Bentonite Slurry Trench
Grout Curtain
Sheet Piling Cutoff
Grout Bottom Sealing
Passive Trench Vents
Passive Trench Barrier
Gas Extraction Wei Is
Chemical Fixation
Chemical Injection
Excavation and Reburial
Leachate Recirculation
Treatment of Contaminated Water
S/m2 A
S/m2 A
$/m2A
$/m2 A
S/m2A
S/m2A
$/m3
S/ha
S/m32 B
$/m
$/m2 B
$/m2B
$/m28
$/m2B
$/m28
S/.28
$/m2A
$/m2 8
$/m2B
$/m28
$/m2A
$/m3C
$/m3 C
$/Ha A
$/l Her/day
Overhead
Allowance
(Percent)
25
25
25
25
25
25
25
25
25
25
25
25
25
25
25
25
25
25
25
25
25
25
25
25
25
Contingency
Allowance
(Percent)
35
35
35
35
10
15
15
20
20
25
30
25
30
30
30
25
40
20
20
30
35
30
40
20
40
Life Cycle (10 Year)
Cost/Unit
U.S. Low
18.59
9.63
6.73
9.59
1.43
1.63
1.75
N/A
N/A
107.29
28.60
59.50
1,760.00
61.41
937.00
73.00
52.96
16.70
47.96
13.9
8.25
2.16
116.00
19,675.00
2.52
U.S. High
27.00
16.33
9.27
16.57
1.81
1.99
3.63
N/A
N/A
152.70
37.17
69.41
1,785.00
106.05
1,881.00
108.00
102.24
24.34
73.03
23.3
14.51
3.81
120.00
24,030.00
4.38
Newark, NJ
22.10
13.50
8.36
13.66
1.70
1.90
3.13
N/A
N/A
139.40
34.48
66.53
1,777.00
95.53
1,643.00
95.00
89.23
21.79
64.75
20.4
12.95
3.38
119.00
22,662.00
3.86 j
i
i
* Assumes a medium-size (5.41 ha) landfill.
m\ = 1.2 yd2
or1 35 ft3
A Surface Area
B Intercept Face
C Landfill Volume
N/A Not Applicable
Derivation of Component Costs
In computing specific remedial action unit operation costs, each
unit operation first had to be broken down into its specific com-
ponent requirements. In addition, each component cost item could
be further defined in terms of subcomponents, with specific costs
assigned to labor, materials, and equipment/supplies. For this pro-
ject, component cost items for each unit operation for both the
landfill and surface impoundment were divided into capital and
operation and maintenance (O&M) cost subcomponents.
Price lists were developed to itemize material, labor, and equip-
ment costs for each of the components used within a unit oper-
ation. For the most part, the 1980 Means and Dodge Guides were
used to obtain the costs needed. These price lists were adjusted
to obtain revised material and labor costs for upper U.S. average,
lower U.S. average, and Newark, N. J. estimates.
Hypothetical Problem
This section has been included to demonstrate the combination
of remedial action unit operations into a total cleanup scenario
appropriate for a given pollution situation. In this example, an
abandoned hazardous waste site has been investigated and found to
have contaminated surface and groundwater. It is decided that the
site must be isolated by: (1) preventing surface runoff from en-
tering the stored hazardous waste, (2) preventing groundwater mi-
gration to the site, and (3) implementing a monitoring program
to confirm the effectiveness of the steps taken. Specific unit oper-
ations required are as follows:
•Contour grading and surface diversion
•Surface sealing
•Bentonite slurry trench cutoff wall (monitoring is included in
this unit operation)
For purposes of this example, it was assumed that the hypo-
thetical site has the following dimensions:
•Surface area = 4 hectare
•Site is square, at 200 m each side
•Average depth of bentonite slurry trench cutoff wall must be 10 m
to impervious material
•Hydrogeologic investigation indicates that the bentonite slurry
trench cutoff wall must extend around three sides of the site to
cutoff ground water migration through the site.
The procedure for estimating the cost of this hypothetical remed-
ial action scenario is as follows:
•Refer to the pertinent tables in the project report and list the ap-
propriate capital and O&M components of the unit operations
selected. This step is shown in the first column of Table 4.
•Refer to the price list in the Appendix of the project report and
determine what units will be required to make sure the cost of
each component (e.g., hectare, m2, etc.). This step is shown in
the second column of Table 4.
•Calculate the number of units of each cost component required
for the site. This step is shown in the third column of Table 4
for the hypothetical site used in this example.
-------
PERSONNEL SAFETY
313
Table 3.
Remedial Action Unit Operation Costs—Constant Units*
Remedial Action
Surface Sealing with Synthetic Membrane
Surface Sealing with Clay
Surface Sealing with Asphalt
Surface Sealing with Fly Ash
Revegetation
Contour Grading
Surface Water Diversion Structures
Sedimentation Basins and Ponds
Dikes
Well Point Systan
Deep Well System
Drain System
Injection System
Bentonite Slurry Trench
Grout Curtain
Sheet Piling Cutoff
Grout Bottom Sealing
Passive Vents
Passive Trench Barriers
Active Gas Extraction Wells
Chemical Fixation
Chemical Injection
Excavation and Reburial
Leachate Recirculation
Treatment of Contaminated Water
U.S. Low
Initial
Capital
$
185,900
96,300
67,300
95,900
13,400
15,300
First
Year
O&M
$
-0-
-0-
-0-
-0-
105
114
Life
Cycle
Cost
$
185,900
96,300
67 3QO
95,900
14,300
16,300
Included in Contour Grading
647
23,120
9,900
3,490
7,830
109,100
1,858,000
146,100
5,282,470
17,150
27,830
19,560
69,100
46,425
12,751,100
5,270
123,800
-0-
1,930
1,820
1,620
20,250
1,600
1,600
-0-
1,600
20
140
1,100
1,600
6,400
1,600
1,700
9,600
647
39,660
25,375
17,150
178,900
122,600
1,871,000
146,100
5,295,970
17,300
29,015
28,900
82,500
59,920
12,764,600
19,700
204,900
U.S. Hiqh
Initial
Capital
$
270,000
163,300
92,700
165,700
16,500
17,900
1,028
36,650
16,240
5,090
9,140
197,900
3,741,000
215,350
0,209,110
24,850
41,690
29,900
130,000
90,830
3,176,000
8,360
209,600
First
Year
O&M
$
-0-
-0-
-0-
-0-
185
240
-0-
2,083
1,970
1,770
20,400
1,750
1,750
-0-
1,750
44
295
2,160
1,750
7,000
1,750
1,850
17,440
Life
Cycle
Cost
$
270,000
163,300
92,700
165,700
18,100
19,900
1,028
56,450
32,980
20,010
181,475
212,600
3,755,800
215,350
0,223,867
25,220
44,180
48,200
145,000
105,590
3,190,000
24,000
356,850
Newark, NJ
Initial
Capital
$
221 ,000
135,000
83,600
136,600
15,700
17,260
909
34,050
14,160
4,570
8,650
176,300
3,264,700
190,300
8,908,930
22,280
37,040
26,620
115,050
79,180
13,059,300
7,300
183,970
First
Year
O&M
$
-0-
-0-
-0-
-0-
157
207
-0-
2,046
1,930
1,730
20,360
1,713
1,713
-0-
1,713
36
253
1,890
1,713
6,850
1,713
1,813
5,470
Life
Cycle
Cost
$
221,000
135,000
83,600
136,600
17,000
19,000
909
51,540
30,600
19,200
180,680
190,700
3,279,150
190,300
8,923,375
22,590
39,170
42.640
129,500
93,630
13,073,760
22,660
314,620
* All units in dollars oer hectare of site surface area.
ha = 0.41 acres
•Refer to the price list in the Appendix of the project report and
list the unit cost of each cost component required. A decision
will have to be made on whether to use upper U.S. average,
lower U.S. average, or Newark, N.J. costs. For this example
scenario, upper U.S. costs were used and are shown in the fourth
column of Table 4.
•The final cost calculation required for multiplication of the num-
ber of units (Step 3 above) by the unit cost (Step 4 above) as
appropriate, and summation of the cost components to arrive at
a total cost. For this example, see the fifth and sixth columns
of Table 4.
For the hypothetical remedial action scenario used here as an ex-
ample, the estimated costs were calculated as follows:
Total Capital Costs =$1,006,300
Total O&M Costs During 10 Years = $ 177,200
Total 10 Year Life Cycle Cost: = $1,183,500
HEALTH AND SAFETY COST CONSIDERATIONS
Identification of Cost Components
Ten basic health and safety cost components were identified
(Table 5) through a combination of previous experience and ob-
servation on hazardous waste sites, contractor discussions, and lit-
erature reviews.5-6'7
Degree-of-Hazard Impacts on Health and Safety Costs
The primary factor in determining the cost of health and safety
for a unit operation on an abandoned hazardous waste site is the
degree-of-hazard associated with the site. As the degree-of-hazard
increases, concern for providing protection of health and safety
also increases. Personnel protective equipment, medical surveil-
lance/services, decontamination, site security, and environmental/
personnel monitoring must be sufficient to prevent excessive ex-
posure. Personnel training, emergency preparedness, and insurance
must be suitable for addressing potential releases of hazardous ma-
terials. Manpower inefficiencies increase with the use of personnel
protective equipment and clothing due to the physically limiting
characteristics of their use. Finally, recordkeeping must provide
documentation necessary to increase information available for
planning appropriate remedial actions, complying with regula-
tions, limiting liabilities, and addressing legal considerations.
The degree-of-hazard associated with an uncontrolled hazardous
waste site is a function of the toxicology of the waste and its phy-
sical and chemical properties in combination with the physical and
environmental characteristics of a site and remedial action activities
on site (Fig. 3). The materials released at an uncontrolled site are
a function of the physical properties of the materials, the environ-
mental and physical characteristics of the site, and the site manage-
ment and control activities. The physical properties and environ-
mental and physical characteristics of the site will determine the
pathways of dispersion. Site management activities will determine
-------
314
PERSONNEL SAFETY
120
UJ
(X
<100
UJ
u
80
0.
o
1C
at
O
u
UJ
u_
_l
UJ
<
cr
ui
60
20
UPPER U.S
LOWER U.S.
12
16
20
WALL FACE AREA IN M (lOOO'S)
Figure 1.
Unit Costs of Bentonite Slurry Trench at Five Hypothetical Landfills
900
ui
IT
< 700
UJ
o
0.
UI
u
cc.
z
« soo
ft.
o
cr
ui
a.
300
_
O
>
o
U,
U.
_i
ui
_L
*OO 800 1 .2OO
INTERCEPT FACE AREA IN M*
1 , 600
Figure 2.
Unit Costs of Well Point System at Five Hypothetical Impoundments
Water Solubil Uy
4 Boil inq Point
«, Melt mq f-oini
1
Factori Which Affect Potential Pathways of Dispersion
2. Soil nd Bedrock Permeability
4 Veqet lion
5 Tempe alure
7 Prec i itat ior
8 Sorfa e Water
A
Factors Which Affect the Amount of Wastes Released
3. Spill Response and Cleanup'
1
of Wastes Released
Chemical Properties and Toxicology
1
2.
)
4
b.
6.
Corrosiv
fteactivi
Flannabi
ftddioacl
Uplosi.
Flash Po
y
ty
Ity
y
nt
;
8.
9.
10.
11.
12
Autoignition Point
Auto Toxicity
Chronic Toxicity
Route of Entry into
Bioaccumulation
Synerglstlc Effects
the Body
Degree-of-Hazard
Health and Safely Hazards
Associated With the Site
Unit Operation
A Spec if ic Remedial
Activity
Action
Methods for Protecting Worker Health and Safety
Procedures, Equipment, Work Practices, and
Related Activities Designed to Identify
and Control Exposures to Hazardous *nd
To* ic Substances
Figure 3.
Factors Impacting Health and Safety Costs
-------
PERSONNEL SAFETY
315
Table 4.
Sample Calculation of Remedial Action Costs for Hypothetical Landfill
Unit Operations and Components
Unit Operation 1. Contour grading and surface water
diversion.
1. Excavation, and recontouring of site, labor plus
equipment
2. Excavation and grading soil, labor plus equipment
3. Diversion ditch
4. Capital cost (subtotal)
5. Overhead allowance (25 percent)
6. Contingency allowance (15 percent)
7. Total capital cost
8. Maintenance and repair cost, diversion ditch two
times per year
9. Total O&M cost (present value of outlays for 10 years)
10. Total life cycle cost
Unit Operation 2. Surface sealing .
1. Excavation and grading waste (Included in unit
operation No. 1; thus not duplicated here)
2. Excavation and grading soil (Included in unit
operation No. 1; thus not duplicated here)
3. Surface seal, bituminous concrete cap (0.08 m thick)
installation and materials
4. Capital cost (subtotal)
5. Overhead allowance (25 percent)
6. Contingency allowance (35 percent)
7. Total capital cost
8. Maintenance and operation cost (none)
9. Total life cycle cost
Unit Operation 3. Bentonite slurry tench.
1. Geotechnical investigation
2. Slurry trench excavation and installation of
bentonite slurry
3. Bentonite, delivered
4. Capital cost (subtotal)
5. Overhead allowance (25 percent)
6. Contingency allowance (30 percent)
7. Total capital cost
8. Maintenance and operation , sample collection
9. Maintenance and operation, sample analysis
10. Total O&M cost (present value outlays for 10 years)
11. Total life cycle cost
Upper U.S. Total Unit
No. of Component Cost Operation
Unit of Units This Cost Per Unit Component Total
Measurement Scenario ($) ($) ($)
w3 0.5 m x 1.83 36,600
40,000 m2
20,000 m3
m3 0.3 m x 1.05 12,600
40,000 m2
12,000 m3
m3 4 m x 2 m 3.63 5,800
x 200 m =
16 ,000 mj
55,000
13,750
8,250
77,000
_ •)
m3 1 ,600 m 3.60 11 ,520
x 2 =
3,200 m3 97,000
174,000
m2 40,000 m2 4.55 182,000
182,000
45,500
63,700
291,200
291 ,200
lump sum 1 6,500 6,500
m3 220 m x 3 x 54.5 359,700
10 m x 1 =
6,600 mj
tonnes 6,600 m3 x 177 45,500
0.039 tonnes/
m3 257
tonnes
411,700
102,900
123.500
638,100
hour 96 hrs/yr 16.60 1 ,600
sample 24 samples/yr 330 7,900
80 ,200
718,300
-------
316
PERSONNEL SAFETY
1. PERSOWEL PROTECTION
a. Levels o' protection
b. individual protection
boots
glove*
eye goggles
Overgarments
head gear
two- -ay conn
cartridge ma
full face na
air-supplied
SCBA
safety line
unlcatlons
sks
Sk
helaets
Tiblc 5.
Worker Safety and Degree-of-Hazard Cost Components
DecOHTWIINtTICM
Personnel
Protective 9eir
Cleanup equipment
StBVlCES/SlFVElLLANCE
prior to SHf work
periodic
follov-up to site work
unscheduled due to Illness, accident, or exposure
On-site
first aid (explosions, animal/insect bites, falls, exposure, etc.)
medical personnel
rescue equipment
emergency showers/eye washes
emergency medical facilities/equipment
periodic or continuous monitoring while working
ff-site
transportation (ambulance)
medical facilities
fire department
life squad
coordination with health/medical services and authorities
3. PERSONNEL TRAINING
a. Waste handling
b. Emergency procedures
c. Medical
d. Protective gear/equipment
e. Monitoring equipment
f. Corrmunicatlon equipment
4, fWowER INEFFICIENCIES
a. Restricted mobllHy
b. Waste handling procedures
c. Monitoring requirements
5. RECORD KEEPING
a. Management requirements
b. Labor union requirements
c. Governmental
d. Medical
g. Special/seasonal equipment
h. Rehearsals
1. In-house training
j. Outside programs
k. Safety and health practices
Buddy system
Pre- and post- work activities
Heat stress
e. Training
f. Work history
g. Site safety/maps
Monitoring equipment
Required facilities, equipment.
structures, etc.
Change rooms/facilities
ITE SECURITY
Restriction out of access by personnel or vehicles
Signs and tags
Security personnel
Security systems/equipment
Coordination Mlth law enforcement agencies
NSURANCE
Medical
Liability
Workman's compensation
Unemployment
EMERGENCY PREPAREDNESS
a. Fire fighting
• extinguishers
t turnout gear
Spill containment/control
• absorbents
t oil booms/containment devices
10. HA2ARD ASSESSMENT
a. Sample packaging and shipping requirements
special containers
special packing materials
labels, markings, and placards
authorized transporters
b. Monitoring equipment
oxygen detectors
combustion gas
organic vapor analyzers
radiation detectors
field test kits
specialized laboratory equipment
amp]ing equipment
disposable sampling equipment
special material construction
specialized equipment
nalytlcal costs (field and contract laboratories)
qualitative
quantitative
evlew and Interpretation of data
establishment of levels of protection
assessment of contaminant migration
source Identification
the amount of materials released. The materials released in com-
bination with the chemical properties and toxicology of the mater-
ials will determine the degree-of-hazard associated with the site.
Actions taken at an uncontrolled hazardous waste site during the
course of remedial action unit operations will require a level of
worker protection appropriate for the degree-of-hazard conditions
existing on the site. The cost of worker protection and related
component cost items (e.g., medical surveillance, personnel train-
ing, etc.) will increase as the degree-of-hazard is not a static con-
dition and may change with environmental conditions (e.g., air in-
versions, temperature variations, and precipitation), site activities
which generate increased releases (e.g., spills, dusting, and vapor
releases), and combinations of both environmental changes and site
activities (e.g., reducing pressure within a container by venting off
vapors during an air inversion).
Other factors which can impact health and safety related costs
include: (1) social and political awareness and concern, and (2)
protected environments (e.g., nature preserves, national parks,
state parks, and historical sites). For the purposes of this project,
only the costs of worker health and safety protection are being
examined. By examining the costs at the point of greatest con-
tamination (i.e., on-site) most, if not all, of the cost increases
should be incorporated into the cost calculations.
DEVELOPING INFORMATION FOR
HEALTH AND SAFETY COSTS
Availability of Health and Safety Cost Information
Contacts were made with hazardous waste site cleanup contrac-
tors, regulatory officials, and other consulting firms involved in
related USEPA projects. The information obtained was primarily
general cost information on equipment and materials. Specific
prices with reference to suppliers and manufacturers were pro-
vided for equipment and clothing costs. Costs associated with
equipping an individual worker for four levels of personal protec-
tion as defined by the Interim Standard Operation Safety Pro-
cedures developed by the USEPA Emergency Response Team is
found in Table 6.8-9 Some preliminary estimates for health and
Table 6.
Costs for Equipping a Worker at Four Levels or Protection!*.]
Level A—$1,854
Includes:
Fully-encapsulating chemical re-
sistant suit
Pressure demand, self-contained
breathing apparatus
Gloves, inner, chemical resistant
Cloves, outer, chemical resistant
Boots, chemical resistant, steel
toe and shank
Hard hat
Boots, outer, chemical resistant,
disposable
Level B—$1,136
Includes:
Chemical resistant clothing-
jacket, bib overalls
Pressure demand, self-contained
breathing apparatus
Cloves, inner, chemical resistant
Cloves, outer, chemical resistant
Boots, chemical resistant, leggin
Boots, chemical resistant, steel toe
and shank
Boots, outer, chemical resistant,
disposable
Hard hat
Level C—$736
Includes:
Sara-coated disposable suit
Full face air purifying cannister
respirator:
-front belt mounted
-back belt mounted
Gloves, inner, chemical resistant
Gloves, outer, chemical resistant
Boots, chemical resistant, leggin
Boots, chemical resistant, steel
toe and shank
Boots, outer, chemical resistant,
disposable
Escape mask
Hard hat
Level D—$260
Includes:
Disposable coveralls
Gloves, outer, chemical resistant
Boots, chemical resistant, steel toe
and shank
Escape mask
Hard hat and face shield
-------
PERSONNEL SAFETY
317
safety costs for individual cost components are shown in table 7.'°
Determining the health and safety costs for actual site cleanups
was difficult. The number of variables that enter into the degree-
of-hazard and required health and safety considerations (Fig. 3)
make comparisons of costs difficult. Additionally, information on
costs has traditionally not been collected in a way that enables
health and safety costs to be distinguished from other costs. Sim-
ilarly, more recent cost information available from documents does
Table 7.
Preliminary Estimates of Health and Safety Components Costs [10]
Health and Safety Cost Components
•Baseline Physical
•Medical Coverage
•Personnel Protection (includes clothing
and monitoring equipment)
•Decontamination (capital)
•Decontamination (operational and
disposal)
•Site Security (general)
•Site Security (guard service)
•Training—new employee
•Training—after 6 months
•Recordkeeping
•Manpower Efficiencies—losses (pri-
marily affected by temperature
extremes and level of protection)
•Insurance—Liability
•Insurance—Workman's Compensation
•Insurance—Bonding
Preliminary Estimate of Costs
$ISO/person
3 to 10% of base pay
$4,100to$5,000/person
$12,000 to $15,000/site
$100/day
1 to 5 % of bid price
$10/hour
5 to 10% of activity
1 to 5 % of activity
1 to 5% of working hours
25 to 40%
$lto$2/$100ofsales
$1 to$25/$100ofwages
20 to 100% of bid price
not provide sufficient detail to identify specific health and safety
cost items. As indicated by the variations in bids submitted for a
site cleanup under Superfund, various components appear to be
included under different categories by different contractors. Bid
quotations submitted for the site in question are given in Table 8.
As a result, the approach used on this project was developed to
obtain the necessary cost information in a consistent format which
would allow for comparisons between different sites by controlling
as many variables as possible.
Development of Uncontrolled Hazardous Waste Site Scenarios
Hazardous waste site scenarios were developed to be represen-
tative of three basic types of sites: (1) subsurface burial, (2) sur-
face impoundments, and (3) above grade storage. Whenever pos-
sible these scenarios were developed by reference to actual clean-
up operations either completed, in progress, or planned for the
near future. This approach was adopted partially to ensure that
the scenarios would reflect realistic conditions, and partially to
assist in identifying organizations and individuals experienced in
particular types of cleanup operations.
Each scenario was composed of a number of distinct unit oper-
ations. For each scenario, site characteristics (e.g., topography,
size, weather conditions, hydrogeology, etc.) were defined to pro-
vide a detailed profile of the site. Similarly, the characteristics
of the waste at the site were defined so that the levels of protec-
tion could be determined for each unit operation. The levels of
personnel protection used for the project were those defined by
USEPA in their Interim Standard Operating Safety Procedures. A
brief description of the conditions associated with the levels of per-
Table 8.
Bid Quotes for Cleanup of a Superfund Site
Bids
A
B
C
0
E
r
G
H
I
J
K
Average
1
Insurance
Bonds,
Permits
$18,143
$13,625
$21 ,500
$16,000
$19,000
$ 1,500
$26,470
$20,000
$11,250
$10,000
$60,000
$19,767
2
Other
Project Start-
up and Site
Services
$
$
$1
$
$
$
$
$
$
$
$
$
193,762
236,628
,048,679
600,000
552,000
446,300
190,000
992,248
475,581
101,600
185,000
456,527
3
Site
Preparation
$161.680
$ 80.310
$ 34,958
$345,835
$290,000
$ 60,000
$130,000
$296.500
$187,431
$112,100
$207.753
$173.324
4
Liquid
Material
Disposal
$135,927
$493,072
$165.590
$199,312
$266.812
$154,677
5
Solid
Material
Disposal
534,640
423,392
536,683
522.760
532,000
659.334
$173,250 $1,280,200
$ 33,025 1
$203,960 1
$211,272 )
$336,247 J
$215,740 J
161,950
297,522
442,340
241 ,000
511,984
6
Salvage
of
Motors
$ 2,327*
$ 4,310*
$ 3,590*
$ 431*
$ 4.310*
$ 8.620*
$ 4,310*
$ 1,077
$ 8.620*
$ 2.004
$43,100*
$76,537*
7
Salvage
of
Scrap
Steel
$ 9
$ 4
$ 5
$
$ 2
$ 4
$ 3
$ 2
$ 2
$ 2
$ 4
$21
,400
,000*
,000*
200*
,000*
,000
,000*
.000
,000*
,000
,000*
,600*
8
Disposal
of Drummed
Material Cur-
rently on Site
$ 1
$ 1
$ 1
$ 1
$ 4
$
$ 1
$ 4
$ 1
$
$ 4
$22
,080
,045
,179
,440
,200
720
,680
,200
,308
900
,500
.252
$1
$1
$1
$1
$1
$1
$1
$1
$1
$
$
$1
Total
,052,305
,239,762
,799,999
,684.716
.657.702
.309,911
.794,420
.511,000
.166,432
882,216
987.400
,371.442
* Credit
Table 9.
Conditions Associated with Levels of Personnel Protection [9]
1. Level A—require full encapsulation and protection from any body con-
tact or esposure to materials (i.e., toxic by inhalation and skin ab-
sorption).
2. 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 stand-
ards (i.e., below harmful concentrations).
3. 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.
4. Level D—no identified hazard present, but conditions are monitored
and minimal safety equipment is available.
5. No hazard—standard base construction costs.
sonnel protection is given in Table 9.
Contractors were identified with particular experience in each
type of scenario. Cost information was then solicited from se-
lected contractors based on their particular experience, availability
of qualified personnel, and their willingness to provide the neces-
sary information. The information initially requested was a break-
down of health and safety costs (using the cost components shown
in Table 5) for each unit operation in a scenario, along with the
baseline construction costs excluding any such health and safety
costs. This approach was intended to obtain incremental health
and safety costs for specified levels of personnel protection asso-
ciated with the unit operation and the waste being handled.
To extend the usefulness of the data from each scenario,
further information was sought on health and safety costs for each
unit operation at different levels of protection. For example, if
the scenario under consideration included the unit operation of ex-
-------
318
PERSONNEL SAFETY
cavation and reburial for dioxin wastes at a level of protection
designated "A", then the additional costs developed would be for
the same operation involving less hazardous wastes at levels desig-
nated "B", "C", and "D".
A further important refinement was to consider the effects of
variations in site conditions, particularly weather conditions.
Where the impacts of varying site conditions were identified as
significant for extremes of heat and cold, these incremental costs
were also investigated as an extension of the scenarios.
The data collected using these scenarios will be used to estab-
lish a health and safety incremental cost factor for each unit oper-
ation identified at each of the four degree-of-hazard levels asso-
ciated with the four levels of personnel protection. The incremen-
tal cost factors can then be used to adjust the remedial action con-
struction costs determined in the first project to properly reflect
health and safety cost considerations.
CONCLUSIONS
In current decision making processes, concern for cost-effec-
tive use of funds is a major issue in political, economic, and
budgetary considerations. The limitation of available Superfund
monies to address the problems associated with uncontrolled haz-
ardous waste sites necessitates cost-effective evaluation and selec-
tion of remedialf actions. This need has been reflected in the re-
quirements of the Superfund Act to provide methods for an analy-
sis of the relative costs of alternative remedial actions. The stud-
ies discussed in this paper were designed to assist in fulfilling these
requirements.
The primary result of these studies will be a costing methodology
which can be consistently applied to the various unit operations
and reflect variations due to geographic location, facility size,
and health and safety concerns. The resulting cost estimates would
enable regulatory officials, consultants, and contractors to: (1)
compare costs for alternative unit operations performing the same
function, and (2) compute cost estimates for combinations of unit
operations comprising complete remedial action plans. Other use-
ful applications might include:
•Design of a specific unit operation. For example, cost data gen-
erated by this project could help determine the location of a
bentonite slurry wall. Should it be placed on-site where excava-
tion could require installation at Level C personnel protection?
Or would it be more cost-effective to move further down-grad-
ient where a longer wall would be required to contain a dispersed
plume, but personnel protection is not required?
•More study versus start of cleanup actions. For example, should
remedial action proceed immediately without the benefit of de-
tailed waste analyses, assuming that worst-case Level A protec-
tion is required? Or should further sampling and analysis be per-
formed if there is a possibility that results will allow a lessening of
personnel protection to Level B or C?
ACKNOWLEDGEMENTS
The projects described in this paper were performed under
USEPA Contracts Nos. 68-01-4885, Directive of Work No. 12, and
68-03-3028, Directive of Work No. 14. The authors would like to
thank the USEPA Project Monitors, D. Ammon and D. Banning,
of the USEPA Municipal Environmental Research Laboratory,
Solid and Hazardous Waste Research Division in Cincinnati, Ohio.
REFERENCES
1. A.W. Martin and Associates, Inc., "Guidance Manual for Min-
imizing Pollution from Waste Disposal Sites," EPA-600/2-78-I42,
USEPA, Cincinnati, Oh., 1978.
2. Fred C. Hart Associates, Inc., "Analysis of the Technology, Prev-
alence, and Economics of Landfill Disposal of Solid Waste in the
United States," Volume II. USEPA, Washington, D.C., 1979.
3. Geraghty and Mifler, Inc. "Surface Impoundments and their Effects
on Ground Water Quality," USEPA, Washington, D.C. 1978.
4. JRB Associates, Inc., "Remedial Action for Waste Disposal Sites:
A Decision Makers Guide and Technical Handbook," USEPA, Wash-
ington, D.C., 1980.
6. National Safety Council, Fundamentals of Industrial Hygiene, Chi-
cago, II., 1977.
7. "The Industrial Environment—Its Evaluation and Control," Publica-
tion No. 614, U.S. Department of Health, Education, and Welfare,
U.S. Government Printing Office, Washington, D.C. 20402, Stock
No. 017-001-00396-4 1973.
8. USEPA, "Emergency Response Team Interim Standard Operating
Safety Procedures," 1982.
9. Nash, Robert S., personal communication, IT Corporation, Cin-
cinnati, Oh'., Sept. 1982.
10. Dalton, T.F., personal communication, Hazchem Services, Garwood,
New Jersey, Aug. 1982.
-------
HAZARDOUS WASTE SITE INVESTIGATIONS:
SAFETY TRAINING, HOW MUCH IS ENOUGH?
STEVEN P. MASLANSKY
Geo-Environmental Consultants, Inc.
Port Chester, New York
INTRODUCTION
With the exception of some state and Federal employees, special-
ity consultants and spill cleanup contractors, many individuals
associated with inspection, evaluation, and remedial actions at un-
controlled hazardous waste sites lack adequate safety training. The
investigation and possible subsequent cleanup of a site containing
known or suspected hazardous material, whether funded privately
or with governmental monies, is a complex task involving a diverse
group of engineers, scientists, technicians, and contractors. Safety
training and other operational considerations, including medical
monitoring, are often perfunctory or non-existent for those in-
volved in sample collection, geophysical surveys, or subsurface in-
vestigations.
With greater emphasis being placed on remedial action by private
industry, local government or through the "Superfund" program,
more and more of the construction trades will be found on hazard-
ous sites. The employers have both a legal and moral obligation to
ensure that their employees are prepared to undertake work at un-
controlled sites. In particular, employers or their representatives
should satisfy themselves that those they have employed have been
thoroughly screened and trained for on-site work. In order to max-
imize the needs of the individual and minimize budget and time re-
strictions a tiered training approach is usually the most efficient.
TRAINING APPROACH
Many individuals employed by the government or speciality con-
sultants and contractors have been given some degree of training
in accordance with their level of responsibility. Others receive only
on-the-job training (OJT).
The USEPA, for example, now requires that all of their em-
ployees engaged in field activities undergo a training program and
be certified to a level commensurate with the degree of anticipated
hazards. Their training program is divided into three categories.
The Basic Level requires a minimum of 24 hours of health and
safety training which includes emergency help and self-rescue, and
the safe use of field and personnel protective equipment. The In-
termediate Level adds an additional 8 hours or more training in
such areas as safety plan development and decontamination. The
Advanced Level includes an additional 8 or more hours with em-
phasis placed upon management of restricted and safe zones, pro-
cedures for dealing with the public and the media, and the safe use
of specialized sampling equipment. The training is further en-
hanced by a required OJT period and supplemented by additional
training as deemed necessary.
For others who work with hazardous materials, however, these
training programs, as well as other more intensive training pro-
grams, are either unavailable or unobtainable. Also, although
numerous volumes are available to outline textbook approaches to
safety during field investigations, oftentimes the procedures out-
lined for optimum personnel protection may not be workable in
actual field situation. This is especially true from the standpoint of
minimizing the potential conflict between the safety of the inves-
tigators and any undue anxiety among individuals of neighboring
residential or industrial areas. Organizations with limited training
budgets and/or resources must approach the training require-
ment from a pragmatic standpoint.
All individuals working at hazardous waste sites have different
roles and levels of responsibilities. Organizations perform different
functions at sites; and site hazards and conditions vary greatly both
between sites or portions of the same site. Training must there-
fore be organized for the specific needs and levels of compre-
hension of the individual, their organizations, the work assign-
ment of that organization, and the specific requirements of a par-
ticular site.
BASELINE CONSIDERATIONS
All individuals required to have access to a site containing known
or suspected hazardous materials must undergo baseline medical
profiling. The specific elements of the medical program are a func-
tion of the individual's duties and the need to maintain a balance
between site, safety and an individual's right to privacy. How-
ever, it is not unusual for a worker to feel that his job may be in
jeopardy by the disclosure of medical test information.
Medical profiling normally includes the following tests: pul-
monary function, blood chemistry, urine analysis and liver and
kidney functions. These tests, in conjunction with a thorough
medical history, can serve as indicators of possible preexisting ab-
normal conditions or unusual sensitivities to certain types of toxic
compounds. Other tests which might be appropriate include chest
X-ray, electrocardiogram and specific tests for exposure to certain
chemicals or physical agents. Medical profiling should be repeated
yearly (as a minimum). Exit physicals should be given to any work-
er terminating his employment.
OVERVIEW TRAINING
The first level of a recommended training program is overview
training. Overview training serves as an introduction of the worker
to the nature and types of hazards he or she might encounter in
the field. A number of generalized hazardous material short
courses are given by government agencies, educational institu-
tions, professional societies and consulting organizations.
Although normally well received, these courses, in many in-
stances, fall short of satisfying the basic needs of individual work-
ers outside of the sponsoring agency. For example, a hazardous
material emergency response course given by FEMA's National
Fire Academy is excellent for firefighters; however, it is not ap-
propriate for those engaged in field investigations or remedial
measures.
The overview training then, even though basic in nature, must
be designed for a specific audience, its needs and its level of com-
prehension, in order to be successful. Course length is a func-
tion of the needs of the organization. The workers of a geophysi-
cal firm may need less training than those employed by a well drill-
ing contractor, due to the nature of anticipated hazards and ex-
posure risk. Of course, budgetary constraints are an unfortunate
consideration.
Course length for overview training normally runs from 2 to 5
days, with little or no time for hands-on training. Overview train-
ing programs for emergency response and military personnel, how-
ever, run two weeks or longer in order to familiarize the student
with protective equipment before intensive training commences.
Class size is limited by the organization's needs or available
resources. It is best to try to limit class size to a maximum of 25
students.
Large organizations in industry or government who oversee or
participate in investigations or remedial actions should have all
personnel involved in the program attend the training. In addition
319
-------
320
PERSONNEL SAFETY
to field workers and supervisors, administrative, legal, and tech-
nical support personnel should be involved in order to better appre-
ciate problems encountered in the field.
The contents of a typical four day overview course for an organ-
ization with responsibility in overseeing remedial actions at aban-
doned uncontrolled hazardous waste facilities is given in Table 1.
Most of the topics can only be touched upon within the time frame
allowed. Since only one day has been allocated for personnel pro-
tection, no more than a preliminary treatment of equipment selec-
tion, fitting, use and maintenance can be given.
The course is designed so that a variety of the organization's
employees from field personnel to specification writers can profit
from the course. Although the topics are geared to the organiza-
tion's responsibility and protective' equipment already available,
much of the topics covered will be found in existing courses. For
many, overview training or overview training with a little OJT with
a more experienced worker is all that is needed. For others, disci-
pline intensive training followed by detailed site specific training
may be required.
DISCIPLINE INTENSIVE TRAINING
Certain supervisory, technical support, or field personnel may re-
quire intensive training in specific areas such as incident manage-
ment, respiratory protection, radioactive material handling, safe
sampling and decontamination techniques, etc. It is normally be-
yond the capabilities of most organizations to develop their own
courses but fortunately, just as there are numerous basic courses,
there are also many excellent intensive training programs given by
many of the same groups. On the other hand, many of these
courses are restricted to the sponsoring organization employees
or to only government employees.
As an aid in determining what courses are available, many organ-
izations are preparing directories and catalogues. The USEPA,
FEMA, Corps of Engineers, and the National Environmental
Training Association are presently preparing listings of available
hazardous material courses. Although for use primarily by the
respective organizations, the information, particularly the
USEPA's, which is the most comprehensive, should be made avail-
able to interested parties.
SITE-SPECIFIC TRAINING
The final required training, and probably the most important,
is site-specific training. During this phase the individual receives
detailed training in the actual conditions and hazards which may
be encountered. Depending on the amount of information regard-
ing on-site material, the worker may only be instructed in the haz-
ards and appropriate responses to a limited set of conditions. For
example, a site may be known to contain only PCB contaminated
soil. The workers' site specific training would stress the hazards of
PCBs and the degree of protection required for anticipated con-
centrations to be found in water, soil and surrounding air. It is,
however, assumed that all employees have received some degree
of overview training prior to being tasked with a field assignment.
Site-specific training may last just a few hours for a simple site or
several days for a complex site preparing field workers to properly
utilize and maintain site measuring devices and protective clothing.
Actual respiratory protection devices to be worn are fit-tested and
the individual given sufficient training in their use under both
normal and emergency working conditions during this phase. It is
important that site-specific training be given well in advance of
actual site work for a complex or high hazard site. Not everyone
can wear or tolerate wearing protective equipment, particularly
the fully encapsulated suit. Adequate time must be budgeted be-
tween training and actual on-site work to secure and train replace-
ment workers if necessary.
Some site-specific safety training must be accomplished at the
site, such as review of emergency escape routes, emergency com-
munications, and location of emergency equipment. During this
phase the worker should also become familiar with the site's Per-
sonnel Decontamination Station, with rehearsals conducted using
non-toxic materials.
ADDITIONAL CONSIDERATIONS
Two other critical elements in safety training are first aid and
physical fitness. All on-site workers, including support personnel,
should complete a basic first aid program prior to on-site work. As
a minimum, each worker should pass the American Red Cross or
equivalent 8 hour standard First Aid course and the Cardiopul-
monary Resuscitation Course. These courses are readily avail-
able and inexpensive. They can be incorporated into an overview
training program or taken separately. First aid procedures for an-
ticipated site specific materials are given during the site specific
training phase.
Finally, even though a person has passed medical profiling and
all prerequisite training, he or she may still be unsuitable for on-
site work. Field investigation and remedial action work is in itself
both physically and mentally demanding. Work in cumbersome
protective clothing can be performed safely only by those in top
physical condition. It is critical that individuals engaged in on-site
work be encouraged to maintain a program of physical fitness.
CONCLUSIONS
How much training is enough? Obviously one cannot be over-
trained when dealing with hazardous materials. One might argue
that a worker can never be fully trained when dealing with mixed
hazardous wastes on a complex site. Training must progress
through a tiered approach, evaluating the needs of the individual,
his job function and level of comprehension; the organization he
works for and their project responsibility; and the degree of known
or anticipated hazards for a specific site. Personnel training needs
are satisfied through a program of overview, discipline intensive
and site specific training, designed to accommodate time and bud-
getary restrictions.
Prior planning and proper training can generally prepare per-
sonnel for unexpected conditions while mitigating the potential for
serious accidents. Training in the proper and appropriate use of
specialized protective equipment and site entrance and egress con-
trols can also ameliorate the possible conflict between the safety of
the site worker and any undue anxiety among individuals of neigh-
boring residential or industrial areas.
Table I.
Remedial Measures Training Overview Course
Day 1 Introduction—Defining the Problem
Introductory Remarks and Course Overview
CERCLA and the Organization's Role
Superfund Sites, What Are They? How Were They Determined?
Physical-Chemical Properties of Hazardous Materials
Toxic Properties of Hazardous Wastes
Hazardous Chemical References and Their Use
Base Line and Personnel Considerations
Day 2 Personnel Protection
Measuring Devices
Respiratory Protection
Protective Clothing
Practical Exercises
Day 3 Operational Considerations
Site Entrance and Decontamination
Sampling Techniques
Dealing with the Public and the Media
Practical Exercises
Day 4 Remedial Measures
Hazardous Waste and Hydrology
Remedial Techniques
Defining the Scope and Extent of Remedial Actions
Specifications
Contingency Planning
-------
ADDRESSING CITIZEN HEALTH CONCERNS DURING
UNCONTROLLED HAZARDOUS WASTE SITE CLEANUP
GREGORY A. VANDERLAAN
U.S. Environmental Protection Agency, Region V
Chicago, Illinois
INTRODUCTION
The Midwest Industrial Waste Disposal Company, Inc., (Mid-
co I) uncontrolled hazardous waste site is located in Gary, Lake
County, Indiana. The surrounding area consists mostly of wetlands
with some usage by light industry. Across a limited access highway
approximately one-half mile to the west, is the Hessville neighbor-
hood in the City of Hammond, Indiana (Fig. 1). Approximate-
ly 20,000 people live within a one-mile radius of the site, mostly
within the corporate boundaries of Gary and Hammond, Indiana.
Operations that began in 1975 consisted mainly of temporary
bulk liquid and drum storage of various industrial wastes that
came primarily from industries located in the greater Chicago and
Northern Indiana area. Some solvent reclamation activities were
undertaken at times. Other operations included neutralization of
acids and caustics.
Waste handling activities began in the mid-1970s at two Mid-
Co facilities. Midco II, a sister site to Midco I is also located in
Gary and began operations after Midco I was destroyed by fire.
Midco I was incorporated on July 21, 1975. The primary direc-
tor and registered agent was Mr. Ernest DeHart, a Crown Point,
Indiana resident. The site is on land owned by Mr. DeHart and
several others. The operation of Midco I utilized DeHart's land and
with or without permission land owned by others adjacent to his
property.
Actual activities at the site started Apr. 21, 1975. DeHart con-
tinued operation until Dec. 21, 1976, when a fire destroyed 14,000
55 gal drums of waste material on site. The Gary, Indiana fire
marshal determined the fire was the result of a chemical reaction
that occurred when acids and solvents from leaking barrels co-
mingled. Evidence of the fire remained on the site prior to
USEPA's cleanup actions; the drums that were damaged or des-
troyed had never been removed.
Mr. DeHart reportedly sold the business logo, telephone num-
ber and customer list of Industrial Techtonics. Operations at the
facility were begun again by this firm in 1977. At that time, the
site contained 14,000 burned out 55 gal drums. Industrial Tech-
tonics (Intec) leased approximately 0.5 acres of the Midco I site
from Mr. DeHart. During the course of their activities, an addi-
tional 2,000 drums accumulated at the site.
As mentioned above, this site is located near a highly populated
area in northwest Indiana. During periods of unusually intense
rainfall, runoff from this entire Gary, Indiana area often flooded
the adjacent Hessville neighborhood of Hammond, Indiana (Fig.
2). The entire area has a history of past industrial use and midnight
dumping activities. The main concern of the Hessville residents
was the possibility that, during future heavy rains, chemically-
contaminated runoff would flow from the Midco I site and general
area into their streets and sewers, flood their basements and
threaten their health.
The residents were also worried about the hazards of children
playing in the vicinity of the site. In fact, concern was so great that
after a storm event in June, 1981, a sand and concrete barrier
This paper has not been peer and administratively reviewed within USEPA. It pro-
vides the viewpoints of the author only and does not represent official Agency
positions, policy or guidance.
was placed across a connecting street between Gary and Hammond
to stop the flow of runoff from the large open area, under Cline
Avenue into the Hessville area. The dike that blocks the main
thoroughfare between predominantly black Gary and predom-
inantly white Hammond, for the stated purpose of diverting flood-
water, has been a matter of disagreement between the two com-
munities (Fig. 3).
Need for Agency Action
On Dec. 21, 1976, a fire destroyed drums and bulk tanks stored
at Midco I. This fire, however, did not create a lasting concern
among citizens; the site was not recognized as a continuing haz-
ard. On Mar. 14, 1981, a 14 year old Hessville boy suffered leg
burns while playing near the site; his parents attributed the burns to
Figure 1.
This aerial photograph depicts the general location of Midco I
(encircled) in relation to the Hessville area of Hammond
located below Cline Avenue.
Figure 2.
View looking northwest from the site toward the 9th Avenue bridge
underpass at Cline Avenue.
321
-------
322
OFFSITE SAFETY
Figure 3.
Close-up of the sand dike built by Hammond authorities to
keep run-off waters from entering Hessville.
chemicals from the Midco I site. Shortly afterward, on June 14,
1981, exceptionally heavy rains sent floodwaters through the Hess-
ville neighborhood. Many residents complained of rashes and
burns, blaming chemical runoff from the Midco I site.
After the alleged chemical burn incident, an On-Scene Coordi-
nator from Region V inspected the site and obtained samples.
Based on the location of the site (a wetland area with a predom-
inant drainage pattern toward the Grand Calumet River), Section
31 l(k) funds were obtained from the USCG to construct a security
fence around the site.
Responsible Parties and Enforcement
Three companies successively operated at the site between 1975
and 1979. Each provided temporary bulk liquid and drum storage
for various wastes and reclaimable material. In 1979, all operations
ceased.
In 1980, a preliminary injunction was granted, ordering various
owners and operators to abate the imminent and substantial danger
posed by the site. The parties subject to the order failed to abate
the hazard.
Transporters and generators of the waste were then requested
to undertake response actions. Upon failure of these parties to
undertake the necessary action, initial response actions were taken
under authority of the Comprehensive Environmental Response
Compensation and Liability Act of 1980.
Most of the owners and operators of the site, and the owners
and operators of the reclamation businesses which used the site,
have been in contact with the USEPA since 1979, when the gov-
ernment filed suit to enjoin operations there. Mr. Ernest DeHart,
the principal Figure in the Midco I company, was not located un-
til July 1981. Criminal charges are pending against DeHart in Lake
County, Indiana, in connection with his Midco operations. The
state also brought a civil action in Lake County and an injunc-
tion concerning limited work at another site. Civil and criminal
proceedings were brought against DeHart in Grant County in
connection with a third site in Upland Indiana.
The responsible parties identified were asked to undertake the
cleanup. Owners and operators indicated they could not afford to
clean up the site. Transporters and generators responded with vary-
ing degrees of cooperation. One company removed approximately
800 drums. Several companies indicated an interest in monetary
settlement; many reserved judgment concerning responsibility. A
few companies have refused to seriously consider settlement.
The Agency's First Planned Removal
After the Hessville boy suffered the burns, allegedly attributed
to chemicals at the site, and the storm that sent runoff and flood-
waters into the Hessville neighborhood, individual Hessville citi-
zens and Hammond officials began to press for cleanup of Midco I.
Considerable program uncertainties existed at this time, since it
had been merely six months since passage of Superfund.
Region V's assessment of the site indicated that the greatest
immediate threat to public health and environment was associated
with waste material contact. Once the security fence had been con-
structed, USEPA's concerns involved the potential for indirect
contact via runoff and floodwaters, and the longer term environ-
mental problems associated with groundwater and subsurface con-
tamination. After the fence had been completed, USEPA felt con-
fident there would be time to obtain some level of competition for
the procurement of contractor services to initially perform a sur-
face cleanup.
Although the National Contingency Plan had not yet been pub*
lished, planned removal program guidance had been developed suf-
ficiently enough to provide a mechanism for funding approval and
contract procurement. Much of this guidance was based on Reg-
ion V's novel approach for waste material removal involving use of
limited solicitation to obtain contractor services for bulk waste re-
moval at the Seymour Recycling Center, Seymour, Indiana.9
Cost estimates on an immediate removal basis for a surface
cleanup from reputable contractors ran as high as $2.5 million.
The range of costs among those contractors who submitted pro-
posals was from $450,000 to approximately $1.7 million. After the
evaluations had been made, the low cost contractor was selected
from among those considered equal technically. Ultimately, the
total cost for the entire cleanup effort was approximately $900,000.
Additional costs were incurred due to an emergency runoff con-
trol project and an expansion of the project scope to address drums
brought onto the site by Intec (Table 1).
Table 1.
Cleanup Costs at Midco I
Task
Removal and disposal of drums destroyed by fire
Sampling and disposal of Midco drums
Sampling and disposal of Intec drums
Soil removal
Emergency runoff containment and treatment
Costs
$192,696.72
151.381.47
250,735.71
278.577.47
29,267.43
$902,658.80
In order to fully evaluate the accuracy of the original scope of
work and cost estimates, it is necessary to debate the costs asso-
ciated with the Intec materials. When USEPA formulated the scope
of work, the Agency assumed successful negotiations with the com-
pany and, therefore, did not include this aspect of cleanup in the
overall cost estimate. Comparing the cost of cleanup without Intec
materials with original contractor estimates reflects a 45* cost
overrun (Table 2).
Table!.
Comparison of Cleanup CosU
Cost of cleanup less Intec material
Contractor estimate based on Agency scope of work
Apparent overrun
Costs
$65 1 ,923.09
450,000.00
201,923.00
Specific reasons for the cost increase involved underestimate* of
quantities (volumes) of fire-destroyed drums and contaminated
soils. This underscores the importance of primary site investiga-
tions and of obtaining the basis on which to provide the best vol-
ume of material and cost estimates possible. Obviously, the need
for action and the time frames with which it is to be taken mutt be
considered before committing to a long term investigative effort.
-------
OFFSITE SAFETY
323
The Community Relations Plan
USEPA's community relations plan for this planned removal ac-
tion was prepared in Jan. 1982. The community relations officer
was remarkably accurate and perceptive in identifying the issues
and potential problem areas before, during and after the cleanup
activities. The plan provided an accurate history and pinpointed the
time period and specific incident which stimulated the local push
for a cleanup. The plan was extremely helpful in identifying the
following key issues and concerns:
•Gary-Hammond relations
•Health concerns
•Additional hazardous waste dumps
•Confusion over legal actions
•Possible delays
Prior to cleanup activities, the community relations plan pointed
out three primary health concerns among the Hessville residents:
(1) chemically-contaminated runoff, flooding their basements, (2)
hazards to children playing in the vicinity, and (3) contaminated
groundwater spreading from the site and seeping into their base-
ments. This last concern seemed to be of secondary importance to
the residents during the time prior to commencement of our
planned removal. People in the area who were interviewed were
aware that groundwater contamination might remain after
USEPA's planned removal was completed but did not appear
alarmed by this possibility. This ostensible public understatement
for a potential serious health effect proved to be one of several
major difficulties we encountered as we proceeded toward comple-
tion of our planned removal.
PROBLEMS ENCOUNTERED
Four specific public health incidents occurred prior to and during
our site cleanup operations (Table 3). Two of these events, which
have been discussed briefly above, served to focus Agency re-
sources on the site and developed need for the planned removal.
Table 3.
Problems Occurring at Site
Time of Occurrence
Mar. 1981
June 1981
Mar. 1981-Apr. 1982
Mar. 1982
Event
Children allegedly burned by chemical waste
while playing near the site
Runoff waters from the site entering Hessville
neighborhood
Recurring Fontanyi family illnesses allegedly
attributed to the site
Indiana highway department employee com-
plaints of eye and skin irritation, nausea and
dizziness allegedly due to planned removal
cleanup activities at the site
Prior to cleanup, the attitude of involved citizens and local offic-
ials was that, if they had not attracted media attention to the burn
and flood incidents, and had not exerted political pressure for a
cleanup, the government would not have planned a response for
the Midco I site. While it is difficult to identify the fine line that
distinguishes program actions from political actions, the arousal of
local officials as a result of these things and the subsequent media
coverage required the USEPA to look at the site in closer detail.
Storm of June 1981
A dike constructed by City of Hammond employees during the
storm that resulted in substantial flooding of the Hessville area of
Hammond is shown in Fig. 3. During the flooding at least 11 resi-
dents were treated for chemical burns and skin irritations after
coming into contact with flood water believed to be contam-
inated by hazardous substances from the Midco I site. Also, ten
Hammond employees who helped construct the dike were treated
for suspected chemical burns and skin irritations.
Later that week, the Hammond Air Pollution Control Depart-
ment requested USEPA's assistance in investigating numerous
residential complaints of organic vapors emanating from basement
sumps. These people believed that wastes from Midco I were car-
ried by flood waters into the Tennessee Street area of Hessville.
USEPA's Technical Assistance Team (TAT) provided Ham-
mond with this assistance and inspected basements in ten houses
whose owners had lodged complaints with the Hammond Air
Pollution Control Department. No organic vapors were detected
olfactorily in any of the basements. At only one house was a
reading above background registered on the H-Nu meter. This
reading was taken in a sump where there were numerous tar
globules. Here, the owner had stated that he had tarred his foun-
dation about five days earlier and that the rains had probably car-
ried the material into his sump before the tar had dried. Through-
out the home to home survey, residents repeatedly complained of
illness and injury due to chemical contact or smell, including ref-
erence to the city workers mentioned above. At the time of the
TAT survey, these complaints comprised the primary motivation
for response activities because sample analyses indicated that the
huge volume of stormwater that inundated the area most likely di-
luted any water quality problems.
Fontanyi Family Health Problems
Mr. Frank Fontanyi, his wife, daughter and son were residents
of Hessville living on Tennessee Avenue during the flooding which
occurred in June 1981. Mr. Fontanyi led a verbal assault on a
variety of governmental agencies after the floodwaters entered the
area. Mr. Fontanyi claimed his seven year old son, Frank Jr., was
burned from contact with the water. When Regional field per-
sonnel arrived in Hessville to investigate the situation, Fontanyi
waved home grown vegetables from his garden under their noses.
He dared the inspectors to eat them; they politely declined.
In Aug. 1981, Fontanyi claimed a series of strange afflictions
affected his family and continued for a period of six months, ulti-
mately forcing him to abandon his home and neighborhood. The
afflictions, affecting himself, his son and his daughter, included
nausea, fatigue, eye irritation, headaches, muscle spasms, vomit-
ing and inflammation of the throat and tonsils. Fontanyi had also
been diagnosed by doctors as having had a slight stroke in the fall
subsequent to the flooding. He believes it was due to exposure to
the hazardous substances in the floodwaters. Fontanyi's children
were also apparently being affected. His daughter was hospital-
ized three times with kidney infections over a period from Jan. to
Mar. 1982. Fontanyi's son periodically needed to have his ears
drained on an out patient basis and suffered a 20% hearing loss.
Conventional medical doctors could not find the cause of the
Fontanyi family ailments. Fontanyi turned to a doctor who pio-
neered a field called clinical ecology and beginning in Mar. 1982,
he and his daughter underwent six weeks of testing in a procedure
called chemical detoxification. Both father and daughter were ad-
mitted to an "ecology center", in Chkago, one of half a dozen
such oases in the United States where such detoxification is carried
on. The center is a special world of aluminum wall paper, ceramic
tile and filtered air where floors are scrubbed with baking soda.
Plastics, synthetic fibers, cigarets, cosmetic and cleaning fluids are
banned. Books are isolated in glass boxes and patients reach in
with gloves to turn the pages.
Fontanyi and his daughter only drank spring water. They were
re-tested for allergic reactions to water, food and respirable partic-
ulates such as pollen to double check previous tests. Where tests
proved negative, further tests were administered for possible resid-
ual chemical contamination of the body. The clinical ecologist at-
tending Fontanyi told him prior to testing that if anything was
found Fontanyi and his family should not return to their house.
Clinical ecologists believe the body's internal defense system can-
not tolerate the world technology has created. They are not worried
about the major assaults, or the chemicals that everyone knows
will poison, mutate genes or cause cancer. It is, in their estima-
tion, the small, almost imperceptible insults they believe that come
-------
324
OFFSITE SAFETY
from polyester clothing, sipping disodium guanylate in soup or soft
drinks, those barely noticeable fumes from no-wax floors, office
copiers, and perfumes or even common foods eaten in excessive
quantities. The founder of clinical ecology attributes a vast array
of ills to "allergies" caused by these silent insults: nausea, diarrhea,
headaches, blurred vision, dizziness, fatigue, confusion, cramps,
wobbly knees, asthma, fevers, "brain fog", anxiety, schizophren-
ia, arthritis, alcoholism.
Individuals who had experienced years of tests, drugs, and psy-
chiatric treatment found relief through clinical ecology. Some re-
searchers in allergy and immunology are not convinced that chem-
icals are the source of these problems. Clinical ecologists provoke
from these researchers the same response laetrile elicits from oncol-
ogists. The head of the immunology and allergy division at San
Diego's Scripps Clinic and Research Foundation describes clinical
ecology as being "more a religious cult than science." He and
other researchers see the ecologists patients as gullible people seek-
ing external causes for their inability to cope with the world. The
American Academy of Allergy considers ineffective or unproven
the methods used by some clinical ecologists—urine injections and
dropping diluted chemicals under the tongue.
Highway Department Employee Health Complaints
Immediately west and adjacent to the site is the State of Indiana
Highway Department Gary subdistrict garage facility (Fig. 4). Dur-
ing the latter half of Mar. 1982, with the cleanup operation six
weeks old, the area experienced a typical midwestern early spring
wanning trend. As expected, volatilization rates of the compounds
on site increased significantly. Almost immediately, workers at the
highway department facility began to complain of adverse health
effects. On Mar. 18, seven employees went home sick, complaining
of nausea, dizziness, coughing and a burning sensation on their
skin. These individuals had been working inside the garage with the
doors open less than a block from major staging and sampling ac-
tivities on the site (USEPA's site safety plan required level C pro-
tection and all staging and sampling activities were in full view by
highway department employees).
The On-Scene Coordinator (OSQ was concerned, but rather
puzzled, that ambient air concentrations affected by the cleanup
activities were causing these kinds of acute responses among the
highway department employees. The OSC's twice daily total organ-
ic compounds measurements showed no levels greater than 3 ppm.
However, one Mar. 19,1982, an additional 25 employees were sent
to a local hospital for tests. All of the workers were complaining
of burning sensations in the eyes, nose and throat. None of the
workers required hospitalization.
Again on Mar. 24,1982,20 more workers were forced from their
workplace apparently due to ambient air concentrations of volatile
organic compounds emanating from the site. At that point state
highway department officials temporarily moved all of the state
employees to another facility until the magnitude and seriousness
of the problem was better defined. Still, daily on-site air quality
readings did not seem to reflect the symptoms and complaints be-
ing displayed and voiced by the highway department workers.
APPROACH FOR PROBLEM RESOLUTION
Community wide public health issues associated with direct con-
tact to hazardous materials on the site and contaminated flood-
waters entering the Hessville neighborhood were addressed by ob-
taining funding in Oct. 1981, for a planned removal surface clean-
up. While it was impossible to determine the proportionate impact
the site may have had on the quality of the floodwaters that en-
tered the Hessville area, it represented without question a source
of highly contaminated runoff and groundwater contamination
(Figs. S and 6).
It was clear to USEPA that no one in the Agency would be able
to credibly address the health issues raised by Mr. Fontanyi and
the highway department employees. Fortunately, the Agency had
already established a dialogue and worked with on a site specific
basis representatives from the Centers for Disease Control (CDC)
Figure 4.
View of the Midco 1 site showing a portion of the adjacent highway
department facility in the foreground.
Figure 5.
Conditions at the site prior to our cleanup activities.
Figure 6.
Conditions at the site prior to our cleanup activities.
Superfund Implementation Group. After USEPA forwarded baric
information on the site to CDC, a medical lexicologist was as-
signed to provide support and direction in attempting to resolve
these problems.
Shortly after the Indiana State Highway Department relocated
the Gary subdistrict employees to Crown Point, USEPA designed
-------
OFFSITE SAFETY
325
Figure 7.
Agency field representative setting the meteorological station,
initiating the ambient air monitoring program.
and implemented a compound specific ambient air monitoring pro-
gram (Fig. 7). Because total organic vapor measuring devices were
being used, one could not say with certainty that the high reading
of three ppm did not contain one ppm of benzene and, therefore,
did not exceed the threshold limit value (TLV) for that substance.
Data from the compound specific ambient air monitoring pro-
gram confirmed the OSC's professional intuition. Concentrations
of organic pollutants were 10 to 100 times below the TLV's set for
the work place. It was still puzzling how such minute concentra-
tions could cause such wide spread ill effects.
When the highway department workers were still reluctant to re-
turn to Gary, a meeting with them was organized by the Agency,
the State of Indiana Highway Department and CDC. The assigned
physician attended this meeting; his presence and comments were
pivotal in getting the employees back to their original workplace.
The main point of logic he used to convince the workers to return
was rooted in benzene exposure. The CDC physician calculated
that additional gas consumption necessary to travel a greater dis-
tance to work increased their exposure to benzene via greater gas-
oline consumption. They realized they were breathing greater levels
of benzene because of an increase in the number of gasoline tank
fiilups necessary for them to drive greater distances to work. Most
of them returned to the Gary subdistrict garage.
The Agency's ability to address the Fontanyi family health issues
was constrained considerably by the inaccessibility of Mr. Fontanyi
and what appeared to be uncooperativeness by Mr. Fontanyi and
his clinical ecologist physician. The CDC doctor was able to dis-
cuss with Fontanyi's physician his general procedures and the basis
of his therapy. He refused at that time to discuss Mr. Fontanyi's
case until the government had been given a release for access to his
medical records. Unfortunately, after Mr. Fontanyi and his
daughter were released from detoxification, they became inaccess-
ible.
Repeated attempts were made by the Agency and CDC to con-
tact Mr. Fontanyi so it could discuss and better understand his
problems and determine the potential for others in the area to be
affected. Apparently, after Mr. Fontanyi had completed the detox-
ification process his doctor advised him not to return to his home
or the Hessville neighborhood because the contamination in the
soils and air was causing his medical problems. Finally, after many
weeks, the Agency contacted Mr. Fontanyi and received his per-
mission to review his medical records. They were just recently
turned over to CDC for review.
Later, in a newspaper interview, Mr. Fontanyi's physician made
the disturbing comment that part of the Hessville area is "unfit for
human habitation" and that homes there should be evacuated.
"The whole area is saturated with chemicals and wash-off from the
dumps. This is a highly toxic area and if you continue to live there
you are going to get very sick." The Agency would have hoped that
this physician would have shared his data and findings with CDC
and local health authorities prior to these kinds of public
statements.
Recommendations
The community relations plan cannot consider all possibilities
while undertaking cleanup. Because of that, standard bylines for
community relations plan development and implementation are
that: (1) it must maintain flexibility, and (2) it must be updated
whenever necessary.
While the community relations plan for this cleanup was one of
the better ones, the OSC and program office must, whenever neces-
sary, utilize all resources available to them. Most Regions now
have assigned to them a CDC health representative responsible
for interpreting and implementing health guidance and resources
regarding possible adverse health effects associated with a site. If
necessary, the health representative can seek a CDC physician in
cases similar to Midco I. However, there will be instances when
a doctor will be necessary. The problem with the highway depart-
ment workers is a good example. Any person with a technical back-
ground could have used the benzene in gasoline example but it ob-
viously meant much more to those affected individuals hearing it
from a medical doctor.
Finally, at sites near areas of residential or commercial develop-
ment, compound specific ambient air monitoring should be per-
formed prior to cleanup activities and for a one to two week period
during cleanup efforts. This will provide substantive data or air
quality impacts and can be used to defend total organic compound
measurements taken on a daily basis.
REFERENCES
1. "Emergency Action Plan—Midco I Gary, Indiana", Region V, USEPA
Technical Assistance Team, Chicago, II., Apr .1981.
2. "Community Relations Plan for Planned Removal at the Midco I Haz-
ardous Waste Site, Gary, Indiana", Office of Public Affairs, Region
V, USEPA, Chicago, II., Jan. 1982.
3. "Continuum" OMNI, U.S. 5, Number 1, Oct. 1982, ISSN 0149-8711,
p. 47.
4. "Ambient Air Monitoring for Volatile Organics, Midco I, Gary, In-
diana", Environmental Services Division, Region V, USEPA, Chi-
cago, II., May 1982.
5. Vanderlaan, Gregory A. "A Fast Track Approach to Impact Assess-
ment at Uncontrolled Hazardous Waste Sites", Proc. Second National
Conference on Management of Uncontrolled Hazardous Waste Sites,
Oct. 1981.
-------
ESTIMATING VAPOR AND ODOR EMISSION RATES
FROM HAZARDOUS WASTE SITES
ALICE D. ASTLE
RICHARD A. DUFFEE
ALEXANDER R. STANKUNAS, Ph.D.
TRC Environmental Consultants, Inc.
East Hartford, Connecticut
INTRODUCTION
Odorous emissions from hazardous waste sites are often a major
factor influencing community reaction. While in most cases odors
are not in themselves a serious threat to health, they are directly
sensed by the affected population and serve as a reminder (and a
warning) that the site is emitting volatile compounds.
In this paper, the authors present a review of some of the factors
influencing volatile organic emissions and present a methodology
that can be used to determine the local impact of odors from a
waste disposal site. Five areas of activity are addressed: emissions
factors, odor emission measurement, community surveys, at-
mospheric transport and dispersion and community response ser-
vices.
EMISSION FACTORS
Odor problems at hazardous waste disposal sites can generally be
equated to the emission of volatile organic chemicals. For con-
taminated soil systems, the most important mechanisms are direct
volatilization (evaporation) and diffusion of vapors through the
soil. The emission rate is a function of a number of factors: the area
of contamination and concentration of the waste, the vapor
pressures and gaseous diffusion coefficients of the various waste
components, .the depth and porosity of any overburden, mass
flow of air and other gases, temperature, and the ambient concen-
tration of waste components in the air above the site.
In cases where the waste material is at the surface of the soil, such
as in a seep or a freshly uncovered contaminated soil layer, the
dominant mechanism will be evaporation. The source strength can
be approximated by:
Q = Km (P-Pa)/RT
where:
(D
Q = mass flux
P = vapor pressure of the material
Pa = partial pressure of the material in the air over the
surface
Km = transport coefficient of the atmosphere directly above
the material
T = temperature of the liquid
R = gas constant
The transport coefficient can be estimated from:
Km = CU°-78X-o.» (2)
where:
C = an empirical constant
X = liquid pool diameter
U = wind speed
However, care must be exercised when attempting to employ this
approach for multicomponent, viscous mixtures like those found at
most waste disposal sites. The composition of the exposed liquid
surface will not necessarily be replenished by mixing with the bulk
of the material and emission of volatile components may be quite
different than expected. Formation of films of relatively non-
volatile materials that greatly reduce evaporation is commonly
seen.
In sites where there is significant production of gases within the
landfill, the emission of volatile compounds can be significantly af-
fected by the mass flow out of the soil. The emissions for systems
with significant outgassing can be estimated from:
Q =
(3)
where:
Q
K
U
A
L
Ap
mass flux
soil permeability
viscosity coefficient
effective area
effective thickness of soil cover
pressure differential between landfill gas and the
atmosphere
concentration of material in the gas phase
However, the toxicity of most wastes disposed of in hazardous
waste sites effectively prevents the microbial action most often
responsible for landfill gas production and this mechanism is rarely
dominant.
The most common situation at waste disposal sites is one where
the volatile waste is buried under a semi-permeable layer of soil. In
such cases, diffusion through the capping soil is the dominant
mechanism of emission. The emission rate for this process can be
estimated by:
Q • -oa
where:
Q
Da
Ca
- C
10/3 2
. /Pt
(4)
mass flux from the soil surface
effective diffusion coefficient in air
concentration of vapor in air above the surface
concentration of vapor in air of soil pores below
the surface
soil depth
air filled porosity of the soil
total porosity of the
L
Pa
Pt
The major drawback to estimating emissions by this technique is
the requirement for data on the diffusion coefficients and vapor
pressures of all the materials of interest, as well as subsurface
porosity and temperature. The method also assumes that the soil
porosity itself is fairly uniform and that macro defects, such as
cracks, root channels and animal burrows are not present in signifi-
cant numbers. Finally, the estimates of vapor concentration in the
326
-------
OFFSITE SAFETY
327
soil pores (Cs) are subject to the same sources of error for
multicomponent, viscous mixtures that evaporation calculations
are.
This is not to say that models of emission rate are not useful.
They can provide a reasonable starting point for estimating poten-
tial impacts. Moreover, they can be used to extrapolate how the ex-
isting rate would change as a function of site modification or
change in sampling conditions.
However, for odor problems, where a whole ensemble of
chemical compounds, each with its own unique set of physical and
chemical properties must be considered, modeling alone, is of
limited usefulness. It is much more practical to combine the models
of vapor behavior with a limited set of on-site measurements.
ODOR EMISSION MEASUREMENTS
Odor is not something that can be measured without reference to
subjective response. That is, no instrument can tell you how bad
something smells. This is especialy true for complex odors related
to mixtures of chemical compounds such as those found at hazar-
dous waste sites. The most effective method of detecting and quan-
tifying odorous emissions is to smell them.
Panels of odor judges, generally composed of a cross-section of
the population in order to obtain a reasonable estimate of the inten-
sity and objectionable nature of any particular odor as perceived by
the community exposed, are used to quantify odors. The use of this
technique is surprisingly reproducible and accurate in making
quantitative measurements of a subjective phenomenon.
Samples for sensory odor evaluation are taken using the bag-in-
a-drum method shown schematically in Fig. 1. A Tedlar sample bag
is installed inside an aluminum sample drum. A vacuum is pulled
on the drum, drawing the sample through Teflon tubing into the
sample bag. When the bag is partially filled, the drum is pressuriz-
ed, emptying the bag. The vacuum is then reapplied and the bag is
completely filled. The partial filling and subsequent emptying of
the bag serve to "precondition" the bag walls by equilibrating the
inner surface with the sample gas, thereby minimizing losses caused
by adsorption.
Measurements of odor source strength can be taken with the aid
of a portable wind tunnel, as shown in Fig. 2. The wind tunnel pro-
vides a constant flow of odor-free air at a known velocity and
isolates the area being measured from emissions from other areas.
The tunnel consists of a variable flow blower with an activated car-
bon filter on the inlet and a three-sided tunnel on the outlet. The
support feet of the tunnel are lined with acid-resistant foam to seal
over uneven spots in the surface. The blower is set to deliver the
desired tunnel velocity and a sample is drawn from the tunnel at
various distances "downwind" using the bag-in-a-drum technique.
Odorant or volatile organic samples can also be collected by ad-
sorption on Tenax or other sampling media, depending on the
nature of the waste. Samples can be adsorbed directly from the
wind tunnel or they can be transferred during sensory evaluation
using a system such as that shown in Fig. 3.
Odor samples are evaluated for detectability (dilution-to-
threshold level) and intensity using a panel of judges who have been
screened for their ability to detect odor qualities typical of the
odors to be measured. There are many techniques available for
measuring both detectability and intensity. TRC uses the forced-
choice triangle olfactometer5 for measuring detectability and the
eight-point butanol reference scale6 for evaluating intensity. The
measurements are done in a mobile laboratory equipped with a car-
bon filtration system which maintains an odor-free environment.
Measurement of both properties is necessary because, although
odor annoyance is more closely related to intensity, dispersion
modeling is best done in terms of detectability.
The forced-choice triangle olfactometer uses carbon-filtered air
to make six simultaneous dilutions of the odor sample. Each dilu-
tion level is presented by means of a cup containing three glass-sniff
ports. Two ports have only odor free air while the third has the
diluted odor. Panelists begin with the most dilute sample and at
each stage must choose one of the three ports whether they can
detect an odor or not. The median threshold for the panel, that is
the point at which 50% of the panel correctly and reproducibly,
selects the diluted odor, is thus determined.
«irl»c to
Control
II
Spied
Sopllng Probe for
Tr«nch
Lead fetgtit
to Aid Sc
-------
328
OFFSITE SAFETY
The butanol intensity olfactometer provides eight simultaneous
dilutions of a standard reference odorant, 1-butanol. Each level
doubles in butanol concentration from the preceding level. The
scale provides objective reference points for comparison of odor in-
tensities. Each panelist compares an odor level from the triangle
olfactometer to the butanol scale and select the butanol port which
is perceived to have the same intensity. The choices are averaged to
derive a mean value for the panel.
The detectability and intensity measurements can be used to
estimate the objectionability threshold for the odors emanating
from the site. A log-log plot of detectability vs. intensity of each
sample, similar to a Steven's Law function, is used to determine the
detectability level corresponding to an intensity of 50 ppm of
butanol vapor equivalent to 2.5 on the 8-point scale. This value has
been shown in previous work to be in most cases a reasonable
estimate for the upper limit for an acceptable odor level.8 This level
provides the target for evaluating control requirements and the
potential for creating an odor nuisance during mitigation activities.
In addition to direct olfactory measurements, samples can be
characterized by a gas chromatography odorogram. In this pro-
cedure, the materials adsorbed on the Tenax tubes are desorbed in-
to a gas chromatograph (GC). The column is equipped with an ef-
fluent splitter which sends a portion of the elution from the column
to a flame ionization detector (FID) and the remaining portion to a
sniff port. As peaks emerge, odor judges describe the odors in the
sniff ports and their observations are recorded on the
chromatogram. This technique helps identify which of the peaks
are odorous, and a skilled judge can even provide a fairly accurate
estimate of the relative contribution of each component to the
overall odor.
However, this procedure does not identify the components
chemically. To accomplish this, a second sample is analyzed by
GC/mass spectrometry and the odorous chemicals are identified by
mass number. The chromatographic system used for the GC/(MS is
identical to that used for the GC/odorogram so peaks can be mat-
ched directly by retention time. Of course, any of the peaks,
whether odorous or not, can be identified by GC/MS, if desired.
Once the odorants or volatile organics of concern are identified
chemically, they can be quantified on the GC using external stan-
dards. The resulting concentration data can be used to calculate an
emissions rate for any particular chemical.
COMMUNITY SURVEYS
Community odor surveys should be done simultaneously with
on-site sampling, and at various other times of the day. The
primary purpose of these surveys is to provide data which can be
used to calibrate an odor dispersion model for the specific site in
question. The surveys also help document the geographical extent
of the odor problem and the odor levels that occur under various
meteorological conditions.
Community surveys are done by experienced odor judges driving
or walking in the neighborhood around the site and noting the loca-
tion, time, quality, intensity and detectability of any odors
detected. Wind speed and direction, ambient temperature, and
other meteorological conditions are also noted. Intensity
measurements are made by butanol referencing. Detectability is
measured using the Scentometer, a simple field instrument shown
schematically in Fig. 4.
ODOR DISPERSION MODELING
Although the sensory measurements, chemical analyses, and
community surveys all document the impact and characterize the
emissions from a hazardous waste site, it is the use of an at-
mospheric transport and dispersion model which ties all the data
together and allows predictions of the community impact of an
odor source under any given set of conditions.
TRC uses a proprietary model designed specifically for use in
odor situations. Odor is not a steady state or time averaged
— Air Inlet to filter d/2")
Activated
chaccoal bedt
pieces
Figure 4.
Scentometer
response. Accordingly, the model is a "fluctuating puff" model
that determines dilution rates of odor emission puffs and distribu-
tions of plume centroids, from which it predicts instantaneous odor
levels. Calculated results are presented as a frequency distribution
of predicted odor levels at specified receptor locations. It can be us-
ed to determine peak odor level during a one-hour period and
percentage of occurrence at detectable odor during a one-hour
period.
The model is first calibrated using simultaneous on-site emission
and community response data. The source parameters are input
along with the meteorological conditions existing at the time of the
survey. Odor levels are predicted for the locations at which odors
were actualy detected. If the predicted and observed values are in
good agreement, the model is considered appropriate for that site.
If there is a significant deviation, the input parameters are adjusted
until agreement is attained.
Once calibrated, the model can be used to predict the odor im-
pact under any given set of meteorological or site conditions. For
instance, if an open site is to be partially covered, the source dimen-
sions are changed in the model input. If parts of the site are to be
excavated, the emission terms are modified accordingly.
COMMUNITY RESPONSE SERVICE
Establishment of a community response service can be helpful in
assessing the odor impact of a hazardous waste site, but is not
essential. In most cases, a response service functions as a clearing
house for odor complaints by community residents. When a com-
plaint is received, a response team is sent to the location as quickly
as possible. On-site observations of the meteorological conditions
and nature and intensity of the odor by a trained odor judge are
coupled with sampling of the odor for later analysis where ap-
propriate.
The complaining resident is then asked to complete a question-
naire designed to evaluate the intensity and odor quality at the time
of the complaint. If possible, matching information is solicited in
the area immediately up and downwind of the affected area. The
information obtained not only helps define the conditions that the
community considers objectionable, but the act of setting up such
an operation often has a beneficial public relations effect
-------
OFFSITE SAFETY
329
EXAMPLE OF AN APPLICATION
OF THE METHODOLOGY
Results from a study in which the methodology described above
was applied to a hazardous waste site used for petroleum wastes are
shown in Tables 1 and 2 and Fig. 5. Since the time the site was ac-
tive, the neighborhood around it has been developed so that now
there are houses within 150 ft of major emission points.
Table 1.
Sample Results of Sensory Odor Measurements
a. Surface Samples
Location
Sump 1
Sump 2
Sump 3
Sump 4
Sump 5
Sump 6
Sump 2
Sump 4
Sump 1
Sump 2
Sump 3
Sump 4
Sump 5
Sump 6
Sump 4
Sump 3
Sump 2
Sump 1
Sump 5
Sump 6
Sample
Core Hole A
Core Hole A
Core Hole A
Core Hole A
Core Hole B
Core Hole B
Core Hole B
Core Hole B
Soil
Conti-
tion
Dry
Dry
Dry
Dry
Dry
Dry
Dry
Dry
Wet
Wet
Wet
Wet
Wet
Wet
Drying
Drying
Dryi ng
Drying
Dryi ng
Dryi ng
4'/! ft
9 !/! ft
20 ft.
30ft.
Soil
Temp.
Length
of Wind
Tunnel
(°F) (meters)
60
69
69
88
79
76
64
63
67
63
73
75
66
78
80
76
80
84
b.
. BWS'
. BWS
BWS
BWS
4
8.
8.
8.
4
4.
8,
8
4
8,
8.
8.
4
4,
8.
8,
8.
4
4.
4,
.9
,6
.6
.6
.9
,9
.6
.6
.9
.6
.6
,6
.9
.9
.6
.6
.6
.9
.9
.9
Subsurface
7 ft. BWS
\OV, ft. BWS
20ft.
27 ft.
BWS
BWS
ED50
4,782
4,427
16,293
18,334
12,661
14,099
18,597
19,077
Detect- Intensity
Tunnel ability ED
Velocity (ED ) At 50
(m/sec)
1
1
1
1
1
1
2
2
1
1
1
1
1
1
1
1
1
1
1
1
.0
.0
.0
.0
.0
.0
.5
.5
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
50
11
38
20
70
33
39
17
17
13
12
30
:10
:10
15
:10
74
19
53
19
39
1.
4.
2.
3.
2.
97T
9
2
9
1
1
3.1
1.
2.
1.
1.
2.
0.
0.
2.
0.
2.
1.
2.
1.
2.
9
5
5
6
0
6
9
4
8
5
6
7
7
6
Samples
Odor Level
in Cup 6
ED50
250
19.1
17.7
65.2
73.3
50.6
56.4
74.4
76.3
Intensity
al ED50
250
4.9
4.2
5.3
5.4
4.5
4.5
4.9
4.9
•BWS = Below waste surface
Results of sensory odor measurements made on both surface
samples taken from the wind tunnel and subsurface samples taken
from a flux chamber lowered into core holes are shown in Table 1.
Surface sampling showed that the odor levels varied with soil
moisture conditions and soil temperature, but not with wind tunnel
length, after a relatively short run. This implied that the concentra-
tionof odorants in the air above the soil did not need to be very high
for a diffusion equilibrium to be established. In this particular case,
the odor levels at the surface were fairly low, but because of the size
of the source, the overall emission rates were high enough to cause
objectionable odors in the community under stable meteorological
conditions. Subsurface samples showed very high odor level poten-
tial if excavation was attempted.
The relationship between detectability and intensity for the sur-
face and subsurface samples are shown in Fig. 5. For the surface
samples, an odor would begin to be considered objectionable at
about 8 times its threshold level. This estimate was confirmed by
the results of the community response questionnaire, which in-
dicated that at an odor level 8 times higher than the threshold con-
centration complaints began.
However, the results of the subsurface sampling indicated that
the nature of the odor beneath the surface was quite different. Such
an odor would be considered objectionable as soon as it was detec-
table. Therefore, if the site was to be excavated, it would have to be
done in such a way as to maintain community odors at less than
their detectability threshold.
The odorogram indicated that the mixture of wastes was highly
complex, with as many as 70 peaks, of which over 30 were strongly
odorous (Table 2). Considering this complexity, the semi-empirical
source measurement program clearly represents the easiest way to
quantify the problem.
Table 2.
Chemical Compounds Found in a Typical Surface Sample
1,1 -Dibromo-2-chloro-2-
fluorocyclopentane
Oxirane
Methyloxirnane
Methylene chloride
2-Methylpropanal
Hexane
Trichloroethene
Methylcyclopoentane
1,1,1-Tpichloroethane
Benzene
2-Methylhexane
2,3-Dimethylpentane
3-Methylhexane
1,3-Dimethylcyclopentane
2-Ethyl-l-hexanol
1,2-Dimethylcyclopentane
Heptane
Methylcyclohexane
2,5-Dimethylhexane
Ethylcyclopentane
1,2,4-Trimethylcyclopentane
2,3,4-Trimethylcyclopentane
2,3,3-Trimethylpentane
2,3-Dimethylhexane
Toluene
3-Methylheptane
1,3-Dimethylcyclohexane
2,2,5-Trimethylhexane
1,2-Dimethylcyclohexane
Octane
1,4-DimethyIcyclohexane
Tetrahydrothiophene
2,3,5-Trimethylhexane
2-Methyloctane
Ethylcyclohexane
1,1,3-Trimethylcyclohexane
1,2,4-Trimethylcyclohexane
2,3-Dimethylheptane
Ethylbenzene
2-Methyloctane
m,p-Xylene
o-Xylene
Nonane
1 -Ethyl-4-methylcyclohexane
Octahydroindene
Propylcyclohexane
l,l'-Oxibisoctane
2,2,5,5-Tetramethyl-3-hexane
2-Methylnonane
1 -Ethyl-1 -methylcyclohexane
octanol
Decane
1-Ethyl-1-hexanol
1 -Ethyl-3-methylcyclopentane
1-Decanol
2,2,5-Trimethylhexane
Decahydronaphthalene
2,2,3,4-Tetramethylpentane
Undecane
Table 3.
Comparison of Predicted and Observed Odor Levels
for Daytime Conditions
Observed Odor Level
(D/T)
Threshold to 8
Threshold to 32
None
Threshold to 8
Threshold to 32
Threshold to 2
Predicted Peak
Odor Level
(D/T)
8
12
0.1
(undetectable)
8
8
1
Predicted %
Time Detectable
Odor Levels
88
99.5
0
96.5
74
0
Receptor
Location
A
B
C
D
E
F
The documentation of odor impact that can result from com-
munity surveys is shown in Fig. 6. These data are very helpful in
assessing the geographic extent and severity of the odor impact, as
well as in evaluating the effects of local topography and
meteorology on odor dispersion patterns for scheduling mitigation
activities.
The calibration of a site-specific odor dispersion model is shown
in Table 3. The odor emission data shown in this table were put into
-------
330
OFFSITE SAFETY
10
8
6
0.8
INTENSITY = 1.36(ED5o)°-40
r2= 0.50
4 8 8 10 20
OETECTABILITY (E050)
Surface Samples
40 6080100
INTENSITY = 2.88(EDj0>°-13
r* r o.as
6 8 10 20 40
DETECTABILITY (ED50)
Subsurface Sataples
6080100 200
Figure 5.
Relation of Detcctability and Intensity
Patterns are based upon both location of complaints residences
and TRC surveillance.
Patterns are highly Influenced by local topographical as well as
meteorological conditions.
Figure 6.
Odor Impact in Vicinity of Site Under Existing Conditions
the model along with meteorological data recorded during a survey.
The predicted odor impact agreed very well with that observed, and
the model was considered calibrated for this particular site.
CONCLUSIONS
Despite the complexity of the mechanisms involved, there is a
practical methodology available which will provide the data
necessary to evaluate odor impacts of hazardous waste sites, as well
as the potential impact of mitigation activities. The method in-
volves odor and volatile organics sampling, sensory odor
measurements, speciation and quantification of volatile organics,
community surveys, and dispersion modeling. Establishment of a
service to respond to community complaints can be helpful, but is
not essential for the successful application of the method.
REFERENCES
1. Sutton, O.O., "Micrometeorology," McGraw Hill, New York N Y
1953.
2. Farmer, W.J. et al.
3. "Land Disposal of Hexachlorobenzene Wastes: Controlling Vapor
Movement in Soils," Fourth Research Symposium: San Antonio, TX,
EPA-600/9-78, Aug., 1978.
4. Mackay, D. and Matsugu, R.S., "Evaporation Rates of Liquid Hydro-
carbon Spills on Land and Water," Canadian J. of Chemical Engineer-
ing, 51, Aug. 1973.
5. Dravnieks, A., Prokop, W.H. and Boehme, W.R., "Measurement of
Ambient Odors Using Dynamic Forced-Choice Triangle Olfactometer,"
JAPCA, 28, 1978, 1124-1130.
6. ASTM E544-75, Standard Recommended Practices for References
Suprathreshold Odor Intensity, Philadelphia, PA., 1975.
7. Cain, W.S., "Preliminary Data Presented at the Building Ventilation
and Indoor Air Quality Program Annual Technical Review, Lawrence
Berkeley Laboratory, Berkeley, CA., Oct. 1979.
8. Duffee, R.A. and Astle, A.D., "Evaluation of Odor Annoyance Prob-
lems," Paper 82-28.1, 75th Annual Meeting Air Pollution Control
Assoc., New Orleans, LA., June 1982.
-------
AIR MODELING AND MONITORING FOR SITE EXCAVATION
BRIAN L. MURPHY, Ph.D.
TRC Environmental Consultants, Inc.
East Hartford, Connecticut
INTRODUCTION
Usually at an abandoned hazardous waste site, groundwater con-
tamination or surface runoff are the major concerns. During site
excavation, however, air impacts due to contaminated fugitive
dusts and liberated volatile compounds may be the major source of
risk. In general there are three steps in a risk minimization program
for air impacts:
•Planning site activities (principally modeling)
•Monitoring (principally during remedial action)
•Dealing with community and agency concerns (integration and
interpretation of monitoring and modeling results)
In addition to allaying concerns at the time, proper documenta-
tion of actual exposure levels can prevent future liabilities.
Generally both monitoring and modeling activities are concerned
with the determination of ambient concentrations at the point of
worker exposure, at the property line and in populated nearby
areas. The parameters on which concentration and exposure de-
pend are shown in Table 1. Predicted or monitored concentration
levels can then be compared with threshold limiting values
(TLVs), OSHA standards (adjusted for the fact that they pertain
to an 8-hour day) or other appropriate requirements. If concentra-
tions or exposure are deemed unacceptable a number of avenues of
recourse are available.
Fugitive dusts can be controlled by a number of practices at the
site including fogging or misting. Wetting the ground can also
reduce volatile emissions. In addition, pre-planning site activities as
to season, area of disturbance, and duration can influence accep-
tability.
Table 1.
Parameters of the Air Pollution Problem at Hazardous Waste Sites
Concentration depends on:
•Chemical compound
•Wind speed
•Ambient temperature
•Overburden (volatiles)
•Soil moisture
•Soil porosity
•Dust control measures
Exposure depends on:
•Concentration
•Population movement in space and time
•Area and duration of excavation
•Indoor/outdoor air relationships
MODELING
The models of interest in this paper estimate emissions. Having
an estimate of the mass rate of emissions, conventional urban and
industrial source complex dispersion models can be used to
calculate downwind concentrations for compounds which are not
rapidly reactive in the atmosphere.
Both fugitive dust and hazardous vapor emissions will be in-
fluenced by geometry as shown in Fig. 1. Material at the bottom of
a pit deeper than it is wide (in the windward direction) is shielded
from atmospheric turbulence, volatiles can only escape via
molecular diffusion, a slow process and hence a low emission rate.
For a shallow pit (wider than it is deep) the excavation acts as an ad-
ditional roughness element as shown in Fig. 1 b. Enhanced tur-
bulence due to a stationary vortex will increase emissions. A similar
phenomenon occurs at piles (Fig 1 c). In the lee of the pile near the
peak a vortex system is responsible for fugitive dust emissions.
Molecular
Diffusion
(la) PIT
Enhanced
Turbulence
(Ib) PIT
Enhanced
Turbulence
(Ic) PILE
Figure 1.
Influence of site geometry
The fact that fugitive dust emission rates in particular depend on
such local factors is important from a measurement standpoint. It
means that single monitor measurements are likely to be misleading
since a small shift in wind direction will shift the monitor reading
off centerline or perhaps cause the plume to be missed entirely
VOLATILES
First volatile compound emissions will be considered. The ge-
ometry is shown in Fig. 2. Note that geometrical effects discussed
331
-------
332
OFFS1TE SAFETY
Air k
T
• Ground Surfac*
Soil
Mitt
Upper Level of
* Bulk Wastes
Upper Level of
volatile:
Figure 2.
Volatile compound emissions from buried waste
earlier are neglected. Volatiles must diffuse through the bulk waste
then through the soil then into the lower atmospheric boundary
layer. Conservation of flux F gives:
ID
r -
(Co -
* "
(C2-C)'
where the k's are mass transfer coefficients in each regime and the
C's are concentrations at the locations shown. Any significant wind
will rapidly remove material once it enters the windstream, hence
C <). ID
dt o w
which can be combined with Equations 3 through 5 to give:
2 t H
(MT)
which together with Equations 3 through 5 now provides a com-
plete description for the emissions. Several comments are in order
concerning these equations:
•As overburden is removed hs decreases, ks and becomes unim-
portant in Equation 3. The formula then becomes the same as
for evaporation from bulk waste at the surface.
•When there is significant overburden soil moisture makes a large
difference through the factor (PJ10 3 in Equation 6. The flux can
decrease by an order of magnitude between wet and dry soil.
•The dependence of concentration on wind speed will range from
V"1 to V~ -22 depending on whether or not ka is controlling rather
than kw or ks. Thus low wind speeds will produce peak concen-
trations although emissions are greatest under high wind speeds.
FUGITIVE DUST
Equation 7 for ka also applies to fugituve dust. Its usefulness
however is negated by two factors. First, the amount of suspend-
able dust is a priori not known. Second, dust particles are bound to
each other and to the substrate by a variety of forces. Because of
these difficulties one falls back op a few rules of thumb, although
these have not been developed spe'cifically for excavation sites. Two
such rules are:4'5
•Fugitive emissions are negligible if soil moisture content is more
than 4% by weight.
•Emission rates range from 0.1 to 0.3 Ib/ton of material handled
at 1 % moisture content based on experience with crushed gravel
and sand.
Figure 3.
TRC MEDUSA fugitive dust monitoring system
-------
OFFSITE SAFETY
333
Clearly there are opportunities for substantial improvements in
modeling fugitive dust emissions due to remedial activities at hazar-
dous waste sites. Activity specific measurements are badly needed
particularly with controls such as misting and fogging in use.
MONITORING
Some of the presently available methods for monitoring are
shown in Table 2. The large scale (9.4 m horizontally or vertically)
"medusa" hi-vol array developed at TRC for fugitive
measurements is shown in Fig. 3. With this array it is possible to tell
whether the plume centerline has been captured or not unlike the
use of a single monitor. The advantage of a hi-vol system as com-
pared to a nephelometer is that concentrations are available for
subsequent laboratory analysis. This is important since a variety of
emission sources are likely to be present, e.g., background or
equipment emissions.
The TRC portable wind tunnel used for volatile emissions
measurements is shown in Fig. 4. It can be assembled where terrain
permits in sections up to 8.8 m, a long fetch being desirable to
achieve an equilibrium solution. The wind tunnel method for
estimating volatile emissions is only valid when k^> ks, kw, i.e.,
appreciable wind speeds.
COMMUNITY CONCERNS
Dispersion modeling may be desirable in conjunction with am-
bient monitoring to extrapolate to between monitor values, to
larger distances or to different meteorological values. Particular
emphasis may be placed on densely populated regions or potential-
ly sensitive receptors as schools or hospitals.
Table 2.
Monitoring Methods
Method
Fugitives
Multiple Hi-Vols
Nephelometer
Volatiles
Sorbent Filters
Portable Wind Tunnel
Comment
Area monitoring needed to.
capture source
May be confounded by diesel
equipment emissions or
fogging/misting operations
Ambient measurement
Direct emissions measurement
Varlac to
Control
Bloxtr
Speed
Sampling. Probe for
Trench Samples
Lead Height
to Aid Sea)
3-Inch Thick
•cld-Reslstant
Foam Cushion
for Seal
Figure 4.
Schematic of portable wind tunnel
Odors and toxicity are not synonymous. Public information pro-
grams which distinguish odor and health impacts may be useful.
Separate odor modeling and analyses as described in the paper by
Duffee and Stankunas6 at these proceedings may be necessary.
ACKNOWLEDGEMENT
Portions of this work dealing with choice and application of
mass transfer or k values were carried out as a consultant to
GCA/Technology Division under EPA Contract No. 68-02-3168,
Technical Service Area 3, Work Assignment No. 82.
REFERENCES
1. Millington, R.J. and Quirk, J.P., "Permeability of Porous Solids,"
Trans Faraday Soc. 57, 1961.
2. Farmer, W.J., Vans, M-S., Letz, J. and Spencer, W.F., "Land Dis-
posal of Hexachlorobenzene Waste," USEPA Report EPA-600/2-
80-119, August 1980.
3. Sutton, O.G., Micrometeorology, McGraw-Hill, New York, 1953.
4. Cowherd, C. et al., "Development of Emission Factors for Fugitive
Dust Sources," Midwest Research Institute, NTIS NO. P.B.-238-262,
1974.
5. "Compilation of Air Pollutant Emission Factors," Third Edition,
USEPA, AP-42, August 1977.
6. Duffee, R.A. and Stankunas, A.R., "Estimating Vapor and Odor
Emission Rates from Hazardous Waste Sites," these Proceedings.
-------
SAMPLING TECHNIQUES FOR EMISSIONS MEASUREMENT
AT HAZARDOUS WASTE SITES
CHARLES E. SCHMIDT, Ph.D.
W. DAVID BALFOUR
ROBERT D. COX, Ph.D.
Radian Corporation
Austin, Texas
INTRODUCTION
Emission measurements at hazardous waste sites provide a
mechanism by which existing and potentially hazardous conditions
can be assessed and evaluated. Emission measurements are used to
estimate the amount of a specie(s) emitted from a given surface area
to waste exposed to the atmosphere over time. These data are
necessary to estimate the population exposure through modeling ef-
forts. The exposure level can be subsequently used to quantitate
risk and assess potentially hazardous conditions at or near waste
facilities. Involvement in hazardous waste facility assessment,
evaluation and remedial action has necessitated improved sampling
techniques for emissions measurements.
A number of sampling approaches,' both direct and indirect, are
reported in the literature that could be used to obtain emissions
data at hazardous waste sites. An example of an indirect method is
the concentration-profile technique described by Thibodeaux.2'3
Emission measurements are obtained by measuring the concentra-
tion of specie(s) at various heights above the waste surface. The
concentration data, along with meteorological data are used to
estimate the emissions from the waste based upon a micrometeoro-
logical model. An example of a direct emission measurement is the
isolation flux chamber technique.4'5'6'7 Here, an enclosure is placed
on a waste surface and emission measurements are made directly by
measuring the concentration of specie(s) from a known sweep air
flow rate for a given surface area of waste.
Both of these sampling approaches provide quantitative emis-
sions data needed for hazardous risk assessment. In this paper, the
authors present examples of both indirect and direct emission
measurement sampling techniques that have been used successfully
by Radian to measure emissions from waste facilities. Sampling
methodology design, materials and equipment used, advantages
and disadvantages encountered, as well as example results are
discussed.
INDIRECT EMISSION MEASUREMENTS—
CONCENTRATION-PROFILE TECHNIQUE
Organic emissions from waste lagoons have been measured in-
directly using the concentration-profile technique.2'3 As originally
conceived, this technique involved collecting organic vapor samples
in liquid oxygen cooled glass bead traps at 6 logarithmically spaced
heights above the water surface. Wind speed and temperature pro-
files were simultaneously obtained and then were used for calcula-
tion emission rates using a micrometeorological model. The con-
centration profile technique has been used by Radian to measure
organic emissions at two wastewater treatment facilities for the
organic chemical manufacturing industries.8
Methods and Materials
A 6m pontoon boat was outfitted to perform sampling on lagoon
surfaces. The boat contained a 3m sampling mast, on-board com-
puter, power supply and breathing air (Fig. 1). The sampling mast
consisted of six wind speed and temperature sensors, six air sampl-
ing probes and transfer lines, all at logarithmically spaced intervals
above the water surface, and one wind direction sensor.
Since cryogenic traps are somewhat impractical for field use,
evacuated stainless steel canisters were used to collect integrated air
samples over a 20 min period. After collection, the samples were
pressurized with ultra high purity nitrogen to approximately 15 psi.
The air collection system did not require power, and alleviated pro-
blems due to contamination and moisture condensation, which are
commonly encountered with pumping systems.
MAST 10* HIQH
SENSOR ARM—'
TEMPERATURE
SENSOR
1 WIND 1
I SENSOR |
1
AIR SAMPLING 1
PROBE |
WIND SPEED
TRANSLATORS
, D D
1 D D
J TEMPERATURE [^ )l(
TRANSLATORS U LJ
, D D UI'"
LJ
1 D D
LUrUJ
j 3 WAY VALVE -\_ SOURCE
1 1 r\ V h
/ \J " 1
STEEL TUBING REGULATOR II
( 1
CANIS1ER
Figure 1.
System Block Diagram for the Concentration-Profile Technique
Field Sampling
Meteorological data and air samples were collected continuously
over 20 min periods. In addition, water samples, water
temperature, and relative humidity were obtained during each
sampling period. To properly fit the concentration-profile micro-
meteorological model, the boat was positioned at least 50 (and
preferable 200) times the distance of the height of the dike or other
significant wind obstruction, away from the edge of the pond
(along the direction of average wind). Profile measurements taken
closer than this cannot reflect the logarithmic profile and conse-
quently emission data may not be valid. Also, data were only col-
lected during stable atmospheric conditions (Richardson numbers
greater than -0.1).
Results
Results of three concentration-profile tests performed at each of
two sites are presented in Table 1, while a typical species profile i»
shown in Fig. 2. The logarithmic wind speed profiles, chemical con-
334
-------
OFFSITE SAFETY
335
centration profiles, and resulting flux rates are presented. Standard
errors, correlation coefficients and T-values are also reported for
estimates of slopes.
Flux rates (indirect emissions) for benzene, diethyl-ether, indene,
styrene, and total aromatics were measured at the first site which
was an aerated treatment basin. Flux rates for benzene, acetone
and cyclohexane were measured at the second site, which was a
S1te-2 Site Pos1t1on-3 Compound-Cyclohexane
C 0.0025
E
O 0.0015
N
] 1.5 30 3.5 40 4.5 so
Log Sensor Height (CM)
Figure 2.
Specie Profile, Plot of Concentration-vs-Height Above
1 Emission Surface
Site 2—Site Position 3 (Compound-Cyclohexane)
nonaerated basin. In general, standard errors for measured flux
rates were significant, due to the low levels of organic species which
were measured in the air.
Discussion
The concentration-profile technique has provided emissions
measurement data over a range of natural conditions, with minimal
disruption of natural processes. There are, however, disadvantages
of this technique.
The sampling apparatus is bulky and expensive. It is difficult to
accurately determine temperature profiles within 3m of a surface at
ambient temperature. Also, since organic species concentrations
decrease logrithmically above the sampling surface, accurate
measurements at upper heights can be difficult due to analytical
sensitivity limitations. Finally, restrictions of the Concentration-
Profile model do not allow sampling on small lagoons which poten-
tially could produce serious emissions. Meteorological restrictions
of the model also limit its utility. Despite these limitations, indirect
emissions measurements using the concentration profile technique
and emissions estimate provide reliable emissions data for risk
assessment for applications such as these which are difficult or im-
possible to sample by other conventional techniques.
DIRECT EMISSION MEASUREMENTS-
ISOLATION FLUX CHAMBER TECHNIQUE
Isolation flux chambers of various size and shape have been used
by Radian. Each of the chambers described below was designed for
a specific measurement application, at a waste site. However,
all types can be discussed in general terms (Fig. 3).
An isolation flux chamber isolates a given soil/waste surface
area. Clean, dry sweep air is added to the chamber at a fixed, con-
Table 1.
Calculated Chemical Flux Rates and Statistical Regression Parameters Using Indirect Emission Measurements
Ulndspeed (cm/sec) va Chemical Cone (ng C/cm1) vs . 2
Ln Sensor lit (cm) Ln Sensor lit (cm) l"8 C/Cm scc'
Slope Std.
Site/Sample n Estimate Error r T-value Species
FB-1 5 73.0 7.7 0.98 9.5 Benzene
Dlethyl Ether
Indene
Styrene
Total Aromatics
FB-2 5 73.0 7.7 0.98 9.5 Benzene
Dlethyl Ether
Indene
Styrene
Total Aromatics
FB-3 5 56.2 17.4 0.88 3.2 Benzene
Dlethyl Ether
Indene
Styrune
Total Aromatics
TB-1 6 99.5 12.6 0.97 7.9 Benzene
Acetone
Cyclohexane
TB-2 6 123.7 7.2 0.99 17 Benzene
Acetone
Cyclohexane
TB-3 6 114.3 16.3 0.96 7.0 Benzene
Acetone
Cyclohexane
n
5
5
4
5
5
5
5
3
3
5
6
6
6
6
6
4
5
4
2
5
2
6
5
5
Slope
Estimate
-0.0054
-0.0048
-0.0039
-0.0010
-0.0104
-0.0066
-0.0103
-0.0048
-0.0026
-0.0192
-0.0030
-0.0039
0.0007
-0.0003
<-0.0042
-0.0007
-0.0008
-0.0005
-0.0003
-0.0026
-0.0006
-0.00004
-0.0027
-0.0008
Std.
Error
0.0030
0.0034
0.0010
0.0002
0.0045
0.0024
0.0017
0.0008
0.0035
0.0031
0.0023
0.0014
0.0005
0.0008
0.0057
0.0004
0.0010
0.0001
A*
0.0009
**
0.0004
0.0010
0.00005
r
0.73
0.63
0.95
0.94
0.80
0.84
0.96
0.99
0.59
0.96
0.55
0.82
0.57
0.16
0.35
0.77
0.44
0.95
1.0 •
0.84
1.0
0.04
0.84
0.99
T-value
-1.8
-1.4
-4.1
-4.8
-2.3
-2.7
-6.0
-6.4
-0.73
-6.3
-1.3
-2.8
1.4
-0.33
-0.74
-1.7
-0.85
-4.5
—
-2.7
—
-0.08
-2.7
-17
Est Ima te
0.0679
0.0604
0.0357
0.0104
0. 1169
0.0761
0.1207
0.0410
0.0250
0.2001
0.0336
0.0442
-0.0054
0.0025
0.0425
0.0077
0.0121
0.0051
0.0032
0.0397
0.0058
0.0004
0.0422
0.0087
Std.
Error
0.0544
0.0543
0.0163
0.0044
0.0643
0.0500
0.0582
0.0230
0.0440
0.0914
0.0426
0.0558
0.0045
0.0090
0.0902
0.0049
0.0152
0.0015
0.0227
0.0138
0.0215
0.005J
0.0207
0.0019
**A standard error of 0.002 ng c/cm1 was used In computing the standard error of the flux rate estimate.
Notations; n number of observat ions
r correlation coefficient
T-value is the t-stallstlc for testing the null hypothesis that the slope equals zero.
-------
336
OFFSITE SAFETY
trolled rate. The volumetric flow rate through the chamber is
recorded and the concentration of the species of interest is
measured at the exit of the chamber.
The emission is calculated using the following equation:
E =
(Q)
where:
Ex = emission rate of species x, /tg/m2 min ~ '
Cr = measured concentration of species x, ppmv converted to
/ig/m3
Q = sweep air flow rate, mVmin
A = exposed surface area, m2
Clean
sweep
air
vol.
unit time
Species
Cone 'n
^
mass
unit vol.
Species emission
(mass/unit time/surface area)
Figure 3.
Concept of an Isolation Flux Chamber
The chamber shape or geometry determines the chamber volume
and source exposed surface area. The chamber volume must be
small enough and/or the sweep air flow rate high enough that the
response time of the chamber is short. The response time is typical-
ly characterized by residence time. The chamber resident time (T) is
a function of chamber volume (V) and sweep air flow rate (Q). The
quotient of volume and flow rate (V/Q = f) is the theoretical resi-
dence time. Three to five residence times are needed to establish
steady-state conditions in the chamber at which time representative
sampling can occur.
The compromise between chamber design and operating
parameters is also influenced by the expected species concentration
and the lower limit of the analytical method of detection. The
source surface area permitted along with the sweep air flow rate
and emission source strength determines the concentrations of
emissions species of interest. Thus, low source emission can be
compensated somewhat by chamber design and operating condi-
tions.
Methods and Materials
Undisturbed and disturbed surface isolation flux chamber
measurements were made using the surface flux chamber shown in
Fig. 4. A clear, acrylic, commercially available sky light was used as
the chamber body. The chamber had a volume and surface area ex-
posure (once placed on the soil/waste surface) of 26.0 liters and
0.319m2, respectively. The chamber was stirred with an 8 bladed,
8.9 cm diameter impeller driven by a 12-volt DC motor. The sweep
air was introduced from a bottle supply, regulator and rotometer
through 0.64 cm Teflon® tubing, 0.64 cm stainless swage bulkhead
fitting, and a .15 cm Teflon® inlet line extending to a corner of the
chamber in close proximity (but not venting on) the undisturbed
soil/waste surface. The chamber output manifold consisted of a
0.64 cm stainless swage bulkhead fitting on 0.64 cm Teflon® tub-
ing leading to the instrument manifold. The entire internal surface
area of the chamber and associated components were Teflon*
coated. The chamber material, prior to coating, had a transmit-
tance rating of 92% for visible light and 85% for solar energy
(manufacturer published values). The range of sweep air flow rates
used was 3.1 to 23 1/min. resulting in a residence time of 1.4 to 1.1
minutes).
The surface flux chamber had a narrow (0.64 cm) lip as a footer
on the bottom edge of the chamber. No attempt was made to force
seal the chamber on the soil/waste surface. Any attempt to do so
would either introduce an unwanted effect (i.e., contaminant
source or sink) or constitute a surface disturbance. Therefore, the
output manifold was operated on negative pressure relying on the
combined instrument sampling pump motive. The total consump-
tion of output gas was less than 1 1/min with a larger input flow.
The excess chamber gas was vented at the chamber/surface inter-
face without effect on emission rate measurements, since the
chamber is a well-mixed system.
Downhole emission measurements were performed at various
depths inside a hollow-stem auger. A schematic of the downhole
chamber is shown in Fig. 5. The chamber was fabricated from 6.4
cm I.D., 7.6 cm O.D. Plexiglass pipe with a 0.64 cm thick, 7.6 cm
diameter Plexiglass flat cemented on the tube top constructing the
chamber. The exposed surface, once the chamber was placed on the
soil/waste surface, was 0.00318 m2. The sweep air was introduced
from a bottle supply, regulator and rotameter through 0.64 cm
Teflon® tubing, 0.64 cm stainless swage bulkhead fitting, and a 18
cm Teflon® inlet line ending in a 90 ° glass bend. Thus, the clean
sweep air was introduced at near bottom of the chamber in close
proximity (but not venting on) the core hole surface.
The chamber output manifold consisted of a 0.64 cm stainless
swage bulkhead fitting and a 0.64 cm Teflon® tubing leading to
the instrument manifold. The chamber input and output lines were
13.m in length facilitating flux measurements to a 9m (Below Land
Surface) depth. Both Teflon® coated and noncoated chambers
were used in the study with no detectable difference observed
(shown by quality control simulations).
The range of sweep air flow rates used was 1.5 to 12 1/min
monitored by a calibrated rotameter. The top of the chamber was
weighted which reduced raising and lowering difficulties in the
hollow auger. The chamber was attended by a 0.64 cm steel support
cable.
Although the chamber was weighted, no attempt was made to
force seal the chamber on the core hole exposed surface. Rather,
the output manifold was operated on negative pressure relying on
the combined instrument sampling pump motive. The total con-
sumption of output gas was less than 1 1/min with a larger input
flow. Thus, the excess chamber gas was vented at the chamber/sur-
face interface.
Subsurface emission measurements were made at shallow depths
(0.6 m and 1.6m BLS) using a slightly different technique. Subsur-
face emission rates were obtained by driving specially designed
ground probes into the soil/waste fill and passing clean sweep air
0.28* INPUT
LINE, TEFLON
0.25* OUTPUT
LINE. TEFLON
Figure 4.
Schematic Diagram of the Surface Isolation Flux Chamber
-------
OFFSITE SAFETY
337
0.25' OUTPUT
LINE, TEFLON
WEIGHTS
T LENGTH OF
TEFLON (0.25')
. SAMPLING:
Tip dlaaatar la aaaa aa
plpa O.D., forming vapor
during driving. Shapa cantara
tip during driving, avoiding
loll entry to plpa.
Tuba la raisad 2"
to aeca*a soil
vapora. Tha 2"
diffaranca la
enaurad by dlf-
feranca batvaan
S plpa and cantral
I staal rod.
s
Figure 5.
Schematic Diagram of the Downhole Isolation Flux Chamber
Figure 6.
Schematic Diagram of the Ground Probe
through the probes. The concept behind the shallow mapping in-
vestigation is identical to that of the isolation flux chambers
described earlier.
The ground probes were fabricated out of 2m lengths of
Teflon® -coated schedule 40 galvanized steel pipe. The probe and
associated equipment is shown in Fig. 6. Teflon® -coated iron drive
heads with support cables were fabricated to afford probe installa-
tion and monitoring at 0.6 m and 1.6 m BLS depths. The probes
were fit with drive heads, capped, and driven into the soil/waste
manually. Once in place (i.e., 0.6m) the probes were pulled 5 cm
off the drive heads, exposing a small surface of material at said
depth. A Teflon® input line, 1.6m long so that sweep air was add-
ed close to and exposed surface. The bottled sweep air flow rate
was controlled by a regulator and monitored with a rotometer. The
output line consisted of a 0.64 cm Teflon® tube which extended
2.5 to 5.1 cm into the probe and delivered the output to the output
manifold and instruments.
SO2 and THC were monitored using the analyzers described in
Table 2. The range of instrument(s) response (SO2 : 0.005-10 ppm,
1.0-5,000 ppm and THC : 1-100,000 ppm) provided for emissions
monitoring over a wide range of waste types and emission levels.
Ambient and waste surface temperature were also monitored con-
tinuously using K-type thermocouples with direct digital tempera-
ture readout. The SO and THC instrument response was recorded
using portable Soltec strip chart recorders while the temperature
was manually recorded.
Grab samples were collected for hydrocarbon and sulfur specia-
tion by GC, GC/MS analytical speciation and quantitation. The
hydrocarbon/sulfur speciation sampling and analytical methods
used are described in detail in the accompanying manuscript in
Table 2.
Description of Analyzers
Manufacturer
Model
Technique
Precision
Sensitivity
Response Time
Range
Power Supply
SO2
Theta'Sensor, Inc.
1041
Electrochemical11
1% F.S.
1% F.S.
60 seconds
0-500 ppmv
0-l,500ppmv
0-5,000 ppmv
AC
THC
CSIa
SA 165-3
Flame Photo-
metric
1% F.S.
10 ppb
15 seconds
0-1.0 ppm
0-.5 ppm
DC
Century System
OVA-1
GC/FID
10% for Standard
Analyses
1 ppm (methane)
1 second
1-10,000 ppm
1-100,000 ppm
logarithmic
DC
Service Life
(continuous
use/charge)
Weight
6.8kg
4 hours
9.1 kg
8 hours
6.4kg
a. Sensitive to all "single atom sulfur species" noted as SO2*, response as ppm 202-
b. Interferences from hydrocarbons are avoided using a proprietary scrubber/prefilter.
these proceedings entitled: "Screening and Analysis Techniques for
Organic Vapor Emissions for Hazardous Waste Disposal," by
R.D. Cox, K.J. Baughman, Radian Corporation, Austin, Texas.
Field Sampling
The generic sampling procedure for emission measurements us-
ing the enclosure technique is summarized below:
-------
338
OFFSITE SAFETY
•Locate equipment instruments at the sampling location and docu-
ment location time, conditions, air and surface temperatures
•Begin sweep air flow and address instruments and recorders
•Interface enclosure to waste, record time
•Monitor emissions documenting steady-state concentrations after
3-5 residence times
•Record temperatures
•Collect grab samples if desired
•Remove enclosure, end determination
•Relocate to next sampling location
Appropriate quality control both for analytical instrumentation
and chamber operation was included to document the accuracy and
precision of results.
Results
A theoretical range of measureable emission rates for the three
sampling techniques described are presented in Table 3. The
reported theoretical emission rates are based on practical sweep air
flow rates for field measurements (with consideration given to
enclosure residence time). The design of enclosure greatly affects
the emissions range of the technique. The analytical sensitivity used
for these estimates are those of the instruments used during field
monitoring. Design of the enclosure, field operation of the device
and the range of analytical response can be attenuated to specific
sampling needs to specialize the sampling technique for certain ap-
plications.
Table 3.
Theoretical Range of Emission Per Enclosure
Isolation
Chamber
Surface
Downhole
Ground
probe
SO2(mg/m3 min-»
Mln Max
THC(mg/m3 min-
Min Max
2.2 x 10-'
5.3 X 10-'
2.6 x 102
1.1 x 106
5.0 x 107
4.6 x 109
5.4
1.3 x 10'
6.5 x 103
5.4 x 10s
2.5 x 107
2.3 x 109
An example emissions profile for a gas species over a solid waste
site using the surface isolation flux chamber is given in Fig. 7. The
rise in species concentration up to about time 3r is a characteristic
build-up of species concentration approaching steady-state condi-
tions in the enclosure. After approximately 3r, the fluctuations
observed represent variations in emission due to several factors in-
cluding: natural emission processes, temperature, sky cover (ultra-
violet radiation), surface disturbances, etc. An average emission
value and uncertainty can be used to represent the emission for the
isolated waste.
Downhole emissions data using the downhole isolation chamber
can be correlated to drilling data in solid waste disposal facilities
generating vertical survey information. Core hole data plotted with
emissions data can provide subsurface information that can be
useful in remedial action activities. Data from two drilling explora-
tions are shown in Fig. 8a and 8b for two different waste sites. Data
for organic and inorganic species are shown in both plots. The
variability in emission levels and relative emissions between species
of interest, as shown here, are often quite large depending on the
waste type investigated.
Tim* (Residence Time Units)
Figure 7.
Emissions Profile of Gas Species from a Waste Disposal Facilty
Measured using the Isolation Flux Chamber Technique
Ground probes can be used to obtain cost effective shallow sub-
surface emissions data over a large waste area. Results from survey
studies can be used to map out waste boundaries in solid waste
disposal facilities. Again this data is valuable for remedial action
activities.
10* Of EMISSIOHS (NUS/SUVACE AU*.
a *
a
Figure 8(a).
Emissions Results for Vertical Profile in Waste Deposit
LOG OF EMISSIONS (MSS/SJRFKE WE*. TIKI
Figure 8(b).
Emissions Results for Vertical Profile in Waste Deposit
Other data that can be obtained using the probes describe the
emission potential of waste over long periods of time. If ground
probes are left sealed in solid waste beds and monitored daily for
long periods of time, an indication of prolonged emission potential
can be assessed. A plot of relative emission versus time, as shown in
-------
OFFSITE SAFETY
339
g
la'"1
UJ
UJU
SI
»l
PPP.
p p p p p
-t-
-t-
10
15
25
Peak
20
DAYS
S - Steady State
30
35
Figure 9.
Plot of Emissions Versus Time From Ground
Probe Emission Monitoring
Fig. 9 illustrates this application. Each daily measurement shown
reports an instantaneous emission value or peak (P) value and a
steady-state (S) value representing an average reading. The trend in
emission potential indicates, in this application, a rather steady (but
slightly decreasing) emission potential over an extended duration.
Depending on the desired information, the ground probe emission
monitoring technique can be a valuable sampling tool.
Discussion
The single most important advantage to monitoring emissions us-
ing enclosures is that a direct measurement is made for an
"isolated" emission source where all the necessary parameters,
physical and chemical, are controlled. This allows for data that is
representative of the emission process under set conditions affor-
ding intercomparison of data. Further, the concentration of species
of interest in the enclosure outlet stream can be varied somewhat by
controlling the chamber sweep air flow rate and exposed source
surface area. Another advantage is that meteorological parameters
which influence other sampling methodologies have little effect on
enclosure sampling because the experimenter has control over the
experimental parameters. The technique can be made specific for a
sampling program depending on the emission source type
(chamber/source interface) and can be designed for specific opera-
tional considerations or sampling requirements. Thus, the
enclosure method of direct emissions monitoring offers a specific,
reproducible approach to emissions monitoring for applications,
such as this one, and is capable of providing a comparable,
representative data set.
The disadvantages of the technique are concerned with possibly
creating, through the act of monitoring, an artificial emissions con-
dition. The enclosure design and operation must minimize the
potential of altering the natural emissions process (by covering the
emissions source and introducing a sweep air to the chamber) and
the interaction of the emissions output with the enclosure, output
manifold and sampling lines.
REFERENCES
1. Sekulic, T.S., and Delaney, B.T., "Assessing Hazardous Waste Treat-
ment Facility Fugitive Atmospheric Emissions." 4th Symposium on
Fugitive Emission Measurement and Control, 1980, 119-135.
2. Thibodeaux,, L.J., Parker, D.G., and Hack, H.H. Measurement of
Volatile Chemical Emissions from Wastewater Basins. USEPA, Cin-
cinnati, Oh.
3. Thibodeaux, L.J., and Hwang, S.T., "Landfarming of Petroleum
Wastes—Modeling and Air Emissions Problem." Environ. Prog. I:
1982, 42-45.
4. Zimmerman, P., "Procedures for Conducting Hydrocarbon Emission
Inventories of Biogenic Sources and Some Results of Recent Investi-
gations." Presented at the 1977 USEPA Emissions Inventory/
Factor Workshop, Raleigh, NC, Sept. 1977.
5. Adams, D.T., Pack, M.R., Bamesberger, W.L., Sheppard, A.E., and
Farwell, S.O., "Measurement of Biogenic Sulfur-Containing Gas
Emissions from Soils and Vegetation." Presented at the 71st Annual
Meeting of the APCA, Houston, Tx, June, 1978.
6. Hill, F.B., Anega, V.P., and Felder, R.M., "A Technique for Measure-
ment of Biogenic Sulfur Emission Fluxes." J. Environ. ScL, Health,
A13, 1978, 199-225.
7. Adams, D.F., "Sulfur Gas Emissions from Flue Gas Desulfurization
Sludge Ponds," JAPCA, 29, 1979, 963-968.
8. Cox, R.D., Steinmetz, J.L., and Lewis, D.L., "Evaluation of VOC
Emissions from Wastewater Systems (Secondary Emissions)," Vol I.
USEPA Contract No. 68-03-3038/SBR04. Radian Corporation, Austin,
Tx, 1982.
-------
PUBLIC PARTICIPATION IN HAZARDOUS WASTE
SITE CONTROL—NOT "IF" BUT "HOW"
RICHARD A. ELLIS, Ph.D.
Regional Waste Management Program
Tennessee Valley Authority
Chattanooga, Tennessee
ROBERT W. HOWE, Ph.D.
U.S1. EPA Instructional Resources Center
The Ohio State University
Columbus, Ohio
INTRODUCTION
Environmental problems requiring engineering solutions usually
have one thing in common: they are site specific. Site-specific ac-
tions have a factor in common: they have local publics which are
directly and unequivocally impacted by the action. This is true of
any project producing environmental impacts, from highway con-
struction to dam building. For intensity of impact perceived by the
public, however, chemical waste disposal facilities—and especially
the discovery and attempted reclamation of illegal chemical waste
dumps—lead the list.
In undertaking to plan and implement projects with impacted
publics, it is wise to take heed of what was stated during a Senate
hearing on public participation:
...policies formulated without consulting the interests involved
may lead to decisions which are so far outside the reigning or
conceivable political consensus, or which make such unac-
ceptable demands on a particular interest group that the re-
sistance will thwart the action entirely.'
The overriding emotion and residual distrust which frequently
characterize hazardous waste facility siting or dump cleanup, the
loss of time, effort, and money invested in unimplementable plans,
and the damage to the credibility of the planners may result in
serious delays in effective cleanup or establishment of environmen-
tally acceptable disposal facilities.
Everyone who has seriously approached the examination of in-
stitutional factors associated with hazardous waste site establish-
ment, cleanup, or control has encountered the exhortations of
"shoulds" and "oughts" of how to involve publics effectively. Un-
fortunately there is rarely any empirical observation to back up
these prescriptions. As Wengert2 has chronicled:
The phrases "public participation and citizen involvement"
have many meanings and connotations depending on the situa-
tion to which applied.... much of the literature (on public
participation), especially that related to governmental pro-
grams, has tended to be prescriptive and hortatory, abounding
with rhetoric and polemics and resting on unanalyzed premises
and assumptions.
Projects involving the manipulation of hazardous wastes involve
planning. Where such materials are dealt with, a frequent public
perception is that a real and present danger to life and health exists
over a wide radius from the proposed activity. Failure to plan ade-
quately can result in accidents or incidents which threaten the abili-
ty to implement beyond the time of the mishap. This type of error
occurred during an attempt to retrieve a pocket of drummed waste
at a defunct landfill which had been accepting hazardous wastes,
and which has since been included on USEPA's hotspot list for
Tennessee.
Bumpass Cove Landfill
The Bumpass Cove Landfill is located in upper east Tennessee
near Johnson City. It is in a rural, mountainous area. After a
lengthy period (reported by the press to be about 10 years)1 in
which local citizens had fought the dump and petitioned the State
for assistance, the owner of the defunct landfill was ordered to ex-
cavate an area where the citizens claimed a number of drums were
buried in shallow trenches, possibly threatening both groundwater
and surface streams. In the process of excavating in mud and winter
conditions, several of the barrels encountered were ruptured. Nine-
teen persons were evacuated with several reported being treated for
blistered lips, while others reported respiratory problems.4 Officials
were quoted as downplaying the incident,3 but eventually admitted
to error in planning the excavation for winter conditions.3 A local
State representative blasted the operation as lacking in good sense.1
Bumpass Cove has been a particularly good example of polariza-
tion of publics versus implementors in planning levels from the ex-
ample "micro" level all the way through regulatory policy concern-
ing the site. More discussion on Bumpass Cove will follow. The key
point is understanding how such situations can be better planned,
how citizens can be tapped for the knowledge they have that the
planners do not, and how planners can proceed to make the
technical decisions they are paid to make without imposing value
decisions in which they have a right to no more voice than any other
citizen, and perhaps less right than the citizen who must absorb the
impact of the decision.
PUBLIC PARTICIPATION RESEARCH
A research6 project was implemented to analyze public participa-
tion variables for their relationship to environmental project out-
come variables. These variables were identified as being common to
a majority of 105 case studies of environmental project planning in
wastewater management and water development (i.e., flood con-
trol, navigation, and reservoir projects, and highways or road con-
struction). Although hazardous waste management siting or
cleanup cases were not among the cases studied, personal ex-
perience with hazardous waste management, and careful perusal of
the hazardous waste case study literature discloses the presence of
the same variables in most hazardous waste site-specific controver-
sies. Furthermore, repeated statistical analyses produced no indica-
tion of significant differences of participatory characteristics or
correlations among the various project types studied. In other
words, project planners in these three diverse areas encountered
essentially the same planning problems when faced with dealing
with public involvement.
Data Gathering
For this study, the authors first undertook a thorough review of
the post-1965 literature. From this, a list of 161 variables wa»
generated to include factors which the literature described as
representing the possible social contexts of project planning, the
possible factors involved in the process of project planning with
public involvement, and the possible outcomes of project planning*
The variables collected by this inductive approach were then sorted
into clusters representing categories of similar variables. These wet*
340
-------
PUBLIC PARTICIPATION
341
then incorporated into systems components of a conceptual
framework of the public participation process.
The conceptual framework with its sets of variables was con-
verted into a worksheet by which the presence or absence of the
variables could be documented from case studies. Each variable
was given a specific definition for reference during analysis of case
studies. After this step, 105 case studies were analyzed. Data col-
lected from the case analyses were analyzed using standard com-
puterized statistical methods. A correlation matrix of all 161
variables computed against each other was prepared along with
descriptive statistics and frequency information.
Data Analysis
The data generated from computer analysis were then used to
screen variables to identify the strongest associations between
variables identified as independent variables (relating to process
and context of public participation) and dependent variables
(relating to outcomes of project planning). The variables were
screened to assure statistically acceptable frequencies of occurrence
for further analysis. Variables showing statistically significant cor-
relations at p <.05 in quantities greater than predicted by chance
were retained for analysis and only those independent variables ex-
plaining at least 9% of the variance (r 20.30) of at least two depen-
dent variables (or for dependent variables, those exhibiting
r 50.30 with at least two independent variables). The independent
variables and their strongly correlating dependent variables surviv-
ing these screenings were then arrayed to disclose their associations
(Table 1).
Results of Study
It was found from the analysis that project personnel actions that
isolate planning from interactions with the social environment and
inhibit communications tend to result in outcontes that are general-
ly unfavorable to the implementor's interest. It was noted that the
array of independent variables showing strong positive correlations
with, desirable project planning developments (Table 1) was
characterized by the presence of effective information exchange
mechanisms.
It was also found that those variables associated with restriction
of information flow correlated strongly negatively with desirable
outcomes.
The importance of exchange of information between project
managers and public during project planning is indicated by these
findings. Exchange of information functions as a feedback loop
whereby information gathered from the project's institutional en-
vironment (the potential "stoppers") permits the project system to
adjust its actions to enhance its chances of implementation and sur-
vival. A project frequently in the courts defending itself against
elements of society which have arrayed themselves with the intent
of closing it down has allowed many clues from the institutional
and social environment to go past unnoticed or unheeded. A few
trips to the courtroom, even if cases are won, builds distrust and
causes the attraction of new publics to the fray as multi-party in-
terests gravitate toward a polarized conflict of two alliances.'
A slight favoring of information flow from public to project, if
unavoidable and guileless, will not prevent planners from incor-
porating informed public desires in planning. This is not to suggest
that project planners may be cavalier about dissemination of infor-
mation to the public. It was concluded from the study that lack of
preparation of the public for participation (i.e., failure to provide
them with the ability to use pertinent information) frequently is
associated with undesirable reactions on the part of the public. It is,
however, necessary for the planner to actively seek knowledge con-
cerning the public's desires. If these public desires are thwarted, the
public will react with or without company-supplied information.
It seems, from the data, to be particularly important for the im-
plementor to include assessments of public attitudes and needs in
its planning activities, and to seek input in identifying alternatives
for project plans. This garnering of information seems to enjoy a
synergistic enhancement if the agency arms the public with infor-
Table 1.
Correlations of Variables Drawn from a Sample* of Case Studies
of Project Planning with Public Participation
Dependent Variable
Desirable Developments
Information critical
to project planning
becomes available to
agency
Public contributes
information probably
available only from
public
Public needs and
goals identification
Inform and educate
public
Broad socioeconomic
participation
Enhancement of
planning dialogue
High degree of
compromise
Undesirable Developments
Public opposition
develops
Participants perceive
participatory program
to be too limited
Inadequate education
effort—pubic
unprepared to participate
No resolution
r Independent Variable
-0.31 Redistributive policy
-0.31 Small project: high social significance
-0.34 Agency intends to 'sell' premade
decision
0.35 Agency seeks project alternatives
0.43 Agency seeks information on public
needs
-0.44 Redefinition of public demands
0.50 Agency seeks public attitude data1
0.38 Agency seeks attitude data2
0.33 Meetings
0.32 Agency seeks public needs data
0.32 Pamphlets
0.55 Agency seeks public needs data2
0.49 Agency seeks public attitude data'
0.44 Agency seeks project alternatives2
0.34 Meetings
0.34 News releases
0.33 Agency endeavors to inform public
0.49 Agency seeks public needs data2
0.39 Meetings
0.39 Pamphlets
0.36 Agency seeks public attitude data2
-0.33 Hearings
0.32 Workshops
0.32 News releases
0.32 Meetings
0.32 News releases
0.31 Agency seeks public needs data2
0.39 Agency seeks needs data
0.37 Agency seeks public attitude data!
0.36 Workshops
0.35 Meetings
0.33 Agency endeavors to inform public
-0.31 Hearings
0.41 Agency endeavors to inform public
0.35 Agency seeks public attitude data2
-0.34 Redefinition of public demand
0.33 Newsletters
0.36 Agency intends to 'sell' premade
decision
— 0.34 Low temporal relevance'
0.56 Redefinition of public demands
0.44 Agency seeks to 'sell' premade
decision
0.39 No education of participants
-0.31 Low temporal relevance1
0.32 Small project: high social
significance
0.32 No education of participants
0.46 Agency seeks to 'sell' premade
decision
0.39 Redefinition of public demand
-0.36 Temporal relevance low'
0.34 Redistributive policy
0.34 Small project: high social
significance
Agency seeks public attitude data2
-0.31
•N - 105.
1. Project will set no precedents for subsequent projects (highly routine project).
2. Relative to the planned project.
mation that enables the public to return more coherent input to the
agency. The data suggest that the planner consider use of news
releases, pamphlets, and other information dissemination techni-
ques. Care must be taken to stress information rather than persua-
.sion.
-------
342
PUBLIC PARTICIPATION
The planner should consider the hypothesis that information
disseminated and gathered during real-time, two-way communica-
tion episodes such as workshops and open meetings may be the
most valuable. The data show strong correlations between these
techniques and desirable outcomes to project planning. Not only is
information in these instances couched in the immediacy of the ex-
pressed needs of both parties, but the atmosphere of free informa-
tion exchange often enhances the trust shared among participating
parties. If initial information exchanges are perceived by the public
to come at a point after crucial decisions have already been made,
these meetings can be most unpleasant and trust can be difficult to
recoup.
Some variables were noted to be associated with undesirable out-
comes to project planning. Projects arising from redistributive
policy, that is, policy that seeks to take the possessions in the do-
main of one group to enhance the domain of another group, tend
not to be characterized by good flow of information and tend to
reach stalemate where no resolution seems forthcoming. Such
issues tend to be high conflictual and all parties tend to be drawn
into what becomes, inexorably, a conflict characterized by two
camps. As stated before, multi-party conflicts tend to become two-
sided conflicts' and, because a government agency cannot opt out
of the conflict involving projects it must permit or regulate, it may
be forced to choose sides. This forecloses the agency's capability
to manage the conflict.
Data from these cases showed a high association of undesirable
outcomes when redistributive policy was present; however, there
may be some promising techniques for controlling polarization of
the public if action is taken early. The data were not sufficient to
document efficacy, but the technique of using project information
clearinghouses to dispense accurate information to both camps' or
the use of third-party mediation' may help reduce all but the most
rancorous polarization when redistributive policy governs the pro-
ject. Recent contact with regulatory personnel of the Tennessee
Valley States and with some innovative private sector firms
underscore this promise.
Another influential variable couched in the context of the project
is operative in the situation where a small project is characterized
by high social significance. This is often the precise situation
characterizing the hazardous waste disposal site cleanup or
establishment. In such an instance, an unprepared planner could
find literally nationwide attention drawn to a small project. The
data show that, for the sample of cases, this situation tended to
reduce the flow of coherent information and tended to be
associated with stalled projects.
The best approach under these circumstances may be to provide
the maximum possible amount of information about the project,
especially early in the planning stage. It would seem essential that
the public be enlisted in identifying alternatives to what may be a
repugnant, although necessary, project.
Redefinition of demand is an action shown by the data to be
strongly related to numerous undesirable outcomes, and negatively
related to desirable outcomes. When the public states a demand,
and the agency responds not in answer to the demand, but with ac-
tions that seem to imply, "We know best what you really need," it
is reasonable to anticipate a public enraged by patronization, or
convinced that the agency simply is not listening. Further, suspi-
cion that the agency is attempting to manipulate public desires will
surely poison the communications atmosphere. This action exhibits
such broadly undesirable consequences that it should be avoided
religiously.
A factor that tended to be associated with undesirable public
perceptions of the participatory process was the failure to educate
the public to at least the rudimentary technical basis for the project.
This deprives the citizen participant of the ability to critically ap-
praise the project and contribute to its planning. Review of failed
siting attempts in the Tennessee Valley and conversations with
engineers and scientists involved in the attempts reveal a firm belief
by these implementors that they have, in fact, provided all the rele-
vant in formation necessary to show their sincerity in ensuring that
as much as possible has been done to ensure safety and mitigate en-
vironmental problems. This is true, but only works in the unlikely
situation that all the members of the public are of the appropriate
scientific or engineering discipline.
The lay citizen can no more use the elegant information of the
engineer than a human being can digest the cellulose in wood, even
though it has food value to other animals. The information must be
useable and, lest the engineer become arrogant, he or she should
review several of the newspaper accounts such as those about Ten-
nessee's Bumpass Cove, where those who ignored the local citizens'
advice (couched in readily understandable terms), were later moved
to state for the record that they had "erred."'
Especially in projects of a highly technical nature, such as haz-
ardous waste management, education is the keystone of effective
participation. The instances are well documented of failed par-
ticipatory efforts giving way to court battles, or extra legal actions
such as demonstrations, pickets, and sit-ins. These situations exact
great prices in prestige and good will enjoyed by the implementor.
Further—and this has been particularly true in the TVA
region—geologically and economically feasible locations for haz-
ardous waste facilities are limited. Options for cleanup of aban-
doned or illegal dumps are limited. Siting and cleanup attempts
which fail "spend" the limited resources of sites or cleanup op-
tions.
A community mobilized against the presence of a hazardous
waste facility because of a bungled citizen involvement process will
remain "cocked" for remobilization should another siting attempt
be made later at the same site, no matter how sophisticated the se-
cond approach may be. Cleanup activities, once perceived as
threatening a community interest, create mobilized opposition
which for years may forestall effective response, and sustain suspi-
cions that elements of the cleanup team are "in cahoots" with in-
terests antagonistic to the community's welfare. Community
mobilization may completely "steamroller" hazardous waste
management plans—even though opposition is inspired by the
wildest misconceptions about the effects of project implementa-
tion.
An additional word should be mentioned about the research fin-
dings relating to an old favorite for filling the "public involvement
square," i.e., the hearing. Hearings have long been a tool used by
agencies to comply with administrative law concepts of "right to be
heard." In the cases sampled, though, it was noted that formal
hearings had strong negative correlations with several desirable out-
comes to participation efforts. The usual conduct of a typical ad-
ministrative procedure hearing is characterized- by a "we-they"
bifurcation between public and implementor.
Dialogue is procedural, suppressed, and stilted, if present at all.
If large groups are present, the individual with the best insight on a
project issue, if timid, will likely never be heard. For the planner
truly interested in hearing the public, this technique should be
avoided where at all possible. Workshops and meetings, in which
favorable group dynamics operate, show promise of producing suc-
cessful citizen input. Formal hearings are usually antithetical to the
promotion of good two-way communication, the key to effective
citizen involvement in planning.
TVA's Concern
With this research as a backdrop, it has been possible to view
from TVA's vantage point the institutional processes accompany-
ing attempts to clean up abandoned or illegal hazardous waste
dumps within the State, and to site new facilities for sound manage-
ment of hazardous waste.
TVA is a Federal regional resource management agency with re-
sponsibility for integrating the economic development of the region
with wise management of the seven-State area's .natural resources.
It is not a regulatory agency. It has many characteristics of an in-
dustry in that it constructs, maintains, and operates the largest elec-
trical generation capacity in the U.S.—a mix of hydroelectric, coal-
fired, and'nuclear power plants. TVA's involvement in hazardous
waste management results from: (1) a need to manage its own in-
dustrial waste, (2) its interest in the region's ability to attract in-
dustry by providing adequate waste-handling facilities, and (3) its
-------
PUBLIC PARTICIPATION
343
practice of making TVA expertise available to situations where our
assistance can (1) further or (2) enhance the protection of environ-
mental resources of the Valley.
TVA has provided technical assistance in several abandoned
dump cleanups in the Valley, and has had the opportunity to
observe the ebb and flow of conflict around these attempts.
Bumpass Cove, mentioned before, is one of the abandoned il-
legal dumps at which TVA has offered specific technical assistance.
The controversy at Bumpass Cove, ostensibly a sanitary landfill,
began in the mid-1970s when residents claimed hazardous wastes
were being supprptitiously dumped in the landfill. Some com-
plained of health effects. The situation reached a crisis in 1979
when heavy rains uncovered barrels and strong odors and
suspicious substances were noted in the creek at the base of the
landfill. The exasperated citizens, apparently unable to establish a
credible dialogue with the landfill operator or regulators, blocked
the roads to the landfill, preventing trucks from reaching the fill.
Acts of sabotage were dealt vehicles using alternate routes.10
In August of 1980, after the earlier-mentioned evacuation of
residents and discovery of buried drums, the Tennessee State
Health Department ordered the former operators to remove barrels
of toxic chemicals from the area, along with soil proven to be con-
taminated. TVA was called upon as a neutral party to chair a task
force of State agencies, the USEPA, and Bumpass Cove citizens."
Meetings were held, but true to predictions from research, the
lateness of the information exchange has had deleterious effects.
Residents have accused health officials of being liars when they
claimed lack of knowledge of illegal dumping, but other citizens
have been quoted, regarding the task force, as saying, "all we need
is someone we can believe in.'"2
The situation at Bumpass Cove is now chaotic. The State recently
approved a closure plan that, according to press accounts, has not
included nor responded to public demands. A rift has formed in the
citizen's group such that lawsuits have been filed among the
citizens. The developments over the past decade and the current
acrimony which promises to continue are all in keeping with predic-
tions from the research.
Factors which, in this situation, predict undesirable outcomes
have included small projects with high social significance; redefini-
tion of public demands; and perception by the public that involved
action agencies intent to sell a premade decision.
This last item should be noted particularly. It does not state that
responsible parties, in fact, intend to conduct sham public involve-
ment as a post-hoc validation of a decision. It refers, however, to
the public's perception that this is what is happening. What the
public perceives is the public's reality. What is real and threatening
or offensive to the public, the public will act on.
It is absolutely essential that credible information be made
available concerning the sincerity of the implementor's desire to in-
volve the public. The public must then have available information,
interpreted for their use, which puts them on a factual footing that
permits them to appreciate (or reject) the engineer's plans. Im-
pacted citizens must also be able to judge plans and situations from
an informed position if they are to play one of their most important
roles—providing insight, information, and data which may simply
be unavailable to the planner, or which would take great expense to
gather or retrieve. Citizen-provided information of this type has
certainly played a role in the Bumpass Cove situation.
Hazardous waste managers and disposers have been tarred with
the brush of Love Canal, Valley of the Drums, and others. The vast
majority of professional hazardous waste managers which TV has
dealt with have been scrupulously honest, meticulous in safety and
environmental protection, and very conscious of their responsibili-
ty to society. TVA could not continue to operate if they were not,
even with the small amounts of hazardous wastes which it employs
contractors to handle for us yearly.
New facilities in the Valley would greatly reduce the opportunity
for midnight dumpers to tempt customers now faced with expense
of long-haul shipping to distant facilities. New facilities would
reduce the environmental and safety problems involved in risks of
long-haiil transport of hazardous wastes. These, among others, are
compelling reasons for facility siting, yet we have observed sincere,
competent engineering firms stymied by public opposition to siting
of facilities in areas where risks posed by sound operation in
geologically favorable sites would be minimal.
There are a few reasons for this which often seem to be neglected
by planners. First among these is the failure to recognize that no
matter how safe or unobtrusive a hazardous waste facility is, it is
till a noxious concept to a lot of people. Those who live near it are
subject to an increased risk, no matter how small, and may perceive
it as a moderate to great risk (which is, therefore, their reality).
They are asked to endure this concentrated negative for a benefit
dispersed to a wide public not asked to make the same sacrifice.13
Citizens must either be given some assurance that their imbalance
will be mitigated or redressed (and some interesting examinations
of siting incentives aimed at this problem are ongoing),14 or provid-
ed information such that they can participate in making decisions
as to what manner and in what form they will accept this risk on
behalf of society.
What is often observed in retrospect following a failed siting ef-
fort is recriminations by the routed planners that the press
destroyed their chances, or a hysterical public never understood.
The bottom line, however, is that public opposition renders siting
(or any other aspect of hazardous waste management) as infeasible
as many technical flaws, regulatory impasses, or physical condi-
tions.
A recent siting effort was documented" in a paper by a candid
and articulate consulting engineer. It described the intricate care in
determining location for best probably physical attributes for a
hazardous waste management facility in the Tennessee Valley. It
described how the company made no attempt to hide its presence as
it made its tests and surveys of a proposed site. It also plainly states,
however, that it made no attempt to involve the public or inform
the public of its intentions. What occurred was a rapid and unex-
pectedly sophisticated mobilization of public opposition when the
purpose of the survey was discovered. The press covered the
developments heavily, receiving most dramatic quotes from the
citizen opponents.
Eventually, following heated public meetings, disappearance of
previous support from local and State government and opinion
leaders, the company abandoned its interest in the proposed site, at
a loss of investment in purchase option and survey costs. What is
interesting is that the explosion of opposition was blamed in part in
the account on too much public participation, too early. The writer
described the ongoing controversy as a forum for selfish interests.
The research, however, predicts the strong probability of harsh
public reaction if the public perceives that decisions within the
public domain have been made with only post hoc involvement of
the impacted public. Further, it does no good to bemoan the pub-
lic's in ability to understand the fine points of safety features and
sound operating practice. The ability of the public to understand
these points is at least 50% the responsibility of the communicator
to couch them in terms understandable by the public. To assume
that broad operating and construction concepts cannot be com-
municated to the public by an effective education process is an
elitism of its own guaranteed to enrage segments of the public, and
ensures the public is unprepared to participate in a planning process
where citizens are called upon to separate wheat from chaff in both
proponents' and oppontents' arguments.
The beginning of technical reliability and legal permitting or
licensing of processes in getting facilities sited has been underscored
by the efforts of Pyrotech, Inc., of Tennessee. The Pyrotech pro-
cess for disposal of PCBs has been certified for operation by
USEPA and has been tested by independent laboratories such as
Battelle, with very favorable results. It cannot find a home,
however. It has lost the support of local officils in Pyrotech's home
city, where once it had their support. When public opposition
developed, political support swung with it. When the company at-
tempted to site in rural Grundy County, it was turned away by
unexpectedly well-organized public opposition. It then turned to
the outskirts of Chattanooga, an already heavily industrialized city,
as a potential site, and again a rapidly mobilized public played a
-------
344
PUBLIC PARTICIPATION
role in forcing Pyrotech to abandon its hopes to site at its chosen
location.
It is difficult to disbelieve that Pyrotech's perceived failure to
conduct informational meetings and open discussions as promised
contributed to this rejection." Opponents have gone on record as
approving of the effectiveness of the Pyrotech process, but have
argued against the hazards of truck traffic bearing PCBs into the
area with its attendent spill potential, and against the "foot-in-
door" possibility of other "noxious" facilities gravitating to the
area.
Pyrotech is continuing to explore siting possibilities in Chat-
tanooga, and has opened a dialogue with environmental groups in
the area. It will be interesting to follow what appears to be an evolv-
ing siting approach.
Again, citizens not trained in applicable science and engineering
disciplines do not arrive at conclusions concerning complex
technology a priori. The citizens must, if necessary, be taught how
to interpret the information before information is presented.' Also,
as in the Pyrotech situation, if citizens present concerns decrying
transportation-related hazards, the research predicts the hazards to
effective public participation of redefining this demand by respon-
ding with additional assurances of process safety.
A final example of problems with public involvement in waste
management involves the inability to engender public involvement
when it is needed. This situation exists at this time in a Valley coun-
ty characterized by several large, unmanaged open dumps. The
county is rural and poor, but close to an affluent larger city.
TVA has several interests in assisting the local governments in
this county to attack the dumping problem: (1) the dumps are
disincentives for in-migration of workers and industry which would
improve the depressed economics of the county; (2) one of the large
rural dumps is located in an isolated area and suspected of being a
midnight dumping site; and (3) a TVA structure is also threatened
by burning and corrosion from the nearby dump. TVA has worked
with very dedicated county and State health departments to attempt
to enlist local government assistance in cleaning up the dumps and
designing an effective rural solid waste collection system. The
public response, which at first surprised and now frustrates, is the
admonition that the dumps have always been there, and always will
be, and cleanup attempts will not work.
The apathy has characterized both government and much of the
citizenry. Public interest groups in this community are virtually
nonexistent. Traditional advocacy organizations such as the
Chamber of Commerce are small and avoid public attention of any
kind. The situation reflects what opinion leaders in the area
describe as a lack of community pride. Meanwhile, the dumps
threaten community water supplies with potential pathogens and
toxic wastes.
TVA is approaching this problem by first presenting to the local
government the opportunity for a TVA program offering selected
communities assistance in comprehensive economic planning. The
strategy in this case is to target the elevation of confidence and
pride in the community's resources and potentials. A major aspect
of this is technical assistance in waste management to eliminate the
dangerous and ugly dumps, and institute collection systems. TVA
economists and community experts working with the TVA
Regional Waste Management program are beginning to observe a
snowballing of interest in the community's potential. Because this
situation of resistance by essentially the entire spectrum of the com-
munity toward involvement is uncharted territory so far as predic-
tive models go, we are learning as we progress. Predictions for suc-
cess at eliminating the dumping sites are risky at this time, but pro-
spects appear much more favorable than the null prospects prior to
this TVA assistance program.
CONCLUSIONS
The research described has alluded to tentative relationships bet-
ween how an implementor or planner approaches citizen involve-
ment in environmental project planning. Its fundamental finding is
that exchange of information early, often, and all the way through
a project is essential. It points out that this information must be
usable by the public, and, if the technologist cannot completely
convert his science to terms understandable by the man on the
street, he must make the effort to provide enough background in-
formation to educate the citizen to a level of understanding where
the citizen may evaluate the merits of the planners' product without
needing to evaluate the fine points of each equation.
The bottom line is that attempts to avoid public participation or
to offer sham exercises is counter-productive for the planner. If the
citizen is denied the capability to participate, or the opportunity to
participate under conditions which respect the viewpoints of all, he
will participate in other ways. In the arena of hazardous waste
management and facility siting, the odds are on the project oppo-
nent's side. The next stop for the engineer may be on the witness
stand before a nonengineer judge hearing suit over a partially com-
pleted project. Instances are also well-documented of extra-legal in-
terventions which exact great prices in corporate prestige and ex-
tremely adverse publicity.
The authors have cited some observations in this paper of siting
and cleanup efforts here in the Tennessee Valley. They have been
characterized by uncertain or undesirable outcomes. The failures
spend the best locations by assuring the victorious citizen op-
ponents that they were correct in their assumptions about haz-
ardous waste management, and assuring that their model For
response to siting of management facilities is a mobilized, un-
flinching opposition. In an area such as the Tennessee Valley,
suitable areas for hazardous waste/ management are limited. It is
troubling to observe potential sites "spent."
A reason there are no success stories documented in this paper is
that there really are not many to draw on. In preparation of this
paper, several of the success-story disposal facilities' managements
were called. They were asked about how they had sited their
facilities without apparent conflict. The prevalent reply was that
procedural requirements of current hazardous waste laws were not
in place, and they just "did" it. Such would not be the case today.
Students of environmental controversy in the Tennessee Valley
have observed surprising awareness and sophistication in mobiliz-
ing opposition to a wide variety of environmentally impacting pro-
jects, even in some sparsely populated places.
TVA's role in hazardous waste management in the Valley in-
volves a commitment to manage its own hazardous wastes in an en-
vironmentally sensitive way, and in accordance with law.
TVA also maintains a surveillance of public attitudes concerning
where resources should be applied. TVA documents calls to TVA's
Citizen Action Line on topics relating to hazardous waste. TVA
also recognizes the relationship adequate hazardous waste manage-
ment facilities have to the potential for Regional industrial expan-
sion. Moreover, TVA has actively assisted State, local, and Federal
agencies in locating, monitoring, and cleaning up hazardous waste
dumps. Citizen participation in these same issues remains a process
which TVA supports.
REFERENCES
1. Frank, R.A. et al., "Public Participation in the Policy Formulation
Process." Center for Law and Social Policy. Paper submitted for the
record during hearings before the Subcommittee on Administrative
Practice and Procedure of the Committee of the Judiciary, U.S. Sen-
ate, 95th Congress, 1st Sess., on S270 (1977).
2. Wengert, N., "Citizen Participation: Practice in Search of a Theory."
Natural Resources J., 16, 1976.
3. Elizabethton Star, Elizabethton, Tn. Feb. 17, 1980.
4. News-Free Press, Chattanooga, Tn. Feb. 16, 1980.
5. News Sentinel, Knoxville, Tn. Feb. 17, 1980.
6. Ellis, R.A., "An Analysis of the Impact of Public Participation Ac-
tivities in Water and Transportation Projects." Ph.D. Dissertation.
Ohio State University. ERIC Data Base access SE 032 957; Uni-
versity Microfilms (1980).
7. Cobb, R.W. and Elder, C.D., "Participation in American Politics:
The Dynamics of Agenda Building." Johns Hopkins University Press,
Baltimore, Md. 1972.
-------
PUBLIC PARTICIPATION
345
8. Cutlip, S.M. and Center, A.H., "Effective Public Relations."
Prentice-Hall, Inc., Englewood Cliffs, NJ, 1971.
9. Cifrano, Deborah, "Tearing Down the Wall Through Environ-
mental Mediation," Conservation News, 43, 1978, 19.
10. Press Chronicle, Johnson City, Tn. July 27, 1980.
11. Cox, D.B., Milligan, J.D., Carson, H.T., and Hyfantis, G.J., "In-
teragency Coordination in the Investigations and Cleanup of Chemi-
cal Waste Sites." Proceedings, ASCE National Conference on En-
vironmental Engineering. 1981.
12. Journal, Knoxville, Tn. Aug. 5, 1980.
13. Peelle, Elizabeth, Statement of Elizabeth Peelle, Social Impacts As-
sessments Group, Energy Division, Oak Ridge National Laboratory
to the Subcommittee on Rural Development, Rural Development,
Senate Committee on Agriculture, Nutrition, and Forestry. Social
Impact Mitigation and Nuclear Waste Repository Siting. Aug. 26,
1980.
14. Carnes, S.A., Copenhaver, E.D., Sorenson, J.H., Soderstrom, E.J.,
Reed, J.H., Bjornstad, D.J., and Peelle, E., Incentives and Nuclear
Waste Siting: Prospects and Constraints. Oak Ridge National
Laboratory. Energy and Health and Safety Research Divisions.
Dec. 1981.
15. Wilson, L.E. The Hickman County Experience. The Trials of Locat-
ing a Site for a Hazardous Waste Disposal Facility. A Case History.
Paper presented to the WATTEC National Energy Conference and
Exhibition, Knoxville, Tn., Feb. 1982.
16. Times, Chattanooga, Tn. July 9, 1982.
-------
PUBLIC INVOLVEMENT IN RESOLVING HAZARDOUS WASTE
SITE PROBLEMS
HUGO D. FREUDENTHAL, Ph.D.
JEAN A. CELENDER
H2M/Holzmacher, McLendon and Murrell
Farmingdale, New York
INTRODUCTION
In 1978, a rather innocuous ditch which had been used as a
chemical waste disposal site became the battleground for the in-
itial skirmishes of what was to become the major environmental
war of the seventies. Its name, Love Canal, was to become syn-
onymous with public involvement in hazardous waste landfill
issues.
Looking back and trying to analyze the Love Canal conflict,
one tries to assess what went wrong and gain insight into better
ways of handling future situations. The area had been used for
chemical disposal, as was the accepted technology of the day; later,
the land was developed as a school and residential area, with-
out full appreciation of the potential hazard. Love Canal is signif-
icant because it represents a total failure of technologists, regula-
tors, and impacted citizens to communicate effectively with each
other.
As the dangers of Love Canal became evident, the conduct of
public officials led to an environment of mistrust. Adeline Levine,
in her book on Love Canal, states that, "Although residents had
been informed by (government agencies), they were not consulted
as people who might have important contributions to make about
the history of the area or their own experiences, or might have
opinions about what should be done to and for them. What the
residents viewed as secrecy and their exclusion from vital decision-
making was to result in their suspicion that there was a cover-
up."1
In retrospect, it is difficult to accuse the public health officials
of misconduct or insensitivity. They did their job in accordance
with the accepted standards of performance of the day. A signif-
icant social conflict in the era of technology is the relationship
between the professional and the lay public. Does the patient
have a vote in the therapy which the physician deems necessary to
save his life? Do the poor have any input into the manner in which
a benevolent government seeks to raise them to a better life? Can
children voice opinions about how they should be educated? Can
passengers tell pilots how to fly a plane, and can homeowners in-
fluence toxic waste cleanup, when their lives and property values
are equally part of the package?
When society licenses or puts trust in a professional, it signifies
a willingness to accept that individual as a saviour in technical
issues. But society looks at the record and also realizes that the
professional is only a human being, and bridges collapse and peo-
ple die on the operating table in spite of diplomas. As a result,
the legislative/judicial system accepts the concept that the public
does have a valid right to participate in the decisions which di-
rectly affect them.
In this paper, the authors will try to outline some of the mechan-
isms which can be used to ensure that the public has an input into
hazardous waste issues. The process is called PUBLIC IN-
\OL\EMENT and the authors urge its adherence for two rea-
sons:
•Public participation is a proper, socially-mandated procedure,
respecting the rights of the people who are impacted and pay the
bill
•Failure to involve the public in a timely manner can lead to de-
lay, increased cost, and unpleasant legal complications
PUBLIC INVOLVEMENT
The public involvement process can be divided into three phases
(Fig. 1). The first is problem identification, the second is strategy
development, and the third is the public participation program.
Problem Identification
Strategy Development
Public Participation Program
Figure 1.
Public participation process
Problem Identification Phase
The problem identification phase is normally considered to be
the responsibility of technical and regulatory specialists. The prob-
lem generally presents itself in either of two ways. In the first, water
or soil is found to be contaminated and the source traced back to
a hazardous waste disposal site. This process involves hydro-
geologist, engineers, analytical chemists, environmental scientists
and other technical people. In the second, the problem site is
listed as a violation of federal or state regulations by virtue of past
known or alleged dumping activities. This situation involves regu-
346
-------
PUBLIC PARTICIPATION
347
latory officials, legal personnel, inspectors and others with appro-
priate specialized expertise.
There is, or should be, a public portion to this first phase. The
activity seeks to resolve initial public concern about the site and
the related problems. The danger is, that a little knowledge, either
on the part of the technical side or the public, can easily burst into
inflammatory perceptions. Therefore, the social scientists who do
this phase must be extremely careful in what they say and what they
do.
It is best to identify responsible civic leaders early in the pro-
cess and begin the social problem identification with them. The
lowest levels of local government must become involved and kept
informed. Experience has shown that these bodies can become
powerful forces for progress or obstruction. The worst thing that
can be done is to ignore a local group, such as a civic association,
or bypass local lines of government structure.
This involvement can be accomplished best by a consultant who
has intimate knowledge of the local political forces. In the absence
of such a person, the local newspaper or chamber of commerce or
bank may be good places to develop a pattern of local leadership.
Such leadership may occur in places not commonly suspected.
For example, the local volunteer fire department has been the seat
of village political power, and, after all, they are logically the
protectors of public safety.
A clear differentiation must be made between how the public
perceives a hazardous waste issue and how it is addressed by a
regulatory agency. The technologist is paid to achieve compliance
with standards, laws, regulations, and engineering practices. The
police officer who tickets for speeding is not interested in why, but
only that a rule was broken.
The local resident is concerned about himself and the people and
things he holds dear to himself. He fears cancer and the loss of
equity in his most valuable possession, his house. The names of
organic chemicals sound to him as the name of horrible deities
which early man held in awe, and their minute concentrations
are equally as invisibly terrifying. The public is going to react out
of fear, and fear is a very powerful force.
The end product of phase one is shown in Fig. 2. A common
problem is viewed by two forces with different concerns. Phase
two, the strategy phase, seeks to bring these sides closer together
toward a common solution.
The planning for public participation must begin early in the
process, before issues become inflamed by media and political
forces. It takes time. There is a tendency to spend money on con-
struction and to economize on the social issues in a problem. There
should be a monument to the phrase, "If we had only listened."
The strategy process involves numerous meetings between civic
leaders and the technologists.
Strategy Development Phase
Another phrase which has become synonymous with the seven-
ties is "cover-up." The public must never again have cause to sus-
pect that a hazardous waste problem is being so treated, the in-
dustry must learn from the past. The strategy development phase
is more than just a planning process. It is also a mechanism for
bringing the public into the problem. It is a means of educating
them as to their responsibilities, too, because with the right to de-
cision-making participation goes the privilege of being part of the
consequences of the remedial action.
Conflict can not be avoided. It is a basic human process. Accept
conflict as a reality, and plan for it using the techniques addressed
in phase three, the public participation program.
A public participation program promotes conflict resolution by
providing opportunities for individuals and opposing groups to
explore compromise solutions. Six levels of public participation
are identified here. Each represent an increasing intensity of public
involvement. Which level or levels are used is a function of the
degree of anticipated public interest (Fig. 3).
Public
Issues:
• Property values
• Present and future
individual health
• Nuisances
Technologists
Issues:
• Cost of remedial action
• Compliance with standards
• Limitation of liability
Figure 2,
Identification of problem and conflicting issues to be resolved
Level 1—Public Information
The public information component involves merely a dissem-
ination of information to the public through project orientation
devices, such as fact sheets, newsletters, and brochures, with
distribution in a local area. Press releases and public information
announcements may also be prepared for local newspapers and
broadcast media.
The purpose of this informational level is to inform the public
of hazardous waste issues and to report the progress on projects
involving these matters. Although this step is often favored by
the technologists as a cost-effective method, it is a one-way flow of
information devoid of public input.
Because such a program is one-sided and autocratic, a subse-
quent public reaction with strong negative feelings about the direc-
tion of the project can ensue. Residents reading about a closure of
a well near a landfill due to contamination may perceive a greater
health threat than actually exists. In the case of the Port Wash-
ington, N.Y., municipal solid waste landfill, where media releases
concerning methane explosions in homes near the landfill and the
presence of toxins in a well and a local school caused residents
and civic groups to request a "60 Minutes" investigation which was
aired on a major television network. The residents distrusted town
and county officials' reports contending that the evidence of leach-
ate and gases in the landfill represented chemicals normally found
in a variety of consumer products and did not present a signif-
icant health threat.
Residents have alleged, however, that industrial wastes have been
dumped at the landfill. Their concerns that the landfill may be
"hazardous" and a public health threat has pressured local health
departments and the USEPA to conduct additional tests at the
landfill. The release of information in this case only created
greater public opposition and distrust that the conflict could be re-
solved.
The other major defect about the one-sided program is that the
technologists may be totally unaware of a festering public reaction
until it explodes upon them.
4
Degree
ol
Public
Participation
Environmental
Mediation
Public Meeting
Task Force
CAC
Public Survey
Public Information
Complexity ol Problem
Figure 3.
Six levels of public involvement
-------
348
PUBLIC PARTICIPATION
Level 2—Public Survey
The second level of public involvement is a public survey. A pub-
lic survey can consist of a written or oral questionnaire which is
mailed, conducted by telephone, administered door-to-door or pre-
sented at a central location, such as a local supermarket.
The public survey level provides greater opportunity for public
involvement than the information level, since it has the ability to
input citizens' views. But, public surveys also have several defic-
iencies. Telephone surveys often cannot be geographically matched
to a location. In the case of interviewing persons by telephone who
are living near a suspected impact area, it is necessary to also have
their address. Consulting a telephone directory for an address is
time-consuming and may not resolve the problem.
Responses to mailed questionnaires are usually a small per-
centage of the total sent, with 10 to 20% representing a normal
state of return. Furthermore, the respondent has complete freedom
whether or not to respond and as a result, the returned responses
may not represent a cross-section of public views, but only those
individuals with strong feelings one way or the other.
While surveys provide a method of obtaining some public opin-
ion, they are usually a one-time event. They neither track public
perception as a function of time, nor do they produce a mechan-
ism for a dialogue on issues. The public is suspicious of data collec-
tion in this manner.
Level 3—Citizens Advisory Committee
The citizens advisory committee (CAC) is the third level of pub-
lic involvement. The membership of a CAC should represent a
broad case of community interest including residents, elected
officials, representatives of special interest groups, and technical
and environmental experts. The CAC may make recommenda-
tions, assist in the development of a public participation work-
plan, and participate in public meetings, workshops and seminars.
If properly balanced and adequately staffed, a CAC may ensure
functional two-way communication and an on-going link between
citizens and agencies involved in resolving a problem. A CAC
can be an effective public participation mechanism since com-
munity values and goals can be reflected during the planning
stages.
The criticisms leveled at CAC's have basically involved two
problems. First, CAC membership, instead of being broad-based,
is often dominated by persons with the loudest voices and/or those
individuals with a special interest in the issues under consideration.
Secondly, the role of a CAC is strictly advisory in the decision-
making process. Therefore, in order to be effective, this advisory
role of the CAC must be clearly defined to the members and the
general public. Otherwise, they may perceive their powers to be
much greater than they are legally allowed, creating greater polar-
ization of the parties and issues. This may result in the mushroom-
ing of issues beyond a technical level into the political arena.
In the final analysis of the project, the loss of control over tech-
nical issues can leave the participants with a greater-than-ever
feeling of frustration. At this point, they are organized and have
tasted power. They then may seek help in the political arena and
when this happens, several consequences follow.
These are:
•Project delay
•Increased costs
•Potentially greater health risk due to lack of immediate remedial
response
•Reduction of options
•Mandated remedies which may be neither feasible nor econom-
ically desirable.
Level 4—Task Force or Ad Hoc Committee
A task force or ad hoc committee is the fourth level of public
involvement. This body is usually a small group of people who
have been assigned to research a specific problem in a limited time
frame. Its membership should consist of those with the expertise
necessary for the specific problem. In this respect, it does not rep-
resent the broad-based public participation which is desired.
The problem with a task force is that once goals and tasks are
decided upon, it may not have the flexibility to anticipate other
potential problems. This was a criticism cited by Levine of the Gov-
ernor's Love Canal Interagency Task Force. Once work routines
were established for the task force, she found that production took
precedence over flexibility. There seem little, if any, intent to deal
with major problems which "might occur after these decisions were
made and announced.2
A task force is usually advisory in its capacity. It has no author-
ity to make binding decisions. While it is more specialized than a
CAC, its impact is little better.
Level 5—Public Meetings and Hearings
The fifth level of public participation is public meetings and
hearings. Public meetings and hearings can vary from an informal
workshop to a formal, stenographically-recorded hearing. Both af-
ford the opportunity for concerned citizens to present their views,
often as part of a project's permanent record or file. Public hear-
ings are sometimes a legal requirement under certain project re-
view and permitting procedures. The New York State Environ-
mental Quality Review Act allows formal opportunity for public
comment concerning overall hazardous waste-related programs
and policies as well as specific proposals.
Although in theory public hearings are scheduled just before de-
cision-making, in reality they usually take place after major con-
clusions have already been formed. All too often, the burden is .on
the public to prove that a different conclusion is warranted. This
is not an easy task for citizens. The time frame for citizens to
respond to the public record is much shorter than the time taken
by consultants and agencies in preparing the programs and pro-
jects under review. In an adjudicatory hearing for example, cit
izens must present expert witnesses to testify "points-of-fact" re-
buttals for issues which are predetermined by an administrative
law judge. Further, decisions which are made under the adjudica-
tory process may be binding.
Public hearings have been criticized not only because of the
heavy burden of proof on the public to change agency course,
but also that they occur too late in the planning process. Officials
responsible for public hearings should consider holding a hearing
earlier in the decision-making process than is required by regula-
tion. For example, the New York State Department of Environ-
mental Conservation held voluntarily a public meeting in the Town
of Ft. Edward to review with local citizens and hear comments
on the General Electric Company's engineering plan for remedial
actions at the Fort Miller PCB dumpsite.3
Level 6—Environmental Mediation
The highest level of public participation is environmental med-
iation. Mediation is a process in which conflict is legitimized.
It provides a mechanism for the formal settlement of conflict.
However, mediation is entirely a voluntary process. It can occur
only when all parties have a genuine desire to find a solution to a
problem. The several parties cannot be forced into arbitration, nor
can they be forced into agreement. The mediator has no authority
to impose a settlement; settlement occurs only when all parties
agree. Nevertheless, if the public participation programs previous-
ly described have set the stage for a genuine desire to find accep-
table remedies, mediation can be effective.
However, conventional methods of conflict resolution are the
antithesis of mediation. In court and at hearings, parties are repre-
sented by attorneys, whose job it is to be advocates and who are
trained to be adversaries. It is extremely difficult to "bring sides to-
gether after the field of combat has been joined. The entire public
participation effort should be directed toward setting an environ-
ment for mediation, if this level of conflict resolution is ultimately
needed.
Gerald Cormick of the University of Washington's office of En-
vironmental Mediation has made some astute observations about
-------
PUBLIC PARTICIPATION
349
the mediation process, and these should be incorporated into any
public participation strategy.4 Applying these to hazardous waste
situations, the following principles should apply:
•Mediation requires some relative balance of power. Typically,
the strength lies with government, as it controls the funds, the
permitting authority and the technical expertise. The real power
•that the public holds is in numerous voices, the press and the ears
of politicians. Disaster makes news, and the public participation
program should strive to assure that the public does not have to
emphasize the potentiality of disaster. This can be done by sup-
porting their side with access to technical data, moral support,
and perhaps even financial support or other tangible assistance.
•Mediation must result in compromises. Each side must establish
priorities, and be prepared to surrender some issues to gain others.
The problem is that it is legally or politically unwise to com-
promise standards. At present the nation has adopted water qual-
ity standards which have large margins of safety built into them
to compensate for poor public health impact data. Nevertheless,
the public perceives these standards to be minimal, and taxpay-
ers regard them as being completely accurate when arguing dam-
age claims. Thus, standards are like a rachet, they are easy to slip
in, but impossible to slide out.
If chemical concentration standards cannot be compromised,
what is left? Frequently the impact is excessive cost or social dis-
location.
•Environmental mediation need not wait, as in labor negotiations,
until an impasse has been reached. Environmental issues frequent-
ly have a wide range of alternatives, and the doors to all viable al-
ternatives should be kept open until there is agreement on a plan
of action.
The environmental mediation process should begin, or at least
be part of the planning, as soon as public perception of a poten-
tially dangerous situation becomes significant enough to warrant
attention. Reflecting on the landfills on Staten Island, it appears
that the public was aware of alleged hazardous waste dumping
long before it was recognized or confronted by regulatory agen-
cies. Thus the public was already developing a position and, there-
fore, they should be part of any technical solution that is devel-
oped.
•A third party, impartial and acceptable to both sides, is essen-
tial in the mediation process. Neither side may perceive the party
as an advocate for either side.
Finding such a party for a hazardous waste negotiation is not
easy. The mediator must be more than just an interested party, he
must also be technically competent. This means a good working
knowledge of chemistry, health effects, engineering, and a com-
plex legal/regulatory system which few fully understand. On top
of this, they need mediation skills, and the ability to inspire con-
fidence.
"Shuttle diplomacy" is a term which has become popular in re-
cent years. The Near East has the same volatility as toxic organ-
ics. For every hour that the mediator spends in joint meeting, he
may spend ten hours in private conferences with each side, es-
pecially if the issues are very polarized. Someone has to pay for
the mediator's services. This money should be programmed into
the public participation program immediately in inflammatory
issues, and set aside as contingency funds in less controversial
matters.
SUMMARY AND CONCLUSIONS
In all of the various mechanisms for public involvement, the
public is an important part of the package on hazardous waste
issues. Although the concept of working with the public is abhor-
rent to the technological professional, it is a fact of life. Per-
haps many of the problems of society are due to technologists
neglecting their social responsibility.
There are no clearly defined black and white solutions to haz-
ardous waste issues. The dangers are frequently poorly defined.
People look at the risks as an isolated event, instead of in the con-
text of the general risk of living. They react from fear of the un-
known. They crave professional help, and find it lacking and un-
caring.
No two hazardous waste situations are alike and therefore the
public participation program required will be different and tailored
to the specific problem. A good public involvement program must
be incorporated into every hazardous waste issue involving land
usage. Failure to do so can be dangerous to the social health of
the community and the viability of the project and its sponsors.
REFERENCES
i
1. Levine, A.C. Love Canal: Science, Politics and People. Lexington
Books, Lexington, Ma., 1982.
2. Ibid.
3. New York State Department of Environmental Conservation. Haz-
ardous Waste Disposal Sites in New York State—1st Annual Report,
1980.
4. Cormick, G., "Environmental Mediation." Presented at the 1980 An-
nual Conference, National Association of Environmental Professionals,
Washington, D.C., Apr. 1980.
-------
CITIZEN PARTICIPATION IN THE SUPERFUND PROGRAM
CATHERINE NEUMANN
BEN DRAKE, Ph.D.
Environmental Defense Fund, Toxic Chemicals Program
Washington, D.C.
INTRODUCTION
Since the passage by Congress of the Comprehensive Environ-
mental Response, Compensation, and Liability Act (CERCLA),
better known as "Superfund," citizen participation in the pro-
gram has been a much-studied issue.1 This paper will discuss the
shifting in USEPA policy from "citizen participation" to "public
relations" and compare current USEPA policy on community re-
lations to guidelines for public participation developed by con-
cerned citizens living near Superfund dumpsites.
Unlike other programs administered by the USEPA, Super-
fund is not directly mandated by Congress to provide for public
participation. However, USEPA recognized at an early date the
need to develop a policy toward public participation for both legal
and public policy reasons. Public participation is, in the Agency's
opinion, legally mandated by other legislation, such as NEPA.2
Moreover, events at Love Canal, Stringfellow Acid Pits and other
sites had shown that the success of the Superfund program was
dependent on effective citizen involvement.3
DEVELOPMENT OF AGENCY POLICY
Development of the Agency's community relations policy began
in Apr. of 1980, well before the existence of Superfund, when
the Office of Analysis and Program Development (OAPD) out-
lined what it termed the policy's theoretical parameters and goals.
The parameters originally suggested by OAPD included local media
and press relations, public participation, local government and
local interest group relations, and public information and educa-
tion. The major goal of the community relations program was to
ensure adequate communication between government and local
communities in order to facilitate the implementation of cost-
effective solutions to hazardous waste problems. Four secondary
goals were advanced in order to achieve this primary goal. They
were:
•To ensure the local community a meaningful voice in those im-
plementation decisions that the community considers most impor-
tant
•To establish a program of public information and media rela-
tions appropriate to the degree of interest and concern about
the site
•To anticipate potential conflict and attempt to avoid conflict
whenever possible
•To establish standard operating procedures that ensure exten-
sive interaction between local government officials and federal/
state officials'
The stated policy explicitly emphasized the importance of citizen
participation in decision-making in furthering the Agency's pri-
mary goal.
In the summer of 1980, the community relations policy was be-
ing tesied and expanded by a study of community involvement at
21 hazardous waste emergency operations and Clean Water Act 311
emergency actions.' An interim guidance document which reflected
this analysis was distributed to the Regional Administrators on
Feb. 25, 1981. The document required regions to adhere to the
following principles:
•To empathize with local concerns, learning about the local com-
munity
•To avoid the generation of unrealistic expectations
•To be open and forthright with information
•To anticipate the formation of ad hoc citizen groups
•To coordinate actions with local officials
•To assign community relation coordinators whenever possible
•To use a variety of participatory techniques
•To consider the establishment of citizen advisory committees at
sites and spills having a high degree of citizen concern
•To provide adequate training for Regional staff6
Two principles, (7th and 8th), take public involvement into ac-
count by asking Regions to consider techniques which enable cit-
izens to participate. This guidance suggests citizen advisory com-
mittees may be established and recognizes the importance of pro-
viding citizens with timely information. These principles require
that regional offices give citizens straight-forward information—
though it is not specified whether the information be given prior
to decisions. Thus, citizen participation is still viewed as an impor-
tant component of the community relations program.
In the summer of 1981, the National Contingency Plan (NCP),
the cornerstone of the Superfund Program, was drafted. This ver-
sion of the NCP, unlike the final, provided guidance on citizen
participation. This guidance was contained in an annex entitled
"Community Relations," which outlined the necessary compon-
ents of community relations plans (CRPs) at each phase of the
Superfund process.
The contents of the CRPs were to be determined by two fac-
tors: the degree of technical complexity at. the site and the level of
citizen concern. While the specific plans would vary from site to
site, all plans were required to include mechanisms which ensured
some elements of community participation. The Regions had to
reflect the emphasis local communities place on remedy selection,
and a public hearing at this stage was required. While the Annex
structure may not have been elaborate, it provided some handle for
citizens to ensure that their concerns would have to be identified
and addressed by the Agency.
In July-Aug. of 1981, a draft handbook entitled Community
Relations in Superfund—A Handbook was sent out to the Regional
Administrators, with a disclaimer that it need not represent official
EPA policy. In the introduction, this handbook states, "The com-
munity relations program should provide the community with a
means of influencing response actions but should not be used to
alarm a community."7 This handbook further developed the idea
of basing CRPs on the level of citizen concern and degree of tech-
nical complexity at the site by providing a matrix which would be
used to determine the appropriate contents of different CRPs. The
handbook also provided information on various information dis-
semination and gathering techniques and discussed participatory
techniques such as public consultations, hearings, and meetings.
However, there were no requirements in subsequent guidance doc-
uments on the extent to which this handbook should be used.
Therefore, the significance of the document is questionable.
350
-------
PUBLIC PARTICIPATION
351
THE SHIFT IN AGENCY POLICY
On Nov. 18, 1981, William N. Hedeman, Jr., Director of the
Office of Emergency and Remedial Response, sent a memorandum
to Regional Administrators entitled, "Superfund Community Re-
lations Police and Guidance." This memorandum outlines the ob-
jectives of community relations plans—objectives which are quite
different from those of the Interim Guidance document issued in
Feb. 1981. The objectives are, in their entirety:
•To establish at each site some means of learning the commun-
ity's concerns
•To let the affected public know Superfund actions are limited
by budget constraints, so that unrealistic expectations do not
develop
•To deal constructively with public response to Superfund ac-
tions—not to provide undue concern through unnecessary hear-
ings or inflammatory publications
•To contribute to the Superfund program's overall success by de-
creasing the likelihood of costly delays, cost overruns, and polit-
icization of purely technical issues
•To establish a preventive program—to lessen or avoid public
confusion about Superfund remedies
•To stress the interaction of federal, state, and local government
in solving local problems'
These\>bjectives totally remove public participation from the ob-
jectives of the community relations programs. The component
of public participation has changed to a public relations objec-
tive designed to diffuse, rather than address, public concerns.
The language of this memorandum is instructive in its overridingly
negative attitude: concern is expressed about "unrealistic expec-
tations," "undue concern," "unnecessary hearings," and the
avoidance of "costly delays," and "public confusion." The
agency asks each regional office to find its own way to "deal con-
structively" with what is clearly viewed as the potentially hos-
tile force of the citizens most intimately affected by its decisions:
citizens who would attempt "the politicization of purely technical
issues."
The limitations of the guidance document are not merely those
of tone and general purpose, however. The cover letter states,
"There is no set of 'requirements' or a required 'formula' to be
applied at each site." The adequacy of each CRP therefore de-
pends on the importance each Regional Administrator assigns to
public involvement in the Superfund program, and there is no as-
surance that there will be consistent and adequate public participa-
tion at each site.
The document recommends that, with adequate prior prepara-
tion, a public meeting be held at the point when the on-scene
coordinator is "ready to recommend a permanent remedy for the
site." This recommendation is clearly contrary to the idea that
the public should participate in the decisions affecting the site,
since the hearing will be held after the decision is already made.
At no other point is any technique recommended whose inclusion
would allow citizens to become involved in the decision-making
process.
A final and central deficiency with this guidance document,
though it has existed from the inception of the program, is that
there are no provisions whatsoever for citizen involvement during
responsible party cleanups or enforcement actions. This lack of
guidance further excludes citizens from involvement in the Super-
fund program, and in precisely those circumstances where, in all
probability, most remedial actions will take place. Such exclusions
have been encountered at sites such as Woburn, Ma., where cit-
izens were wholly removed from the decision-making process after
a responsible party was identified even though relatively good cit-
izen participation techniques had previously existed.
The National Contingency Plan appeared in final form on July
16, 1982. In this document, the Annex which contained specific
provisions for citizen participation has been completely deleted.
Only two brief sections of the NCP deal with the subject, or with
"community relations," as public participation is now labeled. The
first, Section 300.34(e), merely authorizes certain help to the on-
scene coordinators in public information and participation during
"major responses.'" The second, Section 300.61(c)(3), reads: "In
determining the need for or in planning or undertaking Fund-
financed action, response personnel should, to the extent practic-
able,....be sensitive to local community concerns (in accordance
with applicable guidance)"'° [emphasis added]. In effect, these
two statements comprise the whole of the Agency's new public
participation policy.
The phrase "to the extent practicable" makes it quite clear that
even "sensitivity to community concerns" and the use of "applic-
able guidance" about public relations are matters of choice. Even
more revealing, the term "applicable guidance" has no defined
meaning and no legal or regulatory status. Thus the guidance may
be defined by whatever policies, written or oral, the Agency
chooses to call "applicable guidance" at any given time. This ver-
sion of the NCP shows that the Agency has abandoned the earlier
goal "to ensure the local community a meaningful voice in...im-
plementation decisions," which had appeared in the initial OAPD
statement. The plain truth is that neither the NCP nor the Novem-
ber policy document obligates the Agency to anything in the way of
citizen participation.
Currently, the term "applicable guidance" refers to the Hede-
man memo of Nov. 18, 1981." Since the Hedeman guidance doc-
ument directs Regions to manage community relations programs
on a site-by-site basis, we turned to the Community Relations
Plans (CRPs) actually developed in the field to gauge Agency pol-
icy. These give some documentary evidence, though it is virtually
impossible to determine the extent to which the plans are actually
being carried out at the specific sites, since the Agency has no in-
ternal checking system.
ANALYSIS OF COMMUNITY RELATIONS PLANS
EDF examined 23 of the 25 CRPs submitted and approved at
Headquarters. (Appendix 1.) The stated objectives of each plan
differed greatly, varying from merely informing local officials to
stating that citizens would be ensured a meaningful voice in the de-
cision-making process.
Despite explicit requirements for submitting community relations
plans,12 EDF found a number of sites where community relations
plans have not been developed in a timely manner. For example, a
community relations plan is just now in the draft stage for Bruin
Lagoon, Bruin, Pa., although there has already been a feasibility
study and a remedial option selected at this site. The ad hoc pro-
cedures which were followed without such a plan included, as we
have been informed by local residents, a public hearing announced
in a newspaper from a city over 30 miles away and a period for
review and comment of the feasability study by the public of only
9 days. A detailed screening for those techniques which are com-
patible with an active rather than a purely passive role for citizens
showed:
•Out of the 23 plans reviewed only one made specific provisions
for notification/comment periods. (Woburn, Ma.)
•There were only three sites where a citizen's advisory committee
or task force was established by the State. (Woburn, Ma.; Tar
Creek, Ok.; Luminous Processes Site, Ga.)
•Only ten sites had any provisions for public input before a final
remedial option was selected through public consultations, meet-
ings or hearings. (Woburn, Ma.; Olean Fields, N.Y.; Luminous
Processes site, Ga.; Gratiot County Landfill, Mi.; Colbert Land-
fill, Wa., Aidex, la.; Chem-Dyne, Oh.; Tar Creek, Ok.; French
Limited site, Tx.; Bio-Ecology, Tx.)
•Three other sites stated that a public meeting would be optional
before the selection of a remedial option. (Motco, Tx.; Arkan-
sas City, Ak.; Summit National Services, Oh.)
•At three other sites where plans include public meetings, it was
impossible to determine at wnat point in the process they would
be held. (Winthrop, Me.; Keefe Environmental Services, N.H.;
Gilson Road site, N.H.)
•Only four plans made any provisions for documenting citizens
-------
352
PUBLIC PARTICIPATION
concerns and the Agency response to these concerns. (Woburn,
Ma.; Winthrop, Me.; Tar Creek, Ok.; Colbert Landfill, Wa.)
The picture presented by these plans is that USEPA's attitude
toward public participation in each Region is highly variable. In
general, the plans contain a minimum of specific techniques which
invite any actual participation. The CRPs at two sites stated that a
public meeting would be held only after the selection of a remed-
ial option. (Ellisville, Mi.; Burnt Fly Bog, N.J.) The plans as a
whole indicate that without an explicit structure for citizen involve-
ment, with a set of minimum guidelines to be adhered to by each
Region, citizens will continue to be excluded from participating in
the decisions of the Superfund program. Jt was this that inspired
one group of citizens active at Superfund sites to develop its own
guidelines for citizen participation.
ORIGIN OF THE CITIZEN STATEMENT
On June 4-7, 1982, Environmental Defense Fund (EOF) held a
national conference entitled, "Superfund: Promise and Reality,"
attended by 31 citizen representatives from 22 Superfund sites
across the country. The participants met with USEPA's Admin-
istrator Anne Gorsuch to discuss the situation at their respective
sites. The major concern aired by the participants was their lack of
involvement in the decisions being made about their sites. They
noted, among other things, that technical reports were unattain-
able, that public hearings were either announced in cities far from
their sites or were not timely, and that there was little real com-
munication between the Regional Offices and local citizens.
Furthermore, the citizens noted that they were not even aware the,
Agency had a policy on community relations or that specific com-
munity relations plans had been developed for their sites. After
hearing these concerns, Mrs. Gorsuch invited the group to develop
its own set of guidelines for public participation in the Superfund
program. In a letter sent on June 7th, the participants formally
accepted her invitation and outlined nine basic principles of effec-
tive citizen participation.
In the following three months, the citizens continued work on
the document in small regional meetings, by phone call, and by let-
ter. On Sept. 7, 1982, their "minimum guidelines for citizen par-
ticipation," were sent to Administrator Gorsuch, where they now
await disposition.
SPECIFIC GUIDELINES FOR PUBLIC PARTICIPATION
In this section, the citizen's nine principles of public participa-
tion are presented. They are the best available index of what cit-
izens living near Superfund sites wish to see—in order better to
identify the deficiencies in the Agency's existing policy on com-
munity relations. Under each principle, the minimum require-
ments specified by the citizen's needed to ensure effective involve-
ment are identified. Following this, the pertinent provisions under
the Agency's existing community relations policy are listed."
Principle 1: Participation from beginning to end of the decision-
making process.
•Minimum Requirements—The citizens want on-going educa-
tional workshops, public meetings, and public hearings to be held
upon their reasonable request. This includes the liberal use of
comment periods on important materials.
•Agency's Existing Requirements—The only provision for public
input in existing USEPA guidance is the discretionary public
hearing to be held before a remedial option is selected.
Principle 2: Timely notification of all steps in the Superfund
process before they are taken, with reasonable time allowed for
comment and other response.
•Minimum Requirements—Citizens want to be notified at least 30
days in advance of all significant actions. Such actions may in-
clude the completion of reports, scheduled workshops, meetings,
hearings, and pending Administrative Orders or voluntary settle-
ments.
•Agency's Existing Requirements—There are no requirements for
notification and comment periods in the existing guidance docu-
ments.
Principle 3: Explicit structure for public participation in devel-
opment of the CRP.
•Minimum Requirements—Citizens want to be consulted when
site-specific community involvement plans are being developed.
Suggested techniques are, for instance, that the Agency hold
workshops to discuss the plan prior to its development, or that
citizens may request a public meeting to discuss the plan before
it becomes final.
•Agency's Existing Requirements—Existing guidance does not
provide any structure for public participation since the only re-
quirement is that a plan be developed. Furthermore, commun-
ity 'relations plans are being developed without citizen input in
many cases. Citizen interviews are suggested by the Agency to
determine the public's concerns, but there are no requirements
that the public be involved in the actual development of the plan.
Principle 4: Opportunity to participate in each decision, includ-
ing settlements with responsible parties, before finalization.
•Minimum Requirements—Citizens want to be involved during
responsible party cleanups. This involvement may include citizen
observers at negotiation sessions, citizen access to data and other
information exchange between the agency and a responsible party,
and citizen comment on Administrative Orders or voluntary
agreements before they become operative.
•Agency's Existing Requirements—Neither the NCP nor existing
guidance document address the issue of citizen involvement dur-
ing a responsible party cleanup.
Principle 5: Early, immediate, and complete access to all in-
formation.
•Minimum Requirements—Citizens want the Agency to set up a
local depository at each site where at least one copy of all past
and present reports will be placed.
•Agency's Existing Requirements—Existing guidance does not dis-
cuss whether any reports or information about the site should
be publicly available. The draft handbook provides assistance on
information dissemination techniques but does not discuss what
material should be disseminated. Furthermore, there are no re-
quirements that a local depository be established at each site.
Principle 6: Access to appropriate technical expertise through
government or other sources in advance of government action.
•Minimum Requirements—Citizens want access to State, Agency,
and responsible party experts through telephone and/or workshop
contact. Also, when technical issues are in dispute, citizens want
the ability to hire independent expertise and be reimbursed by
the Agency.
•Agency's Existing Requirements—The role of the Agency's tech-
nical consultants is not discussed in existing guidance on com-
munity relations beyond that of assisting the Agency during a
public meeting. The extent to which a technical consultant could
participate in citizens meetings on technical issues is unknown.
Principle 7: Convenient, timely public hearings held at loca-
tions reasonably accessible to concerned citizens prior to any de-
cisions.
•Minimum Requirements—Citizens want public hearings normal-
ly to be conducted only after there has been a thorough back-
ground educational effort via small workshops, and informal
meetings. These hearings should be held throughout the Super-
fund process in response to reasonable citizen request. As a min-
imum, there must always be a hearing held before a remedial op-
tion is selected.
•Agency's Existing Requirements—Existing guidance suggests that
a public hearing be held before the selection of a remedial op-
tion. This is the only discussion of the time at which a hearing
-------
PUBLIC PARTICIPATION
353
should be held.
Principle 8: Provide citizens with progress reports on site activ-
ity. Make provisions for citizen comment on the reports.
•Minimum Requirements—Citizens want monthly reports which
summarize past work and details of upcoming work at the dump-
site.
•Agency's Existing Requirements—There is no mention of
progress reports in any of the existing guidance or policy docu-
ments, let alone the ability for citizen comment.
Principle 9: Provide complete listing of government contact
persons (with telephone numbers).
•Minimum Requirements—Citizens want the on-scene coordina-
tor, public affairs office personnel, consultants, responsible par-
ties and project officers at Headquarters to be identified along
with a telephone number where they can be reached. The citizens
also want a toll-free Regional hotline to be established.
•Agency's Existing Requirements—Existing guidance does not re-
quire that a contact list be developed or made publicly available.
There are no toll-free Regional hotlines where citizens can call
to voice their concerns or complaints.
CONCLUSION
In this paper, the authors have documented an explicit change
in the nature of USEPA's community relations program from a
program whose stated objectives were to involve the public in the
decisions of the Superfund process to one whose objectives are to
diffuse public reaction to that process. Guidelines and criteria for
public participation developed by citizens living near Superfund
dumpsites have been presented.
The differences between the citizen guidelines and the Agency's
existing community relations policy clearly show that the policy
does not meet, and does not promise to meet, the needs of citi-
zens liying near Superfund dumpsites as they perceive them. The
central guidance document of the Agency states: "All available
evidence indicates that the success of the Superfund program de-
pends in large measure on the Agency's ability to implement an
effective community relations program."14 It is difficult to con-
ceive of an effective program which does not meet in some sig-
nificant measure the perceived needs of the population toward
whom it is principally aimed.
REFERENCES
1. The extensive work undertaken by the Agency is examined and in
part referenced by Cohen, S., Ingersall, T., and Janis R. "Insti-
tutional Learning In A Bureaucracy: The Superfund Community
Relations Program," Proc. of the National Conference on Man-
agement of Uncontrolled Hazardous Waste Sites, Oct. 1981, Haz-
ardous Materials Control Research Institute, Silver Spring, Mary-
land, 405-410. In the same volume appear: Goggen, B., Rappa-
port, A. "The Community Hazardous Waste Coordinator Pro-
gram," pp. 411-414, and Shaw, L. Milbrath, L. "Citizen/Govern-
ment Interaction At Toxic Waste Sites: Lessons From Love Canal,"
415-420.
2. Internal memoranda indicate that the Agency recognized the legal
reasons for requiring public participation in the Superfund program.
An August 17, 1981 memorandum from William N. Hedeman, Jr.
entitled, "Guidance on Superfund/NEPA Policy: Areas of Respon-
sibility," equates the studies and reports which are generated for
clean-up in the Superfund program to an environmental impact
statement. The memo then states that the public participation re-
quirements under NEPA should be carried out in accordance with
existing guidance on community relations in the Superfund program.
Likewise, a September 1, 1982 memorandum from the Office of
General Counsel to Administrator Gorsuch and Regional Adminis-
trators ("Public Participation in Remedial Actions Under the Com-
prehensive Environmental Response, Compensation, and Liability
Act of 1980 (CERCLA) states, "...in order for the functional equiv-
alent exception to apply, it will be necessary for remedial actions
to incorporate comment on environmental issues before the final
selection of a remedial alternative. One such standard is public par-
ticipation in decisions regarding remedial action." Furthermore,
Headquarters realized that RCRA requirements for public participa-
tion may be invoked when hazardous waste is generated, treated,
stored or disposed as a consequence of the remedial option se-
lected.
3. ICF, Incorporated, Analysis of Community Involvement In Haz-
ardous Waste Problems, A Report to the Office of Emergency and
Remedial Response, U.S. EPA, July 1981.
4. Cohen, S., Ingersall, T., and Janis R. "Institutional Learning In A
Bureaucracy: The Superfund Community Relations Program," Proc.
of the National Conference on Management of Uncontrolled Haz-
ardous Waste Sites, Oct. 1981, Hazardous Materials Control Re-
search Institute, Silver Spring, Maryland. 405-410.
5. ICF, op. cit.
6. Cook, M., Deputy Assistant Administrator, Office of Hazardous
Emergency Response, "Interim Community Relations Guidance for
Site Clean Up." EPA Memorandum to Regional Administrators,
Feb. 25, 1981.
7. Community Relations In Superfund—A Handbook—Draft USEPA,
Sept. 1981. p. 1.
8. Hedeman, W. Jr., Director, Office of Emergency and Remedial
Response, "Superfund Community Relations Policy and Guidance,"
USEPA Memorandum to Regional Administrators, Nov. 18, 1981.
9. Federal Register, 47, No. 137, July 16, 1982, 31210: "The USCG
Public Information Assist Team (PIAT) and the EPA Public Af-
fairs Assist Team (PAAT) may help OSCs and regional or district
offices meet the demands for public information and participation
during major responses. Request for these teams may be made
through the NRC."
10. Op. cit., 31214. The responsiveness of the Agency itself to public
comments does not sort well with the final version of the NCP.
In its discussion of the comments on public participation the Agency
states, "Several commentators questioned the adequacy of [this sec-
tion] and pointed out that it is important to keep the public in-
formed and to include them in the decision-making process. Spe-
cific comments included...strong advocacy of greater emphasis on
public participation....In order to indicate that the Agency has issued
guidance in this area....that it is necessary to be sensitive to local
concerns 'in accordance with applicable guidance' " (Federal Regis-
ter, 47, No. 127, July 16, 1982, 31198). As we have seen, the guid-
ance can not be thought to be responsive to a "strong advocacy of
greater emphasis on public participation" nor of including the pub-
lic in "the decision-making process." The addition of the stringent-
ly qualifying phrase that sensitivity to community concerns be "to
the extent practicable" is also passed over in silence in the Agency's
explanation.
11. Personal communication with William N. Hedeman, Jr., Director,
Office of Emergency and Remedial Response, Apr. 30, 1982.
12. Hedeman, W. Jr., Director, Office of Emergency and Remedial
Response, "Superfund Community Relations Policy and Pro-
cedures," USEPA memorandum to Regional Administrators, March
31, 1982.
13. The complete set of guidelines can be obtained by writing to: Toxics
Chemical Program, Environmental Defense Fund, 1525 18th Street,
N.W., Washington, D.C. 20036.
14. Hedeman, November 18, 1981, op. cit., p. 1.
Appendix 1: Community Relations Plans Included In Analysis
Region 1 Winthrop Town Landfill, ME
Keefe Environmental Services, NH
Gilson Road Site, NH
Mark Phillips Trust, MA
Region 2 Olean Well Fields, NY
Burnt Fly Bog, NJ
Region 3 McAdoo Associates, PA
Region 4 PCB Spill, NC
Bluff Road Site, SC
Luminous Processes, GA
Whitehouse Waste Oil Pits, FL
Region 5 Summit National Services, OH
Chem-Dyne Site, OH
Fields Brook, OH
Gratiot County Landfill, MN
Region 6 Motco, TX
French Ltd.
Disposal Site, TX
Bio-Ecology Systeir %
Inc.,TX
Region 7 Tar Creek, OK
Aidex Corporation, IA
Arkansas City
Dumpsite, KS
Ellisville Site, MI
Region 8
Region 9
Region 10 Colbert Landfill, WA
-------
PROGRESS IN MEETING THE OBJECTIVES OF THE
SUPERFUND COMMUNITY RELATIONS PROGRAM
BARRY H. JORDAN
ANTHONY M. DIECIDUE
U.S. Environmental Proltection Agency
Washington, D.C.
JAMES R. JANIS
ICF Incorporated
Washington, D.C.
INTRODUCTION
The USEPA's Superfund program is an ambitious attempt on
the part of the federal government to help solve the problems
created by uncontrolled hazardous waste disposal sites in the
United States, as well as to respond to the thousands of releases of
hazardous substances into the environment each year. These prob-
lems have grown steadily throughout this century with increases
in the production, distribution, storage, and use of hazardous
chemicals.
It is difficult to estimate the size of the hazardous waste prob-
lem, but the Agency has been notified of a significant number
of sites which are at least potential problems. While the Super-
fund remedial program will not address most of these sites, those
that present the greatest threat to public health and welfare will re-
ceive some kind of response action. In addition, Superfund will
provide the resources needed to respond promptly and effectively
to emergencies caused by releases of hazardous substances during
their transportation, storage, or disposal.
Early in its administration of the Superfund program, USEPA
realized that special attention would need to be given to public
concerns and fears in communities located near hazardous waste
sites or releases that have been targeted for response. According-
ly, USEPA has established a community relations program de-
signed to provide communities with accurate information about
problems posed by releases of hazardous substances and at the
same time, to seek local officials' and citizens' input into the de-
cision-making process used to select a technical solution to site
problems. The Superfund community relations program, in short,
is an information and communications program aimed at meeting
the special needs created by hazardous substance problems.
USEPA insists that an effective community relations program must
be an integral part of every Superfund-financed action.
PROGRAM DEFINITION
Why a Community Relations Program is Important
A release of hazardous substances poses both human and tech-
nical problems. While the issue to the engineer may be one of the
technical feasibility of a proposed response, to concerned citizens
living near the site the issue is their health and the health of their
children. Citizens may also see the value of their property as well as
the threatened image of their community. Consequently, any haz-
ardous substance problem is inherently sensitive. It must be man-
aged with attention given to public attitudes and fears. If a pro-
posed response to a release of hazardous substances is to be suc-
cessful or even, sometimes, to be completed at all, the local com-
munity musl have an accurate understanding of the nature of the
threat and of the various alternates that may be presented for
dealing with it.
Local opposition to government actions may be motivated by
fear of the effects the hazardous substance problem will have on
the qualiu of life in the community and by distrust of industry
and government. Often, the local media and local public officials
will state these same fears and concerns. All too frequently, the
result is public excoriation of the company(ies) involved, rejec-
tion of reasonable response plans (or unrealistic expectations from
them), obstructionist tactics, and even the forced closure of exist-
ing sites.
The level of sensitivity and public interest need not be related
to the technical complexity of the problem. In addition, it can be
intensified by limitations in the resources available for a single
Superfund response. People who have lived with a hazardous waste
problem for years are likely to want—and expect—the federal and
state governments to remedy the problem immediately. Removal
actions or initial remedial measures will sometimes result only in
interim, on-site solutions. Furthermore, cost-effectiveness stipula-
tions in the Superfund statute, as well as National Contingency
Plan, requirements for detailed analysis of remedial alternatives,
must be satisfied before off-site remedies can occur. An effective
community relations program will help explain these kinds of con-
siderations to the concerned local public, and will help the citizens
understand the technical and financial limitations of the national
response program.
Objectives
Specifically, then, the objectives of the Superfund community
relations program are as follows:
•To collect information about the concerns of the community.
A community relations program provides a vehicle of exchange
among USEPA, the state, the public, and local government. It
enables USEPA and state staff to identify citizen leaders, public
concerns, and relevant social and political considerations. Some-
times it can also yield technical information useful in planning a
solution to a site's problems.
•To inform the public of planned or ongoing actions.
The program should inform the public of the nature of the en-
vironmental problem, the remedies under consideration, and the
progress being made.
•To give citizens the opportunity to be involved in decision-
making.
The program should enable citizens to express opinions about
decisions that will have long-term effects on their community.
A course of action stands a greater chance of public acceptance
if citizens have had a voice in its planning.
•To focus and define issues and to help resolve conflict.
Conflict may be unavoidable in some circumstances, but it can
be constructive if it brings into the open alternative viewpoints
based upon solid reasons for criticism or dissent. A community
relations program channels conflict into a forum -where it can
serve a useful purpose.
PROGRAM REQUIREMENTS
There are three kinds of Superfund response actions:
•Immediate removals are undertaken to prevent or mitigate immed-
iate risk of harm to human life or health or to the environment.
354
-------
PUBLIC PARTICIPATION
355
Immediate removals may last for as little as a few days.
•Planned removals are either a continuation of an immediate re-
moval or a separate cleanup action at an unranked site, limited
in time and cost.
•Remedial actions are those responses to releases at national prior-
ity sites that require longer term and possibly more expensive
cleanup efforts.
Because every hazardous waste site is likely to have unique cir-
cumstances and issues of local concern, requirements for designing
and implementing the community relations program have been
kept minimal and flexible. The program relies chiefly on the ability
and perceptions of regional and state staff. The key requirement
is the submission, by the USEPA Regional Office or the state, of a
"community relations plan" for each remedial action and for each
removal action which last more than a few days. The community
relations plan must carefully consider the specific needs of an in-
dividual site and must ensure that community relations activities
are closely coordinated with technical work being contemplated or
performed.
For example, while a particular site may be the focus of a
high degree of vocalized public concern, the health or environ-
mental threat it poses may not be so acute as to demand an immed-
iate removal. Instead, there may be sufficient time to plan and
implement a more lengthy remedial action. Clearly, such a situation
would require early extensive community relations activities preced-
ing any selection of alternatives in order to explain the more time
consuming, albeit prudent, approach.
All community relations plans should be based on on-site dis-
cussions with local officials, community leaders, and involved cit-
izens to identify local concerns or sensitive issues. USEPA should
be aware of any lengthy history of problems at the site that may
provoke public concern or influence local attitudes. For example,
in developing a community relations plan for a site during the past
year, USEPA found that the site owner had threatened nearby resi-
dents and made them reluctant to seek help from the government.
The importance of the on-site visit cannot be stressed too strong-
ly. From the beginning of the Superfund program, USEPA has
recognized that a local community's concerns cannot be adequate-
ly measured from a distance. All of the best examples of USEPA's
community relations plans are based on an accurate understand-
ing of community concerns gained through personal visits by
USEPA staff and its contractors.
When the USEPA decides to fund a Superfund response, the
Agency develops a cooperative agreement or a contract with the
state; this agreement establishes responsibility for information and
communications activities at the site. The responsible agency then
develops a community relations plan that includes four key points:
•How citizen concerns will be identified at the site
•How accurate information on problems associated with the re-
lease of hazardous substances will be distributed and explained
to the community
•How citizens will have an opportunity to comment on ongoing
and proposed site work
•How the proposed technical solution will be explained to the
community
Specific methods listed in the plan for soliciting citizen input
and distributing information will vary from site to site, depend-
ing upon the level of citizen concern and the nature of the site's
technical problems. However, smaller scale, informal methods,
such as living room gatherings with citizens and local officials,
are generally emphasized. The responsible agency will then imple-
ment the plan in close coordination with other interested agencies.
CURRENT STATUS OF THE PROGRAM
In the past year, there has been some progress in all areas of
the community relations program. Some major accomplishments
are represented by the following:
•More than 40 community relations-related site visits have occur-
red during the preparation of community relations plans and re-
lated analyses since the program's inception
•A policy for community relations in Superfund has been com-
pleted by Headquarters and communicated to all the Regions
•A detailed handbook on community relations in Superfund has
been published; the handbook lays out specific options which
might be followed in carrying out community relations programs
•Twenty-five community relations plans have been reviewed at
Headquarters and approved
These successful developments mainly represent continued pro-
gress in publicizing the program, refining program objectives,
and accumulating information. While all of these steps are neces-
sary to ensure a good program, adeuqate plans have not yet been
produced for all scheduled remedial actions.
Of the accomplishments listed, the site visits represent the most
visible indication to the public that USEPA is giving concerted at-
tention to community concerns. The visits show concrete move-
ment toward a solution of the locality's problem. Overall, such
visits enhance the public image of USEPA's commitment to the
Superfund program.
In addition, site visits provide significant educational benefits to
staff who will conduct response activities. Interviews with citizens
and local officials compel Regional Office and state staff to think
carefully about community concerns and needs—preventing
blunders with sensitive local issues.
Site visits have been useful for identifying or resolving policy
issues. For example, the visit to Commencement Bay, Washing-
ton, highlighted a major Superfund policy issue: how to treat haz-
ardous waste sites located on Native American lands. The problem
involves the responsible management of these lands as well as
their classification for the purposes of Superfund.
The site visits also confirmed the value of the decentralized,
flexible nature of the community relations program. Through site
visits, the USEPA or state officials conducting community rela-
tions activities at the local level have obtained an immediate aware-
ness of what methods are likely to be most useful. This local
orientation should increase in effectiveness as officials expand their
community relations experience.
Another significant step was the development of a policy and
guidance document. The policy establishes USEPA goals and pro-
cedures for the community relations program, and requires Reg-
ional Project Officers to carry out a program that is consistent
across the nation.
The program is supported by the community relations hand-
book. Having gained greater confidence as a result of field ex-
perience, USEPA now has a detailed, step-by-step guide to com-
munity relations activities and methods that can be followed in any
kind of response action. Specific techniques are presented, show-
ing instances when they might be most appropriate. For instance,
the timing of briefings, citizen group meetings, community infor-
mation interviews, and media appearances are discussed as they
relate to various problems and to the different stages of a partic-
ular cleanup activity. For example, community interest often peaks
at the stage of a remedial action in which cleanup alternatives are
developed and selected. Thus, a sensitive community relations
effort is especially important at this point. The handbook recom-
mends briefings for local officials and small, informal meetings as
among the appropriate activities.
USEPA has implemented several Superfund community rela-
tions programs, but there remain goals yet to be achieved. One
goal is a full briefing of Regional USEPA staff on the design
and implementation of community relations programs. At the same
time, Headquarters will conduct a detailed program review to de-
termine whether requirements are being met, and to determine
whether any unforeseen problems have been encountered at sites.
PROBLEMS AND IMPLEMENTATION
Apart from specific cases where insufficient attention has been
paid to community relations during a response, the Superfund
community relations program has faced several general problems
-------
356
PUBLIC PARTICIPATION
that arise from the structure of the program. Institutional bar-
riers present the program with difficult management and policy
problems. The fact that different levels of government have
different concerns, different pressures, and different constituencies
causes some slowdowns and inefficiencies even in the smoothest
of relationships.
A second problem which remains manageable, but is difficult
nonetheless, arises in cases of enforcement against responsible par-
ties. Public release of information can complicate negotiations
with responsible parties to finance cleanup operations. If USEPA
reveals too much information in response to community requests,
the responsible party may feel threatened and become intransigent,
closing off continued negotiation. USEPA and Justice Depart-
ment attorneys are justifiably reluctant to disclose anything about
negotiations. The result of secrecy, however, may be public sus-
picion and increased requests for information. In short, providing
citizens with information about legal proceedings (or even about a
situation subject to legal proceedings) can complicate those legal
actions. Yet citizens have a right to know what is going on. A bal-
ance needs to be maintained.
A further point where balance is necessary is in the amount of
influence citizens may have on the decision-making process. While
citizen input is necessary (indeed, that is the point of the com-
munity relations program), USEPA's response decisions must rest,
in the end, on the judgment of qualified experts in engineering,
public health, and environmental sciences. Furthermore, the prob-
lems addressed by Superfund are nationwide in scope: the pro-
gram's costs and benefits are distributed across the country. The
concerns of a locality, however, are usually more narrow. The
responsible agency must consider local concerns and recommenda-
tions and respond to them. The Agency does not believe, however,
that Congress intended for citizens and local officials to make final
decisions on response actions. USEPA must balance its require-
ment to respond to a local situation with its responsibility as a
national agency with national management requirements and de-
mands.
In early September of this year, the Agency received a set of
community relations guidelines prepared by citizens who live near
certain sites around the country. The guidelines contain specific
recommendations from these citizens on how to improve the com-
munity relations program. The Agency appreciates these comments
and is now reviewing the recommendations to see where they may
make useful additions to the current program.
Such "grass-roots" input is of significant value both in pro-
gram planning and insofar as it symbolizes the essence of the Super-
fund community relations program: that hazardous substance re-
leases cannot be adequately dealt with without a real cooperative
effort among all concerned. In the past, communications between
citizens and cleanup officials have at times been inadequate, caus-
ing difficulty for both groups and interfering with effective re-
sponse activities. Similar situations must be avoided in the future.
The progress of the Superfund community relations program in-
dicates that USEPA does understand the complexities of the
challenge and is working toward solutions.
-------
SPILL INCIDENTS AT HAZARDOUS MATERIAL STORAGE
FACILITIES: AN ANALYSIS OF HISTORICAL DATA
FROM THE PIRS AND SPCC DATA BASES
E. HILLENBRAND
B. BURGHER
JRB Associates
Waste Management Department
McLean, Virginia
INTRODUCTION
Spills of hazardous materials from storage tanks and associated
equipment may result in significant environmental damage through
the direct effects of the material on the environment and through
fires and explosions precipitated by the spill event. Through
analysis of a large number of reported spill events from storage
tanks and their associated equipment, it should be possible to deter-
mine:
•The relative probabilities of failure events attributed to different
causes or points of failure
•The relationship of spill size to cause or point of failure
As part of a current project being performed by JRB Associates
under contract to the USEPA, Office of Solid Waste (OSW), over
4,000 reported spill events from storage tanks and associated equip-
ment were analyzed for spill size and cause or point of failure. The
resulting data are to be used, in conjunction with other information
being developed under this project, by OSW in further refining
regulations for tank storage of hazardous waste.
DATA SOURCES
Two primary data sources, containing information on a suffi-
cient number of storage-related failure incidents to allow for a pro-
babilistic analysis, were identified. These data sources are the Spill
Prevention, Control, and Countermeasure (SPCC)' data base,
maintained by USEPA, and the Pollution Incident Reporting
System (PIRS)2 data base maintained by the U.S. Coast Guard.
These data bases contain information on all spills of oil and hazar-
dous substances reported to the Federal government under Section
311 of the Federal Water Pollution Control Act (FWPA)3 Spills to
inland waters are reported to EPA and are entered into the SPCC
data base. Spills to coastal waters and navigable inland waters are
reported to the Coast Guard and entered into the PIRS data base.
While both systems are similar, they do differ in the types of infor-
mation available, accessibility, and format.
Pollution Incident Reporting System
Spills that would affect coastal waters and navigable inland
waters (such as the Mississippi River and Great Lakes) are reported
to the Coast Guard under the FWPA and incorporated in the PIRS
system. Additionally, spills reported to USEPA that eventually re-
quire litigation are also incorporated in PIRS, since the Coast
Guard is responsible for determining extent of damage in such
cases. Spills that would be reported include those resulting from
transportation, loading, and unloading operations, production
processes, and material storage.
Computer files for approximately 2000 "Onshore Bulk Storage
Facility" spill cases reported to the Coast Guard for the year 1975
through 1980 were accessed and analyzed. Each case file included
the following information:
The format of the PIRS data base made it possible to access only
those spills which originated from storage operations (tanks and
ancillary pipes, pumps, and valves). The ability to distinguish be-
tween spills originating from pipes, pumps, and valves associated
with storage and those associated with a specific process or with
transportation is obviously important to an analysis of this type.
The data available from the PIRS data base have a number of
important limitations affecting the reliability of the data:
•Only spills which may affect surface water bodies are required to
be reported to the Coast Guard (or USEPA) under Section 311
of the FWPA.
•It is not possible to determine what percentage of spills are un-
reported, and whether percentage of. reporting is biased toward
any particular type or cause of failure. Additionally, it is likely
that small spills in particular are underreported, biasing results
toward large spills, since small spills are generally more likely to
be contained.
•Spills occurring within the confines of a secondary containment
structure are unlikely to be reported.
•The data lack any description of such important factors as tank
capacity or tank construction materials.
•Approximately one-third of the retrieved cases lacked a critical
piece of information (e.g., spill size) and had to be dropped from
the data pool.
•There is no means of identifying the size of the universe of po-
tential respondents.
•Size determination is based generally on the judgment of the
spill investigator.
Spill Prevention, Control, and Countermeasure
Data Bases
At the time of this analysis the SPCC data base contained bet-
ween 20 and 30% of all spills affecting inland waters reported to
EPA between 1975 and 1980. USEPA is entering the remaining in-
cidents for that period as resources permit; the order of entry is
random for the parameters of concern in this analysis, and
therefore, should not bias the results.
Incidents reported include transportation, production, and
storage related spills. Printouts were received from USEPA's
SPCC Office for all 10 Regions (data provided for Region 7 was in-
complete) containing those cases reported as onshore non-
transportation spills of petroleum products, hazardous chemicals,
and other substances. Spills occurring at industrial plants,
marketing distributors, refineries, power plants, hazardous waste
sites, and other storage facilities have been included Information
included in the data base includes:
•Date of spill
•Location
•Size of spill
•Substance spilled
•Cause
•Type of operation
•Facility name and locations
•Data and time of spill
•Water body affected
•Material spilled
Unfortunately, the design of the SPCC data base does not allow
for retrieval of information on spills from storage tanks and an-
•Source of spill
•Cause of spill
•Size of spill
357
-------
358
DATA BASE
ciliary equipment due to the categorization used. The category ac-
cessed (onshore non-transportation spills) encompasses spills from
process tanks, pumps, piping, valves, etc., as well as for storage
tanks and ancillary equipment. This is indeed a limitation of these
data in view of the purpose of this analysis.
In addition to this limitation, the SPCC data base has similar
limitations to those of the PIRS data base.
METHODOLOGY
Both the PIRS and SPCC computer printouts were manually
reviewed to select all applicable cases. From these cases, all data on
size and cause of spill were entered onto the PIRS or SPCC
Regional Data Compilation Form designed to facilitate later quan-
tification of probability. Cause descriptors were grouped into four
"failure point" categories:
•Containment device—structural failure of tank or drum
•Operations—all human error with tank overflow highlighted as
the most frequent and germane
•Ancillary Equipment—structural/equipment failure other than
tanks or drums
•Other—fire, explosion, vandalism, acts of nature
The PIRS and SPCC data include spill reports on many different
materials, such as petroleum, petroleum fuel and non-fuel pro-
ducts, and organic and inorganic chemicals. In order to obtain a
sufficient number of data points for analysis, spills of all materials
were combined under the assumption that the occurrence and size
of spills was independent of the material stored. To check this
assumption, SPCC spill data from USEPA Region IV (one-third of
the total storage-related SPCC spills currently in the data base)
were broken out to compare spills of bulk stored petroleum pro-
ducts used as fuel with spills of chemicals and non-fuel petroleum
products.
The differences between results for the two groups were com-
pared and found to be statistically but not highly significant (\2 =
11.16, 3df.). The percentages for chemicals and non-fuel petroleum
products were then compared to those obtained for all SPCC
recorded storage spills for all regions (Table 2) and again dif-
ferences were found to be not significant (x2 = 6.51, 3df.). Given
the uncertainties and limitations associated with the data (as previ-
ously discussed), it was decided that fuel and chemical spills should
Table 1.
Fuel/Chemical Breakdown of Region IV SPCC Spill Incidents
Amount Spilled (gallons)
Cause
0- 50- 100- 250- 500- 1000-
49 99 249 499 999 10,000 >10,000 TOTAL
CHEMICALS & NON-
FUEL PETROLEUM
PRODUCTS
Containment
Operations
Ancillary
Other
5
21
19
30
1
3
10
4
5
12
18
5
1
8
n
5
2
13
11
0
8
20
46
6
2
3
14
4
24
80
129
54
8
28
45
19
TOTAL
75 18 40 25 26
80
23 287 100
BULK STORAGE OF
PETROLEUM
PRODUCTS USED
AS FUELS
Containment
Operations
Ancillary
Other
TOTAL
iRANO *OTAL
14
95
87
5
201
276
8
31
42
7
88
106
12
63
71
12
158
198
5
34
32
4
75
100
3
38
46
8
95
121
14
46
80
16
156
236
3
4
5
11
23
46
59
311
363
63
796
1083
7
39
46
8
100
be regrouped and treated as one data set for the remainder of the
analysis.
Finally, the data from all regions were combined and pro-
babilities (percent of total number of spill incidents) were
calculated and graphed for PIRS and SPCC.
ASSUMPTIONS
Due to limitations in the descriptive ability of the PIRS and
SPCC data bases, and deviations in reporting, certain assumptions
were developed to enable analysis of the data. The most salient of
these are as follows:
•That, due to the fact that failures in pipes, pumps, and valves
reported in the SPCC system could not be attributed in all cases
to storage, the PIRS data only would be relied upon for final
analysis. SPCC data would be presented for comparison only.
•That tank overflow is a result of human error and, therefore,
belongs under Operations. As explained by investigators' re-
ports, the majority of overflows are due to operator error, but
this category will likely contain some incidents caused by me-
chanical failure as well.
RESULTS OF THE ANALYSIS
The existence of two systems, designed to record spill incidents
reported under the FWPA, offers a unique opportunity to compare
the two sets of data. The SPCC system contains inland water-
related spills while the PIRS system contains coastal navigable in-
land water-related spills plus those USEPA cases carried to litiga-
tion (USCG sets civil penalties in litigated cases). The exact degree
of this overlap is not known and, therefore, the data were not com-
bined. The systems are treated as two separate data sets expected to
correlate strongly, except that the SPCC data would be expected to
show more failures related to pipes, pumps, and valves since these
items were not limited to storage-related uses.
Comparison of Bulk Stored Petroleum Products Used
as Fuels and Storage of Chemicals and Non-Fuel
Petroleum Products
Region IV provided approximately one-third of the SPCC data
for hazardous materials storage. Spill incidents from this Region
were grouped to compare data on spills from bulk storage of
petroleum products used as fuels with data on spills from storage of
chemicals and non-fuel petroleum products (Table 1). Of the
Region IV spills, 73 % involved petroleum products. In Table 2, the
petroleum and non-petroleum cause-of-failure percentages are
compared.
The higher percent of spills during operations for fuel storage
reflects a higher relative number of tank overflows. The higher
"Other" percentage for chemicals reflects a higher relative number
of fires and explosions.
Data Presentation
Raw data for the failure incident reports are presented in Tables
3 (PIRS) and 4 (SPCC). These data show that, by an overwhelming
amount, spills are most likely to be the result of operational errors
or ancillary equipment breakdown.
The data were analyzed to determine if spill size could be related
to point of failure. Initially, the percentage of failures in each cause
category were developed for each size. These data are presented in
Tables 5 (PIRS) and 6 (SPCC). From these percentages, probability
curves were developed for certain cause categories of regulatory in-
terest.
Table 2.
Summary of Spills by Cause
Cause Fuels Chemicals
Containment 7% 9%
Operations 39% 28%
Ancillary 46% 45%
Other 8% 19%
-------
DATA BASE
359
Results of this analysis for failures of the containment vessel are
presented graphically in Figs. 1 and 2. From these graphs, it can be
determined that the mean size failure for containment devices is in
the 300 to 350 gal range, and that only 9 to 10% of spills from this
source exceed 5,000 gal in size.
The results of this analysis for operations (divided into tank
overflow and other categories), are shown in Figs. 3 and 4. From
these graphs, it can be determined that the mean size for overflows
is 150-175 gal, and that only between 2 and 5% of spills from
overflows exceed 5,000 gal. For other operations, it can be deter-
mined that the mean is 90-150 gal, and that between 2 and 5% of
these spills are over 5,000 gal.
The results of a similar analysis for ancillary equipment are given
in Figs. 5 and 6. For pumps, the mean spill was between 80 and 180
gal, and between 4 and 10% of spills were above 5,000 gal. For
pipes, the mean was between 35 and 250 gal, with between 3 and
5% of spills over 5,000 gal. For valves, the mean ranged from 200
to 400 gal, with between 7 and 9% of spills above 5,000 gal. For the
other.category, the mean ranged from 25 to 150 gal, with between 2
and 3% of spills over 5,000 gal.
The other category, as shown in Figs. 7 and 8 has a mean size of
between 30 and 300 gal for fires, explosions, flooding weather, and
other natural disasters. This spread is to be expected since the
Table 3.
Spill Incidents from PIRS Data Base (1974-1980)
Amount Spilled (gallons)
categories do not exactly correspond with the PIRS and SPCC
systems for these areas. Spills of over 5,000 gal accounted for bet-
ween 9 and 13% of spills for these categories. For the vandalism
category, the mean size was approximately 500 gal, with 10% of
spills over 5000 gal.
Table 4.
Spill Incidents from SPCC Data Base (1975-1980)
Amount Spilled (gallons)
Cause 0- 50- 100- 250- 500- 1000-
49 99 249 499 999 9999 >1 0,000 TOTAL (%)*
CONTAINMENT DEVICE
Tank Rupture/Leak1 10 1 5 4 7 9 3 39 3
Tank Corrosion 213 23 2 13
SUBTOTAL 12 2 8 4 9 12 5 52 4
OPERATIONS
Tank Overfill 32 12 27 21 14 27 3 136 10.5
Other2 177 48 67 40 28 88 24 472 36.5
SUBTOTAL 209 60 94 61 42 115 27 608 47
ANCILLARY
Pipes3 125 16 41 22 17 33 9 263 21
Pumps* 146933 7 3 45 3
Valves5 111210 7 9 14 4 67 5
Secondary
Containment 3 14
Other? 87 12 15 12 13 28 167 13
SUBTOTAL 240 46 75 44 42 83 16 546 42
OTHER
Fire/Explosion 1 2 3
Adverse Weather 4161 4 1 171
Natural Disaster 31 11 6
Other8 21 5 10 1 5 22 11 75 6
SUBTOTAL 25 9 17 2 6 28 14 101 7
TOTAL 486 ll7 194 111 99 238 62 1307 100
Percent of Total
Spill Incidents 37% 9% 15% 8% 8% 18% 5%
'Relative Probability
1. Minor damage; other casualty, design fault; material fault; structural failure.
2. Personnel error; improper maintenance; overpressurization; capsizing; overturning; collision;
grounding; improper installation; hose, pipe or loading arm cut/twisted; improper valve operation;
flanges improperly secured.
3. Minor damage; excessive wear; corrosion; design fault; material fault; other.
4. Minor damage; excessive wear; corrosion; material fault; design fault; other
5. Minor damage; excessive wear; corrosion; material fault; design fault; other.
6. Dike rupture or leak due to material fault or design fault.
7. Gasket failure due to minor damage, excessive wear, material fault, corrosion, other equipment
failure.
8. Container lost intact or other structural failure due to other or unknown factor.
ANCILLARY
Pipes2 181 94 192 131 130 232 43 1003 33
Pumps3 31 10 12 9 12 16 10 100 3
Valves* 30 24 29 24 37 68 19 231 8
Secondary
Containment5 6 2 3 5 16 1
Other6 31 9 16 6 8 25 2 97 3
TOTAL 279 137 251 170 187 344 79 1447 48
OTHER
Fire 12 3 11 1 1 8 3 39 1
Explosion 27 1 9 1 6 44 1
Vandalism 20 10 14 13 16 40 11 124 4
Flooding 21 4 10 5 4 14 3 61 2
Other7 2368106 17 5 75 3
TOTAL 103 24 52 29 28 79 28 343 11
GRAND TOTAL 715 284 529 305 352 686 163 3034 100
Percent of Total
Spill Incidents 24% 9% 17% 10% 12% 23% 5%
*Relative Probability
SPCC INCIDENT FOOTNOTES
1. Valve handling, collision with vehicles, improperly secured flange, bucket spilled, student broke
jar, punched drum, car hit pump hose, truck knocked valve open, line overload, cut hose line, con-
struction work, excavation, open valve, tank cleaning, improper hose connection, improperly sealed,
carelessness/improper/inadequate handling, negligence, loose cap, spillage, trash can tipped,
poor/improper maintenance, improper cleaning, overflow spillage due to inattention, cleaning, per-
sonnel, operations, overturned drum, emptying tank, oil bucket overturned, backhoe dug up
pipeline, pump was run over, forklift poked hole in tank, tank trailer dropped, pumping tank out,
pump left on, tractor ran over pipe, drop tank while moving, uncoupling, off loading, improper
valve handling, overpressurization, tank plug missing, unloading error, operator error, human/per-
sonnel error, overfilled container tank overturn, forklifl punctured drum, hand truck failure, hit by
crane, backflow, poor housekeeping, welding spark, oil put in wrong tank, overturning.
2. Frozen line, regulator malfunction, pipeline corrosion, hose/pipe/line rupture, collar broke, line
ing system leak, plugged siphon line, injection line rupture, cooling line coll crack, clogged pipeline,
leaky union, pipeline flange leak, cracked/faulty line, plugged/line/leak, kinked hose, pipe
blockage, pipeline leak.
3. Pump failure rupture, leaking seal, broken/ruptured casing, leaking packing, cracked siphon,
hyd. pump, broken pump, ruptured pump seal, frozen pump.
4. Defective valve, blown valve packing, installing valves, valve rupture, leaking stop.
5. Defective sump pit, dike failure/leak, dike washout/broke, drainage system failure, overflow
catchment bod, broken dike, sump overflow, sump pump.
6. Automatic shutdown, alarm failure, break in coupling, nozzle malfunction, other connection,
broken fitting, gauge malfunction, loose plug, control malfunction, pressure release, welding spark,
pressure blowoff, leaking flange, broken coupling, faulty/gasket/blow, equipment malfunc-
tion/failure, broken hose coupling, coupling leak/failure, loose fitting, fitting failure/leak, broken
joint, tub rupture, disc rupture, gasket rupture, gasket leak/failure, seal failure, seal blowout, nozzle
failure, gauge failure/leak, metering system, alarm failure, bushing failed.
7. Weld failure factory/field, freezing, heavy winds/rains, lightening, hot weather, natural disaster
(rockslide, lightening, earthquake, landslide), seam failure, other rupture, corrosion.
-------
360
DATA BASE
Table S.
PIRS Spill Percentages (1974-1980)
Amount Spilled (gallons)
Cause
0- 50- 100- 250- 500- 1000-
49 99 249 499 999 10,000 >10,000
CONTAINMENT DEVICE
Tank Rupture/Leak'
Tank Corrosion
OPERATIONS
Tank Overfill
Other2
ANCILLARY
Pipes^
Pumps4
Valves5
Secondary
Containment5
Other?
26
15
24
38
48
31
16
75
52
3
3
9
10
6
13
18
7
13
23
20
14
16
20
15
9
10
15
8
8
7
10
7
18
15
10
6
6
7
13
8
23
23
20
19
13
16
21
25
17
8
15
2
5
3
7
6
OTHER
Fire/Explosion
Adverse Weather
Natural Disaster
Other8
24 6
50
28 7
4
17
13
17
7
33
24
16
29
67
6
15
Table 6.
SPCC Spill Percentages (1975-1980)
Amount Spilled (gallons)
Cause
0- 50- 100- 250- 500- 1000-
49 99 249 499 999 10,000 >10,000
CONtAINMENt DEVICE
Tank Corrosion
Tank Rupture
Tank Leak/Seeping
Drum/Rupture/
Leak/Rotted
Underground
Leak/Tank
18
15
29
45
25
5
7
n
18
25
23
12
22
18
50
7
7
6
15
8
9
9
30
33
19
9
Z
18
3
Leaking Bladder 100
OPERATIONS
Tank Overflow
Other1
ANCILLARY
Pipes*
Pumps3,
Valves4
Secondary
Containment'
Other6
OTHER
Fire
Explosion
Vandalism
Flooding
Other7
27
31
18
31
13
38
32
31
61
16
34
31
10
11
9
10
10
9
8
2
8
7
8
21
14
19
12
13
13
16
28
20
11
16
11
10
8
13
9
10
6
3
10
8
13
11
11
13
12
16
8
3
2
13
7
8
19
21
23
16
29
19
25
21
32
23
23
3
4
4
10
8
31
I
8
H
9
5
7
S 10 IS 20 » 40 SO 60 70 SO 85 90 » 96*
5 10 IS 20 30 40 50 60 70 80 B* 90
Figure 1.
PIRS Containment Failure Probability Curve
Figure 2.
SPCC Containment Failure Probability Curve
-------
DATA BASE 361
5 10 15 20 30 40 50 60 70 80 85 90 95
Figure 3.
PIRS Operations Failure Probability Curve
Figure 4.
SPCC Operations Failure Probability Curve
Figure 5.
PIRS Ancillary Equipment Failure Probability Curve
Figure 6.
SPCC Ancillary Equipment Failure Probability Curve
-------
362
DATA BASE
Analysis of Results
PIRS data were analyzed to determine if spill size distribution
was correlated to cause or failure point.
For containment device failures, spill size was not found to vary
significantly by cause (tank rupture or corrosion) (x2 = 2.104,
5df.). For pipe, pump, and valve failures, spill size was also not
found to vary significantly by point of failure (x2 = 12.95, lOdf.).
However, for operation failures, size distribution for tank overfill
and other categories were found to vary significantly (x2 = 17.75,
5df.).
Further comparisons were made of data on containment device
failure, tank overfills, and pipe, pump, and valve failures. The size
distributions for these categories were found to vary significantly
by cause (x2 = 45.4, lOdf.).
"Other" categories (except "other" under operations) were not
further examined because they represented a very large number of
different causatory events, each with a very low probability.
"Other" failures under operations represented a large number of
unusual failures directly attributable to human error.
CONCLUSIONS
Based upon this analysis of historical spill data for tank storage,
the following conclusions can be made:
•The major causes of spills from tank storage operations are
operator error and ancillary equipment failure, which together
account for 89% of storage related spills in the PIRS data base.
•Spills from the tank itself, due to tank failure, overfilling, ad-
verse weather, or fires and explosions, account for only 16.5%
of spills. However, tank failures and spills due to adverse weather
or fire and explosion (5% of all spills) are significantly larger on
average than other spills.
•The large number of spills (55.5% of the total) falling into the
"other" subcategories within the containment device, operations,
ancillary equipment, and other categories, indicated the ex-
tremely diverse nature of accidents.
•The size of the spill is at least partially dependent upon cause or
point of failure. Unfortunately, the data base does not have in-
formation on the capacity of the storage equipment involved
in the failure, so it does not allow for analysis of spill size de-
pendence on this factor. However, it is likely that spill size is
primarily dependent on capacity of the storage system and only
secondarily dependent upon cause or point of failure.
The implications of these results in terms of developing regula-
tions to protect human health and the environment from adverse
effects of hazardous waste tank storage have not been fully anal-
yzed in light of other information being developed on this project.
However, these data do indicate that good design of primary con-
tainment systems (tanks, pipes, pumps, valves) may not solely pro-
vide adequate protection for the environment due to the numerous
possible causes of spills which are largely unforeseen at the time of
design. This situation reveals a need for mitigative measures, such
as secondary containment around storage areas and the limitation
of ancillary equipment to the secondary containment area to the ex-
tent practical, to adequately protect human health and the environ-
ment from potential adverse effects of hazardous waste storage.
ACKNOWLEDGEMENTS
This work was supported by the U.S. Environmental Protection
Agency under Contract No. 68-03-3115. The assistance and advice
provided by Ms. Penny Hansen and Mr. John Heffelfinger of the
Office of Solid Waste are gratefully acknowledged. Additionally,
the authors would like to acknowledge the efforts of the JRB sup-
port staff in reviewing over 5,000 damage reports.
REFERENCES
1. "Spill Prevention, Control, and Countermeasure Database." USEPA,
Office of Emergency and Remedial Response, Emergency Response
Division, Data obtained between Dec. 1981 and Feb. 1982.
2. Pollution Incident Retrieval System, Department of Transportation,
U.S. Coast Guard, Data obtained between Dec. 1981 and Jan. 1982.
3. The Federal Water Pollution Control Act Amendments of 1972, 35
U.S.C. 1251 et seg., as amended by the Clean Water Act of-1977,
P.O. 95-217. Section 311(b).
Figure 7.
PIRS Olher Failure Probability Curve
Figure 8.
SPCC Other Failure Probability Curve
-------
THE HAZARDOUS MATERIALS TECHNICAL CENTER
DAVID A. APPLER
Defense Logistics Agency
Cameron Station, Virginia
MURRAY!. BROWN
U.S. Army Environmental Hygiene Agency
Edgewood, Maryland
TORSTEN ROTHMAN
Dynamac Corporation
Rockville, Maryland
INTRODUCTION
The Hazardous Materials Technical Center (HMTC) was estab-
lished in June 1982 by the Defense Logistics Agency (DLA) to pro-
vide a center of expertise on technology and regulations related
to handling, storage, transportation and disposal of hazardous ma-
terials. The need for an HMTC began with the assignment of re-
sponsibility to DLA for managing most of the hazardous wastes in
the Defense Department (DOD). To help in carrying out this re-
sponsibility DLA decided to establish a contractor operated Haz-
ardous Materials Technical Center. Through competitive procure-
ment the Dynamac Corporation of Rockville, Maryland was se-
lected to establish and operate this Center.
This presentation discusses the purpose, potential users and
functions of the HMTC. The Center is an Information Analysis
Center operated for DLA with technical supervision by the U.S.
Army Environmental Hygiene Agency and DLA.
The purpose of the HMTC is to provide a single location as a
source for information on all aspects of hazardous materials tech-
nology and regulatory requirements. It will consolidate many scat-
tered information resources and other pertinent data which are cur-
rently available. The initial users will be primarily DLA and the
Military Services with information provided to other Federal, state
and local government agencies and the commercial/industrial sec-
tor on a non-interference basis.
The specific functional activities and the potential informational
needs that will be met are shown in the following table:
Activity
Occupational Safety
and Health
Shipping, Receiving
and Traffic
Warehousing
Need(s)
Information on all aspects of safety,
health, handling and regulatory com-
pliance for safety and health surveys
Information on containerization, stor-
age compatibility, transportation, spill
cleanup and recoupment
Information on containerization, stor-
age compatibility, spill cleanup and re-
coupment
Information on handling and disposal
Information on disposal, spill cleanup,
recoupment and regulatory compliance
Facilities
Environmental
FUNCTIONS
The HMTC is organized along two functional lines: develop-
ment of the Disposal File for the Hazardous "Materials Informa-
tion System (HMIS) and other products and services. The first
functional area to be discussed is the HMIS Disposal File.
The data files and the data elements currently in HMIS are
described in the following overview, setting the stage for discus-
sion of the proposed data elements for the Disposal File.
AN OVERVIEW OF THE EXISTING HMIS
The HMIS currently has two data files:
•Safety and Health File—for which the information comes pri-
marily from the Material Safety Data Sheets (MSDS) provided
by the supplier of the material
•Transportation File—for which information is prepared by the
person developing the data for the HMIS based on the MSDS
and other data
A third file for the HMIS, primarily concerned with disposal
information, will be developed by the HMTC in close coordina-
tion with the users of the HMIS.
Starting in 1976, Federal government agencies were required to
obtain Material Safety Data Sheets (MSDS) for all procured haz-
ardous materials. The MSDS contains health and safety informa-
tion pertaining to that particular hazardous material. To provide a
comprehensive and organized repository for the data in the
MSDS's, DLA established the Hazardous Materials Information
System (HMIS) in 1978. HMIS was also intended to assist in com-
pliance with pertinent regulations in the areas of safety, health and
transportation. The HMIS is operated and maintained by the De-
fense General Supply Center (DGSC) in Richmond, Virginia with
input from the several agencies within DOD.
The Safety and Health File is comprised of four major sections:
(1) Identification and Logistics, (2) Chemical Properties, (3) Safety
and Health Properties, and (4) Storage, Spill, Leak and Disposal
Procedures. The Identification and Logistical data section contains
key items for products (such as National or Local Stock Number,
Federal Supply Code For Manufacturers, NIOSH Code, Focal
Point Indicator, etc.) which are unique to the product and may
serve as links to the other files (Transportation and Disposal).
These data elements may also serve as a basis for retrieving in-
formation from the Safety and Health File. The Chemical Prop-
erties section provides the file with information about the chemical
components and the physical and chemical properties of the ma-
terial.
The Safety and Health Properties section contains information
on such items as explosive concentrations, threshold limit value
(TLV), first aid procedures, hazardous decomposition products,
protective equipment, etc.
The Storage, Spill, Leak and Disposal Procedures section pro-
vides the user of the product with guidance on action to take in
storing the product and how to handle a spill of the product and
then in what manner the contained, spilled or leaked material
should be disposed. The waste disposal referred to here is for the
disposal of materials used to clean up spills and is very general in
nature. A typical instruction is "Place material in suitable con-
tainer for shipment to disposal area". Obviously, this instruction
is not enough to insure compliance with regulatory requirements
for routine hazardous materials disposal.
363
-------
364
DATA BASE
Similarly, the Transportation File has an Identification and
Logistical set of data elements in addition to Transportation data
elements. The Identification and Logistical data in the Transpor-
tation File serve the same purpose as those data in the Safety and
Health File. However, the Transportation data pertains exclusive-
ly to the manner in which hazardous materials are shipped. Be-
ginning with the product measurement data and special chemical
classes in regard to shipping and ending with the different avenues
of shipment (road, water or air), the transportation data provide
the user with a comprehensive source of information. This infor-
mation is used to determine transportation restrictions, shipping
modes, packaging requirements, labeling requirements, manifest-
ing requirements, etc.
The final data file is the Disposal File which will be developed
by the HMTC in close coordination with the using activities and
based on their experience.
HMIS—PROPOSED
It is proposed that five major sections of data will comprise the
Disposal File. These sections include:
•Mandatory Data Elements
•Disposal Criteria
•Special Characteristics
•Transportation Data
•Supplementary Data
The Mandatory Data Elements Section will be comprised partly
of those data elements which are similar to the Identification and
Logistical data for the existing Safety and Health file and Trans-
portation file of the HMIS. These data elements will be included
in this file so that it will be able to stand alone. There will also be
several new data elements in this file. For the purposes of this dis-
cussion, only the new proposed data elements are presented for
consideration.
The first new data element proposed [Mandatory Data Ele-
ments] is:
•Is it reportable to Defense Property Disposal Office (DPDO)
by DTID (Disposal Turn in Document)?
DOD policy has established eight categories of hazardous ma-
terials such as:
•Biological/Chemical Warfare Agents
•Unique R&D wastes
•Municipal wastewater sludges, etc.
that are not reported to DPDO for disposal. DOD policy is that
these materials will be the responsibility of the generator of the ma-
terial. This first new Mandatory Data Element would indicate
whether or not the material is in any of those categories.
The second data element [Disposal Criteria] is:
•Is it reportable to DPDO for services only?
This information will indicate those items for which DPDO will
provide disposal assistance even though they are not required to
formally accept the item, i.e., a service contract to dispose of in-
dustrial sludges.
The third data element [Special Characteristics] is:
•Can hazardous property bypass all or part of disposal cycle?
This information is of prime importance to materials which are
considered hazardous waste upon generations. Therefore, if the
answer is yes, it will normally be followed by a qualifier statement
in the supplemental data field such as:
"It is a waste upon generation", e.g., spent solvents.
"Do not sell or donate", e.g., PCB transformers.
The next data element [Transportation Data] is:
•The EPA Hazardous Waste Code
This code is assigned by EPA in the subtitle C, Hazardous Waste
Regulations for implementation of RCRA and is needed on man-
ifests to identify a specific type of waste. Some examples are:
PO37 for dieldrin
KO47 for red water from TNT operations
Similarly, the next data element [Supplementary Data] is:
•The EPA Hazardous Characteristic(s)
This data element gives the characteristic that caused the material
to be declared hazardous, i.e., ignitability, corrosivity, reactivity
or toxicity.
A closely related data element that is proposed is:
•Is it an Acute Hazardous Waste?
Designation as an Acute Hazardous Waste results in requirements
for special and more restrictive disposal methods and handling of
storage containers.
The final proposed new Mandatory Data Elements are:
•Is an Environmental Impact Statement (EIS)/Environmental As-
sessment (EA) available:
•Has a Categorical Exclusion been granted?
Certain chemicals such as DDT has had an EIS prepared for its
disposal. Certain wastes may have Categorical Exclusions from the
EIA/EA requirement granted by the Council on Environmental
Quality.
The next major section in the Disposal File is Disposal Criteria.
The first proposed data element in this section is:
•Disposal Restrictions
This data element will be a narrative that describes policy and regu-
latory restrictions such as:
•Can the material/waste only be sold to licensed persons? e.g.,
pesticides
•Can the material be resold overseas but not in the United States or
vice-versa?
•Special pre-processing needs may be identified, e.g., fixation of
an inorganic heavy metal sludge prior to disposal.
•Qualifier's for IFB's (invitation for bid).
•Is the hazardous material/waste covered by another regulation(s),
e.g.,TSCA,FIFRA?
•Special DOD or DLA policy restricting disposal of certain items.
The second proposed data element in the Disposal Criteria is:
•Technical Disposal Code
This will be an alpha numeric code which will reference a DOD
handbook, which will be prepared for technical disposal instruc-
tions. The complete disposal procedure is not included in the data
element since only a limited number of disposal options are avail-
able and the inclusion of these options in each data file would re-
sult in excessive repetition. In addition, some instructions may be
very lengthy and would be better referenced. Furthermore, this
data element would also have information on preferable and accep-
table methods for disposal and any technical restrictions in regard
to disposal methods or procedures.
The next section in the Disposal File is Special Characteristics.
This section highlights those characteristics of a hazardous waste
which require special attention in the data elements such as:
•Handling/Storage Precautions:
-avoid heat or cold or water that could change the characteristics
of the material/waste and make it more hazardous
-special caution for materials that change on aging
-spill residue handling which would indicate the material used in
cleanup becomes a RCRA controlled hazardous waste
•Storage Compatibility
-chemical characteristics to be aware of that would require stor-
age restrictions, e.g., not storing acids near poisons, etc.
•LD^/LC^ Information
-this data would provide additional information on those ma-
terials designated as Acute Hazardous Wastes
These elements are the proposed new data for the HMIS Disposal
File. The information for the data elements in the Transportation
section of the Disposal File will primarily come from the existing
-------
DATA BASE
365
HMIS Transportation File, and the information in the Supplemen-
tary section will be used to amplify any information in the other
data elements regarding further clarification.
As was previously discussed, this approach to the Disposal File is
a starting point. Work will be done with the users to fully define
their needs and how they can best be met.
Although the data elements proposed for the Disposal File ap-
pear primarily administrative and logistical in nature, they are syn-
thesized from extensive scientific and technical information on the
hazardous materials to ensure the procedures recommended are
technically sound, meet all environmental and health regulations,
and are implementable.
For example, the following factors are considered in the develop-
ment of these data elements:
•Potential degradation due to long term storage, e.g., picric acid
will produce explosive crystals
•Environmental transport mechanisms and environmental fate
determined by the chemical/physical properties of the material
and specific site characteristics
•Treatment technology evolutions such as biodegradation, micro-
organism acclimatization, chemical neutralization, fixation and
solidification, slow and rapid oxidation
•Industrial hygiene and engineering control technology as parts of
a safety and health program for disposal workers
These are some examples of the scientific and technical considera-
tions which go into the developing of the Disposal File.
OTHER PRODUCTS AND SERVICES
The second major functional area of HMTC is Other Products
and Services. This function includes the development of products
such as:
•Handbooks—general guidance documents on broad topics
Potential topics might be: Organic Solvent Disposal
Pesticide Disposal
Storage of Hazardous Materials
•State-of-the-Art Reports (SOAR)—more specific and limited than
handbooks and can be considered monographs targeted to an ex-
perienced technical audience. Some potential topics are:
Hazardous Waste Exchanges
Disposal of PCP (pentachlorophenol)—
impregnated ammo boxes (which might
include the identification of potential
substitute chemicals)
•Critical Reviews/Technical Assessments—are similar to State-of-
the Art Reports but even more specialized.
•Abstracts and Indices—prepared quarterly by HMTC. The Center
will compile all the new abstracts of the pertinent literature added
to the Center's repository on a quarterly basis. HMTC will also
prepare an annual publication of all the abstracts for that year.
In addition, the HMTC will provide detailed responses to tech-
nical inquiries which can be made by phone or letter. Further-
more, the Center will prepare bibliographic reports upon request.
Finally, a catch-all category of special studies is included in case
some project does not fit into any of the other categories for
products or services.
To provide rapid access to the information HMTC has devel-
oped, an IBM System 34 will be used to catalog and store the data.
HMTC will not duplicate existing data bases or information net-
works, but will make use of them. The Center's system will be lim-
ited to information which is not available from other easily access-
ible sources.
PRODUCT/SERVICE DELIVERY
The delivery of these HMTC products and services begins with
the processing of a customer request received by mail or phone. A
routine technical inquiry or request for readily available material
will be processed in three days. A technical inquiry requiring
preparation of a custom tailored response and requests for bib-
liographic reports will be completed in ten days. The generation
of a new product such as a handbook or SOAR may take a year or
longer. For this effort, a detailed scope and cost estimate are pre-
pared for the client's approval before proceeding.
-------
DOCUMENTATION FOR COST RECOVERY UNDER CERCLA
R. CHARLES MORGAN
BARBARA ELKUS
U.S. Environmental Protection Agency
Office of Waste Programs Enforcement
Washington, D.C.
INTRODUCTION
Section 107 of the Comprehensive Environmental Response,
Compensation, and Liability Act of 1980 (CERCLA)1 authoritizes
the Federal Government to recover cleanup costs, costs of Fed-
eral employees' efforts, and costs of damage to or loss of natural
resources from a responsible party. Documentation to recover
these costs begins with the notice that release of a hazardous sub-
stance has occurred.
Section 103 of CERCLA requires that persons in charge of ves-
sels or facilities, as well as owners or operators, report releases of a
hazardous substance and the conditions of the release. Section
103 of CERCLA provides for response actions consistent with the
National Contingency Plan (NCP) to remove and remedy the re-
lease as necessary to protect public health, public welfare or the
environment.
Whenever the government undertakes a response action, various
information gathering activities must be included in order to docu-
ment the existence and extent of the release, the source and nature
of the hazardous substance, the extent of hazard to the public or
the environment, and the responsible party. CERCLA §107 author-
izes recovery of funds spent on response actions from a respon-
sible party and such a cost recovery action will be anticipated in
all cases by the collection and preservation of appropriate evi-
dence and documentation. While not all costs will be recoverable
(e.g., cleanup of a federally-permitted release), all costs will be doc-
umented pending a later legal determination.
In the documentation of critical facts and response costs, it is
important to demonstrate:
•Conditions that existed at the site at the time of release
•The association between the site and the owner/operator, trans-
porter and/or generator
•The rationale and physical evidence justifying a removal and/or
remedial action
•The development of a remedy not inconsistent with the NCP and
its cost-effectiveness
•The total costs including Federal employee expenses paid to clean
up the site
These facts will be included in a site file which is established at
the outset of any identification of a release of a hazardous sub-
stance. The file will be organized by phases and contain the doc-
uments which record the major decisions made during each phase.
Each of the documents will be signed and dated by the preparer
for authentication. Those involved in data collection and decision-
making will be identified by name, qualifications, and how they
may be contacted. Although the original of each document will
be collected, copies of pertinent documents are acceptable as long
as the preparer can substantiate that a signature had been affixed to
the original. At each step or phase, records which are required
to support a cost recovery action will be added to the file. The
compilation of facts in this manner provides both a complete file
and efficient document control (chain of custody of the docu-
ments).:
The phased approach for documentation of a cost recovery ac-
tion described here follows the phased approach used by the pro-
gram office at USEPA to investigate sites and design remedial ac-
tions. Generally, surveys, studies, and information searches are in-
itially conducted to document the responsible party and the areal
and vertical extent of contamination. Based upon these data, a re-
moval and/or remedial program is devised, for the purpose of iso-
lating and/or cleaning up the contamination to protect the public
health, welfare and environment.
In order to determine the effectiveness of the remedy, a de-
tailed monitoring program may be implemented. If required to in-
sure the continued effectiveness of the remedy, operation and
maintenance may continue. Health and safety programs may also
be implemented. Documentation of the decisions made in each
phase are made a part of the site file as they are generated.
These phases are outlined in Subpart F of the NCP.3 In this
paper, USEPA tracks those phases, not so much to describe the
procedures employed, but to describe the documentation necessary
in each phase. These documents become the factual and evidentiary
information to support a cost recovery action. Documentation in-
cludes not only descriptions of what was done but also the ration-
ale for a particular choice of response action, and how it is consis-
tent with the NCP. The documentation will include records that a
release is threatened or has occurred, conditions at the site, the
link between a responsible party (if any) and the site, the technical
justifications for the removal and/or remedial action, and the total
costs and claims paid to cleanup^the site consistent with CERCLA
and the NCP.
Cost recovery may begin after the completion of any phase.
Therefore, it is essential to maintain documentation at the outset
of the investigation and as the cleanup proceeds.
The development and preservation of evidence for a cost recov-
ery action require a concerted technical effort that must be accom-
panied by good bookkeeping and record collecting techniques.
Standardization of data collection, use of site specific account
numbers for all Federal expenses related to site work, chain-of-
custody and document control such as those described by the
USEPA National Enforcement Investigations Center's document
control system, will greatly aid the Agency's ability to recover
funds spent.
PHASE I—DISCOVERY OR NOTIFICATION
In the discovery or notification phase, record of a release of a
hazardous substance will be documented. The discovery may take
place as a result of a compliance investigation conducted by gov-
ernment authorities in accordance with the inspection authority
of CERCLA or other statutory authorities pertinent to govern-
mental agencies. Discovery may also occur as the result of inventory
efforts or random incidental observations by government agencies
or the public.' Notification may occur in accordance with section
103(a) and (c) of CERCLA which require persons in charge of a
vessel or facility and owners or operators of a facility to notify the
National Response Center (NRC) and USEPA respectively, of any
366
-------
DATA BASE
367
DISCOVERY OR NOTIFICATION
NOTIFICATION
COMPLIANCE
INVESTIGATION
(SIMM CEHCLAI
S3013 RCRA
RANDOM
OBSERVATIONS
S103I.I
OR Icl CERCLA
PERMIT
REQUIREMENT
• STATE OR LOCAL
INSPECTION REPORTS
• FEDERAL INSPECTION
REPORTS
• INVESTIGATIVE REPORTS
• ANALYTICAL REPORTS
• PHOTOGRAPHS
I
• TELEPHONE RECORDS
• CORRESPONDENCE
• PHOTOGRAPHS
• INVENTORY RECORDS
• PERSON IN CHARGE
MUST NOTIFY-TELE
PHONE RECORD INRCI
• OWNER/OPERATOR
MUST NOTIFY WRITTEN
DEPORTS IEPAI
• PERMIT COMPLIANCE
REPORT
• TELEPHONE RECORD
WITH FOLLOWUP
,
Figure 1.
Discovery or Notification Flow
release. Notification may also occur as a result of Federal or State
permit requirements.
This discovery or notification scheme is depicted in Fig. 1 and
examples of records necessary to document the discovery or notif-
ication are included in the flow. In the case of a compliance in-
vestigation, documentation of a release may be presented in local,
State or Federal compliance inspection reports or in reports pre-
pared by owner/operators of facilities. Accompanying the inves-
tigative report may be the results of chemical analyses and an
analytical report describing the constituents of the waste. The in-
vestigative report may also include a statement on the toxicity of
the waste and its association with the site. Of course photographs
of the site and its conditions are a vital record to document a re-
lease.
Documentation of random observations may include records
such as telephone conversations, correspondence from facility
employees or private citizens, private photographs, and inventory
records. These random observations will need to be corroborated
by evidence provided by an appropriate governmental agency or
the respondent.
Notification under section 103(a) of CERCLA would normally
be documented by a telephone record at the NRC and record of
conveyance of the information to appropriate governmental agen-
cies. Notification under section 103(c) would normally be by a
written report submitted by the owner or operator of the facility. It
should describe the existence of the facility that is or has stored,
treated or disposed of a hazardous substance.
Notification may also come from a Federal or State permit
holder as a compliance requirement of the permittee to notify the
appropriate governmental agency of the release of a hazardous
substance. This normally will be in the form of a specified, peri-
odic report or an immediate telephone call followed by a report
submitted to the appropriate governmental agency.
Once discovery and notification documents have been compiled,
a decision will be made as to whether further investigation of the
release of a hazardous substance is needed. Reports from the NRC,
the On Scene Coordinator, and investigative agencies will be used
to make that decision.
The next phase assesses the data compiled during discovery, de-
fines additional data needs, identifies the responsible party and
determines whether a response action is necessary.
PHASE II—PRELIMINARY ASSESSMENT
The preliminary assessment will be based on information col-
lected during data and record reviews, investigations and inspec-
tions. The assessment depicted in Fig. 2 is designed to evaluate the
magnitude of the hazard, identify the source and nature of the re-
lease, identify the responsible party, and evaluate whether immed-
iate response action is necessary. Additional information may be
needed to complete the assessment, and data gaps will be identi-
fied and filled during this phase.
In order to evaluate the magnitude of the hazard, documen-
tation of management practices employed at the site, literature
searches on the physical and chemical properties of the hazardous
substance, their lexicological characteristics, and demographic or
ecological data on areas adjacent to the site are necessary. To aug-
ment these data, interpretation of historical aerial photographs
and results of on-site and off-site inspections are desireable.
To further focus the assessment, an evaluation of the source
and nature of the release will be necessary. Field notes, photo-
graphs of the site, hydrogeological reports, environmental samples
and laboratory analyses will all be valuable. Witness accounts
from employees, private citizens and even owners or operators re-
corded as signed statements or affidavits will aid in the evaluation
of the conditions at the site that might have led to a release. In the
event of an off-site spill, sample results correlated with available
inventory records, bills of lading or manifests will document that
the hazardous substance originated at the site. It should be empha-
sized that collection of this information will be initiated immed-
iately when a response action is anticipated. Photographs before,
during, and after the response action will be desireable to des-
cribe the course and results of the response action.
-------
368
DATA BASE
PRELIMINARY ASSESSMENT
1
t
( EVALUATION OF \ /IDENTIFICATION OF SOURCE\ ( IDENTIFICATION OF "\
^ MAGNfTUDE OF HAZARD ) \ AND NATURE OF RELEASE j \ RESPONSIBLE PARTY J
• STORAGE .
TREATMENT.
MANIFACTURING.
DISPOSAL RECORDS
• UTERATURE SEARCHES
PHYSICAL 6 CHEMICAL
PROPERTIES
• LITERATURE SEARCHES
TOXICOLOGY
• DEMOGRAPHIC OR
ECOLOGY DATA
• AERIAL PHOTOS
• ON AND O
=F-SITE
INSPECTIONS
• EXTENT OF ANY ACUTE
HAZARDS
• FIELD NOTES
• ON AND OFF SITE
PHOTOS
• HYDROGEOLOGICAL
REPORTS
• SAMPLES AND ANALYES
• SIGNED STATEMENTS
• AFFIDAVITS
• INVENTORY RECORDS
• BILLS OF LADING
• MANIFESTS
1
(PLANNED REMOVAL \
^/NE
1
• DEED OR LEASE
• DIRECTORY OF CHEMICAL
PRODUCERS
• SEC RECORDS
• GOVERNMENT RECORDS
• DUN ft BRADSTREET
• MOODY S INDUSTRIAL
MANUAL
• DRUM LABELS. LOT NOS .
SAMPLES
• UTILITY RECORDS
• VEHICLE INFORMATION
• INTERVIEWS
• FINANCIAL INSTITUTIONS
• POLICE El FIRE
DEPARTMENT RECORDS
//X!D\
ED^\
|No
AND/OR REMEDIAL V- ^
-------
DATA BASE
369
LEGAL ADDRE
OR
DESCRIPTIO
*
^_ COUNTY
^" RECORDS
<.
FILE REVIEW
EPA/STATE
Minn rndtirl 1
*lAop(ovai by Re^'onai
i>,n<,B1 ftt-rj""*"11
FACI
NAM
LOCA
^
>
__ s*
SS NO vxxOVM
N ^sj
>
LITY
F g.
TION
^niufB^ iNfORMillON
i <,HP L, ^ ^p°/ftH,EfR,^\ ^ ( SEARCH
1 1 ^^DEmifiE^/ ,flEPrfln
I /
ER- ^s^ YES IMITIAI /
ATOP ^ ^ INITIAL r
ATOR^P ^ CONTACT
*
iNFflBiunnN
SEARCH
NO
PUBLICLY C°"PANY PRIV.IEIY
OV«ED 9jmoFs OWN£0
1 i
CHE£S?A"LV SEC * ,^1 Gm£BNMENI °"NS *
PRODUCERS "ETOBIS ™C'L "ECOBOS BS.DSIREET
1
SYNTHETIC .H0|.« „,
ORGANIC ORGAN ZATION RFmTFR r?im ORCJNIiAT ON
.^_ CHEMICALS REGISTER ESTATE
' .^^ 1
"^ OTHER RELAIED CHANGES* SID 4 POORS UN FORM
> INFORMATION ' INOUS'"£S "ENI ™o™ ™'1
SEARCH
FINANCE
>
N
IDENTIF
>
FIE
INVESTK
fel"PWci l^e 'esi >s manually searcnea
IflAOE
PUBt CAIlONS
1
D
CATION
FIELD
NVESTIGAI ON
I
ONSIIE OFFS IE
J
L , I II I
> POTENTIAL P01ENIIAL POTENTIAL I P01ENNAI I POIENIISL I POTENTIAL
OENERMOR OWN1ROPER TRANSPORTER GENERATOR OWNER OPER TRANSPORTS
"- ID ID ID | ID ! It) | 10
5ATION || | j
I
'TOP . ™ ^™rTr'Ai^N '" ,! CONM«wr.pN
\^JtN HE^x^ i -3a
Figures.
Procedure to Identify Responsible Parties"
-------
370
DATA BASE
IMMEDIATE REMOVAL
PLANNED REMOVAL AND/OR
REMEDIAL ACTION
CONSIDERATIONS FOR
DEFENSIVE ACTIONS
• ALTERNATIVE WATER
SUPPLIES
• LIMIT ACCESS
• CONTROL SOURCE OF
RELEASE
• MONITORING
• MOVING HAZARDOUS
SUBSTANCE
• DETER SPREAD
• EVACUATION
• EXECUTE DAMAGE
CONTROL
'NEED"
'FOR PLANNED-^ Yes
REMOVAL AND/OR
REMEDIAL _>^ Rgure 5.
XCTIOf
Figure 4.
Immediate Removal
agreements, contract documents with contractors, USEPA action
memoranda, and feasibility studies. Estimates of costs of response
will be outlined in such documents.
The NCP indicates that "an inspection will be undertaken to
assess the nature and extent of the release and to assist in deter-
mining the priority for fund-financed response". In accordance
with section 104(b) and (e) of CERCLA, Federal and State offic-
ials may conduct such an inspection to investigate, monitor, sur-
vey, test and conduct other information gathering. The record of
this inspection should clearly document the confidence in the data
collected by outlining the quality assurance and quality control
used in collecting and evaluating the data. Documentation of such
information as analytical detection and quantitation limits as well
as precision and accuracy limits should be expressed. During this
inspection, particular attention should also be paid to situations
which may represent an immediate danger to persons living or
working near the site.
The documentation of the decision to use a particular response
action must be in accordance with provisions of the NCP; that is,
the action must be a cost effective response, and the decision to
use a particular response action must be balanced between the
need for protection and availability of the fund to respond to other
sites. Such decisions are currently documented in "record of de-
cision" memoranda. Supporting documentation is also needed.
Supporting documentation may come from records collected by
the Hazard Ranking System used to set priorities for use of the
CERCLA fund. Documentation and preservation of evidence is
critical during this phase.
I CHEMICAL ANALYSES
OF TAP WATER
t CONDITION AND
ORIENTATION OF BULK
CONTAINERS
I REPORTS OF CONTAMI-
NATED SURFACE SOILS
. RECORDS AND TESTS OF
FIRE AND EXPLOSION
HAZARD
I WEATHER CONDITIONS
THAT ENHANCE
MIGRATION
I HAZARDOUS PROPER
TIES
• HYDROOEOLOOICAL
FACTORS ft CLIMATE
• EXTENT OF MIGRATION
• PAST EXPERIENCES
• ENVIRONMENTAL
WELFARE CONCERNS
CONSTRUCTION )
IPfeunO
• INSPECTION AND IN-
VESTIGATION REPORTS
• MONITORING. SURVEY
TEST RESULTS AND
REPORTS
• FEASIBILITY STUDIES
• ENVIRONMENTAL
INFORMATION
DOCUMENTS
• PROBABLE OR ACTUAL
COST RECORDS
• BALANCE OF FUND
DOCUMENTS
• DEQREE OF ENDANQE -
MENT REPORTS
• MEETING MINUTES OR
NOTES ON URGENCY
Figures.
Planned Removal and/or Planned Remedial Action
Documentation must be compiled to show that the action was
within the category of those permitted by sections 101(23) and
(24) of CERCLA defining removal and remedial actions, respec-
tively. The documents must show that the action was not incon-
sistent with the NCP is performed by a governmental agency or was
consistent with the NCP is performed by a third party. The docu-
ments must show that the action was a cost effective response.
Supporting documents include environmental information docu-
ments prepared by contractors which describe the technical feas-
ibility for various remedial approaches and the probable costs
associated with those approaches. Other documents include the ac-
tual use and costs of similar response actions at other sites, and the
balance of the fund in light of these other response actions.
The degree of endangerment and the imminence or urgency of a
release need to be documented in order to justify the response ac-
tion. Expenditures for this phase, including costs, bills and con-
tractor tasks for inspections and response design must also be docu-
mented. These studies, reports, letters, memoranda notes and
meeting minutes will provide documentation for this phase as well
as the response actions for planned removals and/or planned re-
medial action.
-------
DATA BASE
371
PHASE V—PLANNED REMOVAL
Planned removal is not unlike the immediate removal phase
from the standpoint of the documentation that is compiled to
justify the cost recovery action. Several records necessary to jus-
tify a planned removal include: documentation of the threat of
direct contact with hazardous substances by a nearby popula-
tion, chemical analyses showing contaminated drinking water at
the tap, photographs, chemical analyses, and inventories of haz-
ardous substances in bulk containers that are in such condition
and orientation to pose a serious threat, chemical analyses, photo-
graphs, and reports that document the existence of contaminated
surface soils that pose a threat to the public health, welfare or
environment, records and tests that document a serious threat of
fire or explosion, and, description of weather conditions that may
enhance migration of the hazardous substances. These facts will be
documented by the cost and scientific/engineering records.
As during the immediate removal phase, documentation of the
action taken and the costs associated with the action need to be
compiled. Documents involving the planning, legal, fiscal, eco-
nomic, engineering, architectural, and other studies or investiga-
tions necessary or appropriate to plan, direct, conduct, and en-
force the response action are needed.
PHASE VI—REMEDIAL ACTION
Remedial actions are those consistent with permanent remedy
that are taken instead of or in addition to removal actions. The
purpose of the remedial action is to prevent, minimize or mitigate
the release of a hazardous substance into the environment so that
it does not migrate to cause a hazard to present or future pub-
lic health, welfare or the environment.
During this phase, documentation will be compiled that assesses
a limited number of alternative remedial measures and investi-
gates the feasibility of employing those measures. Three major
types of remedial measures will be considered; initial remedial
measures, source control measures, and off-site measures.
An investigation may be initiated to gather the facts outlined
under each major type of planned remedial action. This investiga-
tion will seek to assess whether the release can be controlled at or
near the source or whether off-site remedies will also be necessary.
A feasibility study outlining alternative approaches will be devel-
oped. This study will assess the costs of each alternative, the
effects of the alternatives, and the availability of acceptable en-
gineering practices to achieve the desired results.
Documentation for initial remedial measures track those that
have been discussed previously for an immediate or planned re-
moval. Documentation of the consideration of such measures as
drainage ditches for effective drainage control, an alternative water
supply to eliminate contaminated tap water, capping or excava-
tion, and drum removal needs to be compiled.
Source control remedial measures will be required if inadequate
barriers exist to retard migration of a hazardous substance from
where they were originally located. Off-site measures may be re-
quired when the hazardous substances have migrated beyond the
area where the substance was originally located. Considerations
that need to be documented include: the extent of hazard which
encompasses risk to the population, amount and form of the haz-
ardous substances, hazardous properties, hydrogeological factors,
and climate; extent of migration or natural containment; historical
experiences with the remedial measures at other sites; and, environ-
mental effects and welfare concerns.
CONCLUSIONS
Documentation for the various phases of the response and par-
allel cost recovery action must be compiled at each step or phase
of the action. The overwhelming number of documents discussed
in this paper can not be effectively compiled at the end of the
response action in preparation for a cost recovery action. Docu-
ments are lost, and memories lapse. The site file must be com-
piled sequentially in order to provide efficient document control.
Use of the" management system outlined in this paper will provide
for effective document control in support of a cost recovery ac-
tion under CERCLA.
REFERENCES
1. Public Law 95-510, 94 STAT. 2767-2811, Dec. 11, 1980.
2. NEIC Procedures Manual for the Evidence Audit of Enforcement In-
vestigations by Contractor Evidence Audit Teams, USEPA National
Enforcement Investigations Center, Denver, Colorado, Sept. 1981.
3. Subpart F—Hazardous Substance Response, National Oil and Haz-
ardous Substances Contingency Plan, Federal Register 47, No. 137,
July 16, 1982/Rules and Regulations, 31213-31218.
4. Procedures for Identifying Responsible Parties for Uncontrolled Haz-
ardous Waste Sites—Superfund, USEPA, Office of Legal and Enforce-
ment Counsel, Washington, D.C., Feb. 1982.
-------
DEVELOPMENT OF A FRAMEWORK FOR EVALUATING
COST-EFFECTIVENESS OF REMEDIAL ACTIONS AT
UNCONTROLLED HAZARDOUS WASTE SITES
ANN E. ST. CLAIR
MICHAEL H. MC CLOSKEY
Radian Corporation
Austin, Texas
JAMES S. SHERMAN
Radian Corporation
McLean, Virginia
INTRODUCTION
Many uncontrolled hazardous waste sites across the nation are
currently being "cleaned up" or are slated for remedial action im-
plementation in the near future. Some of these remedial actions
are being funded by the parties responsible for the site, some by
state governments, and others by the federal government under
the Comprehensive Environmental Response, Compensation and
Liability Act (CERCLA).
An initial step in any remedial action plan at hazardous waste
sites is identification of potential remedial alternatives, followed
by the selection of the most appropriate one. The formality re-
quired in this identification/selection phase may be dependent on
the source of funding for the remedial action. For example,
CERCLA requires that remedial action conducted using Superfund
monies be demonstrated to be the cost-effective alternative that
adequately protects human health and the environment. Whether
state agencies or private contractors use a cost-effectiveness assess-
ment per se, some procedure is necessary to select the most ap-
propriate plan from a list of potential alternatives.
Radian Corporation, under contract to USEPA's Municipal
Environmental Research Laboratory in Cincinnati, Ohio, is devel-
oping a methodology for assessing cost-effectiveness of remedial
action alternatives. The project, initiated in June 1982, will be
completed in early 1983. In this paper, the authors discuss pro-
gress to date and current thinking relative to methodology devel-
opment. At completion, the project will provide a convenient,
accurate and readily implemented methodology for conducting
cost-effectiveness assessments that can be used by regulatory agen-
cies, decision-makers, and others involved in remedial action se-
lection.
In the first part of this paper, the authors discuss the concepts
related to cost-effectiveness and their applicability to assessment
of remedial action alternatives. After reviewing these concepts,
the development of a specific methodology for systematic, accur-
ate assessments of potential remedial action plans is outlined.
In developing a methodology for assessing cost-effectiveness, the
objectives and criteria which must be met should be well defined.
One major purpose of the cost-effectiveness framework is to pro-
mote consistency in decision-making while maintaining applic-
ability to widely varying situations. The methodology must be sim-
ple to apply, requiring only minimal instructions on its use. In this
regard, the analysis should be based on the smallest possible num-
ber of independent variables that address all relevant concerns.
Because of the nature of the problem at uncontrolled hazardous
waste sites, the methodology must not be dependent on large
amounts of information, either on site conditions or on the remed-
ial alternatives being considered. However, the method must read-
ily allow for consideration of newly obtained information if it be-
comes available. In addition, it should not be overly quantita-
tive or impose a precision on the analysis that is inconsistent with
the degree of knowledge about the problem or expected results.
Finally, the methodology should incorporate a means for deter-
mining the sensitivity of the analysis to judgements made in ap-
plying it.
RELATED APPROACHES
Risk assessment, cost/benefit analysis, cost-effectiveness analy-
sis, decision tree analysis, trade-off matrices, and sensitivity an-
alysis have all been applied to various types of alternatives eval-
uation, including assessment of environmental controls. Certain
elements of these evaluation techniques may be applicable to cost-
effectiveness assessments of remedial action plans, and they are
characterized below.
However, systematic assessment of remedial action alternatives
for uncontrolled hazardous waste sites presents challenges which
are not characteristic of other applications of these analytical ap-
proaches. Probably the most significant of these challenges is the
lack of a well-defined clean-up objective by which the remedial
alternatives can be compared on a consistent basis. USEPA de-
fines the cost-effective alternative as "the lowest cost alternative
that is technologically feasible and reliable and which effectively
mitigates and minimizes damage to and provides adequate protec-
tion of public health, welfare, or the environment". Although
this guidance promotes consideration on sites on a case-by-case
basis, it does little to pinpoint what will be considered an accep-
table level of cleanup. Further complicating the evaluation is the
fact that the ability to predict the expected results of a partic-
ular alternative is limited by several factors including:
•lack of information on waste characteristics
•lack of information on the physical system
•lack of knowledge of long-term performance of remedial tech-
nologies
•inability to quantify expected results
Remedial alternatives may .have widely varying timeframes over
which they are functional, and, in many cases, this timeframe may
be unknown due to a lack of experience with some technolo-
gies. Finally, cost-effectiveness evaluation of remedial alternatives
must incorporate a number of factors which are subjective, diffi-
cult to quantify and for which a common metric, such as dollars,
cannot be established.
Risk Assessment
A risk assessment involves the definition of the risks to the en-
vironment and human health of continued pollution from the site.
The most inexpensive remedial action that reduces the risk to an
acceptable level could be considered the most cost-effective. Tech-
niques similar to risk assessments have been used in the regulation
of pesticides and food additives.
The use of risk assessments for assessing the cost-effectiveness
of remedial action alternatives would be a straightforward tech-
nique if risk levels to the environment and populace could be de-
termined quantitatively for the remedial actons under considera-
tion. For example, if a risk assessment were being used to assess
cost-effectiveness of remedial action alternatives at a site contain-
372
-------
REMEDIAL COST
373
ing PCBs, the levels of risk to human health (or to a receptor pop-
ulation) associated with differing levels of emission of PCB-con-
taminated leachate would have to be estimated for the alterna-
tives being considered.
One alternative for remedial action at the site might involve con-
tainment of the site using a slurry wall, whereas another might con-
sist of extraction wells to intercept the contaminated plume. An es-
timate of the levels of leachate emission from the two techniques
would be required as a first step in a risk assessment. The slurry
wall would allow some leachate to enter the environment, and the
extraction wells might not intercept all of the contaminated flow.
The level of emission and its characteristics, the size and charac-
teristics of the receptor population, pollutant pathways, the types
of harm the leachate might do, the probability the leachate will
cause that harm, and the consequences of that harm are all esti-
mates that would be required to conduct a formal risk assess-
ment.
Although one might infer, from the previous discussion, that re-
lease of contaminants from a disposal site and the resulting con-
sequences are caused by a single event, this is not necessarily the
situation. Often a series of events will lead to the release, exposure
and impact of emissions from a site. To fully characterize this series
of events, a decision or event tree may be necessary. The probabil-
ity of occurrence of each event, the exposure to any release from
that event, and the impacts of that exposure must all be identified.
Thus, a risk assessment of an alternative remedial action would
actually involve the identification of several emission release scen-
arios and an assessment of each one.
The costs associated with each remedial action would then be cal-
culated and the action with the highest acceptable risk and the low-
est cost would be considered the most cost-effective. Unfortunate-
ly, the use of risk assessments has several disadvantages including:
•Detailed information about the probability of release of emis-
sions, the quantities of emissions, characteristics of emissions,
effects of emissions on population-at-risk, size of population-
at-risk, characteristics of population-at-risk, etc. must be known
so that a formal risk assessment can be conducted; or many as-
sumptions must be made to fill in data gaps.
•Developing the data or making assumptions and conducting the
formal risk assessment is time-consuming and requires extensive
knowledge by the decision-maker of risk assessment techniques
and the functional relationship between the probability of a neg-
ative consequence and the value of that consequence.
•Someone must identify and quantify the highest acceptable risk.
(Which may be difficult for a public agency in that it is similar
to placing a value on human life.)
If a series of assumptions is required to fill data gaps so that a
formal risk assessment can be conducted, it is important to have a
sensitivity analysis for documenting how changes in those assump-
tions affect the overall cost-effectiveness rankings. Because of the
nature of formal risk assessments, the conduct of such an analysis
would be tedious.
Cost Benefit Analysis
Cost/benefit analysis involves the identification of all conse-
quences of an action so that comparisons between its costs and
benefits can be made. It is most often used as a tool for allocating
public and private expenditures. In classic cost/benefit analysis, a
project is considered appropriate if its benefits are equal to or
greater than its costs. In order to make that judgement, all positive
and negative consequences of an action must be identified, meas-
ured and translated into a common standard of measure, usually
dollars. The remedial action alternative with the lowest cost/ben-
efit ratio, or the greatest difference between benefits and costs
could be considered the most cost-effective.
A distinctive trait of cost/benefit analyses is the fact that they
can be conducted on a single alternative, independent of any other
alternatives. A true cost-effectiveness analysis assesses alternatives
m relation to each other, such that it is meaningless to conduct a
cost-effectiveness analysis on one remedial action plan. The use of
a technique similar to a cost/benefit analysis would be burden-
some because of the necessity to identify and assign dollar values
to all consequences of each action. For example, there will be
benefits that accrue because of the remedial action, such as re-
duced health risk to the adjacent population, for which assign-
ing accurate dollar values would be extremely difficult.
Cost-Effectiveness Analysis
Cost-effectiveness analyses discussed in this section are related to
the studies done to develop New Source Performance Standards
for reduction of pollutants in air emissions, and wastewater dis-
charges. This type of cost-effectiveness assessment can be con-
sidered a modification of cost/benefit analysis. It usually compares
the costs of alternatives with some measure of their benefits. The
alternative that produces the most benefit for a given cost, or pro-
duces a desired benefit for the lowest cost is considered most cost-
effective.
In environmental applications, cost-effectiveness assessments
have typically been concerned with situations in which alterna-
tive treatment scenarios for a given waste stream (either air or
water) were being assessed. In these cases, the timeframe for
needed treatment, the waste stream characteristics, and the amount
of pollutant to be removed was the same for all alternatives.
Estimates of the amount of pollutant removed by each of the al-
ternatives and the associated capital and operating costs can be
made readily. With those estimates available, it is a relatively
simple matter to express cost-effectiveness as the dollars required
to remove a pound of pollutant, or the dollars required to treat a
colume of waste stream. Selection of the most cost-effective al-
ternative is then a straightforward matter of identifying the alterna-
tive with the minimum cost for pound of pollutant removed or
volume of waste stream treated.
For most situations involving uncontrolled disposal sites, the
timeframe of the various remedial action alternatives functionality
may not be the same. Furthermore, for amny of the alternatives
there may be no treatment per se, it may be difficult to quantify
the amount of pollutant that is being emitted to the environment,
or the volume of waste stream to be treated. These obstacles make
it difficult to apply cost-effectiveness analysis involving calculation
of cost per pound of pollutant removed or volume of waste stream
treated to analysis of uncontrolled disposal sites.
Trade-Off Matrix
A fourth technique for comparing alternatives and identifying
cost-effective options is the trade-off matrix. It has been used by
many different disciplines in various industrial and governmental
applications. In this method, various alternatives are rated against
each other relative to various measures of effectiveness or cost.
These ratings are made by the evaluator or a team of evaluators.
If the team of evaluators is used (for assessing remedial alterna-
tives, this team may consist of the individual responsible for the
preparation of the cost-effectiveness assessment plus associates
with expertise in hydrogeology, civil engineering, environmental
engineering, public health, etc.), then it is probably most appro-
priate to use a method of obtaining and refining judgments re-
quired from a group, such as the Delphi technique.
The use of a team of experts is particularly valuable when deal-
ing with situations in which exact knowledge is not available and,
therefore, professional judgments are needed. Use of the technique
usually involves acquiring responses from the team members (in
this case rating of the various alternatives relative to effectiveness)
and the compilation of a single response from the group responses.
A step that sometimes is inserted between the two mentioned above
is that of an interaction or feedback step. The feedback step may
occur after the team leader or head evaluator reviews the inputs
and decides that a particular input is inconsistent, unexpected, or
an extreme value. Consultation among the evaluation team is car-
ried out to clarify the situation and establish whether that individ-
ual had a special insight that would be valuable to share with the
entire team.
-------
374
REMEDIAL COST
The single response compiled from group responses can take the
form of identifying the median or the mean of the responses. The
median is closer to the true parameter than at least half of the in-
dividual estimates, whereas use of the mean places greater empha-
sis on extreme estimates.
Sensitivity Analysis
Regardless of the methodology selected for use in cost-effec-
tiveness analyses for disposal site remedial action plans, the results
of the assessment will be subject to criticism. An important con-
sideration, therefore, when selecting a methodology, is the ease
with which sensitivity analysis can be carried out to account for un-
certainty or judgments made in the process. Any methodology de-
veloped will require that judgments be made. A well conceived
methodology/sensitivity analysis combination can quickly assess
the effect of changing any of the judgments on the overall cost-
effectiveness ranking of the remedial action alternatives being con-
sidered.
o 3
0 0
ii §
?
. X CD X (D X — W
;H — & — a) — -CD — -
(DO) ~ 3 ~ tu 5. IO
£2 §.£ law a
so ^ o Ero5 S£
^5 "g "g-5 53
55 2. 5-™ CT ™ 5 o 5
x J» i .» ° ." = -S "
1 Weighhng Factois
>
(0
^
u
<'
ID
U)
Construction _
O
Operation & "
Mamtamence 30
fi>
^«
Other 5'
(Q
W)
V Cost Ratings
Level of Cleanup/
Isolation Achievable
Time to Achieve o
Clean-up/Isolation *
A
Technology Status j^-
m
Usability of Land 3;
After Action J?
**
Capability of Action to <'
Minimize Community Impacts 3
During Implementation Jj
V)
Capability of Action to _
Minimize Adverse Health & j
Environmental Impacts o>
During Implementation J2
v Effectiveness Ratings
S Effectiveness Ratings
v Cost Ratings
Table 1.
Example Trade-Off Matrix for Cost-Effectiveness Assessment
of Remedial Actions it Uncontrolled Hazardous Waste Sites
-------
REMEDIAL COST
375
Methodology Framework
Based on an examination of the advantages and limitations of
the techniques described above, some form of trade-off matrix ap-
proach appears to offer the best means of addressing the objec-
tives and problems associated with evaluation of cost-effectiveness
of remedial alternatives at uncontrolled hazardous waste sites. One
advantage of the technique is the fact that the trade-off matrix
approach requires the least "reeducation" of the evaluator, be-
cause it requires only that he rate various alternatives based on his
knowledge of the situation.
The trade-off matrix, completed by a team of experts and in con-
junction with a sensitivity analysis, is the methodology that ap-
pears best overcomes the disadvantages inherent in the other tech-
niques. It is as useable for situations in which few data are avail-
able as in cases where extensive background data exist. In fact, it
can be used at more than one point in the project to update assess-
ments as more data become available. For example, if new infor-
mation becomes available which would change some ratings in the
matrix, the rankings of alternatives can be easily re-evaluated to
measure the sensitivity of the rankings to those changes. Finally,
the technique does not require that a common measure be placed
on all benefits, so that in essence a comparison of "apples" and
"oranges" can be made based on experts' judgments.
METHODOLOGY FRAMEWORK
In developing a trade-off matrix for cost-effectiveness evalua-
tion of remedial alternatives, the elements of the matrix must be
selected carefully in order to ensure that the evaluation considers,
all relevant concerns. However, equal care must be taken to avoid
selection of evaluation criteria that are not independent, such that a
specific consideration can affect ratings for several of the criteria,
possibly creating a significant bias in the analysis. The matrix
should include both cost and "effectiveness" measures against
which each of the alternatives under consideration can be rated or
scored. An example of a trade-off matrix that might be used in
assessing cost-effectiveness of remedial alternatives is given in
Table 1.
Cost Measures
Evaluation of remedial action alternatives at most sites will be
conducted before detailed design information or cost data are
available. However, because of the importance of cost in the eval-
uation, it is necessary to develop cost estimates of sufficient ac-
curacy for comparison of alternatives. The trade-off matrix in-
cludes construction cost and operation and maintenance costs as
separate evaluation criteria.
In addition, other cost factors can be considered if site-spe-
cific considerations warrant. For example, in some cases it may be
appropriate to include consideration of the effect on property
values as a measure of cost of various alternatives.
Generic Effectiveness Measures
Certain criteria for assessing "effectiveness" of a remedial plan
can be considered applicable to the large majority of situations.
These generic effectiveness measures are identified in the example
trade-off matrix. These measures have been selected to address
the range of concerns which must be considered in selecting a re-
medial plan, including:
•How clean will the site be after completion?
•How long will cleanup take?
•How well will the remedial action reduce health and environ-
mental impacts?
•What are the impacts of performing the remedial action?
•Can the site be used after cleanup?
•Is the technology proven and feasible?
•What is the risk of failure?
•What would be the impacts of failure?
A cost-effectiveness assessment can be performed for most sites
using onjy these cost and generic effectiveness measures. How-
ever, the methodology is designed to allow incorporation of addi-
tional criteria to address specific conditions dictated by a partic-
ular case.
Site-Specific Effectiveness Measures
In certain cases, there may be factors dictated by conditions or
clean-up objectives unique to a specific site which are considered
significant enough to be used as distinct effectiveness measures.
For example, if a waste site has resulted in severe impacts to a
downstream lake and cleanup of the lake is considered to be a
major objective of the remedial action, the degree to which the al-
ternatives mitigate impacts to the lake may be added as a measure
of effectiveness. This measure may be used in addition to the gen-
eric factors or may substitute for one or more of them.
Weighting Factors
The trade-off matrix technique enables the user to conveniently
account for the fact that some measures of effectiveness are more
important than others. This can be done through the use of weight-
ing factors that are applied to each of the measures of effectiveness.
These weighting factors typically range from 0.1 to 1.0 and will
be multiplied by the ratings that each alternative has relative to each
measure of effectiveness.
EVALUATION PROCEDURE
There are several distinct steps involved in conducting the cost-
effectiveness assessment using the method outlined above. These
steps are:
1. Compile background data on the site.
The first effort for any assessment of remedial actions involves
collecting information about the site. This information should in-
clude descriptions of the types and volumes of waste disposed,
geologic aspects of the site (e.g., depth to groundwater, soil types,
lithology and depth to confining stratus), and any other data that
may exist. This information will be needed when defining appro-
priate remedial action alternatives and in building a data base to
aid in judging the effectiveness of the alternatives.
2. Identify appropriate remedial actions.
Based on the outputs of Step 1, viable remedial action alterna-
tives are identified. There is a possibility that some or most of these
alternatives will have been defined before the evaluator becomes in-
volved in the assessment process.
3. Identify appropriate measures of effectiveness.
The measures against which the various alternatives will be assessed
are then selected by the evaluator. The generic measures, because
of their nature, will probably be identified for the evaluator. The
site-specific measures will depend on the site characteristics, waste
type, surrounding environment, and alternatives being considered.
4. Assign weighting factors.
The relative significance of all the measures of cost and effective-
ness is established, typically by a concensus of the evaluators.
Appropriate weighting factors are then assigned to the various
measures to reflect their importance.
5. Rate alternatives.
Each alternative is rated (e.g., 1 to 5) relative to its ability to satisfy
each measure of effectiveness. The evaluators will depend primar-
ily on their expertise and knowledge of the situation to make appro-
priate judgments. Additional technical information can be pro-
vided to the evaluation team prior to the rating exercise.
6. Calculate final ratings.
The final rating for each measure of effectiveness for each alterna-
tive is then calculated by multiplying the ratings by the appropriate
weighting factor. This step enables the evaluator to factor the rela-
tive importance of the various measures of effectiveness into the
analysis.
7. Sum the final ratings for each alternative.
The sum of all the effectiveness ratings for each alternative is cal-
culated. This step involves summing all the final ratings in a row,
each row corresponding to a different alternative (Fig. 1).
-------
376
REMEDIAL COST
8. Estimate costs.
The costs for construction, operation and maintenance, and any
other identifiable costs, are estimated. These estimates may be
developed independently by the evaluator or they mayt be derived
from available engineering studies or standard references.
9. Calculate cost ratings.
Cost ratings are calculated by expressing the cost estimates in
millions of dollars (e.g., a cost estimate of $753,000 has a corres-
ponding cost rating of 0.753). Appropriate weighting factors may
be applied as required. Ratings for each cost measure (e.g., con-
struction, and operation and maintenance) should be calculated.
10. Sum the cost ratings.
Cost ratings for each alternative are summed, as was done for the
effectiveness ratings.
11. Calculate cost-effectiveness "scores".
The cost-effectiveness "score" for each alternative is calculated by
dividing the sum of the effectiveness measures (Step 7) by the sum
of the cost ratings (Sept 10) for each alternative. The results of
this step provide the final overall cost-effectiveness score for each
alternative. The alternatives with the highest scores are those that
are most cost-effective.
12. Recalculate trade-off matrix to check sensitivity.
There will be instances in which it will be desirable to assess how the
overall cost-effectiveness rankings change if certain elements with-
in the trade-off matrix change. A recalculation of the affected
scores will give an indication of the sensitivity of those scores to the
changes. Elements of the trade-off matrix that may change are
costs, individual ratings or weighting factors. These changes may
be induced by an increased knowledge on the part of the evaluator
or member of the evaluation team, different assumptions, or the
investigation of the sensitivity of changes may be necessary to
alleviate concerns of members of the public. The use of such a
sensitivity analysis to assess the changes in overall cost-effective-
ness rankings of the alternatives should be considered an essential
element of the methodology.
Selection of Preferred Alternatives
After all remedial alternatives have been rated by each member
of the evaluation team, there are several procedures available to
identify the most cost-effective alternative. The mean or median
ratings for each effectiveness measure can be identified and then
those values summed for each alternative. Or the individual rat-
ings for each member of the evaluating team can be summed, from
which the mean or median sums can be determined for each altern-
ative. Either method will result in a composite effectiveness score
for each alternative which is then divided by the sum of the cost
ratings for each alternative to calculate a composite cost-effec-
tiveness score.
In principle, the alternative with the highest score can be con-
sidered most cost-effective. Actually, however, it is probably
worthwhile to consider several of the highest-ranking alternatives
as equivalent based on cost-effectiveness assessments, particularly
if the numerical scores are almost equal.
CONCLUSIONS
The trade-off matrix offers a flexible but valid approach to
cost-effectiveness evaluation of remedial action alternatives at un-
controlled hazardous waste sites. The techniques can be used in
widely varying situations but can be tailored to meet site-specific
needs. Because of its simplicity, the technique is not so time-con-
suming as to preclude examination of a relatively large number of
alternatives. The approach is easily understood by non-technical
persons and does not require extensive evaluator education. By re-
lying heavily on the judgments of experts, the method is not de-
pendent on extensive site data although a good data base un-
doubtedly improves the confidence level of the evaluation. In
addition, the methodology readily allows examination of the sensi-
tivity of the analysis to judgment or assumptions made in the eval-
uation process.
A possible drawback to this approach to cost-effectiveness eval-
uation is that the technique is essentially subjective and therefore
may be subject to evaluator bias. However, care in selection of the
evaluation team and use of the Delphi technique to deal with olut-
liers in the ratings can minimize this problem. In addition, the
method is not mathematically rigorous which may affect its cred-
ibility in some cases. However, in general the information base
and prediction capability supporting decisions about remedial ac-
tions are insufficient to warrant rigorous quantitative evaluation.
BIBLIOGRAPHY
1. Cost Effectiveness Analysis Guidelines, 40 CFR Part 35, September
4, 1973.
2. Dower, R.C. and Maldonado, An Overview: Assessing the Benefits
of Environmental, Health and Safety Regulations, prepared for the
U.S. Regulatory Council, May 1981.
3. Engineering Science Inc., Water Quality Management Planning Meth-
odology for Municipal Waste TReatment Needs Assessment, pre-
pared for Texas Department of Water Resources, March 1977.
4. ICF Incorporated, RCRA Risk/Cost Policy Model Project, Phase 2
Report, Office of Solid Waste, USEPA, June 15, 1982.
5. Kufe, C. and Kilpatrick, M., Rating the Hazard Potential of Waste
Disposal Facilities, JRB Associates and USEPA.
6. USEPA, Municipal Environmental Research Laboratory, Handbook,
Remedial Action at Waste Disposal Sites, EPA-62516-82-006, June
1982.
7. National Economic Research Associates, Inc., The Business Round-
table—Air Quality Project—Cost-Effectiveness and Cost-Benefit
Analysis of Air Quality Regulation, IV, New York, N.Y., Nov. 1980.
8. Pound, C.E., Crites, R.W. and Griffes, D.A., Costs of Wastewater
Treatment by Land Application, EPA-403/9-75-003, June 1975.
9. Proc. of National Conference on Risk and Decision Analysis for
Hazardous Waste Disposal, Baltimore, Maryland, Aug. 1981.
10. Radian Corporation, Cost-Effectiveness of New Source Performance
Standards, prepared for Office of Environment, U.S. Department of
Energy, Jan. 1981.
11. SCS Engineers, Cost-Effectiveness of Remedial Actions at Uncon-
trolled Hazardous Waste Sites, prepared for USEPA, Solid and Haz-
ardous Waste Research Division, MERL, July 1981.
12. Unites, Dennis, Possidento, Mark and Housman, John, Preliminary
Risk Evaluation for Suspected Hazardous Waste Disposal Site in
Connecticut, TRC, Connecticut 208 Program, and Connecticut De-
partment of Environmental Protection.
-------
NEGOTIATIONS: THE KEY TO COST SAVINGS
WILLIAM R. ADAMS, JR.
Perkins Jordan, Inc.
Reading, Massachusetts
INTRODUCTION
The stakes, both legal and financial, are high in a Superfund or
hazardous waste study or cleanup project. The consulting engineer-
ing profession has long prided itself on its ability to develop sound
engineering and technical solutions which are cost-effective. A con-
sulting engineer or scientist often spends days and sometimes weeks
in performing calculations and evaluating alternatives to determine
the most cost-effective technical solution to a problem. However,
the profession has not been as aggressive in sharpening its skills to
deal with the public, political figures, and agencies to assure that
the investigative studies, work plans and remedies that may be im-
posed on a client are reasonable and necessary.
The author believes that proper evaluation and negotiation of the
scope of technical studies can provide cost savings of the same
order of magnitude that a thorough technical evaluation can pro-
duce. In this paper, he provides strategies and guidance to assure
that a client is properly represented in the negotiations of an accep-
table proposal or work plan as well as in the actual technical evalua-
tion of a project.
Proposed Problem
Assume that a hazardous waste problem or site has been
detected. This could be the discovery of an abandoned disposal
area or the determination that past industrial practices have created
problems at the plant site. Contamination of the environment has
been detected and a regulatory agency, either the USEPA or the
State Environmental Agency has determined that an engineering
evaluation and potential corrective action is necessary. This deter-
mination to take action at a particular site may, or may not, have
been influenced by the keen interest of a citizen group in the area,
and/or the interest and attention of political leaders, at the local,
state and perhaps federal levels. Usually, the generator of the waste
has been identified, has been notified he is being held responsible
and has indicated a desire to conduct reasonable investigations and
possibly cleanup activities.
The proposed investigation and potential cleanup will be super-
vised and finally approved by a government agency. The agency
will determine the nature, extent and thoroughness of the assess-
ment, investigation, the need for cleanup, and the type of remedial
action to be implemented. The agency will also be responsible for
answering the ultimate question associated with site cleanup:
"How clean is clean?"
The potential for a classic confrontation now exists. On one side
is a government agency that is demanding a proper and complete
work plan. An agency which is driven by laws and its own regula-
tions and which must be responsive to political and public
pressures. On the other side is a company who recognizes the need
for a responsible work plan but is being driven in the opposite
direction by costs and other concerns.
Reaching agreement may appear hopeless, but upon careful ex-
amination, the situation actually provides a favorable climate for
.negotiations. An element present that both the governmental agen-
cy and the client desire, is an efficient and effective work plan
which solves the problem at hand. The consulting engineer must
have the confidence of both parties if he is to build from the com-
mon element a study which will meet the needs of his client as well
as the agency.
To properly protect a client and to establish a technically ade-
quate program, exclusive of unreasonable requirements, the con-
sulting engineer must understand the regulatory agency, its people,
and its pressures, i.e., an agency adopts its own regulations and re-
quirements but the implementation and interpretation of these
regulations is accomplished by staff who frequently are motivated
by the political community and the public at large. The agency en-
visions itself as the public's protector and wants to be seen as tough
but fair.
In the case study proposed here, the cost for the investigation
does not come from the agency's budget and hence there may be a
tendency to require more, instead of less, work. Conversely, the
client may, with an eye on costs, and perhaps the fear of uncover-
ing damaging information, tend to propose less, instead of more,
work. Moreover, in most regulatory agencies, the costs for pursu-
ing litigation often come from departmental budgets outside of the
agency or at least a section of the budget not controlled by technical
personnel and can be viewed by some as a beneficial option—either
as an example to others or as a measure of strength.
Negotiations can only be successful if a common goal exists, i.e.
to design a reasonable work plan, and each party respects the other.
This may be an oversimplification, but it will be used in this paper.
The key to successful negotiation depends, in large'part then, not
only on the consultant's technical expertise, but also upon his abili-
ty to deal with the agency personnel. On the basis of considerable
personal experience in government service it is the author's opinion
that the general public and consulting engineers, in particular, may
tend to stereotype government employees as being difficult to deal
with, unreasonable and sometimes unmotivated to solve problems.
Some government employees fit that description, but the author's
service in government has convinced him the opposite is true. Most
government employees are intelligent and motivated; they are in-
terested in solving problems and in seeing positive results; they are
motivated'by the same issues that move people in the private sector;
they are anxious to excel, develop professionally, and to bring
about successful solutions to problems. However, frequently they
lack hands-on technical experience. Most consultants have exten-
sive experience and therefore must fully utilize this quality.
A hypothetical problem has been presented and the basic
qualities of the agency staff who are responsible for its solution
have been outlined. Based on this scenario, a procedure to suc-
cessfully manage the negotiations required for the project will now
be developed. Manage is the key word and will be the principal
thrust of this paper in which the author assumes that the consulting
engineer has all of the basic skills and experience necessary to effect
a successful remedy; therefore the author treats the aspects of
negotiations rather than the technical details of the solution. It is
also assumed that the agency personnel are reasonable and anxious
to find an acceptable solution.
377
-------
378
REMEDIAL COST
Before the consulting engineer can discuss a project with the
regulatory agency, he must have the confidence of his client and
should be a true partner in the effort to find a solution. He must
understand any limits on the client's participation in the project
and the latitude given to him to negotiate on the client's behalf. He
must also have a sound understanding of the legal liabilities and
ramifications of the case.
A five step procedure is proposed for negotiating with the
regulatory agency. These steps can represent five separate meetings
or can be consolidated into one meeting depending upon the
magnitude of the problem and the complexity of the situation.
Depending upon the dynamics of the situation, various elements of
these steps can be interchanged or omitted. The five basic steps are:
•Preliminary project evaluation
•Initial agency meeting
•Presentation of a proposal
•Reception of agency feedback
•Final negotiations
PRELIMINARY PROJECT EVALUATION
Before meeting with the agency, the engineer should do as much
investigation as possible to assure he has a complete understanding
of the site and its history. He must visit the site, review whatever
has been written about the site and discuss this information with
knowledgeable individuals, both within and outside the company.
He should make a preliminary evaluation of the extent of con-
tamination and the potential for off-site migration. It will be useful
to understand the nature of the public and political interest in the
project and the individuals who are leaders in these efforts. The
engineer should also have a sound understanding of the agency's
position including the people and the forces which will be directing
its decisions. This preliminary evaluation of the problem will form
the basis for a meeting with the agency staff.
INITIAL MEETING
On his first approach to the agency, the consultant should at-
tempt to constrain the discussions to technical issues. He must
avoid threats and posturing which would lead the negotiations
toward legal, rather than the technical solutions. He must convince
the staff that his client is seriously seeking a solution and is willing
to accept reasonable requirements.
The engineer, at this meeting, should also present his qualifica-
tions to do the work required. This should be done in a positive
manner, and in such a way that agency personnel accept his creden-
tials as a measure of experience and competence and not one of ar-
rogance. The engineer should take every opportunity to refer to
previous work he has accomplished, and to point out the similarity
of past work with the problems at hand. Gaining the agency's con-
fidence at the onset of negotiations is extremely important and
must be accomplished in a manner which is not considered
threatening to agency staff.
The consultant should solicit as much information as possible.
This should include any reports, studies, correspondence or other
documentation available as well as any on-site experience agency
personnel may have had. He should evaluate the agency's urgency
and direction. The engineer should solicit advice and attempt to ob-
tain as clear an understanding as possible of what the agency con-
siders an adequate solution to the problem. If the agency does not
articulate a strong position, it may be wise at this time, to allow the
staff to be vague to prevent the need for a change of its position
later. This meeting is not intended to make final decisions as to
what is, or is not, needed for the study. Its basic purpose is to
understand the situation, the agency and its people.
PRESENTATION OF A PROPOSAL
It is extremely important that the presentation of a proposal to
the ageno b> the engineer should be well planned and executed.
During this phase, the engineer should reinforce his previous
presentations of technical qualifications and competence, again
referencing similar successful work.
The proposal should be complete and contain sufficient detail to
allow the agency to properly evaluate it. It should include maps,
charts, and other visual aids to facilitate understanding and accep-
tance. The proposal should be structured to provide answers to the
questions the agency is asking. An effective way of evaluating the
many unknowns of a hazardous waste project is to provide in-
termediate decisions points so that both the client and agency will
have an opportunity to decide, after evaluation of data, whether
additional information is required. The program should be con-
structed so that it builds upon itself and therefore have various
stages so that information from each step can either trigger an addi-
tional step or allow an activity to cease. The engineer should be
prepared to give detailed explanations of his proposal, the technical
reasons for each decision and the choices available at each in-
termediate decision point.
After the presentation there should be time allowed for questions
to clarify the engineer's intent. Evaluation and judgment is best
delayed, to permit the agency to conduct a comprehensive and
thoughtful review. The agency should be given sufficient time for
proper review, but a deadline for comments should be established
at the onset. The engineer's willingness to clarify points over the
telephone should be expressed.
AGENCY FEEDBACK
After the agency has had an opportunity to carefully review the
proposal, a meeting should be arranged to discuss and evaluate the
agency's comments. Recommended additions or deletions to the
proposal which will not impact technical viability or costs should be
quickly accepted and incorporated into the proposal. If the agency
drastically alters the proposal, or demands additional costly re-
quirements, the engineer must determine the reasonableness of
these modifications.
It is essential that the engineer listen carefully to the reasons
given for these changes and show respect and understanding for the
agency's position. He should also attempt to evaluate the technical
background of the person proposing the change. Skillful handling
of this situation can avoid a confrontation, or the need for a face-
saving concession. Quiet, objective listening at this point will
facilitate the negotiations which must follow. It may be wise to ac-
cept the agency's comments for further study and postpone
substantial discussion to the negotiation session which is to follow.
FINAL NEGOTIATIONS
The entire thrust of this meeting is to build cooperation instead
of irritation. The engineer must approach the agency with a feeling
of confidence in his own abilities and respect for the abilities of
agency personnel. Each issue raised by the agency must be ad-
dressed. It is frequently beneficial to refer to simple visual aids such
as maps, schedules and schematics, to assure that agency personnel
have a full understanding of the engineer's plan. The engineer must
be certain that he understands why the issue was raised and how
deeply committed the individual or agency is to the point. Working
on a non-negotiable issue is senseless.
A schedule of issues ranging from the simplest to the most dif-
ficult is often advantageous so it may be wise to reorder the agenda
so that non-controversial and inexpensive issues are discussed first.
This builds understanding and confidence in both the agency staff
and engineer in solving problems. It develops a sense of com-
promise, and allows the engineer to educate agency personnel in
areas where technical weaknesses may have been observed.
As these issues are resolved, the engineer should also develop a
sensitivity that will allow him to anticipate and address questions
and concerns before they are raised. By working from the easiest to
the most difficult issues, agency personnel will acquire a better
understanding for the sophistication and experience of the
engineer. Dangers in ordering the negotiations to address the
simplest problems first, include the risk of mounting frustrations
on both sides and using up time scheduled for the negotiation ses-
-------
REMEDIAL COST
379
sion before the important issues are addressed. Thus, the engineer
should maintain sufficient control over the session to prevent either
or these situations from occurring.
In resolving the most important conflicts, care must be taken to
assure the agency personnel that their position is worthy of discus-
sion and will not be dismissed out of hand. If major conflicts are
identified, the engineer is responsible for presenting his arguments
and reasoning in a logical and technically sound manner. Hopefully
the basic premise that both sides desire a reasonable, workable,
cost-efficient plan, will prevail. If, however, the engineer is unable
to convince the agency that his proposal is adequate, then the fall-
back position of accomplishing additional work in increments, with
intermediate decision points, should be utilized. If possible, objec-
tive and measurable criteria for a "go" or "no go" decision should
be established at this time.
In summary, to effectively deal with an agency it is necessary that
the engineer: know the site, its history, and the problems envision-
ed by the agency, understand the agency's motivation and the
pressures that it is experiencing, understand and respect individual
staff members, their needs, experience and motivation, be compe-
tent and be thoroughly experienced in the area of discussion, and
prepare a plan which recognizes the problem and provides the
means to find a solution.
The engineer's proposal should be complete and should include
as many of the agency's concerns as are reasonable. The plan of
study should be methodical and developed to allow each step to
build on the previous steps, thereby allowing the engineer and agen-
cy to curtail or abandon investigations which are unnecessary. The
plan must provide intermediate decision points to allow agency
staff sufficient flexibility to accommodate unknowns.
The strategy described here has been greatly simplified in the in-
terest of time and space. Obviously, there is much more to suc-
cessful negotiation than has been discussed. Any professional per-
son, with the experience and background which qualifies him for
such important negotiations, has learned a great deal over the
years. He is skilled in anticipating concerns, interpreting people's
reactions and body language, and exerting the self-control and
discipline needed in any bargaining session.
CONCLUSIONS
How effective will the scheme described here be? Let me recount
a particular situation that I have been involved with.
Perkins Jordan, Inc. was requested by a large industrial firm to
represent them at a particular hazardous waste site. The agencies
involved had determined that hazardous waste contamination ex-
isted, had notified the company it was responsible and had
established an overall budget of nearly $650,000 for the necessary
engineering investigation. In following the procedures previously
outlined, with emphasis on the collection of available information,
the need to become thoroughly familiar with the site, we were able
to develop an investigative study to evaluate the situation.
The study was divided into several phases. For Phase I, which
had been previously estimated to cost over $208,000, an investiga-
tion, basically using the same rates, but costing less than $104,000
was proposed. Subsequent negotiations between the engineer and
the agency, as described in this paper, resulted in additional work
being added to the original proposal, bringing the total cost to ap-
proximately $117,000. The agencies responsible for this project
agreed that the proposal, as finally negotiated, was adequate and
would fulfill their needs. Though this project did not proceed
because of reasons unrelated t»' those discussed here, it is an ex-
cellent example of how sound engineering, skillful proposal
preparation and negotiation can produce substantial savings and
satisfy the needs of all concerned.
-------
APPLICATION OF ENVIRONMENTAL RISK TECHNIQUES TO
UNCONTROLLED HAZARDOUS WASTE SITES
BARNEY W. CORNABY, Ph.D.
KENNETH M. DUKE, Ph.D.
L. BARRY GOSS, Ph.D.
JOHN T. MC GINNIS, Ph.D.
Battdle's Columbus Laboratories
Columbus, Ohio
INTRODUCTION
Many problems have been created by inadequate disposal of
hazardous wastes: literally thousands of disposal sites in the U.S.
and other countries have been filled with waste materials of un-
known origin and composition. Although some of these wastes
may have been carefully containerized, the containers usually de-
teriorate with time releasing the wasted materials into the environ-
ment where they may travel via groundwater, soil, and sediments
and ultimately endanger human health and ecological systems.
How is a problem diagnosed? Where does the problem exist?
What is being affected—vegetation? animals? humans? other parts
of the ecological system? What actions need to be taken? Such
questions can best be answered by a methodically conducted in-
vestigation in which risk is quantified and can be compared. More
precisely, risk assessment is the process of determining the prob-
ability that certain activities will produce certain adverse effects.
These and related aspects of risk are discussed in such sources as
Whyte and Burton, " and Cornaby.' Risk assessment has become
a commonly practiced method of evaluating actions affecting
human health and safety. Numerous pieces of Federal legislation
currently require some form of risk assessment, several of them be-
ing the Clean Water Act, the Toxic Substances Control Act, and
the Resource Conservation and Recovery Act.
In this paper, the authors describe the need for an integrated
system for quantifying environmental risk. Such a scheme is pre-
sented and briefly explained. This scheme consists of the develop-
ment of three interrelated elements for human and nonhuman as-
pects: exposure assessment, effects assessment, and risk assess-
ment.
NEED FOR INTEGRATED ENVIRONMENTAL
RISK ASSESSMENT
Environmental risk assessment for uncontrolled hazardous waste
sites has three major interrelated elements: (1) exposure assess-
ment, defining the movement of source terms through their ulti-
mate environmental fate, (2) effects assessment, defining the im-
pacts associated with varying levels of stressors for both humans
and nonhumans, and (3) quantification of risk, defining the prob-
ability that certain activities will produce certain adverse environ-
mental effects (Fig. 1). The existing methodologies for exposure
and effects assessments are reasonably well advanced and include
a number of complex models.2'8'10'11'14
Effects assessment is less well conceptualized especially at higher
levels of biological organization (communities and ecosystems).
Adequate quantification of risk techniques is virtually nonexistent.
The paramount concern is that no known general methodology is
available for conducting overall environmental risk assessment,
risk assessment that includes both humans and especially non-
human or ecological receptors. In fact, the terminology in the liter-
ature is noi always clear, suggesting that our views and knowledge
of environmental risk assessment are still evolving.
NEW SCHEME FOR ENVIRONMENTAL RISK
A scheme that integrates human and nonhuman or ecological
aspects of environmental concerns has been developed. Ecologi-
cal and human risk assessment requires the categorization and
definition of receptors and possible stressors in a manner so that an
analysis can be logically pursued to a definitive end and yield values
which are computable and comparable. An approach to organiza-
tional aspects of environmental risk assessment (including health
and ecological concerns) is shown in Fig. 2. The overall con-
ceptual approach expands on the three major elements of environ-
mental risk assessment (exposure assessment, effects assessment,
and quantification of risk in Fig. 1) to more completely define
the major components and the interactions among them neces-
sary for environmental risk assessments.
The fate information is used quantitatively and directly (as sug-
gested by the solid line) in the exposure/hazard adjustment step
of the integration of risk. Also, fate information is used qualita-
tively (as suggested by the broken line) to guide the environmental
classification step in the effects assessment component. Other solid
lines mean that the results of definitive data collection and analy-
sis are transferred to the next step.
Not all hazardous waste pollutants need to undergo risk assess-
ment. For pollutants with legal standards it is a matter of com-
paring the source term, exposure, or other measurement with the
standard.4 If the standard is exceeded, no further risk research
would be needed unless there was a reason to contest the stand-
ard. If there is no standard, the conduct of a risk assessment is
desirable to determine the levels of risk associated with various al-
ternative actions faced by managers of uncontrolled hazardous
waste site(s).'
Exposure Assessment
There are three interrelated steps necessary to evaluate exposure
assessment: (1) source, (2) transport, and (3) fate.
EXPOSURE ASSESSMENT
Defines Source Terms
Through Their Ultimate
Environmental Fate
EFFECTS ASSESSMENT
Defines the Impacts
Associated with Varying
Levels of Stressors
QUANTIFICATION OF RISK
Defines the Probabilities
That Certain Activities Will
Produce Certain Adverse
Environmental Effects
Figure 1.
Overview of the Three Basic Elements of Environmental Risk Assessment
380
-------
RISK/DECISION 381
Aquatic
'
_
Environmental
Classification
•
Kev
Selection
Recovery
i
Dose
Response
I
{Terrestrial I '
1 /
Food Chain
I
I
I
Demographics
Dose
Response
Figure 2.
Overall Environmental Risk Assessment Scheme
Regarding source, the location, type of release, and rate of re-
lease are all important aspects of this characterization. An uncon-
trolled hazardous waste site in the middle of a large city would pre-
sent a risk different from the same facility located in a remote for-
est.- The type of release also determines the type of subsequent
work that would follow. For example, a situation of an occasional
release of a compound of low toxicity dictates an approach dif-
ferent from a situation where highly toxic materials are being re-
leased continuously into not only the surface water and air but also
the groundwater. Of course, the quality of pollutant and the rate at
which it leaves the source are other important aspects of the source
characterization.
Pollutant releases from hazardous waste sites can range from
those resulting from explosions and accidents to those associated
with day-to-day operation and maintenance. Sources can be further
classified by size of the site, type of hazardous material, soil char-
acteristics and efficiency of control systems.
The transport or movement of a pollutant from its source
through the environment to a receptor can be in the air, water,
and/or soil. The pollutant can also move from one medium to an-
other. For example, a pollutant can first be released into and trans-
ported by the groundwater, only to be used on land (e.g., irriga-
tion) and subsequently transported back to the groundwater or
evaporated into the air. Many of these interactions are discussed
in such sources as Home.'
Because hazardous waste sites can result in transport-through-
soil problems, the narrative will focus on soil. Transport through
soil is the least understood of the three media. Nevertheless, there
are processes about which some knowledge is available. These pro-
cesses are leaching, percolation chemistry, percolation volume and
time, groundwater movement, percolation chemistry, percolation
volume and time, groundwater movement, percolation mixing, and
groundwater chemistry. Models of movement of materials in soil
range from the relatively simple to the relatively complex.7 For ex-
ample, the least complicated models assume one-dimensional
movement relative to a general water table, slope, gradient, and
soil permeability whereas the most complicated are time-dependent
and use three dimensions. In these models,, the soil is organized
mto spatial and temporal boxes, each with input/output functions.
More research to improve models of material movement in soil
and especially to expand our knowledge of the reactivity of pollu-
tants with soil per se will result in improved transport models for
the soil medium. Bloom et al.' discuss other transport models.
Another important aspect of transport involves transformation
models. Transformation models focus on complex changes in
molecules as they move in the environment. These models are more
difficult to organize because of the requirement for knowledge
about these complex changes in molecules, especially organic mole-
cules. This lack of data (as opposed to lack of mathematical prow-
ess) is a stumbling block, especially for pollutants that interact
with carbon-based substrates. The simplest modeling approach to
the question of transformation of pollutants is to bypass it, by lim-
iting the model to materials where transformation is not impor-
tant, e.g., heavy metals. For materials where chemical or other
transformation cannot be ignored, the simplest approach is to as-
sume a constant time rate of transformation. If this is not desir-
able, reliance needs to be placed on complex models which in-
corporate what is known about transformation. These models can
require a considerable amount of information about chemical re-
action times and media conditions. Transformation is a difficult
step to complete, especially for organic molecules that react with
many other substances.
The third step in exposure assessment is the determination of
the ultimate fate of the hazardous material. Fate refers to the: (1)
final environmental concentration, (2) geographical area exposed,
and (3) duration of the exposure for the pollutant(s). The first ac-
tivity in the fate component is to determine the estimated environ-
mental concentration (EEC) of the chemical at its point of contact
with receptors. Knowledge of the amount and volume of chemical
being discharged, whether the discharge is point source, and the
discharge rate are available from the step on source and must be in-
tegrated with information on migration and dilution, removal from
original medium to alternate media, and chemical or biochemical
transformations.
Two additional aspects of fate are the geographical area exposed
and the duration of the exposure. Geographical area means the dis-
tribution of the EEC through space while duration means the distri-
bution of the EEC through time. A small geographical aiea of a
-------
382
RISK/DECISION
few square meters presents a different problem from an area of
many square kilometers. Likewise, duration is an important aspect
of the risk assessment process because a one-year exposure could
lead to greater environmental degradation than a one-day exposure
at a greater concentration.
After the EEC for a pollutant(s) has its spatial and temporal
boundaries defined, the fate data can be used as a guide for dose
range in the effects assessment element. For example, mercury
once transferred from its source and transformed to methyl mer-
cury becomes dangerous to organisms at certain environmental
concentrations." Chromium undergoes valence changes and many
organic pollutants are transformed as they move through the en-
vironment and become more or less toxic.4 The fate of the pollu-
tant provides clues about the nature, extent, and magnitude of the
possible biological effects. But it is the next component—ecologi-
cal and human effects assessment—where the fate data are used in
this way.
ECOLOGICAL EFFECTS ASSESSMENT
There are three basic steps necessary to evaluate ecological ef-
fects of uncontrolled hazardous waste sites: (1) environmental
classification, (2) key receptor selection, and (3) dose-response re-
lationship. Each step is completed in proper sequence. The com-
pletion of the effects evaluation will result in the inputs needed for
the actual quantification of environmental risk of the pollutant(s)
of concern. The following narrative defines the variables and dis-
cusses the approach for each step in the ecological effects eval-
uation process.
By knowing the geographical location and areal extent of ef-
fluents released from uncontrolled hazardous waste sites, the types
of environments exposed can be identified. Within each environ-
ment the actual ecosystems (terrestrial and aquatic) affected can
also be determined. Classifications for habitats are numerous and
are avajlable in the work of Odurn.12'13
Once the geographical area which will be exposed has been clas-
sified into one or more of various types of environments, it is nec-
essary to develop weights for these environments; this will facili-
tate the integration of risk (a step later in this integrated scheme).
The weighting needs to be done from the human-welfare or value-
to-man viewpoint. The value of collective ecological attributes of
each affected environment is expressed on an appropriate scale.
Attributes to be considered in the development of weights include
not only agricultural productivity, recreational resources, and
commercial fisheries, but also the ability of the systerh to assimilate
and process wastes, protect threatened species, and act as a life-
support system for man (e.g., oxygen generation by green plants).
Weighting is best accomplished through a review of the specific
component ecosystems of each environmental type and the values
associated with each component system tallied and then summed.
The components of each environmental type which will be ex-
posed must be identified. In contrast to human risk assessment,
which deals with risk to the individual and starts at the popula-
tion level and worked down to lower levels of organization (e.g.,
organs, tissues, and cells), ecological risk assessment is concerned
with effects to whole populations or organisms and must start at
the population level and work up to higher levels. Three basic
components of ecological organization are recognized: popula-
tions, communities (multiple populations), and ecosystems (com-
munities plus the nonliving environment and the processes such as
nutrient cycling which bind them together into a functioning
system).
The key receptors for the pollutant of interest must be carefully
selected for each aquatic and terrestrial ecosystem. These receptors
must be representative of the important components and processes
a! risk in each system and their importance must be scientifically
defensible. Once key member population and communities and
ecological processes at risk for each aquatic and terrestrial eco-
system are identified, appropriate indicators of effects to these key
receptors can be chosen. Indicators are used simply because it will
likely be impossible to obtain data on all the key receptors within
the time and cost constraints imposed by the risk assessment pro-
cess. In addition, data collection techniques are standardized and
validated for a relatively small number of receptors. The use of
nonvalidated protocols and populations or processes lacking a
good interpretive data base can seriously compromise ecological
effects evaluation.
The third step in ecological effects assessment is the dose-re-
sponse assessment. Dose-response assessment involves the measure-
ment of the change in ecological effect(s) in relation to the change
in concentration and duration of the stressor to which the popula-
tion or process is exposed.
Two basic approaches to dose-response data collection are recog-
nized. The first is the simultaneous approach where a full set of
tests are implemented essentially simultaneously on the pollutant(s)
of concern. This approach is thorough, usually provides highly
reliable results, but is time consuming and expensive, often includ-
ing redundancy among the test results. A second approach, called
the phased or tiered approach, uses a sequence of testing activ-
ities rather than simultaneous implementation.'
In the phased approach, each subsequent phase is designed us-
ing the results of the preceding step. Generally, short-term acute
exposures using single-species or process tests are employed at the
first tier or phase. These inexpensive procedures can provide a
coarse, quick, relative assessment of hazard." The results are used
in the design of the next phase of testing with emphasis placed
on those indicators in the first phase which showed the highest
level of hazard.
The higher levels of testing (Phase 2 and above) generally em-
ploy longer duration or chronic exposures and may use multi-
species or process techniques. Data collection procedures have four
key attributes: (1) type of endpoint, (2) conditions of exposure,
(3) complexity of experimental measures, and (4) form of experi-
mentation. The endpoint is the effect that is actually measured in
the experiment. Commonly used endpoints include survivability
or lethality, reproductive success, rates or products of chemical
reactions, growth, and development time.3
The objective of dose-response experiment would be to deter-
mine how the endpoint changes with increasing concentration of
the pollutant. Exposure conditions include both the means and
length of exposure. The key receptors may be exposed via their
food or through air- or water-borne pathways. They may receive
a single dose or a continually renewed dose designed to maintain
the concentration of the pollutant constant during the experiment.
The duration of the test can be very short or extend to a partial
or the full cycle. Multiple generations may also be used."
The complexity of the experimental measure can range from a
study of highly controlled single process or study to a highly com-
plex multiple species and process (microcosm) test. The latter tends
to be more indicative of real-world effects but is a more compli-
cated test procedure."
Closely related to complexity is the form of experimentation.
Laboratory tests are highly controlled and have the best chance of
giving unequivocal results, but they are less indicative of real-world
conditions and effects than tests conducted in the field. Thus, it is
clear that dose-response data collection for ecological effects can
vary from laboratory exposure of a simple acute, static single
species or process through complex, lengthy field tests. The over-
all technical approach or philosophy will determine just which pro-
cedures are selected and how they are used to define biological
effects.
Another important aspect of dose-response assessment is the re-
covery potential of populations and processes after exposure steps.
Populations able to recover rapidly after exposure to a particular
stress are subject to less damage overall than populations which
are slow to recover or never recover after a similar exposure. Re-
covery potential data for a population or process has not routine-
ly been measured and no widely accepted protocols exist. Tech-
niques need to be developed to indicate recovery potential.
-------
RISK/DECISION
383
Human Effects Assessment
Human risk assessment models are much further developed than
ecological risk assessment models. The subjects at risk—humans—
are more valuable and important than nonhumans and, therefore,
more attention has been given to the development of risk systems
for human protection. Human risk assessment models rely on at
least three basic types of extrapolations: (1) high-dose-to-low-
dose, (2) species-to-species, and (3) acute-to-chronic-effect ex-
trapolations.
The concept of interspecies extrapolation is implicit in both high-
dose-to-low-dose and acute-to-chronic-effect extrapolation. The
models are usually developed for estimating human risk from ani-
mal data; they can, in some instances, be applied to other popu-
lations in a general way. However, the concepts, but not the re-
sults of human and ecological effects and risk assessments, are
comparable at present. The results could be made comparable
through improvement of population models and the development
of community and ecosystem level risk models. Because the thrust
of this paper is on the nonhuman aspects of risk, no further treat-
ment will be given to human effects assessment.
The final step in risk integration is the combining of the human
and ecological risk values. This involves summing the two values
with the ecological value being multiplied by a 0-1 weighting fac-
tor before the addition. This weighting factor is an estimate of the
relative importance of ecological risk in reference to human risk.
The output of this step of risk assessment is a single fully inte-
grated estimate of risk associated with the action or pollutant of
concern.
The relative risk evaluation step culminates the risk assessment
scheme. Its objective is to develop overall risk assessment values
which can be used directly in the decision-making process. It fac-
tors together the risk value and economic costs of each alternative
action (including no action) associated with the control and cleanup
of a hazardous waste site and permits the comparison of alterna-
tives so the "best" one can be selected. This is, perhaps, the most
difficult step in the risk assessment process and the one for which
sound, widely acceptable approaches are lacking. However, efforts
are underway to develop the formulas or algorithms to solve this
step.
QUANTIFICATION OF RISK
The third component in the overall risk assessment scheme is
the synthesis of environmental fate information with effects data
to obtain an overall assessment of risk. This process applies to
both human as well as ecological risk assessment and can be broken
into three basic steps: (1) exposure/effects adjustment, (2) inte-
gration of risk, and (3) relative risk evaluation. Each step success-
ively integrates the data until a single measure results.
The development of the two risk estimates, one for human and
the other for ecological, is done in step 1 and the first part of step
2. These two estimates are then combined in the last part of step
2 to give a single integrated risk value. The third and final step
(comparative risk assessment) gives perspective to the work through
relative comparisons among hazardous waste locations and control
alternatives.
The objective of the exposure/effects adjustment step in the
risk assessment scheme is to synthesize the dose-response data for
each indicator with the environmental fate information. This es-
timate of the level of effect is modified by the duration of the
exposure, the potential for recovery, and the area of the eco-
system or the number of humans actually exposed. The duration
of the exposure is a function of the pattern and rate of release to
the environment, the rate of movement along the pathways to the
target ecosystems or human populations, and the retention of
time.
Recovery is the'ability of the receptor to either adapt to the level
of exposure or return relatively quickly to pre-exposure conditions
after cessation of exposure. Ecological populations or processes
with short generation times or half-lives can recover more quickly
than those with long generation times.12 For example, planctonic
algal populations subject to chemical stressors can often return
to pre-exposure levels within days and weeks after the stress
ceases. Rooted plants may take significantly longer to recover. A
third modifier of the quantitative effect level derived from the
dose-response curve is the area exposed. The larger the area of ex-
posure, the greater the magnitude of effect, i.e., a larger pro-
portion of the ecosystem or human population is affected.'3
The integration of risk involves the synthesis of the individual
receptor effect estimates, into a single risk value. The integrated
ecological risk value is obtained by summing the risk values of each
individual receptor. This sum is then multiplied by the importance
value or weight assigned to that particular environmental type.
The resulting weighted values are then summed to give the overall
ecological risk assessment value for the pollutant and action of
concern. The development of the human risk estimate involves a
similar activity.
CONCLUSIONS
Risk assessment of uncontrolled hazardous waste sites are incom-
plete without ecological considerations. Not only could some risk
assessments be based totally on nonhuman health issues but results
of ecological risk assessments could be used to differentiate be-
tween options where there are alternatives of equal or similar health
risk. Concepts and techniques for ecological risk assessment lag
behind concepts and techniques for human risk assessment. This
is especially true for higher levels of biological organization (eco-
logical communities and ecosystems) which are less conceptualized
than the population level. The individual (expressed in terms of a
population statistic) is the primary element of concern in human
risk assessment whereas the population is the lowest level of con-
cern in ecological risk assessment.
Both human and ecological risks have now been put together in
a new logical scheme to aid in characterizing uncontrolled haz-
ardous waste sites. This scheme recognizes three interrelated com-
ponents: exposure assessment, effects assessment (which includes
parallel human and ecological effects steps), and quantification of
risk. The steps in each of these components has been identified
and explained. The actual risk assessment component includes
steps that integrate and compare risks to facilitate the decision-
making process.
REFERENCES
1. Bloom, S.G., Cornaby, B.W., and Martin, W.E., "A guide to mathe-
matical models used in steam electric power plant environmental im-
pact assessment," Biological Services Program, Fish and Wildlife
Service, FWS/OBS-78-01, 1977, 153 p.
2. Cairns, J., Jr. and Dickson, K.L., "Field and laboratory protocols
for evaluating the effects of chemical substances on aquatic life,"
J. Test.Eval. 6, 1978,81-90.
3. Casarett, L. J. and Doull, J., Toxicology, the Basic Science of Poisons,
Macmillan Publ. Co., Inc., New York, N.Y., 1975, 768 p.
4. Cleland, J.G. and Kingsbury, G.L., "Multimedia environmental goals
for environmental assessment," Vols. I and II. EPA-600/7-77-13a,
and -b. USEPA, Research Triangle Park, North Carolina, 1977,
336 and 451 p.
5. Cornaby, B.W., "Biological pathways, transformations, and eco-
system effects associated with power plants." Proc: Environmental
risk assessment, how regulations will affect the utility industry, Elec-
tric Power Research Institute, EA-2064, 1981, 4-43 to 4-64.
6. Cornaby, B.W., Sharp, D.A., and Smithson, G.R., Jr., "Using
technology to manage industrial wastes: a team approach." Battelle
Technical Inputs to Planning. Report No. 25, 1981, 36 p.
-------
384
RISK/DECISION
7. Duguid, J.O. and Reeves, M., "Material transport through porous
media: a finite-element Galerldn model," Oak Ridge National Lab-
oratory, ORNL-4928. Oak Ridge, Tn., 1976, 201 p.
8. Duke, K.M. and Merrill, R.G., Jr., "Development of new bio-
assay protocols," Management of Toxic Substances in our Ecosys-
tems. B.W. Cornaby, ed. Ann Arbor Science Publishers, Inc., Ann
Arbor, Mi., 1981, 101-120.
9. Home, R.A., The Chemistry of Our Environment, John Wiley and
Sons, New York, N. Y., 1978, 869 p.
10. Kenaga, E.E., "Test organisms and methods useful for early assess-
ment of acute toxicity of chemicals," Environ. Sci. Tech. 12, 1978,
1322-1329.
11. Maki, A.W., "An analysis of decision criteria in environmental haz-
ard evaluation programs," Analyzing the Hazard Evaluation Pro-
cess, Proceedings of a workshop held in Waterville Valley, N.H.,
Aug. 1978. K.L. Dickson, A.W. Maki and J. Carins, Jr., eds. Water
Quality Section, American Fisheries Society, Washington, D.C.,
1978,83-100.
12. Odum, E.P., "The strategy of ecosystem development," Science,
164,1969,262-210.
13. Odum, E.P., "Fundamentals of Ecology," W.B. Saunders Co., Phil-
adelphia, Pa., 1971, 574 p.
14. Parkhurst, M.A., Onishi, Y., and Olsen, A.R., "A risk assessment
of toxicants to aquatic life using environmental exposure estimates and
laboratory toxicity data," Aquatic Toxicology and Hazard Assess-
ment, D.A. Branson and K.L. Dickson, eds. ASTM STP 737. Amer-
ican Society of Testing and Materials, Philadelphia, Pa., 1981, 59-71.
15. Schroeder, H.A. and Mitchener, M., "Toxic effects of trace elements
on the reproduction of mice and rats," Arch. Environ. Health 23,
1971, 102-106.
16. Tsubaki, T. and Irukayama, K., eds., "Minamata Disease, Methy-
mercury Poisoning in Minamata and Niigata, Japan." Kodansha
Ltd. and Elsevier Scientific Publishing Co., Amsterdam, 1977, 317
pp.
17. Van Voris, P., O'Neill, R.V., Emanuel, W.R., and Shugart, H.H.,
Jr., "Functional complexity and ecosystem stability," Ecology, 61,
1980,1352-1360.
18. Whyte, A.V. and Burton, I., eds., Environmental Risk Assessment.
John Wiley and Sons, New York, N.Y., 1980,157pp.
-------
EXPOSURE-RESPONSE ANALYSIS FOR SETTING
SITE RESTORATION CRITERIA
G.W. DAWSON
Battelle Pacific Northwest Laboratories
Richland, Washington
D. SANNING
U.S. Environmental Protection Agency
Municipal Environmental Research Laboratory
Cincinnati, Ohio
INTRODUCTION
As with most endeavors, the selection and development of
remedial action alternatives for site restoration is most productive
when targeted for specific environmental objectives. In general, the
objective is to reduce associated risks to an acceptable level. The
National Contingency Plan speaks of a balanced consideration of
risks and cost to select appropriate levels of actions.
Ideally, one would eliminate all risks, but experience has shown
that there is no zero risk site. Indeed, prudent observers have noted
that risk minimization becomes exceedingly costly at low levels of
residual primary risk and may in so doing increase secondary risks
associated with producing the resources required for implementing
greater degrees of reduction. When residual risk becomes in-
distinguishable with normal or background risk, the added costs
(including the cost of secondary risks) can not be justified. Thus,
remedial action assessment is best based on the evaluation of alter-
natives in the context of acceptable levels of residual risk. This ap-
proach is currently phrased in the vernacular as "How clean is
clean enough?"
Resolution of the "How clean is clean" question is directed at
two elements of the site mitigation process: 1) the designation of
those areas within a site which must be addressed, and 2) the selec-
tion of mitigation alternatives capable of achieving designated
levels of restoration. With respect to the former, one would have to
designate a threshold level of contamination such that environmen-
tal media (waste, soil, sediments, etc.) containing hazardous
residual at that level or greater would be subjected to
restoration, while media containing less would be left undisturbed.
Only those remedial action alternatives capable of bringing all en-
vironmental media below the designated thresholds would be con-
sidered for implementation.
THE EXPOSURE-RESPONSE APPROACH
A method has been developed for the derivation of cleanup
criteria using environmental risk assessment techniques in a mode
designated as Exposure-Response analysis. This approach was
taken recently in a study of the LaBounty Landfill at Charles City,
Iowa. The methodology is described here followed by excerpts
from the Iowa work to illustrate how the application is accomplish-
ed.
The most direct approach to setting restoration goals would be to
use existing media criteria. However, no criteria presently exist for
soils or sediments. For nearly all contaminants, there are no soil
threshold limits defining when hazardous effects will begin to be
evidenced. In part, this reflects the fact that there are no simple
standard tests to which a contaminated soil can be subjected for
designation as hazardous or nonhazardous.
Fortunately, there is still a relatively simple approach available
tor establishing criteria. For most contaminants, soil residuals are
ot concern because of the ultimate ability to contaminate the at-
mosphere (through volatilization or resuspension) and the
Hydrosphere (through leaching and runoff). Hence, for those con-
taminants, hazardous levels in soil can be defined as those which
will sponsor hazardous levels in air or water.
Criteria and guidelines for ambient air and water have been sug-
gested for a number of chemicals. By working backwards with
minimal data on dilution potential and distribution coefficients,
criteria can be established. Given these, one can determine the
subset of alternatives which can meet objectives and the extent of
restoration required.
This approach to establishing restoration goals, requires
knowledge of three basic components:
•The air and/or water criteria which are not to be exceeded
•Intrinsic properties of the contaminants involved which will de-
termine their fate and migration from source to receiving at-
mosphere/hydrosphere
•The algorithms required to account for soil, air, and water dy-
namics which will dictate the transport of the contaminants
The criteria level (RC) is derived very simply from the above in-
puts utilizing the relation:
RC = (S) x (A) x (D)
where
S = the standard or criteria for the receiving water or atmosphere
(concentration)
A = the attenuation or loss of contaminant during transport de-
fined in terms of the ratio of the chemical on the soil to that
in the air or water at any given time—the distribution con-
stant
D = the dilution factor during transport defined in terms of the
ratio of the concentration at the source to the concentration
in the receiving water or atmosphere
Dimensionally, the result RC is provided a concentration term in
/tg/1 in the soil.
The text following this section is directed to the selection of
restoration goals (threshold criteria) for the LaBounty Landfill in
Charles City, Iowa. Preliminary studies have determined that the
major migration route of concern is the generation of con-
taminated leachate with subsequent transport to the Cedar River.
As a consequence, damage to human health and aquatic life are
hazards of interest upon which restoration should be based. Possi-
ble interconnection of the upper and lower Cedar Valley aquifers
raises concerns for potable wells located in the latter. More restric-
tive standards would be in order if development of the Upper Cedar
River aquifer at the landfill site were contemplated in the future.
PERTINENT STANDARDS
The LaBounty Landfill received a wide variety of materials in-
cluding municipal refuse and pharmaceutical wastes from the
Salsbury Laboratories. As a consequence, a number of discreet
chemical compounds were disposed at the site and may ultimately
escape into ground and surface waters. For the purpose of the cur-
386
-------
387
RISK DECISION
rent study, however, attention is focused on the five constituents
identified in Table 1 as major components of LaBounty waste. For
the purposes of illustrating the Exposure-Response approach,
details will be provided for the arsenic only. Results will then be
given to establish criteria for all five contaminants.
Table 1.
Estimate of Major Components of LaBounty Waste'
Chemical
Arsenic
O-Nitroaniline
Nitrobenzene
1,1,2-Trichloroethane
Phenol
AmountQcg)
2,750,000
680,000
127,000
32,000
12,300
Available data on time-dose relations for ingested arsenic are
summarized in Fig. 1. The lowest detrimental concentrations
reported relate to the potential for initiation of skin cancer in
humans.
Much of the literature on arsenic poisoning stems from an in-
cidence in Taiwan where contaminated water was consumed over
an extended period. It is likely that this source of arsenic was sup-
plemented by locally grown foods. Hence, extrapolation for
designation of carcinogenic risk using a straight line, no threshold
relation yields a very low value of 220 ng/1 for the 10~4 risk level.
This is well below the primary drinking water standard of 50 /ig/1
and the findings of a large-scale survey in Taiwan wherein popula-
100 p •
HUMAN
LETHAL ZONE
*
z
u
z
u.
o
o
U
Z
o
u
- *s CHRO
0.1
0.01
0.001 —
0.0001
110 REPRODUCTIVEV/7/77,
IMPAIRMENT IN DAPHNIA
43 SKIN CANCER
HRONIC
TO AQUATIC UFE
IS REPORT CHRONIC POISONING-
»15 CANCEROUS LESION IN 0.01%
OF POPULATION
»7 PRIMARY DRINKING WATER STANDARD
m .LOWEST LEVEL WHERE CANCEROUS
LESIONS WERE DETECTED IN USERS
NO CANCER NOTED IN CHRONIC USERS
»« 10 CANCER RISK LEVEL
»5 10-5 CANCER RISK LEVEL FOR EATING
_ AQUATIC ORGANISMS FROM THE WATER
, , . I , Mil I , ,,l
10
100
DAYS OR EXPOSURE
Figure 1.
Time-Dose Universe for Arsenic in Water
tions consuming 1 to 17 /ig/1 arsenic could not be found to evidence
any cancer. Skin cancer appeared only in those inhabitants con-
suming water with 50 to 1820 /tg/11 As.
Based on these contrasting values, it is believed that a criteria
value of 10 /ig/1 in Cedar River water should provide ample protec-
tion for human health and aquatic life. This is roughly a factor of 5
lower than the arsenic levels currently reported in the Cedar River
under mean flow conditions.
Similar evaluations were made of toxicological data for the four
organic contaminants found in the LaBounty Landfill. The selected
receiving water standards are summarized in Table 2.
Table 2.
Selected Threshold Criteria in Receiving Waters
Surface
Waters
Ground
Waters
Contaminant
Arsenic
1,1,2-Trichloroethane
O-Nitroaniline
Phenol
Nitrobenzene
10
0.6
10,000
300
30
10
0.6
100,000
300
30
Basis
human health
10 ~6 cancer risk
aquatic life/human health
organoleptic
organoleptic
Notes to Figure 1
1. Estimated lethal dose11 [70 to 180 mg] for 1 I/day in a 10 kg child, this
equates to 70 mg/1, since arsenic has a half-life in the human body of 280
days24 a simple model of the concentration of arsenic in the human body'
yields a chronic lethal drinking level of AG = 70(ln2)/280 = 0.17 mg/1.
Thus, a consumption schedule for lethality would be
Days of Consumption Concentration (mg/1)
1 70
2 3.5
oo 0.17
2. Daily ingestionof 3 mg As for 2 to 3 weeks led to skin and nervous
system disorders." This equates to consumption of 3 mg/1 water for the 10
kg child model'.
3. Chronic exposure (assume >1 year) to <1 mg/1 in water led to skin
cancer, keratosis and hyperpigmentation.7
4. EPA estimates that chronic exposure to 220 ng/1 of As will increase
cancer risk at the 10~4 level exposure.'
5. EPA estimates that chronic consumption of aquatic organisms from
water with 175 ng/1 As will raise the cancer risk at 10~5.'
6. EPA estimates a chronic limit for freshwater fish of 440 ;tg/l.'
7. Primary drinking water standard for As is currently 0.05 mg/1.
8. Lowest reported level for chronic poisoning 0.21 mg/1.1
9. LC50 for daphnia 7.4 mg/1 over 96 hr exposure
LC50 for daphnia 2.85 mg/1 over 21 days exposure.1
10. Reproductive impairment in Daphnia was 50% and 16% at 1400 and
520 jig/1 respectively.'
11. 96 hr LC50 values for bluegill, channel catfish and fathead minnows
reported as 44.8, 18.1 and 15.1 mg/1 respectively."
12. 16 week exposure to 0.7 mg/1 As resulted in 18% survival for formative
bluegills.'0
13. Residents using water with 50 to 1820 jig/1 As showed excessive cases of
cancerous epithelial lesions.10
14. Residents using water with 1 to 7 jig/1 As had no cancer.10
15. Consumption of water with 80 (tg/1 As yield cancerous lesions in 0.01%
of population."
INTRINSIC PROPERTIES
The migration of contaminants in the groundwater is largely a
function of solubility, the affinity for the constituent of interest to
associate with soil particles and degradation mechanisms. With
respect to the former, the primary issue is the relation of solubility
levels to pertinent criteria. If a contaminant's solubility is less than
the criteria, no risk is posed in the way of leaching with subsequent
contamination of receiving waters. On the other hand, if solubility
exceeds criteria, the potential exists for leachate to bring receiving
-------
RISK/DECISION
388
waters to unacceptable contaminant concentrations. Similarly, if
the ratio of solubility to criteria is less than the dilution factor,
solubility constraints will prevent the criteria from being exceeded
in the receiving water.
The affinity contaminants have for soil affects their mobility by
retarding transport. The more strongly a chemical is attracted to
the soil, the more its transport is attenuated, and hence, the slower
its migration is relative to the groundwater itself and the resultant
equilibrium water concentration is lower. Retention on soil may be
due to ion exchange or physical adsorption. If migration carries the
contaminant into different geochemical environments, precipita-
tion of less soluble salts may also occur. Soil retention mechanisms
are often quantitatively described with a distribution coefficient,
kd, which is defined as the ratio of the concentration of the con-
taminant on the soil to its concentration in associated waters. For
those organic contaminants whose attenuation results from sorp-
tion on organic matter in soil, the distribution coefficient is referred
to as koc. The latter is a kd adjusted to organic soil content. Hence,
_
adsorbed contaminant/g of soil organic carbon
jig dissolved contaminant/g solution
Degradation mechanisms serve to reduce both soil and ground-
water contaminant concentrations over time. They may result from
hydrolysis, photolysis, biodegradation and volatilization. Ground-
water conditions are typically such that only hydrolysis plays a ma-
jor role in contaminant loss. The following review provides order
of magnitude estimates for the pertinent intrinsic properties of
arsenic.
Arsenic solubility is dependent on the form (valence state) and
presence of other ions. In general, arsenic is soluble at 100 to 500
mg/1. This is well above the criteria level of 0.01 mg/1 and
therefore, solubility will not limit the ability of leachate to contami-
nant receiving waters.
Arsenic is one of the few anionic species that ties up in soils.
Arsenic attenuation in soil is believed to be a result of adsorption
on metal oxide species and coprecipitation. In Sharpsburg and
Menfro soils, Hess and Blanchar'4 found a kd for arsenic of 50 at
solution concentrations sO.3 mg/1 and 200 at solution concentra-
tions 70.3 mg/1. The odd shape of the results when plotted as a
Freudlinch isotherm suggested that precipitation was the controll-
ing mechanism and not monolayer adsorption.
In liner studies at the University of Arizona, Fuller" found ap-
proximate retardation factors of 5 to 10 in soil and 50 to 75 in soil
with limestone. By definition, the retardation factor equates to the
sum 1 + 3(kd). Hence, these values correspond to kd values of 2 to
25 depending on soil pH. Since Charles City is underlain by
dolomite, it is not unreasonable to assume that the soil attenuation
will lie midway between these values rather than at the low end.
Therefore, a kd value of 10 is reasonable for the purposes of deter-
mining where a soil is contaminated to the point of requiring
mitigation.
No loss mechanisms are important for arsenic. In some cir-
cumstances, biological action can sponsor the production of arsine
which will volatilize. However, in flowing groundwaters, bacteria
are typically filtered out so that this mechanism is not likely to oc-
cur at Charles City.
From the above data and other assumptions, it is possible to
derive the necessary parameters to predict movement of arsenic
from the LaBounty Landfill site. A similar evaluation resulted in
pertinent values for the four organic contaminants. The selected
values are summarized in Table 3.
TRANSPORT DYNAMICS
The final data required to set restoration objectives relates to the
dynamics of the migration route. In the case of the LaBounty
Landfill, the factors of importance will be the dilution factors en-
countered when affected groundwaters enter the Cedar River and
when the affected groundwaters enter Upper Cedar Valley aquifer.
These dilution factors can be estimated from the water flux
values employed in calibrating the groundwater model developed
Table 3.
Intrinsic Properties Affecting the Fate and Migration
of Contaminants at LaBounty Landfill
Contaminant
Arsenic
1,1,2-Trichloro-
ethane
O-Nitroaniline
Phenol
Nitrobenzene
('vratio)
Solubility
Threshold
Criteria
10'
10'
10'
Kd
10
3.7
1.5
0.17
1.7
Half-life
6 mo.
long
long
long
Decay
Mechanism
None
hydrolysis
biological
biological
biological
for the site. In particular, it was found that the flow through the
waste area which generates the contaminant plume is 55.2 mVday.
The total aquifer flow in the region is 11,694 mVday. Hence,
aquifer dilution is 11,692 •*• 55.2 or approximately 200. Similarly,
low flow in the river (7 day, 10 year value) is 230,893 mVday which
yields a river dilution factor of 230,893 -H 55.2 or approximately
^4,000.
CRITERIA FORMULATION
As noted previously, the pertinent criteria (RC) for a site is
estimated from the relation:
RC = S x A x D
The necessary input values for each contaminant at LaBounty are
summarized in Table 4 along with the calculated criteria. Several
comments are appropriate:
Table 4.
Input Values and Restoration Criteria for LaBounty Landfill
Slandard-S
Contaminant
Attenuat'n
Factor-A*
Dilution
Factor-D
Arsenic
1,1,2-Trichloro-
ethane
O-Nitroaniline
Phenol
Nitrobenzene
10
0.6
10,000
300
30
10
3.7
1.5
1
1.7
Arsenic
1,1,2-Trichloro-
ethane
O-Nitroaniline
Phenol
Nitrobenzene
10
0.6
100,000
300
30
Upper Cedar Valley Aquifer
200
200
200
200
200
4,000
4,000
4,000
4,000
4,000
Cedar River
10
3.7
1.5
1
1.7
Soil Criteria
RC-Oig/g)
20
0.4
3,000
60
10
400
600,000
1,200
200
•The 7 day 10 yr low flow value may be conservative for use of
chronic hazard values. Mean flow values (1.9 million mVday)
may be more appropriate. This would result in soil criteria
roughly 10 times as high for the Cedar River.
•The dilution is proportional to the size of the contaminant plume
source. As a consequence, the dilution factor goes up as source is
removed. Hence, development of maximum criteria would re-
quire interactive consideration of reduced source term size as
the criteria level is raised. The values provided here are order of
magnitude only.
It is clear that this approach to setting criteria will be conser-
vative. As noted above, partial reduction of the source term in-
creases the dilution term and therefore would allow a smaller
volume of more concentrated soils to remain without exceeding
receiving water standards. This does not invalidate the approach, it
merely suggests that if remedial action is going to be very expensive,
more effort could be effectively employed in refining the criteria
for the site through an interactive evaluation recalculating dilution
-------
389
RISK/DECISION
factors for each source term reduction until outflows do not exceed
the receiving water standard.
From Table 4, it is readily apparent that contamination of the
Cedar River will be the controlling concern in setting cleanup
criteria. Since no good data are available on soil contamination
contours, it is not possible to determine the extent of cleanup that
would be required as a result of applying these criteria. In general,
however, the contaminant plume extends under the landfill area to
the river. Hence, in this application one need only address the area
of the landfill itself. Indeed, as noted above, removal of the waste
materials themselves will increase the dilution factor significantly.
This, in turn, will reduce the amount of contaminated soil which
must be removed. A few measurements of soil contamination levels
downflow of the wastes would quickly identify the extent of the
area to be addressed.
A cursory review of the criteria suggests that arsenic and tri-
chloroethane will be the determining contaminants in setting
restoration criteria. While these materials have the highest attenua-
tion factors of the five contaminants, neither value is very large. As
a consequence, the groundwater concentration profiles will yield a
fairly accurate estimate of the zone of soil which will exceed the
cleanup criteria. The solubility to toxicity ratio for O-nitroaniline is
on the same order of magnitude as the aquifer dilution factor. It is
an order of magnitude less than the river dilution factor. As a con-
sequence it would appear that O-nitroaniline standards are not like-
ly to be exceeded even if no remedial action is taken. However, cau-
tion is necessary here since there are few toxicological data
available on this chemical. If the toxicity level is found to be much
lower, the solubility ratio may exceed dilution.
REFERENCES
1. Biesinger, K.E., and Christensen, G.M. "Effects of Various Metals
or Survival, Growth, Reproduction and Metabolism of Daphnia
Magna." J. Fisheries Research Board of Canada, 29, 1972, 1691.
2. Brighenti, L., "Azione di Derivati Fenoliec Sull'accrese del S. c
erevisial ed Effetta di Usi Fattore. E patico," Ricerca Sci, 25, 1955,
3333.
3. Burruss, R.P., Jr., and Sargent, D.H. Technical and Microeconomic
Analysis of Arsenic and its Compounds. NTIS PB-253 980, 1976.
4. Cardwell, R.D., el al., "Acute Toxicity of Selected Toxicants to Six
Species of Fish." U.S. EPA Ecological Research Series, EPA 600/
3-76-008, 1976.
5. Clemens, H.P., and Sneed, K.E. "Lethal Dose of Several Chemi-
cals for Fingerling Channel Catfish." U.S. Fish, Wildlife Special
Sci. Report Fish., No. 316, 1959.
6. Cole, C.R., "Evaluation of Landfill Remedial Action Alternatives
Through Groundwater Modeling," Proc. of the Eighth Annual Re-
search Symposium Land Disposal of Hazardous Waste, USEPA,
E8A-600/9-82-002, Mar. 1982.
7. Dawson, G.W., "The Chemical Toxicity of Elements." BNWL-
1815, UC-70, U.S. Atomic Energy Commission, 1974.
8. Federal Register, 5, No. 231, 1980.
9. Fuller, W., "Lines of Natural Porous Materials to Minimize Pol-
lutant Migration." NTIS PB81-221863, USEPA report EPA-600
12-81-122, 1981.
10. Gilderhus, P.A. "Some Effects of Sublethal Concentrations of Sod-
ium Arsenite on Bluegills and the Aquatic Environment." Transac-
tions of the American Fisheries Society, 95, 1966, 289.
11. Goring, G.A.I., and Hamaker, J.M., Organic Chemicals in the Soil
Environment. Marcel Dekker, New York, N.Y., 1972.
12. Hansch, C., et al. Tabulation of Solvent-Salvent Partition Coeffici-
ents, Pomona College, Calif., 1980.
13. Handy, R., and Schindler, A. "Estimation of Permissible Concen-
tration of Pollutants for Continuous Exposure." USEPA Research
Triangle Park, North Carolina. EPA-600/2-76-155, 1976.
14. Hess, R.E., and Blandran, R.W. "Prediction of Water Soluble Ar-
senic in Sewage-Soil Systems," Missouri Water Resources Center, A-
068-MO, 1975.
15. Huang, J.C., and Gloyna, E.F. "Effect of Organic Compounds on
Photosynthetic Oxygenation. I. Chlorophyll destruction and Sup-
pression of Photosynthetic Oxygen Production," Water Research,
2, 1968, 347.
16. Iowa State College of Agriculture and Mechanical Arts-Agricultural
Experiment Station. "Soil Survey of Iowa-Floyd County," Soil Sur-
vey Report No. 48, Ames, Iowa, 1927.
17. Judis, J., "Studies on the Mechanism of Action of Phenolic Disin-
fectants, II." J. Pharm. Sci., 52, 1963, 126.
18. Kopperman, H.L., et al., "Aqueous Chlorination and Ozonation
Studies. I. Structure-Toxicity Correlations of Phenolic Compounds
to Daphnia magna." Chem. Biol. Interactions, 9, 1974, 245.
19. Lammering, M.W., and N.C. Burbank, "The Toxicity of Phenol, 0-
Chlorophenol and O-Nitrophenol to Bluegill Sunfish, Proc. of the
Fifteenth Ind. Waste Conference, Purdue Univ., Lafayette, Ind.,
1960, 541.
20. Little, A.D., "Chemical Hazard Response Information System,"
U.S. Coast Guard, Washington, D.C., 1974.
21. Lutkenhoff, S.D., and Bruins, R.J.F., "Human Health Hazard Posed
by Chemicals Leading from LaBounty Dump, Charles County, Iowa."
ECAG-CIN-82-0, USEPA OFficial use document, 1981.
22. McKee, J.E., and Wolf, H.W., Water Quality Criteria, U.S. Public
Health Service, Publication D-A, 1971.
23. Neely, W.B., Chemicals in the Environment, Marcel Dekker, New
York, N.Y. 1980.
24. Ng, Y.C. et al., Prediction of the Maximum Dosage to Man from the
Fallout of Nuclear Deirces. IV. Handbook for Estimating the Maxi-
mum Internal Dose from Radionuclides Released to the Biosphere.
TID-4500, UCRL-50163 Part IV, 1968.
25. NIOSH. "Registry of Toxic Effects of Chemical Substances," U.S.
Dept. of HEW, Washington, D.C., 1977.
26. Oil and Hazardous Materials Technical Assistance Data System
maintained by USEPA, Washington, D.C.
27. Pickering, Q.H., and Henderson, C. "Acute Toxicity of Some Im-
portant Petrochemicals to Fish." JWPCF, 38, 1966, 14.
28. Ruesink, R.G., and Smith, L.L., Jr., "The Relationship of the 96-
hour LC50 to the Lethal Threshold Concentration of Hexavalent
Chromium, Phenol, and Sodium Pentachlorphenate for Fathead
Minnows, (Pimephales promelas Rafmesque)." Trans. Am. Fish
Soc., 3., 1975, 567.
29. USEPA, "Chlorinated Ethanes: Ambient Water Quality Criteria.''
Draft criteria document, 1979.
30. Yeh, S., "Skin Cancer in Chronic Arsenicism," Human Pathology,
4, 1973, 469.
31. Yllner, S., "Metabolism of l.l^-Trichloroethane-l^-'tC in the
Mouse." Acta Pharmacol Toxicol., 30. 1971, 248.
32. Zaldivar, R., "Arsenic Contamination of Drinking Water and Food-
stuffs Causing Erdemic Chronic Poisoning." Beitr. Path Bd., 151,
1974, 389.
-------
PERSPECTIVES OF RISK ASSESSMENT FOR
UNCONTROLLED HAZARDOUS WASTE SITES
TERRY ESS
Consultant in Risk Analysis & Computer Application
Glenside, Pennsylvania
CHIA SHUN SHIH, Ph.D.
Division of Engineering
University of Texas at San Antonio
San Antonio, Texas
INTRODUCTION
To answer the prime question of risk assessments, "Is this risk
acceptable?", requires that one first know the magnitude of risk
that exists. Mathematically risk can be defined as a function of the
probability of a negative consequence occurring and the value of
that consequence. Therefore, the information required to estimate
risk is the joint probability of a series of events leading to a conse-
quence, the value of this consequence and the functional relation-
ship defining risk. The concepts connected with the valuation of
consequences wilal be covered in a later discussion. In this paper,
the authors concentrate on the estimation of probabilities of occur-
rence and the definition of the functional relationship.
BASICS
To understand the risk estimation process, one must first delve
into some of the characteristics of risk. The first property of risk
that one should note, is that it can be modeled as a chain or series
of events. As indicated in Fig. 1, the events that occur in this path
are hazard, outcome, exposure and consequence. An example of
each type of event is provided in the diagram. One would normally
expect a multiplicity of interconnected hazards, outcomes, etc., so
one should properly think in terms of risk paths.
The second property that should toe considered is that in modern
technological risk problems, hazardous waste included, many of
these events will be relatively rare occurrences. This means that a
good base of historical statistical data on event occurrence frequen-
cies is limited or nonexistent. This problem is further complicated
by the final property which needs to be considered. Many of the
elements of risk pathways, such as the specific toxic substance in-
volved, are new and therefore unknown.
Event
Example
Hazard —
Tree struck
by lightning
» Outcome — » Exposure — >
Tree falls Man walking
in woods
Figure 1.
Pathway Concept
Consequence
Injury to man
The probability of a consequence occurring is normally:
P(H«0«E«C) = P(H)»P(O/H)»P(E/H)O)»P(C/H»OE) (1)
where:
P(H) = probability of a specific hazard
P(0/H) = probability of an outcome given that a specific
hazard has occurred
"(E/H 0) = probability of exposure given a hazard and out-
come
0 E) = probability of a specific consequence given an ex-
posure, outcome and hazard
In the case of mutual independence this reduces to:
P(H»OE«C) = P(H)»P(O)»P(E)»P(C) (2)
Therefore, to determine the desired probability, requires that one
know the conditional probabilities along a specific risk path or in
the case of mutual independence the a priori probability of occur-
rence for each event. However, as indicated earlier, in many cases
this will not be easy to accomplish since one will be dealing with
events for which little or no historical data exist. Thus one needs a
technique that will allow one to estimate those probabilities where
historical data are lacking.
Nor is this the only difficulty to be encountered. How does one
determine all the likely risk paths (or at least the significant ones) in
an "acceptable manner"? Considering the potential sensitivity of
the final risk estimate to the completeness of this information,
answering this question becomes critical to the solution search.
To address these problems, special efforts involving both
technical experts and community based citizen groups must be or-
chestrated. Four suggested actions seem to enjoy the widest accep-
tance: (1) use of a multidisciplinary team headed by a generalist and
composed of knowledgeable personnel, (2) allowance for some type
of peer review of the technical data base developed, (3) some
mechanism for meaningful public involvement, and (4) a suitable
analytical methodology.
Most of the remainder of the paper will be devoted to this
"suitable analytical methodology". This methodology will need to
possess the following capabilities:
1. Means to provide estimates of risk probability
2. Procedures to facilitate systematic thinking
3. Processes which allow for the incorporation of inputs from
multiple individuals and disciplines
4. Easy to review
In short, what is required is a quantitative analysis method which is
both structured and flexible and which imposes a reasonably strin-
gent standard of documentation.
A continuum of methods, from totally informal undocumented
intuition to very formal fault tree analysis, are possible. However,
only the formal end of the scale has the potential to meet all the re-
quirements listed above. In particular, fault tree analyses possess
each of the capabilities mentioned. Fault trees allow one to logically
break down a problem until one reaches a level at which one knows
or can readily determine occurrence probabilities.
RISK ACCEPTABILITY
Risk acceptability is concerned with the determination of what
level of safety is required (or what level of risk can be allowed) by
society for specific risk situations. The problems of risk acceptabili-
ty can be summarized in two questions. The first is "How safe is
safe enough?" The second question, "Acceptable to whom?" is
not nearly so often heard as the first. There are no organized
techniques to handle this problem. It is of course impossible to real-
ly answer the first question without first answering the second, but
390
-------
391
RISK/DECISION
often the answer to the second question is only implicitly stated if at
all. Regretfully, the authors also will have to bypass the second
question with the tacit understanding that the individual decision
maker involved in hazardous waste management has a clear idea of
who the "whom" is.
RISK PERCEPTION
If every individual perceived the world around him the same way
there would be no difficulty in assessing the acceptability of a par-
ticular risk situation. In the real world people often fail to perceive
reality very clearly or in the same way. The risk problems of interest
to the authors, are, in many cases, neither well established nor
documented. They are often surrounded with a large degree of
uncertainty resulting from such diverse causes as limited knowledge
and restricted measurement capabilities. Compounding these
limitations is the complexity of the problem, not just in terms of the
multiplicity of risk pathways but also because risk does not exist by
itself. It is only one of many problems which must be considered as
simply one factor in a morass of benefits and costs, direct and in-
direct, that surround any public decision problem.
The intuitive and cognitive ability of the normal individual are
simply swamped by this complexity, thereby forcing him to rely on
simplified rules of thumb. These simplified information-straining
and decision making rules often produce erroneous judgments. For
instance, one such hueristic, judging the probability of a risk based
on the ease with which instances can be brought to mind, can ob-
viously lead to unjustified biases. This hueristic at least partially ac-
counts for the media's ability to distort the public's perception of
risk. Under these circumstances, it is not surprising that it is often
difficult, if not impossible, to assess a risk's public acceptability.
Often one simply does not even know what the general subjective
perception of the risk is until after the fact.
An anatomy of human perception and its effect on choice
behavior based on experimental evidence is beginning to be
generalized in prospect theory.' Under this theory one can no
longer use expected value (i.e., probability consequence) to
describe the preference ordering of options. Instead, one must also
incorporate functions which account for the differences in percep-
tion due to the problems framing (i.e., the observers conception of
the problem, consequences, and contingencies). As a result instead
of the familiar expected value formulation for risk one gets:
R = [TI(p)][v(c)] (3)
where:
no
R
= decision weight associated with probability of
occurrence
= values associated with consequences
= risk
Hypothetical value and decision weight functions are depicted in
Figs. 2 and 3. If n and v were linear throughout, an individual's
preference between choices would be independent of the problem's
framing. However, due to the characteristic nonlinearities, dif-
ferent frames can lead to different choices even though the ex-
pected values of the options remain the same. Prospect theory is
amendable to incorporation in multiattribute utility measurements,
a topic which will be covered elsewhere.
Besides the theoretical and experimental work on prospect
theory, a great deal has been done to determine inferred or intuitive
factors in the development of perception. One of the most com-
plete analyses, at least for the specific area of risk assessments, has
been provided by Rowe." His factors for transforming objective
reality in to subjective perception are summarized in Table 1. A
brief description of these factors follows:
Table 1.
Objective to Subjective Transformation Factors4
Factors involving type of consequence:
•Voluntary or involuntary
•Discounting of time
•Identifiable or statistical risk taker
•Controllability
Factors involving nature of consequence:
•Position in hierarchy of consequence
•Ordinary or catastrophic
•Natural or man originated
Other factors:
•Magnitude of probability of occurrence
•Propensity for risk taking
•Voluntary or involuntary—perception appears to be markedly
effected by whether the risk is incurred by choice or not. For in-
stance, one normally expects a worker at a hazardous waste site
to be much more tolerant of risk than the surrounding inhabi-
tants.
•Discounting of time—events happening now tend to be valued
higher than the same event sometime in the future. This obviously
dovetails with the long held financial concept that a dollar today
is worth more than the same dollar a year from now.
•Identifiable or statistical risk taker—whether a risk will be taken
(or imposed) on individuals or groups with which we identify or
just a "number in the crowd" affects one's perception. A classi-
cal example of this can be seen in the huge, expenditures of money
undertaken to rescue trapped miners who have become identifi-
able while begrudging support to routine safety budgets.
l.O-l
o
LLJ
3
2.5-
. 5
STATED PROBABILITY P
Figure 2.
Hypothetical Probability Function'
I. 0
VALUE
Figure 3.
Hypothetical Value Function'
-------
RISK/DECISION
392
•Controllability—people appear to accept much higher risk when
they feel that the situation is well controlled such as when they
are driving the car.
•Position in hierarchy of consequences—the wish to avoid a con-
sequence depends heavily on the perceived undesirability, i.e.
position in a desirable-undesirable hierarchy, of the conse-
quence (see Table 2). As a result, one would normally expect the
threshold for noticing risk to be much lower for fatality situa-
tions than ones involving risk to security.
•Ordinary or catastrophic—large numbers of fatalities, etc. in a
single accident has a much more pronounced impact than the
same number of fatalities spread over a number of small acci-
dents. For example, a much higher risk tolerance is expressed by
the public in the case of auto accidents (which are normally
ordinary) than in commercial aviation accidents (which tend to
be catastrophic).
•Natural or man originated—risk imposed by natural causes tends
to be much more easily tolerated than man-made risks probably
because there are few if any alternatives to accepting the natural
risk.
•Magnitude of probability of occurrence—the perceptions of a
consequence are nonlinearly influenced by the magnitude of the
Table 2.
Consequence Hierarchy
Lowest Priority
Highest Priority
Self-Actualization
Egocentric
Belonging/Love
Security
Exhaustable Resources
Survival Factors
Illness & Disability
Death
probability of that consequence. This often results in very low
probabilities being overstated and high probabilities understated.
•Propensity for taking risk—the inherent level of risk taking ac-
ceptable to an individual or group can reasonably be expected to
vary. As a result we see instances such as the nuclear power con-
troversy where equally rational groups have extremely divergent
opinions on the acceptability of a risk activity.
It seems fairly obvious that one is dealing with a very complex
human phenomenon. Further, if one accepts the factors listed (or a
similar set), then it is necessary to reject the use of aggregated risk
to make decisions. If one wishes to use subjective perceptions in
one's assessments, then risk must be deferentiated into classes
which are commensurate with the difference between reality (or our
best objective estimate of it) and these perceptions.
TECHNIQUES FOR DETERMINING
RISK ACCEPTABILITY
A number of possible techniques for addressing the question,
"How safe is safe enough?" have been proposed. Three basic ap-
proaches can be readily identified. The first is the formal analysis
approach. The principal methods included in this category are
benefit/cost analysis and decision analysis. This approach relies
heavily on formal logic and optimization principles.
The next technique is the comparative analysis approach. It is
composed of three distinct methods: revealed preference, expressed
preference and natural standards. In each case a more or less ab-
solute acceptable risk limit is devised against which the estimated
risk of an activity can be compared.
The last major category is professional judgment. This ap-
proach, of course, relies principally on the intuitive intelligence of
the professional community. An in-depth comparison of each
technique using five key characteristics, decision making criteria,
locus of wisdom, principal assumptions, decision attributes possi-
ble and data requirements, is provided in Table 3.
Table 3.
Comparison of Techniques
Technique
Formal Analysis
•Benefit/Cost
Analysis
•Decision Analysis
Comparative
Decision Making
Criteria
Economic
optimization
Utility optimization
•Revealed
Preference
•Expressed
Preference
•Natural Standards
Professional
Judgment
Preservation of
historical balance
Current preference
Biological wisdom
Professional
judgment
Locus of Wisdom
Formalized
intellectual processes
Formalized
intellectual processes
Societal decisions
during recent past
Societal decisions now
Long term species
survival
Intuitive intellectual
processes
Principal Assumptions
•Man is or should be a
rational economic
maximizer
•Decisions should be
purely objective
•Man is or should be a
rational utility
maximizer
•Decisions should use
decision makers value
judgments
•Past decisions were
essentially optimal
•Little or no change in
circumstances
•Public understands & has
well articulated preferences
•The optimal level of
exposure to pollutants is
characteristic of conditions
during species evolution
•Professionals understand &
have well articulated
preferences
• Professionals always
exercise free will
Decision Attributes
Anything which can
be converted to money
Anything; any number
Risk only
Risk only
Risk only
Anything but limited
number
Data Requirements
Possible
•All significant economic
events & consequences
•Accurate probabilities &
magnitudes for each
•All significant events &
consequence
•Accurate probabilities &
magnitudes for each
•Current risk
•Historical risk
•Current risk
•Current preferences
•Current risk
•Risk magnitudes during
geologic time
Almost anything
-------
393
RISK/DECISION
Each of these techniques has strengths and by themselves some
serious weaknesses (Table 4). The formal analysis methods,
especially decision analysis, provide structure, flexibility and the
potential for easy and meaningful review. Benefit/cost analysis is
seriously constrained, however, since any attribute which cannot
readily be converted into economic terms is ignored. Both formal
techniques suffer from the requirement for large amounts of detail-
ed and reliable data and the inability to incorporate public subjec-
tive perceptions of risk into the formulation.
All the comparative analysis methodologies have the advantage
of establishing "absolute" risk limits. This benefit, however, is off-
set by the fact that all three are only intended to address risk, they
cannot handle the entire decision problem, i.e., the tradeoffs be-
tween risk and the other decision dimensions such as benefits. In
addition, both revealed and expressed preference methods are
subject to the limitations of society and its citizens.
Finally, professional judgment enjoys the benefit of being easily
implemented. However, it is severely constrained by its limited
review capabilities and potentially significant errors in judgment in
the case of complex problems. Probably even more important is the
apparent recent loss of public trust in professional judgments.
Table 4.
Technique Strengths and Weaknesses
Technique
Formal
Analysis
Benefit/Cost
Decision
Analysis
Comparative
Analysis
Revealed
Preference
Strength
•Systematic
•Ease of review
•Handles all decision
dimensions
•Systematic
•Ease of review
•Flexible
•Handles all decision
dimensions
•Incorporates decision
maker's judgment
•Handles uncertainty
well
•Establishes absolute
limits
•Incorporates historical
experience
Weaknesses
•Discounts attributes which
cannot be readily converted
to economic terms
•Large data requirements
•Cannot handle subjective
value judgments
•Large data requirements
•Cannot handle public
perceptions of risk
Expressed
Preference
Natural Standard
Professional
•Establishes absolute
limits
•Allows for widespread
public involvement
•Establishes absolute
limits
•Not subject to limita-
tions of society
•Handles all decision
dimensions
•Flexible
•Easy to implement
•Well established
•Past decisions often were not
always optimal
•Circumstances changing rapidly
•Disaggregated historical base-
line hard to establish
•Does not address whole
decision
•Subject to inherent limita-
tions of society and its
citizens
•Dependend on ability to get
unbiased survey
•Public does not always under-
stand or have a preference
•Does not address whole
decision
•Subject to inherent limitations
of society and its citizens
•Many of today's pollutants
did not exist before
•Does not address whole
decision
•Geological time baseline hard
to establish
•Does not allow for tradeoffs
•Hard to impossible to review
•Bias due to employer
•Stretched intuitive & cognitive
skills can lead to erroneous
judgments
•Professionals appear to be
losing public support
A QUANTITATIVE REVEALED
PREFERENCE METHOD
Of the techniques discussed, probably the two which have en-
joyed the most recent exploration in a risk context are revealed
preference and decision analysis. The remainder of this paper will
deal with a quantitative methodology for determining risk accep-
tability based on revealed preference which is based on the work of
Rowe.4 Rowe's technique is composed of four principal parts:
•Devise an appropriate risk classification scheme
•Determine an absolute risk reference for each class in the scheme
•Using risk references as a base, calculate risk referents that act
as the acceptability limits for specific situations
•Compare the estimated risk with the appropriate risk referent. If
the estimated risk is within an order of magnitude of the referent
then it can be considered acceptable.
The first two steps are intended to be completed only once,
thereby creating an absolute reference base. The third step is
repeated for each new activity and allows for the modification of
the reference base to fit the situation specifics. As indicated in
Table 5, these steps explicitly include all the objective to subjective
transformation factors that were mentioned earlier except for one,
the effect of the magnitude of probability of occurrence.
Table 5.
Transformation Factor Utilization in Risk Referents4
Factors involving type consequence:
•Voluntary or involuntary (1)
•Discounting of time (1)
•Identifiable or statistical risk taker (1)
•Controllability (2)
Factors involving nature of consequence:
•Position in hierarchy of consequence (1)
•Ordinary or catastrophic (1)
•Natural or man originated (1)
Other factors:
•Magnitude of probabilty of occurrence
•Propensity for risk taking (2)
(I) Explicitly included in determination of absolute risk reference
(2) Explicitly included in determination of risk referent
If risk assessments are to be useful, one cannot treat risk as an
aggregate. One must break it into component parts commensurate
with one's understanding of the factors that lead to subjective
perception. The basic classification scheme advocated by Rowe is
shown in Table 6. In conjunction with this scheme, four classes of
consequence are delineated: (1) fatalities, (2) injuries and morbidi-
ty, (3) property damage, and (4) reduction in years of life expectan-
cy. Additional consequence classes would be desirable in order to
adequately cover the entire range of the consequence hierarchy
(Table 2) but the data just do not exist to justify any other classes.
Table 6.
Classification of Absolute Risk4
Classification
Immediate statistical
1- Natural
a. Catastrophic
b. Ordinary
2. Man triggered
a. Catastrophic
b. Ordinary
3. Man originated
a. Catastrohpic
b. Ordinary
Immediate identifiable (1)
Delayed statistical (1)
Delayed identifiable (1)
(I) Same as immediate statistical
Voluntary Regulated Involuntary
X
X
X
X
X
X
X
X
X
X
X
X
-------
RISK/DECISION
394
Once a classification scheme is adopted, an absolute risk
reference must be established for each class. These are estimated
directly from historic, societal risk data (Rowe selected reference 1
as his primary source for man-originated accidents) as revealed
preferences. Where data do not exist for a specific class, estimates
are based on an analogous set of risk classes for which data exist.
The risk references determined by Rowe for immediate statistical
accidents are shown in Table 7.
Risk Classification
Table 7.
Summary of Risk References4
Class of Consequence
Fal/Yr
He/Yr
$/Yr
Yr
Naturally occurring
Catastrophic
Ordinary
Man originated
Catastrophic
Involuntary
Voluntary
Regulated voluntary
Ordinary
Involuntary
Voluntary
Regulated voluntary
Man triggered
Catastrophic
Involuntary
Voluntary
Ordinary
Involuntary
Voluntary
Regulated voluntary
IxlO'6
7xlo-5
IxlO-7
2xlO-6
3xlO-5
5xlO-6
6X10-4
IxlO-4
2xlO-7
4xlO'6
lxlO'5
IxlO-3
2X10-4
5xlO-6
4X10-4
5xlO'7
2xlO'6
3xlO'6
3xlO-5
3x10-'
6xlO-2
IxlO'6
4xlO'6
0.02
3
2x10-2
0.4
0.4
1
200
30
4x10-2
0.8
3xlO'2
0.2
3X10-4
6xlO-3
6xlO-2
IxlO-2
1
.1
6x10-4
6x10-3
3xlO-2
2
0.2
Conversion of a risk reference into a risk referent requires four
steps:
•Determine the appropriate risk proportionality factor, i.e., the
fraction of existing societal risk (risk reference) that would be
considered acceptable in a situation where there was a very favor-
able indirect benefit/cost balance, for both voluntary and in-
voluntary risk. (Fl)
•Determine a factor, the risk proportionality derating factor,
which modifies the risk proportionality factor in those situations
where the indirect benefit/cost balance is not as favorable as in
the first step (F2)
•Determine the modification factor associated with the degree of
risk controllability. (F3)
•Using the three factors determined above, calculate the risk
referent:
Risk referent = Risk reference x Fl x F2 x F3
(4)
The first two factors address the inherent propensity of specific
individuals/groups to take risks and incorporates the additional
decision dimension of indirect benefits/costs. This acknowledges
the tendency for people to accept a higher level of risk if the benefit
to them more than offsets the imposed risk or to be increasingly
risk adverse if it does not. All three of these factors are based on
value judgments. The specific numbers in Table 8, risk propor-
tionality and proportionality derating factor, and Table 9, con-
trollability factor, are based on the "straw man" values originally
posed by Rowe.
A brief explanation of Table 9 is probably in order. The overall
controllability factor is the result of the multiplication of four sub-
factors (Fl =C1 x C2 x C3 x C4). The four sub-factors are: (1) con-
trol approach (i.e., the type risk control management used), (2)
degree of control (i.e., effectiveness of risk control), (3) state of im-
plementation, and (4) basis for control effectiveness (i.e., whether
risk change calculations are related to an activity index, relative, or
not, absolute).
Example Problem
An example that may help illustrate this process concerns an
aquifer in San Antonio, Texas. In this illustration, only ordinary
fatalities of either the public or workers will be considered, since it
is assumed that there is no possibilty of a catastrophic accident and
that fatalities are the only type risk impact.
Since both the risk reference for a given risk class and the risk
proportionality factor are constants, the first factor one needs to
determine is the proportionality derating factor. Even without a
detailed indirect benefit/cost study it is possible to make a good
estimate of what the public and workers are likely to perceive given
a specific disposal site plan. For instance, if the plan (call it alter-
native 1) was to just dig a hole, dump in 55 gal drums of waste and
cover them with dirt then one could expect that both the public and
workers would perceive this as a much worse balance than another
plan (call it alternative 2) which includes some advanced type of
container design that incorporated a monitoring system and exten-
sive precautions for worker protection.
Table 8.
Risk Proportionality and Risk Proportionality Derating Factors4
Factor
Proportionality factor
Derating factor
Balance
Favorable
Marginal favorable
Indecisive
Marginal unfavorable
Unfavorable
Involuntary Regulated
Risk Voluntary
0.01 1.0
1.0 1.0
0.1 0.2
0.01 0.1
0.001 0.02
0.0001 0.01
Table 9.
Controllability Factor4
Control
Approach
Systematic
Control
1.0
Risk Mgt. System
0.8
Special design
0.5
Inspec. &
regulation
0.3
No scheme
0.1
Degree of
Control
Positive
1.0
Level
0.3
Uncontrolled
0.1
Slate of
Implementatn.
Demonstrated
1.0
Proposed
0.5
No action
Basis for
Control
Effectiveness
Absolute
1.0
Relative
0.5
None
Table 10.
Hypothetical Indirect Benefit/Cost Example
Class
Involuntary (public)
Regulated voluntary
(worker)
Alt.
1
2
1
2
Balance
Marginally unfavorable
Marginally favorable
Indecisive
Favorable
Value
0.001
0.1
0.1
1.0
-------
395
RISK/DECISION
Table 11.
Hypothetical Controllability Example
All.
Control
Approach
0.1
0.5
Degree of
Control
0.1
0.3
Stale of
Implement.
—
0.5
Baste for
Effective.
—
0.5
Factor
Value
0.01
0.0375
Table 12.
Risk Referent Calculation Summary
Risk dusiflciUon
Involuntary,
ordinary, fatal
Regulated
voluntary
ordinary, fatal
ReferencePropor- Alt.
tlon.
Derate. Control. Reference
0.1
IxlCH 1.0
0.001
0.1
0.1
1.0
0.01 5x10-12
0.0375 2x10-9
0.01
0.0375
1x10-7
4xlO-«
This hypothetical example is shown in Table 10. It is also fairly
obvious that the controllability factor would be significantly dif-
ferent for each of these two alternatives (Table 11). All the risk
referent factors and the resulting risk referents are summarized in
Table 12. These referents would be compared to risk estimates for
each alternative and if the estimates were no more the one order of
magnitude higher than the appropriate referent then the risk would
be judged "acceptable".
As indicated before, this approach makes some heroic assump-
tions about the optimally of past risk acceptance decisions made in
the market place. By itself, therefore, it may be of limited
usefulness. However, when incorporated with another technique
—the quantitative revealed preference method—it may serve the
important task of eliminating totally unacceptable alternatives
from consideration.
REFERENCES
1. CONSAD Research Corporation, "Consequences and Frequencies of
Selected Man-Originated Accident Events", USEPA, 1975.
2. Fishhoff, B. et al., "Approaches to Acceptable Risk: A Critical
Guide," Oak Ridge National Laboratory, ORNL/Sub-7656/1, 1980.
3. Fishhoff, B., Slovic, P. and Lichtenstein, S., "Weighting the Risks,"
Environment, May 1979.
4. Rowe, W., An Anatomy of Risk, John Wiley & Sons, New York
N.Y., 1977.
5. Salem, S., Solomon, K. and Yeskey, M., "Issues and Problems in In-
ferring a Level of Acceptable Risk", RAND, R-2561-DOE, 1980.
6. Tversky, A. and Kahneman, D., "The Framing of Decisions and the
Psychology of Choice", Science, Jan. 30, 1981.
-------
ABANDONED SITE RISK ASSESSMENT MODELING
AND SENSITIVITY ANALYSIS
BRIAN L. MURPHY, Ph.D.
TRC Environmental Consultants, Inc.
East Hartford, Connecticut
INTRODUCTION
In this paper, the author presents an objective calculational pro-
cedure (OCP) for evaluating site specific risks due to hazardous
waste migration in the environment. The procedure is based on
describing environmental pathways in terms of series and parallel dif-
fusion and advection elements. Analogies with electric circuit theory
then permit one to identify the controlling elements and to compute
losses along a pathway. A simplified equation results for the risk
presented by a particular chemical to a particular cohort group, that
is a group of individuals exposed along a particular pathway.
The Hazard Ranking System (HRS) described in the National
Contingency Plan also can be reduced to a single equation (for air,
groundwater and surface water pathways) and hence comparisons
with the OCP are possible if one assumes that the HRS also measures
risk. Three observations can be made as a result of this comparison:
•The HRS attempts to express multiple viewpoints in terms of the
relative importance of near term versus long range future risks as
well as near field versus far field risks. The OCP can only express
a single viewpoint consistent with the input assumptions and
parameters.
•For two sections of a pathway through the environment in series
for the OCP describes the one offering the most resistance (i.e.,
the most secure) as controlling. The HRS description is opposite,
the least resistive or least secure portion contributes most to the
final score.
•Sensitivity analysis reveals that the HRS final score is most sensi-
tive to the least important scoring areas. For the OCP, a per-
centage change in toxicity, release rate or population density will
produce the same percentage change in risk.
The principal conclusion to be drawn from the observations above
is that the HRS embodies values other than degree of risk. The first
observation, for example, can be interpreted as a willingness to
sacrifice consistency in return for including multiple viewpoints as to
what constitutes "risk." One can interpret the second observation
that the HRS focuses on things that are "wrong" at a site in the sense
of poor design or practice rather than things that are "right," ir-
respective of how these influence risk. The last observation can be at-
tributed at least in part to the parameters in the HRS being arranged
and combined based on user ease rather than strictly according to the
logic of risk evaluation.
AN OBJECTIVE CALCULATION PROCEDURE
One can assume risk to a human population to be defined by:
Y P C
(1)
cohorts
chemicals
where /3 is the potency of the chemical in question, 7 is a factor
relating ambient concentrations to inhalation or ingestion rates, P is
the population exposed and Ce is the exposure concentration level.
One also defines a cohort group as individuals who share a common
pathway of exposure. The remainder of this section is devoted to fur-
ther development of Ce.
A diagram such as Fig. 1 illustrating the migration of chemicals
through the environment can be viewed as diffusion and advection
processes in series and parallel arrangements. To estimate flux an
electrical circuit analogy can be exploited to produce simplification
of complex diagrams.
Assume a quasi-steady-state. It cannot be a true steady state
because the ultimate source region (e.g., the buried wastes) must be
diminishing or otherwise changing in character. Using the same nota-
tion as in Fig. 1, one then has for conservation of flux, F, in a series
diffusion-advection process (e.g., diffusion of components from
bulk waste interior into groundwater flow):
"l Al (Co - C>
C.
(2)
k being the mass transfer coefficient, A the flux area, C the concen-
tration and V the advection velocity. Eliminating C gives:
K c
where
kl\
V2 A2
(3)
(4)
For a series diffusion process (e.g., diffusion through laminar
surface water and air microlayers during chemical evaporation) the
analogous expressions are:
- c
(5)
If Cl =
C2 and C<
-------
397
RISK/DECISION
Emplacement
Advection from
Site in Groundwater
Ingestion
Exposure in
Well Water
Advection in
Surface Water
Ingestion
Exposure
via Biota
Direct
Ingestion
Exposure
Diffusion into
Atmosphere
Advection
in Wind
Inhalation
Exposure
Diffusion of
Volatiles to
Bulk Waste Surface
Direct Contact
Exposure at
Ground Surface
Advection
Diffusion
I 1= Mode of Migration
(—)= Type of Exposure
Arrows refer to the process
to which they lead
Diffusion through
Soil to
Ground Surface
Diffusion into
Atmosphere
Advection
by Winds
Inhalation
Exposure
Figure 1.
Sample diagram for migration of hazardous wastes in the environment.
water concentration. The process
c v
produces a flux out of the main flow at rate k A C (assuming as
before C'<< C). Suppose material in the main flow is contained in
a layer of constant thickness L. Then:
dC
dt
F - k A C - -LA
(9)
or
where
(10)
(ID
Other types of loss process, e.g., where L diminishes and C is
constant or where there is a bifurcation in the flow can be modeled
in similar fashion. Equation 8 can be changed to read:
c K A
o • •
(12)
where r is the duration of the loss period.
Human exposure occurs at the end diagrams of Fig. 1; the
subscript e is used to denote this location. Then the flux is F = C
Vr Ac and:
(13)
where Ve has been written instead of Kc since it appears that all im-
portant human exposure occurs in advection locations (ground-
water, surface water and air flows).
The analysis procedure implied by Equation 8 requires one to
estimate k, H and V at various points along the flow. V is assumed
to be part of the description of the problem and H is either
tabulated or chemical properties to calculate it are in handbooks
for many substances. The mass transfer coefficient k can be
estimated from the sources shown in Table 1.
Finally, substituting Equation 13 in Equation 1, one obtains the
equation for total risk over time t,, the release lifetime or other
suitable time horizon:
B f
C K A
o m «
(14)
cohort!
where the ratio ^ is the population density of the cohort group in
the exposure area and (C0 K,,, A,,,) is an effective emission rate.
Table 1.
Estimated k Values
k Value Situation Reference
k L liquid side of interface, lagoon Mackay'
k L liquid side of interface, stream Owens el al.'
kA air side of interface in earth's Sutton'
lower boundary layer
kg diffusion through soil for vapor Farmer*
transport
-------
RISK/DECISION
398
CORRESPONDENCE WITH THE
HAZARD RANKING SYSTEM
In the HRS "parameter" scores S^p are combined with multi-
pliers MRAP to form an "area" score SR*. The area scores are then
combinedinto "route" scores SR. Finally, these are combined into
an aggregate groundwater, surface water, air score S. Specifically:
1 SB <15)
SR '
RA
(16)
where KR is a constant for each route, ir is the product operator
and,
RAP
or combining the above:
.59
[l < (I 5 v. ••»)']
1/2
(18)
While Equations 15 through 18 accurately express the form of the
HRS, they do obscure the fact that an observed release produces an
override of the route characteristics and containment areas. It is the
parameters P whose scores are S^p that we wish to identify with
the quantities used in Equation 19. This correspondence is shown
for the groundwater route in Table 2. Similar tables can be con-
structed for the surface water and air routes and comments similar
to those which follow would ensue.
Table 2.
Correspondence Between Equation 14 (OCP) and
Equation 18 (HRS) Parameters
HRS OCP
R = 1 Groundwater
A = 1 Observed release C0 Km Am
A=2 Route characteristics Km Am
P = 1 Depth to aquifer of concern
P=2 Net Precipitation
P = 3 Permeability
A=3 Containment Km Am
A=4and A = 5*
A=6 Waste characteristics
P = l Physical state Km
P=2 Persistence n
P=3 Toxicity/infectiousness /3
A=7 Hazardous waste quantity C0 Km Am t]
A=8 Targets
P = 1 Groundwater use y
P = Distance to nearest
well downgradient A^, T
P = 3 Population served by
groundwater within 3-mile
radius P
•These steps simply choose the larger score between A = l and (A=2) x (A = 3) for subsequent
calculation.
There is no transformation which will map the HRS into the
OCP or vice versa. Perhaps this is not surprising since parameters
in the HRS appear to be arranged by topical area for use conve-
nience. In any case, consider the parameters p (persistence), 0 (tox-
icity) and P (population exposed). In the HRS the scores combine
in the form (M^ + M^S^Sp. In the OCP the parameters occur as
the factor 0 P e"'"'. One transformation of Equation 14 is of par-
ticular interest. It is possible to generate a scoring system for a par-
ticular chemical and cohort group by taking the logarithm of Rt:
S • Sg + S + S- + SH -S -liT (19)
where S^ = In 0, for example, P is population density and W is
quantity of waste, t! C0 Km Am. Alternatively one can set t] = 1 to
consider risk per unit time and replace Sw by SQ, where Q is the
release rate (C0 Km Am).
In fact the HRS Equation 18 is a combination of additive and
multiplicative factors, a sort of hybrid between Equations 14 and
19.
CONCLUSIONS
Examination of Equations 14 and 18 in light of Table 2 yields the
following conclusions:
•The OCP makes no attempt to discount future risks although this
could be done. Thus if there is a release the total risk is simply
proportional to the amount of hazardous material involved (C0
Km Am t,). The HRS on the other hand "strikes a balance" by
combining scores for both total amount of material and release
rate (Km Am) parameters. As a related issue, it is noted that for a
linear no-threshold dose response only the factor e"'"' in Equa-
tion 19 can contain the risk within a limited area Ae. For a per-
sistent chemical therefore the distances involved can become very
large. Again, the HRS takes a middle course by scoring on the
basis of persistence but (for groundwater) only including popula-
tion within a 3-mile radius.
•In the absence of an observed release the HRS multiplies scores
for containment and route characteristics. Thus the precise value
of either scoring area is important no matter how large (or
small) the score. The OCP on the other hand stresses the area
with what amounts to the smaller score since containment and
route characteristics are serial elements with the more resistive or
secure one controlling. Put differently, the OCP tends to con-
sider what is "right" about a site and the HRS what is "wrong"
about a site.
•The particular structure of Equation 18 leads to an interesting
result for the sensitivity of the HRS results to input scores. From
Equations 15 through 18, one can compute the derivative B (S)2/d
SRAP which is a measure of the sensitivity of the final score to
the parameter score SRAP, since ASRAP may be taken as ± 1 (par-
ameter scores only take integral values). Of course for the case
of an observed release or not one must compute S2 (observed)
— S2 (not observed).
.70 sr
(20)
The significance of SRA occurring in the denominator of this result
is that the final score is most sensitive to the least important area
scores.
ACKNOWLEDGEMENT
Portions of this work dealing with the selection and combination
of mass transfer coefficients or k values were performed as a con-
sultant to GCA/Technology Division under USEPA Contract No.
68-02-3168, Technical Service Area 3, Work Assignment No. 82.
REFERENCES
1. Mackay, D. "Environmental and Laboratory Rates of Volatilization
of Toxic Chemicals from Water," Hazard Assessment of Chemicals, 1,
1981, 303.
2. Owens, M.F Edwards, R.W. and Gibbs, J.W., "Some Reaeration
Studies in Streams," Int. J. Air Wat Poll., 8, 1964, 469.
3. Sutton, O.G., "Micrometeorology," McGraw-Hill, New York, 1953.
4. Farmer, W.J., Yang, M-S. and Letz, J., "Land Disposal of Hexa-
chlorobenzene Wastes," USEPA Report: EPA-600/2-80-119 under
Contract No. 68-03-2014, August, 1980.
-------
ASSESSING SOIL CONTAMINATION AT LOVE CANAL
GLENN E. SCHWEITZER
U.S. Environmental Protection Agency
Las Vegas, Nevada
INTRODUCTION
No chemical waste site has caused more controversy and concern
than Love Canal. Located in the midst of a residential area, this
permanent repository of large quantities of discarded industrial
chemicals and chemical wastes has been a center of attention, local-
ly and nationally, both for environmental groups interested in
waste disposal policies and for individuals directly impacted by
nearby waste sites. The Canal has also attracted the attention of
many of the Nation's leading scientists concerned with the technical
complexities of determining the impact of toxic chemicals on
human health.
During the latter half of 1980, the USEPA carried out an inten-
sive study of environmental contamination in the area near Love
Canal. While several limited investigations of contamination in the
area had been previously conducted by Federal and State agencies,
none approached the scope or magnitude of this 1980 effort.
The geographical area of special concern during the study, name-
ly, the Declaration Area, is shown in Fig. 1. This U-shaped area
was defined by a Presidential state of emergency order on May 21,
1980, following the earlier evacuation of the two rings of houses ad-
jacent to the Canal inside the U-shaped area which are referred to
as Rings 1 and 2. Within the Declaration Area are approximately
700 residences, some occupied and some evacuated.
A cross-section of the Canal and the subsurface features in the
immediate vicinity of the Canal are shown in Fig. 2. As discussed
below, this subsurface profile provides important clues as to the
containment capability of the terrain around the Canal.
The study included a hydrogeological investigation, involving
groundwater pumping tests and geophysical investigations, and a
monitoring program involving the collection and analysis of a large
number of air, soil, sediment, water, and biota samples. The
magnitude of the monitoring program which included analyses of
more than 6000 environmental samples and about the same number
of quality assurance samples, is illustrated in Table 1.
Figure 1.
Love Canal Study Area
CLAY CAP, 1ft. THICK (PERMEABILITY: 10 7cm/«l
SILT FILL-PERMEABILITY >10~*cm'i
BASEMENT
IILTY tANO - PERMEABILITY. >10'*on,/l
Mill IT
- -
A W
PERMEABILITY
10'7 la 10 8em*i
— BIIOUND Lf VIL
— ij-2.sn.
_ 4.0 - S.6 ft
— S.O fl.
— 12.0 H.
Figure 2.
Vertical Profile of the Canal and Surrounding Area
LEGEND: BURIED UTILITIES A*f
S - STORM SEWER
A - SANITARY SEWER
W - WATER MAIN
399
-------
RISK/DECISION
400
Table 1.
Magnitude of the Love Canal Monitoring Program
Sampling Areas
Water
Drinking
Ground:
A shallow
Bdeep
Sanitary sewer
Storm sewer
Sump
Surface
Soil
Sediment
Sanitary sewer
Storm sewer
Stream
Sump
Air
Basement
Living
Outside
Transport study
Occupied/Unoccupied study
Sump/Basement-Air Study
Biota
Crayfish
Dog hair
Maple leaves
Mice
Oatmeal
Potatoes
Worms
Declare- Canal
tion Area Area
31
Control Peripher-
Areas al Area Total
44
112
1
18
4
9
55
8
3
0
1
20
14
5
12
11
4
19
13
0
3
13
—
24
0
4
—
3
1
6
1
—
0
9
...
—
6
2
2
2
2
11
15
0
1
1
5
9
0
1
5
—
0
4
0
...
0
0
1
15
11
2
4
3
'3
79
57
1
2 28
7 54
10 19
28 171
J
1 24
9 18
3
10
65
9
5 5
4 7
9
2
35
31
9
18
16
9
The field aspects of the program were, of necessity, confined to a
three-month period. Therefore, there was no attempt to acquire
and interpret preliminary field data to guide a more intensive ef-
fort. Such a two-phased approach is often desirable to help reduce
monitoring costs through more deliberate selection of sampling
sites. Also, the acquisition of data during different seasons was not
possible. However, given the stable condition of the clay cap on the
Canal and the sampling approach, it is unlikely that seasonal varia-
tions reflected in different patterns of surface runoff would
significantly affect the monitoring results.
Study Objectives
The objectives of the overall study effort were:
•To determine the extent of environmental contamination in the
Declaration Area attributable to Love Canal as of the Fall of
1980
•To assess the short-term and long-term implications of any de-
tected groundwater contamination
•To assess the relative environmental quality of the Declaration
Area
•To provide environmental data which could be used to determine
the habitability of residences within the Declaration Area
A central consideration in achieving these objectives was an
assessment of the chemical contamination of the soil in the area.
Discussed below are the soil monitoring program that was carried
out, the significance of data that were obtained from the program,
and the relationship of the data to data obtained from the other
study efforts.
Target and Non-Target Chemicals
About 150 chemicals including metals, pesticides, and industrial
organic chemicals were selected as target chemicals for the soil,
sediment, and water monitoring program. In making this selection,
particular attention was given to chemicals that were identified in
Table 2.
Chemicals Disposed at Love Canal by Hooker Electrochemical
Company (1942-1953)*
Total
Est.
Quant.
Physical Type of Waste State (Tons) Container
Misc. acid chlorides other than liquid & 400 drum
benzoyl—includes acetyl, solid
caprylyl, butyryl, nitro
benzoyls
Thionyl chloride and misc. sul-
ful/fhlorine compounds
Misc. chlorination—includes
waxes, oils, naphthalenes, aniline
Dodecyl (Lauryl, Lorol) mercap-
tans (DDM), chlorides and misc.
organic sulfur compounds
Trichlorophenol (TCP)
Benzoyl chlorids and benzo-
trichlorides
Metal chlorides
Liquid disulfides (LDS/LDSN/
BDS) and chlorotoluenes
Hexachlorocyclohexane
(7-BHC/Lindane)
Chlorobenzenes
Benzylchlorides—includes benzyl solid
chloride, benzyl alcohol, benzyl
thiocyanate
Sodium sulfide/sulfhydrates solid
Misc. 10% of above
Total
liquid &
solid
liquid &
solid
liquid &
solid
liquid &
solid
liquid &
solid
solid
liquid
solid
liquid &
solid
500
1,000
2,400
200
800
400
700
6,900
2,000
drum
drum
drum
drum
drum
drum
drum
drum
& nonmetallic
containers
drum and
nonmetallic
containers
2,400
2,000
2,000
21,800
drum
drum
*Interagency Task Force on Hazardous Wastes, Draft Report on Hazardous Waste Disposal in Erie
and Niagara Counties, New York, March 1979. Hooker Electrochemical Company is now known as
the Hooker Chemicals and Plastics Corporation.
previous monitoring efforts at Love Canal, were detected in
chemical analyses of leachate from the Canal, and/or were known
to have been deposited in the Canal (Table 2). Also, consideration
was given to selecting chemicals of special lexicological concern
(e.g. radionuclides, dioxin) and chemicals with a range of
physical/chemical properties, such as solubilities and partition
coefficients, that might influence their rates of environmental
migration.
In addition to the target compounds, the analytical laboratories
were required to identify up to 20 additional organic chemicals pre-
sent in each sample if the chromatogram of the sample indicated
the presence of such non-target chemicals. This requirement was to
help insure that if other organic chemicals from whatever sources
were present in the Declaration Area, they would be detected.
Sampling Strategy
The soil sampling program was based on both random and
directed sampling. Random sampling was carried out on the basis
of a grid, with grid squares measuring 440 ft on each side,
throughout the Declaration Area. In the absence of reliable
preliminary sampling data, it was not possible to develop a
statistical basis for determining the optimum size of the grid
squares. Therefore, the dimension of 440 ft was selected taking into
account .the constituency of the soil, the pattern of man-induced
disturbances of the soil and of possible pollutant pathways through
soil, soil sampling experience at other locations, and budgetary
limitations. In retrospect, the data that were acquired confirmed
that from a statistical viewpoint, the size of the grid squares was
reasonable.
-------
401
RISK/DECISION
Figure 3.
Swales in the Love Canal Area
Directed sampling was carried out:
•Along swales, or geological depressions which over the years had
been filled in and which were suspected of providing surface run-
off pathways for pollutant migration (Fig. 3)
•In areas associated with wet basements which were believed to be
areas of more heavily contaminated soil
•In a known sand lens near the Canal which might have provided
a preferential pathway for chemical migration
To provide a basis for statistical comparisons, samples were
taken in three areas: (1) in the area immediately adjacent to the
Canal, (2) in the Declaration Area, and (3) in several control areas
one to two miles from the Canal. Also, a few soil samlples were
taken at "base" sites throughout the Declaration Area where
multimedia sampling was conducted in efforts to clarify possible
transport pathways (Fig. 4).
Special soil sampling efforts were undertaken in the search for
possible problems related to dioxin and radionuclides. These ac-
tivities revealed no indication of dioxin contamination in the soil
nor radiation levels above world-wide background and are not
discussed further.
Sampling Techniques
The basic soil sampling scheme at each site is shown in Fig. 5.
Five cores, 6 ft long and 1.375 in. in diameter, were composited at
the site and subsequently analyzed for semi-volatile organic
chemicals, including pesticides, and for metals. Two additional
cores were taken at each site and immediately sealed for analysis for
volatile organic chemicals. The sampling equipment was carefully
cleaned between samples to avoid contamination, and a number of
field blanks were analyzed as a further precaution to insure that
contaminants due to sampling procedures were not misidentified as
pollutants in the soil.
A coring depth of 6 ft was selected since this depth included most
of the more permeable soil layers in the area although the area was
generally characterized by a predominance of relatively im-
permeable class. A shallower depth might have provided a better
1. Outdoor air
2. Indoor air. ground floor
3. Indoor air. basement
4. Tap water
5. Basement sump
6. Soil
7. Groundwater
Figure 4.
Multimedia Monitoring at Base Sites
picture of contamination due to surface water runoff in the area in
recent years while greater depths conceivably might have provided
better correlations between the shallow groundwater aquifer and
soil contamination. Budget and time constraints prohibited multi-
depth sampling, and composite sampling of the top six feet of soil
was considered a reasonable compromise in determining possible
pollution migration patterns as well as the pollution saturation of
the area as one indicator of future habitability of the area. Had
1.37 Inch
Diameter
KH
I Corn for chflmicali that can evaporate quickly.
' (2 separate samptos)
> Cores for other chemicals.
(5 cores which are mixed together into one sample)
Figure 5.
Soil Sampling Scheme
6
Fwt
-------
RISK/DECISION
402
preliminary sampling at different depths been possible, a statistical
basis for determining the most appropriate depth might have been
developed. In any event the soil data obtained was generally consis-
tent with data acquired by the State of New York which collected
soil samples using a trenching technique.
ANALYTICAL METHODS
Selecting the Methods
The selection of appropriate methods for the analysis of soil and
sediment samples was one of the important challenges of the Love
Canal program. Over the years, USEPA has had experience in the
analysis of soil samples for most of the pesticides and metals of in-
terest, and this experience was reflected in the analytical methods
that were chosen as indicated in Figs. 6 and 7. On the other hand,
prior to 1980 neither USEPA nor the scientific community as a
whole had much experience in analyzing soil samples for the other
organic compounds of interest, and no analytical methods had
been evaluated in interlaboratory comparative studies.
Figure 6.
Method for Pesticide Analysis
Soil
Sample
Nitric Acid
Digestion
Oxidation With
Hydrogen
Peroxide
Sb, As, Ba, Be,
Cd, Cr, Cu, Pb,
Mi, Se, Ag. Ti, Zn
Acid Treatment,
Filter. Adjust
Volume.
AA Analysis
Figure 7.
Method for Analysis of Metals
Problem:
Artifacts
Problem:
Phase Separation
Figure 8.
Candidate Methods for Extraction of Semi-Volatile Organics from Soil
The three candidate procedures that were considered for extrac-
ting semi-volatile chemicals from soil are shown in Fig. 8. Two con-
tractor laboratories were selected for a quick evaluation of these
procedures. These laboratories prepared extractions using each of
the approaches from a limited number of triplicate samples.
Analyses of the samples indicated that the steam distillation tech-
nique resulted in the formation of chemicals that were not
originally present. Also, these analyses indicated that the acetone/
hexane technique resulted in phase separations that impeded
analysis. Therefore, the extraction technique using methylene
chloride which did not exhibit comparable problems was selected as
the most appropriate.
For analyzing the extracts the gas chromotography/mass spec-
trometry method that had been successfully used by USEPA in pro-
grams under the Clean Water Act (Method 625) was selected.
However, one very significant modification of the method was in-
troduced, namely, the use of fused silica capillary columns in lieu
of packed columns. This technology had only recently become
available and had demonstrated excellent results during USEPA
evaluations. It reduced considerably the time and costs of analyses.
Three other modifications of Method 625 were the use of a high
speed mechanical stirrer, centrifuging to separate phases, and op-
tional gel permeation chromatography procedures for heavily con-
taminated samples.
With regard to analyses for volatile organic compounds,
USEPA's proposed Method 624 developed for the water program
was used. Minor modifications of the method included placing a
mixture of soil and reagent water in a vial and purging as in Method
624, but heating the mixture to 55 °C during the purge to facilitate
rapid equilibration of analyses between the sorbed and liquid
phases.
Quality Assurance/Quality Control Procedures
This aspect of the program was designed to insure that the field
collection teams and the analytical laboratories performed at the
highest scientific level possible, that all data included in the data
base met minimum acceptance criteria, and that the variability of
the data was set forth. Among the key elements of the approach
were requirements for:
•Sample collection, preservation, and holding times
•On-site sample system audits and personnel performance audits
•Analytical methods, calibrations, and control chart usage
•External analytical quality assurance programs, including the
use of EPA performance evaluation and quality control samples
•Internal analytical quality assurance programs, including refer-
ence compounds, method blanks, laboratory control standards,
laboratory duplicates, and surrogate compound spikes
•The collection and analysis of replicate field samples and field
blanks
•Splitting field samples between laboratories
The procedures worked. They uncovered deficiencies during the
program and provided opportunities to correct the deficiencies. For
example, prompt review of the data indicated that one of the
laboratories had improperly analyzed 70 samples. The data that
had been reported were rejected, and the activities of this
laboratory were halted until the problems were corrected. In
another case, laboratory contamination was identified through the
quality assurance program, and correction factors were applied to
the data.
Data Review and Validation
In addition to the data review carried out by the USEPA contrac-
tor managing the overall program, USEPA specialists also reviewed
all data. Each data point was individually "validated" by USEPA
prior to acceptance into the data base. The validation process asked
four aspects:
•Were the recoveries of the surrogate chemicals introduced into
each soil sample within acceptable limits?
-------
403
RISK/DECISION
•Were the recoveries of laboratory control standards within an
acceptable range?
•Were the holding times of samples prior to analysis within ac-
ceptable limits?
•Were there adequate adjustments in the reported data in response
to indications of background contamination in the field or re-
agent blanks that were analyzed?
Only data which met these criteria were "validated" and entered
into the data base.
The variability of the data attributable to the inherent variations
in the analytical procedures was of special interest. It was recog-
nized at the outset of the program that in applying state-of-the-art
methods to an operational monitoring program for the first time
without the benefit of interlaboratory comparative studies, there
would be considerable uncertainty in this area.
Initially, precision and accuracy of the analytical data were to be
based on the results of analyses of duplicate samples. However, the
number of positive finding was so small that this approach was not
feasible. Therefore, attention was directed to the performance of
the laboratories in analyzing standard reference materials provided
by the National Bureau of Standards. Unfortunately, soil or sedi-
ment reference materials for many of the chemicals of interest were
not available, and rigorous precision and accuracy determinations
were not possible. Nevertheless, using the limited information that
was developed, rough estimates were made of the variability of the
inorganics data attributable to the extraction and analytical pro-
cedures and were 15 to 30% and for the organic data were factors
of 2 to 3.
Also of interest were the method detection limits, particularly for
organic chemicals, since most of the data points were "below detec-
table." The recovery data for the sediment reference materials that
were available and for the surrogate spikes together with a review
of the performance of the analytical laboratories using aqueous
samples provided important insights in this regard. The best
estimates were that the instrument detection levels determined on a
daily basis by each laboratory using laboratory control standards in
an aqueous form were of the same order of magnitude as the detec-
tion capability of the methods employed to extract and measure the
same chemicals from a soil matrix. These instrument detection
limits were in the 10-20 ppb range for most organic chemicals.
FINDINGS OF THE SOIL MONITORING PROGRAM
Presentation of the Data
The findings were presented in several formats to ease the inter-
pretation process. One approach (Table 3), was to compare
statistically the extent of contamination in the Canal, Declaration
Area, and Control Areas on a pollutant-by-pollutant basis. A se-
cond approach (Table 4), was to compare statistically the concen-
tration levels of the few contaminants that were found in the three
areas. Another approach was to present in tabular form data on the
number of positive and trace findings in ten subareas of the
Declaration Area. The final, and in many respects the most in-
teresting, approach was to plot individual data points for selected
chemicals on a map of the area and then to examine carefully the
maps, individually and collectively, for possible pollution gradients
or other patterns. Two examples* f such maps are shown in Figs. 9
and 10.
T-Ti
>-Mo» Oration
N-NoA/Wyn
Figure 9.
Soil Sampling Resulis (ppb), Lindane
Figure 10.
Soil Sampling Results (ppb), Cadmium
-------
RISK/DECISION
404
Table 3.
Significant Differences Observed in Extent of
Soil Contamination at Love Canal
Percent Detect
(No. of Samples)
Compound/Element
Phenanthrene
a-BHC
5-BHC
•y-BHC (Lindane)
Heptachlor epoxide
Endrin
Decl.
23.8
(105)
8.3
(109)
10.1
(109)
6.4
(109)
0.9
(109)
9.2
(109)
Control
44.4
(9)
0.0
(9)
0.0
(9)
0.0
(9)
0.0
(9)
0.0
(9)
Canal
39.1
(23)
26.1
(23)
39.1
(23)
21.7
(23)
8.7
(23)
26.1
(23)
Comparison*
Canal
Decl.
No(a =
0.108)
Yes
Yes
Yes
Yes
Yes
Decl.
Control
No
No
•No
No
No
No
DDT
1,1-Dichloroethene
Chloroform
3-Chlorotoluene
Chlorobenzene
Cadmium
5.5
(109)
2.3
(213)
19.2
(213)
0.0
(213)
1.4
(212)
4.6
0.0
(9)
0.0
(17)
41.2
(17)
0.0
(17)
0.0
(17)
0.0
21.7
(23)
17.8
(45)
42.2
(45)
4.4
(45)
6.7
(45)
39.1
Yes
Yes
Yes
Yes
Yes
Yes
No
No
No
No
No
No
•Comparisons are based on a one-tailed difference of proportions test (a = 0.10), using Fisher's exact
test, for the areas indicated, and in the order presented.
Contamination Patterns
In general, the few patterns of soil contamination that were
observed, suggested that contaminants had migrated from the
former Canal to the immediate vicinity of certain Ring 1 residences.
These included residences that were suspected of having been sub-
jected to the overland flow of contaminants from the landfill prior
to emplacement of the clay cap over the Canal and residences that
had been constructed in the vicinity of more permeable soil
pathways. In particular, soil contamination was prevalent at soil
sampling sites located closest to the known sand lens and located at
the former major swale that crossed Love Canal.
Table 4.
Significant Differences in Levels of Soil Concentration
SUBSTANCE
LEAD
NICKEL
MERCURY
SILVER
THALLIUM
ZINC
ALPHA-BHC
BETA-BHC
GAMMA-BHC
DELTA-BHC
PERCENT OF SAMPLES
ABOVE DETECTION LIMIT/
STO. DEVIATION OF 7. /
(NUMBER OF S/MPLES)
CONTROL
100.0
0.0
( 9)
100.0
0.0
( 9)
100.0
0.0
( 9)
0.0
0.0
( 6)
0.0
0.0
( 9)
83.9
10.5
( 9)
0.0
0.0
( 9)
0.0
0.0
( 9)
0.0
0.0
( 9)
0.0
0.0
( 9)
CANAL
91.3
5.9
( 23)
100.0
0.0
( 23)
69.6
9.6
( 23)
52.2
10. 4
( 23)
4.5
4.4
( 22)
100.0
0.0
( 23)
26.1
9.2
( 23)
21.7
8.6
( 23)
21.7
8.6
( 23)
39.1
10.2
( 23)
DEC.
98,1
1.3
( 108)
100.0
0.0
( 108)
92.5
2.5
( 107)
61.7
4.7
( 107)
3.7
1.8
( 108)
100.0
0.0
( 108)
8.3
2.6
( 109)
13.8
3.3
( 109)
6.4
2.3
( 109)
10.1
2.9
( 109)
DIFFERENCE IN
PERCENT OF SAMPLES
ABOVE DETECTION LIMIT
(PROBABILITY VALUE)
CANAL -1 DEC. -
CONTROL [CONTROL
-9
(1.000)
0
(1.000)
-30
(1.000)
52
(0.026)
5
(0.710)
11
(0.261)
26
(0.111)
, 22
(0.167)
22
(0.167)
. 39
(0.029)
-2
( 1.000)
0
(1.000)
-7
(1.000)
62
(0.004)
4
(0.723)
11
(0.077)
8
(0.477)
14
(0.281)
6
(0.565)
10
(0.401)
CANAL -
DEC.
-7
(0.983)
0
(1.000)
-23
(0.999)
-10
(0.859)
1
(0.611)
0
( 1.000)
18
(0.025)
8
(0.249)
15
(0.035)
29
(0.002)
MEDIAN CONCENTRATION 1 DIFFERENCE IN
AND 1 MEDIAN CONCENTRATIONS
90TH PERCENTILE 1 (PROBABILITY VALUE)
1
CONTROL! CANAL I DEC. ICANAL -I DEC. -ICANAL -
1 1 ICONTROLlCONTROLl DEC.
1 1 1 1 1
190001 130001 220001 -60001 30001 -9000
220001 310001 570001(0.943)1(0.017)1(0.999)
1 1 1
i i i
1 1 1
200001 200001 220001 0
300001 260001 290001(0.589)
1 1 1
1 1 1
341 311 35! -3
62l 23001 851(0.783)
1 I 1
1 1 1
Bl 5601 8601
Bl 11001 1600K . )
1 1 1
1 1 1
Bl Bl Bl
Bl Bl Bl( . )
1 1 1
1 1 1
630001 560001 640001 -5000
920001 740001 1120001(0.996)
1 1 1
1 1 1
Bl Bl Bl
Bl 381 B|( . )
1 1 1
1 1 1
B| Bl Bl
Bl 1431 7|( . )
1 1 1
1 1 1
Bl Bl Bl
Bl 361 Bit . )
1 1 1
I 1 1
Bl Bl Bl
Bl 801 Tl( . )
1 1 1
1 1 1
2000
(0.639)
1
(0.522)
.
( . )
r
( . )
1000
(0.511)
.
( . )
.
( . )
t
( . )
( . )
-2000
(0.971)
-4
(0.968)
-300
(0.988)
.
( . )
-6000
(0.965)
( . )
( . )
.
t . i
( . i
-------
405
RISK/DECISION
However, there was no indication that contaminants had
migrated into the Declaration Area. Gradients or other patterns
which could relate contaminants to the Canal were not present.
While a few relatively high values of inorganic chemicals were
reported in the Declaration Area, they appeared to be clearly
anomalies surrounded by lower values which were generally "below
detectable" values. No evidence of Love Canal-related contamina-
tion that had migrated preferentially through former swales into
the Declaration Area was found, nor were residences with wet
basements found to have a higher degree of contamination than
"dry" residences.
Comparison with Control Areas
Neither the extent of soil contamination nor the concentration
levels observed in the Declaration Area were statistically different
than the extent or levels of contamination in the Control Areas.
The presence of a number of heavy metals in both Areas, such as
mercury, zinc, copper, and lead, was consistent with findings in
soils in many other parts of the country. Similarly, the occasional
occurrence in soil of pesticide contaminants associated with
household applications is not unusual in many residential areas.
Consistency with Other Aspects of the Study
The findings of the individual media monitoring efforts were
generally consistent and mutually reinforcing. Specifically, the fin-
dings of the shallow aquifer and sump monitoring efforts were con-
sistent with the soil findings. Each indicated probable contamina-
tion within Rings 1 and 2 directly attributable to the Canal within
the Declaration Area. Within Rigs 1 and 2 the highest values for the
individual media tended to be centered at the same residences which
were located along suspected pathways through permeable soils.
Similarly, the findings of the hydrogeology program
demonstrated that there is little potential for migration from the
Canal into the Declaration Area. Specifically, the geological in-
vestigations indicated that the extensive clay deposits with very low
permeability throughout the area offer little opportunity for pollu-
tant movement through soil. Also, the shallow aquifer meanders
very slowly with little opportunity for distant transport of
pollutants. There is no evidence that this aquifer is hydraulically
connected with the bedrock aquifer which does move with
somewhat greater speeds. Finally, man-made barriers such as elec-
trical and sewer conduits and road foundations effectively block
lateral migration of pollutants through the shallow aquifer or soil.
CONCLUSIONS
The soil monitoring data clearly support the principal conclu-
sions of the overall study, namely, contamination which had
migrated from the Canal into residential areas was confined to
localized portions of the area within Rings 1 and 2. There was no
evidence that the soil in the Declaration Area contained pollutants
which had migrated from the Canal. In addition, it is unlikely that
undetected contamination exists in the Declaration Area above the
detection levels, because the target compounds and many of the
sampling sites were intentionally selected to maximize the pro-
bability of detection of both individual pollutant points and migra-
tion patterns.
The soil data are consistent with the information obtained from
the hydrogeological program and the groundwater monitoring ef-
fort. This information indicates that given the subsurface structure
of the area and continued effective operation of the existing barrier
drain system around the Canal the lateral migration of con-
taminants through the overburden will not occur in the near term
and there is little likelihood of lateral or vertical migration over
the long term.
The anomalous data points of relatively high values found in the
Declaration Area cannot be explained by migration from the
Canal. According to local residents there were cases of soil from the
Canal being used as fill in earlier years, and this might explain some
of the anomalies. Also, habits of individual consumers in using
household chemicals and in maintaining their residences are known
to result in different types of contamination.
Finally, with regard to the environmental quality of the Declara-
tion Area, the soil is no mote contaminated than the soil in the
Control Areas. This level of contamination which is in large
measure attributable to naturally occurring chemicals is probably
little different than contamination levels in many other industrial-
ized areas of the country.
REFERENCES
1. USEPA Environmental Monitoring at Love Canal, Volume I, II, and
III, USEPA-600/4-82-030a, May 1982.
2. USEPA Environmental Monitoring at Love Canal Interagency Re-
view, Office of Research and Development, May 1982.
3. JRB Associates, The Groundwater Monitoring Program at Love Canal,
Sept. 1981.
4. Technos Inc., Geophysical Investigation Results, Love Canal, New
York, Dec. 1980, Project Number LC-80-128.
5. GeoTrans, Inc., Final Report on Groundwater Flow Modelling Study
of the Love Canal, New York, Jan. 1981, Project Number LC-1-619-
026-14-FR-OOOO.
6. GCA Corporation, Love Canal Monitoring Program, GCA, QA/QC
Summary Report, Jan. 1982 (draft report).
7. GCA Corporation, Quality Assurance Plan, Love Canal Study, Pro-
ject Number LC-1-619-206, (draft report).
8. Federal Register, 44, No. 233, Dec. 3, 1979, p. 695532 (For Method
624).
9. Federal Register, 44, No. 233, Dec. 3, 1979, p. 69540 (For Method
625).
-------
USES AND LIMITATIONS OF RISK ASSESSMENTS IN
DECISION-MAKING ON HAZARDOUS WASTE SITES
IAN C.T. NISBET, Ph.D.
Clement Associates, Inc.
Arlington, Virginia
INTRODUCTION
Many decisions on the management or cleanup of hazardous
waste sites require a more or less explicit balancing of costs against
risks. For example, the decision as to whether or not to clean up an
uncontrolled hazardous waste site that is leaking involves weighing
the costs of cleanup against the risks resulting from continued leak-
age. The decision to site a new facility in one place rather than
another involves weighing the relative costs of the two sites against
their relative risks. The decision by an operator to seek insurance
coverage for a site involves weighing the risks posed by the site
against the cost of premiums and loss control programs. Although
these and other decisions are subject to governmental regulations
and permits, the regulations themselves are established by a process
that involves at least some modicum of cost-benefit analysis. Al-
most every aspect of operating a hazardous waste site, from select-
ing the site through design, permitting, construction, insurance,
operation, public relations, and closure, to remedial cleanup ac-
tion and settlement of liability claims, requires some attempt-
formal or informal—to assess risks.
Risk assessment for toxic chemicals is a fafrly well-established
scientific procedure which requires comparison of the degree of ex-
posure of individual persons to a toxic chemical with the exposure
levejs known to cause toxic effects. The degree of exposure is cal-
culated from knowledge of the distribution of the chemical in the
environment, which is derived from dispersion models, from mon-
itoring data, or from some combination of the two. Toxicity in-
formation is derived, where possible, from human experience, but
more frequently is obtained from experimental studies in labora-
tory animals. The exposure and toxicity data are combined with in-
formation on the size of the exposed population to yield overall
estimates of the number of people likely to be affected, and the
nature and severity of the effects.
Risk assessments for toxic chemicals are almost always difficult
to conduct and give uncertain results, among other things because
exposure is usually variable and poorly characterized, and because
toxicity information is difficult to extrapolate from animals to
humans. The problems are much greater for hazardous wastes, for
several reasons. First, the wastes are often poorly characterized
chemically, especially when they result from disposal in the "bad
old days" of the past. Even when the chemicals present in the
wastes are reasonably well known, they are usually present as com-
plex and variable mixtures, which makes it very difficult to assess
either exposure or toxicity. In uncontrolled landfills, for example,
where wastes are leaking into subsoil and groundwater, the vari-
ability of underground media makes it very difficult to calculate
rates of dispersion, and the variability of underground concentra-
tions makes it difficult to measure rates of dispersion without elab-
orate and expensive monitoring programs. The toxicity of mixtures
is very rarely measured, and is difficult to assess theoretically be-
cause so little is known about interactions and synergisms. And
without knowledge of the rate of dispersion, even the size of the
population at risk is difficult to estimate reliably.
QUALITATIVE RISK ASSESSMENT
For these reasons, risk assessments for hazardous waste sites are
frequently made only in a qualitative way. A typical qualitative risk
assessment involves a number of steps:
•An engineering survey of the site, including an assessment of the
propensity for scheduled and unscheduled releases;
•An inventory of the materials stored or disposed of at the site;
•An assessment of geological, hydrogeological and meteorological
data, to assess the propensity for transport of materials away from
the site;
•A monitoring program for groundwater, air, and other media, to
measure the ambient concentrations of chemicals being trans-
ported away from the site;
•A survey of the distribution of the human population and other
sensitive targets subject to exposure;
•A review of the toxicity of each of the major components of the
material subject to release;
•Finally, an assessment of potential risks resulting from the ex-
posure, including characterization of the uncertainty in this assess-
ment.
Unless a very detailed monitoring program is carried out, it is
rarely possible for such an assessment to be quantitatively reli-
able. In many cases, the risk assessor has to settle for much less—
often a ranking or scoring procedure for factors controlling inher-
ent hazard, release, environmental transport, and population at
risk. At best, such a procedure can yield a ranking of risk on a
qualitative or semi-quantitative scale; for example, on a scale from
1-10, or from "low" to "high". This is sufficient for many pur-
poses—e.g., for priority setting, for permitting, or for decisions on
loss control programs or insurance.
QUANTITATIVE RISK ASSESSMENT
There are at least three situations, however, in which semi-quan-
titative assessment of risks is not sufficient, and the risk assessor is
forced to venture into the risky area of quantitative assessment.
These three areas will be the subject of the remainder of this paper.
Defining Boundaries of Areas of Habitability
Once contamination has been identified as posing a significant
hazard to local residents, the decision-maker is faced with an
immediate problem: how is the boundary to be drawn between
areas of habitability and uninhabitability? Clearly, the places where
hazardous exposures have been identified are uninhabitable: if
immediate cleanup cannot be effected, the inhabitants must be
evacuated. Clearly, areas far enough from the site are habitable.
But how is the boundary to be determined? By definition, it corres-
ponds to the margin between acceptable and unacceptable risk.
This margin is difficult to determine in the best of circumstances,
but in circumstances of multi-chemical contamination the diffi-
culty is compounded by the lack of information about both ex-
posure and toxicity. Initial decisions on evacuation usually have to
be taken at short notice, under conditions of public pressure or
panic. However, subsequent decisions may not be much easier, as
experience at Love Canal illustrates. Although the U.S. Environ-
mental Protection Agency spent over $5 million on an extensive
multi-media monitoring program to document its finding that the
"Declaration Area" was habitable after remedial work at the dis-
posal site, this finding was vigorously disputed by other interested
parties. It is noteworthy that this decision was not based on a
formal assessment of risks, or exposures, or even of transport of
406
-------
407
RISK/DECISION
chemicals away from the site. Indeed, EPA's report suggests that
the data were not sufficient to make quantitative assessments,
despite the scope and expense of the monitoring program.
Establishing the Limits of Cleanup
Once a decision is made to clean up a site, the issue immediately
arises as to the extent of cleanup. Although a large fraction of the
hazardous materials at a site may be concentrated, some are usu-
ally widely dispersed, and the marginal cost per unit of material re-
moved usually rises steeply as the cleanup progresses. For rational
allocation of resources, the marginal costs of cleanup should not
increase much beyond the marginal value of the reduction in risk
thereby achieved.
To implement this economic efficiency criterion, however, re-
quires fairly precise calculation of the marginal reduction in risk,
not to mention agreement on how to express the value of risk re-
duction in dollars. Calculation of marginal risk reductions requires
numerical estimates of risk for each degree of cleanup, which in
turn requires precise estimates of exposure and toxicity as a func-
tion of exposure. A classic illustration is provided by the contam-
ination of the Hudson River with PCBs. Despite fairly detailed in-
formation on exposure and on the toxicity of PCBs, cost-risk cal-
culations did not command sufficient acceptance to serve as the
basis for public policy decisions. EPA has still to make final de-
cisions on the degree of cleanup required for PCB spills, despite
the immediate importance of such decisions for many affected
parties.
Award of Damages in Personal Injury Suits
In the event that material leaking from hazardous waste sites ac-
tually leads to injury to individuals, their final recourse is to seek
redress through the courts. The award of compensatory damages is
a third area in which quantitative risk assessment should play an
important role. Except in cases where exposure has led to short-
term injury, clearly attributable to a hazardous waste site, the es-
tablishment of liability and the assessment of damages require the
proof of risk (i.e., probability of injury). This involves the same
process of estimation of exposure and toxicity that has proved
difficult in other contexts. Although some personal injury cases
have been settled out of court after presentation of scientific evi-
dence about the magnitude of risk, I know of none which has gone
to trial and has been resolved on the basis of such evidence.
SUMMARY
Enough has been said to illustrate the limitations of risk assess-
ment in resolving critical issues involving hazardous waste sites.
Although quantitative risk assessment is theoretically essential for
certain types of decision-making, in practice it is rarely, if ever,
used for these purposes. One practical problem is that quantitative
risk assessment requires a degree of comprehensiveness in monitor-
ing and toxicity testing that is usually prohibitively expensive. But
even where comprehensive data have been obtained, they have
either not been used for risk assessment (as in the case of Love
Canal) or the risk assessments have not been used in decision-mak-
ing (as in the case of the Hudson River). In these and other cases,
perceptions about the uncertainties in quantitative risk assessments
have led to their results being discounted, so that public policy de-
cisions were based on more traditional, more subjective weighing of
the information.
All the effort is not wasted, however. At the least, the exper-
ience gained from these attempts to make quantitative risk assess-
ments contributes significantly to their practitioners' ability .to
make the reasonable scientific judgments involved in qualitative
and semi-quantitative risk assessments. With sufficient knowledge
and experience, these assessments can be reliable and useful tools
in decision-making for most hazardous waste issues. It is impor-
tant, however, to recognize the limitations of risk assessments, and
to understand that quantitative calculations of risk are only justi-
fied if unusually extensive information is available.
-------
MULTIATTRIBUTE DECISION-MAKING IMBEDDED WITH
RISK ASSESSMENT FOR UNCONTROLLED
HAZARDOUS WASTE SITES
CHIA SHUN SHIH, Ph.D.
Division of Engineering
University of Texas at San Antonio
San Antonio, Texas
TERRY ESS
Consultant in Risk Analysis & Computer Application
Glenside, Pennsylvania
INTRODUCTION
Many modern technology related societal problems, especially
those involving risks to human health and safety are becoming
more complex and uncertain. Large amounts of data and multiple
conflicting objectives require the incorporation of subjective
judgments. In case the adversary positions (i.e., a politicized con-
flict) exist, it usually becomes a problem so politically volatile that a
decision maker must rely on more than just intuition. The problem
area of sludge and compost utilization is just such a situation. It
calls for a systematic technique which would allow decision makers
to adequately address the complex issues involved and develop
viable solutions based on both objective analysis and subjective ad-
justments.
Decision analysis that incorporates a multiattribute utility func-
tion is an apparent, effective tool for this type of problem since it is
highly flexible, incorporates methods to handle uncertainty, and
multiple objectives and is also a well developed technique.
However, it is not without faults. The most glaring, as pointed out
by Rose,3 is its inability to properly treat the subjective nature of
risk. Thus, an integrated method incoiporating multiattribute deci-
sion analysis with a quantitative risk assessment technique is needed
to handle such problems as sludge and compost utilization. The
basic steps entailed in such a unified approach include:
•Construct a decision tree for the specific situation
•Complete a detailed risk analysis
—Determine the objective risk of each decision branch
—Determine the appropriate risk referents
—Use an objective risk versus risk referent comparison to de-
termine if a decision branch requires modification or should
be eliminated from consideration
•Conduct a sensitivity analysis of the risk comparison in order to
determine which branches are only "marginally" acceptable
•Complete the multiattribute decision analysis with the prunned
tree
•Conduct a sensitivity analysis of the "solution"
In the remainder of this paper, the authors explore this integrated
technique in more detail.
DECISION TREE CONSTRUCTION
The organization and construction of a decision tree is essentially
the first task in the decision making process. In general the follow-
ing steps should be utilized to construct a tree:
•Generate an objective hierarchy which terminates in the attributes
(which includes risks) and attribute measurements (a possible
hierarchy for hazardous waste problems is shown in Fig. 1)
•Determine the viable courses of action available
•Determine the possible chance events (i.e., failure events, out-
comes etc.) resulting from a decision
•Arrange the decision options and resulting chance events in
chronological order (a generalized structure for problems which
largely involve risk is shown in Fig. 2)
•Evaluate the specific probabilities for each chance event
•Evaluate the magnitude of each attribute
A number of key areas in this process require a more detailed ex-
planation. First, it is necessary to take a closer look at the meaning
of "viable courses of action". The term implies that some pre-
decision tree criteria are used to eliminate "nonviable" alternatives
from entering into the decision making process. In the hazardous
waste area, the principal criterion to serve this purpose is im-
plementation time. The discovery of an uncontrolled dump site for
highly toxic substances would in all political reality require initial
positive action which took a minimum time to implement. The
number of viable options would probably be small. In the case of
planning a new controlled disposal/storage site, this constraint
would probably be greatly reduced, allowing a much wider scope of
options to be considered.
A closer look is also required in the area of event space. A quick
glance at Table 2 would cause some to think that what is being
discussed is one "success" event and a few (one or more) "failure"
event(s). This is of course far from the truth in anything but the
most simplistic cases.
In most "real-life" type problems the failure event shown cor-
responds to the top event of an appropriate fault tree. A fault tree
is another type of analytical tree which allows one to depict the
Protect biological
systems (especially
man) in the most
efficient acceptable
manner. .
Acceptable
protect ion
Efficient
manner
Human
effects
Non-human
effects
f-
Risk
fa al-
i t ies
Aeste-
tics
acres
set aside
for waste
Lead
time
(yr.)
# induced anomalies
m indicator species
(animal £ plant)
Figure 1.
Sludge & Compost Utilization Objective Hierarchy
408
-------
409 RISK/DECISION
D«cl«ion Baiard Outcoa* Kxporar* COBM- Attribute
qoanc* Magnitude
o
Figure 2.
Generalized Decision Tree Structure
Table 1.
Possible Risk Classification Scheme*
Risk Description
ediate
Catastrophic
Involuntary
Regulated Voluntary
Ordinary
Involuntary
Regulated Voluntary
Class of Consequence
Fatalities Morbidity Property
Damage
Delayed
Catast
Invo
Regu
Ordian
I nvo
ophlc
untary
ated Voluntary
untary
Regulated Voluntary
Table 2.
Objective to Subjective Factors Summary1
Factors involving type of consequence
•Voluntary or involuntary (1)
•Discounting of time (2)
•Controlabilily (2)
Factors involving nature of consequence
•Position in hierarchy of consequences (1)
•Ordinary or catastrophic (1)
•Natural or man originated (1)
Other factors
•Magnitude of probabilily of occurrence
•Propensity for ri*.k. taking (1)
in htpluiiK mJuJcd in deicrminjuon ol absolute risk reference*
l2i txpluisU truluJed in Jcicrmmjticn ol rnk rcfcrenl
logical interrelationships between basic, events that lead to a
undesired event (i.e., the failure event). Using fault trees, Boolean
algebra and various statistical techniques,' it is possible to deter-
mine failure event probabilities. In many cases the probabilities
determined are only order of magnitude estimates. The success
events shown in Fig. 2 represent the summation of all possible suc-
cess events for a specific decision path. In practical terms the pro-
bability of this success is equal to one less the sum of all the com-
puted failure event probabilities for the decision path.
This discussion leads right into the next area of decision tree con-
struction which requires elaboration, the significant potential inac-
curacies in both probability and attribute assessments. As indicated
in the prior paragraph, even when "objective" methods are used
the information obtained may be only order of magnitude. When
value judgments are involved, which is often the case in these
assessments, then even more inaccuracies can be expected. In fact,
the level of accuracy can often be so low that a single
"correcf'answer cannot be determined. This does not negate the
value of using quantitative techniques but should caution one to
refrain from making unjustifiable claims of accuracy. Because of
this, it is almost as important to obtain variance information about
assessments as it is to know their location. This variance informa-
tion especially becomes important in the subsequent step of sen-
sitivity analysis.
RISK ANALYSIS
As indicated by Rowe,3 risk cannot be meaningfully analyzed in
an aggregate; it must be differentiated into appropriate classes. A
possible classification scheme f<5r hazardous waste problems is pro-
vided in Table 1 . After agreeing on an appropriate scheme, the ob-
jective (modeled) magnitude of risk of each class on each decision
tree must be determined. In general this would be formulated in the
following manner for each decision branch when a decision tree
format is used:
= E(PCij)(aij)/ET
(1)
where:
j
a specific path in a decision branch
PC|J conditional probability along path j
Aj a specific risk class
3jj the consequence magnitude of risk Aj on path j
(i.e., number of fatalities, etc.)
E total population exposed to risk A,
T time in years
(Note: This provides a measure of risk which is commensurate with
the manner of data presentation used by Rowe.)
The next step is to determine appropriate risk referents. The pur-
pose of a risk referent is to serve as the measure of risk acceptability
(incorporating subjective perspectives). Rowe has proposed a
methodology of calculating risk referents which is composed of two
steps:
•Using historical data as a base, absolute risk references are de-
termined for each risk class
•These absolute references are modified to fit the specific situation
being analyzed producing risk referents
A clearer understanding of the details of this process can be gain-
ed by examining Table 2. In this table, the primary factors mention-
ed by Rowe as effecting the subjective (i.e., public) perception of
risk are summarized and the step in which they are explicitly includ-
ed in the calculation of risk referents is indicated. Rowe's assess-
ment of the accuracy of this method indicated potential variances
in excess of one order of magnitude, the area of prospect theory'
provides a formalized explanation of the psychological rationale
behind the determination of risk referents and provides some hope
for future refinement of this process. At present, however, this is
only a hope.
The final phase of the risk analysis can now be accomplished.
This is nothing more complicated than a comparison of a decision
branch's risk (for each class of risk) with the appropriate risk
-------
RISK/DECISION
410
referent. If the risk does not exceed the referent by more than one
order of magnitude then the risk is considered publicly acceptable.
This order of magnitude comparison is used due to the inherent in-
accuracies in the risk information used. If this criterion cannot be
met then the choice remains to either modify the decision branch
(which will probably effect some or all of the non-risk attributes be-
ing considered such as cost) or to eliminate that branch. One of the
modification steps involves, when the difference between objective
and subjective risk is large, the education of the public as to the
"actual" risk. In many cases, this may be a difficult, costly and
time consuming path but in some cases it may be the only option
other than complete abandonment of a project.
MULTIATTRIBUTE DECISION ANALYSIS
This process can be broken into two principal parts:
•The development of a multiattribute utility function
•The calculation of expected utilities
The first is by far the more complicated procedure. It basically
involves the assessment of attribute independence and then the
development of appropriate mathematical formulations which
allow the analyst to combine the utility function. A simplified sum-
mary of the independence assumptions and resulting formulations
in multiattribute utility theory is provided in Table 3. The concept
of preferential independence involves attributes under the condi-
tion of certainty while utility independence is specifically concerned
with uncertainty. This process is developed according to the deci-
sion makers' perspective of the attributes not according to some set
standard rules. Some general observations about the independence
perceptions likely to be held by most decision makers in hazardous
waste problems seems in order, though. First, the risk attributes
normally will be both preferential and utility independent of the
other attributes. Second, each risk attribute would normally be
both preferential and utility independent of each other.
The second part of the process, the calculation of expected
utilities, is basically just a mechanical process of "averaging out"
and "folding back". Averaging out involves nothing more than the
computation of £ PcU (U-utility magnitude) for a decision node.
Folding back entails the elimination of the less desirable paths at a
decision node. Unlike most of the other aspects of the integrated
analytical procedure, this portion of the process is purely "objec-
tive".
Table 3.
Summary of Multiattribute Utility Theory1
Independence Definitions:
Preferential Independence (PI)—attribute X is PI of attribute Y if
preference for consequences (x,y2) with y2 held fixed do not depend on the
amount of y2
Utility Independence (UI)—attribute X is UI of attribute Y if preference for
lotteries on x,y2) with y2 fixed do not depend on the amount of y2
Additive Independence (AI)—attribute X and Y are AI preferences for lot-
teries (x,y) depend only on the marginal probability distributions on x and y
Utility Formulations:
IfXiUIXi, i = 1,2,.., nthen
u(x) = £kjUj(x) + kjj U, (Xi)Uj(Xj) + ... + k,...n nu, (x,)...un(xn)
If X ,Xi PI Xy.i =2,3..., n and X! UI X,, then either
1(1 + ku(x) = n 1 + kkjU^Xi) or
2)u(x) = Ekiui(xi)
IfXi AlXi, i = l,2..., no then
u(x) = Ekiui(xi)
SENSITIVITY ANALYSIS
Sensitivity analysis is used at two points in this integrated pro-
cedure: (1) after the risk analysis, and (2) after completing the deci-
sion analysis. A sensible question at this point is what is sensitivity
analysis and why is it necessary? Sensitivity analysis is a post-
solution technique, intended to indicate how much trust can be
placed in solutions, when one knows that all or most of the
parameter values (probabilities, attribute assessments etc.) are not
certain.
A significant level of inaccuracy is inherent in the type of pro-
blems being confronted here. Therefore, sensitivity analysis has a
very important part-to play in this analysis. Using this technique
after risk analysis allows one to ascertain if any of the "acceptable"
decision branches is really only marginally so. One would have a
great deal of doubt if one of these marginally acceptable branches
was the solution or a part of the solution chosen during decision
analysis. In that case, one would probably be inclined to look at the
problem with even greater intensity.
The same type of function is provided by the sensitivity analysis
conducted after decision analysis. Sensitivity analysis does not have
any rigid rules about what specific techniques to use. Within the
context of the problems being addressed, there are two possibilities.
The first is simply to change the value of selected parameters and
see if the "solution" arrived at differs. A more useful approach in-
volves the substitution of variables for selected parameters (nor-
mally one at a time), then solving the decision problem in terms of
the variable. By changing the variable, one can graphically depict
the effect it has on the desirability of each possible decision. This
allows one to quickly determine which parameters are really signifi-
cant and therefore require close attention.
APPLICATION
In order to illustrate the use of the integrated methodology, the
case study of Denny Farm 1' will be utilized. In this case an uncon-
trolled chemical dump site containing TCDD along with other
substances was discovered in Missouri. The problem posed is how
best to eliminate the health risk at and around this site. Four viable
alternatives were suggested: (1) leave the site as is, (2) install and
maintain a ground water monitoring system, (3) excavate the dump
and restore in a controlled manner on site, and (4) excavate the site
and transport liquids and residues via truck to Syntex, an approved
hazardous waste storage site. Since joint probabilities for each
chance path have been delineated in the study, the simplified deci-
sion tree shown in Fig. 3 is used. Due to the lack of information
provided by the study, only four attributes can be used: two classes
of human risk, fatalities and morbidity, cost and lead time. The
division of human impacts between fatalities and morbidity is made
possible by assuming that 1.0% of all possible harmful human ex-
posures determined in the study will result in fatalities. In reality,
this assumption would have to be verified and revised as necessary.
The next step is to proceed with a risk analysis. The risk data pro-
vided by the study are summarized in Table 4. In order to use these
data, it is necessary to express them in terms commensurate with
the risk referents that will be calculated. This requires that a time
duration in years for long term risk be established and that
estimates of the total population exposed be determined. For the
purposes of this paper, the long term risk duration is assumed to be
evenly distributed over a span of 30 years.
Estimates of the total exposed population were calculated using
data provided in the study. In some, but not all cases, this could be
reasonably assumed to be equal to the maximum exposed numbers
derived in the study. The objective risk was then calculated using
the equation provided in the "Risk Analysis" section of this paper.
These risk estimates are summarized in Table 7.
With objective risk determined, the next step was to calculate
risk referents. This was accomplished using the procedure outlined
by Rowe.3 In order to do this some reasonable assumptions about
the public's and worker's perception of the indirect gain-loss
balance and the controllability of each alternative had to be made.
These assumptions are shown in Table 5. A summary of the factors
-------
411
RISK/DECISION
BUBU Rl»k Cmt Lead Time
Fatal Morbid. (10 S) (no)
Table 5.
Risk Referent Assumptions
i .0' u uno
^-~___ , 120
.£—-—-" ° °
< jj™-" i wo
- 1 120
^^ ° °
J&T.^ ° 3
^^-- ° 50
>^^-i£i___ 71 70
^^^Sp"?-- ° *°
^\ 1 120
* 0 0
j££ '• i
IgS^ i
^fe^ \ i»
x^ ,• ,
\ 0 10
Figure 3.
Initial Decision Tree
Table 4.
Denney Farm Risk Data Sumn
0
0
0
.75
.75
.75
3.0
3.0
3.0
3.0
3.0
3.0
3.S
3.5
3.5
3.5
3.5
3.5
3.5
3.5
0
0
0
3
3
3
a
8
8
8
8
8
10.5
10.5
10.5
10.5
10.5
10.5
10.5
10.5
Indirect gain-loss balance:
Class Alt Balance
Involuntary 1 Marginally unfavorable
(public) 2 Indecisive
3 Marginally favorable
4 Favorable
Regulated Voluntary
(workers)
Controlability:
3/4 Favorable
Value
0.001
0.01
0.1
1.0
1.0
Control Degree of Stale of Basis of
Alt Approach Control Implement. Effect.
1 0.1
2 0.3
3/4 1.0
Risk Classification
Involuntary, catastrophic,
fatal
Involuntary, ordinary,
fatal
Involuntary, catastrophic,
health effect
Involuntary, ordinary,
health effect
Regulated voluntary,
ordinary, fatal
Regulated voluntary.
0.1
0.3 0.5 0.5
1.0 1.0
Table 6.
Risk Referent Calculation Factors
Risk Ref. Risk Proportion.
Prop Derating
F«c. Factor
\xtQ-~1 O.I All 1 0.001
Alt 2 0.01
5x10 ~6 0.1 Alt 1 0.001
Alt 2 0.01
Alt 30.1
Alt 4 1.0
SxlO"7 0.1 All 1 0.001
Alt 2 0.01
3x10-' 0.1 Alt 1 0.001
Alt 2 0.01
Alt 30'.l
Alt 4 1.0
IxlO-4 1.0 1.0
6xlO-2 1.0 1.0
Control.
Factor
Alt 1 0.01
Alt 2 0.015
Alt 1 0.01
Alt 2 O.OIS
3/4 1.0
AIM
All 1 0.01
All 2 O.OIS
Alt 1 0.01
Alt 2 O.OIS
3/4 1.0
1.0
1.0
lary Note: All risks are treated as immediate
Joint
Invol.
Probability Fatal
Alternative
1 . Leave buried
2. Install & maintain a
grounduater mom-
lormg system
3- Excavate & store
material on site
4 Excav ale & transport
liquids and residues
via truck lo Syntex
(Term)
0.01
(long)
0.9
(long)
3.3x10-"
(long)
0.45
(long)
0.2
(short)
3. 2x10 "5
(short)
0.04
(long)
O.I
(short)
2.5xlO"2
same as
3 plus:
3 5x10- '
(short)
3. 5x10 -1
14
1
4
1
1
0
_-
1
0
Invol Reg.
Morb. Fatal.
1400
120
370
120
50
70
0.04
120
10
Vol.
Morb.
...
3
40
...
t
used in the risk referent calculation is shown in Table 6. A com-
parison of objective risk and risk referents is provided in Table 7. A
quick glance at this comparison indicates that none of the proposed
alternatives is "acceptable" to the public and that alternatives 3
and 4 are only marginally "acceptable" to the workers in terms of
fatalities.
Alt
1
2
3
4
4
Mod
Table 7.
Risk Comparis
Risk Classification
Involuntary, catastrophic, fatal
Involuntary, ordinary, fatal
Involuntary, catastrophic, health
Involuntary, ordinary, health
Involuntary, catastrophic, fatal
Involuntary, ordinary, fatal
Involuntary, catastrophic, health
Involuntary, ordinary, health
Involuntary, ordinary, fatal
Involuntary, ordinary, health
Reg. voluntary, "ordinary, fatal
Reg. voluntai-y, ordinary, health
Involuntary, ordinary, fatal
Involuntary, ordinary, health
Reg. voluntary, ordlanry, fatal
Reg. voluntary, ordlanry, health
Involuntary, ordinary, fatal
Involuntary, ordinary, health
Reg. voluntary, ordinary, fatal
Reg. voluntary, ordlanry, health
on
Objective
Risk
3.3xlO'6
3.0x10"*
3.0xlO~*
2.7xlO"2
l.lxlO"7
1.5x10'*
9.9xlO"6
1.3xlO"2
1.7xlO"5
l.SxlO'3
3.2x10"*
3.3xlO"2
2.5xlO"7
l.SxlO"3
3.2x10"*
3.3xlO"2
4.1xlO'8
1.7xlO"7
3.2*10"*
3.3xlO"2
Risk
Referent
l.OxlO"'3
S.OxlO'12
S.OxlO"12
3.0X10"'1
1.5X10-'2
7.5,10-"
7.5X10-'2
4.5xlO-'°
S.OxlO'7
3.0xlO'7
l.OxlO"3
e.oxio'2
S.OxlO"7
3.0X10'7
l.OxlO"3
6.0»10"2
S.OxlO"7
3.0»10"7
l.OxlO"3
e.oxio'2
-------
RISK/DECISION
412
3
Mod
Table 8.
Sensitivity Analysis
Risk Classification j Objective Risk
i Risk Referent
1
Involuntary,
fatal
Involuntary,
fatal
Involuntary,
health
health
catastrophic.
ordinary,
catastrophic,
8.4xlO~6 7.5xlO"5
7.5xlO"5 B.OxlO"3
7.5xlO~5 S.OxlO"4
A IT
7 6x1 0 -3 3x 1 0~
l.OxlO"14
S.OxlO"13
S.OxlO"14
3 Oxl 0
- l.OxlO"1
- 5.0x10"'
- 5.0x10"'
- 3 .0x1 0
Involuntary, ordinary,
fatal !
Involuntary, ordinary 3.9x10"" - 2.8xlO"J
hea 1 th
Reg. voluntary, ordinary, i 7.9x10" 5 2x10
fatal I
Reg. voluntary, ordinary, j 8.3x10" - 7.5x10"
hea I th
Involuntary, ordinary,
fatal
Involuntary, ordinary,
health
Reg. voluntary, ordinary,
fatal
Reg. voluntary, ordinary,
health
-5
5.0x10" - 5.0x10"
3.0x10 - 3.0x10"
1.0x10" 1.0x10"
1.0x10"° - 9.0x10
9.0xlO"8 - S.OxlO"5 j
Human Risk Coat Lead Time
Fatal Morbid. (10 $) (mo)
0
3
SO
40
3.0
3.0
3.0
3.0
3.0
Figure 4.
Revised Decision Tree
Unfortunately, no information was provided in the study to
enable a sensitivity analysis. Illustrating this procedure will require
that some reasonable variance information be postulated. It will be
assumed that:
1. The parameters used in the decision tree, joint probabilities and
attribute magnitudes, can vary by 50% within the limits imposed
by the exposed population size and the maximum possible
probability of 1.0.
2. The risk referents can vary by one order of magnitude.
1.0-
00
oS
ai
8
00
.5 .6-
.2-
j 0 0
,// ° 3
£/ /
//*&- l
^^
~----
X*i\
X^-,^ 0 2
1 9
3.
3.
3.
3,
3.
3.
3.
5
5
5
5
5
5
5
10.
10.
io.
10.
10.
10.
10.
5
s
5
S
5
5
5
J.
S
i
QJ
O
e
J
o
BU
.u-
. 8-
.6-
.4-
.2-
.8-
.2-
5,0 10.0
Fatalities
Figure 5.
Fatalities Utility Function
15,0
30
60
Morbidity
90
120
Figure 6.
Morbidity Utility Function
1.0 2.0 3.0 4.0
Cost ($ Millions)
Figure 7.
Cost Utility Function
Using these assumptions, the information in Table 8 was
calculated. The data show the full range of variance in both objec-
tive risk estimates and risk referents which could be possible. This
provides a sound basis for assessing which alternatives have a
realistic chance of being acceptable.
-------
RISK/DECISION
Lead Tine (Bpnthi)
Figure 8.
Lead Time Utility Function
Relative Heights .38
Human Risk Cost Lead Total
Fatal Morbid. 110 S) Time
.12
(no)
.25 .25
.3 .1
.3 .1
.3 .1
.3 .1
.3 .1
1.0
.6
.6
0.54
0.57
.10
0
0
0
0
0
0
0
0
0
0
0
0
.5
.5
0.44
0.47
0.36
.50
0.95 .92 0 0 0.47
Figure 9.
Problem Solution
At this point it is fairly clear that alternatives 1 and 2 will, under
no foreseeable circumstances, approach an acceptable level of risk
so they can be eliminated. It appears potentially feasible and worth-
while to modify alternatives 3 and 4 by eliminating the risk to the
public of significant release of residual TCDD after excavation, by
either eliminating most of this residual during the cleanup or by
some form of encapsulation. This modification will of course cause
both the cost and lead time of each alternative to be increased. No
other risk modifications seem realistic with the information given.
As indicated in Tables 7 and 8, with the suggested modification
alternatives 3 and 4 would become acceptable. This suggests that it
is time to seriously consider the modified approach or assemble
some other alternatives. Possible other approaches would need to
include means to either mitigate or eliminate the pubb'c exposure
due to tornados and contaminated workers. Since no in-depth in-
formation is provided, assumptions that alternatives No. 3 and 4,
as modified above, are the best that can be developed for further
analysis. In real life this is possibly an outcome that could
materialize, in which case the public must be informed in detail.
With the risk assessment completed as above, a revised decision
tree including only alternatives 3 and 4 (modified) is needed for fur-
ther analysis. Such a tree is illustrated in Fig. 4. However comple-
tion of the formal analysis required utility functions for each in-
dividual attribute and a multiattribute utility function. The utility
functions used for each attribute are diagrammed in Figs. 5 to 8.
Figs. 5 and 6, fatalities and morbidity, provide an example of risk
neutral functions while Fig. 7, costs, is slightly risk prone and Fig.
8, lead time, is risk averse. These functions would be a direct result
of the perceptions of the decision maker for the problem. The
original cost and lead time for the modified version of alternatives 3
and 4 are being used. Due to the similarity of the two options and
the likelihood that the changes would be relatively minimal, this
should not create any deviations for the resolution of data. In order
to combine the four separate utility values for different attributes
into a single utility value, a multiattribute utility function is re-
quired. As in the case of the individual utility functions, this
multiattribute utility function is directly related to the perceptions
of the decision maker in question. For the purposes of this paper
the use of an additive function seems reasonable due to the general
observations mentioned previously and the small relative difference
between the two alternatives being considered.
Finally, expected utility for each path is calculated using the
mechanics of decision analysis to arrive at a "solution". The deci-
sion tree with all calculated values is diagrammed in Fig. 9. Alter-
native 3 (modified) is obviously preferable to 4 (modified) in this
particular case. Since both alternatives are so similar it is unlikely
that sensitivity analysis would indicate any significant different
choice within any reasonable limits.
CONCLUSIONS
The proposed method provides a quantitative tool that is system-
matic but flexible, is capable of handling uncertainty, multiple con-
flicting objectives and the subjective judgments of decision makers,
and addresses the highly subjective nature of public risk precep-
tion. With the prudent use of a "viable option" criteria and risk
assessment, the number of options that must be fully considered
can be effectively limited. On the other hand, this approach will
help pinpoint requirements for considering a wider -scope of alter-
natives when necessary.
The process is obviously powerful and, also, appropriate for use
with multiattribute problems. Given the potential complexities in-
herent in sludge and compost utilization the use of this technique
appears justified. This method of problem solving does not try to
eliminate the subjective judgments, but does provide the judgments
an opportunity of being scrutinized. It is the authors' attempt to
bring together a set of powerful tools and apply them to the site
management of uncontrolled hazardous wastes. More refinement
and improvement will still be required to enable the application of
this method to larger scale problems.
REFERENCES
1. Buchanan, J., et al., "Technical Study and Remedial Action for Den-
ney Farm Site 1, Auroria, Mo. (Final Report)". Ecology and Environ-
ment, Inc., 1980.
2. Keeney, R. and Raiffa, H. Decisions with Multiple Objectives: Prefer-
ences and Value Tradeoffs. John Wiley & Sons, 1976.
3. Rowe, W., An Anatomy of Risk, John Wiley & Sons, 1977.
4. Shih, C., "Decision Analysis and Utility Theory", Seminar on Risk
and Safety Assessment, HMCRI, 1981.
5. Tversky, A. and Kahneman, D., "The Framing of Decisions and the
Psychology^ Choice." Science, Jan. 30, 1981.
6. Vesely, W., et al.. Fault Tree Handbook, U.S. Nuclear Regulatory
Commission, NUREG-0492, 1980.
-------
U.S. ARMY CORPS OF ENGINEERS ROLE IN
REMEDIAL RESPONSE
BRIGADIER GENERAL FORREST T. GAY, III
NOEL W. URBAN
JAMES D. BALLIF
U.S. Army Corps of Engineers
Washington, D.C.
INTRODUCTION
The USEPA under Executive Order 12316 was assigned primary
responsibility for implementation of the Comprehensive Environ-
mental Response, Compensation, and Liability Act of 1980 (i.e.,
CERCLA or Superfund). The Superfund program consists of two
parts: (1) emergency response (removal action) to hazardous sub-
stance spills and uncontrolled sites, and (2) remedial response to
cleanup problem sites. Remedial response consists of the follow-
ing four major activity phases:
•Investigation of the problem
•Feasibility study to select the most effective and cost efficient
cleanup alternative
•Final design of cleanup action
•Implementation (construction) and related tasks
States may elect to manage and direct all or part of the remedial
response activities, otherwise, EPA will take the lead. In either
case, the state in which the site is located will be required to pro-
vide 10% or 50% cost-sharing.
The Army Corps of Engineers Agrees to Support USEPA in
the Management of Superfund Work.
The USEPA and the Army Corps of Engineers signed an inter-
agency agreement on February 3, 1982. Under the agreement,
upon USEPA request, the Corps of Engineers will manage de-
sign and construction contracts and provide technical assistance to
USEPA in support of remedial cleanup of hazardous waste sites.
USEPA has a three-tiered process that will determine the extent
of Corps assistance under Suprefund. Under this process, USEPA
will: (1) determine whether a private entity is liable for cleanup
and approach that entity to perform the necessary tasks; if that
does not develop, then (2) determine whether the state can/will
do the cleanup; if not, then (3) determine that Federal cleanup is
appropriate and request that the Corps undertake design and con-
struction.
CORPS RESPONSIBILITIES UNDER THE INTERAGENCY
AGREEMENT
The Corps of Engineers primary responsibilities under the in-
teragency agreement are as follows:
Serve as contract manager for design and construction
•Review Designs
•Monitor Construction
•Provide Technical Assistance to EPA
•Review State Plans upon EPA Request
Superfund Site Management Plan
Prel.miOHfy
rwesligalion
Voluntary
Action tiy the
Restjonsible Parly
/
rCorps Mission Assignmenl
EPA Assigns Projects and
Provides Funds
C T A Corps Technical Assislance
C H - Corps Responsibilily
Figure 1.
Superfund Site Management Plan
414
-------
415
STATE PROGRAMS
At sites where USEPA has primary responsibility for cleanup
(Federal lead), the Corps will contract out and manage actual de-
sign and construction work, once a remedial concept is approved
by USEPA and the Corps (Fig. 1). Overall program guidance,
policy, and funding for Corps support will originate with USEPA.
The Corps will provide technical assistance to USEPA, as
needed, during the remedial investigation and feasibility study
phases. This assistance will be of necessary scope to assure that
the proposed remedial action selected by USEPA can be engineered
and constructed. The Corps will also assist USEPA in the review
of projects undertaken by the states as to their suitability for bid-
ding and construction. In any case, USEPA will not assign a
remedial action to the Corps for management if the Corps de-
termines that the action is not reasonable to design, construct,
operate, and maintain.
ACTIVITIES OF OTHERS
USEPA, the states, local interests, or others will be responsible
for the following:
•Establishing Priorities
•Selecting Sites
•Cost Recovery
•Public Involvement
•State Assurances
Maintenance
Cost Sharing
Disposal Sites
Environmental Impact Statements
Obtaining Permits
Legal Determinations
Obtaining Real Estate Rights
CORPS OPERATING PRINCIPLES
The Corps of Engineers, in executing its EPA Superfund
assignment, has adopted the following operating principles:
Be Responsive to USEPA's Program Needs
•Timely Response
•Single Point of Contact Concept
•Support USEPA's Cost Recovery Program
Provide Highest Technical Competence in Engineering and
Construction; Provide Assertive Technical Assistance
Adopt Highest Standards in Safety and Health Aspects
Provide Cost Control
•Cost-Effective Solutions, Design, Construction
•Implement as Many Sites as Possible with Available USEPA
Funds
Support EPA's Community Relations
•Promote Public Confidence in Corps and USEPA
Streamline Corps Management Structure
•Design Center-Centralized Engineering Resource
•Construction Managed by Lead Districts Along State Lines
•Maximum use of Private Sector
THE CORPS MANAGEMENT STRUCTURE
The Corps will utilize its existing nationwide decentralized man-
agement structure, integrating its Superfund responsibilities into
the existing program. The Chief of Engineers designated the
Director of Civil Works to perform executive direction and man-
agement activities in a similar manner to the Corps traditional civil
works missions, except for the absence of a direct interface with
the Office of Management and Budget, and the Congress, on bud-
get and authorization activities. The Office of the Chief of Engi-
neers (OCE) interfaces with EPA in determining Superfund issues,
polio, funding, priorities, research needs and national program
direction within the interagency agreement. OCE will assign pro-
jects to Corps division engineers, provide field guidance, perform
program management activities, conduct Washington-level pro-
gram reviews and coordination, and provide design and construc-
tion oversight. The Chief, Engineering Division, OCE has been
assigned programmatic responsibility by the Director of Civil
Works.
Project Management and Program Coordination
The Chief, Engineering Division, OCE, is the Corps national
program manager. The project management and program coor-
dination function is assigned by state and will be based on the loca-
tion of the hazardous waste site in relation to the state-lead
district assignment (Table 1). The Division Engineers, through
their staff designees, are regional Corps program managers re-
sponsible for overall project and program management, providing
a project manager for each hazardous waste site designated by
USEPA.
Design Review
Nationwide, the Missouri River Division (MRD) Engineer has
been assigned the responsibility for the contracting, review, and
coordination of project design. All actual design will be per-
formed by private architect-engineer firms contracted by the Kan-
sas City and Omaha Districts within MRD. MRD will coordinate
the design contracting, resolve design problems, verify design cost
estimates, coordinate the design review, and approve the de-
signs. A flow diagram depicting how the engineering and design
activity would normally function is shown in Fig. 2. The "con-
Usual Process
Design District
MRD
Lead District
A-E
Prepare
Plans &
Specifications
Design
District
Prepare
Bid Package
Design
District
Advertise
Bid
Opening
at
District
Design
District
Prepare
Amendments
Award by
Design
Distnct
Design
Distnct
Post
Amendments
Transfer of Bid Package
District
for
Construction
Figure 2.
Design Review
-------
STATE PROGRAMS
416
structing" or lead district will coordinate design management ac-
tivities with MRD (see Table 1). MRD will prepare construction
bid packages for competitive award to private industry con-
tractors. All division engineers will perform regional coordina-
tion within their respective EPA Superfund boundaries.
Table 1
USEPA Superfund Project Management and
Construction Responsibilities
Corps
Division
New England
North Atlantic
Ohio River
South Atlantic
Lower Missis-
sippi Valley
North Central
Southwestern
Missouri
South Pacific
North Pacific
Pacific Ocean
States
ME
VT
MA
NH
RI
CN
NY
NJ
PA
DL
MD
VA
DC
WV
KY
TN
IN
OH
VI
PR
NC
sc
AL
GA
FL
MS
LA
MN
WI
MI
IL
NM
TX
OK
AR
ME
IA
MO
KS
MT
ND
SD
WY
CO
NV
AZ
CA
UT
WA
OR
ID
AK
Amer
Samoa
Guam
HI
Corps
Lead District
New England
New England
New England
New England
New England
New England
New York
New York
Baltimore
Baltimore
Baltimore
Norfolk
Baltimore
Huntington
Louisville
Nashville
Louisville
Huntington
Jacksonville
Jacksonville
Wilmington
Charleston
Mobile
Savannah
Jacksonville
Vicksburg
New Orleans
St. Paul
St. Paul
Detroit
Chicago
Albuquerque
Ft. Worth
Tulsa
Little Rock
Kansas City
Kansas City
Kansas City
Kansas City
Omaha
Omaha
Omaha
Omaha
Omaha
Los Angeles
Los Angeles
Sacramento
Sacramento
Seattle
Portland
Walla Walla
Alaska
Pacific Ocean
Pacific Ocean
Pacific Ocean
USEPA
Region
I
I
I
I
I
I
II
II
III
III
III
III
III
III
IV
IV
V
V
II
II
IV
IV
IV
IV
IV
IV
VI
V
V
V
V
VI
VI
VI
VI
VII
VII
VII
VII
VIII
VIII
VIII
VIII
VIII
IX
IX
IX
VIII
X
X
X
X
IX
IX
IX
Construction Management
The implementation or construction activity will be fully in-
tegrated into the existing construction management structure at
Corps districts, divisions, and at the Office of the Chief of
Engineers. Lead districts will be responsible for the construction
management phase (see Table 1 for state breakout.) The "con-
structing" district will execute contracts let by competitive bid to
private industry and will provide contract administration and con-
struction management activities, including financial management
and reporting activities. The "constructing" district engineer, or
his designee(s), will have contracting officer authority, using bid
packages prepared by MRD. The construction effort will be
managed by the lead district, which will turn over the completed
project to the respective USEPA regional office.
Simplified USEPA-Corps Activity Breakdown
Figure 3 is a simplified remedial response activity flow diagram
that shows the activity interface between USEPA and the Corps.
Other Corps Responsibilities
The Corps will be responsible for developing a site safety plan
based on information contained in the remedial investigation and
feasibility study. The plan will cover the health and safety of per-
sonnel involved in site design and remedial actions, as well as
populations in the immediate site area. Implementation of the plan
will be shared between USEPA and the Corps. Corps responsibility
will be limited to design and remedial action personnel, and
USEPA will coordinate all actions relating to off-site populations.
In addition to development of the site safety plan, the Corps will be
responsible for environmental monitoring during the design and
construction phases; preparation of site operation and maintenance
manuals; facility start-up; operator training; and assisting USEPA
in the implementation of community relations plans.
DESIGN AND CONSTRUCTION CONTRACTING
Design
Architect-engineer firms and construction contractors can get in-
formation on upcoming work by:
•Keeping their current DA Form 254's on file with MRD and the
Kansas City and Omaha districts for design, and with the
geographic lead district for construction.
•Watching Commerce Business Daily announcements
•Keeping in contact with the geographic lead district procurement
office
Construction
In addition to the above, each District Engineer maintains a list
of prospective construction bidders who have expressed interest in
specified types of procurement that may occur within his assigned
geographic area. Annually, in February, each division engineer
publishes, for distribution to the construction industry and sup-
pliers, a schedule of major construction procedures expected to be
advertised for bids over a 20-month period.
PROCUREMENT PROCEDURES
•The Corps will use its standard contracting and procurement pro-
cedures
•Small business set asides will be in accordance with the criteria set
forth in the Federal Procurement Regulations at EPR
l-l-706-5(a).
-------
417 STATE PROGRAMS
EPA
Figure 3.
Remedial Response Activity Flow Diagram
-------
STATE PARTICIPATION UNDER SUPERFUND
HARRY P. BUTLER
State and Regional Coordination Branch
Hazardous Site Control Division
U.S. Environmental Protection Agency
Washington, D.C.
INTRODUCTION
Effective hazardous materials cleanup requires a strong partner-
ship between the USEPA, private parties, and the States. The
strength of that coalition has been developing in recent years, and
promises to continue to grow stronger. This paper focuses on the
developing relationship between USEPA and the States, with
primary attention to the joint participation of States and USEPA in
taking remedial action at uncontrolled hazardous waste sites to
date under the Superfund program.
A PARTNERSHIP UNDER SUPERFUND
A strong partnership has developed between the States and
USEPA over the first two years of the Superfund program. On the
projects that have been initiated at 58 sites in 29 States across the
country, USEPA and the States have assumed joint roles for the
management of site activities and for decision-making. As the pro-
gram has matured, the States have shown an increasing desire to
take the lead role in managing the response actions at their sites.
Mechanisms of Participation
Planning and executing the cleanup of a hazardous waste site is
both technically and administratively complex. CERCLA and the
National Contingency Plan specify a number of determinations
and assurances that must be made before work can begin. They
also specify a number of procedures and requirements that must be
followed while carrying out preparatory planning and remedial ac-
tions. In addition, other Federal laws and regulations place re-
quirements on either the State or USEPA, depending upon which
has the lead role for the project.
Because of these requirements, it is imperative that there be a
clear delineation of the responsibilities of all parties. The two
documents that are being used to define USEPA and State roles are
the cooperative agreement and the State Superfund contract. The
cooperative agreement is used when the State takes the lead role,
while the State Superfund contract is used when USEPA leads the
effort. The responsibilities of the States and USEPA under the two
instruments are summarized in Fig. 1, and the following sections
provide more details.
Cooperative Agreements
When the State desires the lead management role for remedial
planning and implementation at a site, the State must submit an ap-
plication for a cooperative agreement. A cooperative agreement is
much like a grant, in that money is transferred from USEPA to the
State. A key distinction, however, is that there is more substantial
USEPA involvement in a cooperative agreement than under a
grant.
The cooperative agreement application contains the State's work
plan, schedule, project budget, and the assurances required by
CERCLA. The work plan and schedule set forth specific details on
how and in what time frame the State will accomplish the remedial
action. The budget shows the expected cost of each major activity.
It is also broken down by categories such as personnel, travel, con-
tractual services, and equipment. The assurances that are required
by section 104(c)(3) of CERCLA are that the State will share in the
cost of the action; that approved capacity is available for any
necessary off-site treatment, storage, or disposal, and that the State
wilal assume responsibility for all future maintenance of the
response action. The application also documents how the State will
comply with other applicable Federal laws and regulations.
USEPA reviews the application upon receipt, and drafts special
conditions to the award for those aspects that are not adequately
addressed in the application. An offer of award is then made by
USEPA.
When the State signs the offer, USEPA sets up a letter of credit
account in the amount of the award. The State may draw down
from its account to meet its expenses. Using funds in the letter of
credit, the State may procure contractors' services to conduct the
work called for or use in-house resources. The costs of State per-
sonnel that are working on site activities are also included. It is the
State's responsibility to see that the activities specified in the work
plan are carried out according to schedule and within budget.
Even though the State has the lead role under a cooperative
agreement, USEPA also has responsibilities. The Agency's main
responsibilities are to monitor the State's progress throughout the
project and, in accordance with section 104(c)(2) and (4) of
CERCLA, to select the appropriate remedial action to be taken at
the site, after consultation with the State.
State Superfund Contracts
When USEPA takes the lead for a project, the work is done by a
USEPA contractor or through the Corps of Engineers. Conse-
quently, there is no transfer of funds from USEPA to the State, as
RESPONSIBIUTIES
1 . APPOINT PROJECT OFFICER TO
COORDINATE AND LEAD ACTIVITIES
2. HANDLE ALL CONTRACTUAL MATTERS
RELATING TO THE PROJECT
3. DEVELOP SCOPE OF WORK INCLUDING
COST ESTIMATES AND SCHEDULES
4. OVERSEE AND DIRECT ALL PROJECT WORK
5. REVIEW AND COMMENT ON WORK PLAN
AND COST AND TIME ESTIMATES
6. DEVELOP AND IMPLEMENT COMMUNITY
RELATIONS PLAN
7. PREPARE AND SUBMIT REPORTS ON
PROGRESS AND EXPENDITURES
8. MAKE STATUTORY REQUIRED
ASSURANCES
9. PREPARE SITE SAFETY PLAN
10. ASSURE SAMPLING AND ANALYSIS
QUALITY
COOPERATIVE
AGREEMENT
EPA
•
•
STATE
•
•
•
•
•
•
•
•
•
SUPERFUND
CONTRACT
EPA
•
•
•
•
•
•
•
STATE
•
•
•
Figure 1.
Major Responsibilities of States and EPA under
Superfund Contracts and Cooperative Agreements
418
-------
419
STATE PROGRAMS
there is with a cooperative agreement. However, with a USEPA-led
project, as with a State-led project, there is a requirement that the
State make the CERCLA 104(c)(3) assurances that are discussed
above. The mechanism the State uses to make those assurances
must be a formal legal document; thus a non-procurement contract
between USEPA and the State—the State Superfund contract—is
used.
The scope of work for the activities to be performed is, like the
assurances, a major element of the State Superfund contract. The
scope of work is prepared by USEPA, and along with the re-
mainder of the State Superfund contract, it is reviewed by the
State. The contract is signed by USEPA and the State when both
agree to the terms.
If remedial planning activities—that is, a remedial investigation
or feasibility study—are to be done, a USEPA contractor performs
the required work. The contractor reports its findings to USEPA,
and USEPA in turn reviews the information with the State. By con-
trast, for remedial design or remedial construction projects,
USEPA transfers funds to the Corps of Engineers through an In-
teragency Agreement. The Corps of Engineers then contracts for
and supervises the work.
Throughout the project, USEPA involves the State in major
decisions that must be made. Foremost among these is the deter-
mination of the appropriate remedial action. USEPA is required by
section 104(c)(2) of CERCLA to consult with the State before mak-
ing this decision. There is also a substantial involvement of the
State in reviewing contractor bids, contract documents, project
work plans, and progress reports.
Record of State Participation
To date, States have shown a keen interest in participating under
Superfund. A total of 41 States have sites on the Interim Priority
List of 160 sites, and as of the end of Fiscal Year 1982 (Sept. 30,
1982), 29 States had initiated some kind of Superfund remedial ac-
tivity. The lack of remedial action at the remaining 12, however,
should not be construed as a lack of State involvement. In some
cases, the State's site(s) had been cleaned up using emergency
funds, and no further remedial action was necessary. In other
cases, there was active negotiation for private party cleanup taking
place between USEPA of the State and the responsible party. In
still others, the responsible party was cleaning or had cleaned the
site.
States have also demonstrated an increasing desire to lead
remedial actions. The data in Table 1 shows that in Fiscal Year
1982, over 44% of the remedial projects that were initiated were
State led, up from 27% in the first year of implementation. Fur-
thermore, State led projects represented approximately 67% of the
money obligated for remedial projects in Fiscal Year 1982, up from
42% in Fiscal Year 1981.
There are a number of reasons for a State to decide to have
USEPA take the lead. In many cases, it is a question of timing.
USEPA's contracts already have been negotiated, and therefore,
USEPA led projects can begin shortly after the State Superfund
contract is signed. Conversely, with State led projects, the State
must go through the full procurement process to hire contractors, a
process which may take from two to three months. In other cases,
States have not had sufficient staff to take the lead role for all of
their sites. They have chosen to have USEPA take the lead on some
sites, while they take the lead on others.
Table 1.
Superfund Obligations for Remedial Actions
Total
Source of 8 SUt» Ltd Subtotal (tl EP> Led Subtotal (il Project! Totil (}|
ROW Projects
Ottwr Projects
SF/FY 81 Projects
To til FV 81 Sites
SF/FT 82 Projects
Toul nr 82 Sites
SF/Subtoul
Projects
29*
31
39
1.760.609
0
2.305,600
34.0Z1.740
36.327,340
38.087,945
17
4
3
20*
Grand Totil
i». of States 20*
"lot additive. SUM have botn EPA-led and Sta
•.391.000
5*0.000
726,000
17.098.140
17,824,140
22,755.140
23
4
4
26*
46*
72
101
•6,151,105
540,000
3,031,600
S1.119.MO
54.151.4W
«0,84),«
29*
-led projects.
As States and USEPA work together under Superfund, more
and more projects will be initiated, and greater sums of monies will
be expended. As shown in Table 1, in Fiscal Year 1981, using
Resource Conservation and Recovery Act (RCRA), Clean Water
Act, and Superfund monies, the States and USEPA initiated 33
projects at 26 sites. The total amount obligated was almost $10
million for a site average of approximately $370,000. By contrast,
the total number of projects initiated in Fiscal Year 1982 doubled
with 68 projects begun at 46 sites. Total funds obligated were over
$51 million, and the average obligation per site in the second year
of Superfund implementation was over $1.1 million, three times
that of the first year.
The average obligation per site increased because the type of pro-
ject being initiated at many of the sites was changing. Most of the
projects initiated in Fiscal Year 1981 were planning pro-
jects—which are relatively low cost. In Fiscal Year 1982, a number
of the planning activities that were initiated in Fiscal Year 1981 and
earlier (those initiated by the States using their own funds) were
completed, and more costly remedial actions were begun.
CONCLUSIONS
USEPA and the States have worked together for almost two
years on Superfund activities. The record of those two years shows
that the total number of actions undertaken has increased from the
first to the second year. Furthermore, State involvement in those
activities has also increased.
In the future, a larger number of sites will be eligible for Super-
fund attention. The Interim Priority List, the National list of
priority sites for Superfund attention, was expanded on July 23,
1982, from 115 sites to 160 sites. The National Priority List of 400
sites will be announced in the fall of 1982. An increasing number of
States will gain experience with Superfund activities through work-
ing with USEPA. This increased experience will be reflected in
a greater number of State staff capable of managing future Super-
fund activities. It is expected that, as a result of the working rela-
tionship that has been established in the past two fiscal years, the
partnership between USEPA and the States will continue to grow
in Fiscal Years 1983 and 1984.
-------
FEDERAL/STATE COOPERATION ON SUPERFUND:
IS IT WORKING?
GAILTAPSCOTT
State Regulation Report, Business Publishers, Inc.
Silver Spring, Maryland
INTRODUCTION
This attempt to assess the success of the Federal/State partner-
ship in implementing the Comprehensive Environmental Compen-
sation and Liability Act of 1980 focuses on problems and suc-
cesses in both policy formulation and practical application. The
information presented is based on telephone or in personal inter-
views with major figures involved in this implementation process
at the federal level (both headquarters and regional personnel)
and at the state level. In addition, major documents published in
the Federal Register or otherwise made available by the Environ-
mental Protection Agency have been reviewed. Also several major
speeches or pieces of testimony by key Administration figures that
relate to the state role in CERCLA implementation have been ex-
amined.
Persons interviewed were assured that they would not be quoted
directly and that all conclusions drawn would be based on a broad
survey of opinions. All the interpretations and conclusions in the
paper are those of the author and are based on an evaluation of
general trends found in interviews with persons representing a
broad spectrum of views.
HISTORICAL FRAMEWORK
CERCLA popularly known as Superfund was designed to pro-
vide a coherent federal response to the growing problems caused by
abandoned hazardous waste sites.
As it was developed during the waning days of the 96th Con-
gress, CERCLA was conceived of primarily as a federally oper-
ated program. General thinking among the drafters of the legisla-
tion seemed to be that states were lacking both the financial and
technical ability to adequately address the problem.
After Superfund became law and the USEPA began implemen-
tation, it rapidly became clear that state governments were not
anxious to relinquish complete control of hazardous waste cleanup
within their borders.
State interest in playing an active role in both policy develop-
ment and practical implementation of Superfund was given a big
boost by the commitment of a new Republican Administration to
return as much regulatory power as possible to the states. Under
the Reagan Administration banner of "new Federalism", state
officials sought and quickly found a hearing in USEPA.
In what many observers said was an effort to anticipate the de-
sires of the new Administration, Walter Barber, acting adminis-
trator of USEPA during the transition period, moved to set up a
dialogue with states. In Mar. of 1981, Barber, working jointly with
the National Governors' Association (NGA), the National Con-
ference of State Legislators (NCSL), the National Association of
Counties (NACo), the National Association of Attorneys Gen-
eral (NAAG) and the Association of State and Territorial Solid
Waste Management Officials (ASTSWMO) formed an ad hoc com-
mittee to advise USEPA's Superfund Task Force.
The ad hoc advisory group met on two occasions—Mar. 6 and
Apr. 14 and 15—with members of USEPA's Superfund Task
Force, headed by Gary Dietrich of the Office of Solid Waste.
Purpose of the first meeting was to delineate state concerns, while
the Apr. meeting was designed to seek state input and reaction to
a preliminary draft of a Superfund Strategy prepared by USEPA
staffers.
The ad hoc committee prepared a detailed response to the draft
strategy in which it said in part, "the inevitable conclusion is that
the program for implementing superfund should be in every way
possible delegated to units of state government."
State officials urged Congress to repeal Section 114 of Super-
fund, dealing with preemption of state funds. On the compen-
sable costs issue, states urged that, contrary to strategy's intent,
direct costs for state personnel working on a site should be com-
pensable on grounds this would not only be cost-effective but
would also provide an incentive for states to develop and expand
their own response.
States expressed concern about the strategy's proposal for an an-
nual site prioritization process, taking the position that this ap-
proach could jeopardize the continuation of cleanup programs. On
the issue of selection of prime and subcontractors, the states ex-
pressed strong support for being allowed to choose their own sub-
contractors.
Although a final Superfund Strategy was never publicly re-
leased, USEPA officials insisted that the ad hoc committee posi-
tion paper joined other USEPA-prepared documents on Adminis-
trator Anne Gorsuch's desk as she began on the job of adminis-
tering the agency. Agency officials seemed to feel that the view ex-
pressed in the draft agency strategy and the state dissenting pa-
per made their point. They noted that early drafts of headquarters
guidance to USEPA regional offices reflected a desire on USEPA's
part to create a real partnership with states.
Although the ad hoc advisory committee was originally con-
ceived as a potential ongoing advisory committee, this has not turn-
ed out to be the case. However, the staff of the Natural Re-
sources and Environment Section of the NGA sees itself as playing
at least a monitoring role vis-a-vis the federal government as Super-
fund implementation continues.
Concurrent with but distinct from the ad hoc committee, a State
Task Force on Superfund composed of hazardous waste regula-
tory officials was created under the umbrella of the Association of
State and Territorial Solid Waste Management Officials. The
ASTSWMO Task Force headed first by Don Lazarchick of Penn-
sylvania and now by Dale Wikre of Minnesota participated in some
meetings with USEPA to review preliminary drafts of the agency
policy statements on Superfund Cooperative Agreements with
states. Task force members provided comments on the draft and al-
though some state officials found fault with the final document,
they did not provide official comments on the final package.
In addition, at the ASTSWMO meeting in Utah on Sept. 1-3,
1981, the Task Force set itself several additional tasks in the area of
monitoring Superfund implementation. The group made plans to
file detailed comments on the National Contingency Plan proposal.
Indeed when the NCP was proposed in Mar. 1982, the group as a
whole as well as individual states provided in depth and somewhat
critical comments on the proposal. With the final promulgation of
the NCP in June 1982, many states continued to express their dis-
tress with several aspects of the plan which will be detailed in a later
section of this paper.
The Task Force also agreed to monitor the activities of the other
federal agencies involved in Superfund implementation, such as the
U.S. Coast Guard and the Treasury Department to see that they are
conducting themselves according to the spirit of the law.
In addition, the group expressed interest in monitoring the first
few Cooperative Agreements with states reached under Superfund
for such items as apportionment of leadership and consistency of
policy among USEPA regions and to see how USEPA addresses
420
-------
421
STATE PROGRAMS
the "how-clean-is-clean?" issue on a case-by-case basis.
The Task Force is an ongoing group but it has not been very ac-
tive since its group effort to file comments on the C/A Guidance
Package and the NCP in April. The Task Force is set to meet on
Sept. 10 and 12 before the National Solid Waste Management
Association meeting in Salt Lake City. At that time Wikre said he
will poll Task Force members and other state officials to get a bet-
ter feel for a state consensus of how Superfund implementation is
evolving.
MECHANISMS FOR STATE LIAISON
In order to facilitate a dialogue with the states, USEPA's organ-
izational structure for the Office of Emergency and Remedial Re-
sponse includes a branch devoted entirely to state and regional co-
ordination. Although the existence of such a unit does suggest a
commitment to state involvement, the number of people actually
working on state issues at the headquarters level is limited to less
than 10 full time professional persons out of approximately 90 peo-
ple in the headquarters office.
The major division with state liaison responsibility is the Haz-
ardous Site Control Division. Major functions of the division are:
(1) policy guidance, technical studies, contracts and program man-
agement regarding uncontrolled hazardous waste sites, (2) discov-
ery and assessment, remedial containment and control activities at
sites, and (3) state liaison and coordination.
The formal organization of the division includes three branches:
Discovery and Investigation Branch, Remedial Action Branch, and
State and Regional Coordination Branch.
RCs branch is broken down into two sections. One section is
State Programs and the other is the Remedial Implementation Sec-
tion. To facilitate direct contact with the USEPA regional offices
and to provide for an organized distribution of labor, the coun-
try has been divided into three zones, each with a Zone Manager re-
porting to a Section Head.
As currently organized under the severely limited staffing con-
straints, Zone I consists of Regions I, II and IV. Zone II covers
Regions III and V, and Zone III contains Regions VI through X.
The State Programs Section is responsible for providing guid-
ance to the regional offices on how to conduct Cooperative Agree-
ments (C/As) with states on remedial actions. The section is also
in charge of major technical review of each state application for a
C/A. The C/A application goes to the regular USEPA Grants
Administration Office where the formal award takes place. How-
ever, the bulk of the evaluation process will be in the Superfund
Office.
The Remedial Implementation Section provides input in the de-
velopment and later the implementation of the NCP. It will address
the state role in the NCP and give guidance to states on their even-
tual part in the plan's operation.
The Emergency Response Division in OERR is also vitally con-
nected to the state's interest. This group coordinates the National
Response Team (NRT) and its Regional Response Teams (RRT)
which often work closely with state emergency personnel in the case
of spills, fires and explosions. When USEPA goes forward with
current plans to develop a generic state C/A on Emergency Re-
sponse action under Superfund, the C/As will be administered
under this office.
At the regional level, the early stages of Superfund implemen-
tation were conducted largely on an ad hoc basis. By early sum-
mer of 1981, each regional office had on person known as a Super-
fund coordinator designated as the person to interact with both
states and headquarters. Some of the more sophisticated regions
like Region II have a staff of several people working exclusively on
Superfund activities. In other regions like Region VIII such ac-
tivities are carried out by persons who have many other duties.
In the wake of a Sept. 15, 1981 directive from USEPA Admin-
istrator Anne Gorsuch, regional handling of Superfund and related
issues became more uniform. Regional offices choose between two
organizational formats, both of which created greater consolida-
tion of activities regarding Superfund, the Resources Conserva-
tion and Recovery Act and permits issuance and monitoring.
The major line of communication between USEPA state liaison
staffers at headquarters and regional Superfund personnel takes
the form of regional guidances which set down general para-
meters for dealing with the states.
A final version of the USEPA Guidance on Cooperative Agree-
ments under CERCLA was made available on Mar. 11, 1982 at the
same time the proposed National Contingency Plan was unveiled.
This version differs in some substantial ways from earlier public
drafts and has given rise to many state objections that will be dis-
cussed in a later section of this paper.
The final C/A guidance notes four areas that are of particular
importance to federal/state relations. The first is the requirement
that federal remedial actions should be undertaken only after con-
sultation with the affected party. Agreements between a state and
USEPA are documented in either a C/A or a Superfund State con-
tract.
The second provision related to assurances that the affected
states must provide prior to any remedial action. The state must
assure that it will: (1) assure operation and maintenance costs for
all remedial and removal measures that are implemented, (2) pro-
vide for a facility for off-site disposal if necessary, and (3) share in
the costs of the remedial action.
The third and fourth provision for coordinating state and federal
actions related to the granting of a credit to an affected state for
costs expended or obligated at a remedial site between Jan. 1,1978
and Dec. 11, 1980. The credit is used as part of the states share
of costs for federally funded response at the site.
According to the final guidance a C/A will be used when a state
is expected to take the lead in remedial work and funds are ac-
tually transferred to the state. A Superfund State Contract will be
used when USEPA is taking the lead.
On the state assurances issue, with regard to O & M costs, the
guidance says the state must make a firm commitment to assure
this responsibility. To satisfy this requirement the state must: (1)
identify the organizational unit responsible for administering 0 &
M ACTIVITY, (2) identify the state financial mechanisms for
funding O & M and (3) identify milestones for assuming respon-
sibility.
In probably the most controversial change from the earlier draft,
the final guidance notes under the cost sharing provision that at
privately owned sites, states must share 10% of both remedial plan-
ning (investigation, feasibility studies and design) and remedial im-
plementation.
If the state credit is not sufficient to satisfy the state's share of
the costs, either state funds will be used simultaneously with fed-
eral funds under a C/A or payment terms will be negotiated in a
state contract.
Prior to remedial design activity, the state must either through a
C/A or a new or amended state contract, make a firm commit-
ment to provide funding for remedial implementation. A state may
satisfy this assurance by (1) authorizing the reduction of a state
credit to cover its share of costs, or (2) identifying current avail-
able funds earmarked for remedial implementation or (3) submit-
ting a plan with milestones for obtaining necessary funds.
The final guidance notes that state credits must be documented
on a site specific basis for state out-of-pocket non-federal eligible
response costs between Jan. 1, 1978 and Dec. 11, 1980.
For USEPA-led remedial action where there is no state credit or
the credit is not sufficient to cover state costs, payment terms will
be negotiated between USEPA and the state and documented in the
state contract.
For state-led remedial actions where there is no state credit or the
credit is not sufficient to cover state costs, the C/A will cover only
USEPA's share of the costs. USEPA will provide the award
amount to the state through a letter of credit. States are required
to match federal funds as work progresses.
On the controversial issue of how state trust funds may be pre-
empted by federal Superfund, the final guidance explains that
USEPA has determined that Section 114(c) of CERCLA does not
apply to state funds which are used for the following purposes: to
-------
STATE PROGRAMS
422
finance the administrative costs of a state fund, to finance the
purchase or preposition of hazardous substance response equip-
ment and other preparations for responding to releases within a
state, to finance the cleanup of petroleum discharges, to pay the re-
quired state contribution to cleanup actions financed by the
CERCLA Trust Fund, to compensate claims to the cost of restor-
ation and replacement of any natural resources damaged or des-
troyed by a release of a hazardous substance, to advance funds to
remove or remedy releases of hazardous substances eligible to be
financed by the Hazardous Substances Response fund if a Cooper-
ative Agreement or Contract has been issued by the USEPA, and to
compensate damage claims and to remove or remedy releases of
hazardous substances eligible to be financed by the fund but for
which no federal reimbursement is provided.
MAJOR CURRENT STATE PROBLEMS WITH SUPERFUND
The picture vis-a-vis major state problems with Superfund im-
plementation has already changed several times during the less than
two years that the Act has been law.
Starting with initial state euphoria that they were going to be
able to have a more active part in Superfund planning than has
historically been the case in other environmental law development,
the situation then went into a phase where most state officials in-
dicated the view that USEPA was moving much too slowly and
husbanding the fund so closely that only a "trickle" of money was
getting actually transferred to the states to deal with individual
problem sites.
During the middle phase, the National Governors Association
at its Feb. 1982 meeting in Washington, B.C. voted on a Superfund
resolution through its Committee on Energy and the Environ-
ment. The resolution stated that in view of the absence of program
guidelines and cleanup standards, the national program appeared
to be stalled. It noted that states had also been reluctant to take
action.
NGA urged USEPA to issue as soon as possible a national
cleanup plan which would incorporate the following features:
frequent consultation between NGA and USEPA on all aspects of
Superfund implementation; inclusion of each state's first priority
site in the National Priority List, rapid delegation of cleanup re-
sponsibility to willing capable states for remedial action and emer-
gency response, annual appropriation request from the Adminis-
tration to Congress to provide financial assistance to states for
preparation of a national waste site investory, federal payment of
100% of site investigations, feasibility and design costs for remed-
ial actions; state payment 10% of construction costs and remed-
ial action sites (or 50% if publicly owned during the time of dump-
ing), management of remedial action and emergency response by
willing and capable states, including selection and supervision of
contractors and subcontractors, state payment of 100% of the
operation and maintenance costs to remedial action sites with
Superfund paying the balance for the life of the fund, close co-
ordination between the Department of Justice and USEPA
attorneys and their state counterparts in enforcement and litiga-
tion efforts.
Clearly the NGA position is in almost direct opposition to many
of the positions set out in the final C/A package as well as to some
of the provisions of the final NCP.
Regulatory officials who deal with the nitty gritty aspects of
working with the USEPA regions share many of the governors
concerns and have also developed other problems with USEPA's
state policies.
. A 1980 General Accounting Office report called "Federal-State-
Environmental Programs—The State Perspective" delineated
four major problems that seemed to crop up in the implementa-
tion phase of most federal environmental programs requiring state
participation. General problems fell in these categories: (1) de-
layed and inflexible regulations, (2) excessive USEPA control over
state programs, (3) inability to fill state vacancies, and (4) delayed
and uncertain federal funding.
This next section of the paper will attempt to look at how prob-
lems with Superfund at the state and regional level fall into these
categories.
Delays have plagued the Superfund program much as they have
most other environmental programs. The first three major initia-
tives whose delay caused distress for states were the site notifica-
tion list, the interim priority list, and the revised National Con-
tingency Plan.
Many states had hoped to wait for the site notification list which
is a compilation of known sites around the country before submit-
ting preferred sites for the interim priority site list that appeared in
late Oct. 1981. The states were very upset about the delay in the in-
terim site list primarily because they were under great pressure to
get on with cleaning up problem sites but felt they could not do so
until they were assured the site would appear on the list and qualify
for Superfund credit.
Once the interim list was out, states developed an even greater
concern about what they saw as the potential delays that would be
forthcoming in developing the 400 site National Priority List. State
officials expressed great distress about the amount of information
USEPA wanted on each site before it would even consider it for in-
clusion on the list. They noted that in many cases they would have
to spend a great deal of money to even evaluate a site sufficiently to
give it a score on the Hazard Ranking Model. States were also con-
cerned about what they saw as the inflexibility of USEPA in want-
ing to make sure that there were detailed assurances as to the level
of participation the states will promise on each site nominated for
the 400 list.
In reality at least part of the state fears have proved to be an
overreaction. Talks with state officials in early Sept. revealed that
most of them have been able to gather the necessary data for haz-
ard ranking more easily than thought. Many states attended
USEPA-sponsored technical workshops on how to use the system
and other states took a great deal of advantage of the technical ex-
pertise offered by USEPA in the form of its Field Investigation
Teams.
A USEPA headquarters official observed that there were many
different levels of participation in the scoring process by states but
said most states took at least some type of role. Several state peo-
ple remarked that although the ranking system was not entirely to
their liking it had worked out better than they had expected.
Headquarters spokesmen insisted that USEPA is keeping to the
schedule it set itself in the June 20 Guidance for establishing the
National Priorities List and will possibly announce the list by some-
time in Oct. At any rate, as of early Sept. most of the ranking had
been completed and was in USEPA's hands.
States were also distressed about the EPA delays in getting out
the NCP. State Officials were afraid to spend much money until
they had a clear idea of just what kind of role they would be play-
ing under the new NCP. After reviewing the final NCP many
state spokesmen indicated that they think it has a reasonable role
for state and local governments. They are still concerned, how-
ever, that the NCP does not take a clearcut position on the "how-
clean-is-clean?" issue. There is widespread state concern that
USEPA may come in and do a "quick and dirty" job of cleanup
and leave the huge longterm operation and maintenance costs to
the states.
Several states that are geared up for emergency response-type
activities were unhappy that USEPA had not moved more quickly
to involve states in emergency response actions. They feel that
states are in the best positions to determine what constitutes an
emergency within their borders. Many regional officials think in
contrast that the regional office is in the best position to assess
what constitutes an emergency that qualifies for Superfund action.
USEPA officials note that a C/A fpr emergency Response has been
prepared under contract to Booz Allen and Hamilton and that a
pilot program is underway to test the program in Region VI.
USEPA Assistant Administrator for Superfund, Rita Lavelle,
noted in an Aug. 6 speech that one of the major issues that arose
during the revision of the NCP was the development of the state's
role in emergency response. She noted that the final plan encour-
ages state involvement and delineates state roles in emergency re-
-------
423 STATE PROGRAMS
sponse actions. Lavelle noted that these roles include working with
on-site coordinators in determining the appropriate extent of rem-
edy and participation in Regional Response Teams. Also included
is a receipt notification of releases. A new section indicates how a
state may get into cooperative agreements or contracts for re-
sponse actions pursuant to CERCLA.
Although Superfund does not operate under the massive and de-
tailed regulations that states often complain about, the states were
initially very disturbed about what they saw as inflexibility and in-
equality in the original Mitre Model for the initial ranking of
Largely in response to state concerns and also at the behest of an
Office of Technology Assessment study, the model was changed
significantly and finally emerged as the Hazard Ranking System
set down in the June 20 Guidance for development of priority
sites. Although still not sure the system is the best possible method
to rank sites according to the real level of danger they pose, most
states admit the system is the best thing available at the moment.
Excessive USEPA control over state programs was a complaint
in some states but not in others. A number of state officials indi-
cated that the headquarters guidance on cooperative agreements
has proved helpful and that their negotiations with regional of-
ficials were working well. A typical turnaround time on a C/A
from the start of serious negotiations to USEPA headquarters
sign off has tended to be around two months. Part of the reason
that some states have few complaints about USEPA control is that
due to the site-by-site nature of Superfund, the states and USEPA
negotiate each C/A and Superfund contract individually and the
precise federal/state duties and responsibilities are spelled out in
detail. Some early concern about the possibility of USEPA exer-
cising too much domination over the contracting process seems to
have abated somewhat but problems will probably arise in situa-
tions where USEPA is taking the lead under state contracts. Sev-
eral state officials expressed resentment of the fact that EPA-nego-
tiated contracts often run higher costs than state-negotiated con-
tracts. They said that it was only natural that states would resent
having to pay their 10% of costs they felt were excessive to
start with.
Although most states originally had expressed considerable skep-
ticism about the major role that USEPA had carved out for the
Corps of Engineers under Superfund, things don't seem to be as in-
flexible as the states may have originally feared. In all situations
where the state is taking the lead on a site under a C/A they are
being allowed the flexibility to choose their own contractors. A
USEPA spokesman said that in the event that a state wanted to
use the corps for the construction phase of a remedial action they
could probably do so even though they are forbidden to negotiate
directly with any USEPA contractors. In all USEPA situations, the
Corps will be in charge of the construction phase of the work and
will determine what contractors will handle the work. Many state
officials anticipate some problems with the Corps as work at state
Superfund contract sites moves into the construction phase.
Inability to fill state vacancies is proving to be a problem in
Superfund implementation just as in other programs. In many
states, already overworked staffs are being asked to handle Super-
fund activities in addition to trying to shoulder responsibilities for
state hazardous waste programs under the Resource Conservation
and Recovery Act. In several states such as New Jersey, New
York and Pennsylvania, some of the same people are also involved
in writing the technical criteria for siting disposal facilities. Many
of the less industrialized states also have limited expertise in emer-
gency response and tend to rely heavily on Regional Response
Teams to address emergencies. Several states have also had major
personnel changes in the last year and many are currently under-
staffed.
Probably the most serious state concerns about the way Super-
fund implementation is unfolding revolve around the funding is-
sue. Although states have expressed some general concerns about
how long it takes EPA to approve allocations to the states, the even
bigger concern revolves around how the states themselves are go-
ing to meet the financial burden imposed on them under USEPA's
interpretation of the law vis-a-vis the state financial role.
Two major changes that became apparent last Mar. in the way
USEPA is interpreting Superfund have resulted in what might al-
most be termed "paranoia" on the part of the states that they will
wind up paying the lion's share of the final construction costs on
sites not to mention the potentially vast burden of long term opera-
tions and maintenance costs. As discussed earlier, USEPA will not
negotiate with a state unless it can provide certain assurances re-
garding the long term operation and maintenance costs. Many
states see the current USEPA position that the states must take vir-
tually total responsibility for O & M as both a policy reversal and a
betrayal.
Several state officials expressed the fear that under this policy
USEPA might choose to fund a lot of projects that involve min-
imal front end remedial work but will require complex and expen-
sive long term monitoring of groundwater, surface water and air.
Of almost equal concern to states is USEPA's decision to re-
quire states to pay 10% of all the costs involved in remedial plan-
ning including preliminary feasibility and design studies. Although
states have more or less reconciled themselves to this policy they
are still predicting dire consequences down the pike as a result of
the decision.
Shortly after the decision was made public, Dale Wikre, who
heads the Minnesota hazardous waste program and also chairs an
ASTWMO task force on Superfund, said that the timing issue pre-
sented by this unexpected requirement was almost as crucial as the
money involved. He noted that given the legislative constraints in
many states with regard to obtaining matching money, there will be
long delays before even the preliminary work can start.
Although USEPA officials feel they have given the states a lot of
flexibility in how they can provide assurances for their share of
funds, Wikre expressed skepticism about whether most state legis-
lators or governors will buy the package. He said that in some ways
what states are being asked to do is tantamount to writing a blank
check.
Noting that 1985, when Superfund sunsets, is not that far away,
Wikre said the delays that could result from the 10% of everything
requirement could last until Superfund expires in many cases. He
expressed the view that at least a few states would not be able to ob-
tain adequate assurances and that some who do attain them will not
be able to really come through. He expressed the personal view that
USEPA seemed to be making it hard for states to participate while
saying they wanted states to be active.
The outspoken Minnesota official went so far as to suggest that
the top echelon as USEPA sees its job under Superfund as protect-
ing the fund rather than protecting the environment.
CONCLUSIONS
It is impossible to even begin to address the real story of fed-
eral/state coordination under Superfund in a short paper since
there are 50 distinct and separate situations. To try to access a gen-
eral picture of how the process is going requires making general
statements that might be rather misleading when applied to an in-
dividual state.
Some states like California, New York and New Jersey for ex-
ample have clear funding mechanisms of their own but have so
many problem sites that the possibility of them getting what they
see as adequate funding under Suprefund is virtually nil. Other
states have almost no sites that they feel are requiring federal at-
tention. Still other states may have real problem sites but will not
submit them because they are not sure of getting the 10% match.
Some observers say that this development could result in a situa-
tion where many dangerous sites do not become Superfund can-
didates until so late in the process that most of the money will have
already been committed.
Ultimately the real success of federal/state coordination can not
be determined until far in the future when it becomes clear how
many sites actually got cleaned up and how well states are able to
handle the long-term operation and maintenance costs. It will also
hinge on the actions of state legislators who must decide whether
to allocate funds for the state share.
-------
THE TCE RESPONSE IN ARIZONA
PAMELA JANE BEILKE
Arizona Department of Health Services
Division of Environmental Health Services
Phoenix, Arizona
INTRODUCTION
The contamination of groundwater by hazardous and toxic
wastes has become the environmental issue of the '80s. Throughout
the nation, this valuable underground resource, once believed to
be pristine and resilient, has been degraded as the result of man's
casual disposal of wastes. Historically the disposal of wastes has
not been closely monitored or regulated. Industrial wastes have
been dumped into unlined surface impoundments, shipped to
landfills, poured down abandoned wells or simply dumped in the
dark of night in remote locations.
With the realization that the quality of the groundwater had been
jeopardized by these activities, a national assessment of waste dis-
posal sites was initiated. In Dec. 1979, The Arizona Department
of Health Services (ADHS) completed a statewide surface im-
poundment assessment.' The report evaluated the potential for
groundwater pollution by leachate from numerous surface im-
poundments and assessed the priority of these sites for further in-
vestigation. The Hughes Aircraft facility (Hughes) located in the
area' of Tucson International Airport was a high priority and had
reportedly caused the contamination of nearby wells with chrom-
ium, cadmium and arsenic.
As a result of the surface impoundment assessment, a USEPA
field investigation of selected uncontrolled hazardous waste sites
was conducted inn Arizona. The investigation of the Hughes facil-
ity in March 1981 revealed the presence of various contaminants
in wells at and near the facility. The preliminary report2 con-
cluded that there was evidence of groundwater contamination
and recommended further investigation. Additional sampling con-
firmed the presence of trichloroethylene (TCE),l,l-dichloroethy-
lene, 1,1,1-trichloroethane and chromium. The most predominant
contaminant found was TCE, a degreaser used extensively by a
variety of industries since the 1920s.
Subsequent groundwater monitoring in the Tucson International
Airport area revealed a plume of groundwater contamination ex-
tending several miles, affecting 8 wells belonging to the City of
Tucson (Tucson Water), four industrial wells at the Hughes facil-
ity and several semi-public or private wells. The highest concen-
trations were found at the Hughes facility. They exceeded 13,000
|ig/l. Concentrations in the municipal wells ranged from 1 ug/1 to
240 ug/1.
The discovery of TCE in Tucson led to testing in other areas
of the State. The City of Phoenix initiated a TCE scan of distribu-
tion system samples in conjunction with quarterly trihalomethane
analyses in Oct. 1981. This screening process detected a contam-
inated area near Indian Bend Wash in east Phoenix. Further test-
ing confirmed the presence of TCE in seven municipal wells be-
longing to three different cities (Phoenix, Scottsdale, Tempe) and
four irrigation wells. TCE concentrations in the Indian Bend Wash
area ranged from 5 ppb to 1,000pg/l.
ADHS, the regulatory agency with primary responsibility for
groundwater protection, safe drinking water and control of haz-
ardous wastes, has acted as the lead agency in coordinating all
local, State and federal TCE-related activities. The Departmental.
goals underlying the development of the TCE response were the
protection of the public's health first, and secondly the protec-
tion of the groundwater quality in the State. In the absence of
State or federal standards and regulations for volatile organic con-
taminants in drinking water or groundwater, ADHS has estab-
lished a State action level for TCE and developed guidelines for
managing TCE in public water supplies.
A State action plan was developed to identify the basic steps to
be taken to clearly define the extent of groundwater contamina-
tion; to identify contaminated drinking water supply wells; and to
evaluate potential solutions. A State "Superfund" known as the
Water Quality Assurance Revolving Fund (WQARF) was estab-
lished by the State Legislature to provide funding for the removal
or reduction of man-made pollutants in drinking water supplies.
STATE ACTION LEVEL
Federal drinking water standards have not been established for
TCE or any of the volatile organic contaminants, although many
of these contaminants are suspected carcinogens, thus the desirable
concentration would be zero. USEPA has provided guidance based
on health risks for selected contaminants. The guidance, known
as Suggested No Adverse Response Levels (SNARL), identifies
the concentration ranges that would increase the risk of excess can-
cer in a given population exposed over a lifetime.
ADHS has established an action level of 5 /ig/1 which is equiv-
alent to USEPA's SNARL3, for TCE. It is the concentration at
which one may expect to observe one additional case of cancer in a
population of one million people consuming 2 liters of water per
day over a 70 year lifetime. The action level is not an enforce-
able standard, but the level at which the public water supplier is
asked to voluntarily remove a well from service while the magni-
tude of the problem is assessed and health impacts are evaluated.
GUIDELINES FOR TCE IN PUBLIC WATER SUPPLIES
In Mar. 1982, ADHS released the final Guidelines for TCE in
Public Water Supplies* The guidelines were developed in response
to requests and inquiries made by the impacted communities. The
public water suppliers actively participated in the development pro-
cess since it represented a concerted, cooperative effort to ensure
the safe quality of the public water supplies in the State. The guide-
lines are not mandatory, nor are they enforceable.
The guidelines define acceptable levels in drinking water, re-
quirements for initiation of testing and follow-up testing, pro-
cedures for sampling, reporting and public notification; steps to
ensure safe drinking water supplies; and allowances for blending
or short term use of water exceeding 5 ug/1. Although the guide-
lines specifically address TCE, they were structured so that they
can be expanded to incorporate other organic contaminants as
needed.
Initial testing may be conducted by ADHS or by the public water
supplier in areas identified by ADHS to have a high risk of TCE
contamination based upon an evaluation of land use and waste
424
-------
425
STATE PROGRAMS
disposal practices. Samples should be collected from all drinking
water wells if possible, although representative points within the
distribution system may be sampled in the preliminary testing. If
TCE is detected in a distribution system sample, all wells con-
tributing to that sampling location must be tested.
Wells containing less than 5 jig/1 may remain in service with per-
iodic testing to ensure that the TCE level has not increased. Any
well that exceeds the action level is resampled and an analysis for
the purgeable priority pollutants is completed. A sample is also
collected from the nearest consumer's tap for the same analysis.
When it is verified that a well exceeds the action level, the public
water supplier is advised to take it out of service while the neces-
sary investigations are conducted and alternatives for future use are
evaluated. Before the well is put back on-line, the water supplier
should submit a written plan of action which defines the steps to
be taken to ensure that the drinking water delivered to the cus-
tomer will not exceed the action level.
The guidelines allow for limited delivery of water containing
TCE in concentrations greater than the action level. Assuming that
no other contaminants are present at levels that pose a health
threat and the TCE concentration is less than 5 ug/1 for the re-
mainder of the year, the water supplier may deliver water con-
taining between 5 and 10 ug/1 for no more than 6 months, or 10
to 20 jig/1 for no more than 3 months, or 20 to 40 pg/1 for no more
than 1 !/2 months, or 40 to 50 jig/1 for no more than 1 month.
Finally, the guidelines discuss the alternatives for use of wells
which exceed the action level. Some of the possibilities are as
follows:
•Replacement by alternate source
•Blending with acceptable source
•Seasonal usage of up to 50fig/l
•Well modification to seal off contaminated zone
•Treatment to remove contaminants
WATER QUALITY ASSURANCE REVOLVING FUND
In April 1982 the Arizona State Legislature enacted a bill, H.B.
2207,' which established the Water Quality Assurance Revolving
Fund (WQARF). The monies appropriated to the WQARF,
$500,000 for 1982, and all monies recovered by civil penalty against
dischargers of hazardous wastes, were designated for the clean-up
or removal of man-made pollutants from groundwaters of the State
used for the purpose of supplying water for human consumption
by any political subdivision.
The monies in the WQARF may be used in two ways: (1) to
assist in the clean-up or removal of contaminants from ground-
water, or (2) to provide State matching funds to the federal super-
fund projects for groundwater-related remedial actions.
The legislation set forth the basic criteria for selection of sites to
receive funds which included the level of pollution present in
groundwater; the availability of alternative sources of water for
human consumption; the length of time the pollution has existed;
and the recent or future degree of impact on the public health and
welfare. The specific conditions for eligibility that have subsequent-
ly been defined are as follows:
•Pollution of the groundwater must be the result of a man-made or
man-induced alteration of the chemical, physical, biological or
radiological integrity of the groundwater
•The level of pollution present in the groundwater must exceed a
State standard as defined therein
•The polluted groundwater must have been used for human con-
sumption
•The applicant must be a political subdivision of the State in ac-
cordance with Arizona Revised Statutes
•The applicant must own or operate the water system which sup-
plies potable drinking water for human consumption
•The applicant's proposed project or clean-up/removal plan must
remove or reduce the pollutant to within tolerable State standards
•The applicant must be willing and able to commit local match-
ing funds at least equal to the total amount sought from the
WQARF
The Legislation designated ADHS the lead state agency respon-
sible for administering the WQARF. The Water Quality Control
Council (WQCC), established by statute to develop water quality
standards, effluent limitations, pretreatment standards, etc., was
charged with developing and implementing criteria for project eval-
uation and prioritization.
The priority system' adopted by the WQCC in Aug. 1982 des-
cribes the mechanism for the evaluation of project eligibility, the
adoption of the priority list, the procedures for additions and mod-
ifications to the priority list, the interface with the Comprehen-
sive Environmental Response, Compensation and Liability Act
(CERCLA); and the distribution of funds.
The following criteria were established to assess the priority
ranking of each applicant:
•Degree to which the pollutant in the drinking water source ex-
ceeds State standards
•Areal extent of contamination
•Estimated length of time the pollution has exceeded State stand-
ards
•Level of increase or decrease of pollutant
•Percentage of total existing potable water supply which will be in
jeopardy within the next 10 years
•Percentage of total existing potable water supply which currently
exceeds State standards
•Total number of persons which could be served by the potable
water source if the pollution were removed or reduced
•Percentage of total existing potable water supply which could be
replaced by developing an alternate source
•The ratio of the cost of developing an alternate source to the cost
of proposed clean-up project
•Years of service life of proposed project
Priority lists will be developed and adopted by the WQCC each
fiscal year that funds are available. ADHS will review all applica-
tions, clean-up proposals, and plans and specifications to verify
compliance with WQARF guidelines and State rules and regula-
tions. Construction of approved projects must be initiated within
the funding year.
STATE ACTION PLAN
ADHS has developed a State action plan which identifies the
basic steps to be taken by the State in response to the discovery
of TCE in groundwater. The objective is to define the magnitude of
the problem, isolate potential'sources and evaluate feasibility of
remedial actions.
Step 1. Monitor all public water supply wells in the vicinity of
the contamination. The purpose of this monitoring is twofold:
(1) to locate all drinking water wells that exceed the action level
and, (2) to define the horizontal extent of the plume.
Step 2. Monitor all other wells—semi-public, private, indus-
trial, and agricultural—in the area to further define the ex-
tent of contamination and characteristics of the plume.
Sept 3. Compile and evaluate all available information on the
wells such as well driller's logs, details of well construction,
pumping capacity, static water levels and historical water qual-
ity data, in addition to local hydrogeological data to facilitate
definition of the 3-dimensional characteristics of the plume.
Step 4. Define a study area based upon available informa-
tion. Review all information on historical and current land use
within the study area to identify potential sources of contam-
ination, including a survey of surface impoundments, a review
of old aerial photos, a summary of underground injection wells,
NPDES discharge monitoring reports, landfill- monitoring re-
ports and complaints of illicit activities.
Step 6. Inspect and survey industries in the study area to ver-
ify that current practices for handling, storage and disposal of
hazardous wastes are in compliance with State rules and regula-
tions, to determine the types and quantities of solvents and
chemicals used by the particular industry and to obtain informa-
tion on past disposal practices.
-------
STATE PROGRAMS
426
Step 7. Collect shallow soil samples in suspected source areas
to determine potential for contribution to groundwater con-
tamination.
Step 8. Conduct deep soil borings at those locations where po-
tential for groundwater contamination exists, as indicated in
shallow soil samples.
Step 9. Based upon the results of the deep soil analyses, con-
struct monitoring wells around the confirmed source area to de-
fine the vertical and horizontal distribution of the plume.
Step 10. Evaluate the feasibility of remedial action. Some of
the alternatives may include the elimination of sources still con-
tributing to groundwater contamination, rehabilitation of the
highly-contaminated regions of the aquifer, application of treat-
ment at the well for drinking water purposes, and utilization
of the contaminated aquifer for other purposes.
IMPLEMENTATION
The response to TCE in Arizona has been a cooperative effort.
Many parties have combined their resources and expertise to ac-
complish the same objectives addressing a mutual concern. The
various municipalities have contributed by performing pump tests
on contaminated wells, renovating and sampling abandoned wells
within the problem area, and reviewing historical water quality,
well construction and water consumption records. The Pima Coun-
ty Health Department initiated an extensive monitoring program
identifying and sampling private and semi-public wells in the
Tucson International Airport area. The USEPA provided tech-
nical assistance, conducted extensive field investigations, and con-
tributed analytical support. Their assistance was instrumental
in getting the Tucson site onto the USEPA's Interim Priority
Superfund List. The Arizona Department of Water Resources
has conducted well inventories in the problem areas and provided
well driller's logs and well specifications.
ADHS has been the central agency responsible for the over-
view and coordination of all TCE-related activities. Since the
organizational structure and fiscal planning process had not
allowed such an intensive "emergency response", resources and
manpolwer had to be borrowed from existing programs, leaving
obligated tasks unfinished. In the past year-and-a-half, responsi-
bilities have shifted back and forth between safe drinking water,
hazardous waste and groundwater protection programs. A special
core group has now been specifically set up to manage the con-
tinuing investigations and implementation of remedial action in the
TCE problem areas and to oversee the utilization of WQARF and
CERCLA monies.
In the process of implementation of the programs discussed
earlier, ADHS has encountered numerous obstacles and uncov-
ered new issues. Among these are analytical capability and qual-
ity control, relative health risks of alternate water supplies, effects
of well closure on contaminant migration, and resource alloca-
tion.
At the time that TCE was first discovered in Tucson, the State
Laboratory was in the initial stages of developing the capacity to
perform these analysis. Since that time, the laboratory's analytical
capabilities have been expanded to encompass all 32 purgeable
priority pollutants. The State Laboratory, however, is the only
facility within the State with these capabilities; therefore, the ma-
jority of the analytical work in the TCE investigations has been
performed by them.
The lab has a certification program for commercial laboratories
in the State. To date this program has only applied to those
analyses required by regulation; i.e., bacteriological, inorganic
chemical and a limited number of organics in surface water
samples. In order to help meet the State's needs, the laboratory
is in the process of developing a certification program for volatile
organics. Additionally, some of the affected municipalities have or
are in the process of developing their own analytical capabilities
to minimize the long term cost of surveillance.
In July of this year, the City of Phoenix closed two wells on the
west side of the city in accordance with the TCE Guidelines. The
wells contained relatively low concentrations of TCE (9 to 30 ug/1)
but fed directly into the distribution system with no blending.
To make up for lost production the City utilized the adjacent
distribution system which contained a mixture of groundwater and
surface water with a trihalomethane level of 87 fig/1. This trihalo-
methane level corresponded to an excess cancer risk of 295 per one
million, significantly greater than the excess cancer risk of the
TCE contaminated supply. ADHS is in the process of reevaluating
the Guidelines for TCE in Public Water Supplies to determine if
and how they should be revised to address this fissue.
Another issue that has been raised repeatedly, yet remains unre-
solved, is the questionable effect of well closure on subsurface con-
taminant movement. In the Tucson area, where some wells have
been closed for more than a year, significant increases in TCE
concentrations have been observed in downstream wells. It has
been suggested that the closure of a contaminated well allows the
contaminant, once contained by pumping, to migrate further. In-
creased pumping of nearby wells to make up for lost production
may actually promote the migration of the contaminant. The ve-
havior of TCE within the aquifer and the regional effects of pump-
ing upon the contaminant plume are not well understood. In the
absence of strong technical arguments, ADHS has elected to en-
courage closure of contaminated drinking water wells.
Finally, resources are always an obstacle when developing and
implementing new programs. The legislation which created the
WQARF, for example, did not allocate any funds for its adminis-
tration. And, although the fund may be supplemented by monies
recovered by civil penalty against hazardous waste dischargers,
enforcement is difficult with rapidly diminishing resources. It is
hoped that with a more efficient organization, the availability of
WQARF and CERCLA monies and new legislation to allocate ad-
ministrative funds the "TCE Response" will be much improved.
FUTURE ACTIONS
The response to groundwater contamination in Arizona has just
begun. ADHS will be involved in various activities over the next
several years related to both corrective and preventative measures.
The investigations and implementation of remedial actions in
areas of groundwater contamination will be continued to accom-
plish the following goals:
•To restore all regions of the aquifer used for drinking water to
acceptable levels for all contaminants;
•To minimize further movement of contaminants within the en-
vironment;
•To eliminate sources of contamination that are or may still be
contributing to groundwater pollution;
•To provide substitute water supplies to any impacted entity if
necessary for present and future water demands;
•To conduct investigations and monitoring activities to evaluate
all known'or potential sources of contamination; and
•To ensure that all costs for groundwater contamination assess-
ment and remedial action are borne by those responsible for the
contamination.
In the Tucson International Airport area, a considerable amount
of investigative work has already been completed, particularly at
the Hughes Aircraft facility. Since the facility is located on prop-
erty owned by the Air Force, the Department of Defense's In-
stallation Restoration Program (IRP) has been instituted at the site.
Interim remedial actions already have been proposed and are ex-
pected to be implemented within the next year.
Final remedial actions at the Hughes facility will be developed
after the total regional solution has been defined. It is likely that
all of the contamination in the area did not originate from a
single source. The City of Tucson, the State, USEPA, the Air
Force and Hughes are negotiating a memorandum of agreement
which will define their respective roles and responsibilities in the
source investigations, problem assessment and solution evaluation.
-------
427
STATE PROGRAMS
The Tucson International Airport area has been designated Ari-
zona's top priority site on USEPA's Interim Priority List for
Superfund (CERCLA), and it is also number one on the priority
list for Arizona's WQARF. These monies (WQARF = $250,000)
will be spent on the continued investigation of potential sources
and solution definition, development and implementation.
Several monitoring wells already have been installed in the area
to help define the vertical extent of the plume. Solute-transport
modeling will be utilized to project the extent of contamination
caused by a particular source and to predict the effect of pro-
posed remedial actions upon the regional aquifer.
Monitoring in the study area will serve a variety of purposes:
to verify the predictions of the models, to provide additional data
for further modeling, to determine the extent of contamination
caused by known sources, to identify other sources, to evaluate
the effectiveness and impact of interim remedial actions, and to
identify and evaluate final remedial actions.
If all sources of funding (DOD-IRP, EPA-Superfund, Arizona-
WQARF) are made available as anticipated, final remedial ac-
tion may be underway in the near future.
Two municipalities in the Indian Bend Wash area (near Phoenix)
have received the remainder of the WQARF monies. Both cities
will be required to submit proposals for removal or reduction of
TCE in their contaminated supplies unless the Indian Bend Wash
site is named to the USEPA's National Superfund Priority List.
If the site does make the final Superfund List, WQARF monies
could be used as State match.
Most field activities in Indian Bend Wash thus far have been
the product of a task force including representatives from the three
affected municipalities (Phoenix, Scottsdale, Tempe), the Ari-
zona Department of Water Resources, ADHS, the University of
Arizona and Salt River Project (an irrigation/utility district).
In view of limited resources, each of the participants will con-
tinue to contribute their individual expertise to approach the com-
plex problem of ground water contamination.
Groundwater contamination is a problem of which the nation
has only recently become conscious. Subsequently, it is not well
understood nor is it easily managed. Since groundwater moves
slowly, with little mixing or dispersion, pollutants remain in a rela-
tively well-defined plume which is highly concentrated and exists
for a long period of time. An underground plume is very diffi-
cult to define, requiring the installation of numerous expensive
monitoring wells. Although the plume may be contained by pump-
ing, clean-up of an aquifer is expensive, if it is possible at all.
For these reasons, prevention of groundwater contamination is
the best policy. The State strategy for the protection of ground-
water quality is many faceted, involving water quality control and
waste control programs. ADHS is in the process of developing
the following regulations and programs:
•Underground injection control
•Permitting of municipal and industrial discharges to groundwater
•Water quality standards
•Hazardous waste control
•Hazardous waste management facility
•Pretreatment program
Implementation of these new programs will require additional re-
sources. With diminishing federal funding, the success of these pro-
grams is questionable. The solicitation of increased financial sup-
port from the State Legislature will be an essential activity. Pro-
posed revisions to State statute would enable the ADHS to collect
fees for various activities. These fees would be placed in a fund
which would be used solely for the environmental programs.
CONCLUSIONS
Many have speculated that the discovery of TCE in the ground-
water is only the "tip of the iceberg"; it is feared that other more
harmful chemicals are present in the aquifer or will soon appear.
In all of the study areas, samples have been collected from the most
highly contaminated TCE wells to be analyzed for the 129 priority
pollutants. The results have not indicated the presence of contam-
inants other than those commonly found in association with TCE.
These contaminants are also industrial solvents, or components
thereof, which were used in association with TCE: dichloroethy-
teHe, trichloroethane, tetrachloroethylene, chloroform.
It is difficult to predict whether or not other contaminants will
be discovered in the future. ADHS will continue to evaluate in-
formation on land use and waste disposal to identify areas where
the potential for groundwater contamination exists and monitor
for a variety of contaminants. Through the development and im-
plementation of the strategy discussed in this paper, ADHS has
gained valuable experience. The necessary framework to respond
to any groundwater pollution crisis has been established.
REFERENCES
1. ADHS, Arizona Surface Impoundment Assessment, prepared by the
Division of Environmental Health Services, December 1979.
2. USEPA, Preliminary Site Inspection Report, Hughes Aircraft Com-
pany, USAF Plant #44, prepared by Ecology and Environment under
EPA Contract No. 68-01-6056, June 1981.
3. USEPA, SNARL for Trichloroethylene, Health Effects Branch, Cri-
teria and Standards Division, Office of Drinking Water, December
1979.
4. ADHS, Final Guidelines for TCE in Public Water Supplies, prepared
by the Division of Environmental Health Services, March 1982.
5. House Bill 2207, Arizona Revised Statute Title 36, Chapter 16, April
1982.
6. ADHS, Arizona Priority System for the WAter Quality Assurance Re-
volving Fund, prepared by the Division Environmental Health Services,
August 1982.
-------
IMPLEMENTATION OF A STATE SUPERFUND PROGRAM:
CALIFORNIA
ROBERT D. STEPHENS, Ph.D.
California Department of Health Services
Division of Toxic Substances Control
Sacramento, California
THOMAS E. BAILEY
Site Cleanup and Emergency Response Section
Sacramento, California
INTRODUCTION
The primary goals of the California hazardous waste manage-
ment program are to protect public health and the environment
by ensuring proper and safe handling, storage, transport, and dis-
posal of hazardous waste materials and cleanup of hazardous waste
sites requiring remedial action. The program is committed also to
the conservation of land, materials, and energy resources through
the promotion and support of waste reduction, waste treatment,
and resource recovery activities.
The hazardous waste management program in California is ad-
ministered through the Division of Toxic Substances Control in
the Department of Health Services. The program has experienced
rapid growth of responsibility since its beginnings in 1973. Legis-
lation since 1980, especially through the 1981 and 1982 legisla-
tive sessions, has added a number of key elements to California's
overall toxics management program.
The element to be discussed herein is implementation of the
California "Superfund" program which was established through
enactment of State Senate Bill 618 in September 1981.
CALIFORNIA SUPERFUND
Senate Bill 618, Statutes of 1981,' established a 10-year, $10
million per year, Hazardous Substance Account and a multitude
of program activities to be funded from that account. The total
amount in the account is to be raised through a tax on the gen-
erators who submitted hazardous and extremely hazardous waste
for disposal offsite, or who disposed of, on-site, such hazardous
wastes, including wastes from the extraction, beneficiation, and
processing of ores and materials; including phosphase rock and
overburden from mining of uranium ore.
The base tax rate is established following receipt and tabula-
tion of reports on the total amounts of waste disposed in the cate-
gories specified in statute. The account is limited to a maximum
funding of $10 million per year and accordingly, reductions to the
tax rates are made when unexpended funds are carried over from
prior fiscal years.
Program Elements
The key provisions for which the account is to be used include
the following:
•Funds to cleanup state hazardous waste sites that qualify as
Superfund candidate sites
•Ten percent state matching funds to meet federal Superfund elig-
ibility requirements
•Up to $1 million per year for assisting local agencies in respond-
ing to emergency spills of hazardous chemicals
•Funds to support local and state agencies response to hazardous
materials spills through the purchase of emergency response
equipment
•Up to $500,000 per year for health effects studies undertaken
regarding specific sites or specific substances at specific sites
•Up to $2 million per year for compensation of certain losses
caused by the release of hazardous substances
Program Staffing
A multidisciplinary team of engineers, chemists, toxicologists,
epidemiologists, physicians, waste management specialists, geolo-
gists and lawyers has been assembled to carry out the California
Superfund program. A multiplicity of inherent conflicts occur dur-
ing the conduct of a program such as Superfund. Complexities
are inherent in all phases of site impact assessment in establishing
criteria and protocols for remedial response in evaluating remedial
alternatives against selected mitigation criteria, in designing of
remedial response plans, and in assuring adequate public health
and environmental protection during and following site cleanup
activities. Resolution of these conflicts requires both competent
and aggressive technical and management staff. The 40-person
staff assigned to fulfill program responsibilities are located primar-
ily in the Toxic Substances Control Division in the following sec-
tions: Site Cleanup and Emergency Response, Hazardous Materials
Laboratory, Epidemiological Studies, and Air and Industrial Hy-
giene Laboratory.
SITE IDENTIFICATION
California's regulation of hazardous and nonhazardous waste
disposal to land was initiated in about 1950 through the orig-
inal agencies of the California Regional Water Pollution Control
Boards. However, regulation was limited to "official" disposal
sites and did not necessarily include onsite disposal, storage, treat-
ment and handling facilities. Therefore, while many significant dis-
posal sites were regulated, more were not at all or were regulated
based upon less stringent environmental and health standards than
exist today.
The state's proactive hazardous waste management program that
was established in 19732 has improved efforts to locate, regulate
and monitor sites used for disposal and/or sites of releases of haz-
ardous materials. Through the program's enforcement activities
and a special Abandoned Site Project search effort, sites were
identified as candidates for Superfund consideration. The pro-
cedures used in the Abandoned Site Project were reported prev-
iously.3
In summary, the project's methodology provides for the gather-
ing of all available data on a site to determine need for further
evaluation. Extensive review of business directories, tax rolls,
phone directories, and industrial association lists is made to lo-
cate inactive or potential abandoned sites. A survey questionnaire
is sent to all companies possessing certain standard industrial
classification (SIC) codes. The questionnaire seeks information on
disposal methods used prior to 1972 and on locations where waste
may have been disposed of prior tp that year. Leads to other sites
are developed through contact with community groups, individ-
uals, and searches of records of other public agencies.
Further evaluation includes a site "drive-by" inspection to de-
termine present use of the site, proximity of residents, evidence
of material disposed or leaving the site, and security of the site.
Evaluation may also include aerial photograph interpretation of
site history. Additionally, an on-site inspection may be conducted
428
-------
429
STATE PROGRAMS
and samples collected for laboratory analyses.
Results of the project indicate that in the 12 counties com-
pleted, 8,280 potential abandoned waste sites have been investi-
gated by staff of the project. Of those, 63 have been determined
to in fact be abandoned waste sites that would present a poten-
tial significant public health and environmental hazard. In the six
counties currently under investigation, 6,203 possible abandoned
waste sites have thus far come under investigation. Additional
potential sites are expected to be identified and contracted in those
counties in 1982. No projection as yet can be made on how many
of those sites may be determined to actually be abandoned haz-
ardous waste sites which present a health hazard. From the results
of the investigations conducted in the 12 completed counties, it
would be expected to be somewhat less than 1%. However, ex-
perience gained in the first counties investigated has educated
project staff in avoiding unproductive investigations and should
increase the ratio of actual sites to potential sites.
SITE RANKING
California's review and evaluation of the federal Hazard Rank-
ing System4 revealed that certain apparent deficiencies existed in
the method used for establishing priorities for remedial action
based on a total estimation of potential for environmental dam-
age or public health hazard. Additionally, California's law' re-
quires that the Department of Health Services adopt by regulation
the criteria for the selection and priority ranking of sites and pub-
lish a priority ranked list of sites targeted for remedial action. Ex-
penditures of state Superfund monies are to be based upon this
ranked list.
Accordingly, the California criteria were developed using the
federal system with significant modifications and additions to meet
the requirements of state statutes and the Department's mandate
for protection of public health. The resultant criteria6 produced a
relative ranking score which consists of the cumulative scores from
the federal method of ranking on the basis of the Migration Hazard
Mode, Fire and Explosion, and Direct Contact. In addition, two
categories, Toxic Hazard and Potential Hazard were developed
by state staff to represent variable waste concentrations and the
potential impact caused by releases of more than one chemical
waste.
The state considers concentration as a function of three routes:
acute toxicity, persistence and bioaccumulative toxicity and car-
cinogens.
The acute toxicity route is derived from the toxicological haz-
ard assessment in Dangerous Properties of Industrial Materials
by N. Irving Sax and from the concentration at the site. The per-
sistent and bioaccumulative toxicity route is based on the Cali-
fornia Assessment Manual for Hazardous Waste, which lists trigger
concentrations for inclusion as hazardous and extremely hazardous
wastes. The carcinogen route uses the Registry of Toxic Effects
of Chemical Substances as a standard reference to identify car-
cinogens and uses the concentration as a multiplier. This route has
an acceptable exposure limit of zero. Sites with any carcinogens,
no matter how small the concentration, will receive a score for this
route. The criterion further distinguishes between suspected ani-
mal carcinogens, known animal carcinogens, and known human
carcinogens.
The approach taken in the Potential Hazard criterion was in-
cluded to reflect the estimated potential for release of hazardous
substances into the air. This criterion addresses a specific problem
where the Mitigation Hazard Mode failed to provide sufficient
consideration. For example, zero points are assigned for potential
release of toxic gases or fumes. Under the state system sites receive
a score based upon this potential hazard.
REMEDIAL RESPONSE
Once the annual priority list of sites is created, the process of
administratively managing the list for remedial response in in-
itiated. The top 15 sites are selected for evaluation and for prep-
aration of site reports. The site reports include:
•Site description summaries
•Status of enforcement
•Description of the problem
•Identification of land ownership or responsible parties
•Proposed actions with tentative time schedules
From the site reports program strategies are formulated on the
course of actions to be proposed at each site, including a time
schedule and tentative estimate of costs. These strategies are used
to prepare the Superfund program plan and the annual budget.
All site remedial response activities flow through four phases:
(1) remedial investigation and site characterization, (2) develop-
ment of mitigation criteria and analysis of remedial alternatives,
(3) selection and design of the most effective remedial solution,
and (4) implementation.
Remedial Investigation
This phase involves all tasks necessary to determine the degree
of risk to public health and the environment, to characterize the
nature of wastes onsite and their areal extent and potential or ex-
isting migration and to collect data for establishing mitigation
criteria, evaluating alternatives and preparing the final remedial re-
sponse design. Activities in this phase include:
•Soil borings in sufficient depth and numbers to estimate the size
of the polluted area, and the contaminants on-site
•Quantitative chemical analysis of toxic substances in on-site
wastes and in surface and groundwater
•Air monitoring data to determine current emissions and possible
emissions during excavation
•Geophysical and hydrogeological site data
•Toxicological risk assessment of chemicals on-site
The conduct of this phase involves extensive coordination with
other public agencies involved in environmental regulation and
public health protection. Without complete interagency coordina-
tion and concurrent in the remedial response actons proposed to
be initiated, serious conflicts occur. Such conflicts slow down im-
plementation and, in severe cases, can stop progress.
Similarly, adequate public involvement must be planned for all
phases, starting with the initial site characterization field studies.
There is always pressure for an expedient cleanup response and
often the pressure is to proceed without a comprehensive assess-
ment of the extent of site problems and impact. The local pub-
lic must be informed and be afforded an opportunity to share in
the decisions relative to site mitigation. A fully informed, par-
ticipating citizenry can be a significant assistance to resolving prob-
lems associated with site mitigation. Such assistance is particular-
ly valuable when conflicting public agencies' opinions exist on
the procedures to be used to protect the public health. In Cal-
ifornia, citizen participation is a principle element of the remedial
response strategies.
Alternatives Assessment
Phase 2 of the process to attain cleanup involves a number of the
most fundamental steps required to assure adequate worker saf-
ety and public health protection. This phase includes an interpre-
tive analysis of the remedial investigation results, development of
criteria for site mitigation, and identification and evaluation of all
feasible remedial alternatives that lead to the selection of the least
environmentally sensitive, most cost effective solution.
Mitigation criteria will relate typically to levels of odors, con-
centrations of organic and inorganic chemicals in the soil, limits on
allowable air emissions for chronic and acutely toxic chemicals,
and for limits on potential chemical contamination of surface wat-
ers and groundwaters. Additionally, criteria are developed to ap-
ply to the short-term worker and public health protection and en-
vironmental problems that might be caused during actual site re-
medial response activities.
All alternatives to be considered must meet all public health
and environmental protection requirements. The first screening of
alternatives will include estimating costs to a 50% level of accuracy
and preparing an environmental assessment to consider major is-
sues of obvious primary significance and all other environmental
-------
STATE PROGRAMS
430
issues that might be of secondary influence on the final environ-
mental impact report prepared for the selected alternative.
Remedial Design
Subsequent to selection of the "best" solution for mitigation of
a site, the design phase is initiated. The consultant or staff must
prepare specific plans and specifications, including safety and
transportation plans, and construction documents complete
enough to be used in receiving and evaluating contractor's bids
for the site work.
The plan shows the staging of construction phases as required,
schedules for coordination with utilities, regulatory and public
health agencies and the public. Drawings are comprehensive and
detailed to show all elements of earthwork, areas of treatment,
neutralization or encapsulation, depth to background soils, in-
itial and final grades, location and design of monitoring stations
for water and air, location and measures for protecting environ-
mentally sensitive areas, location and design of air scrubbing sys-
tems, and measures to be implemented for short-term protection
during site disturbance activities. Selection and development of the
plan for final mitigation is done in close coordination with local
health, environmental agencies, and affected communities. Accep-
tance of the "best" remedial solution is difficult without such
broad participation.
Implementation
The contractor(s) selected to carry out the remedial response
plan must have experienced personnel and specialized equip-
ment in most hazardous waste site cleanups. In dealing with haz-
ardous waste, nothing is simple. Precautions during construction
must be backed up with support systems capable of reacting to on-
site emergencies. Also the contractor(s) must be able to mobilize
in a manner consistent with the safety plan and must have a site
specific safety training plan for workers.
Typically, the overall site safety plan includes provision for the
on-site representative to shut down or curtail site activities if
either worker safety or public health is threatened. In California
such on-site controls range from curtailment of some site activ-
ities-if for example, a low level of adverse air emission is measured,
up to a total shutdown of the day's site activities if the established
maximum allowable short-term air concentrations are exceeded at
a perimeter monitoring station.
The entire exercise of site remedial response design and imple-
mentation must be predicated on the principle that public health
protection is the controlling factor for both short-term on-site pro-
ject activities and long-term site mitigation criteria.
MITIGATION CONSIDERATION
Final selection of sites subject to remediation'can include cri-
teria other than those relating to public health and environmental
impacts. They may well include such factors as feasibility of remed-
ial action as well as political impact concerning a specific site in
the jurisdiction of a legislator. As mentioned earlier, the Depart-
ment was required by law to develop a set of criteria and a prior-
ity ranking of sites within the state which would be used for ex-
penditures of Superfund dollars. During this legislative session,
however, at least four bills were introduced which directed the De-
partment to expend a specific amount of monies for remediation
of a particular site.
Legislative "ranking" of sites for remedial action is seldom
based purely upon environmental and public health criteria. Leg-
islative direction may also mandate selected mitigation technolo-
gies without sufficient consideration for feasibility. This problem
results from the intense public focus on abandoned waste sites
and a legitimate response to these public concerns by local legis-
lators. This potential difficulty can be largely overcome by a
concerted effort in public, community, and legislative education
regarding Superfund programmatic activities.
Selection of alternative mitigation involves consideration of
many factors. Given that some remedial action is necessary at
the high priority sites, options for mitigation involve some form of
on-site containment or excavation and remote treatment and/or
disposal, or some combination of the two. Primarily, decisions
on which options are to be carried out are based on a technical eval-
uation of the effectiveness of selected mitigation technologies and
the relative cost of these technologies. Comparative evaluations of
on-site containment versus remote disposal and treatment are very
difficult for a variety of reasons:
•Uncertainties in the long-term behavior of many substances in
soils, liners, and aquifers. On-site containment may well involve
long-term operation and maintenance costs of such features as
pumping, leachate wells, and treatment plants.
•Excavation of problem wastes is confronted by such problems as
uncommonly high costs, and potential for excessive short-term en-
vironmental and public health impacts
•A significant problem exists for the remote disposal of excavated
material and the possibility of creating future problem sites from
such land disposal
•Significant problems are also incurred by the transportation of
large volumes of excavated materials
•Difficulty of treatment of contaminated soils
•Excavation of contaminated soils from problem sites causes par-
ticular difficulty in California in light of the newly published reg-
ulations on the restriction of a wide variety of materials from
land disposal. In view of the fact that these restrictions include
concentrated metal wastes and certain volatile wastes. Commonly
excavated wastes from problem sites may include these cate-
gories and therefore if excavated, must be treated prior to disposal
in the land.
•Migration of toxic substances through the air or water results in
low, subacutely toxic levels of a large number of substances.
The environmental and public health impact of these complex
mixtures is extremely difficult to assess. Assessment requires ex-
trapolation of high dose acute toxicity data to low dose chronic
exposure, assumption of toxicological interactions, and relative
sensitivities of environmental and public health targets. The pub-
lic and their representatives usually take a conservative health
view toward such uncertainties. Fund managers, be they public
or private, more commonly take a less conservative position.
When searching for a balance between cost and effectiveness,
cost estimates are in general much more solid than are effec-
tiveness estimates as measured in units of public health protec-
tion.
On a different level, commonly remediation is influenced by a
strong public community feeling that remedial action must in-
volve complete removal and remote treatment or disposal (i.e., re-
moval of the wastes from my backyard). Communities are very re-
luctant to consider on-site containment as an effective solution
and the concept of cost-effective solutions is not generally accep-
table. Strong community feelings toward remote disposal or treat-
ment can be translated into a legislative action mandating a par-
ticular remedial option.
California has approached these science, technology, public pol-
icy difficulties by making the whole process of site evaluation,
mitigation assessment, and final resolution as public as possible.
The uncertainties, technological, and fiscal limitations are openly
discussed. This approach to date has minimized the confronta-
tion between departmental program activities and the public.
REFERENCES
1. Carpenter-Presley-Tanner Hazardous Substance Account Act, Chapter
6.8, Division 20, Health and Safety Code, California Legislature,
Sacramento, Ca., Sept. 1981.
2. Hazardous Waste Control Law, Chapter 6.5, Division 20, Health and
Safety Code, California Legislature, Sacramento, Ca., 1972.
3. Casteel, Sawn, el al., "A Methodology for Locating Abandoned Haz-
ardous Waste Disposal Sites in California", California Department of
Health Services, Sacramento, Ca., July, 1980.
4. USEPA National Oil and Hazardous Substances Contingency Plan,
Appendix A, 40 CFR Part 300, Federal Register 47, No. 137,
July 16, 1982.
5. California Legislature, Chapter 327, Statutes of 1982.
6. State Superfund Site Selection and Ranking Criteria, California De-
partment of Health Services, Toxic Substances Control Division,
Mar. 1982.
-------
AN INTERNATIONAL STUDY OF CONTAMINATED LAND
M.A. SMITH
Director NATO/CCMS Study on Contaminated Land
Department of the Environment, Building Research Establishment
Garston, Watford, United Kingdom
M.J. BECKETT
Secretary, Interdepartmental Committee on the Redevelopment of Contaminated Land;
UK Representative on CCMS Study Group
Department of the Environment, Central Directorate on Environmental Pollution
London, United Kingdom
INTRODUCTION
In recent years many developed countries have faced problems
from land that has become contaminated with substances haz-
ardous to human health and the environment. International aware-
ness has been kindled by well publicized cases such as Lekkerkerk
in the Netherlands, Love Canal in the USA and Shipham in the
United Kingdom. These sites have attracted interest not only from
a technical, but also from a social viewpoint.
Each country has its own particular problems arising from a
wide range of hazardous substances that have contaminated the
land. In the USA hazardous, uncontrolled waste disposal sites have
been of greatest concern; in the UK it is the many former indus-
trial sites often located in urban areas where the demand for land
is high that have been discarded through economic and technolog-
ical changes.
Each country has its own national economic, social and polit-
ical priorities for dealing with contaminated land problems, and
must work toward their solution within the constraints of national
legislation. The OECD Waste Management Policy Group provides
an international forum for discussion of such issues in relation to
waste disposal sites. But, whatever the origin of the problem or
the context within which it arises, there is a common thread in
the technical requirement for effective, long-term remedial meas-
ures.
The variety of problems encountered in the UK suggested that
other countries might share similar experiences and in turn might
have complementary experience and solutions to contribute. Ac-
cordingly, the UK made a proposal in 1980 to NATO/CCMS for
a pilot study on contaminated land. In this paper, the authors
describe how the study group has set about its task and outlines
work in progress.
THE STRUCTURE OF THE STUDY
An inaugural meeting was held at the Building Research Sta-
tion, Watford, England in November 1981. Six countries were rep-
resented: Canada, Denmark, Federal Republic of Germany, The
Netherlands, UK and USA. A further planning meeting was held in
The Hague in April 1982 at which these countries confirmed their
interest in, and support for, the study.
Work to date has comprised the establishment of priorities and
work programs. A first meeting of experts to discuss progress on
each of the five active study areas (see below) took place in Wash-
ington immediately following this conference.
At the start of the study, each country outlined its own percep-
tion of contamination problems and identified topics of particular
inierest. In discussion, the group agreed that the study should
concentrate on land that could be regarded as contaminated ac-
cording to the following definition proposed by the UK:
"Contaminated Land" is land where substances are found that,
it present in sufficient quantity or concentration, could be haz-
ardous to workers, or to the eventual users or occupiers of the
site, or to a wider population due to transport of the substance
from the site, for example by wind action or pollution of water.
It also embraces the presence of substances that may be harmful
to plants and animals, pathogenic organisms, and substances
that are aggressive to building materials.
This definition hinges on the presence of hazardous substances,
and not the past use that has given rise to the existence of con-
tamination. It is therefore wide enough to encompass the uncon-
trolled hazardous waste sites of particular concern in the USA, as
well as the many types of former industrial sites (for example metal
mining and processing, chemical production and coal gas produc-
tion) of particular concern in the UK and Western Europe. It also
includes land that has become contaminated by over-application of
sewage sludge rich in "heavy metals" or other contaminants.
All participants, irrespective of any national policy context, saw
the need for identification or development of adequate techni-
cal solutions. It was agreed that a NATO study on the technical
aspects of tackling contamination would provide a useful comple-
ment to the OECD Waste Management Policy Group's work on
administrative and legislative issues. The most pressing require-
ment identified by all members was the need for remedial meas-
ures of proven long-term effectiveness and capable of meeting a
range of applications. Participants were agreed that such solu-
tions should be identified and, where they did not already exist,
developed taking into account both the variety of hazardous sub-
stances and the type of exposure, its targets, and risk after treat-
ment.
A two or three year pilot study could not hope to take, or bene-
fit from, new research initiatives. However, the value of the group
as a forum for exchange of expertise and experience meant that
substantial benefits might flow from evaluating existing national
research effort. The product would then be a practical review of
available remedial techniques and methods of assessing their per-
formance. Such a review would be of value by increasing the stock
of knowledge available to individual countries in the group and,
if published would benefit the wider international community.
The first meeting of participating countries agreed on the topics
that should be covered by the Pilot Study. They comprise two main
groups (Table !):(!) priority topics for active study, and (2) topics
of lower priority, to be restricted to an exchange of information.
Other topics were also considered for inclusion in the study—prin-
cipally questions of identification and assessment. However, con-
siderable published information already existed both from the
USEPA1'2 and the OECD.3'4
With limited resources and time, it was agreed that progress
could best be made by countries with particular national problems
setting out an initial review on a priority topic of particular inter-
est to themselves as a basis for wider debate in the CCMS study
group. Five projects are at present in progress following this ap-
proach and in the remainder of this paper, the authors introduce
briefly these five study areas. A tentative offer to lead a sixth pro-
ject (H) has been made by the USA.
431
-------
INTERNATIONAL
432
A In-situ treatment of contaminated sites—Lead country—USA
Methods of treating without excavation the bulk material on a contam-
inated site by detoxifying, neutralizing, degrading, immobilizing or other-
wise rendering harmless contaminants where they are found.
B On-site processing of contaminated soils—Lead country—The Nether-
lands
Methods of decontaminating or otherwise reducing the potential environ-
mental impact of the bulk of contaminated material on a site by: excava-
tion; treatment to detoxify, neutralize, stabilize or fixate; and usually re-
deposition on-site.
C Covering and barrier systems—Lead country—United Kingdom
To study the design of systems to prevent the migration of contaminants
vertically or laterally, or to prevent ingress of surface or ground water into
contaminated sites.
D Control and treatment of groundwater and leachate—Lead country-
Canada
Primarily concerned with those operations designed to control &r treat the
liquid phase on contaminated sites including design of cut-off systems,
hydrogeological modelling and leachate treatment.
Table 1.
Priority Projects
E Organic Chemicals and Plants
To study the tolerance to, and the uptake of, organic chemicals by
plants and their implications for site assessment, human health and site
reclamation. Initially the study would concentrate on a few substances
considered to be of general interest, e.g., creosols and chlorinated hydro-
carbons.
F Rapid on-site methods of chemical analysis
To find methods of chemical analysis allowing determinations to be made
on "soil", water and air samples on-site so as to speed and reduce the
costs of site investigations.
G Long term effectiveness of remedial measures—Lead country—Fed-
eral Republic of Germany
To collect information on examples of remedial and restoration actions
that can be demonstrated to have worked for a number of years and
methods for the monitoring of sites for long term effectiveness of remed-
ial measures.
H Toxic and Flammable gases—Lead country—USA (tentative offer)
To be concerned with problems in the measurement and assessment of the
amount of gas being produced in relation to the design of buildings to
prevent ingress of toxic and flammable gases.
I Social impact of contaminated sites
To collect information on the impact on, and the response of the publk
to contaminated sites.
Table 2.
Information Exchange Topics
ham that have attracted world-wide interest. It is considered valuable to
have reliable information on sites that are often referred to as examples of
the penalty of development of contaminated sites.
J Soil quality criteria
Collection and dissemination of information of national and other guide-
lines relating levels of contamination to acceptability of land for partic-
ular end-uses including site specific examples. It is not intended that any
attempt should be made to propose common criteria.
K Multi-national register of "important" sites
Collection in standard format of data on sites judged by participant coun-
tries to be of importance because of the nature of contamination or remed-
ial measures adopted.
L Information on key sites
The collection and dissemination of brief but comprehensive authorita-
tive statements about sites such as Lov6 Canal, Lekkerkerk and Ship-
M Contamination and specific industries
Collection and dissemination of information on the contamination char-
acteristics of specific industries or processes.
N Research register
Collection and dissemination of information on relevant research in a
standard format.
O Site identification in urban areas
Method of identifying contaminated sites in urban areas as opposed to
identification of specific site types, e.g., hazardous waste "problem"
sites on a national or regional basis, a problem that is being dealt with
adequately elsewhere.
STUDIES IN PROGRESS
In-situ Treatment of Contaminated Sites (A)
This project will deal with methods of treating, contaminated
material in bulk without excavation. Contaminated sites have typi-
cally been treated either by the excavation and removal of the
offensive material (for disposal elsewhere or permanent destruc-
tion) or by containment using cover often in conjunction with lat-
eral barriers. Both approaches may be expensive to implement.
Excavation and disposal of large volumes of contaminated ma-
terial may pose considerable logistical and environmental prob-
lems; many factors may affect the long-term effectiveness of cover-
ing systems and are the subject of further investigation. Hence
means of treating the mass of material without its excavation are
attractive. Existing technology in civil engineering, mining and
waste disposal offer a number of possibilities using injected fluids
to stabilize, neutralize or leach out contaminants. Other techniques
such as deep ploughing or electro-osmosis may also be applicable.
As a first stage the US are preparing a review of stabilization tech-
niques drawing on local experience.
On-site Processing of Contaminated Soils (B)
Project B is concerned with methods of decontaminating or
otherwise reducing the environmental impact of contaminated
material by treatment on-site. As pointed out above current meth-
ods for dealing with contaminated sites or "problem" hazardous
waste sites can present a number of technical and environmental
difficulties. In-situ techniques may also have considerable limita-
tions, principally of quality control to ensure, for example, that all
the contaminated material has been in effective contact with in-
jected fluids. A further option to current practice may rest in exca-
vating contaminated material and treating it on site prior to re-
deposition in a safe and stable form.
The Project will draw on experience at Lekkerkerk5 and on re-
search currently in progress in The Netherlands on other on-site
techniques. It will encompass simple mixing of clean and contam-
inated materials to reduce the concentration of the latter as well
as chemical processing and incineration techniques already em-
ployed for the disposal of hazardous wastes. The project will look
at the practical considerations of applying these technologies on
site as a means of avoiding substantial transport and disposal
costs.
Covering and Barrier Systems (C)
The objective in treating contamination by the technique of
containment is to prevent both the upward and outward migra-
tion of contaminants and the ingress of water. Much of the cur-
rent practice is adapted from experience with landfill sites where
they are used to limit the movement and volume of leachate.
A wide variety of different materials and techniques have been
employed from simple covering with soil materials to the use of
natural materials such as clays as "break" or "barrier" layers
-------
433
INTERNATIONAL
and impermeable membranes or "geotextiles". Often these tech-
niques have to be linked with adequate site drainage and inter-
ception of ground and surface waters.
Control and Treatment of Groundwater and Leachate (D)
Project D is predominantly concerned with those operations de-
signed to control or treat liquids on contaminated sites but also
includes some work on modelling of contaminant migration.
Pollution of surface and ground waters is often associated with
contaminated land. Many cleanup operations have required chem-
ical or microbiological treatment of considerable volumes of leach-
ate or polluted ground water. If they are to be effective, remed-
ial measures must control the volume of water and leachate and
incorporate adequate provision for the interception and treat-
ment. Whilst the physical and chemical role of barriers has been
subjected to some study, specific design requirements to meet
hydrogeological conditions are also sometimes necessary.
Long-term Effectiveness of Remedial Measures (G)
Project G seeks to identify examples of remedial and restora-
tive actions that can be demonstrated to have worked for a num-
ber of years. It will also identify suitable methods to monitor
the long term effectiveness of remedial measures on sites.
Reclamation schemes for contaminated sites are generally pre-
pared on the basis of current knowledge and professional judge-
ment. Schemes are often required to be effective for long periods—
from a minimum of 20 years to perpetuity. However, because
reclamation technology is still in its infancy there is little evi-
dence on the long-term effectiveness of treatment, and it is diffi-
cult to simulate field conditions in laboratory or other accelerated
testing regimes. As further treatment options become available,
monitoring of their performance will be required if both compara-
tive and absolute effectiveness are to be assessed.
CONCLUSIONS
Before setting out on any major new initiative concerned with
the protection of the environment, it is important to share ex-
perience. Only by doing this will real problems become apparent,
so that the world's resources of expertise and research can be de-
ployed to maximum advantage. The problem of contaminated land
is indeed one of the "challenges to modern society" on an interna-
tional scale and it is therefore fitting that NATO/CCMS should
have accepted the subject for this pilot study.
The work is still in its early stages, but it is to be hoped that
the experience of individual members of the group, some of which
is described in other papers to the conference, will highlight the
potential benefits of the study. Resources are limited, not least be-
cause the field is very active. Any further offers of assistance, or
requests to participate in the study, would be welcomed—whether
they come from other countries with particular experience to share
or from government's partners in industry who may have exper-
tise to contribute or may be concerned with developing products
for application in the field of land reclamation. Their active sup-
port for the study would be of special value if it broadened the
range of technical solutions available to deal with the problems
now being faced.
ACKNOWLEDGEMENT
This paper has been prepared as part of the research program of
the Building Research Establishment and is published by permis-
sion of the Director.
REFERENCES
1. Proc. of National Conference on Management Uncontrolled Haz-
ardous Waste Sites, Washington, D.C., Oct. 1980, HMCR1, Silver
Spring, Md.
2. Proc. of National Conference on Management Uncontrolled Haz-
ardous Waste Sites, Washington, D.C., Oct. 1981, HMCRI, Silver
Spring, Md.
3. Report of Expert Seminar on Hazardous Waste 'Problem' Sites, Paris,
1980, Organization Economic Co-operation & Development, Paris.
4. Supporting papers to OECD Expert Seminar on Hazardous Waste
'Problem' sites (Ref. 3)—available from OECD, Paris.
5. K. Strybis, Paper to Expert Seminar on Hazardous Waste 'Problem'
Sites, Paris 1980 (OECD Paris 1980) Paper ENV/WMP80Sem 15.
-------
LONG TERM EFFECTIVENESS OF REMEDIAL MEASURES
KLAUS STIEF
eltbundesamt, Federal Environmental Agency
Berlin, Federal Republic of Germany
INTRODUCTION
The long-term effectiveness of remedial measures is a complex sub-
ject, and conclusions cannot be generalized. Different remedial
measures, i.e., in-situ treatment or on-site processing of soil or
waste, covering and barrier systems, and control and treatment
of leachate must be discussed separately. Perhaps, it will even
be necessary in the future to make a more detailed differentia-
tion.
CLASSIFICATION OF REMEDIAL MEASURES
Initially, it seems to be reasonable to group remedial measures
into two categories:
•Remedial actions, where the contamination is changed
•Remedial actions, where migration of contamination is minimized
by artificial measures
Change of Contamination
This situation encompasses in-situ treatment and on-site process-
ing. Questions as to long-term effectiveness include:
•How long will it take until the required decrease of contamina-
tion in the soil is reached?
•Is the change of contamination stable or is remobilization of
contaminants possible? In what time period will this happen and
is it due to external influences?
•Can one restrict the migration of contamination? This condition
encompasses encapsulation of contaminated sites, but also may
include lowering of the ground water table or changing of ground
water flow.
•Does the effectiveness of remedial action change due to external
influences or due to the influence of contaminants?
•When must the remedial action be repeated?
•Is long-term management of remedial action possible?
If the migration of contaminants should be restricted, without
changing the contaminant itself, regular repetition of remedial ac-
tion will be necessary. Time periods between will not be longer
than usual in civil engineering solutions. Possibly they will be
shorter, due to contamination, which is not sufficient for evalua-
tion.
If hindrance of migration is combined with changing of con-
tamination, long-term efficient control is necessary and must be
assessed, of course, with regard to the time of decrease of con-
tamination.
LONG-TERM EFFECTIVENESS
The long-term effectiveness of remedial measures can be de-
termined by monitoring the efficiency of reduction of contam-
inant flow after construction or a given length of time (in the
case of continuous measures). Changes caused by time to the con-
taminants as well as to the measures to restrict migration must
be evaluated on the basis of laboratory and field tests.
In many cases of remedial action, measuring the effectiveness
for the total area under treatment is a significant problem. At
present, effectiveness can only be evaluated by measuring the de-
crease in impacts, e.g., on ground water pollution.
The disadvantages are obvious: (1) delayed evidence of effec-
tiveness or ineffectiveness, (2) uncertainties, due to proper or
wrong location of groundwater control wells, and (3) in case of par-
tial ineffectiveness, e.g., leakage in a cut-off wall, impossibility to
localize the failure.
MONITORING THE EFFECTIVENESS OF REMEDIAL
MEASURES
Effectiveness of Micro-encapsulation
Barry2 discussed, in a report on treatment options for contam-
inated land, the effectiveness and the monitoring of the effective-
ness, of described treatment options, but only in few cases did he
mention long-term effectiveness. It seems reasonable to cite Barry's
remarks or recommendations on "monitoring effectiveness" in
order to establish a basis from which requirements for long-term
effectiveness may be evaluated. With regard to micro-encapsula-
tion by shallow grouting, Barry assessed the possibilities for mon-
itoring effectiveness as follows:
"Standard engineering tests can be used for checking the re-
sultant permeability of treated ground. In many cases this can
be a good indicator of effectiveness. It is not practical, how-
ever, over a large area to monitor vertical permeability, a fac-
tor which will normally be critical in contaminated land".
If this is true, then only groundwater monitoring is an appro-
priate monitoring approach with the well-known disadvantage of
time lag and unreliability between release of contaminants from
the contaminated site and occurrence in monitoring wells.
Effectiveness of Grouting and Ground Leaching
Discussing "grouting in landfill", effectiveness and monitoring
effectiveness is referred to only, as: "...it can be difficult to test
efficiency of the grouting in place".
The importance of effectiveness of grouts or ground injection
to build up barriers against groundwater is perhaps greater if
ground leaching or in-situ detoxification of contaminants is the
remedial measure chosen. In these cases, the flow of contaminants
is large and they must be hindered from escaping into the environ-
ment. Whether long-term effectiveness of containment construc-
tion is necessary depends on the time necessary to decrease con-
tamination. Here, too, some difficulties arise concerning mon-
itoring effectiveness:
"As mentioned earlier when dealing with grouting, there is no
simple way of checking the full effectiveness of the application.
Analyzing leachate gives the average concentration; it does not
define the range of concentration. Similarly the residual con-
tamination can only be measured by soil sampling, at the same
locations as earlier samples".
434
-------
435
INTERNATIONAL
Effectiveness of Cut-off walls
Last, but not least, Barry assesses monitoring effectiveness of
cut-off walls:
"All forms of cut-off are imperfect and some leakage is al-
ways accepted in engineering applications. The extent of this
leakage can, of course, only be tested after the event and to
give any meaning to such tests it is necessary to have measured
the ground permeability before installation. This is particularly
true of grout courtains where, as discussed earlier, the grout
permeation factor is critical".
Effectiveness of Redevelopment of a Hazardous Waste Site
Describing investigation and reclamation at Thamesmead site,
Lowe1 discussed "short term and occasional risks". He stated:
"It would be unrealistic to assume that statutory or local
authority operatives will be kept informed of possible health
risks in the long term so where necessary, recommendations are
made for backfill of trenches to take a form which will reduce
the risk of contact and the spreading of foul material over clean
surfaces during re-excavation".
If Lowe is right in general, some remedial actions, in particular
ones such as containment, must be reconsidered.
With regard to remedial measures adequate for the Thames-
mead site, Lowe required long-term research.
"The most difficult aspect of reclamation at Thamesmead has
been the lack of any authoritative guidelines on such matters
as the upward migration of metals through covering layers and
the effectiveness of different forms of fill as barriers to con-
tamination. It must be said that under such circumstances a
'fail-safe' attitude has had to be adopted in deciding on remedial
measures in the knowledge that this may involve additional and
heavy costs both in advanced civil engineering work and the
construction of buildings.
"It is fortunate that research into these matters is now under
way and gratifying that Thamesmead has been chosen as one of
the sites where long-term research can be conducted. Liverpool
University are currently operating an experimental site".
Whether such an approach is appropriate to answer questions
about long-term effectiveness is discussed later in this paper.
Effectiveness of Redevelopment of former Gas Work Sites
In a comprehensive study on problems arising from the redevel-
opment of gas works and similar sites, Wilson and Stevens3
wrote:
"The effectiveness of the various remedial actions, particu-
larly in the long-term, has not been systematically studied. The
only documented evidence available appears to be at Battle,
case B.I, where analytical results are available both before and
immediately following the remedial work".
The above-mentioned follow-up survey at the Battle site, which
can also be said to monitor effectiveness, is described in the
following section.
Initial Survey
A total of 12 trial pits were dug; of these five were in clean soil and
no samples were taken; one pit hit concrete and was abandoned;
the remaining six pits were sampled...No groundwater was en-
countered in any of the pits.
In addition, two trenches were dug at right angles in the vicinity
of the pond. Heavily contaminated tarry material was found to a
depth of about 3m. Several pipes were observed to terminate in
the pond area and other pipe runs appeared to be heading for it...
The results show patchy contamination with much of the site be-
ing relatively clean. The clay soil has limited migration of the
pollutant. It was recommended that material from the pond area
and from other hot-spots adjacent to buildings be removed from
the site.
Remedial Measures
Contaminated material from "hot-spots" was excavated and
segregated into two heaps according to appearance and smell. The
most contaminated soil was sent to a landfill, while the least con-
taminated was reburied at the base of the excavation. The site
was then regarded and levelled.
Follow-up Survey
A second survey was undertaken with the aim of checking the
effectiveness of remedial measures. As housing with shallow foun-
dations was planned for the site, the pit depth was restricted in the
second survey to 0.3m.
One sample only was taken from the base of each of the 13 pits.
Analysis was restricted to species which were found to be of con-
cern in the first survey.
It was concluded that the remedial action had been successful.
Limitations were noted, however, with regard to the limited depths
of sampling and to the lack of information from areas covered by
concrete. In addition, the Regional Water Authority advised
against piling in order to minimize the possibility of groundwater
pollution.
Surface Sealing of a Landfill Site
Remedial actions proposed to abate and prevent pollution from
the Windham, Conn, landfill included:
•Regrading of the landfill to maximize surface water runoff and
minimize infiltration
•The placement of a 20-mil P VC top seal
•Covering the top seal with approximately 18 in. of final cover
•Revegetation
The remedial actions were designed to be passive to insure min-
imum future maintenance.
A monitoring system was installed; it consisted of suction lysi-
meters and pan lysimeters to determine the movement of mois-
ture through the refuse, a groundwater monitoring system con-
sisting of wells to determine fluctuations in the water table as well
as the rate and movement of leachate in the groundwater, and
surface water monitoring system consisting of staff gages in nearby
ponds. Monitoring of the landfill has continued for several years
establishing complete baseline data, and will continue for two years
following the installation of the remedial action alternatives to
determine their effectiveness.
This monitoring system is mainly to evaluate the effectiveness
of remedial action immediately after action has been finished.
Although that is, of course, important, there are some questions
that must be asked concerning the long-term effectiveness: (1)
When will it be necessary to replace the PVC membrane totally be-
cause of deterioration and subsequently loss of impermeability
against water? (b) How can damage of the membrane by digging
or ploughing, by plant roots, or mice or rats be detected? (c)
How can damage to the membrane due to settlement of the land-
fill body be determined?
The aim is to insure that future maintenance is a minimum be-
tween periods of replacement of the sealing membrane. The re-
medial action yeilds no "eternal" solution because erosion of the
sand and clay cover, subsequent damage to the membrane by
ultraviolet light or pinholes is to be expected. If this happens,
pollution will occur again, as hazardous materials in the landfill
body will be still available for further leaching.
CONCLUSIONS AND RECOMMENDATIONS
The effectiveness of widespread remedial actions can only be
proved in the field by a decrease in contamination, e.g., de-
creasing groundwater pollution.
"Absolute" effectiveness is in practice not achievable.
Inorganic contaminations will not change without specific
measures, e.g., chemical reactions due to injection of agents or due
to leaching processes. Thus encapsulation measures have to be re-
peated ad infinitum. Organic contaminants change very slowly
-------
INTERNATIONAL
436
without acceleration of microbiological processes. Like organic
contamination, organic material used for encapsulation, is not dur-
able forever. The same is true for inorganic materials, e.g., steel,
concrete, bentonite, etc.
Thus the aim of assessment of long-term effectiveness is to
find the point in time, at which a remedial action must be re-
peated. Constructional sealing measures can be expected to be
effective for a maximum period of 50-100 years, which is ac-
cepted as a reasonable depreciation period in civil engineering.
Whether particular chemical or microbiological strains at a con-
taminated site causes a shorter useful life time, must be checked
by monitoring.
Whether particular chemical or microbiological strains at a con-
taminated site causes a shorter useful life span, must be checked
by monitoring.
If the contaminants, which contained by capping or barrier-
system are not changed within this "calculated" lifetime, so that
the hazardous contamination, present at the beginning of remed-
ial action, has decreased to an acceptable threshold, a new re-
medial action scheme has to be designed. However, the degree of
contamination present at this time may require less stringent re-
medial methods.
The objective of monitoring these facilities is to check the
accuracy of the assumption of long-term effectiveness, so that the
remedial measures can be renewed before if needed at the appro-
priate time.
The obligation for continuous monitoring of remedial measures
such as capping or installation of barrier systems forces the devel-
opment of proper legal management, so that monitoring will in-
deed be performed.
How can one ensure that someone is responsible for contam-
inated sites where remedial measures have been taken but where
the actual contamination has neither been changed nor disposed
of?
How can one ensure that contaminated sites never will be for-
gotten but will be attended to forever, the remedial measures
being continually checked, repaired and renewed?
Remedial measures, which could be rendered ineffective by some
defect or other, should not be used unless the site can be perma-
nently monitored to see the containment system is still intact.
If remedial measures which require monitoring are chosen, they
should be such that they require checking only every six months,
or even better, every 1-2 years. In this way, is there a chance that
they will actually be monitored.
Monitoring has to be performed: (1) on site by visual control
and using a checklist, and (2) by analyzing samples of soil, sealing
material and/or groundwater.
When locating monitoring points and deciding how frequently
monitoring is necessary, the mobility of contaminants in ground-
water and soil must be considered. Future users of contaminated
sites and the responsible authorities have to be informed about
the contamination and the remedial measures taken. Appropriate
entries should be made in deeds to the land (in the Federal Re-
public of Germany the "Grundbuch").
If there is a change of usage of restored areas, i.e., a change
from industrial usage to installation of a children's playground,
the prior in-ground construction has to be checked. The obliga-
tion to check long-term effectiveness of a remedial measure (i.e.,
necessity of maintenance, repair and replacement or improvement
of measuring devices) also has to be transferred to the future user
of the site.
The new user of the site has to take over the maintenance, re-
pair and replacement of a remedial measure.
RECOMMENDATIONS
In writing a report for project "G", the author's first attempts
at compiling examples systematically, shows the subproject re-
port would contain, to a large extent, case-studies of remedial
measures of subprojects A, B, C and D, in which measurement
of long-term effectiveness would be particularly stressed, without
sharp definitions of long-term effectiveness. This would have
caused insufficient categorization of monitoring studies. The re-
port for subproject "G" would have become partly a repetition
of subproject reports A, B, C and D, as well as a supplement to
project report K "Register of important sites" or to project re-
port M "Register of key sites".
However, it seems to be important, to delineate questions for
"long-term effectiveness, in general or in an individual case so
clearly, that they can be answered by technical solutions. Answers
must be given in relationship to remedial measures.
What questions must be put to get answers to the question,
"What is the long-term effectiveness of remedial measures for con-
taminated sites"?
Questions on real effectiveness after finishing remedial action:
•Which parameters have to be measured in laboratory or in the
field to describe effectiveness?
•What values (range of values) describe insufficient/sufficient/
good/excellent long-term effectiveness?
•What changes of measured parameters for defining effectiveness
are possible/conceivable?
•What influence on the restored site, but also on single parts of
construction, are necessary to cause changes in the measured
parameters: (1) physical/remedial, (2) chemical, (3) biological,
and, (4) microbiological?
•Are changes in the measured parameters to be expected? Sud-
denly or continuously? If so, at what time and in what functional
relationship of time?
•How can effectiveness of remedial actions be measured in field?
•What relevance do the results of measurement in the field have?
•How can changes in the effectiveness be simulated?
•What relevance have results of simulations?
•What costs result, at what point of time, for measurements/sim-
ulation?
•What reserves are necessary for maintenance, repair and replace-
ment of remedial actions?
Answers to these questions can be collected in a subregister
"Long-term effectiveness of remedial actions" which has to be
connected directly with the register of important sites.
In a separate register "Long-term effectiveness measures" in-
formation, e.g., on durability of plastics, injection agents, mon-
itored remobilization of heavy metals, etc., could be collected,
which might be results of a case study or of an R & D project.
REFERENCES
1. Lowe, G.W., GLC Development at Thamesmead; Investigation and
Reclamation of Contaminated Land", Proc. Con/. Reclamation of
Contaminated Land, Eastburne, 1979, Society of the Chemical Indus-
try, London, 1980, B 7/1-15.
2. Barry, D.L., "Treatment options for Contaminated Land" Report
trising from Research Project of DOE, U.K., Atkins Research and
Development, July 1981.
3. Wilson, D.C. and Stevens, C., "Problems Arising from the Redevel-
opment of Gas Works and Similar Sites", Environmental and Med-
ical Sciences Division, AERE Harwell, Nov. 1981.
-------
LEACHATE TREATMENT AND MATHEMATICAL
MODELLING OF POLLUTANT MIGRATION FROM
LANDFILLS AND CONTAMINATED SITES
V.E. NIEMELA
K.A. CHILDS
G.B. RTVOCHE
Department of the Environment
Environmental Protection Service
Ottawa, Canada
INTRODUCTION
Pollution of both ground and surface water by toxic con-
taminants migrating from dangerous hazardous waste storage and
disposal sites, and particularly from areas of land contaminated by
various industrial activities, is a widespread problem. In this paper,
the authors review and attempt to assess the current remedial
technologies used in NATO/CCMS countries, with special em-
phasis on control and treatment of leachate and of contaminated
groundwater.
This overview includes the engineering design of various cut-off
barriers, grout curtains, clay fill trenches, etc. required to contain
contamination, but does not include an analysis of the physical and
chemical effectiveness of such barriers as it is covered in another
project. The focus of this paper is also on mathematical modelling,
a proven tool in site-specific situations but perhaps a more con-
troversial one from the viewpoint of general applicability.
INFORMATION GAINED BY NATO/CCMS COUNTRIES
The importance of operations designed to control and treat the
liquid phase on contaminated sites is more easily understood if one
remembers that approximately 90% (by weight) of the hazardous
industrial wastes produced in North America are produced as li-
quids; of these, 60% are organic and 40% are inorganic. The re-
maining 10% are sludges, slurries and solids. In other words the
methods used to prevent contamination of both ground and surface
water by toxic contaminants migrating from sites polluted by
hazardous waste are costly and complex not only because the
wastes are hazardous but also because they are mostly liquids,
usually delivered to these contaminated sites in tank trucks or in
drums.
In this paper, "contaminated sites" mean either special hazar-
dous waste disposal sites or landfills where hazardous waste is co-
disposed together with non-hazardous waste, and in one instance
an area of land polluted by a spill resulting from industrial ac-
tivities.
Based on the information and space available, this paper
discusses, in chronological order:
•Prevention techniques designed either to limit the quantity or
the danger of the leachate produced such as, new classification,
segregation, stabilization, co-disposal
•Containment techniques such as barriers
•Passive and active treatment techniques
Prevention Techniques
The form of wastes disposed on land will affect the quantity of
leachate produced. Since hazardous wastes in their untreated form
are predominantly liquid, the leachate they generate consists of
disposed liquids in addition to any component dissolved from other
wastes or the various types of soils found or brought to the site.
Conversely wastes in the solid form are obviously less likely to
generate as much leachate as that from liquid waste disposal. Thus,
stabilization of liquid chemical wastes prior to disposal has been
considered as an option designed to reduce the volume of leachate
generated. Various studies indicate that chemical stabilization or
inorganic sludges prior to disposal is a promising method to reduce
leachate generation. Recent studies indicate that stabilization of li-
quid organic wastes does not work as well as the treatment of in-
organic sludges, however, research is underway to develop pro-
cesses to also stabilize organic hazardous waste. Another approach
which seems promising is encapsulation of solid hazardous waste in
polymeric material so as to prevent the release of toxic constituents
to leachate.
Another management practice that influences the character of
leachate is segregation, that is the establishment of separate cells
within large landfill settings, or separate disposal sites, specifically
to receive particular wastes. Highly reactive wastes that threaten the
health and safety of landfill workers should obviously not be mixed
and disposed together. Similar consideration should also be given
to the disposal of hazardous wastes that generate leachate of an
especially unfavorable or dangerous character.
The disposal areas for different types of hazardous wastes should
therefore be segregated, where possible, to minimize interactions
between wastes, thereby simplifying predictions about the an-
ticipated characteristics of the leachate and how it might behave in
the hydrogeological environment if off-site migration does occur.
However, the co-disposal of compatible hazardous wastes that in-
teract to produce a leachate with favorable characteristics should
occur where feasible. Unfortunately, there is a lack of information
in the literature concerning the co-disposal of compatible wastes.
The above discussion emphasizes the necessity of a joint world-
wide effort to promote and establish some sort of new hazardous
waste classification system in order to assess the types of wastes
which have been or could be co-disposed with municipal wastes,
and to assess the compatibility of the various types of hazardous
wastes for co-disposal. Some types of hazardous wastes may be
considered imcompatible for co-disposal with municipal wastes or
other hazardous wastes due to chemical reactions which could
result in ignition, explosions or the release of harmful gases during
landfilling operations. The identification and differentiation of the
various types of hazardous wastes and an understanding of their
properties are required in order to control the behavior of the
wastes in disposal sites and assess their compatibility.
A classification system for hazardous wastes applicable to co-
disposal should ideally enable operators of co-disposal hazardous
landfill sites to provide a go/no go decision for the disposal of a
particular hazardous waste with municipal wastes or other hazar-
dous wastes. The waste producer, however, will likely identify the
wastes delivered to a disposal site according to their chemical com-
position or disposal properties. A hazardous waste classification
system should relate the various waste producing industries and
types of processes producing the wastes to the general chemical
composition of the hazardous wastes. Knowledge of the chemical
composition of the wastes is required to enable an assessment of the
compatibility of different types of wastes for co-disposal to be
made.
437
-------
INTERNATIONAL
438
A number of classification systems for categorizing wastes by
their origins or by their properties have been reviewed and in-
vestigated for suitability as guides for determining whether specific
industrial wastes may be safely co-disposed and landfilled. These
classifications have not, generally been prepared specifically for co-
disposal purposes, and not surprisingly each shows some inade-
quacy for this use. Any system providing data so that landfill
operators can determine the acceptability of a given waste class will
tend to be either cumbersome (due to the large variety of chemical
wastes and disposal parameters), or oversimplified. An interna-
tional effort in the classification field may be therefore needed as
mentioned before.
Containment Techniques
Restricting Infiltration
Because infiltration of precipitation and surface runoff provides
much of the moisture responsible for the generation of leachate
within a landfill, it is important to minimize the amount of infiltra-
tion into the landfill during and following the deposition of the
wastes. This can be achieved using a variety of cover materials and
engineered designs.
Various cover materials have been used for capping. The most
commonly used covers have consisted of natural geologic materials
such as clays and silts; however, synthetic liners or membranes may
be used to complement, or in place of, soil covers. In normal land-
fill practice, a compacted soil layer is placed over each day's waste
to minimize the influence of wind and infiltration, in addition to
the final thick cover of compacted soil applied prior to close-out.
The choice of cover material and design can also be influenced by
other functions to be served by the cover: minimizing the effects of
surface water erosion, resisting long-term deterioration under par-
ticular climatic conditions, providing a base for vegetative growth,
controlling animal and insect burrowing and minimizing settlement
by maximizing compaction.
Several documented cases of failure of covers of geologic
material on landfills occurred in 1981 in the State of New York.
These failures were primarily a result of poor compaction of waste
materials and the introduction of excessive quantities of liquid
hazardous wastes (or infiltration water) during the active life of the
landfills. Following the close-out and the emplacement of final soil
covers, differential settlement occurred, cracks and fractures
developed and excessive infiltration resulted. Leachate generation
at these sites has been so excessive that leachate had to be pumped
from the landfill and treated following collection.
In the literature, there is little quantitative information on the
performance of covers on landfills, neither with respect to the long-
term infiltration rates through the covers, nor with respect to the
long-term effects of erosion or weathering. A study of soil and
vegetation covers on municipal landfills in Florida indicated that
these covers were only of use in reducing surface infiltration during
severe rain storms.
Restricting Leakage from Landfills
The landfill in a shallow hydrogeologic setting cannot be totally
"secure" in areas of humid climates because, regardless of
engineering design or practice, some contaminants will be released
to the environment, although the rate of release may be very slow
and the quantities released may be very small. By adopting various
engineered controls, however, the introduction of contaminants to
local, off-site groundwater systems may be limited to acceptable
levels.
Leakage to the surface via springs or side seeps can be eliminated
if the "bathtub" effect is not allowed to develop within a landfill.
the "bathtub" effect results as a consequence of the rate of in-
filtration into a landfill exceeding the rate of seepage out. In humid
climatic regions and where a landfill is located in low permeability
material, or has a liner of low hydraulic conductivity, the landfill
can gradually fill with leachate if excessive infiltration occurs
because of defects in the landfill cover. When the landfill fills,
leachate springs or side seeps develop around the perimeter of the
landfill at the ground surface. To prevent this type of contaminant
release, it is essential to maintain cover materials so that the rate of
infiltration does not exceed acceptable rates of seepage through the
base of the landfill to the groundwater.
To prevent unacceptable rates of seepage from the base of land-
fills to the shallow subsurface environment, landfill liner systems
have been used with some degree of success. Liners can be con-
structed of geologic and/or synthetic materials, and will not be
discussed in this paper as their specific properties and effectiveness
are covered in another project.
Grouting is another engineering approach used to stabilize waste
in place. Basically appropriate materials are injected into porous
geological units to reduce hydraulic conductivity which in turn can
restrict the movement of contaminants.
A variety of grouting materials have been used in engineering
practice: cement, cement slag, resin, asphalt and several different
colloidal and low viscosity chemicals. The choice of grout used for
a specific problem would depend upon the desired lifetime and the
hydraulic conductivity of the unit.
In engineering practice, grouting is often used to construct a low
hydraulic conductivity wall or curtain in the subsurface. Because
grouts in unconsolidated materials may not invade the formation
evenly or very far, it may be necessary to grout along closely spaced
lines of boreholes to achieve the required degree of control. En-
vironment Canada believes that grouting techniques are unsuited to
the problem of controlling hydraulic conductivity throughout
several extensive deep geological units which include relatively im-
permeable zones.
Other engineering modifications can improve the level of subsur-
face environmental protection in the vicinity of hazardous waste
landfills. These include slurry walls, vertical grout curtains, in-
terceptor trenches and wells, drains at the surface to intercept
leachate springs, and various pumping schemes to create hydraulic
gradients toward the landfill and help prevent groundwater flow
and contaminant migration away from the landfill site.
All these measures are very useful but probably do not provide a
permanent solution. Today there is uncertainty as to the long-term
effectiveness of "secure landfills." There is doubt that the "secure
landfills" built in the U.S. and Europe will be totally effective for
permanent containment of contaminants. For instance the landfills
in Ontario that have been used almost exclusively for the disposal
of chemical wastes have all caused problems due to the off-site
migration of contaminants. There is evidence that the larger
municipal landfills where the same materials have been placed in
co-disposal with a< variety of wastes have caused less of a problem.
Studies that have been done in Ontario at major landfills in
which large quantities of liquid industrial waste have been dumped
do not appear to have pollution hazards that are significantly
greater or different from those where liquid waste has not been
dumped. There is evidence that highly permeable ground, par-
ticularly ground that has a low lime content, may be undesirable for
deposition' of large quantities of liquid industrial waste. The limy
soils in southern Ontario, apparently have the capacity to handle
large quantities of industrial waste without adverse effect.
Treatment Techniques
Regardless of preventive measures adopted to limit leachate
generation, a leachate is still likely to be formed, especially in areas
where precipitation exceeds evaporation and/or at sites used for
disposal of liquid hazardous wastes. The treatment of that leachate
will be simplified if certain types of wastes were segregated in
separate cells at the site or in certain cases excluded altogether.
Leachate collector systems have recently been designed in con-
junction with lined landfills. These systems allow the collection of
excess leachate and its treatment prior to its ultimate disposal or re-
circulation back through the landfill with the objective of further
attenuation. Such active control requires continuous on-site super-
vision and maintenance and means that such a hazardous waste
landfill is not a disposal facility but a waste storage facility.
-------
439
INTERNATIONAL
Because hazardous waste leachates vary widely in composition
and often contain a diversity of constituents, the actual treatment
technologies are case specific and often comprise a combination of
different treatment processes. However, the review and description
of these processes is beyond the scope of this paper. The USEPA
lists for example 20 possible treatment processes.
Some brief examples of remedial actions taken in Canada and by
some NATO/CCMS countries are listed instead.
At a polychlorinated biphenyl (PCB) spill in Regina, Saskatche-
wan, Canada, contaminated glacial till was excavated for treatment
and disposal. A slurry wall to a depth of approximately 10 m in the
glacial till around the entire spill site was emplaced in an effort to
restrict contaminant migration to the surrounding hydrogeologic
environment. The slurry wall was installed between temporary
metal sheet piles driven into the ground which were subsequently
removed following the pouring of the bentonite slurry. Although
this was not a designed disposal site the remedial actions taken
could be applied in the vicinity of hazardous waste disposal sites.
Another example of remedial action is that in 1981 at the Rocky
Mountain Arsenal waste disposal site in Denver, Colorado. A com-
bination of a slurry wall to impede groundwater flow and a series of
interceptor wells through which contaminated groundwater was
removed and treated prior to its return to the groundwater
through wells down gradient of the slurry wall was used to rectify
the groundwater contamination problem.
As regards treatment of groundwater polluted by contaminants
migrating from soils contaminated specifically by hazardous waste
little data is available, as most of it is related to ordinary solid waste
disposal sites. Four cases of groundwater treatment in the F.R.G.,
are summarized as follows:
•A paper given in 1981 in the Netherlands says that oxidation pro-
cesses in an aquifer caused by injection of oxygen containing
water can improve the groundwater quality and protect the
groundwater against pollutions. The so-called subterranean
groundwater treatment has been applied in several European
countries for some years. Stream and transport mechanisms and
chemical and biological reactions as well were described. The
recharge system, in a most practical manner, uses the pumping
well for injection of oxygenated water into the soil.
•Another investigates the treatment of groundwater contaminated
by methyl chloride by the activated carbon method.
•A third describes the treatment of groundwater polluted by a
spill of mineral oil and by leachate from a chemical waste dis-
posal site. Water from four deep wells is pumped to a reaction
tank, mixed with ozone and then allowed to infiltrate back into
the ground. Two problems with this approach are: (1) infiltra-
tion changes the elevation of the groundwater table and hinders
the flow from the source of contamination to the water catch-
ing area, and (2) oxygen in the water increases biological activity.
On the positive side ozonization decreased the chemical oxygen
demand (COD).
•Finally an article "Seeping and distribution of mineral oils and
chemicals in the ground", published in May 1982, describes the
"purification procedures of saturated aquifers" and gives more
detailed information on the distribution of oil in the groundwater.
MATHEMATICAL MODELLING OF POLLUTANT
TRANSPORT BY GROUNDWATER
The Nature of Mathematic Models
Fried4 has noted that the modelling of groundwater pollution
consists of describing, by mathematical expressions, the following
processes:
•The advection of the contaminant, i.e., its transport by ground-
water flowing at the mean velocity
•The dispersion of the contaminant, i.e., its spreading due to
velocity variations of the groundwater within pore spaces about
this mean value
•The chemical interaction of the contaminant with the solid
matrix of the aquifer system, e.g., adsorption
•The decay of the contaminant due to biodegradation or radio-
active disintegration
These last terms describe the attenuation of the contaminant.
Mercer and Faust6 have pointed out that hydrogeologists have
available to them two types of mathematical models with which to
analyze groundwater flow and contaminant transport:
•Analytical models, in which a simplified form or transport
equation is solved by an exact solution for the initial and boun-
dary values stipulated
•Numerical models, in which the partial differential equation
describing the behavior of a continuous variable are approxi-
mated numerically by finite-difference or finite-element methods
resulting in a finite number of algebric equations which may be
solved by matrix techniques
Since 1960, much interest has been shown in the development of
numerical models of groundwater flow (i.e., the advertive flux) and
contaminant transport (i.e., the advective and dispersive fluxs with
various attenuation terms).
In the following section, the authors give a brief review of the
utility of mathematical modelling for studying both flow and
transport problems at waste-disposal sites. The usefulness of the
procedures depends entirely on acquiring adequate, reliable data
for use in the modelling operation. No more eloquent a caution can
be given than the concluding remarks of Wang and Anderson10 in
their book Introduction to Groundwater Modelling:
"The answers generated using a mathematical model are
dependent on the quality and quantity of the field data
available to define the input parameters and boundary
conditions. Modelling can never be a substitute for field
work. Used in conjunction with good field data, a model
can provide insight into the dynamics of the flow system
and also serve as an invaluable predictive tool."
Groundwater Flow Models
Following the acquisition of a complete set of data defining the
hydrogeologic condition and properties of the groundwater flow
system within which the waste-disposal site is situated (e.g.,
hydraulic heads and conductivities, geometries of the permeable
and impermeable units comprising the flow system), it is possible to
begin to model the pattern of groundwater flow by numerically
solving the groundwater for equation (Mercer and Faust6):
d K
dh
dh|t R = S dh
sdt
(D
Where x, y and z are the spatial coordinates, KM, K^,, KH are the
hydraulic conductivities (L/t) in the x, y, and z directions, h is the
hydraulic head (L), R is a source or sink term (1/t) and Ss is the
specific storage (1/L).
A best fit of simulated, with observed hydraulic heads, is ob-
tained by trial and error adjustment of the geometries and
hydraulic conductivities. The first attempts at modelling will likely
result in a desire to collect further field data so that better fits of
simulated with observed values can be obtained. Consequently,
there will be cycles of field and modelling studies resulting in the
prediction of the hydraulic head pattern within the flow system
from which groundwater and contaminant flow patterns may be
deduced. Furthermore by employing Darcy's Law, which relate the
specific discharge of groundwater (q) to the hydraulic-head gra-
dient (dh/dl) and hydraulic conductivity (K), a map of the ground-
water velocities (V) may be drawn in which:
v = q/n = K dh/dl
(2)
where n is the porosity. The velocities are essential to the next step
of groundwater modelling.
-------
INTERNATIONAL
440
Contaminant Transport Models
During the 1960s, it was customary to study contaminant
transport in groundwater for systems either by the use of analytical
solutions' or by groundwater flow models-in the method described
above3. In the last 12 years, however, great progress has been made
in developing numerical solutions to the advective-dispersion equa-
tion describing contaminant transport (Wang and Anderson10):
3 £C .V dC=d£
t dy2 x dx dt
(3)
where C is the concentration of contaminant in question and DL
and TT are the longitudinal and transverse dispersion coefficients of
that contaminant, which can be written:
D = a, V
L L x
and
(4)
(5)
Therefore, the dispersion coefficients are a measure of the rate at
which the concentration gradients are dissipated. aL and aT are the
longitudinal and transverse dispersivities (Units: L) and can be con-
sidered to be measures of the dispersive capacity of the
hydrogeological system being modelled.
All early numerical modelling of contaminant transport treated
the dispersivity as a constant which could be used as a fitting
parameter in the same way as the hydraulic conductivity is used in
groundwater flow models. Consequently, a wide range of disper-
sivity values were reported, from centimetres in sand aquifers to
tens of metres in fractured basalt rocks. Pickens and Grisak9 have
shown that this is partly due to the heterogeneous nature of
hydrogeologic systems and the dilution of groundwater samples
when large-diameter wells rather than small-diameter sampling
systems are used. Since dispersion, causes dilution of the contami-
nant, any dilution of the contaminant clearing sampling implies a
greater dispersivity.
In recent years, groundwater modelers in the USA, Canada and
France have begun to treat the dispersivity value as a function of
the mean travel distance or travel time of the contaminant.
Stochastic modelling techniques used by Gelhar in New Mexico and
de Marsily in France suggest that the dispersivity is dependent on
the time of travel of the contaminant from the waste-disposal site
and should reach an asymptotic value at large travel times when or
distances where hydrogeological boundaries are encountered or a
significant transverse velocity develops.2
Many contaminant transport problems require a chemical reac-
tion term to describe the adsorption of heavy metals, radionuclides,
organic compounds or other contaminants during their transport
through the groundwater flow system. This can be done by adding
the term - pbKD (dC/dt) to the left-hand side of Eq. 3 in which pb is
the bulk density of the aquifer material and KD is the distribution-
coefficient describing the partitioning of the contaminant between
adsorbed and solution phases (Units: L/mass).
The distribution coefficient is usually measured by laboratory
batch procedures in which care is taken to reproduce the ground-
water environment. The application of the KD approach assumes
that the adsorption process is instantaneous and reversible and that
it is described by a linear adsorption isotherm. When the contami-
nant is present only in trace quantities, the quantity observed fre-
quently does vary linearly with its concentration in solution,
however, the adsorption reaction is not always fast or reversible
and therefore may not reach equilibrium. Much research work in
NATO/CCMS countries is currently diverted at finding improved
ways of measuring adsorption and other chemical reaction terms in
groundwater models, including the development of in-situ KD
measurements which produce operationally useful parameters but
of little or no thermodynamic significance.
Consequently because of difficulties in measuring and inter-
preting values of dispersivity and the distribution coefficient, con-
taminant transport modelling is still at an evolutionary state of
trying to obtain an improved understanding and approximation of
basic phenomena.
Application to Remedial Measures
The nature of groundwater flow and contaminant transport
models have been outlined and the opportunities for their use in
decontaminating aquifers and/or controlling contamination in
aquifer systems have been discussed. In fact their ability to predict
system responses to remedial actions make them essential to any
program of site restoration. Cole and McKown of Battelle1 have
summarized their uses in such situations as:
(1) To aid in the design of the site investigation program
(2) To help assess whether remedial measures are required and
which ones would be the most effective
(3) To assist in the design of a site monitoring program
To date, the authors are aware of the use of mathematical
modelling techniques for remedial-measures work by several
groups in NATO/CCMS countries, e.g. Battelle (PNL, Richland,
Washington, USA), GEOTRANS (Washington, D.C., USA) and
GTC (Ottawa, Canada).
In the future, hydrogeologists will have available to them not on-
ly powerful groundwater flow and contaminant transport models,
but also combined transport and optimization codes for the assess-
ment of waste disposal operations and decontamination-well loca-
tion. However, all models will require a prodigious amount of
reliable hydrogeologic (field) information without which the
modelling will be ineffectual.
SUMMARY
In summary groundwater models provide an indispensable aid
in:
•Evaluating the hydrogeologic conditions at waste disposal
sites and determining the additional field activities (e.g., drilling,
piezometer installation) to be conducted
•Predicting the migration of groundwater contamination (from a
calibrated model)
•Assessing the likely effects and costs of various remedial measures
(e.g., purge wells, grout curtains, etc.) on contaminant migra-
tion and concentration
The model results, however, depend on the quality and quantity
of the hydrogeologic data used. Nevertheless, allowing for the
uncertainties in the data, groundwater modelling permits rational
development of a program of remedial-measures at landfills and
contaminated sites.
REFERENCES
Modelling
1. Cole, C.R. and McKnown, G.L., "The Use of Mathematical Models
to Assess and Design Remedial Action for Chemical Waste Sites".
2. de Marsily, G., Dieulin, A., Ledoux, E., and Goblet, p., "Are We
Able to Measure the Parameters Governing Transport of Solute.in
Porous Media, and thus, to Predict Long Term Migration?" Inter-
national Workshop on the Comparison and Application of Mathe-
matical Models for the Assessment of Changes in River Basins, both
Surface Water and Groundwater, UNESCO, IHP, La Corunna,
Spain, Apr. 1982.
3. Freeze, R.A., "Subsurface Hydrology at Work Disposal Sites", IBM
Journal of Research and Development, 16, Mar. 1982.
4. Fried, J.J., "Groundwater Pollution Mathematical Modelling: Im-
provement or Stagnation?" In Quality of Groundwater, Elsevier
Scientific Publishing Co., The Netherlands, 1981, 807-822.
5. Gorelick, S.M. and Remson, I. "Optimal Dynamic Management of
Groundwater Pollution Sources," Water Resources Research, 18,
Feb. 1982.
6. Mercer, J.W. and Faust, C.R., Ground-Water Modelling, National
Water Well Association, 1981.
7. Ogata, A. and Banks, R.B., "A Solution of the Differential Equa-
tion of Longitudinal Dispersion in Porous Media," U.S. Geol.
Survey Professional Paper 411 A, 1961.
8. Pickens, J.F. and Grisak, G.E., "Scale-Dependent Dispersion in a
Stratified Granular Aquifer," Water Resources Research, 17, Aug.
1981.
-------
441
INTERNATIONAL
9. Pickens, J.F., Jackson, R.E., Inch, K.J. and Merritt, W.F., "Meas-
urement of Distribution Coefficients Using a Radial Injection Dual-
Tracer Test", Water Resources Research, 17, June 1981.
10. Wang, H.F. and Anderson, M.P., introduction to Groundwater
Modelling, Freeman, San Francisco, Calif., 1982.
Others
11. "A case study of a spill of industrial chemicals—polychlorinated
biphenyls and chlorinated benzenes" National Research Council
Canada, NRC 17586, 1980.
12. A. Golwer, "Seeping and distribution of mineral oils and chemical
in groundwater". F.R.G. May 1982, (in German).
13. Conestoga-Rovers & Associates "Identification of Policy Options
Regarding the Discharge of Wastes onto Land", DSS Canada,
02SZ.K204-6-EP100.
14. F. Dietzel "Case study of methyl chloride contamination" F.R.G.
1981 (in German).
15. G. Nagel "Decontamination by ozone of groundwater polluted by
mineral oil spill" F.R.G. 1981 (in German).
16. Colder Associates, "Landfill research activities: the co-disposal of
hazardous wastes" DOE, MOE, Canada, May 1982.
17. Landreth, R.E. "Guide to the disposal of chemically stabilized and
solidified waste", USEPA SW-872 September 1980.
18. Shuckrow, A.J., Pajak, A.P. and Touhill, C.J., "Management of
Hazardous Waste Leachate" USEPA SW-871 September 1980.
19. U. Rott "Protection and Improvement of ground-water quality
by oxidation", F.R.G. 1981 (in German).
ACKNOWLEDGEMENT
This paper is based on recently developed data from Canada and
other NATO/CCMS participating countries. In the field of modell-
ing the authors with to acknowledge the participation and the input
of R.E. Jackson, Ph.D. of the National Hydrogeological Research
Institute, Environment Canada.
-------
DEVELOPMENT OF AN INSTALLATION FOR ON-SITE
TREATMENT OF SOIL CONTAMINATED WITH
ORGANIC BROMINE COMPOUNDS
W.H. RULKENS, Ph.D.
J.W. ASSINK
WJ.Th. VAN GEMERT, Ph.D.
Netherlands Organization for Applied Scientific Research
Division of Technology for Society
Apeldoorn, The Netherlands
INTRODUCTION
Soil contamination is one of the most pressing of the environ-
mental problems existing in the Netherlands. Over the last few
years particularly, much attention has been given to this subject. At
the present time, a large number of hazardous waste sites that need
to be cleaned up have been discovered.
Current methods for treating highly contaminated soil mainly in-
volve excavation of the soil followed by thermal treatment
elsewhere. This method is rather expensive and furthermore cannot
be applied in all cases. Thus there is a need for less costly methods
of cleaning up contaminated soil.
A large number of process alternatives for cleaning up con-
taminated soils can be given. Some examples of these alternatives
are extraction, chemical conversion and biological degradation of
the contaminations in the soil. With the exception of thermal treat-
ment, however, no other method of treatment has been developed
for practical use so far. Much research and development work still
have yet to be carried out with the attendant practical problem that
eacK type of soil contamination is different. This also means that
the method of treatment has to be adapted in each particular case.
The investigation described in this paper deals with the develop-
ment of an on-site treatment method for a soil strongly con-
taminated with organic bromine compounds. The contaminated
site is located in the neighborhood of the Dutch municipality of
Wierden. The main problem is the potential danger of the con-
tamination of groundwater used for the production of drinking
water. To clean up this site, the Ministry of Public Health and En-
vironmental Protection ordered the Dutch research institute TNO*
to investigate the possibilities of an on-site treatment method based
on extraction. This investigation, carried out in cooperation with
HBG**, a Dutch company specializing in transport and handling
of soil, was finished recently.
*TNO: "Netherlands Organization for Applied Scientific Research".
"HBG: "Hollandsche Beton Groep N.V."; P.O. Box 81, 2280 AB Rijswijk, The
Netherlands.
CHARACTERISTICS OF THE CONTAMINATED SITE
A schematic presentation of the contaminated site concerned is
shown in Fig. 1. This figure shows a schematical cross section of the
site at which a factory for the production of organic bromine com-
pounds was formerly located.
The penetration depth of the bromine contaminations generally
varies between 4 and 7 m; the groundwater level is approximately 7
m deep. The total surface area of the contaminated site amounts to
about 300 m2 while the total amount of contaminated soil at the site
is estimated to be some 30,000 tons.
Several types of contaminated soil are distinguishable. The bulk
of the soil is yellowish brown sand with a particle size distribution
from about 40 to 300 /im. The top and sub-top layers contain a
relatively large amount of organic humus-like substances. Besides
sand and humus-like substances the contaminated site contains
BORDER OF THE CONTAMINATED SOIL
GROUNDHATER LEVEL
Figure 1.
Cross Section of the Contaminated Site
several clay tongues. Based on analysis of a large number of
samples taken from the site, it is estimated that the quantity of
humus-like substances and the quantity of mineral particles smaller
than 40 pm both amount to about 1 % of the whole.
The contaminants in the soil almost entirely consist of organic
bromine compounds (Table 1). Some of these compounds is rather
volatile; most of them have a low solubility in water. However, the
concentration of the bromine compounds varies markedly within
the contaminated site. The highest (total) bromine content
measured was about 3000 mg/kg of soil in some locations in the top
and sub-top layer. In the investigation described in this paper the
total bromine concentration in water and soil was always measured
by means of neutron activation analysis. The detection limit of this
method is about 0.6 mg/kg. The soil was considered to be clean if
no total bromine could be detected by this method of analysis.
Table 1.
Average composition of the contamination in three
highly polluted samples of soil
Compounds
Dibromo ethylene
Tetrabromo ethanes
Tribromo ethylene
Dibromo butanes
Dibromo alkanes (C5 - C10)
Other bromine compounds
Concentration
(mg/kg
90
690
170
80
1040
1500
442
-------
443
INTERNATIONAL
LABORATORY SCALE INVESTIGATIONS
Each soil extraction process basically comprises a step in which
soil and an extracting agent are intensively mixed, a step for separa-
tion of the treated soil and extracting agent, a washing step for
removing rests of extracting agent from the treated soil and a step
for the purification of the polluted extracting agent.
Due to the relatively unknown nature of the bromine contamina-
tions, the different adsorption properties of the various bromine
compounds on the soil and the presence of different types of soil
particles, experimental research is needed in order to select the most
suitable extracting agent. This too is valid for the process steps to
be applied in the process for the purification of the polluted extrac-
tant.
Extraction of the Soil
A laboratory test procedure was set up to test the extraction
behavior of potentially suitable extracting agents. In this test pro-
cedure, 0.5 kg of contaminated soil was added to 0.75 kg of extrac-
ting agent in a flask. Soil and extracting agent were intensively mix-
ed by rotating the bottle for a period of about one hour. Subse-
quently, the mixture of soil and extracting agent was poured into a
processing column which was provided with a porous bottom. The
mixture of soil and extracting agent in the column was washed in
upflow conditions with, successively, non-polluted extracting agent
and water. The superficial liquid velocity in the column was about 3
m/h. In this way only a very small fraction of soil particles was
dragged out with the liquid flow. The soil was analyzed for total
bromine content before and after extracting and washing.
Some of the most important results of the extraction experiments
with soil, originating from the bulk are shown in Table 2. In this
table the effect of the type of extracting agent and the composition
of this agent on the cleaning efficiency are shown. One can see that
the soil can be cleaned, reducing the total bromine content of less
than the detectable limit (0.6 mg/kg) by extraction with hot water
or an aqueous solution of sodium carbonate or soft soap. In the use
of these two extraction agents it appears to be important that the
pH value of the extracting agent remains above 7.
The good extraction properties of these extractants can be ex-
plained by the fact that an aqueous solution of sodium carbonate
or soft soap has a dispersant effect. This probably results in a col-
loidal dissolution of a part of the organic bromine compounds.
Furthermore, it can be expected that humus-like substances will
dissolve, especially as higher pH values. The result is that organic
bromine compounds, preferably adsorbed to these humus-like
substances, go into a colloidal solution. Moreover it can be ex-
pected, especialy at high pH values, that a part of the organic
bromine components hydrolyze into inorganic bromide, which
dissolves very well in aqueous media.
Extraction of soil from the top and sub-top layers was carried
out with aqueous solutions of sodium carbonate or sodium hydrox-
Table 2.
Effect of type of extraction agent on the cleaning
efficiency of bulk soil
Extraction agent
Water
Hot water (70 °C)
Aqueous solution of 0.1 %
soft soap
Aqueous solution of 1 %
soft soap
Aqueous solution of 0.1 %
Na,CO,
pH after
extraction
4.1
4.4
4.4
7.0
7.4
Total bromine content
in soil after extraction
and washing
(mg/kg)
1.5
<0.6
3.6
<0.6
tij hrommc .onteni in umreaiej w>il about 120 mg
A mixed sample of the first two types of treated soil, mentioned
in Table 3, was extracted a second time. For this extraction step, an
aqueous solution of 2% sodium carbonate was used. The result was
a reduction of the total bromine content to about 4 mg/kg.
In addition to the use of water and aqueous solutions of sodium
hydroxide, sodium carbonate or soft soap, which may generally be
considered as acceptable agents from environmental hygienic point
of view, other types of extracting agents were tested, among which
were 1-1-1 trichloroethane and an aqueous solution of sodium
hypochlorite. However no positive results were obtained with these
extracting agents.
Among the agents which were found to be suitable for carrying
out the extraction process, an aqueous solution of sodium hydrox-
ide is considered to be the most promising. This conclusion is based
on the fact that to increase the pH value, especially for strongly
buffered systems, sodium hydroxide is more effective than sodium
carbonate. In the matter of the use of hot water as an extracting
agent, the impression gained from the experiments was that in
treating soil of top layer and sub-top layer, this agent is not as
suitable as an aqueous sodium hydroxide solution.
Purification of the Polluted Extracting Agent
After the extraction process, the polluted extracting agent
(sodium hydroxide solution) assumed the appearance of a dark
brown liquid. In addition to the colloidal and dissolved bromine
components, it contained also other organic compounds, such as
humus-like substances and mineral particles with a diameter less
than about 40 /im.
From laboratory scale centrifuge experiments, it was evident that
no separation was obtained between the bromine compounds and
the clay fraction, not even at relatively low g-values. The/eason for
this is probably due to the fact that in addition to mineral particles,
colloidal bromine containing particles also settle in the centrifuge.
Only normal gravity settling and washing of the settled particles
with water created the possibility of separating a part of the mineral
fraction free from bromine compounds.
The extracting agent, from which the mineral particles having
diameters ^between 40 and about 20 pm were removed by settling,
could be p'urified to a total bromine content of less than the detec-
tion level. The treatment process to achieve this comprised the steps
of coagulation and flocculation at neutral or slightly acid pH value,
sedimentation, adsorption of organic bromine components to ac-
-------
INTERNATIONAL
444
Table 4.
Some Results of the Experiments Using Mixer/Settler Apparatus
and Buchner Funnels
CONTAMINATED
SOIL
MIXER/SETTLER
(CONCRETE MIXER)
BUCKNER FUNNELS
Figure 2.
Setup of Experiments Using Mixer/Settler and Buchner Funnels
tivated carbon and removal of inorganic bromides by means of an
ion exchanger. However, an excess of activated carbon and ion ex-
changer was used in these experiments.
SEMI-TECHNICAL AND PILOT-PLANT
SCALE INVESTIGATIONS
On the basis of the results obtained from the experiments carried
out on a laboratory scale, a further study of the extraction process
was made on a semi-technical scale (30 to 100 kg of soil/h) and a
pilot-plant scale (500 to 1000 kg of soil/h).
In order to limit the amount of the extracting agent as far as
possible, it is necessary to apply a countercurrent flow of extracting
agent and soil. In carrying out the experiments two extraction
methods with countercurrent flow were chosen. The first of these
comprised a mixing/settling step, followed by a washing step of the
settled soil in a stationary layer on a vacuum filter. The washing
step applied here is comparable within a vacuum filtration and
washing process on a sieve belt filter. In the second method a
modified screw conveyor or a modified sand screw classifier were
used for extracting and washing.
In addition to studying the extraction process, treatment of the
polluted extractant was also studied. The latter study was carried
out on a scale of 50 to 100 kg extraction agent per hour.
The experimental set up devised for the semi-technical and pilot-
plant scale investigation was primarily destined to acquire the data
necessary for the pre-design of a treatment plant which can be used
in practice. An aqueous solution of sodium hydroxide was used as
the extracting agent in all of the experiments.
Extraction by Means of a Mixer/Settler
and Vacuum Filtration
The experiments were carried out using contaminated soil which
had been pre-treated in a low speed pen mill, to reduce the large
clods in size. Approximately 30 kg of the pre-treated soil and about
20 kg of extracting agent were intensively mixed for a period of 5
min in a concrete mixer. After settling of the soil the extracting
agent was separated. The soil was subsequently washed in counter-
current flow at four large scale Buchner funnels in order to remove
the remaining extracting liquid and part of the bromine compounds
still present. The extracting and washing process is schematically
shown in Fig. 2.
Water was used to wash the soil at the last Buchner funnel. An
aqueous solution of sodium hydroxide was added to the filtrate of
the second and third Buchner funnel to increase the pH value. The
filtrate of the first Buchner funnel was used as the extracting agent
in the concrete mixer.
Sample data obtained from the extraction experiments carried
out with bulk soil samples are given in Table 4. The final total
bromine concentration in the treated soil is below the detection
limit after two washing steps in the Buchner funnels.
Soil from the top and sub-top layers could only be cleaned to a
sufficient degree if the mixing/settling step in the concrete mixer
was carried out twice. The extracting agent was renewed after the
first of these steps. Furthermore, it was necessary to maintain the
pH.value above 11. The necessity to employ a more intensive ex-
traction and washing procedure is clearly caused by the presence of
Process step
Mixer/settler
1st Buchner funnel
2nd Buchner funnel
3rd Buchner funnel
4th Buchner funnel
Extracting agent
Aqueous solution of
0.46% NaOH
Aqueous solution of
0.33% NaOH
Aqueous solution of
0.10% NaOH
Water
Water
Total bromine
content in treated
soil (mg/kg)
5.8
1.2
<0.6
<0.6
Total bromine content in untreated soil: about 40 mg/kg
relatively large amounts of organic humus-like substances in the
top and sub-top layers of the soil and the higher concentration of
organic bromine compounds.
The results obtained from the mixer/settler and Buchner funnel
experiments generally confirm the results of the laboratory scale ex-
periments.
Extraction in a Modified Screw Conveyor
Extraction and washing was carried out in a modified screw con-
veyor. This unit was comprised of a helicoid flight, mounted on
and driven by a shaft within the confines of a trough. The length of
the screw extractor was approximately 2 m and its diameter was ap-
proximately 0.15 m. Throughout the experiment the speed of the
shaft was always 5 rev/min. In order to obtain countercurrent flow,
the extractor was inclined at an angle of about 30 ° to the horizontal
(Fig. 3).
The soil to be treated was fed to the lower end of the apparatus.
At the upper end, an aqueous solution of sodium hydroxide was
supplied as extracting agent. In the majority of the experiments, the
ratio of the mass flows of soil and water was approximately unity.
Washing the soil with water after the extraction step was carried out
in a manner similar to that used in the extraction process itself.
It was also possible to extract and to wash simultaneously. This
could be achieved by feeding pure water at the top of the screw ex-
tractor while adding a concentrated aqueous sodium hydroxide
solution mid-way along the length of the screw extractor.
Here again, in the majority of the experiments with the screw ex-
tractor, the soil had been priorly treated in a pen mill in order to
reduce the size of large clods. The mass flow of the soil through the
screw extractor was 50 to 100 kg/h.
CONTAMINATED
SOIL.
Figure 3.
Longitudinal Section of the Modified Screw Conveyor
for Extracting Experiments
-------
445
INTERNATIONAL
Experiments with soil taken from the bulk showed that it is possi-
ble to reduce its total bromine content from approximately 50
mg/kg to less than that of the detection level in an extracting step
and a separate washing step. The pH of the extracting agent in
these experiments was approximately 11.
For efficient treatment of soil from top and sub-top layers it was
found necessary to pre-treat the soil in a mixer/settler (in this case
the concrete mixer mentioned previously). Furthermore, it was
found necessary to adjust the pH of the extraction agent to a
minimum value of 11. Doing so it is possible to reduce the total
bromine content of the soil from approximately 50 mg/kg to less
than the detectable level.
One experiment with the screw extractor was carried out with a
mixture of soil originated form the top and sub-top layers, and of
soil originated from the bulk, in a ratio of one to nine. The choice
of this mixture stemmed from the fact, that in the performance of
an actual cleanup process of contaminated soil, treatment of mix-
tures of several types of soil can hardly be avoided. The quoted
ratio of one to nine corresponds roughly to the amount of soil from
top and sub-top layers and the amount of soil from the bulk as
found in the actual situation. The total bromine concentration in
this mixture, that was not treated in the concrete mixer, can be
reduced to a value below that of the detectable level.
Extraction in a Modified Screw Classifier
Evaluation of the mixer/settler-Buchner funnel experiments and
those carried out with the modified screw conveyor showed that for
practical purposes of application, the principle of applying the
screw extractor to be the most attractive one, particularly when
large treatment capacities are required. The latter requirement was
the reason why the experiments on pilot-plant scale were focused
on the use of the screw extraction process.
The pilot-plant experiments were carried out using a screw
classifier, an apparatus commonly used for wet classification of soil
and gravel. For the extraction process, this screw classifier was
adapted in order to ensure an efficient countercurrent flow extrac-
tion process and intensive local mixing of soil and extracting agent.
On adaption, the operating principle of this screw extractor was
similar to that of the 50 to 100 kg/h scale extractor. However, the
shape of trough was slightly different and paddles instead of a
helicoid flight were mounted for more intensively mixing at the
lower end of the extractor.
The mass flow ratio of soil to extracting agent in the experiments
was approximately one. The soil flow through the extractor
amounted to 500 to 1000 kg/h. In all cases, clods of soil present
were reduced in size before extraction took place. Soil washing
after extraction was carried out in the 50 to 100 kg/h scale ex-
tractor.
The experiments with the pilot-plant scale screw extractor were
carried out with soil from the bulk. The average total bromine con-
tent in this soil amounted to about 10 mg/kg. From the results of
several experiments it was observed that the total bromine content
in the soil after extraction and washing was always less than the
detectable limit (0.6 mg/kg). In all cases a pH value of the extrac-
tant above 11 was necessary.
Treatment of the Polluted Extracting Agent
On a scale of 50 to 100 kg per hour, the purification of the
polluted extracting agent (originated from the modified screw
classifier experiments) was investigated. Two treatment processes
were studied. The first comprised the steps of neutralization of the
extracting agent with hydrochloric acid followed by coagulation
with iron chloride and flocculation with polyelectrolyte. After settl-
ing, the sludge so formed was separated and dewatered by cen-
trifuging.
The overflow from the settler was then treated over a deep bed
sand filter, in order to remove small particles, and was thereupon
fed to an activated carbon filter for removal of the organic bromine
compounds. The effluent from the activated carbon filter was
subsequently treated in an anion exchanger in hydroxide or
chloride form, to remove (anionic) bromjde. These process steps
however did not lead to an acceptably low total bromine concentra-
tion level. The reason for this was the high chloride concentration
in the effluent from the activated carbon filter. As a result, the ion
exchange process was not very effective in removing bromide ions.
In order to prevent the occurrence of a high chloride concentra-
tion an alternative purification process for the polluted extracting
agent was investigated. Instead of carrying out the steps of
neutralization with hydrochloric acid followed by coagulation with
iron chloride, a direct coagulation of the polluted extracting agent
was carried out using lime. Insofar as the rest of the purification
process is concerned, it was similar to that already described. In this
manner a total bromine level below the detection limit was obtained
in the effluent from the ion exchanger in the hydroxide form.
The total sludge production per ton of treated soil will amount to
approximately 50 kg (with a dry matter content of 25 %).
PRE-DESIGN OF A FULL SCALE
INSTALLATION FOR ON-SITE TREATMENT
On the basis of experimentally obtained results, it is possible to
design a full scale installation for cleaning up soil reducing its total
bromine content to less than 0.6 mg/kg. The extraction step of the
treatment process can be carried out effectively with use of a screw
extractor. In principle other types of extraction units such as belt
filters, mixer/settlers, hydro-cyclones or combinations of such ap-
paratus may be used. However, from the experiments carried out
so far, most experience had been gained from the use of the screw
extractor. The design of the treatment installation and the technical
and economic evaluation of this installation is therefore essentially
based on the application of screw extractors.
Process Scheme
A flow diagram for the process to treat contaminated soil is given
in Fig. 4. In the process unit, the excavated soil is first passed over a
coarse sieve having a mesh width of 100 mm in order to separate
large objects therefrom. Large clods of soil are reduced in size to
about 20 mm or less by means of a crusher. After crushing, the soil
is transported by a belt conveyor to a wet screening device. A se-
cond crusher is located at the end of the belt conveyor.
In the second process unit, the wet soil is run over a screen having
a mesh width of 10 mm. Any clods of soil present there, will be fur-
ther reduced in size by water sprays. Stones and other objects hav-
ing a maximum dimension of larger than 10 mm are washed and
removed.
Following this process, the wet soil from the screen is fed to a
mixing tank to which sodium hydroxide is added to raise the pH.
Then, the soil is fed from the pH adjustment tank to two identical
screw extractors, the first of which is used for the extraction process
proper, and the second is used to wash the soil as a post-treatment
step. The second screw conveyor is fed with a purified extraction
agent. If necessary, sodium hydroxide is added to obtain the
desired pH. The conditions under which the screw extractors
operate are such that particles larger than approximately 40 /»ra in
diameter will remain in the treated soil.
In the sixth process unit the soil is dewatered on a screen; the
water content is reduced to 12.5% and the hydroxide is neutralized
by the addition of hydrochloric acid.
On leaving the first screw extractor, the extraction agent is fed to
a settling tank where the fine mineral fraction, having a sufficient
settling velocity is separated from the water phase and washed with
an aqueous hydroxide solution in a side stream. The overflow of
the settling tank is coagulated with lime in a static mixer and subse-
quently flocculated by polyelectrolyte. The resulting floes are
separated in a parallel plate settler and additionally dewatered in a
centrifuge. The dewatered sludge must be considered as a hazar-
dous waste material because it contains a large fraction of the
bromine compounds, humus-like substances and very fine mineral
particles.
Further treatment of the overflow from the centrifuge is carried
out in process unit nine. Here, the flow passes a deep bed sand filter
-------
INTERNATIONAL
446
SUPPLY
HATER
(lOm'/h)
CONTAMINATED
SOIL (40 tonn/h)
POLY-
ELECTROLYTE
SAND ACTIVATED ION
FILTER CARBON EXCHANGER
FILTERS
Figure 4.
Process Scheme of the Proposed On-Site Treatment Installation
to remove particles still present. The filtrate is subsequently fed to
two activated carbon filters in sequence. In these filters the organic
bromine compounds are adsorbed. Finally the extraction agent is
treated in an ion exchange column containing a strongly basic ion
exchanger. Chlorides, sulfates and bromides are exchanged for
hydroxide ions. Regeneration of the saturated ion exchanger is ef-
fected with an aqueous sodium hydroxide solution.
The purified extraction agent, in which the total bromine content
is assumed to be less than 0.6 mg/kg, is partly recycled and partly
discharged into the sewage system or into surface water after
neutralization has been effected.
Process Conditions and Capacities of the Apparatus
The main starting points and the conditions under which a treat-
ment plant based on the previously considered process scheme
operates, are as follows:
•The handling capacity of the treatment installation is 40 tons of
soil/h. The water content of the soil is 10%.
•A flow of 40 tons of extracting agent per hour is required for the
extraction process using screw extractors.
•An aqueous solution of 0.2% sodium hydroxide is used as ex-
traction agent.
•The average total bromine content in the soil to be treated is
approximately 100 mg/kg.
•The total amount of contaminated soil to be treated contains
less than 1 % of total organic substances and less than 1 % of
fine mineral particles having diameters less than 40/un.
•It is assumed that the total organic fraction, half of the fine min-
eral fraction and a substantial part of the bromine compounds
are concentrated in the sludge resulting from the coagulation/
flocculation process.
•75• of the purified extracting agent is recycled. The remainder is
discharged.
•The total bromine concentration in the treated soil and purified
extracting agent is less than 0.6 mg/kg.
•The amounts of chemicals needed for the process are given in
Table 5.
•The final water content of the soil after passing the dewatering
screen amounts to 12.5%.
The capacities of the most important of the process units and the
main waste streams of the treatment process can be calculated on
the basis of the foregoingly quoted starting points and process con-
ditions (Table 6). The most important waste streams are:
•Sludge (75 % water) 2000 kg/h
•Regenerate from the ion exchanger 500 kg/h
•Activated carbon 16 kg/h
Operating Costs
The total investment costs of an installation for on-site treatment
of soil contaminated with organic bromine compounds would be
approximately 2.5 million Dutch guilders.*
*$1 U.S. = 2.7 Dutch guilders.
Table 5.
Amounts of Chemicals for Treatment of the
Polluted Extracting Agent
Chemical
Lime
Polyelectrolyte
Activated carbon
Aqueous NaOH-solution (10%
for regeneration of ion ex-
changer)
Amount (per cubic meter of
extracting agent)
2 kg/m3
0.05
0.03
0.01
-------
447
INTERNATIONAL
Table 6.
Capacities of the Most Important of the Process Units
Process unit
Coarse sieve
Fine sieve
Mixing tank
Screw extractors
Dcv.aiering screen
Parallel plate separator
Solid bowl centrifuge
Deep-bed sand filter
Column for activated
carbon
Column for ion exchanger
Capacity
solids
(lons/h)
40
40
40
40
40
liquids
(lons/h)
10
10
40
15
55
15
40
Remarks
100 mm mesh width
10 mm mesh width
volume: 2 m3
mod. screw classifier
scroll type
volume: 3.5 m3
40
40
volume: 9 m3
volume: 3.5 m3
Table 7.
Operational Costs of the Soil Cleanup
Depreciation and interest on capital equipment
Erection and disassembling
Labor
Analyses
Insurance, maintenance
Ion exchanger resin
Activated carbon
Chemicals
Energy
Total costs
Costs per ton of treated soil : 120 Dutch guilders
Dutch
Guilders
*1(P
2800
100
250
100
100
75
50
100
25
3600
The operating costs for treating an amount of 30,000 tons of
contaminated soil on-site, are specified in Table 7. More than 75%
of the costs are attributable to interest and depreciation costs of the
installation. The total operating costs are approximately 120 Dutch
guilders per ton of treated soil, assuming the installation has no
residual value after use.
The costs of excavating the contaminated soil, redepositing the
treated soil on the site, disposal of the resulting waste (activated
carbon, sludge, regenerate) are not included in the costs presented
in Table 7. Neither are the costs relating to overhead, safety
measures and profits. Although these costs are not included, it can'
be concluded from the operating costs of 120 Dutch guilders per
ton of treated soil that on-site treatment of the contaminated soil
by means of an extraction process is an attractive alternative to hav-
ing to transport the total amount of soil to a site destined for the
disposal of hazardous waste materials (not generally available in
the Netherlands) or excavation and thermal treatment of the soili
somewhere else.
CONCLUSIONS
•From experimental investigations carried out on laboratory, semi-
technical and pilot-plant scales it appears that soil, contaminated
with organic bromine compounds having a mean concentration
between 10 and 100 mg/kg, can be cleaned, with a final bromine
concentration of less than 0.6 mg/kg.
•An aqueous solution of 0.2% of sodium hydroxide is suitable as
an extraction agent.
•The extraction of the soil can be carried out efficiently by using
a screw extractor in which soil and extractant are flowing counter-
currently and the soil/extractant mass flow ratio is one.
•Based on the experimental results, a design of a full scale in-
stallation for treating 40 tons of contaminated soil/h has been
produced. The treatment process effected by this installation
comprises the following main process steps:
-extraction of the contaminants in a screw extractor with
aqueous sodium hydroxide, followed by washing of the treated
soil
-treatment of the extracting agent for reuse by subjecting it to
a process of coagulation/flocculation, activated carbon ad-
sorption and ion exchange
-post-treatment of the soil by neutralization with hydrochloric
acid
•The total investment costs of the treatment installation proposed
is estimated to be 2.5 million Dutch guilders.
•The operating costs for treating contaminated soil are estimated
to be about 120 Dutch guilders/ton excluding the costs of ex-
cavation and redeposition of the soil, the costs for disposal of
residual waste materials (about 50 kg/ton treated soil), over-
heads, profits and costs for safety measures.
ACKNOWLEDGEMENTS
The authors wish to thank the Dutch Ministry of Public Health
and Environmental Protection, the Province of Overijssel and the
Municipality of Wierden for the financial support they provided
for carrying out this investigation.
Thanks are also due to the Hollandsche Beton Groep N.V. for
the cooperation in carrying out the pilot-plant experiments and the
evaluation of the results of these experiments.
-------
DEGRADED AND CONTAMINATED LAND REUSE-
COVERING SYSTEMS
GRAHAM D.R. PARRY, Ph.D.
ROBERT M. BELL, Ph.D.
A.K. JONES, Ph.D.
Environmental Advisory Unit
University of Liverpool
Liverpool, United Kingdom
INTRODUCTION
As a result of proposals made by the United Kingdom in 1980,
a NATO/CCMS pilot study on contaminated land has been estab-
lished. A range of study areas and subjects for information ex-
change have been selected which is described elsewhere by M.
Smith in these proceedings. Study Area C in this project is UK lead,
and will examine systems designed to prevent the migration of con-
taminants vertically or laterally or to prevent the ingress of surface
or ground water into contaminated sites. In this paper, the authors
discuss the desirable properties of covering systems and by example
describe their use in the UK.
CONTAMINATED LAND
For the purposes of this report the Pilot Study Group have ac-
cepted the following definition of contaminated land:
"Contaminated land is land where substances are found that if
present in sufficient quantity or concentration, could be haz-
ardous to construction workers, or to the eventual users or
occupiers of the site, or to a wider population due to trans-
port of the substance from the site, for example by wind-
action or pollution of water." The definition also embraces
the presence of substances that may be harmful to plants and
animals and may include substances that are aggressive to
building materials.
The hazards to human health may be short term such as those
presented to workers by the presence of aggressive chemicals such
as acids or of flammable gases, or long term such as might be pre-
sented by the uptake of toxic elements such as cadmium and lead
by food crops grown on contaminated soil or by the presence of
carcinogens in soil.
The contamination of sites arises from their previous uses and
examples of sites that are commonly found to be contaminated are
former coal-gas manufacturing plants, sewage works and chemical
plants. Typically also much former railway land and dockyard
land, and land used for secondary metal recovery (i.e., scrap-yard")
is contaminated.
This definition encompasses hazardous waste "problem" sites or
"uncontrolled" hazardous waste sites as defined by the OECD and
the United States Environmental Protection Agency respectively.
Two important points should be noted about this definition:
•The emphasis on the presence of potentially harmful substance
rather than on past use of the land
•That a "problem" is only defined after site investigation, and
evaluation of data on a site-specific basis which takes into account
land-use
COVERING SYSTEMS
The majority of contaminated land sites are dealt with by isola-
tion or encapsulation which normally includes superimposition of
cover material, which may consist of several layers. In the treat-
ment of such land the covering systems are likely to be required to
perform three main functions:
•To prevent exposure of the population at risk (site workers and
end users) to potentially harmful contaminants
•To sustain vegetation
•To fulfill an engineering role such as accommodating self-im-
posed stress, i.e., uneven settlement, or even supporting buildings
The effectiveness of any covering system based on soil or soil re-
lated materials to fulfill the above criteria will depend on a num-
ber of factors. These include:
•Control of upward and lateral migration of contaminants through
the ground
•The ability of the cover material to immobilize pollutants through
chemical and physical absorption
•Its effectiveness to control water ingress
•The interaction between covering systems, the contaminants and
the biology, e.g., plant root systems
•The engineering behavior of the system and its component ma-
terials
Where contaminated land sites are treated by the use of cover
material rather than by removal of the pollutants, the covering lay-
ers take a variety of forms. The cover may range from simple super-
imposition of a soil layer, or a combination of several layers which
can include an impermeable layer of clay or synthetic material.
The properties of covering materials and the design of cover it-
self to incorporate the range of requirements described above have
received much attention in the United States, principally in the
field of new landfill site management and solid waste dis-
posal1'2. To date there has been little systematic investigation of re-
medial and ameliorative systems which might be appropriate to
contaminated land including waste disposal sites in the UK. How-
ever, a number of research projects are underway funded by the
Department of the Environment and the National Coal Board
which will enable guidance for contaminated land treatment to be
produced similar to that for landfill in the United States.'
CONTAMINATED AND DAMAGED LAND OCCURRENCE
In the UK contaminated land and controlled hazardous waste
sites are viewed as just one part of the more general problem of re-
habilitating land made derelict by past industrial use. Although
there is no statutory definition, a working definition of derelict
land adopted in the UK is "land so damaged by industrial or other
development that it is incapable of beneficial use without treat-
ment." As such, it encompasses land which would be considered to
fall within the definition of contaminated provided above. As
many former industrial sites have been contaminated by "in situ"
waste disposal, and many mineral extraction areas are associated
with waste disposal, it is often difficult to distinguish between
the various categories.3
While there are no statistics available solely for the distribution
and occurrence of contaminated land in the UK, there are figures
available for derelict land. Much of this will be contaminated in
some way for the reasons described above.
The total area of officially recognized derelict land in England
448
-------
449
INTERNATIONAL
approximates to 45,000ha4 of which 75% is said to be eligible for
restoration. Approximately 5% of this land has been used prev-
iously for waste disposal.
Encouragement for the treatment of derelict land in the UK has
been provided by the availability of grants to local authorities
under a number of Acts of Parliament.' In addition, the occurrence
of derelict land in England amounts to 0.35% of the land surface.
Comparative studies for the United States in the 1970s revealed that
0.18% of the land surface could be considered to be derelict.1 The
population density of Britain per unit area is of the order of 15
times greater than that of the United States, so the impact of dam-
aged land is much greater on the former population.
As a result of the above pressures which must include a greater
desire by the public for improved environmental quality, many of
the reclamation schemes for contaminated sites have been carried
out on an ad hoc basis. This has generally been related to avail-
ability of cover material rather than its specific properties. In addi-
tion, covering systems have largely been used to solve one or two
immediate problems for a site rather than the whole treatment pro-
posed by Lutton et al.' The need for rapid decision making and the
method of funding reclamation schemes has not facilitated long
term field trials or the monitoring of reclaimed sites in general
and information is not readily available on their continued suc-
cess. A number of UK sites can, however, be considered to be suc-
cessful, or potentially successful, examples of what can be used to
illustrate the use of a variety of covering systems to resolve a range
of environment problems.
Pare Mine, N. Wales
Former metal mining of the North Wales ore body during the
last century has resulted in a number of abandoned tailings spoil
heaps rich in lead and zinc. At one such mine, Pare Mine in the
Conwy Valley, 260,000 tonnes of tailings spoil had been deposited
on the banks of a tributary of the river Conwy.
Continuous stream and gully erosion of the unstable surface re-
sulted in serious damage to some 6ha of agricultural land in the
flood plain 2km distant from the mine by contamination with lead,
zinc and cadmium.' The contamination was sufficient to inhibit
cereal production and grazing on this land. In addition, shellfish
production was also adversely affected by heavy metals from this
source 20km distant at the estuary mouth.'
The total area of the site covered 2.2ha and its visual disamen-
ity was insignificant in the context of its surroundings. However,
some rapid and permanent surface stabilization was required to
prevent the transported pollutant problem.
Steep tip slopes prohibited the use of conventional soil covering
because of the potential for slippage and renewed erosion. Direct
development of a metal tolerant grass sward on the tailings surface
was also rejected because of the serious consequences of even small
areas of sward failure. A combination of the two systems was even-
tually selected. A covering layer of readily available quarry over-
burden of 100mm and 5 tonne/ha of crushed limestone was spread
over the site and seeded with a grass clover sward containing 60%
Fesiuca rubra Merlin. This is a grass developed for its tolerance to
high soil concentrations of lead and zinc.
Unlike many covering layers which are designed to inhibit root
penetration, the quarry overburden allowed the tolerant grasses to
root into the underlying contaminated material. Thus the surface
amendment was bound to the mine tailings beneath. This, together
with the growth of non-tolerant species on the overburden alone,
has resulted in a stable surface and control of the erosion problem
which is still operating 3 years after its establishment.'
The Beckton Gas Works, London
Former sites of town gas production from coal and the by-
products industry present a major source of severely contaminated
land. These sites are usually associated with dense urban areas in
the UK \shere land is at a premium and redevelopment pressures
are high. One such site is the former Beckton Gas Works in London
where over 100 years of production generated the largest works of
its kind in Europe.
Byproduct waste in this case was disposed of on site during its
working life resulting in a solid waste tip covering an area of 5ha
with an average height of 16m. The intrusive nature of this tip
earned it its local name of the "Beckton Alps". The estimated
430,000 cubic meters of waste was made up of boiler ash, clinker,
iron oxides and lime residues. Associated with these deposits were
potentially dangerous concentrations of cyanides, phenols, sul-
phides and other compounds. Large quantities of solid and liquid
tarry wastes had also been disposed of on this tip.
Much of the surrounding site was suitable for conventional re-
development to industry and housing, but .redevelopment of the
tip itself was restricted. Removal of the waste materials was un-
economic because of handling and subsequent disposal problems
and the tip is scheduled for "in situ" development to open space.'
Unlike the former example, open space development on this site
presents a series of afteruse problems. Firstly, there is the major
difficulty of establishing hard wearing vegetation cover on material
which is extremely inhospitable to plant growth. Secondly, there is
a need to isolate the toxic materials from site users and services.
Provision of a cover layer in this case has required specification of
a multipurpose system using available natural material.
The decision to open the site to public access requires sufficient
thickness of protective layer to prevent toxic materials reaching the
surface either directly or via the biological activity of the plants
which are to cover the final surface. This is to be achieved by a
seal of London clay of 1.2m covered with 300mm of topsoil.
Other requirements of the covering system necessitate exclusion
of percolating water into the tip mass. A 225mm thick gravel drain-
age blanket has been proposed immediately beneath the clay to-
gether with a surface drainage system. A further problem arises
because of the complex nature of the wastes. It is likely that earth-
moving and tip moulding will result in mixing of the wastes, re-
sulting in the potential for gas generation. The permeable drain-
age blanket has been designed to also act as a gas drainage channel
to a strategically located venting system.
. McKechnie's Tip, Widnes, Cheshire
An alternative approach to physical separation of contaminated
material from the final restored surface is provided by the use of
chemical barriers. At the above site over 50 years of copper refin-
ing and associated waste disposal resulted in some lOha of tip seri-
ously contaminated with copper, zinc and cadmium, and signifi-
cant quantities of barium from a paint pigment works. On closure
of the works, the waste tip area was acquired by the local author-
ity for redevelopment to a golf course.
Concentrations of copper in excess of 14000 ppm together with a
variety of other readily available heavy metals resulted in a surface
which was extremely inhospitable to plant growth. The presence of
barium further necessitated the provision of an adequate isolating
or covering system.
Top soil sources and other covering materials are at a premium
in this locality and were not available for use on the scale re-
quired. The site is, however, located in an area which was once the
UK center of the Leblanc process for production of sodium car-
bonate from sodium chloride. The waste product of this industry
provided a readily available source of highly alkaline material with
a pH range of 9-12.
The alkaline waste has been used to provide a 0.5m cover over
the metal rich waste. The high alkalinity and salinity of the cover
material in itself will inhibit plant growth, but will reduce the solu-
bility of the metals in the original waste and reduce their potential
for movement. The revegetation requirements for the golf course
have been overcome by provision of a further layer of 150mm of
topsoil. Site restoration has begun and there is no evidence of
sward regression in those areas which have been established during
the past 12 months.
-------
INTERNATIONAL
450
SUMMARY
Treatment of contaminated land and its recognition as an en-
vironmental problem will differ between individual countries. The
United Kingdom, for example, places emphasis on the restoration
of contaminated land so that it can be put to some beneficial use
such as housing, public open space, agriculture or industry. This
emphasis is the result of pressures created by the need to regen-
erate land in urban areas made available by declining industries,
and to conserve the stock of agricultural land. At the other extreme
is the concern in the United States for remedial action often by iso-
lation on uncontrolled hazardous waste sites which can cause ser-
ious environmental damage. Between these two extremes other
countries have experienced a range of problems.
While there has been little systematic investigation of ameliora-
tive measures including covering systems until recently in the UK, a
number of reclamation schemes provide examples of the require-
ments of covering systems. The guidance produced in the United
States for cover for landfill sites is an indication of what can be
achieved once the basic principles and requirements are under-
stood. The NATO/CCMS study group under Project C intends to
draw together the major existing sources of information, deter-
mine the variety of covering system requirements, assess their per-
formance by reference to example and to highlight the areas of
uncertainty and research needs.
REFERENCES
1. Lutton, R.J., Regan, G.L. and Jones, L.W. "Design and construc-
tion of covers for solid waste landfill", USEPA Report 600/2-79-165.
1979.
2. Fung, R., ed., "Protective Barriers for containment of toxic ma-
terials", Noyes Data Corporation 1980, 189-268.
3. Wilson, C.D., Smith, E.T., Pearce, K.W., "Uncontrolled hazardous
waste sites: a perspective of the problem in the UK. Chemistry and
Industry, Jan. 1981, 18-23.
4. Department of the Environment "Survey of derelict and despoiled
land in England" published biennially.
5. Department of the Environment "Derelict Land" circular 17/77 Lon-
don HMSO.
6. Johnson, M.S., Eaton, J.W., "Environmental contamination through
residual trace metal dispersal from a derelict lead-zinc mine", J.
Environ. Qua/. 9 1980, 175-9.
7. Elderfield, H., Thornton, I. & Webb, J.S., "Heavy metals and oyster
culture in Wales", Mar. Pol/ut. Bull., 2. 1971, 44-7.
8. Firth, J., Johnson, M.S., Richards, I.G., "The reclamation of lead
mine tailings at Pare Mine, N. Wales", In Trace substances in En-
vironmental Health XV University of Missouri. Ed. Hemphill, D.
333-339.
9. Netherton, D.W., Tollin, B.I., "Reclamation of the Beckton Alps and
Adjacent Areas", Proc. Sym. on Development of Contaminated Land,
Imperial College of Science and Technology, 1982.
-------
IN SITU TREATMENT OF UNCONTROLLED
HAZARDOUS WASTE SITES
J. BRUCE TRUETT
RICHARD L. HOLBERGER
The MITRE Corporation
McLean, Virginia
DONALD E. SANNING
Solid and Hazardous Waste Research Division, Municipal Environmental Research Laboratory
United States Environmental Protection Agency
Cincinnati, Ohio
INTRODUCTION
Problems associated with contaminated lands—lands used
directly as waste disposal sites or contaminated as a result of in-
dustrial or other activities—are common throughout the NATO
alliance and many other industrial nations. These problems range
from imminent health and environmental hazards posed by con-
taminants, to the increasing pressures to reclaim the land for safe,
beneficial use.
The North Atlantic Treaty Organization's (NATO) Committee
on the Challenges of Modern Society (CCMS) is studying in situ
treatment of contaminated lands. Portions of the study are being
conducted in several nations. The Netherlands are evaluating in
situ treatment applications in gas works and other chemically con-
taminated sites.' The United Kingdom has completed a grouting
feasibility study and is evaluating an experimental study of shallow-
depth grouting problems.2 The Federal Republic of Germany has
treated arsenic-contaminated soils and groundwater by injection of
permanganate solution, and has investigated the effects on ground-
water of silicate-gel injections used to consolidate foundation
soils.3'4'5 The United States has recently completed a study on in
situ techniques of solidification/stabilization using one actual site
as a scenario for evaluating the feasibility of such treatment.' These
studies represent a substantial effort in the evaluation and appli-
cation of in situ treatment techniques.
In the report of the second meeting of the Study Group for the
NATO/CCMS Pilot Study on Contaminated Land,7 the term "in
situ treatment" is defined as "methods of treating, without ex-
cavation, the bulk material on a contaminated site by detoxify-
ing, neutralizing, degrading, immobilizing or otherwise rendering
harmless contaminants where they are found." The aforemention-
ed studies conducted in the Netherlands, the United Kingdom, the
Federal Republic of Germany, and the United States all address
some treatment techniques that fall outside this definition as well
as techniques that conform to the definition. In this paper the
author does not dwell on the "almost but not quite in situ" tech-
niques, but focusses on the strictly in situ, while referring the read-
er to the published study reports for a more complete description
of the scope and content of the respective studies.
Certain of the in situ treatment techniques serve not only to re-
duce the pollution impacts from contaminated soils, but also to im-
prove the properties of the soils for certain end uses of the site.
Selected grouts, for example, not only immobilize specific pollu-
tants, but can increase the load-bearing properties of the soil in a
contaminated area intended for eventual use as a building site.
The same treatment techniques may also adversely affect other end
uses, as discussed later in this paper.
THE NETHERLANDS ACTIVITIES
In response to a directive from the Minister of Public Health and
the Environment, the Laboratorium voor Grondmechanica at
Delft developed a general inventory of "soil reconstruction" tech-
niques. The inventory includes techniques in use or under devel-
opment in several countries, but emphasizes their applicability (or
lack thereof) to situations in the Netherlands. The types or cate-
gories of techniques developed in this study are:
•Civil engineering techniques to restrict the spread of pollution,
•Physical and/or chemical techniques to restrict the spread of
pollution (immobilization),
•In situ cleaning techniques,
•Excavation,
•Removal of polluted soil,
•Mobile installations for the cleaning of the soil, and
•Water purification.
Only the first three of these categories involve in situ techniques.
Specific techniques encompassed by the first category are iden-
tified in Table 1, and those in the second and third categories
are listed in Table 2. The inventory presents a brief description of
the specific techniques, and includes diagrams of some of the more
prominent, such as jet grouting (Fig. 1) and infiltration accom-
panied by leachate pumping (Fig. 2). The report also presents an
Groundwater
Level
Longitudinal Section
Impermeable
Stratum
(^.••: ...oe^
8-1 Ocm *
08
-150m
Cross Section
Figure 1.
Infiltration in Combination with Groundwater Pumping.
interesting summary evaluation for specific techniques in most of
the above listed categories. An example of the summary format,
451
-------
INTERNATIONAL
452
here applied to the vertical screening of the category, is shown in
Table 3.
In this study, the authors present a total of 22 conclusions and
recommendations of which 10 that pertain to in situ techniques are
summarized as follows:
•There are as yet no standards to be met by the reconstruction tech-
niques and the materials used in the reconstruction. Issuance of
such requirements in the near term is recommended.
•Civil engineering techniques for the restriction of contamination
from large sites (screens, etc.) are considerably cheaper than
complete excavation. The opposite is the case for smalter sites.
•In situ cleaning techniques are attractive (when applicable) be-
cause large quantities of soil can be treated at once and because
the waste processing problem is of a fairly limited size. Further
development of these techniques is recommended.
•In situ cleaning techniques, in particular biological and chemical
methods, will in many cases have to be used in combination with
screening vertical barriers in order to limit the risk of pollutant
migration.
•Development of less expensive techniques for vertical screening
is recommended.
•The use of immobilization to restrict the spreading of contamina-
tion at large sites is much more expensive than the application
of civil techniques. Little is known so far about the effectiveness
and durability of the immobilization techniques.
•Soil contamination conditions vary so greatly that it is necessary
to determine which technique should be applied on a site-by-site
basis. As a rule, the combination of techniques may be required.
•The selection of techniques is determined in part by the desired
end use of the site. If construction is to take place, then clean-
ing of the soil will probably have to be undertaken on short notice.
If construction is not involved, then techniques which limit only
the spread of the contamination may be selected.
•In order to minimize the number of required measurements, it
is necessary to find the most sensitive parameters (parameter
sensitivity analysis).
•The continued development of computer simulation models of
contaminant spreading and removal for the characteristic con-
taminants and characteristic soils is desirable. By means of simu-
Table 1.
Soil Reconstruction Techniques in Netherlands Inventory—
Civil Engineering Techniques
•Vertical Screening*
-thin membranes (plastic foil, bituminous)
-thin membranes in combination with draining or channeling machines
-open cuts
-slurry trenches
-thin screen walls (bentonite/cement)
-steel dam walls (sheet piling)
-timber or concrete walls
-cutting piles
-jet grouting
-vertical injections (pressure grouting)
•Well Point Dewatering
•Coverage Against Precipitation
-bituminous membrane or synthetic foil
-clay cover
-bentonite layer
•Horizontal Screening* under Contaminated Soil
•In the context of the Netherlands study, the term "Screen" denotes a physical barrier and "Screen-
ing" denotes the emplacement or installation of such a barrier.
lation it will then be possible to optimize reconstruction tech-
niques for cleaning effect, costs, etc., and to compare alterna-
tive techniques.
UNITED KINGDOM ACTIVITIES
Under a commission from the United Kingdom Department of
Environment, the firm of Atkins Research and Development per-
formed a thorough evaluation of the technical and economic feas-
ibility of a wide range of techniques for treating contaminated
land. The appraisal covered in situ, on-site, and off-site systems.
Emphasis was accorded to the in situ techniques such as grout-
ing and ground injection of chemicals, which offer the potential
for improving the engineering properties of the ground while aid-
ing the control of pollution. Estimates costs of applying the var-
ious techniques are presented, and methods of measuring their ef-
fectiveness are discussed. Specific techniques evaluated in the At-
Source:
a) Infiltration via Wells
Reference 1
b) Infiltration via
Surface of Contaminated Area
Figure 2.
Installation of a Vertical Screen by Jet Grouting
-------
453
INTERNATIONAL
Table 2.
Additional In Situ Techniques in Netherlands Inventory
Physical/Chemical Techniques to Restrict the Spread of Pollution
(Immobilization)
•Applications
-injected as vertical or horizontal screen
-to harden a site
-cover layer on top of contamination
-to restrict above-ground leakage
-to make contaminated sludge layers firm
-to restrict leaching
•Types of Immobilization Agents
-cement-based
-lime-based
-oil absorbents
-silicate-based
-urea formaldehyde
Techniques for In Situ Cleaning of Soils
•Removal of Groundwater by Pumping
•Groundwater Pumping in Combination with Infiltration of Water
(Flushing) via
-wells
-surface of contaminated area
•Groundwater Pumping in Combination with Infiltration of Chemical
Solutions
•Biological Cleaning
•Thermal Treatment
Table 4.
Principal Characteristics and Specific
Techniques of United Kingdom Study
Emphases
•Emphasizes Grouting and Ground Injection
•Benefits of These Techniques
-the ground and ground levels need not be disturbed;
-contamination is in limited contact with those treating it;
-ground conditions (e.g., stability) can be improved at the same time;
-treatment can be localized;
-treatment can be applied at considerable depth;
-treatment can (within limits) be applied after development.
Treatment Techniques Evaluated
•In Situ Techniques
-ground injection techniques—general
-grouting for laterial containment
-microencapsulation by shallow grouting
-grouting in landfill
-ground leaching, chemical stabilization/detoxification
-electrochemical processes
-cutoff techniques
-stabilization (mechanical properties of soils)
•Other On-Site Techniques
-deep or other ploughing
-soil covers, additives, cappings and membranes
-chemical fixation and encapsulation
•Off-Site Treatment and Disposal
-disposal to landfill
-chemical fixation and replacement
kins study are listed in Table 4. The following paragraphs des-
cribe a few of the most interesting and promising of the in situ
techniques.
The Atkins study distinguishes between the term "grouting",
and "ground injection". Here, "...'grouting' is used solely in the
'normal' engineering sense, namely, the injection of appropriate
materials—usually a viscous fluid under pressure—into pores and
cracks of another material so as to decrease permeability of com-
pressability, or increase strength, or a combination of the two. The
term 'ground injection' is used to encompass injection of fluids
Key:
+ - yea
o = sometimes
not
? ™ unknown
Source: Reference 1
Membrane constr. In development
Draining machine
Open cut
Slurry trench
- deep wall screens
- screen walls
- steel dam wall (heavy vibr.)
Steel dam wall (light, Injected)
Cutting piles
Jet-grouting (being developed)
Inject luiib
rH
•H
o
n
•o
M
&
•O
01
X
•rl
4J
*4-4
O
CO
e
e
o
•H
4-1
CO
U
•H
rH
O.
p
•*
+
+
-H-
+
+
o
+
0
o
o
Table 3.
Summary Characteristics of Vertical Screening Techniques
•H
•a
§ :
•HO 01 C <~-
•-1 -H rH O rS
O i-1 J3 -H CO
CA d 3 01 a 4-1 -o
rl 3 0 G -H C 3
oi x -H o -H X -^
c
10
4J
CO
•H
CO
0)
rl
^
rH
rH
0)
o
•H
s
01
.c
cj
+
+
-H-
+
+
+
-
-
-t
O
O
IO
c
cd
rl
01
00
c
o
rH
01
14-1
t-4
rJ
?
+
+
+
+
+
0
-
+
+
CO
•rl
rH
01
as
+
+
+
+
+
+
+
+
o
o
o
4J
01
>
•rl
vH
CO
p
01
en
01
rH
4J
4J
•H
rJ
+
+
+
+
+
-
•*•
O
0
_
u
01
X
01
rl
MH
c
o
•H
4J
0
01
a
CA
c
M
o
+
+
+
+
o
o
0
0
o
o
U)
**~t
•a
co
r>.
•H
CO
CO
01
3
•H
CO
a
M
o
+
+
+
+
+
+
+
+
+
01
T-l
ft
O
CO
0)
01
rH
CA
01
rH
o
10
4J
CO
ft
°
-
-
+
+
+
-
0
o
+
+
(U
c
Tl
01
u
CO
•H
3
01
rH
u
u
•H
rJ
+
+
—
-
-
0
•*•
+
-
0
o
u
a)
CO
B
a>
3
rH
10
>
T
zero
zero
io-7
io"8
10~8
10~8
io"7
io-7
t^
u
•rt
U
a)
(X
0
a
0
•H
O
3
•O
O
d.
400
high
75
125
125
500
200
250
100
200
125
ilders/n2)
3
00
p
o
•rl
a
u
•H
•a
a
01
u
•rl
Wi
PLI
100
30
75
100
150
50
200
65
350
75
350
£
f.
4-1
a.
01
•o
0)
rH
ft
CO
U
rH
a
<
15-20
5
15
50
20
15
30
20
50
rH
^
a)
01
4-1
O
u
CO
01
e
3
o
r!
H
10"3
10~3
1-2
0.6-1.5
0.6-1.5
0.1-0.2
10~2
2.10"3
0.6-1.5
0.1-0.2
1-2
ractors capable of
ruction (both
truction according
r> to £
0 C 0
U O u
0
O CO (-
3 * *'>•
Q 4J (0 U
rl O
44 00 O ^
O C 4J C
•rl u 01
r< *J CO r*
0) 3 rl C
.Q O 4J -rl
3 X O O
Z 01 0 4J
a few
very many
very many
a few
many
many
very many
many
sporadically
sporadically
a few
-------
INTERNATIONAL
454
also where the treatment objectives do not necessarily coincide with
the normal engineering objectives just mentioned: for example,
the use of chemical neutralization agents or forced leaching. The
techniques and equipment are identical to those employed in grout-
ing for engineering purposes...It is treatment objectives, and hence,
treatment agents, that distinguish them."
Grouting in Contaminated Soils
When grouting is applied to treat contaminated soils, it is often
necessary to inject at shallow depths (less than two meters), since
the contaminated volume is generally in the soil mass immediately
below the surface. One of the basic characteristics of grouting is
the need for adequate overburden pressure to ensure that the grout
does not escape to the surface through fissures before it permeates
the target areas. To overcome this problem, two solutions are
suggested:
•Applying a temporary overburden, such as a suitable concrete
block (aided on the surface and moved about, as necessary, by
crane).
•The use of electro-osmosis to improve the receptivity of the soil to
grouting, or as part of the grouting process. It has been shown
that the presence of an electric field reduces the grouting pres-
sures necessary and induces movement of the grout. The electric
energy requirements are not well established but appear modest—
in the range of 3 to 25 kwh/yd3 of material.
Grouting in Landfill
Problems arise from the heterogeneous nature of many landfills
and the on-going chemical and biological activity that may be oc-
curring within the fill. The potential for leachate production and
gas generation must be taken into account.
The principal potential applications of grouting in landfills are
considered to be:
•Control of groundwater entry
•Control of leachate egress
•Extinguishing fires in the landfill
•Inducing anaerobic conditions (thus reducing potential for fires)
•Maintaining top temperatures (thereby improving bacterial activ-
ity—this can also be achieved by capping)
•Control of gas migration
•Improvement in ground stability
The principal disadvantages of grouting are:
•Application may be difficult and uncertain depending on com-
position of the fill material
•It may adversely influence gas generation and leachate quality
•Ultimate stabilization of the fill may be delayed
A related observation, made in the context of a different appli-
cation of grouting, is that when the end use of the site is for con-
struction, the presence of hard, cementitious grout in the soil may
interfere with the installation of foundations and underground
waterpipes, electric lines, and other utility conduits; also, the pres-
ence of certain chemical grouting material may corrode such con-
duits. Advantages and disadvantages of grouting are summarized
in Table 5; most of these apply to ground injection as well.
Ground Leaching; Chemical Stabilization/Detoxification
This approach involves application of a liquid to a contaminated
soil mass, followed by collection of the liquid for disposal or re-
use. The liquid may be water to leach out soil contaminants, or
may contain chemical or biological agents that would react with
contaminants to form innocuous products. This technique is being
used in Sweden to reclaim a former herbicide factory site over a
5 to 6 year period. Here, water is applied by means of perforated
pipe laid in a ditch through the contaminated area. The water
permeates the affected soil, is collected by drainage pipes, then
treated by activated carbon.
Other methods for applying the water or treatment liquid in-
clude: (1) direct application to the surface, and (2) underground
Table 5.
Advantages and Disadvantages of Grouting/Ground Injection
Advantages
•Versatile; useful for wide range of objectives
•Can be used for injecting any fluids
•Can be used to reach specific targets
•Well established, relatively straightforward technique
•Contaminated soil is treated in situ
•Can be used at considerable depth
•Can be used to protect services
•Can be used in tight situations
•Unlikely to affect adjoining areas
•Unlikely to be limited by treatment volumes or site area
•Wide range of fillers and chemical grouts readily available
•Applications can be staged (preventing overdesign)
•Equipment not overly cumbersome or heavy
Disadvantages
•Generally very expensive
•Cannot usually be applied at shallow depth (normally 1.5m)
•Permeation not guaranteed; may require repeated application
•Soil and contamination regimes are likely to be heterogeneous and com-
plex, requires matching of application strengths and viscosities to con-
tamination and geotechnical conditions
•Applications (of grout) can interfere with subsequent groundworks
(e.g., excavations for sewers and other services)
•Durability of grouts not proven in potentially aggressive underground en-
vironments
•Robustness of application has to be proven
•Chemical grouts, solvents can present some operator hazards
•Applications may create supplementary contamination
Table 6.
United Kingdom Activities
Recommendations
•Test forced leaching techniques on sites with 'soluble' contaminants
•Test temporary surcharge and electro-kinetic methods for overcoming
shallow ground injection problems
•Study feasibility of choosing treatment techniques for a specific contam-
inated site so as to develop optimum solutions for different end-uses
•Determine the practical potential of a) dispelling oil by fluid injection
and/or b) the use of filler grouts for stabilization on unstable oil-con-
taminated ground
•Examine the ability of jet-grouting to form impermeable barriers at
relatively shallow depths in various soil types
•Examine the effects of filler grouts in landfills on engineering proper-
ties, leachate control, groundwater influx, and biodegradation
•Determine technical and economic potential of using on-site chemical
fixation plants for treatment of contaminated soil
•Develop procedures for forming effective grout seals around load bear-
ing piles driven through impermeable strata underlying contaminated
materials
pressure injection by use of grouting equipment. The latter is of
greatest interest in this study because it offers the potential for
faster, more localized application, and can accommodate gaseous
treatment fluids (e.g., air or oxygen for aeration) as well as liquids.
In this approach, as with many grouting applications, injection at
shallow depths may be a problem because of inadequate over-
burden pressure. The report outlines a test employing pressure in-
jection of liquid to suitably sized sections (or "cells") of a con-
taminated area, with suitably located drainage lines and a leachate
treatment system.
Recommendations
The Atkins report gives eight recommendations that apply to a
range of land treatment techniques (Table 6). Two of these apply
directly to the in situ treatment topics discussed above.
•Perform Tests of Ground-Injection Forced Leaching
These ground injection forced leaching tests could be extremely
valuable particularly where large volumes of material are con-
cerned and where the source contaminant is soluble.
-------
455
INTERNATIONAL
TaWe7.
In-Situ Treatment of Arsenic Contaminated Ground Water hi the
Federal Republic of Germany
Arsenic Content in Ground water Near Smelter Site
1971 1975 1977 1979
pH (range) 3.1-7.0 4.8-7.0 5.5-7.8 5.8-8.2
EhmV — -110-+20 -110- + 440 -120- + 440
As mg/1 0.01-56 0.01-26 0.01-0.3 0.01-5.6
As average
mg/1 22.7 13.6 0.06 0.4
FEJ+mg/l 0.2-140 0.1-93.3 — —
SO1-mg/1 152-2010 £0-1670 — —
Spec, electrical
conductance
S/cm — 440-2300 600-2250 650-2150
The values vary according lo the site and depth of the observation weils.
Source: Reference 3
•Develop Solutions to Shallow Ground-Injection Problems
Since most contamination is at relatively shallow depths,
normal ground injection procedures must be amended to meet
the essential requirement of overburden pressure. This can be in
the form of a temporary surcharge (such as superimposed ma-
terial or a concrete block) and/or the use of electro-kinetic aids.
On-site testing should be conducted to test and develop these
techniques.
FEDERAL REPUBLIC OF GERMANY ACTIVITIES
The reports describing in situ treatment activities in Germany dis-
cuss the application of specific techniques to specific problems,
and the results obtained.
Reference 3, "In Situ Treatment of Arsenic Contaminated
Groundwater," reports the use of an oxidizing agent, potassium
permanganate, to reduce arsenic concentration in groundwater in
the vicinity of a zinc ore smelter near Cologne. The smelter oper-
ated from 1913 to 1971. In the latter year, arsenic concentra-
tions of as high as 56 mg/1 were detected in groundwater at 20m
depth. The natural background concentration is less than 0.01
mg/1. By 1975, the concentration at the same location had dropped
to about 26 mg/1.
Extensive monitoring was conducted at wells and piezometers
installed within a roughly oval area about 300m by 450m on the
axes. Monitoring data revealed that the arsenic was in trivalent
form in the regions of higher concentration (greater than 1 mg/1),
and in pentavalent form in regions of low concentration (less
than 0.1 mg/1). These results, together with observations of arsenic
compounds precipitated on soil samples, suggested that most of the
dissolved arsenic was present in the trivalent state and that trans-
formation into pentavalent species in the presence of calcium and
ferrous ions would cause an appreciable fraction of the arsenic to
precipitate.
From December 1976 to May 1977, a solution containing 29 kg
potassium permanganate was injected into 17 wells and piezo-
meters. As shown in Table 7, the average arsenic concentra-
tion in groundwater samples decreased from 13.6 mg/1 in 1975 to
0.06 mg/1 in 1977. However, this average value had increased to
0.4 mg/1 by 1979.
Reference 4, "Groundwater Impact on Silicate Gel Injections,"
describes an investigation of the groundwater impacts of silicate
gel injection widely used for chemical soil consolidation. The im-
pacts are reported in quantitative form in the referenced docu-
ment, and are summarized briefly as follows:
•Alkalinity increases as a result of the alkaline components of the
water glass (an aqueous solution of sodium silicates—Na,Sio
and Na2Sio4) or alkaline and alkaline-earth precipitants.
•Organic content increases when organic precipitants are used,
causing O, consumption and reduction of O,-containing com-
pounds, resulting in the temporary occurrence of a strongly re-
ducing environment.
•Heavy metal content increases because of impurities of the water
glass (an aqueous solution of sodium silicates—NajSio and
N^Sio,) or dissolution from the sediment by the action of COr
•Heavy metals precipitate as sulfides in a reducing environment
only.
Reference 5, "Protection and Improvement of Groundwater
Quality by Oxidation Processes in the Aquifer," describes the in-
jection of oxygen-enriched water to improve groundwater quality
and protect against pollution. The suggested mechanism is that
concentration of certain ionic species in groundwater varies with
oxidation and reduction reactions. Biological decomposition of
organic matter in soil and groundwater consumes oxygen and
decreases redox potential. The redox potential can be increased
either by adding oxidizing agents (usually atmosphere oxygen) to
the soil, or by injecting oxygenated water into the groundwater. In
addition to this basically inorganic mechanism, the increased level
of oxygen in soil and groundwater can also promote biological
reactions beneficial to water quality.
UNITED STATES ACTIVITIES
The principal purpose of the U.S. study was to investigate the
feasibility of solidifying or stabilizing hazardous industrial wastes
that had been placed in the ground. In the context of this study,
the term in situ treatment of waste materials means that the treat-
ment is applied while the waste remains in the ground.
The USEPA has been involved in research, development, and
demonstration of methods for the proper management of haz-
ardous wastes. USEPA's Municipal Environmental Research Lab-
oratory (MERL) has sponsored much of the research and devel-
opment on methods for managing solid and hazardous wastes, in-
cluding the study now being reported. The work was performed
under contract by The MITRE Corporation.
Methods of Solidification/Stabilization for In Situ Application
In USEPA's "Guide to the Disposal of Chemically Stabilized
and Solidified Wastes," the term "solidification" implies that the
product of treatment will be in solid form (such as pellets, grains,
blocks, large masses of undefined shape) while "stabilization"
implies that the hazardous components of the waste will be ren-
dered insoluble or otherwise immobile or that their hazardous char-
acteristics (e.g., toxicity) will be neutralized.
In this study, treatment techniques to solidify or stabilize haz-
ardous wastes were identified through examination of the literature
and communication with vendors and researchers. The principal
categories of techniques are shown in Table 8. Technical descrip-
tions of the specific techniques in each category were compiled to-
gether with their associated problems, limitations, and a listing of
advantageous or disadvantageous characteristics. An example of
the latter is shown in Table 8 for one of the major categories (meth-
ods involving crystalline matrices).
Each of the techniques was then assessed in terms of its applic-
ability for in situ treatment of landfilled wastes. In this study, the
primary intent of treatment is the control of environmental pollu-
tion and reduction of hazards from the wastes; little attention was
accorded the engineering properties of the site or treated wastes.
The assessment quickly led to three conclusions:
•The preponderance of research and practical experience with
most of the techniques had involved non-in situ treatment,
either before wastes had been placed in landfills or after they
had been evaluated.
•Most of the listed techniques require thorough mixing of a solid-
ifying or reactive additive with the waste, and the required de-
gree of mitring is generally not achievable during in situ applica-
tions.
•No single technique or combination of techniques is widely applic-
able for most landfills.
-------
INTERNATIONAL
456
There are, however, a few exceptions to the second of these con-
clusions, and these exceptions represent the techniques of greatest
interest in the study. They are:
•Injection or surface application of chemical agents (in combina-
tion with leachate collectio'n and removal)
•Thermal fusion/vitrification
•Macroisolation (that.is, surface capping, bottom sealing, ver-
tical barriers, and other structural procedures)
As one basis for assessing the applicability of alternative tech-
niques, a scenario was established in which the techniques were
conceptually applied. This scenario involved a site at which a high-
ly diverse combination of wastes—demolition debris, toxic inor-
ganic and organic chemicals, lesser qualities of sanitary wastes and
municipal solid wastes—had been placed in an abandoned quarry
over a period of some 20 years. The total volume of wastes de-
posited at the site was about a quarter-million cubic meters, of
which about 10 percent was arsenical material. The bottom of the
quarry, and hence the waste pile, lay below the water table in some
areas, so there was groundwater flow through the wastes. The
filled area covered about three hectares and was about 10 m thick
at its maximum depth. Portions of the fill rested on highly frac-
tured bedrock, which overlies an important aquifer system. The
site was in close proximity to a river and to populated areas. An
approximate cross-section is shown in Figure 3. The objective in
this study was to investigate whether it is feasible to achieve a high
degree of pollution control at the scenario site by means of tech-
niques applied in situ.
The majority of techniques listed in Table 8 were excluded be-
cause they require more thorough mixing than can be practically
achieved in situ, and because of possible interference by waste con-
stituents with the setting actions of cements, polymers, and gels.
Thus, the list was narrowed to three conceptually feasible options:
•Injection of water or reactive chemicals (with leachate removal)
•Thermal fusion/vitrification
•Macro-isolation
Their application to the scenario site was reviewed in detail. This
review took careful account of the highly heterogeneous char-
acteristics of the site with respect to its chemical, physical, and
hydrogeologic characteristics.
Leaching with injected water did not appear to be feasible be-
cause of the insoluble nature of toxic, inorganic constituents
known to be present in large quantity.
Injection of reactive chemical fluids was not considered attrac-
tive because of the variety of chemical pollutants in the waste.
Addition of a particular chemical to attack one type of hazardous
constituent might cause antagonistic or counterproductive reac-
tions with other constituents. For example, an oxidizing agent in-
tended to destroy a specific organic compound might also change
the valence state of a metallic ion, increasing the toxicity or mo-
bility. A secondary reason for rejecting this approach is the not in-
significant problem of collecting the leachate in the specified
hydrogeologic setting, without serious risk of contaminants' en-
tering nearby aquifer or surface waters.
f
USA INVESTIGATION
Precipitation
NW
II
Sic
1010
1000
990
„ 980
970
960
950
940
930
920
910
900
i
Wells and Test Boring Sites
ii ii
SE
River
(Depth Estimated)
973.44
EZD
CZJ
CZI
9
r
Fill
Alluvium
Till
Limestone
Dolostone
Shale
Groundwater Monitoring Point
Top of Cored Material
Shown Only in
Deep Wells
Feet
3T10
Meters 2\ Feet
1 \ 30 100
' i'T /
10 2030
Meters
Heavy arrows indicat?
conceptual water flow
paths
Source: Reference 6
Fig. 3.
Cross Section of Scenario Site
-------
457 INTERNATIONAL
Table 8.
Categories of Treatment Techniques in USA Investigation.
•Solidification
-crystalline matrix (cement based) methods
-lime/siliceous matrix (pozzalianic based) methods
-thermoplastic methods (including bitumens)
-organic polymer methods
-gelation
-thermal fusion/vitrification
-surface encapsulation
-microencapsulation
-macroisolation
•Stabilization without Solidification
-injection/surface application of chemically reactive agents
Example of Comparative Advantages of
Crystalline Matrix Methods
Advantages
High Load-Bearing Solid
Inexpensive Raw Materials
Technology Well Known
Non-Specialized Labor
Dewatering Not Required
Tolerant of Chemical Variety
Product Can Be Sealed by Coating
Disadvantages
Pretreatment of Certain Wastes Required
May Release Ammonia from Wastes
Energy Intensive
Final Product Heavy, Bulky
Uncoated Product May Require Special
Landfill
Organic May Interfere With Setting
The thermal fusion/vitrification technique appeared technically
feasible at first inspection because it accommodates to a great
variety of waste material and produces an inert, strong solid mass
of very low solubility. However, there are some practical limita-
tions to its use. The technique employs electric energy at a rate
sufficient to heat a mass of buried wastes to temperatures above
the fusion point of surrounding soils and rocks. The electric en-
ergy is applied through electrodes inserted in the landfill on either
side of the wastes (or portions thereof) to be fused. The elec-
trodes are placed in the ground or fill by drilling or other appro-
priate means, and a strip of graphite in contact with the fill ma-
terial is connected across the electrodes to act as a "starter" in
melting the fill. A cover is placed over the surface of that portion
of the fill which will be fused at a given placement of the elec-
trodes. The cover is intended to capture gases released during the
fusion. Captured gases are ducted to a treatment unit as neces-
sary.
While the "vitrification by electrification" approach has been
successfully demonstrated on masses of soil to the order of ten
tons, the effective volume of any given application, using cur-
rently available equipment, is a cube six meters on the side. While
the technique may be readily adaptable to wastes buried no more
than 6m below the surface, its application to a relatively deep land-
fill as in the scenario site would appear to require considerable
excavation, and thereby violates our definition of in situ treat-
ment. An ancillary reason for considering this technique infeas-
ible is its substantially high cost relative to the cost of excavation
and non-insitu treatment.
The remaining in situ option—macro-isolation—was considered
to be feasible from a civil engineering standpoint, but environ-
mentally undesirable (although not necessarily unsound) because of
the massive amounts of contaminated soil and rock that would
have to be excavated, whether by mechanical or hydraulic means,
and the potential pollutant-release that could result.
Thus, the study concludes that strictly in situ techniques are
not a feasible approach to controlling pollution at the scenario
site. However, it appears probable that they could be effectively
employed in conjunction with non-in situ control measures.
REFERENCES
1. "Inventory of Soil Reconstruction Techniques," Prepared by the Lab-
oratorium voor Grondmechanica, Delft, for the Ministry of Public
Health and the Environment, Report Number BO-2. Staatsuitgeverig,
The Hague, The Netherlands, Oct. 1981.
2. Barry, D.L., "Treatment Options for Contaminated Land." Prepared
by Atkins Research and Development, Surry, for the U.K. Depart-
ment of the Environment's Central Directorate on Environmental
Pollution, July 1982.
3. Matthess, G., "In Situ Treatment of Arsenic Contaminated Ground-
water" Proceedings of an International Symposium on Quality of
Ground water, Noordwijkerhout, The Netherlands, Mar. 1981, (Studies
in Environmental Science, P. Glasbergen, Editor, Volume 17.) Else-
vier Scientific Publishing Company, Amsterdam, The Netherlands.
4. Aurand, K., et al., "Groundwater Impact of Silicate Gel Injections".
Proceedings of an International Symposium on Quality of Ground
water, Noordwijkerhout, The Netherlands, Mar. 1981, (Studies in
Environmental Science, P. Glasbergen, Editor, Volume 17.) Elsevier
Scientific Publishing Company, Amsterdam, The Netherlands.
5. Rott, U., "Protection and Improvement of Ground Water Quality by
Oxidation Processes in the Aquifer". Proc. of an International Sym-
posium on Quality of Ground water, Noordwijkerhout, The Nether-
lands, 23-27 March 1981. (Studies in Environmental Science, P. Glas-
berger, Editor, Volume 17.) Elsevier Scientific Publishing Company,
Amsterdam, The Netherlands.
6. Truett, J.B., and Holberger, R.L., "Feasibility of In Situ Solidifica-
tion/Stabilization of Landfilled Hazardous Wastes", Prepared by The
MITRE Corporation for the USEPA's MERL, Cincinnati, Oh,
MITRE Working Paper No. WP82W00147. McLean, Virginia. Draft
dated 22 March 1982.
7. Smith, M.A., and Assink, J.W., "NATO/CCMS Pilot Study on Con-
taminated Land; Report on Second Meeting of Study Group," Apr.
1982", Building Research Establishment Building Research Station,
Garston, WATFORD WD2 7 JR, England.
-------
ALLOCATING SUPERFUND LIABILITY
EDWARD I. SELIG, ESQ.
JOANNE RADISH, ESQ.
DiCara, Murphy, Selig & Holt
Boston, Massachusetts and Washington, D.C.
HOW THE PROBLEM OF ALLOCATION ARISES
The Comprehensive Environmental Response, Compensation
and Liability Act of 1980 ("CERCLA"), 42 U.S.C. §9601 et.
seq., popularly known as the "Superfund Act", provides new legal
authority for dealing with the release of hazardous substances to
the environment. As the legislative history of the law makes clear,
its principal goal is to enable the government to abate dangerous
conditions caused by leaching contaminants from inactive disposal
sites.' Under CERCLA Section 104, the United States may directly
undertake cleanup operations at offending sites, using money from
the Superfund, which is created and reserved for such activities by
other provisions of the Act. But if there are known "responsible
parties", they may be compelled to clean up a site at their own ex-
pense in accordance with administrative orders issued pursuant to
CERCLA Section 106.
The term "responsible party" refers to any individual or busi-
ness entity that is a potential defendant in a suit brought under
CERCLA Section 107 to recover: (I) costs incurred by the federal
government or a state, or by "any other person" in responding
to an actual or threatened release of hazardous substances, or
(2) compensation for any ensuing damage to publicly controlled
natural resources. The class of persons subject to Section 107 liabil-
ity includes the current owner or operator of a facility from which
there is a (threatened) release, as well as persons who owned or
operated the facility at the time hazardous substances were de-
posited there. In addition, waste generators who arranged to have
hazardous substances disposed of or treated at the facility, as well
as transporters who selected the facility and carried such substances
to it, qualify as members of this class of potential defendants.
It seems likely that most Section 107 actions will be brought by
the federal government after it is forced to respond to a release be-
cause cleanup orders have been ignored by responsible parties, or
because responsive action had to be undertaken before responsible
parties could be identified. Since defendants are strictly liable
under Section 107, i.e., liable for the costs and damages asso-
ciated with a release even if they took reasonable precautions to
prevent it and complied with all applicable regulations for man-
aging hazardous substances, the odds that the government will pre-
vail in such actions are relatively high. For it need not prove that
the defendants were "at fault" in order to win a case. As plain-
tiff, the United States must carry the considerably lighter, although
by no means trivial, burden of establishing that the defendants
own(ed) or operate(d) the releasing facility, generated and arranged
for the disposal of hazardous, substances there, or selected the
facility and carried such substances to it.
Winning a law suit and collecting a judgment are, however, two
different matters. And one of the most troublesome legal issues
that will be raised in Superfund cases is whether Section 107 liabil-
ity for costs and damages should be apportioned among the mul-
tiple defendants, or whether that liability is joint and several. Un-
fortunately, neither Section 107 itself nor its legislative history con-
clusively resolves the issue.
In general, if the rule of joint and several liability governs the
disposition of a Superfund suit, a successful plaintiff could recover
the full amount of costs and damages from any one of the multiple
defendants. Although the paying defendant may have the oppor-
tunity to seek "contribution" from the others for some part of the
plaintiff's award, his ability to secure contribution depends on the
solvency of his co-defendants. In contrast, if the rule of joint and
several liability is not applied, each defendant would be respon-
sible for paying an apportioned share of costs and damages to the
plaintiff. In this situation, the plaintiff bears the risk of not being
able to recover his entire loss, should one or more of the defen-
dants be insolvent and unable to pay his apportioned share.
The question of how Superfund liability should be allocated is a
hot topic of debate in the legal community2—a debate raising the
policy considerations and issues of statutory construction which
are the focus of this essay. However, since an analysis of these
matters is more easily presented once the reader is better ac-
quainted with the underlying principles of allocation, they are dis-
cussed next.
COMMON LAW BACKGROUND
The principles for allocating liability were developed at common
law, i.e., they were fashioned by judges in the course of deciding
suits brought by plaintiffs seeking damages for injuries to their
persons or property allegedly caused by the "tortious" or wrong-
ful conduct of others. These principles have evolved over the years,
and today are applied differently in different jurisdictions. In gen-
eral, however, the modern trend is that damages are apportioned
among multiple defendants whose concurrent acts have caused a
plaintiff to suffer: (1) distinct injuries, or (2) a single "divisible"
injury. An injury is divisible if there is a reasonable basis for cal-
culating the extent to which each defendant's activity contributed
to its occurrence. Although the plaintiff must introduce evidence
showing that the activity of each defendant was a legal cause of the
harm complained of, once this prima facie case is established, the
burden of proving that apportionment is appropriate is on the de-
fendants. If they fail to carry this burden, each is charged with
responsibility for the entire "indivisible" harm. In other words,
they are held jointly and severally liable, and the plaintiff may
collect his entire damage award from any one of them.3
Evidentiary problems aside, apportionment is undoubtedly prop-
er is this sort of case: Dl disposes of mercury at a site adjacent to
property owned by P, and D2 deposits benzene there. By some
quirk of hydrogeology, one of P's wells is contaminated by mer-
cury and the other by benzene. In this situation, there are two clear-
ly distinct injuries, and Dl should pay P for the loss of the mer-
cury-contaminated well, and D2 should pay for the loss of the
other.
The controversial cases involve concurrent acts that combine to
cause a single harm. Although the rule is that if there is a reason-
able basis for calculating each defendant's share of responsibility
for the harm, joint and several liability will not attach, it is diffi-
cult to predict whether a court will hold that such a basis exists,
given a particular set of facts. Suppose, for example, that P owned
only one well, which was contaminated by 100 ug/1. mercury and
200 pg/1 benzene. Does the fact that each defendant contributed
different amounts of distinct contaminants to the well provide a
reasonable basis for apportionment? Or is the harm indivisible? l
Suppose that for 10 years Dl owns and operates a facility that dis-
458
-------
459
LEGAL
charges mercury into an adjacent stream, and then sells the facility
to D2. D2 continues to operate the facility in the same manner un-
til the date of suit, 15 years after the date of sale. Should the dam-
age sustained by the downstream plaintiffs as a result of using con-
taminated water be divided among the defendants according to
their respective periods of facility ownership, or should they be sad-
dled with joint and several liability?;
As the footnote material indicates, in cases involving situations
similar to the ones described in the last two hypothetical, some
courts find a divisible harm, others do not. In principle, how-
ever, information regarding the quality and/or quantity of pollu-
tants attributable to each defendant obviously provides a theo-
retical basis for apportionment. It is the failure of the defen-
dants to introduce sufficient evidence in this area that explains
some decisions characterizing pollution as an indivisible harm.
Such decisions may also turn on (unarticulated) equitable con-
siderations. Thus, if one or more of the defendants is insolvent
and unable to pay any apportioned share, joint and several liabil-
ity might be imposed in order to ensure that the innocent plaintiff is
made whole.4
Because most jurisdictions now afford a right of contribution to
the defendant from whom a plaintiff has collected damages for a
so-called "indivisible harm", imposition of joint and several lia-
bility does not necessarily mean that the one defendant will ulti-
mately bear the entire cost of the plaintiff's judgment.' Rather,
it means that the paying defendant must bear the risk that all or
some of his co-defendants are insolvent. From the solvent ones, he
may recover money paid to the plaintiff in excess of his fair share,
of the damages. For purposes of a contribution claim, a de-
fendant's fair share is traditionally calculated by dividing the plain-
tiff's award by the number of defendants.
The type of pollution case raising the most difficult allocation
problems for the common law is one that has been characterized by
a prominent commentator as involving defendants with "dissim-
ilar functions,"4 e.g., hazardous waste generators and the owner or
operator of the disposal facility to which they sent their wastes.
There are few cases directly on point.
1. See e.g., Landers v. East Texas Sail Water Disposal Co., 151 Tex.
251, 248 S.W. 2d 731 (1952), where the court held that an oil company
owning a pipeline running near plaintiff's property and a salt water dis-
posal company owning an adjacent pipeline could be held jointly and
severally liable for damages sustained by the plaintiff when both pipe-
lines broke, pouring oil and salt water into his lake. Cf. City of Perth Ant-
boy v. Madison Industries, Inc., N.J. Super. (L-28115-76, Law Div.,
C-4474-76, Cl. Div., July 31, 1981), where liability was apportioned when
a wellfield was contaminated by organic chemicals from one defendant
and by heavy metal from another. Each paid for the different remedial
measures required to correct the pollution caused by their respective con-
taminants.
2. See e.g., State of New Jersey, Department of Environmental Pro-
tection v. Ventron Corp., N.J. Super. (A-1395-79, App. Div., Dec. 9,
1981), where the court imposed joint and several liability upon successive
owners of an industrial facility that had discharged mercury into an adja-
cent stream. It concluded that the harm complained of—mercury contam-
ination—was an indivisible injury, without addressing the point that length
of facility ownership apparently provided a reasonable basis for apportion-
ing damages. Cf. comment c to the Restatement (Second) of Torts, §433A:
"|l]f two defendants, independently operating the same plant, pollute a
stream over successive periods, it is clear that each has caused a separate
amount of harm, limited in time, and that neither has any responsibility
for the harm caused by the other."
3. Bui see Ewe/I v. Petro Processors of Louisiana, Inc. 364 So. 2d 604
(La. App. 1978), cert, denied, 366 So. 2d 575 (1979), where a facility
owner/operator who negligently permitted toxic wastes to leak onto plain-
tiff's property and a generator who continued to use the facility after learn-
ing of the problem were held jointly and severally liable for the contam-
ination of plaintiff's property; Stale of New Hampshire v. Mary Char-
pentier, Hillsborough County Superior Court (June 1982), where a jury
held the owner of a hazardous waste dump site liable for 10% of the clean-
up costs, but found the occupant and manager of the site, a number of
chemical waste haulers and a chemical waste disposal company jointly and
severally liable for the rest of the costs.
In theory, however, damaged caused by contaminants released
from a facility might be allocated among such defendants along
these lines: a court could hold each generator jointly and severally
liable with the facility owner or operator for each generator's
respective share of responsibility for the damages. Shares would be
apportioned among these jointly and severally liable pairs on
the basis of the number of barrels of waste, for example, that par-
ticular generators contributed to the facility. Each jointly and sev-
erally liable defendant from whom the plaintiff collected a portion
of his damages—i.e., either a generator or the facility owner or
operator—would have a right of contribution against the other. In
the absence of sufficient evidence germane to apportionment of
generator liability, or in the face of evidence that the facility own-
er or operator and a majority of the generators are financially in-
secure, all of the defendants might be held jointly and severally
liable for the entire damage award.
SUPERFUND LIABILITY—TEXTUAL AND HISTORICAL
ANALYSIS
Paragraph (a) of Section 107, the operative portion of CER-
CLA's liability provision, does not specify how Superfund liabil-
ity is to be allocated. In gen«ral, it merely defines the classes of
generators, transporters and facility owners or operators—the
responsible parties described at the beginning of this paper—who
are liable for response costs and for damages to natural resources
occasioned by a release of hazardous substances. However, a care-
ful textual analysis of Section 107(a) suggests a possible con-
gressional intent that the liability be joint and several.
In pertinent part, Section 107(a) provides that when "a" haz-
ardous substance is released from a facility, its current owner or
operator, the person who owned or operated it at the time any haz-
ardous substances were deposited there, generators who arranged
for the treatment or disposal of their hazardous substances at the
facility, "and" transporters who selected the facility and carried
such substances to it, "shall be held liable" for "all" response
costs.4 Arguably, the quoted language indicates that each de-
fendant is liable for the entire expense incurred by the plaintiff in
responding to a release, regardless of the extent to which the re-
leased material contains substances that a particular defendant
generated, transported or otherwise managed. In short, the defen-
dants may not be allowed to show that there is a reasonable basis
for apportioning Superfund costs. Consequently, a successful
plaintiff would, as a practical matter, be assured of full recovery
from any solvent defendant.
In the final analysis, however, the interpretation of Section
107(a) as a provision designed to impose joint and several liabil-
ity as a matter of law, i.e., regardless of whether apportionment
is appropriate under common law principles, should be rejected.
The section provides for strict liability, placing the costs incurred
by a plaintiff in responding to a release on specified parties re-
gardless of whether they were at fault for that release. At least
when the solvency of these possibly "innocent" parties is not
seriously in doubt—when the plaintiff can probably be made whole
even if damages are apportioned—it seems only fair to let de-
fendants limit their respective liabilities in the first instance by
showing, if they can, a factual basis for apportionment. Cer-
tainly the cryptic text of Section 107(a) does not provide sufficient
evidence of congressional intent to override such an equitable
consideration in order to insulate the plaintiff from any risk, how-
ever small, that his recovery will be less than 100%.
Indeed, an examination of CERCLA's legislative history indi-
cates that the language of Section 107 suggesting joint and sev-
eral liability (quoted above) is but a vestige of competing bills con-
sidered by Congress before Senator Stafford introduced S. 1480,
which became CERCLA. Some of these bills explicitly provided
for the imposition of strict, joint and several liability.1 Although
4. More precisely, they are liable for all response costs "not incon-
sistent with the national contingency plan." CERCLA Section 105 di-
rects the President to prepare an NCP delimiting the appropriate types
of response measures.
-------
LEGAL
460
S. 1480 dropped this provision, the standard of strict liability was
nonetheless retained, as the legislative history of the Act makes
clear.8 Further, according to what is now CERCLA Section
101(32), the term "liability" is to be construed as it is under Sec-
tion 311 of the Clean Water Act, 33 U.S.C. §1251 et. seq., and
the courts have routinely interpreted Section 311 as imposing lia-
bility in the absence of fault for discharges of oil or hazardous sub-
stances into navigable waters of the United States.9
In contrast, the effect of deleting the earlier language of joint
and several liability from S. 1480 was more controversial. On the
one hand, the legislative history indicates that some members of
Congress were persuaded to support S. 1480 because they believed
it did away with joint and several liability altogether:
The drafters of the...substitute have recognized [the] unfair-
ness and lack of wisdom in eliminating any meaningful link be-
tween culpable conduct and financial responsibility. Consequently,
all references to joint and several liability in the bill have been
deleted.10 Indeed, these statements aside, one could argue that, by
actually considering and then rejecting language expressly in-
corporating the rule of joint and several liability, Congress in-
dicated that this rule should not apply.
On the other hand, the managers of the bill in the House and
Senate made these comments:
[T]he terms joint and several liability [were] deleted with the in-
tent that the liability of joint tortfeasors be determined under com-
mon or previous statutory law.''
[The issue of joint and several liability is left to] traditional and
evolving principles of common law.12 And the Congressional
Record contains legal memorandum from the Justice Department
indicating that imposition of joint and several liability is the com-
mon law trend in pollution cases, as well as a Justice Department
opinion to the effect that joint and several liability is the appro-
priate standard under Section 311 of the Clean Water Act.u
If the Justice Department's interpretation of Section 311 had
been fully accepted by Congress, then the enactment of CERCLA
Section 101(32), which cross-references Section 311, would have in-
dicated an intent to adopt the standard of joint and several lia-
bility. CERCLA's legislative history as a whole, however, suggests
that the views of the Justice Department were discounted or for-
gotten. Moreover, they may have been incorrect.15
COMMON LAW IN THE FOREGROUND AGAIN
In light of its legislative history, the text of Section 107 can-
not be said to reveal a congressional intent that joint and sev-
eral liability should apply in Superfund cases as a matter of law.
On the other hand, neither does CERCLA define the circumstances
under which apportionment would be reasonable. Largely by de-
fault, then, the federal courts have received a mandate to develop
and apply common law principles in allocating costs and damages
in Superfund actions.5 The nature of the federal common law and
the circumstances under which it is applied are highly technical
matters, beyond the scope of this paper. In general, however, when
important federal interests are at stake, "statutory interstices" in
federal acts may be filled by the federal common law." Although
in fashioning this law, federal judges frequently turn to the salient
decisions of various state courts for guidance," these are not bind-
ing. In particular, courts have held that the federal interest in rem-
edying pollution, and in doing so through consistent adjudication
of cases brought under environmental statutes, is sufficiently
strong to warrant the development of such independent federal
common laws as may be needed to fill legislative gaps.18
As for CERCLA, it exhibits what might be called an "inter-
stice" on the issues of allocation arising out of liability for Super-
fund costs and damages. Although the federal common law that
can be expected to address these issues will surely be indebted to
5. CERCLA Section 113(b) provides that the United States District
Courts have jurisdiction over all cases arising under the Act. The action
may be brought in the district in which the release or damages occurred
or where the defendant resides, may be found, or has his principal
office.
older common law principles of allocation as applied by state
courts in comparable pollution cases, it will also undoubtedly
reflect a judicially ascertained federal policy on the matters of joint
and several liability and contribution.
REFERENCES
1. The major House and Senate reports accompanying the bill that be-
came CERCLA are found in 5 United States Code Congressional and
Administrative News 6119 (1980). CERCLA's legislative history is
summarized in Eckhardt, "The Unfinished Business of Hazardous
Waste Control", 33 Baylor L. Rev. 253 (1981), and in Grad, "A Leg-
islative History of the Comprehensive Environmental Response, Com-
pensation and Liability ("Superfund") Act of 1980", 8 Columbia J.
Env. L. 1 (1982).
2. See, e.g., Garrett, "Issues Relating to the Implementation of Super-
fund", 14 Nat. Res. Law Newsl. 5 (1982); Macbeth, "Superfund:
The Comprehensive Environmental Response, Compensation and
Liability Act of 1980", Expanding Liability in Environmental Law
(Harcourt Brace Jovanovich, 1981); Marnell, "Superfund: Con-
scripting Industry Support for Environmental Cleanup", 9 Ecology
L.Q. 524 (1981); Rodberg and Percell, "Joint and Several Liability
in Hazardous Waste Litigation", 3 Chem & Rad. Waste Lit. Rep.
448(1982).
3. For a history of the principles of allocation described above, see W.
Prosser, The Law of Torts Ch. 8 (4th ed. 1971). The current state of
the law is summarized in the Restatement (Second) of Torts, §§433A
and B, and in Rodberg and Percell, cited supra in reference num-
ber 2.
4. See Restatement (Second) of Torts, Section 433A, comment h, "Ex-
ceptional Cases".
5. For an analysis of the law of contribution more detailed than the one
given in the text below, see Rodberg and Percell, "Contribution
Among Defendants in Hazardous Waste Litigation", 3 Chem. & Rad.
Waste Lit. Rep. 591 (1982).
6. Rodberg and Percell, "Joint and Several Liability in Hazardous
Waste Litigation", supra at pp. 459-460.
7. See articles cited supra in reference number 1 for a description of the
bills preceding S. 1480.
8. See remarks of Senator Randolph in 126 Cong. Rec. S14964 (daily
ed. Nov. 24, 1980), and remarks of Representative Florio in 126 Cong.
Rec. HI 1773 (daily ed. Dec. 3, 1980), explaining that Stafford's bill
preserved the standard of strict liability.
9. See, e.g., Burgess v. M/V Tamano, 564 F.2d 964 (1st Cir. 1977);
United States v. Tex,Tow Inc., 589 F.2d 1310 (7th Cir. 1978);
Steuart Transportation Co. v. Allied Towing Corp., 590 F.2d 609
(4th Cir. 1979); United States v. LeBoeuf Brothers Towing Co., 621
F.2d 787 (5th Cir. 1980).
10. Remarks of Senator Helms, 126 Cong. Rec. S15004 (daily ed. Nov.
24, 1980).
11. Remarks of Representative Florio, 126 Cong. Rec. HI 1787 (daily ed.
Dec. 3, 1980).
12. Remarks of Senator Randolph, 126 Cong. Rec. S14964 (daily ed.
Nov. 24, 1980).
13. 126Co«g. Rec. HI 1788-89 (daily ed. Dec. 3, 1980).
14. Ibid.
15. See United States v. M/V Big Sam, 505 F. Supp. 1029 (E.D. Louisiana
1981), which rejected the government's position that third party lia-
bility under Section 311 should be read as joint and several: "We
read the statute as being in the disjunctive and giving a choice of
parties, depending on the existing facts." Id. at p.1033.
16. Illinois v. Outboard Marine Corporation, 619 F.2d 623, 630 (7th Cir.
1980). See also Friendly, "In Praise of Erie and of the New Fed-
eral Common Law", 39 N.Y.U.L.R. 383 (1964), and Clearfield Trust
Company v. United States, 318 U.S. 363 (1943), and its progeny.
17. See Textile Workers v. Lincoln Mills, 353 U.S. 448 (1957); D'Oench,,
Duhme & Co. v. Federal Deposit Insurance Company 315 US
447(1942).
18. See United States v. Solvents Recovery Service, 14 ERC 2010, 2020,
496 F. Supp. 1127 (D. Conn. 1980); Illinois v. Milwaukee 406 U s'
91(1972). ' '
-------
PROTECTION FROM LONG-TERM LIABILITY AS A RESULT OF
SUPERFUND REMEDIAL ACTIONS
D.E. SANDERS
F. SWEENEY
E. HILLENBRAND
JRB Associates, Waste Management Department
McLean, Virginia
FEDERAL EPA STANCE
The potential for civil litigation as a result of accidental dis-
charges by a contractor of wastes secured under Superfund is a
very valid concern, and one that is being addressed at both the
Federal and state levels at this time. The USEPA generally pro-
vides indemnification for contractors in its two Superfund Tech-
nical Support Zone Contracts and in its Technical Assistance
Team contract. They are providing this indemnification in order
to encourage maximum competition for the contracts. The reason-
ing for indemnification is that a number of contractors expressed
concern about the possibility that they will be liable to third par-
ties for damages arising out of their remedial response performance
for USEPA. These contractors stressed that the activities con-
ducted under these contracts may involve significant risk of in-
jury to third parties, the effects of which may not be evident for
years. They expressed reluctance to compete for contracts where
they might be held liable for these long term risks.
USEPA realized that insurance against the risk of such liabil-
ity would be extremely costly, a cost that USEPA would ultimate-
ly bear. Moreover, insurance against certain extraordinary risks ap-
pears to be difficult or impossible to obtain, and even where such
insurance is available, some insurers insist on limiting coverage to
a period of time, perhaps three years. Thus, although the Federal
Tort Claims Act does not require or prohibit indemnification for
contractors, USEPA felt indemnification would be the most effec-
tive method available to ensure best use of the fund and maximum
competition for the contract.
Authority to indemnify is limited by the Anti-Deficiency Act.1'2
USEPA believes, however, that agencies may agree to indemnify
contractors if it is as "necessary expense," if adequate compe-
tition could not be otherwise obtained, or if the agency were faced
with the certainty of reimbursing a large amount of insurance costs,
as contrasted with the possibility of large indemnity payments be-
ing required.2'3
There are two types of indemnity agreements, unqualified and
qualified.4
Unqualified Indemnity Agreements
To create an unqualified obligation to indemnify consistent with
the Anti-Deficiency Act, an agreement must either: (1) provide that
the duty to indemnify applies only to loss or damage to specified
property (in which case the total indemnity could not exceed the
value of the property); or (2) establish a ceiling on the amount the
government is obliged to pay for third-party liability. The problem
with an unqualified indemnity agreement is that it is necessary to
establish a reserve of funds to cover the contingency. The Comp-
troller General has stated that: "Where circumstances are such as
to indicate some likelihood of payment being required, an appro-
priate reserve or obligation would, of course, have to be established
to assure the availability of funds."'
Qualified Indemnity Agreements
In a qualified indemnity agreement, any contract providing for
assumption of risk by the Government for. contractor owned prop-
erty must clearly provide that: (1) in the event the government
has to pay for losses, such payments will not entail expenditures
which exceed appropriations available at the time of the losses; and
(2) nothing in the contract may be considered as implying that the
Congress will, at a later date, appropriate funds sufficient to meet
deficiencies.6 Without inclusion of provisions along these lines, leg-
islative exemption from the application of the statutory prohi-
bitions against obligations exceeding appropriations would have
to be obtained.
The application of this requirement to an agreement to indem-
nify for third party liability has been addressed by USEPA.
The phrase "available at the time of the losses" has been inter-
preted to mean that funds must be available at the time the obliga-
tion becomes certain as to amount.
USEPA has chosen the "Qualified Indemnity Agreement" ap-
proach in the language for the Technical Support Zone con-
tracts. The agency has modified the standard "Insurance-Liability
to Third Persons" clause of 41 C.F.R. 1-7.204-5 to make clear
that reimbursement for liabilities to third persons for loss of or
damage to property, or for death or bodily injury as the result of
contractor activities shall not exceed appropriations available at the
time such liabilities are represented by final judgements or by settle-
ments approved in writing by'the government. The language also
makes clear that an agreement to reimburse the Contractor for cer-
tain liabilities to third persons shall not be interpreted as imply-
ing that the Congress will, at a later date, appropriate funds suf-
ficient to meet deficiencies.
The USEPA Office of General Counsel (OGC) expressed con-
cern that the language for the Zone Contracts does not make clear
that the "appropriations available" would come from the Super-
fund. This raises the possibility of the necessity for creation of a
separate fund for liability payments, which was not the intention of
the Agency. This point will be clarified in future language to make
it clear that "appropriations available" must come from Super-
fund.7 Still, a question may remain as to whether general substan-
tive limitations in Superfund on third party claims are applicable to
similar claims arising out of contractor malfeasance.
The question then arises about liability when the cleanup is per-
formed under a State/USEPA Cooperative Agreement. USEPA's
advice is that Cooperative Agreements, which are required when
the state has lead responsibility for the activity and funds are trans-
ferred to the state, and Grants fall under the same liability re-
quirements, even though grants may be less complicated than many
Cooperative Agreements. (For example, under some Cooperative
Agreements, USEPA may do part of the work; i.e., the design
or the feasibility study, while the state and/or state contractor
performs the remainder of the remedial action.) However, 40
C.F.R. Part 30, the Grant Regulations, holds that normally
461
-------
LEGAL
462
USEPA has no liability after funds are given to the grantee. The
USEPA OGC's position is that USEPA will also not indemnify
the state under a Cooperative Agreement.8
The USEPA OGC also is of the opinion that Section 107(d) of
Superfund probably does not protect either contractors or the gov-
ernment against claims by third parties. Their interpretation of
107(d) is that it is a limitation on the liability to the government
of contractors and others for cleanup costs under Title I of the Act,
and probably does not limit private rights under ordinary tort prin-
ciples.'
However, OGC also stressed that in the event the contractor is
held liable to third parties for damages arising out of Superfund
remedial response activities, and money was not available to pay an
indemnity, the contractor could seek a private bill in Congress.
STATE STANCE
A number of states have awarded contracts for remedial activ-
ities on Superfund sites. Only two of the states which were con-
tacted for this study, New Jersey and California, are using or are
planning to use language in their contracts which partially or total-
ly indemnifies the state from liability for third party claims, there-
by placing the responsibility on the contractor. Both of these states
are also very specific about the types of insurance which the con-
tractor must carry to cover remedial action activities.
In New Jersey's contracts, indemnification is addressed through
the following language:
"The Contractor shall be solely responsible for and shall keep,
save, and hold harmless the State of New Jersey and its em-
ployees from and against any and all claims, demands, suits,
actions, recoveries, judgements, and costs and expenses in con-
nection therewith on account of the loss of life, property, or
injury or damage to the person, body, or property of any per-
son, agency, corporation, or government entity, which shall
arise from or result directly or indirectly from the work and/or
materials supplied by or arising out of the performance of this
contract. The Contractor's liability under this contract shall con-
tinue after the termination of the contract with respect to any
liability, loss, expense or damage resulting from acts occurring
prior to termination. This indemnification obligation is not lim-
ited by, but is in addition to the ifisurance obligation contained
in this agreement."10 >
New Jersey requires contractors to maintain insurance in accor-
dance with the following language from a contract:1'
"The Contractor shall secure and maintain in force for the
term of the contract the following minimum insurance cover-
ages. The contractor shall provide the State of New Jersey with
current Certificates of Insurance certifying coverage and con-
taining the provision that the insurance shall not be cancelled
for any reason except after 30 days written notice to be directed
to the State of New Jersey, Director, Division of Purchase and
Property.
"Comprehensive General Liability Insurance as broad as the
standard coverage form currently in use in the State of New
Jersey which shall not be circumscribed by any endorsements
limiting the breadth of coverage. The policy shall include an en-
dorsement (broad form) for contractual liability, an endorse-
ment for completed operations liability, and shall include the
State of New Jersey as an additional insured. Limits of bodily
injury liability and not less than one million dollars per occur-
rence for property damage liability."
"Comprehensive Automobile Liability Insurance covering
owned, non-owned, and hired vehicles with minimum limits of
one million dollars per occurrence for bodily injury and prop-
erty damage liability combined.
"Workers Compensation Insurance applicable to the laws of
the State of New Jersey and any other state where the contrac-
tor will be active under this contract, and Employers Liability
Insurance with a limit of not less than $250,000. The policy
shall be endorsed to include coverage under the United States
Longshoremen's and Harbor Workers Compensation Act and
any other Federal Workers Compensation Law which may apply
to the Contractor's operations."
The State of California protects itself from liability in three
ways.12 First, they use a standard indemnification clause which
indemnifies the state generally against contractor malfeasance.
Second, the state has developed a specific indemnification clause
for hazardous waste sites. This clause requires the employer to:
•Carry specified insurance
•Notify state and contractor employees of the hazards present at
the site
•Notify the state of any spills
Third, the State Tort Act provides that there must be a showing
of gross criminal negligence on the part of the state before any
state entity can be prosecuted. There is also another section in the
State Tort Act which protects the contractor unless gross negli-
gence on the part of the contractor can be proved.
Both New Jersey and California stated that they do not believe
competition for the contracts was reduced by the provision that the
contractor assumes the long term liability for the remedial actions.
Both states did, however, indicate that they believe there may be
health problems in the future which the liability language and in-
surance requirements will not cover.13
The State of Washington includes language in their contracts
which provides protection from claims by the contractor against
the State, but has not addressed the issue of protection from long
term liability.
The State of Michigan did not address this issue in the Superfund
pre-implementation contract which they have already awarded.
They do, however, plan to include language in future contracts
that will include provisions to hold the contractor liable if engi-
neering work fails, while the State will be liable if failure is the
result of remedial design.14 Louisiana is also drafting an Article
which will address the issue of long term liability for remedial
actions.
The State of Colorado, which awarded a contract for work on
the Denver radium sites, included language which indemnified the
Federal government and the state. They required their architect/
engineer contractor to indemnify itself for one year after the work
is completed. However, the contract reads that the State and Fed-
eral government, through the contractor, will "attempt to reduce
the radiation levels,". It does not provide any guarantee to what
value the radiations levels will be reduced. This, the state believes,
will help prevent a basis for third party liability claims.'!
The other five states which were contacted have not yet ad-
dressed the issue of long term liability, but realize they will need
to in the near future.
The experience of selected states indicates several potential bases
for determining indemnification rules. These are:
•Design versus construction. (Michigan) This pinpoints respon-
sibility for remedial design failures versus responsibility for re-
medial construction failures.
•Standard of care. (California) The standards used are:
-Negligence, where the failure is fairly commonplace
-Gross Negligence, where the failure is of such magnitude that
careful attention should have prevented it
-Knowledge where the contractor had knowledge of the problem,
but took no corrective action
•Monitoring and Disclosure. (California) The contractor must
agree to disclose information and keep the State informed on the
status of activities at the site.
•Remedial Action. This considers the question of responsibility
for the remedial action itself. For example, the government may
assume third-party liability, while holding the contractor respon-
sible for redoing a remedial action which fails. Insurance is an
important consideration in this area of negotiation in order to
ensure site integrity and compensation for damages.
-------
463
LEGAL
•Insurance. (California) This involves specifying what insurance is
carried by which entity, and detrmining the distribution of insur-
ance payments across time. A damage ceiling is an important
point to consider when determining insurance requirements.
•Damage Ceiling. Here, exposure of the contractor is limited to a
fixed sum.
•Types of Damages. These include:
-Remedial Action
-Remedial Costs
-Personal Injury
-Damage to Property
-Environmental Damage
•Site Specific versus General Rule. Contract provisions can be site
specific, depending on such factors as the degree of hazard of
the site, rather than providing a general rule for all sites.
Obviously, these contractual determinants are not mutually ex-
clusive. Accordingly, a state can select any combination of these
determinants to modify existing contractual provisions for the
clean-up of hazardous waste sites.
CONCLUSIONS
There appears to be significant disagreement between USEPA
and the states on the degree to which long term liability rules can
potentially affect contractor competition. It is easy to imagine in-
stances where the choice of an indemnification rule could signif-
icantly affect competition. For example, the standard of care (e.g.,
negligence versus gross negligence) may help determine the pool of
qualified contractors. Also, the imposition of broad long term lia-
bility on contractors may have a differential impact on larger,
more established firms and smaller, more financially marginal
firms. Consequently, the competition issue should be carefully
considered at the time an indemnification rule is established.
Finally, the states should establish a-clear-cut contractual rule
determining the entity responsible for additional remedial action in
the event the original remedial action fails. This is important be-
cause Superfund terminates in September, 1985. Thus, appro-
priations available from Superfund to respond to any liability suits
may be significantly reduced or non-existent after this time.
ACKNOWLEDGEMENTS
The authors are grateful to the following people for their assis-
tance in the preparation of this paper:
Mr. Paul Nadeau, Chief, Design and Construction Section, Office
of Emergency and Remedial Response, EPA, Washington, D.C.
Ms. Karen Clark, Attorney Advisor, Office of General Counsel,
EPA, Washington, D.C.
Mr. Dave Mark, New Jersey Department of Environmental Pro-
tection, Trenton, N.J.
Mr. Bill White, Staff Counsel, State Water Resource Control
Board, Sacramento, Calif.
Ms. Claudia Weaver, Environmental Specialist, Department of
Natural Resources, Lansing, Mich.
Mr. Tom Looby, Office of Health Protection and Environmen-
tal Health, Denver, Col.
REFERENCES
1. 31U.S.C. 665(a).
2. Nantkes, Donnell L., "Indemnification of Superfund Contractors,"
Internal EPA Memorandum, June 26, 1981.
3. 45 Comp. Gen. 824 (1975) and 42 Comp. Gen. 708, (1963).
4. Nantkes, D., op. cit., June 26, 1981.
5. 45 Comp. Gen. 569(1966).
6. 31U.S.C.665(2)and41U.S.C. 11.
7. Personal Communication, Ms. Karen Clark, Attorney Advisor, Office
of General Counsel, EPA, September 16, 1982.
8. Ibid.
9. Nantkes, D., op. cit., June 26, 1981.
10. Department of Environmental Protection, Division of Hazard Man-
agement, Trenton, New Jersey, "Burnt Fly Bog Request for Pro-
posal," July, 1981.
11. Ibid.
12. Perspnal Communication, Mr. Bill White, Staff Counsel, California
State Water Resource Board, December 3, 1981.
13. Personal Communications, White, California and Mack, New Jersey,
Decembers, 1981.
14. Personal Communication, Ms. Claudia Weaver, Environmental Spe-
cialist, Michigan Department of Natural Resources, December 3,
1981.
15. Personal Communication, Mr. Tom Looby, Colorado Office of
Health Protection and Environmental Health, December 4, 1981.
-------
LIABILITY AND INSURANCE ASPECTS OF CLEANUP OF
UNCONTROLLED HAZARDOUS WASTE SITES
PAUL E. BAILEY, J.D.
ICF Incorporated
Washington, D.C.
INTRODUCTION
Cleanup operations at uncontrolled hazardous waste sites can
complicate an already complex terrain of legal liability and insur-
ance. Rather than review the underlying common law and statu-
tory liabilities of responsible parties, the author's focus in this
paper is on how the cleanup operations themselves generate an
independent set of liabilities and how that risk can be reallocated
through various contractual agreements.
Because legal outcomes depend very heavily on the facts of spe-
cific situations, a treatise would be required to adequately discuss
the myriad of potential situations that could arise as a result of
cleanup operations at uncontrolled hazardous waste sites. The
author can only present in this paper the broad outlines of typical
situations and applicable legal doctrines. It is also impossible for
the author to analyze and report here the many variations in ap-
plicable state laws.
This paper has three major parts: (1) an overview of what occurs
during cleanup operations and what can go wrong is presented,
(2) potential types of resulting liability are discussed, and (3) the use
of private agreements to shift financial responsibility for liability
claims is surveyed. Given the complexity of this subject matter,
the author's goal is to clarify basic terms and relationships for the
general professional reader.
CLEANUP ACTIVITIES AND POTENTIAL DAMAGES
Uncontrolled waste disposal sites include: landfills containing
loose or containerized waste, open dumps of unidentified barrels
and drums, contaminated structures, contaminated river or lake
sediments, and unstabilized surface impoundments. Hazardous
wastes may include flammable, explosive, radioactive, carcino-
genic, and infectious materials; pesticides, heavy metals, organic
solvents, waste oils, inorganic acids, etc. Among the key hazards
are explosion and fires, release of toxic fumes, contamination of
private real estate, groundwater, and surface water supplies.
Contractors will play a key role in the cleanup of uncontrolled
hazardous waste sites, usually under the supervision of federal
and/or state government personnel. Often, a prime contractor will
use the services of one or more subcontractors for different as-
pects of the cleanup operation. There are several aspects of oper-
ations at uncontrolled hazardous waste sites that contribute to the
dangerousness of the activity but the lack of knowledge of the na-
ture and amount of hazardous substances present and their dis-
position is the major factor.
Contractors are likely to perform some or all of the following
activities as part of a cleanup effort:
•On-site investigation (preliminary assessments, site inspections,
remedial investigations)
•Remedial action planning and design
•Selection of subcontractors including transporters, disposal sites,
etc.
•Technical services including construction of fences, dikes, berms,
or waste impoundments
•Removal and disposal of wastes, decontamination of structures
and equipment
•Temporary provision of alternate water supplies
•Cleanup and restoration or replacement of affected natural re-
sources (including groundwater)
Contractors may face liability exposure to the extent that dam-
age proximately results from any of these activities.
On-Site Investigation
The contractor may be called upon to perform site investiga-
tions and develop characterizations of the hazardous substances
present, pollutant dispersal pathways, types of receptors, and site
management practices. Where damage results that would other-
wise have not occurred had site investigation been properly per-
formed, liability could attach.
The major concern of the contractor should be the proper iden-
tification of the wastes, particularly those that may be explosive,
reactive, or incompatible. In addition, careful identification and
location of nearby sources of drinking water is necessary. Any
plan for protecting the safety of workers and nearby residents will
only be as good as the data it is based on. The uncontrolled na-
ture of the waste site heightens the risks involved.
Remedial Planning, Design, and Construction
The remedial action phase is critical to cleanup operations. It
may include erecting dikes, constructing trenches or ditches, in-
stalling a clay cover or synthetic liners, segregation of reactive
wastes, dredging or excavations, repair or replacement of leaking
containers, collection of leachate and run-off, physical cleanup of
hazardous substances or their neutralization, treatment and incin-
eration.
While planning is an essential management function, the concept
of proximate cause requires that execution take precedence over
planning when assessing legal causality. Poor planning by itself is
rarely the cause of damages and claimants will face a formidable
burden of proof in this regard. A cleanup contractor has a strong
self-interest in selecting subcontractors with care and in ensuring
proper performance on the part of all subcontractors since a prime
contractor may be liable for harms caused by the actions or omis-
sions of its subcontractors.
Design and construction, however, is a major area of concern.
Improper design and/or construction may result in waste contain-
ment structures that do not work. Breach of an impoundment,
liner, or cover due to poor design and construction could result
in damage claims brought against the contractor.
Cleanup contractors need to exercise special care in designing or
selecting underground storage tanks, leachate collection systems,
gas collection systems, liners, and the like. In the construction
field, architects are frequently sued for negligence in specifying
some new product without adequate testing; cleanup contractors
may face similar suits.
What Can Go Wrong
During the cleanup of an uncontrolled hazardous waste site,
the following damages typically could occur:
•Injury or death to a person on-site, such as a cleanup contrac-
tor employee or a government employee due to explosion of a
55 gal drum during earth moving
464
-------
465
LEGAL
•Property damage resulting from a fire which starts on-site and
spreads to surrounding property
•Spill of hazardous wastes during transport for off-site disposal
•Release of toxic fumes into surrounding community causing min-
or injuries due to acute reactions and exposure to potentially car-
cinogenic compounds
•Continued community exposure to contaminants due to improper
design or construction of waste containment systems
•Injury or death of an emergency response professional such as a
fireman
•Increased cleanup costs or harm to natural resources.
From a legal point of view, it is essential to know why these dam-
ages occurred. It can make a significant difference if the damages
resulted due to: (1) negligence, (2) an intentional act (e.g., arson),
or (3) due to no one's fault in particular. In addition, if an inten-
tional or negligent act is involved, the identity of the actor is im-
portant. Among the possibilities are: (1) the prime contractor and
its employees, (2) subcontractors and their employees, (3) tres-
passers, (4) bystanders, (5) residents, or, (6) government personnel
(federal, state, or local). The possible scenarios are numerous.
LIABILITIES RESULTING FROM CLEANUP
Liability has a variety of meanings, even in the law. At its broad-
est, the term can refer to any obligation. As the term is used in this
paper, it has a narrower focus. By liability, the author means a
legally enforceable responsibility to take action or provide money
for particular kinds of losses, damages, or costs.
There are many sources of legal liability. It can arise under fed-
eral and/or state law, both through statutes promulgated by legisla-
tive bodies and through judicially-developed common law. Liabil-
ity can be civil or criminal. In a legal proceeding, a claimant
will name as many alternative theories of liability as possible in
support of its case.
Liability may also arise through contractual agreements. The
most familiar instance of this is the liability created for an in-
surer through the contract of insurance. An indemnification agree-
ment likewise can create liability. This type of liability is some-
times referred to as "secondary" or "derivative" to distinguish it
from liability imposed directly, without reference to private agree-
ments.
Liability always needs to be defined in terms of whose liabil-
ity, to whom, and for what. This is in addition to specifying
whether the liability is federal or state, statutory or common law.
Whose liability refers to the class or classes of legal persons to
whom the liability can legally attach. For example, sovereign im-
munity protects state and federal governments from certain types
of liability. Waste generators may be treated differently from
common carriers transporting waste. Likewise, liability may de-
pend on what class of legal persons is the claimant. Because of
state workers' compensation statutes, employees may have more
limited options if injured as a result of cleanup activities than do
bystanders or residents. Common law similarly restricts certain
causes of action to certain persons, and injuries outside the "scope
of the risk" may not receive preferential judicial treatment. Final-
ly, liability may only exist for certain kinds of damages or for cer-
tain kinds of conduct.
For the purposes of this paper, several sources of liability need to
be addressed: (1) the federal Comprehensive Environmental Re-
sponse, Compensation and Liability Act ("CERCLA" or Super-
fund), (2) state common and statutory law, and (3) the doctrine
of vicarious liability known as respondeat superior. Liability for
damages resulting from cleanup operations will also be affected by
the legal concept of proximate causation which will be discussed
at the end of this part of the paper.
CERC1 A and Liability for Damages
Section 107 of CERCLA contains most of the key liability pro-
visions of the legislation. In addition to establishing liability for
response costs and natural resource damages, Section 107 of the
Comprehensive Environmental Response, Compensation and Lia-
bility Act also addresses liability for damages resulting from remed-
ial and response cleanup actions. The statutory language reads
as follows:
"No person shall be liable under this title for damages as a re-
sult of actions taken or omitted in the course of rendering care,
assistance, or advice in accordance with the national contingency
plan or at the direction of an onscene coordinator appointed
under such plan, with respect to an incident creating a danger to
public health or welfare or the environment as a result of any re-
lease of a hazardous substance or the threat thereof. This sub-
section shall not preclude liability for damages as the result of
gross negligence or intentional misconduct on the part of such
person. For the purposes of the preceding sentence, reckless,
willful, or wanton misconduct shall constitute gross negligence."
(Section 107(d))
This section only addresses CERCLA liability for response costs
and natural resource damages. It says nothing about third party
damages resulting from cleanup activities. Third party damage pro-
visions were dropped from CERCLA prior to its enactment in
1980. Section 114(a) makes it clear that CERCLA does not pre-
empt any state from imposing statutory liability for third-party
damages from the release of hazardous substances.
Although the language of CERCLA is essentially silent on the
third party damage liability issues of greatest concern to cleanup
contractors, the legislative history of the Act does address this mat-
ter in part. The Senate Report on S. 1480 (Report No. 96-848),
the Senate bill that eventually became CERCLA, discusses third
party property damages resulting from authorized response ac-
tions. It states on page 82:
"It should be clear, as under section 311 [of the Clean Water
Act], if an innocent third party's property is damaged pursuant
to authorized removal or remedial operations those damages are
part of the cost of removal or remedial payable from the fund
and chargeable to the owner or operator."
This suggests that, at least with regard to property damages,
contractors are not liable but, rather, the originally responsible
party is. It is also possible that other third party damages result-
ing from response actions could be dealt with this way, i.e., by
defining them as response costs that are payable by the Response
Fund and chargeable to the owner or operator. Regardless of
whether the legislative history extends to third party damages be-
yond property, there is a larger issue: to what extent is legisla-
tive history operationally useful in protecting potentially liable
contractors? The question of personal injury damages has not
come up under Section 311 of the Clean Water Act. The Con-
gressional intent is not clear, nor is the statutory language.
Because CERCLA is silent on third party liability, state statutory
and common law will control with respect to third party damages
incurred as a result of both the uncontrolled disposal of hazardous
waste and damages resulting from cleanup operations. Section
107(d) quoted above offers no definite protection to cleanup con-
tractors outside of CERCLA's narrow focus on cleanup costs and
natural resources damages.
CERCLA addresses secondary or derivative liability in Section
107(e). This section, at first, seems to contradict itself, but in fact
it does not:
"(e)(l) No indemnification, hold harmless, or similar agree-
ment or conveyance shall be effective to transfer from the own-
er or operator of any vessel or facility or from any person who
may be liable for a release or threat of release under this sec-
tion, to any other person the liability imposed under this section.
Nothing in this subsection shall bar any agreement to insure,
hold harmless, or indemnify a party to such agreement for any
liability under this section.
(2) Nothing in this title, including the provisions of paragraph
(1) of this subsection, shall bar a cause of action that an owner
or operator or any other person subject to liability under this
-------
LEGAL
466
section, or a guarantor, has or would have, by reason of subro-
gation or otherwise against any person."
This is a complex provision that accomplishes several goals.
First, Section 107(e)(l) prevents any shift of primary liability away
from responsible parties. Thus, a generator whose standard waste
disposal contracts include agreements by the disposer to indemnify
the generator for any damages would not be able to thereby avoid
the joint and several liability created by CERCLA. Second,
CERCLA maximizes the number of possible "deep pockets" for
payment of claims by not barring private insurance or indem-
nity arrangements—these contractual provisions act to create a
secondary liability on the part of the insurer or indemnifier and to
assure that certain sums will be available to pay claims. Third,
paragraph (2) fosters the negotiation of private risk sharing ar-
rangements by making explicit that CERCLA is not to be con-
strued as barring insurers, indemnifiers, or guarantors from suing
responsible parties for reimbursement pursuant to subrogation
clauses.'
State Law
State statutory and common law are key sources of liability that
may apply to cleanup operations at uncontrolled hazardous waste
sites. There is much variation in the types and language of exist-
ing state statutes and there is often conflict among the holdings of
state common law.
Statutory Law. Within the past 15 years, an increasing num-
ber of states have promulgated laws imposing liability for harms
caused by releases of hazardous substances.2 The goal of this leg-
islation has often been to create "strict" (i.e., no fault) liability
on the part of responsible parties to answer for damages caused
by releases. Whether or when a cleanup contractor becomes a re-
sponsible party under the terms of these laws is usually not ad-
dressed. It should not be assumed that these laws will not apply to
remedial and removal actions.
Many state statutes have been enacted for the protection of em-
ployees against occupational disease and injury. Violation of statu-
tory requirements generally constitutes actionable negligence. In
addition, statutes adopted in many jurisdictions, such as employ-
ers' liability acts, workmen's compensation acts, and similar stat-
utes have altered, limited, abolished, or supplanted common law
rules regarding the liability of an employer for injuries to an em-
ployee. These statutes do not address third parties.
In general, workers' compensation acts provide .no-fault protec-
tion, i.e., a right to compensation for all injuries incident to the
employment, with certain exceptions, is given. Where such acts are
compulsory, not elective, they supercede existing laws and an em-
ployee cannot recover at common law for an injury which is com-
pensable under the acts. This is known as the exclusivity doctrine.
Common Law. In many instances, injured parties will base their
claims for recovery under applicable state common law. Typical-
ly, negligence, strict liability, nuisance, or trespass will be asserted
as the basis for recovery. The strengths and limits of these doc-
trines have been adequately explored elsewhere and need not be re-
peated here.3 These doctrines are generally used by residents and
bystanders with damage claims but may also be asserted by em-
ployees in certain contexts.
In many instances, cleanup contractors may also be manufac-
turers, sellers, or suppliers, which renders them vulnerable to pro-
ducts liability suits. It is now generally accepted that the liabil-
ity of a construction contractor is the same as that imposed upon a
manufacturer for injuries resulting from defective products. Under
this view, a contractor is held to a standard of reasonable care for
the protection of third parties who may foreseeably be endangered
by his negligence, even after acceptance of the work by the contrac-
tee. Breach of the duty of proper performance which results in in-
jury is actionable. The plaintiff must still prove, of course, the
contractor's negligence, that such negligence is the proximate cause
of the injury, and that the plaintiff came within the scope of the
risk. The liability of an employer as a manufacturer is distinct from
its liability as an employer and may be asserted successfully as the
grounds for a claim by an injured employee.
If the negligence of a contractor causes injury to other contrac-
tors or other employees, to his own employees or employees of his
principal, or to some third person for whose protection the con-
tractor is bound to exercise due care, the contractor may be direct-
ly liable for the injury under state common law.
With respect to its employees, it is the general duty of the em-
ployer to provide a reasonably safe place of work and to furnish
suitable and safe instrumentalities with which to work. The em-
ployer's liability to provide a safe place of work is severely com-
promised when cleanup operations at uncontrolled waste sites are
involved. Nevertheless, the standard of care to be exercised must
be commensurate to the dangers of the business. This means that
the employer must warn the employee of conditions under which he
is employed which may engender disease. The employer may be
held directly liable for occupational diseases which can be trace-
able to some negligence on its part. The employer must also furn-
ish suitable protection from the danger (e.g., respirators), provided
he is in a position to have greater knowledge of the existence of
the danger than the employee.
Where the harm was caused by the negligence of a subcontrac-
tor, the prime contractor may be liable under the theory that the
firm selected the subcontractor without satisfying the necessary
standard of care and is thus itself liable. The alleged negligent se-
lection of subcontractors theory would most likely arise in a suit
naming a "deep pocket" prime contractor for damages due to per-
sonal injury from explosion, fire, or significant release of fumes or
material during on-site operations (including removal for off-
site disposal). The claimant could be an employee of a subcon-
tractor, or a third party bystander or resident, or even an emer-
gency response professional. All these parties could, of course,
proceed against the negligent subcontractor itself.
Vicarious Liability
This type of common law liability is based on a relationship
between parties, though not through a contract or indemnifica-
tion agreement. Rather, vicarious liability is based on social pol-
icy considerations under which it is determined that irrespective of
fault, a party should be held to be liable for the acts of another.
Thus, there need be no negligent act or omission of the party held
vicariously liable. The application of vicarious liability to employ-
ers or contractors occurs through the doctrine of respondeat
superior, discussed below.
It is universally recognized that an employer is civilly liable for
injuries to the person or property of third persons caused by the
torts, negligence, frauds, deceits, concealments, misrepresenta-
tions, and other malfeasances or misfeasances of the employee
which are within the scope of his employment. This rule of vicar-
ious liability is one of long standing. Among the showings required
to establish liability is proximate cause. Of course, if the employer
has been held liable by a third party victim, he has a right to be in-
demnified by the wrongdoer employee.
An employer may be vicariously liable for tortious acts of their
employees and may also be vicariously liable for injuries to em-
ployees, even if caused by third persons. An initial question is al-
ways whether the person either injured or causing the injury is, in
fact, an employee. Whether a person is an employee or an inde-
pendent contractor depends upon the power of control which the
employer is entitled to exercise over the person in question, re-
gardless of the existence of written subcontracts. A "cost plus"
contract is not itself determinative.
To a large extent, the difference between an employee and an in-
dependent subcontractor disappears when the nature of the work
being performed is inherently dangerous—as is the case with clean-
up operations at uncontrolled hazardous waste sites. The general
rule of law prevents an employer from using independent contrac-
tors as a shield against liability for harms. Injuries resulting from
excavation, blasting, and demolition work are frequently handled
in this manner. If the injury that occurs might have been antici-
-------
467
LEGAL
pated as a probable consequence of the execution of work as-
signed to an independent contractor, the employer as well as the
subcontractor may be held liable.
Where the "inherently dangerous activity" doctrine is applic-
able, the law invokes the theory of responded! superior, even
though an employer has attempted to escape liability by employing
an independent contractor. There is much variation in the specific
application of this theory. In many cases, recovery is limited to
physical harm only and presupposes that the subcontractor acted
negligently by failing to take suitable precautions.4 Additionally,
there is a conflict of opinion as to whether this doctrine of respon-
deat superior can be invoked by employees of the employer or is
only available to third parties (e.g., bystanders, subcontractor
employees). In all cases, the employer itself need not have been
negligent or otherwise at fault.
Cleanup operations would clearly seem to be inherently danger-
ous activities because it is an activity which can only be safely car-
ried on by the exercise of special skill and care and which involves
a grave risk of serious harm if unskillfully and carelessly done. If
the nature of the work is classified as "ultrahazardous"—work
which necessarily involves a serious risk of harm to others which
cannot be eliminated by the exercise of the utmost care and which is
not a matter of common usage—then vicarious liability is not con-
tingent on proof of negligence on the part of the subcontractor
but is absolute.
Proximate Causation
Another key issue in determining liability for third party dam-
ages resulting from cleanup actions relates to causation. The law
requires" a showing of "proximate cause" as a precondition to the
attachment of liability.' Generators, other responsible parties such
as transporters, and their insurers will take the position that once
cleanups begins, the chain of causation is interrupted and any
"new" damages are the responsibility of the cleanup contractor
and involved government agencies. Conversely, in certain in-
stances, a cleanup contractor or government agency may want to
implead one or more of the originally responsible parties under
the theory that those parties are legally liable for all the adverse
consequences associated with the waste site, including those asso-
ciated with cleanup. The question is one of proximate causation.
The key legal question is whether the cleanup actions are to be
considered "intervening causes" which relieve the waste genera-
tors, site owners, or other originally responsible parties of liabil-
ity. Alternatively, these actions can be viewed as the natural and
probable consequences of the original wrongful act (i.e., the un-
controlled disposal of hazardous wastes). It is only when causes
are totally independent of each other that the nearest in time
and space becomes the proximate cause. The law recognizes that a
wrongful act can set in motion a chain of events leading to in-
jury. The fact that the actual consequence was one that rarely
follows from the particular act or omission does not afford a de-
fense; it is not necessary that the injury should be the usual, neces-
sary, or inevitable result of the negligence for proximate cause to
be found.
Natural phenomena (winds, rains) of a usual and ordinary kind
are not regarded as independent intervening agencies which will
break the chain of causation. A rainstorm somewhat greater than
usual is not so totally unforeseeable as to act as a superceding
cause.6
These issues have been most thoroughly explored in the law with
respect to personal injury cases. Where injuries have been aggra-
vated by medical or surgical treatment (i.e., a remedial response),
the original tort-feasor may be liable for damages resulting from
the treatment itself where the person injured (or a third party) has
used reasonable care in selecting the physician or surgeon. The neg-
ligence or malpractice of a physician or surgeon, selected with
reasonable care, who aggravates or fails to minimize injuries, is
regarded as a consequence reasonably to be anticipated. The law
regards the wrong of the one who caused the original injury as the
proximate cause of the damages and holds him liable.
Of course, negligent selection of a physician—or, analagously,
a cleanup contractor—acts as an intervening cause; the original
tort-feasor is not generally held liable for the consequences of such
negligence.
Thus, a wrongdoer—whether waste generator, transporter, or
cleanup contractor—is responsible for the natural and probable
consequences of a wrongful act or omission and this rule of law
applies both in contract and in tort. However, the application of
this rule to the circumstances of particular cases is often difficult
and each case will be decided largely on its own specific facts.
Normally, the question of "proximate cause" is one of fact to be
determined by a jury (or other trier of fact) although in some cases
the court may, as a matter of law, instruct the jury that damages
are too remote to be considered. Finally, the question of whether
resulting damage is direct and proximate does not depend on the
anticipation, knowledge, or lack of knowledge of the wrongdoer.
Because cleanup activities have potential third party liabilities, it
will be important for cleanup contractors to shift that risk either
through insurance or indemnification agreements, if possible.
SHIFT OF FINANCIAL RESPONSIBILITY
FOR LIABILITY CLAIMS
Given the potential liabilities associated with cleanup operations
at uncontrolled hazardous waste sites, cleanup contractors have an
incentive to reallocate financial responsibility through private
agreements. Two types of private agreements are of most relevance:
(1) indemnification agreements with the contracting authority (fed-
eral, state, or local governments) and with subcontractors, and (2)
insurance agreements with insurance companies.
Indemnification Agreements
Indemnification operates like insurance and is similarly created
by a contract. Commonly, an indemnification clause is used in a
contract as a means of reallocating risk; the indemnifier is not in
the regular or primary business of underwriting insurance but
agrees to compensate the other party for any loss or liability it
incurs. This is also known as a hold harmless agreement.
In drafting contracts for investigation or cleanup of uncon-
trolled hazardous waste sites, government agencies and private con-
tractors can agree to various provisions relating to liability. The op-
tions include the following:
•Contractor agree's to indemnify government
•Government agrees to indemnify contractor
•Contractor agrees to obtain insurance
Which option is selected will depend on the procurement pro-
cedures in effect and different states are likely to use different pro-
cedures.
Federal government procurement regulations have long required
contractual language requiring the contractor to procure and main-
tain insurance for ordinary risks such as workers compensation
and occupational disease insurance as required by state law, em-
ployer's liability insurance, comprehensive general liability insur-
ance, and automobile liability insurance (1 CFR§l-7.204-5). These
requirements were reflected in the contract provisions for WA 82-
HO71, the USEPA Hazardous Site Remedial Response RFP con-
tract.
USEPA's Article LI addresses insurance and liability to third
persons. It directs the contractor to obtain liability coverage as re-
quired by the Agency and allows reasonable costs of insurance as a
reimbursable contract expense. Subpart H adds that if required or
approved insurance coverage is reduced without the USEPA Con-
tracting Officer's approval, the liability of the government will not
be increased as a result.
Subpart E, however, is the crucial section of Article LI. It con-
stitutes a broad indemnification agreement whereby the Govern-
ment agrees to reimburse the contractor for liability in excess of
that compensated by required insurance. It reads as follows:
"The Government will hold harmless and indemnify the Con-
tractor against claims (including expenses of litigation or settle-
-------
LEGAL
468
ment) by third persons (including employees of the Contractor)
for death, bodily injury, or loss of or damage to property aris-
ing out of performance of this contract, to the extent that such a
claim is not compensated by insurance or otherwise. Any such
claim within deductible amounts of the Contractor's insurance
will not be covered under this article. Reimbursement for such
liabilities to third persons will not cover liabilities for which the
Contractor has failed to insure as required or to maintain insur-
ance as approved by the Contracting Officer."
This broad agreement applies even to the claims contractor em-
ployees may bring against the cleanup contractor itself—thus plac-
ing employees on a par with third parties as defined traditionally.
This is particularly good protection in view of the increasing
breaches of the exclusive remedy interpretation of workers com-
pensation statutes, discussed earlier. The clause does protect the
Government, however, against failure by the contractor to get and
maintain required insurance. Deductibles are not covered, nor any
losses compensated by insurance or otherwise. This indemnity is
analagous to a layer of "excess risk" insurance.
In contrast to the indemnification of the contractor by the fed-
eral USEPA, it has been reported that some state governments
require cleanup contractors to agree to fully or partial indemnify
the state from liability for third party claims.7 Language such as
the following would accomplish that purpose:
"In connection with the services to be furnished hereunder,
the Contractor agrees to be responsible for and to indemnify and
hold the Government harmless for any recovery of damages re-
sulting from death, personal injury, property damages, or other
loss incurred by the Contractor's Agents or employees, or third
persons, including within the latter group, Government em-
ployees, regardless of whether they are directly or indirectly asso-
ciated with the performance of services to be rendered pursuant
to the Terms of this Contract.''
Under such an arrangement, the contractor assumes financial
liability for damages the government may be liable for in addition
to its own underlying liabilities for damage claims. Thus, finan-
cial responsibility for cleanup liabilities can be shifted in this direc-
tion as well.
A prime contractor may also have indemnification agreements
with its subcontractors, with the burden to indemnify going in
either direction. There is an important limit on this, however.
Many state and federal statutes (e.g., Federal Employers' Liability
Act) expressly invalidate any contract which attempts to exempt an
employer from liability for injury negligently inflicted on an em-
ployee. According to a number of cases, agreements between em-
ployer and employee (or employer and subcontractor) attempting
to exonerate the employer from liability for future negligence
(whether of himself or of his employees) are void as against public
policy.
Types of Insurance Coverage
Insurance is a contractual agreement whereby an insurer agrees
to pay certain amounts to the insured if certain events occur. Busi-
nesses typically purchase insurance to protect themselves against
the financial impact of: (1) losses directly suffered (e.g., loss of a
pesticide plant in a fire), and (2) liability for losses suffered by
others (e.g., fire damage to neighboring properties). The latter is
termed liability insurance and applies to specific types of liabilities
up to specified dollar limits.
Insurance in the hazardous waste context has become very im-
portant this past year because of the RCRA financial responsibility
regulations. While those regulations do not apply to cleanup activ-
ities at uncontrolled hazardous waste sites, the Apr. 16, 1982 regu-
lations (effective July 16, 1982)8 have had an impact on the market
for pollution insurance or environmental impairment liability (EIL)
insurance.
Insurance coverage is available in a variety of forms to owners or
operators of hazardous waste facilities to cover liability due to
sudden and/or nonsudden accidents involving hazardous sub-
stances. Nonsudden refers to gradual pollution such as from boiler
or incinerator emissions, contamination of groundwater from
leachate, and similar situations. Depending on the specific fact
situation, a cleanup contractor may want to procure nonsudden
coverage. Sudden coverage is advisable, particularly where explo-
sions or fires or release of fumes or impoundments may occur.
There are several different types of insurance coverage that may
be triggered by damage resulting from cleanup operations:
•Comprehensive general liability (CGL) coverage
•Worker's compensation or employer's liability coverage
•Design and construction liability coverage
•Environmental impairment liability (EIL) coverage
Which type of coverage is involved will depend on the specific
type of damage and the parties damaged.
Comprehensive General Liability
Referred to as CGL, this policy form is designed to provide an
"all hazards" scope of protection, subject to certain exclusions
and conditions specified in the policy form. Nearly all businesses
carry this type of insurance. Since 1970, CGL insurance has con-
tained a Pollution Exclusion Clause which excludes coverage for
bodily injury or property damage arising out of the:
"discharge, dispersal, release or escape of smoke vapors, soot,
fumes, acids, alkalis, toxic chemicals, liquids or gases, waste
materials or other irritants, contaminants, or pollutants into or
upon land, the atmosphere or any water course or body of
water."
This exclusion does not apply, however, if the discharge, dispersal,
release or escape is sudden and accidental. Thus, companies with
CGL insurance are covered for all risks apart from non-sudden or
gradual environmental impairment. This latter exposure is rele-
vant only where the remedial action involves on-site land disposal
of wastes with the attendant risks of leachate contamination of
groundwater, etc.
Worker's Compensation Coverage
While discussion of the many different types of state worker's
compensation systems is beyond the scope of this paper, one im-
portant point must be made: a key premise of worker's compen-
sation has been that in exchange for the benefits of participating
in the system, covered workers gave up all other rights against the
employer for seeking compensation for injuries. This "exclusive
remedy" doctrine was upheld as constitutional in 1917 by the Su-
preme Court.' However, this doctrine is being eroded through a
series of court decisions finding employers liable on a number of
different theories: dual capacity, manufacturer liability, intentional
misconduct, etc. Thus, participation in a worker's compensation
program should not necessarily be viewed as adequate protection
against employee claims for compensation.
Employer's Liability Coverage
Employer's liability insurance complements workers compensa-
tion protection. There are two situations where employer's liabil-
ity coverage is commonly used: (1) for employees exempt from
workers' compensation, such as agricultural workers, and (2) in
states where workers compensation can be accepted or rejected.
Only liability for the death of or injury to employees is covered;
such a policy does not cover injury to employees of independent
contractors or to other third parties. The policy may cover the em-
ployer's liability at common law as well as his liability under work-
men's compensation acts.
In quite a few cases, the question has arisen as to whether an
occupational disease is within the scope of the coverage when the
policy does not specifically address this issue. Coverage depends on
whether the courts view the occupational disease as an "accident"
or not.10
-------
469
LEGAL
Design and Construction Defect Coverage
Liability policies are commonly issued to contractors insuring
them against liability for damages or injuries occurring in connec-
tion with the performance of their contracts. This type of coverage
is analogous to CGL coverage but does not necessarily provide pro-
tection against later claims based on design and construction de-
fects. Court decisions have generally depended on the presence or
absence of specific policy provisions dealing with such contingen-
cies."
Recently, construction-defect coverage has been made available
to design-build firms to protect them against liability for construc-
tion defects.11 Previously, liability coverage has been available only
for design defects, not construction defects. Common policies for
design professionals' liability totally exclude construction errors or
omissions. Because the construction defects liability coverage is
relatively new, it is not clear whether it will be provided for firms
doing remedial design and construction work at uncontrolled haz-
ardous waste sites.
About ten companies now offer policies to protect architectural
and engineering firms from professional liability claims. This mar-
ket first opened up in the late 1950s and shows signs of further ex-
pansion because of increasing court victories against state statutes
of limitation that were enacted in the 1960sz to protect design
professionals.
Environmental Impairment Liability
A cleanup contractor may want, or be required, to purchase
pollution liability insurance for financial protection against bodily
injury, property damage, and environmental impairment resulting
from the discharge, dispersal, release, escape, or seepage of toxic
substances into the environment. This type of insurance coverage
is designed to protect the insured from claims resulting from pollu-
tion or environmental damage losses. Typically, policies cover both
sudden and nonsudden events.
In order to cover the pollution exclusion clause of CGL cover-
age, EIL policies provide a very broad definition of environmental
damage or gradual pollution. The usefulness of EIL coverage for
nonsudden events is vitiated by the fact that currently only "claims
made" policies are available. Under this type of policy, coverage is
triggered only when claims are made during the policy period.
The period coverage may be expanded or restricted somewhat by
incorporation (i.e., purchase) of "Discovery Period" or "retro-
active Period" provisions. The purpose of claims made coverage
is to protect insurers from "trailing" liability for diseases which
manifest after lengthy induction and latency periods. Previous
types of insurance—called "occurrence-based"—have been held to
cover claims based on whether the causal event occurred during the
policy period; this coverage is not available for gradual pollution
occurrences at the present time. Eight firms presently offer pollu-
tion coverage, and the market is seen as a growing one.
CONCLUSION
Cleanup operations at uncontrolled hazardous waste sites pose
many risks, not the least of which are legal risks of being directly,
vicariously, or secondarily liable for the costs of damages result-
ing either during operations or, in the case of gradual damage,
many years later. Legal uncertainties abound. Firms which plan or
hope to have "deep pockets" should strive to protect their assets
through the purchase of appropriate insurance coverage or the
negotiation of protective contract provisions.
FOOTNOTES
1. Subrogation may be defined as the right of one person to stand in the
place of another with respect to legal rights. Conventional subroga-
tion arises through contract. The doctrine of subrogation is universal-
ly applied on behalf of a surety (e.g., insurer) who has been compelled
to pay on behalf of its principal (e.g., the insured).
2. See, generally. Survey of State Liability Provisions (ICF Incorpor-
ated, 1982).
3. See Grad et al.. Injuries and Damages From Hazardous Wastes—
Analysis and Improvement of Legal Remedies (1982), especially p. 81-
115.
4. See Restatement of Torts, 2d §427.
5. Corpus Juris Secundum provides the following definition of proxi-
mate cause: "(P]roximate causews any cause which in natural and con-
tinuous sequence, unbroken by any efficient intervening cause, pro-
duces the result complained of and without which the result would not
have occurred, and from which it ought to have been foreseen or rea-
sonably anticipated by a person of ordinary prudence in the exercise of
ordinary care that the injury complained of, or some similar injury,
would result therefrom as a natural and probable consequence." 65
C.J.S. pp. 1129-30.
6. See Southern Pacific Co. v. City of Los Angeles, 55 P.2d 847; Ely v.
Bottini, 3 Cal. Rptr. 756.
7. Sanders, "Liability for Remedial Cleanup Failures," Waste Age, June,
1982, p. 102-3.
8. See47 FR 16544 (Apr. 16, 1982).
9. New York Central Railroad v. White, 243 U.S. 188 (1917).
10. See Belleville Enameling and Stamping Co. v. United States Casualty
Co., 266 111. App. 586; United States Radium Corp. v. Globe Indent.
Co., 178 A. 271,aff'dl82A. 626.
11. See Koch v. Ocean Acci. and Guaranty Corp., 230 S.W. 2d 893 (later
damages covered); Berger Bros. Electric Motors v. New Amsterdam
Casualty Co., 58 N.E. 2d 717 (later damages not covered).
12. SeeENR, July 1, 1982, p. 88.
-------
NEGOTIATING SUPERFUND SETTLEMENT AGREEMENTS
LAUREN STILLER RIKLEEN, Esq.
U.S. EnvironmentalProtection Agency, Office of Regional Counsel
Boston, Massachusetts
INTRODUCTION
The Comprehensive Environmental Response, Compensation,
and Liability Act of 1980, (CERCLA),1 more popularly known
as "Superfund," represents a new approach in environmental
law. CERCLA does not provide a statutory framework for the
establishment of standards or permits to regulate industry as do
most other major pieces of environmental legislation passed in the
last decade. Rather, CERCLA provides EPA with broad authority
for achieving clean-up at hazardous waste sites and imposing lia-
bility for the costs on the responsible parties. Thus, through its
liability provisions, the statute provides an incentive for greater
diligence and concern in the handling and disposal of hazardous
substances as well as incentives for potentially responsible parties
to resolve their liabilities for past operations which resulted in en-
vironmental contamination through the negotiation of settlement
agreements.
Potentially, CERCLA provides a strong arsenal of legal theories
for the USEPA to use in its pursuit against parties responsible for
environmental contamination. At Superfund sites throughout the
country, the USEPA is pursuing its enforcement options, and in
many of these cases settlement negotiations are already under-
way. These negotiations must, of necessity, take place without the
benefit of Superfund case law as there has not yet been a full trial
of a Superfund case. This lack of precedent makes the negotia-
tion of Superfund settlement agreements particularly difficult.
The following paper outlines the general liability scheme of
CERCLA and discusses major issues likely to arise in the course
of settlement negotiations. In addition, this paper will also examine
how these issues were confronted and resolved in the resolution
of one party's liability for the Superfund site in Woburn, Massa-
chusetts.
LIABILITY PROVISIONS OF CERCLA
Sections 104,2 106,3 and 1074 of CERCLA provide the under-
pinnings for USEPA's enforcement scheme.
Section 104(a)(l) provides that the President is authorized to
act,5 consistent with the national contingency plan,' whenever:
(A) any hazardous substance' is released or there is a substantial
threat of such a release into the environment, or
(B) there is a release or substantial threat of release into the
environment of any pollutant or contaminant8 which may pre-
sent an imminent and substantial danger to the public health
or welfare.
The crucial exception to this authority to take action is if the
President determines that such action will be done properly by a
responsible party. Hence, federal remedial response is predicated
on the inability or unwillingness of the responsible parties to act.
Section 106(a) provides that, upon determining that there may
be an imminent and substantial endangerment due to an actual or
threatened release of a hazardous substance, the President may re-
quire the Attorney General to secure such relief as necessary to
abate the danger or threat. The President is also authorized to take
other actions, including the issuance of administrative orders, to
compel necessary action.
Section 107 sets forth four categories of parties responsible for
any incurrence of response costs consistent with the national con-
tingency plan, including all costs of removal or remedial action in-
curred by the United States Government. Broadly speaking, these
categories are:
•Owners and operators of a facility (or other source of release)
•Owners or operators of the facility at the time of disposal
•Generators or any other person who arranged for the disposal or
treatment of hazardous substances
•Transporters of hazardous substances to the site at which the
release occurred
Section 107(c)(3) of CERCLA allows the United States to seek
up to three times the costs incurred by the Fund if a liable party
fails, without sufficient cause, to comply with an order issued pur-
suant to sections 104 or 106 of CERCLA.
With certain limited exceptions, the functions vested by
CERCLA in the President were delegated to the Administrator
of USEPA by Executive Order 12316 of August 14, 1981.'
INITIATING THE SUPERFUND ENFORCEMENT PROCESS
Consistent with the intent of Section 104, USEPA policy10 is
that USEPA should attempt to secure cleanup by the responsible
parties, rather than the Trust Fund established under CERCLA,"
whenever such clean-up can be accomplished properly and in a
timely manner. Therefore, prior to the expenditure of govern-
ment funds at a Superfund site, notice will generally be sent to
responsible parties.l2
However, that it is the position of USEPA that a notice let-
ter is not a legal pre-condition for a subsequent lawsuit by EPA to
seek cost recovery against a responsible party if Fund money is
spent." There may be a number of reasons why a notice letter does
not issue prior to the expenditure of Fund money. For example,
USEPA may be involved in on-going negotiations with a respon-
sible party or there may have been no evidence which linked the
responsible party to the site known to USEPA at the time Fund
money was spent.
Implicit in the notice letter process is the concept that such a
letter will set into motion an opportunity to resolve informally the
problems at the site, thereby eliminating the need to initiate court
action or to issue a unilateral administrative order; where an agree-
ment is reached prior to litigation, the agreement will be reduced
to writing via an administrative order on consent. Hence, in most
cases where responsible parties are identified, former legal action
at a site will be preceded by a period of time in which informal
settlement negotiations will occur. The mere existence, however,
of a responsible party will not prevent the agency from moving for-
ward with the expenditure of Fund money or formal legal action
if such responsible party is not willing to address the problems at
the site in a proper and timely manner."1
470
-------
471
LEGAL
ADVANTAGES OF SETTLEMENT
Both the government and the responsible parties have much to
gain if the problems at a site can be resolved through negotiations.
From the government's perspective, the statutory objective is
achieved if a responsible party undertakes the clean-up obligations.
Fund money is preserved for those sites which will not be addressed
by a responsible party. In addition, because of the intricacies of
federal contracting and the CERCLA statutory requirements rela-
tive to state participation, it is likely that the site will be cleaned
up more quickly if the responsible party agrees to do so voluntarily.
Finally, while negotiations themselves can be very resource-inten-
sive for USEPA, resolution of a case through settlement will in-
variably take much less time and resources than if the case were
actually litigated.
There are also numerous advantages to settlement for the respon-
sible party. CERCLA provides USEPA with a broad delegation of
authority as to what are appropriate actions for which the govern-
ment can recover costs. In addition, the potential exists for the re-
covery of triple the actual costs expended by virtue of the punitive
damages section. Further, the potential application of a strict liabil-
ity standard places a significantly greater burden on the respon-
sible party in actual litigation. A responsible party avoids these
problems by settling, and gains the additional benefit of partici-
pating in the decisions and final terms relative to the scope of its
involvement in the resolution of the problems at the site.
Another concern to a responsible party at a Superfund site is
cost; a party considering a settlement must balance the projected
costs of a negotiated cleanup against the substantial legal fees
necessary to fight, and possibly lose, a law suit. Even if a respon-
sible party thinks his legal defenses to a law suit will prevail in a
court of law, it is necessary to take into account the enormous
costs of litigating hazardous wastes cases due to the substantial
time needed to prepare legal arguments and expert witnesses. The
limited number of hazardous waste cases that have actually been
litigated to date indicate that such trials can take months."
Finally, a responsible party, particularly in the chemical indus-
try, may find there are public relations advantages to settling.
Rather than being portrayed as an irresponsible party that caused a
pollution problem and refused to help clean it up, a corporation
has the opportunity to be a "good corporate citizen" that volun-
teers to rectify an environmental problem.
MAJOR ISSUES IN SETTLEMENT NEGOTIATIONS
The Woburn, Massachusetts Settlement
On May 25, 1982, USEPA, pursuant to the statutory authority
of section 106 of CERCLA, issued an order on consent to the
Stauffer Chemical Company for the investigative study, cleanup,
and future monitoring of the Superfund site in Woburn, Massa-
chusetts. This site was listed by the Administrator of USEPA as
one of the top ten Superfund sites in the country." The Common-
wealth of Massachusetts, through its Department of Environ-
mental Quality Engineering (DEQE), participated in the settlement
negotiations and was a party to the agreement. The settlement
marked the first use of CERCLA's section 106 administrative
order authority at a listed Superfund site.
The settlement negotiations at this site presented the regula-
tory agencies with a number of difficult problems that had to be
resolved before the parties could enter into a consent order. Many
of the issues confronted are ones likely to arise in the context of
most settlement negotiations at other sites. Successful resolution
of these issues requires a careful understanding and balance of
various USEPA policies relevant to the implementation of
CERCLA, the technical needs of the site, the motivations and in-
terests of the negotiating responsible party, and the concerns of the
affected community.
Joint and Several Liability
Joint and sc\eral liability is a legal theory which, if applicable
in a gi\en fact situation, can result in anv one of a number of
responsible parties being held responsible for all of the damages.
The issue of joint and several liability only arises where the acts of
more than one party produce a single harm."
In a hazardous waste context, a basic example of joint and sev-
eral liability is the multiple generator situation where a number of
companies have shipped barrels of a single type of hazardous
waste to a site, little or no evidence exists relative to the amounts
of the substance shipped by each company, and all of the barrels
leak into and result in the contamination of the groundwater.
Under these circumstances, each of the generators may be held
responsible on the theory that the harm to the groundwater
caused by one of the generators is indistinguishable from the harm
caused by any of the others; therefore one, or all, may be held
fully liable. The concept is best summarized as follows:
Where the tortious acts of two or more wrongdoers join to pro-
duce an indivisible injury, that is, an injury which from its na-
ture cannot be apportioned with reasonable certainty to the in-
dividual wrongdoers, all of the wrongdoers will be held jointly
and severally liable for the entire damages and the injured party
may proceed to judgment against any one separately or against
all in one suit."
How joint and several liability will be applied by a court in the
Superfund context is not known at this time. While USEPA and
the Department of Justice will argue the applicability of joint and
several liability where the facts warrant," industry strongly main-
tains that the theory is not applicable in a CERCLA lawsuit. Ac-
tual resolution of the controversy will not occur until the issue has
been fully litigated, and even then case law may vary depending on
the particular fact situation before the court.
As a practical matter, most successful negotiations will result in
some type of apportionment. In most cases there would be little
economic incentive for a responsible party to be willing to cooper-
ate with the federal government unless some apportionment of
liability is recognized. USEPA's willingness, therefore, to appor-
tion liability in a negotiated agreement where the facts are appro-
priate should not be viewed as a rejection of the applicability of
the joint and several liability theory to the site at hand, but
rather a practical recognition that the apportioning of liability is
an important aspect of encouraging settlements.
Apportionment of liability was a major issue in the Woburn
Consent Order. Stauffer and companies it acquired had manu-
factured glue at the site; Stauffer was willing to assume respon-
sibility for those wastes generated by the flue manufacturing pro-
cess. Other wastes found at the site resulted from the manufac-
turing activities of the other known responsible party. Some of the
wastes from all prior operations had been commingled, due to
the land development activities of one of the present property
owners.
The apportionment issue was resolved as follows: Stauffer
agreed to undertake the full investigative study and to participate
in the cleanup of Stauffer-generated wastes. After the conclusion
of the investigative study, USEPA and the state will determine
the proportionate responsibility, pursuant to a formula set forth in
the agreement, for the contamination at the site which will then
be the basis for apportioning the costs incurred in implementing the
investigative study. Stauffer may then choose whatever legal means
it deems appropriate to recover those funds it expended on the
study beyond its proportionate share. In addition, because the
apportionment determination will be made prior to the implemen-
tation of the clean-up, Stauffer's participation in the clean-up
will be limited to its determined proportionate responsibility.
Handling Indefinite Commitments by Agreement
A comprehensive settlement agreement requires fully address-
ing at the outset the scope of the investigative study, the cleanup
and subsequent monitoring commitments, and the release pro-
vision. However, these commitments are being made at a time
when there is frequently little information available to quantify the
extent of the problem being addressed or the adequacy of the in-
tended actions.
-------
LEGAL
472
Because of the difficulties such a situation imposes, a respon-
sible party may only be willing to negotiate its involvement in
the investigative study, leaving its cleanup liability for resolution
at a later date. Even if USEPA were to be responsive to a pro-
posal for a partial settlement, the negotiating party would obvious-
ly not be able to benefit from a release provision relative to its
cleanup liability, thereby leaving unresolved a major aspect of its
liability and a major motivating factor for entering into negotia-
tions at the outset. In addition, the resolution of the total prob-
lem remains unresolved. It seems, therefore, to be in the best in-
terests of the responsible party and USEPA to enter into compre-
hensive settlements which address all aspects of the site, rather
than piece-meal agreements in which only portions of the problem
are resolved at a time.
The Woburn agreement provides an illustration of how indef-
inite commitments were made relative to future actions and ex-
penditures unknown at the time of negotiations. In general, a
"checks and balance" system was developed; adequate controls
were built into each commitment made by the parties, such that
each party had a self-interest in fulfilling its responsibilities. For
example, in regard to the investigative study commitments,
Stauffer agreed to a phased approach: phase I of the study is
specific in its commitments, thereby allowing Stauffer the oppor-
tunity to project its costs for this phase of the investigation. Phase
II is a more open-ended commitment which provides for additional
study as needed pursuant to stated criteria. This second phase
provides the flexibility needed to insure that the site is adequately
studied.
Following the investigative study, Stauffer will propose a recom-
mended remedial action which EPA and the state may accept or
reject; a period of time for negotiation is provided. If accepted,
Stauffer is obligated to participate as directed in accordance with
its determined proportionate share. If the proposal is rejected,
Stauffer has no further obligations under the agreement, now,
however, is Stauffer released from liability. Thus, both parties have
a strong interest in seeing that an acceptable remedial action is
implemented: USEPA wants the site cleaned up; Stauffer wants its
release from liability.
Releases
The release provision is an issue of great concern to a party en-
tering settlement negotiations. Most responsible parties enter nego-
tiations with the expectation that, if the case is successfully re-
solved through settlement, their potential liability at that site will
be ended.
Although the scope of the release will vary with each fact sit-
uation, there are certain general principles which the negotiating
parties should recognize at the outset:
•The releases run to the settlement of civil claims only; the USEPA
cannot settle criminal liability.20
•The scope of the release should be commensurate with the scope
of the cleanup; a total release should not be granted if the party
is not undertaking its total share of the study and cleanup.21
•The release should be conditioned upon the timely and satisfac-
tory completion of all of the party's obligations under the
agreement.22
•USEPA cannot bind other federal agencies. Therefore, USEPA
should not attempt to bind the "United States" or waive claims
which may be asserted by other federal agencies.23
•USEPA should be very careful to protect its future rights against
other responsible parties who are not participating in the settle-
ment.24
Stauffer's release provision was conditioned on Stauffer fulfill-
ing all the commitments made in the Consent Order. Specifically,
one of the issues raised was the statutes under which Stauffer's
release would be effective. USEPA's position was that the releases
would correspond to the jurisdictional authority of the consent
order. Therefore, since the consent order was issued pursuant to
CERCLA and Section 7003 of the Resource Conservation and
Recovery Act (RCRA),2' and since Stauffer agreed to assume full
responsibility for its own wastes, the release stated that fulfillment
of Stauffer's commitments in the agreement constituted full satis-
faction of USEPA's civil claims pursuant to CERCLA and RCRA.
To protect the rights of the parties entering into the consent
order against other, nonsettling parties, a provision was inserted
in the agreement, consistent with Massachusetts law, clearly stat-
ing that the release to Stauffer in no way affects the liability of
any other responsible party.
Citizen Participation
The presence of a hazardous waste site can cause significant fear
and concern in the local community. Particularly frightening for
the affected citizens are the unknown health impacts such a site
may have. It is, therefore, important to provide as much knowl-
edge as possible to the local community and keep the citizenry
informed regarding the activities occurring at the site.
The National Contingency Plan recognizes the affected locality
by stating that response personnel should be "sensitive to local
community concerns in accordance with applicable guidance."26
This includes the establishment of an effective community relations
program at each site. The goal of this program is to set forth the
various ways in which the agency in charge of the response plans
will communicate with the citizens through public forums, press re-
leases, and meetings.
When the negotiations with Stauffer began in Woburn, a local
Citizens Advisory Committee (CAC) was already in existence and
was meeting twice a month to participate in and track develop-
ments relative to the hazardous waste site. State and federal regula-
tory officials attended most meetings of the CAC. During the
lengthy period of time the group had been meeting, the mem-
bers developed a comprehensive understanding of all issues sur-
rounding the site.
Throughout the course of the negotiations with Stauffer the
regulatory agencies made a substantial effort to keep the CAC in-
formed and to seek their input. Stauffer was very cooperative in
this endeavor. For example, prior to presenting an investigative
study proposal, representatives of Stauffer attended a meeting of
the CAC to hear the specific concerns of the citizens. They also
attended other meetings of the CAC to present their study pro-
posal and to accept comments. The citizens brought forward a very
reasoned approach and the company was willing to listen to and
address their concerns. By maintaining regular communication,
the actual issuance of the consent order was not viewed by the
community with suspicion; rather, it was viewed as an opportun-
ity to finally see the problems of the site addressed.
Federal-State Relations
Most states have their own panoply of environmental laws which
are often similar to their federal counterparts, although few states
have a statute with the breadth of authority as CERCLA. There-
fore, in addition to its federal liability, a responsible party may
face the risk of being the recipient of a state enforcement ac-
tion pursuant to state law. For this reason, it is in the best interests
of the responsible party to attempt to resolve its liability with
both USEPA and the state in which the hazardous waste site is
located. USEPA encourages the involvement of the state in the
negotiation of these agreements.
The Woburn negotiations were marked by a close working re-
lationship between USEPA and the state's DEQE. The result was
the issuance of one document to which EPA, the state, and Stauf-
fer were parties. This resulted in the most efficient use of the
agencies' resources. In addition, the fact that there is only one doc-
ument offers greater assurance that consistency and uniformity in
the implementation of the consent order will result.
CONCLUSIONS
Resolving problems at Superfund sites through negotiated agree-
ments is a difficult and complex undertaking. However, there
-------
473
LEGAL
are significant benefits to both the responsible parties and the
USEPA by settling these cases. Upon entering negotiations, the
parties should be willing to consider all aspects of the issues in
controversy and to develop fair and reasonable solutions that
accommodate each other's concerns without sacrificing the tech-
nical needs of the site. In addition, the concerns of the local com-
munity must be addressed in a meaningful way. Approached this
way, significant headway will be made in the clean-up of the na-
tion's many hazardous waste sites.
REFERENCES
1. 42 U.S.C. §960letseq.
2. 43 U.S.C. §9604.
3. 42 U.S.A. §9606.
4. 42 U.S.C. §9607.
5. Such acts include authorization "to remove or arrange for the removal
of, and provide for remedial action relating to such hazardous sub-
stance, pollutant or contaminant at any time...or take any other re-
sponse measure consistent with the national contingency plan which
the President deems necessary to protect the public health or welfare
or the environment..."
6. Section 105, 42 U.S.C. §9605, requires the publication of a Na-
tional Hazardous Substance Response Plan to establish procedures and
standards for responding to releases of hazardous substances, pollu-
tants, and contaminants. This document was recently published at 47
Fed. Reg. 31180 (July 16, 1082).
7. Hazardous substances are defined in section 101(14) of CERCLA,
42 U.S.C. §9601(14).
8. "Pollutant or contaminant" is defined in section 104(a)(2) of
cercla, 42 U.S.C. §9604(a)(2).
9. "Responses to Environmental Damage," Executive Order, No. 12316,
46 Fed. Reg. 42237 (August 20, 1981).
10. February 23, 1982 Memorandum from Christopher J. Capper, Act-
ing Assistant Administrator for Office of Solid Waste and Emergen-
cy Response, and William A. Sullivan, Jr., Enforcement Coun-
sel, to the Regional Administrators, entitled: "Hazardous Waste
Compliance and Enforcement Program Guidance," at 5.
11. 42 U.S.C. §9631 establishes a 1.6 billion dollar Hazardous Sub-
stance Response Trust Fund for federally-financed responses at
designated sites.
12. supra, note 10at 5.
13. November 25, 1981 Memorandum from Sullivan and Capper to Reg-
ional Administrators, et al., entitled: "Coordination of Superfund
Enforcement and Fund-Financed Clean-Up Activities," at 10.
14. supra, note 10 at 7.
15. e.g., Village of Wilsonville, III. v. SCA Service, Inc., No. 52885
(111. Sup. Ct., May 22, 1981), 2 CHEMICAL and RADIATION
WASTE LIT. REP. 288 (1981)—length of trial was over 100 days;
State Department of Environmental Protection v. Ventron Corp.,
Nos. A-1395-79,-1432-79,-1446-79,-1545-79 (N.J. Super. Ct. App.
Div., Dec. 9, 1981)—length of trial was 55 days.
16. On October 23, 1981, the Administrator'of USEPA released an in-
terim national priority list of 115 sites, pursuant to Section 105(8)(B)
of CERCLA, 42 U.S.C. §9605(8)(B).
17. 74 Am. Jur. 2d §62.
18. Landers v. East Texas Salt Water Disposal Co., 248 S.W. 2d 731
(Tex. S.Ct., 1952) at 734.
19. There is significant legislative history in CERCLA which provides a
basis for USEPA's position that joint and several liability was deleted
from the legislation in order to leave the issue to resolution by the
courts pursuant to the common law. For example, see: Senator Ran-
dolph, Chairman, Senate Environmental and Public Works Com-
mittee and Floor Sponsor, Congressional Record, November 24,
1980 (S. 14964); Congressman Florio, Chairman of House Subcom-
mittee on Transportation of House Commerce Committee and Floor
Sponsor, Congressional Record, December 3, 1980 (H. 11787).
20. December 18, 1981 Memorandum from Sullivan to Regional Coun-
sels, et al., entitled: "Guidance on Hazardous Waste Site Settlement
Negotiations," at 4.
21. id.
22. id.
23. id.
24. id.
25. 42 U.S.C. §6973.
26. 47 Fed. Reg. 31214, (July 16, 1982).
-------
HAZARDOUS WASTE AND THE REAL ESTATE TRANSACTION:
A PRACTICAL AND THEORETICAL GUIDE FOR THE
TECHNICAL CONSULTANT, REAL ESTATE ATTORNEY,
BUSINESS PERSON, INVESTOR, OR ANYONE INVOLVED IN
BUYING AND SELLING LAND
JEFFREY T. LAWSON
Environmental Research & Technology, Inc.
Concord, Massachusetts
BARBARA H. CANE, Esq.
Brown, Rudnick, Freed & Gesmer
Boston, Massachusetts
INTRODUCTION
Statutory schemes and common law tort liability expose past
and present owners of land containing hazardous waste materials
to enormous potential liability. Consequently, all those who deal
with real property transactions need to be apprised of the nature
and scope of the problem. A combined technical and legal ap-
proach addresses these questions: Where does the law place lia-
bility? What are the "red flags" of a potential problem? How,
and at what cost, is technical evaluation of a site made? What
types of remedial action are available? How will a problem affect
the value and uses of the land? Is insurance a solution?
SOURCES OF LIABILITY: THE SCOPE OF THE PROBLEM
Like termites or zoning laws, hazardous wastes are a fact of life
which may profoundly affect the value of real property and ex-
pose parties involved in a real estate transaction of significant lia-
bility. As a result, all those who deal with real property trans-
actions must be apprised of the nature and scope of the problem.
Legal liability for hazardous waste is imposed upon past and
present owners of real property by federal statutes and regula-
tions, state statutes and regulations and common law principles.
The major sources of federal liability are found in (1) the Com-
prehensive Environmental Response,1 Compensation, and Liability
Act of 1980 ("Superfund"), 42 USC Sec. 9601 et seq., which
deals with cleanup and containment of hazardous wastes which
have been disposed of accidentally; (2) the Resource Conserva-
tion and Recovery Act ("RCRA"), 42 USC Sec. 6901 et seq.,
which regulates the generation, treatment, storage, disposal and
transportation of hazardous wastes; and (3) the Toxic Wastes
Substances Control Act ("TOSCA"), 15 USC Sec. 2601 et seq.,
which regulated management of toxic substances, including PCBs
(which until recently were commonly used in the oil contained in
transformers and capacitors).
States have enacted legislation and promulgated regulations spe-
cifically directed at hazardous wastes. For example, Massachu-
setts has enacted the Waste Management Act, M.G.L.c. 21C, and
the Massachusetts Hazardous Waste Regulations of July 1, 1982,
310 CMR 30.000. Other state laws directed at protecting various
parts of the environment such as waterways or wetlands, may
also effectively impose liability for improper handling of haz-
ardous wastes. By federal statute, whenever federal and state
laws diverge, the stricter standard controls.
Common law theories, including nuisance, trespass, negligence,
and strict liability may be the basis of public or private actions
for damages or injunctive relief from harm stemming from haz-
ardous wastes.
The sources of liability are extensively discussed elsewhere.
The salient point for the purposes of this article is that statutory
liability for damages and remedial efforts rests with both past and
present owners and operators of facilities where there has been im-
proper generation, treatment, storage or transportation of haz-
ardous wastes. Statutory liability is joint and several.
Thus, whether an enforcement agency compels a cleanup ac-
tion, or seeks reimbursement for costs of an agency conducted
effort, it may seek satisfaction from past and present owners and
operators in any order it deems appropriate. The agency need not
exhaust the resources of one party before proceeding against the
other. While not absolutely clear, it also appears likely that any
liability predicted on common law theories would also be joint and
several, so that an injured plaintiff could proceed against any one
of a number of possible defendants.
Cleanup efforts can be enormously expensive. For example,
typical costs for excavating, transporting and disposing of con-
taminated soil are on the order of $200 per ton. Other remedial
measures, such as capping, may provide a less expensive alternative
in some instances, but the cost is still great. Remedial measures
are usually developed in consultation with enforcement authori-
ties; in some cases they may be sufficiently broad as to include re-
location of parties endangered by contamination.
Joint and several liability means that while it is easy to buy a
hazardous waste problem, it is extremely difficult to sell one: a con-
veyance out does not relieve the seller of potential liability. This
being so, sale of contaminated land to an irresponsible buyer who
improperly disposes of waste can result in liability that will come
back to haunt the seller. The chain of liability can be extended back
as far as the first party responsible for introducing the hazardous
waste onto the land.
At the present time, hazardous waste insurance is of only lim-
ited utility to property owners and does not eliminate or signifi-
cantly reduce the problem of potential liability. Policies typically
cover off-site accidental personal and property damage. On-site
prophylactic cleanup may be covered upon the insurer's prior
written consent when there is an imminent threat of a pollution in-
cident. Exclusions are numerous. Bodily injury, property damage
or environmental damage which is expected or intended from the
standpoint of the insured is not covered; nor are damages result-
ing directly or indirectly from failure to comply with all applicable
laws. Premiums are very high, and before a policy is written, an
extensive site inspection is conducted.
Thus, the nature of potential liability forces all prudent parties
to a real estate transaction to concern themselves with hazardous
wastes and their proper management. Knowledge, together with a
well-conceived plan for compliance with the law, is the best strat-
egy for avoiding or mitigating liability. The technical expert is cru-
cial to achievement of this end.
THE ROLE OF THE ENVIRONMENTAL PROFESSIONAL
An environmental professional provides necessary information
to the parties involved in the real property transaction based upon
his or her training and experience, and the results of a site-specific
environmental liability assessment. The purpose of professional
consultation is to avoid the need for guarding against or negotia-
ting around the unknown. On industrial/commercial sites the un-
known can be potentially hazardous contamination of air, build-
474
-------
475
LEGAL
ings, soil, groundwater or surface water. Other potentially restric-
tive conditions are proximity to wetlands, coastal zones or drink-
ing water supplies. It is also possible that the site is already sub-
ject to regulatory compliance. The information gathered and in-
terpreted by the technical professional provides a greater level of
confidence for making decisions and adopting negotiating postures
in the transaction. Technical information can provide the basis for
specific language in legal documents, and can help define the
characteristics of an industrial/commercial site that require de-
tailed legal attention. Most important, the informed party gains
assurance that "surprises" do not exist, and consequently he or
she can proceed without fear of the unknown.
THE TECHNICAL APPROACH TO ENVIRONMENTAL
LIABILITY ASSESSMENT
A phased or incremental approach to technical assessment of
environmental liability provides the greatest amount of specifically
directed information in the most cost effective manner. The phased
approach consists of gathering and interpreting information in dis-
crete steps. The information from one step helps determine whether
the next phase is necessary, and forms the basis for planning and
executing the next phase of investigation in an efficient manner.
Phase I
The first phase of a technical environmental liability assessment
may not, at the outset, require the services of an environmental
professional. An informed party engaged in a real property trans-
action can start the process by running through a check list such as
that set forth in Table 1. In this table, typical, but by no means
comprehensive, examples of conditions commonly encountered on
active or inactive commercial/industrial properties which are "red
flags" of potential environmental liability are provided. Should
these or similar conditions exist at the site under consideration,
it is prudent to seek professional assessment of the condition.
These conditions may not indicate actual liability, but still may re-
quire consideration as the sale is negotiated. The better the site con-
ditions are defined, the better informed the negotiators will be.
If no "red flags" are evident, it should not be taken as an indica-
tion that the site is problem-free. A site may be covered with tall
grass and bordered by sturdy trees; rabbits may bound across it,
and hawks may wheel overhead, but it is not impossible that sev-
eral decades' worth of industrial waste sludge lies buried below the
grass and wildflowers. In the scene just described, past disposal
practices and the resulting liability may only come to light through
investigation of the site history. Detailed review of the site history is
the most important part of the initial site assessment process,
and should be among the first activities pursued.
"Red flags" of environmental liability are not peculiar to sites
of heavy industrial activity. The careful party should go through
the initial check list for seemingly benign structures, such as apart-
ment houses that might have old transformers in the basement, or
a dwelling where overly liberal use of a pesticide like chlordane
may have resulted in potentially hazardous accumulations.
The alert party should be aware of site conditions, and site his-
tory. He or she should be suspicious. Review of noted site con-
ditions with an environmental professional at this point may serve
to confirm his or her own observations and conclusions, or to
raise further relevant questions.
Phase II and Beyond
Beyond the do-it-yourself check-list phase of environmental
assessment, a technical specialist is required. The technical con-
sultant is best suited by training and experience to employ appro-
priate investigative techniques, interpret data, and formulate
recommendations which will most effectively satisfy the client's
needs. The environmental professional's involvement starts with a
thorough discussion of the client's requirements so that an appro-
priate level of investigation can be designed and executed. The
level of effort is governed primarily by the value of the transaction
and magnitude of potential liability.
Table 1.
Cheek Lfet »f "Red Flags" of Potential Environmental Liability.
This list provides a summary of conditions sometimes found on active
or inactive industrial/commercial property. These conditions can be "red
flags" or strong indications of potential liability.
•Odorous or visible paniculate air emissions
•Use, storage or transport of known hazardous materials
•Waste disposal areas: this includes disposal of all types of waste in end
dumped piles, lagoons, barrels, swamps, wells, etc.
•Leaking pipes, electrical gear, containers, tanks, barrels, or stockpiles
•"Stories" of accidents, spills, explosions or dumping; complaints by
neighbors
•Fouled surface water standing on-site or flowing off-site
•Odorous or turbid water from a well
•Odorous or stained soil
•Proximity to flood plains, wetlands, coastal zones, or surface-water
bodies
•Proximity to an existing or planned public or private drinking water sup-
ply; this includes surface water and groundwater sources
•Regulatory compliance status
Generally, the technical investigation starts with gathering and
reviewing all available pertinent information about the site history
and conditions, including published information regarding soil,
groundwater, surface water, and air quality conditions, aerial
photographs, past engineering and construction reports, and oral
information from the client and client's attorney. Published com-
pany histories which chronicle activities at the site may be partic-
ularly helpful.
Armed with this information* the technical consultant under-
takes a first-hand inspection of the property under consideration.
The investigator, a trained and experienced observer, uses a de-
tailed check-list approach somewhat similar to that undertaken in
Phase I. This directed reconnaissance phase may also include pre-
liminary air, soil, groundwater, surface water or materials
sampling. Further analyses may be ordered based on the likelihood
of the presence of particular constituents.
Based on analysis and interpretation of the information gained
from review of existing information and the site reconnaissance,
the technical consultant determines whether further investigation
is necessary, and if so, what activities should be pursued. A re-
view of case histories that are representative of varying poten-
tial liability and transaction values is most instructive of what may
lie beyond the do-it-yourself check-list and initial technical recon-
naissance.
CASE HISTORIES
The following case histories represent orders of magnitude dif-
ferences in liability, value of transaction, and cost of investiga-
tion. The first case history involves a straightforward low-level in-
vestigation of a site undertaken at the outset of a sales transaction.
The second case history describes a "late-in-the-game" investiga-
tion which was undertaken too late to prevent significant losses.
The last case history demonstrates the severe penalties of not hav-
ing undertaken an environmental assessment prior to sale of the
property.
Case History 1
The first case history involves the assessment of a 200 acre quarry
and bituminous concrete plant on behalf of a potential purchaser.
The purchaser's attorney provided site history information, aerial
photographs and appropriate details on the proposed use of the site
as they influenced the extent and nature of environmental liability
assessment. Public agencies were contacted for information on ex-
isting or planned downgradient drinking water supplies. Further
topographic, geologic and hydrologic information was obtained to
determine the most likely conduits for potential off-site migration
of contaminants.
An on-site inspection was conducted that consisted of a
thorough on-foot traverse of the entire site in the company of the
quarry manager. In this manner, the environmental professional
-------
LEGAL
476
made a first-hand inspection of the land, surface water bodies,
buildings, equipment, and processes while he interviewed a long-
time employee. The plant manager was able to explain the current
operations and past practices. He could also recall past problems
and complaints. The reconnaissance and interview provided: (1) a
confirmation and an augmentation of the site history, (2) an inven-
tory of past and present operations, (3) an inventory of hazardous
materials handling, use and disposal practices, (4) waste handling
and disposal practices, (5) indication of gross air and surface
water quality, (6) areas of potential concern, and (7) a basis for
recommendations of further action.
By discussing the findings of the reconnaissance, the buyer's
attorney and the technical consultant decided that it would be pru-
dent to analyze the groundwater from an on-site well that is used
for process water. The analysis was recommended because several
thousand cubic yards of waste bituminous-concrete were piled on-
site. Fuel oil, asphalt and solvents were also stored and used. Al-
though there were no indications of spills or disposal of hazardous
constituents evident at the surface, testing the groundwater pro-
vided an inexpensive way to investigate for past spills or disposal.
Limited analyses for solvent and petroleum constituents did not re-
veal any contamination.
This level of technical investigation did not indicate any source
of environmental liability. Based on the technical investigation,
the attorney was able to include specific contingencies in the pur-
chase and sale agreement. The cost of the investigation, includ-
ing meetings, analytical fees and a formal written report, was
$2,000.
Case History 2
This case history illustrates the possible penalties for not seek-
ing qualified assistance in a timely manner. In this case the client
was engaged in the purchase of a 60 year old, 15,000 ft2 electro-
plating facility that was to be converted to office and warehouse
space. The client had pursued standard purchasing procedures, in-
cluding a structural inspection of the building. A purchase and sale
agreement had been signed and a $20,000 deposit had been made.
Within one week of the scheduled closing, an alert employee of
the purchaser called attention to the presence of sludge in the crawl
space under the building. A technical consultant was hastily re-
tained and a fast paced investigation was undertaken.
In order to determine whether or not the sludge constituted a
hazardous waste, the following steps were taken: sampling and
analyzing the process materials and waste, determining the regula-
tory compliance status of the current waste disposal practices,
sampling and analyzing (by EP toxicity and mass analysis methods)
the sludge and underlying soil for cyanide and selected metals in-
cluded in the primary drinking water standards, and sampling
and analyzing the groundwater from an on-site well that was used
for process water. Also, the reactivity of the sludge was analyzed
by subjecting the sludge to acid and analyzing for the release of
hydrogen cyanide gas.
The analyses revealed that the leachate from the EP toxicity
tests on the sludge were enriched in the metals analyzed for, but
that the concentrations in all cases were below the hazardous waste
characterization criteria. The sludge did, however, satisfy the haz-
ardous waste characterization criteria of reactivity by virtue of its
evolving hydrogen cyanide gas in the presence of acid. The well
water did not manifest any enrichment of the constituents analyzed
for.
The results of this "late-in-the-game" investigation were com-
plex, and the resulting costs were high. The purchaser backed
out of the transaction and risked loss of the $20,000 deposit be-
cause there was no pertinent contingency clause in the purchase and
sale agreement. The owner subsequently contracted for removal
and disposal of the sludge for an approximate cost of $50,000. The
fee for the investigation, analyses, and report was approximately
$5,000.
Case History 3
This case history is unusual because of the magnitude of costs
involved. It is worth discussion, however, because it emphasizes
the need for investigating and confronting environmental liabil-
ities before conveying land. The" site is a 30-acre grassy field border-
ed by trees. This large open expanse on a river bank in a residen-
tial/industrial neighborhood appeared to be an attractive site for a
municipal park, and the owner donated the land to a municipal
agency for that purpose. Shortly after donation of the property,
construction of the park facilities began.
Construction of the park included earthwork for foot paths,
roads, and parking areas. Trenches were excavated for utilities
and sewage, and several structures were erected. During the ex-
cavation and earthwork, waste disposal trenches were breeched
that contained sludges, slag, minor amounts of off-specification
pesticides and other industrial wastes. Construction workers com-
plained of various systems while they were working on the site.
The municipal agency retained a consultant to investigate. His
limited investigation revealed the severe conditions at the site which
included, among other characteristics, the presence and migration
of various toxic constituents of the buried sludges. He also iden-
tified pesticides in the waste. As a result of this investigation, the
former owner purchased the site back for an amount equal to the
cost of park construction, which by that time amounted to approx-
imately $3,000,000.
The owner then hired its own consultant and undertook an ex-
tensive site investigation. The investigation started with a detailed
reconstruction of the site history from interviews with employees
of companies which had disposed of waste on site, published and
unpublished documents, and aerial photographs. Review of the
aerial photographs was critical. Their coverage ranged back to the
1940s when the site was a picturesque asparagus farm. Photo-
graphs from ensuing years allowed compilation of waste-type and
waste-pit location maps. Initial review of existing information and
waste-history mapping formed the basis for planning, and gave spe-
cific direction to the field investigation.
The field investigation included installation of 26 groundwater
monitoring wells, 46 gas driven groundwater samplers, 3 real time
groundwater level sensors and transmitting gear, drilling and log-
ging 74 borings, and excavating and logging 26 test pits. Fifty soil
or waste samples and approximately 200 groundwater samples
were chemically analyzed. In addition, grain size analyses, scan-
ning electron microscopy, ultimate analyses and other analytical
work was performed. Reduction and interpretation of the field
data produced a thorough site characterization, and quantified the
amount of potentially hazardous constituents migrating off-site.
The recommended remedial action included securing the site
from unauthorized trespass, periodic groundwater sampling and
analysis for selected parameters, and covering and protecting some
exposed waste. One barrel of pesticide encountered during test
pit excavation was excavated and disposed of at a secure land-
fill. This relatively low level of remedial action was acceptable
to the regulatory agencies because the amounts of potentially haz-
ardous constituents that were moving off-site were extremely small.
The off-site migration was less than the discharges from neigh-
boring industries. Further remedial action may entail excava-
tion and incineration of the sludges. The pits would then be back-
filled with material of predictable bearing capacity.
The direct costs to the owner to date are on the order of 4.7
million dollars, including the $3,000,000 repurchase price and
$1,700,000 for site characterization and initial remedial action.
It does not include the costs of potential liability associated with
continued ownership or the costs of bad publicity.
Hindsight indicates that an initial environmental assessment
carried out prior to any transfer activities would have saved at
least the $3,000,000 repurchase price.
-------
477
LEGAL
ROLE DEPENDENT APPLICATIONS
Let us consider how each party to a real estate transaction may
wish to incorporate consideration of hazardous wastes into his or
her individual strategy.
The prospective buyer's position is the most straightforward.
Not yet committed to purchase, the buyer will want a thorough
assessment of the nature and extent of any hazardous waste prob-
lem which may exist, for such a problem could significantly alter,
or even defeat, the contemplated use of the property. Accord-
ingly, the buyer will want to include a hazardous waste inspec-
tion clause into the purchase and sale agreement if there are
threshold indications that trouble may exist. In addition to the
"red flags" discussed above, the buyer should be alert to an Order
of Condition or other problem markers which may appear in a
routine title run-down.
The inspection clause should take into account who is to bear the
cost of inspection, who will select the technical consultant and
have access to the results, and what options will be available to the
parties when the tests are completed and the results have been eval-
uated. The buyer will want to have the option of terminating or re-
negotiating the deal when he or she has evaluated the informa-
tion gained in light of the planned use of the property.
While the buyer will want to have as much information as pos-
sible about the potential risks associated with the site, and may be
motivated to develop a "worst-case" picture in order to gain nego-
tiating strength, the buyer must nevertheless see that the investiga-
tion is conducted with discretion. In the event that a deal is con-
summated, the buyer will have to live with the site evaluation.
Moreover, accurate rather than sensatkmal information is essen-
tial to making a sound business judgment about the economic ef-
fects of the hazardous waste problem upon the usefulness of the
property. Objective, fair investigation thus will serve the buyer
best.
The seller's position is more complex. Because potential statu-
tory (and perhaps common law) liability is joint and several, the
seller cannot realistically hope to sell the problem to a naive buyer
and walk away from it forever. It is always possible that a prob-
lem may develop in the future, and the seller, as a former owner,
will be answerable to enforcement agencies or other third par-
ties. In some situations, for example if the buyer is a large corp-
oration with extensive assets, and the seller is planning to dis-
solve its interests or move out of the United States, the seller may
find it an acceptable risk to convey with little attention to solu-
tion of a hazardous waste problem. In most cases, however, a
seller cannot hope to evade a hazardous waste problem and pos-
sible liability by transferring the property to an unsuspecting buyer.
Indeed, in most instances the liability imposed by the state and
federal statutes forces the seller to be very particular about the buy-
er's identity. For example, if the site is contaminated, the seller
must be wary of conveying to a buyer who is undercapitalized,
or notoriously irresponsible, or both, because it is possible that
such a party will aggravate any problems, but be financially unable
to assume the consequential liability.
The fact of the statutory liability scheme, taken together with the
unlikelihood that the question of hazardous wastes will simply be
overlooked by a sophisticated buyer in a sizable transaction, may
lead the seller to initiate hazardous waste evaluation prior to the
sale. By taking the initiative, the seller can select the technical con-
sultants, actively participate in designing the investigation strategy,
and have complete access to test results. Based on evaluation of test
results, the seller can consider remedial measures and incorpor-
ate them into a marketing strategy. If the seller possesses and con-
trols information about the problems which exist, it will be easier
to resist possible unfair "scare" tactics which a potential buyer may
try to use.
A seller is in the best position to give an honest and complete site
history to consultants. With good information, the consultants can
efficiently select locations for sampling, and do laboratory tests
only for relevant materials. Guess work, dead ends, and surprises
are reduced; better results are achieved at less cost.
Based on test results, remedial measure and possible uses for the
parcel can be coordinated. For example, a portion of a lot might
be able to accommodate significant new construction only at pro-
hibitive cost because of the necessity of removing contaminated
soil. The same problem area, however, might be capped or other-
wise contained and used for parking or as open space at far less
cost. The seller, together with technical advisors, may be able to de-
vise and market to the propsective buyer plans which are both re-
medial and developmental. If worked out in consultation with en-
forcement agencies, such plans may serve as a basis to limit future
liability.
Thus, the informed seller has the advantage of being able to
acknowledge the existence of a problem, and suggest ways of cop-
ing with it. Because such a seller knows what remedial measures
are available and what they are likely to cost, the hazardous waste
problem becomes simply another fact about the property. It is no
longer a club in the hands of a potential buyer who has commis-
sioned an inspection, and quotes astronomical figures based on a
radical "worst-case" remedial scheme in order to win the advan-
tage in price negotiation.
The prepared seller also gains time. The seller can make haz-
ardous waste information readily available to a serious potential
buyer rather than wait for the buyer to initiate tests. Moreover,
the knowledgeable seller can negotiate the purchase and sale agree-
ment and final sales contract to meet his or her own needs. It may
be possible to bargain over who should undertake remedial work,
what standards of performance must be observed in remedial work,
and so forth. The seller may reserve the right to terminate deal-
ings with a buyer who proves to be financially weak or otherwise
irresponsible. The seller may wish to prevent inappropriate uses of
portions of the land by restrictive covenants, exception deeds or
other mechanisms designed to isolate problem areas. Armed with
knowledge of the site, the seller may be able to limit future liabil-
ity through indemnification agreements.
If nothing else, the seller may hope to limit future liability by pre-
serving a "picture" of the status of the site. Test results create a
record of which wastes were associated with the seller's use and re-
duce the likelihood that the seller will be charged with responsibil-
ity for wastes he or she did not generate.
The potential mortgagee must also be concerned with the possi-
bility of hazardous waste contamination. A prudent mortgagee
will incorporate threshold inquiry into its routine questionnaire for
borrowers. Obviously, property subject to an enormous cleanup
liability is not good loan security. The mortgagee must also be
wary of becoming an "owner" subject to full liability in the event
that it becomes a mortgagee in possession. To reduce risk, lend-
er's counsel may wish to demand a hazardous waste opinion from
borrower's counsel.
Contractors who bid on work connected with renovation of a
site or its adaptation to a new use must also exercise caution. In
the course of construction activities a contractor may become
"transporter" of hazardous wastes subject to appropriate licen-
sing and management procedures, and will be exposed to liabil-
ity if it fails to comply with those standards.
A real estate broker must consider the possibility that delib-
erate concealment of information about hazardous waste problems
could be construed as a misrepresentation of material facts.
Counsel for any of the parties to a real estate transaction should,
of course, be prepared to provide in depth advice regarding the
client's particular situation. Indeed, prudent counsel will routine-
ly open inquiry on the subject of hazardous wastes with his or
her client before participating in the conveyance of land. Routine
inquiry, plus consultation with technical experts should it be in-
dicated, may save a deal, and prevent economic disaster.
------- |