United States
Environmental Protection
Agency
Municipal Environmental Research EPA-600/9-81 -002b
Laboratory March 1981
Cincinnati OH 45268
i'
Research and Development
Land Disposal:
Hazardous Waste
Proceedings of the
Seventh Annual
Research Symposium
Do not remove. This document
should be retained in the EPA
Region 5 Library Collection.
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are
1. Environmental Health Effects Research
2 Environmental Protection Technology
3 Ecological Research
4 Environmental Monitoring
5 Socioeconomic Environmental Studies
6 Scientific and Technical Assessment Reports (STAR)
7 Interagency Energy-Environment Research and Development
8. "Special" Reports
9 Miscellaneous Reports
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
PROPERTY OF THE
OFFICE OF SUPERFUND
EPA-600/9-81-002b
March 1981
LAND DISPOSAL: HAZARDOUS WASTE
Proceedings of the Seventh Annual Research Symposium
at Philadelphia, Pennsylvania, March 16-18, 1981
Sponsored by the U.S. EPA, Office of Research & Development
Municipal Environmental Research Laboratory
Solid and Hazardous Waste Research Division
Edited by: David W. Shultz
Coordinated by: David Black
Southwest Research Institute
San Antonio, Texas 78284
Contract No. 68-03-2962
U.S. Environmental Protection Agency
Region 5, Library (PL-12J)
77 West Jackson Boulevard, 12th float
Chicago.lt 60604-3590
Project Officer
Robert E. Landreth
Solid and Hazardous Waste Research Division
Municipal Environmental Research Laboratory
Cincinnati, Ohio 45268
MUNICIPAL ENVIRONMENTAL RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OHIO 45268
-------
DISCLAIMER
These Proceedings have been reviewed by the
U.S. Environmental Protection Agency and ap-
proved for publication. Approval does not
signify that the contents necessarily reflect
the views and policies of the U.S. Environ-
mental Protection Agency, nor does mention of
trade names or commercial products constitute
endorsement or recommendation for use.
-------
FOREWORD
The Environmental Protection Agency was created because of increasing
public and governmental concern about the dangers of pollution to the health
and welfare of the American people. Noxious air, foul water, and spoiled
land are tragic testimony to the deterioration of our natural environment.
The complexity of the environment and the interplay between its components
require a concentrated and integrated attack on the problem.
Research and development is the first necessary step in problem solution;
it involves defining the problem, measuring its impact, and searching for
solutions. The Municipal Environmental Research Laboratory develops new and
improved technology and systems to prevent, treat, and manage wastewater and
the solid and hazardous waste pollutant discharges from municipal and community
sources; to preserve and treat public drinking water supplies; and to minimize
the adverse economic, social, health and aesthetic effects of pollution. This
publication is one of the products of that research—a vital communications
link between the researcher and the user community.
These Proceedings present the results of completed and ongoing research
projects concerning the land disposal of hazardous waste.
Francis T. Mayo
Director
Municipal Environmental
Research Laboratory
in
-------
PREFACE
These Proceedings are intended to disseminate up-to-date information
on extramural research projects concerning the land disposal of hazardous
wastes. These projects are funded by the Solid and Hazardous Waste Research
Division (SHWRD) of the U.S. Environmental Protection Agency, Municipal
Environmental Research Laboratory in Cincinnati, Ohio.
The papers in these Proceedings are arranged as they were presented at
the symposium and have been printed basically as received from the authors.
They do not necessarily reflect the policies and opinions of the U.S. Environ-
mental Protection Agency. Hopefully, these proceedings will prove useful
and beneficial to the scientific community as a current reference on the
land disposal of hazardous wastes.
IV
-------
ABSTRACT
The Seventh Annual SHWRD Research Symposium on land disposal of
municipal solid waste, hazardous waste, and resource recovery of municipal
solid waste was held in Philadelphia, Pennsylvania, on March 16, 17, and
18, 1981. The purposes of the symposium were (1) to provide a forum for a
state-of-the-art review and discussion of ongoing and recently completed
research projects dealing with the management of solid and hazardous wastes;
(2) to bring together people concerned with municipal solid waste manage-
ment who can benefit from an exchange of ideas and information; and (3) to
provide an arena for the peer review of SHWRD's overall research program.
These Proceedings are a compilation of papers presented by the symposium
speakers.
The symposium proceedings are being published as three separate
documents. In this document, Land Disposal: Hazardous Waste, seven
technical areas are covered. They are as follows:
(1) Hazardous waste characterization
(2) Transport and fate of pollutants
(3) Hazardous waste containment
(4) Land treatment of hazardous waste
(5) Hazardous waste treatment
(6) Uncontrolled sites/remedial action
(7) Economics
-------
TABLE OF CONTENTS
SESSION C - HAZARDOUS WASTE MANAGEMENT
Page
Overview: Current Research on Land Disposal of Hazardous Wastes . . . . ix
Session C-l. Hazardous Waste Characterization
Improved Techniques for Flow of Liquids through Hazardous Waste
Landfills 1
Development of a Solid Waste Leaching Procedure and Manual 9
Session C-2. Transport and Fate of Pollutants
Behavior of Cd, Ni, and Zn in Single and Mixed Combinations
in Landfill Leachates 18
Aqueous Chemistry and Adsorption of Hexachlorocyclopentadiene by
Earth Materials 29
Methods of Soil Hydraulic Conductivity Determination and
Interpretation 43
Evaluation of Molecular Modelling Techniques to Estimate the Mobility
of Organic Chemicals in Soils: II. Water Solubility and the
Molecular Fragment Mobility Coefficient 58
Prediction of Leachate Plume Migration 71
Mechanisms and Models for Predicting the Desorption of Volatile
Chemicals from Wastewater 85
Session C-3. Hazardous Waste Containment
Management of Hazardous Waste by Unique Encapsulation Processes ... 91
Estimation of Pollution Potential of Industrial Waste from Small-
Scale-Column Leaching Studies 103
Organic Leachate Effects on the Permeability of Clay Liners 119
Membrane Liner Systems for Hazardous Waste Landfills 131
Durability of Liner Materials for Hazardous Waste Disposal
Facilities 140
-------
Table of Contents (cont'd)
Page
Installation Practices for Liners 157
Session C-4. Land Treatment of Hazardous Wastes
Identification of Hazardous Waste for Waste Treatment Research. . . . 168
Statistical Analysis of Trace Metal Concentrations in Soils at
Selected Land Treatment Sites 178
Factors Influencing the Biodegradation of API Separator Sludges
Applied to Soils 188
Review and Preliminary Studies of Industrial Land Treatment
Practices 200
Assessment of Hydrocarbon Emissions from Land Treatment of
Oily Sludges 213
The Development of Laboratory and Field Studies to Determine the
Fate of Mutagenic Compounds from Land Applied Hazardous Waste . 224
Closure Techniques at a Petroleum Land Treatment Site 240
Land Treating Tannery Sludges Initiation of a Five Year
Investigation 246
Session C-5. Hazardous Waste Treatment
Inorganic Hazardous Waste Treatment. II 250
Emerging Technologies for the Destruction of Hazardous Waste
Ultraviolet/Ozone Destruction 265
Evaluation of Catalyzed Wet Oxidation for Treating Hazardous
Wastes 272
Registration Requirements for Pesticide Labeling - Land Disposal . . 277
Preliminary Studies Evaluating Composting as a Means for
Pesticide Disposal 283
Session C-6. Uncontrolled Sites/Remedial Action
Top Sealing to Minimize Leachate Generation - Status Report 291
Investigation of the LiPari Landfill Using Geophysical Techniques . . 298
Remedial Actions at Uncontrolled Hazardous Waste Sites 312
Barrel and Drum Reconditioning Industry Status Profile 320
-------
Table of Contents (cont'd)
Page
Bench Scale Assessment of Concentration Technologies for Hazardous
Aqueous Waste Treatment 341
Application of Remote Sensing Techniques to Evaluate Subsurface
Contamination and Buried Drums 352
Session C-7. Economics
Conceptual Cost Analysis of Remedial Actions at Uncontrolled
Sites 366
Regulating Illegal Dumping of Hazardous Wastes 387
-------
CURRENT RESEARCH ON LAND DISPOSAL OF HAZARDOUS WASTES
Norbert B. Schomaker
John V. Klingshirn
Michele A. Gualtieri
U.S. Environmental Protection Agency
26 West St. Clair Street
Cincinnati, Ohio 45268
ABSTRACT
The Solid and Hazardous Waste Research Division (SHWRD), Municipal Environmental
Research Laboratory, U.S. Environmental Protection Agency, in Cincinnati, Ohio, has
responsibility for research in solid and hazardous waste management in the area of waste
disposal to the land. To fulfill this responsibility, the SHWRD is developing concepts
and technology for new and improved systems of solid and hazardous waste land disposal;
is documenting the environmental effects of various waste disposal practices; and is
collecting data necessary to support implementation of disposal guidelines mandated by
the "Resource Conservation and Recovery Act of 1976 (RCRA)" PL 94-580. This paper will
present an overview of the land disposal aspects of the SHWRD Hazardous Waste Program
Plan and will report the current status of work in the following categorical area:
A. LANDFILL DESIGN CRITERIA
1. Waste Leaching and Analysis
2. Pollution Migration
3. Pollutant Control
4. Waste Modification
B. LANDFILL ALTERNATIVES
1. Land Treatment
2. Surface Impoundments
C. UNCONTROLLED SITES/REMEDIAL ACTION
D. ECONOMIC ASSESSMENT
INTRODUCTION
The waste residual disposal research
strategy, encompassing state-of-the-art
documents, laboratory analysis, bench and
pilot studies, and full-scale field verifi-
cation studies is at various stages of
implementation. Over the next 5 years the
research will be reported as criteria and
guidance documents for user communities.
The waste disposal research program is
currently developing and compiling a data
base for use in the development of guide-
lines and standards for waste residual
disposal to the land as mandated by the
"Resource Conservation and Recovery Act of
1976" (RCRA). Permit Writer's Guidance
Manuals which provide guidance for conduct-
ing the review and evaluation of permit
applications are currently being-prepared.
Technical Resource Documents in support of
the Guidance Manuals are also being prepared
in specific areas to provide current tech-
nologies and methods for evaluating the
performance of the applicant's design.
The information and guidance presented in
these manuals will constitute a suggested
approach for review and evaluation based
on best engineering judgements.
The current waste residual disposal
research program has been divided into
four general areas: (a) Design Consider-
ations for Current Landfill Disposal
-------
Techniques; (b) Alternatives to Current
Landfill Disposal Techniques; (c) Remedial
Action for Minimizing Pollutants from
Unacceptable or Inoperative Sites; and
(d) Economic Assessment of Hazardous Waste
Disposal Practices and Alternatives.
The waste residual research program
has been discussed in the previous six
symposia. These symposia describe both
the land disposal of municipal and haz-
ardous waste. The most recent symposium
was on Research on Land Disposal of
Hazardous Wastes: "Proceedings of the
Sixth Annual Research Symposium: March
17-20, 1980, Chicago, Illinois, EPA-600/9-
80-010.
LANDFILL DESIGN CRITERIA
Waste Leaching and Analysis
The overall objective of this research
activity is to provide information on the
toxicity and compatibility of hazardous
wastes and their generated leachates.
Analysis of the contaminants within a
waste leachate sample is difficult due to
interfering agents. Existing USEPA proce-
dures have been developed on an as-needed
basis as part of the SHWRD projects. How-
ever most of this work is specific to the
wastes being studied and separate efforts
were required to insure that more general
procedures/equipment would be developed.
One effort (1) in the waste leaching
and analytical techniques area is directed
towards analytical methods for applica-
bility in leaching analysis. Physical and
chemical methods were evaluated for their
applicability in analysis of contaminated
industrial leachate streams. The results
of this completed effort will be published
in a report entitled "Analytical Methods
Evaluation for Applicability in Leaching
Analysis".
A second effort (1) has developed and
validated recommended test procedures for
toxicity of leachates and extracts of
wastes. This completed effort has been
published in a report entitled "Toxicity
of Leachates" EPA-600/2-80-057, March
1980.
(1) Paranthesis numbers refer to the
project officers listed immediately
following this paper who can be
contacted for additional information.
This area of research activity has
recently been transferred to another USEPA
research group with the "Environmental
Monitoring and Support Laboratory" at Las
Vegas, Nevada.
Standard Sampling Techniques
Standard Sampling procedures, including
collection, preservation, and storage of
samples, do not exist for solid and semi-
solid wastes. Existing procedures for
sampling liquid effluents and soils must be
adapted to a variety of circumstances.
Experience with sampling procedures has
been accumulated as part of several on-going
SHWRD projects.
One effort (2) in the sampling tech-
nique area was an activity to standardize
methods for sampling and analysis of hazard-
ous wastes. The findings of this completed
effort have been published in a report
entitled "Samplers and Sampling Procedures
for Hazardous Waste Streams" EPA-600/2-80-
018, January 1980. This effort has been
expanded to include development of a list
of reactive wastes and test methods for
those wastes thought to cause serious
problems.
This expanded effort (2) has produced
a method for determining the compatibility
of binary combinations of hazardous waste.
This method consists of a step-by-step
analysis procedure after which a compat-
ibility chart can be read. This method,
chart included, has been published in the
report entitled "A Method for Determining
the Compatibility of Hazardous Wastes"
EPA-600/2-80-076, April 1980.
A second effort (2) with objectives
similar to the previous effort has devel-
oped a protocol for field testing of
reactive wastes that may cause serious
problems.
A third effort (1) has determined
which sampling and preservation techniques
should be acceptable as standards for
groundwater sampling. Six landfill moni-
toring wells have been studied using four
different pumping techniques and thirteen
different sample preservation procedures.
The results of this completed effort will
be published in a report entitled
"Monitoring Well Sampling and Preservation
Techniques".
-------
This area of research activity has
recently been transferred to another USEPA
research group with the "Environmental Mon-
itoring and Support Laboratory" at Las
Vegas, Nevada.
Co-Disposal Generated Leachates
The overall objective of the co-
disposal activity is to assess the impact
of the disposal of industrial waste mate-
rials with municipal solid waste. Because
the environmental effects from landfill ing
result from not only the soluble and slowly
soluble materials placed in the landfill
but also the products of chemical and
microbiological transformation, these
transformations should be a consideration
in management of a landfill to the extent
that they can be predicted or influenced by
disposal operations.
One ongoing effort (5) involves a
study of the factors influencing (a) the
rate of decomposition of solid waste in a
sanitary landfill, (b) the quantity and
quality of gas and leachate produced during
decomposition, and (c) the effect of ad-
mixing industrial sludges and sewage sludge
with municipal refuse. A combination of
municipal solid waste and various solid and
semi-solid industrial wastes was added to
several field lysimeters. The industrial
wastes investigated were: petroleum sludge,
battery production waste, electroplating
waste, inorganic pigment sludge, chlorine
production brine sludge, and a solvent-
based paint sludge. Also, municipal di-
gested primary sewage sludge dewatered to
approximately 20 percent solids was ^
utilized at three different ratios. The
results of this effort have been reported
in a paper entitled Co-Disposal of Indus-
trial and Municipal Wastes in a Landfill:
"Proceedings of the Fourth Annual Research
Symposium" EPA-600/9-78-016, August 1978,
pp. 129-151. The updated results of this
effort have been published in a paper
entitled Leachate From Municipal and
Industrial Waste Simulators: "Proceedings
of the Sixth Annual Research Symposium"
EPA-600/2-80-010, March 1980, pp. 203-222.
Pollutant Migration
The overall objective of this research
activity is to develop procedures for using
soil as a predictable attentuation medium
for pollutants. Mot all pollutants are
attenuated by soil, and in some cases, the
process is one of delay so that the pollut-
ant is diluted in other parts of the envi-
ronment. Consequently, a significant num-
ber of the research projects funded by
SHWRD are focused on understanding the
process and predicting the extent of mi-
gration of contaminants (chiefly heavy
metals) from waste disposal sites. Both
laboratory and field verification studies
at selected sites are being performed to
assess the potential for groundwater con-
tamination. The studies will provide the
information required to (a) select land
disposal sites that will naturally limit
release of pollutants to the air and water
and (b) make rational assessments of the
need for the cost-benefit aspects of
leachate and gas control technology. Pro-
gress has been made using soil and waste
characteristics to predict movement of
pollutants in soil, but the results are
not sufficiently comorehensive nor have
they been field tested significantly to
qualify as a basis for disposal regulations
that will be challenged most assuredly in
legal proceedings. Work is in progress on
empiricial predictive techniques using
samples of wastes and soils from locations
of specific interest. This approach
appears to be the most promising in order
to integrate the effect of waste and soil
characteristics on pollutant retention
processes. The precision and accuracy
that will be achieved by this method
remains to be determined.
One on-going effort (4) is the devel-
opment of a Guidance Document addressing:
(a) to what degree leachate concentrations
will be reduced by mixing with groundwater
in the vicinity of waste disposal sites in
a variety of hydrogeologic situations (b)
the direction of movement and shape of
leachate plumes, and (c) the types of
models that are appropriate for predicting
movement of solutes in groundwater at sites
with specific hydrogeologic characteristics,
assuming that solutes do not interact with
aquifer materials.
A second effort (4) examined the
extent to which hazardous substances from
specific industrial and flue gas cleaning
(FGC) wastes would migrate into groundwater
at disposal sites. Procedures for con-
ducting such examinations have been
developed. Leachate from a municipal
solid waste landfill was used to extract
the following industrial wastes: electro-
plating, inorganic pigment, water based
-------
paint, Hg cell chlorine production and Ni-
Cd battery. A sequential batch leaching/
soil adsorption procedure was developed
which provides information comparable to
that obtained from soil column studies, but
in a much shorter time. The results of
this completed effort will be published in
a report entitled "Migration of Hazardous
Substance Through Soil".
A third effort (4) was the development
and testing of a large scale hydrogeologic
simulation model for predicting the move-
ment of nonconservative solutes in saturated
and unsaturated soils. The two-dimensional
model used successfully to study a chromium
contamination problem has been developed
into a three-dimensional model. The results
of this completed effort will be published
in a report entitled "Hydrologic Simulation
on Solid Waste Disposal Sites".
A fourth effort (4) is the on-going
development of user oriented models for
predicting movement in soil as a basis for
improving selection of sites for disposing
of solid and hazardous wastes. Average
movement rates of Cd, Ni, and Zn from
industrial waste leachates in soils will be
measured and the Lapidus Amundson solute
movement model will be used to predict long
term steady state movement rates for each
set of experimental conditions. Long term
movement rates will then be correlated with
soil and leachate properties to develop
regression equations forming a basis for a
simplified predictive tool.
A fifth effort (4) studied vertical
and horizontal contaminant migration of Zn,
Cd, Cu and Pb at three secondary zinc smelt-
ing plants and one organic chemical manu-
facturing plant. Migration patterns were
defined using soil coring and monitoring
well techniques. Soil coring was deter-
mined to be an investigative tool, but not
suitable by itself for routine monitoring
of waste disposal activities. The results
of this study are currently being published
in a report entitled "Field Verification
of Toxic Wastes Retention by Soils at
Disposal Sites."
A sixth effort (4) was the determina-
tion of the attenuation mechanisms and
capacity of selected clay minerals and soils
for hexachlorocyclopropentadial (HCCPD) and
"hex" wastes. Also, effects of caustic-
soda brine on the attenuation and solubility
of HCCPD and the development of a chemical
model to predict HCCPD migration through-
soil were pursued. Mobility of HCCPD in
soils was independently assessed by use of
soil thin-layer chromatography and column
leaching techniques. The results of this
completed effort will be published in a
report entitled "Assessment of Soil, Clay
and Caustic Soda Effects on Land Disposal
of Hydrocarbon Wastes (HCCPD)".
A seventh effort (2) studied organic
contaminant attenuation by selected clay
minerals and coal chars. Initial work was
with Polychlorinated biphenyls (PCBs);
subsequent efforts were with Hexachloro-
benzene (HCB) and Polybrominated biphenyls
(PBBs). The mobility of Arochlors 1242
and 1254, and a used capacitor fluid were
measured by the soil thin-layer chromatog-
raphy technique and additional batch ad-
sorption studies were conducted. PCBs
were found to be strongly adsorbed by soil
materials. The adsorption capacity and
the mobility of PCBs were positively
correlated to the organic carbon content
and surface area of the respective soil
materials. The results of this effort will
be published in a report entitled "Attenu-
ation of Polybrominated Biphenyls and
Hexachlorobenzene by Earth Materials".
An eight effort (3) currently under-
way, aims to develop a guidance manual
identifying wastes that increase the
permeability of soils at a disposal site.
Test procedures for predicting whether a
specific waste and soil will react to
increase soil permeability will be
developed.
A ninth effort (4) is currently under-
way to determine how accurately the EPA
Gas Movement Model predicts the maximum
distance that methane gas will move through
soils adjacent to landfills and how
accurately this model will predict the
relative effectiveness of control systems
(e.g., trenches, wells, barriers) for
minimizing methane gas movement.
The extent of methane movement will be
measured at three selected landfills.
With data from each landfill, the model
will be used to predict the maximum extent
of methane movement. The measured and
predicted methane movements will be com-
pared, reasons for any differences analyzed,
and the accuracy and effort of use for the
model determined.
-------
A tenth effort (5) has resulted in
establishment of guidelines for design of
gas migration control devices for sanitary
landfills. The migration control devices
studied included trenches venting under
natural convection, trenches with exhaust
pumping, trenches with recharge pumping,
barriers, hybrid systems consisting of a
barrier with a pumped (exhaust or recharge)
or unpumped trench on the landfill site,
and pumped (exhaust or recharge) pipe vents.
Computer codes developed under a previous
study (c.f. Moore and Alzaydi, 1977: and
Moore and Rai, 1977) were modified to
incorporate pressure flow as well as dif-
fusional flow. The results of this com-
pleted effort will be published in a report
entitled "Gas Control Design Criteria".
An eleventh effort (4) has evaluated
the conditions that would control the move-
ment of hexachlorobenzene (HCB) out of land-
fills and other disposal/storage facilities
into the surrounding atmosphere. The
volatilization fluxes of hexachlorobenzene
from industrial wastes (hex waste) were
determined using coverings of soil, water
and polyethylene film in a simulated land-
fill under controlled laboratory conditions.
The results of this effort have been pub-
lished in a report entitled "Land Disposal
of Hexachlorobenzene Wastes - Controlling
Vapor Movement in Soils," EPA-600/2-80-119,
August 1980.
Pollutant Control
The overall objective of this research
activity is to reduce the impact of pollu-
tion from waste disposal sites by technol-
ogy that minimizes, contains, or eliminates
pollutant release and leaching from waste
residuals disposed to the land. The
pollutant control studies are determining
the ability of in-situ soils and natural
soil processes to attenuate leachate
contaminants as the leachate migrates
through the soil from landfill sites. The
studies are also determining how various
synthetic and admixed materials may be
utilized as liners to contain and prevent
leachates from migrating from landfill
sites.
Natural Soil Processes
The treatment by natural soil processes
of pollutants from hazardous waste and
municipal refuse disposal sites is being
performed in the controlled lab studies
previously discussed in the section on
pollutant transport.
Liners/Membrane/Admixtures
The liner/membrane/admixture technol-
ogy (3) is currently being studied to
evaluate suitability for eliminating or
reducing leachate from landfill sites of
municipal or industrial hazardous wastes.
The test program will evaluate in a land-
fill environment, the chemical resistance
and durability of the liner materials over
12- and 36-month exposure periods to
leachates derived from industrial waste,
sulfur oxides (SO ) wastes, and municipal
solid wastes. Acidic, basic, and neutral
solutions will be utilized to generate
industrial waste leachates.
The liner materials being investigated
under the hazardous waste program include
five admixed materials and eight polymeric
membranes. Specimens of these materials
have been exposed for more than 3 years to
the following six classes of hazardous
wastes which utilized ten specific types
of wastes: strong acid; strong base;
waste of saturated and unsaturated oils;
lead waste from gasoline tanks; oil
refinery tank bottom waste (aromatic oil);
and pesticide waste. Preliminary exposure
tests have been completed on the various
liner materials in the various wastes in
order to select combinations for long term
exposures. The results of these tests
along with a discussion of the overall
hazardous waste liner material program are
presented in a report entitled "Liner
Materials Exposed to Hazardous and Toxic
Sludges First Interim Report," EPA-600/2-
77-091, June 1977.
A second effort (3) relates to the
types of materials tested for use as liners
for sites receiving sludges generated by
the removal of SO from flue gases of coal-
burning power plants. The volumes of SO
sludges generated in any particular place
will, typically, be much greater than those
for other types of wastes, and therefore
the disposal sites will be large. Con-
sequently, methods of lining such disposal
sites must have a low unit cost for mate-
rials as well as labor. The admixed
materials consisted of the following:
cement; lime; cement with lime; polymeric
bentonite blend-(Ml79); gray powder-quartec
(UF); asphaltic concrete; TACSS 020; TACSS
025; TACSS C400; and TACSS ST. The
Xlll
-------
prefabricated membranes liner materials
consisted of the following: elasticized
polyolefin; black neoprene - coated nylon;
and black neoprene - reinforced fabric.
The spray-on liner materials consisted of
the- following: polyvinyl acetate; natural
rubber latex; natural latex; polyvinyl
acetate; asphalt cement; and molten sulphur.
For this effort, a total of 72 special test
cells were constructed to perform 12 and 24
month exposure tests. An interim report
has been published entitled "Flue Gas
Cleaning Sludge Leachate/Liner Compati-
bility Investigated: Interim Report,
"EPA-600/2-79-136, August 1979. A final
report will be published entitled "Effect
of Flue Gas Cleaning Sludges on Selected
Liner Materials".
A third effort (3) is an on-going
study to assess actual field procedures
utilized in (a) preparing the support sub-
base soil structure for liners and (b)
placing the various liner materials common
to projects requiring positive control of
fluid loss. A fourth on-going effort (3)
is concerned with determining the soil
requirements to act as bedding and protec-
tive covers for flexible membranes. Exper-
iments will be conducted using a variety of
flexible membranes, base, and subbase
materials, thicknesses and densities so
that a design manual, applicable to the
majority of the country, can be developed
which will provide a design engineer with
the data necessary to determine the most
economic soil requirements above and below
a flexible membrane. The results of this
effort to date have been compiled in an
unpublished report entitled "Membrane Liner
Systems for Hazardous Waste Landfills".
A fifth on-aoing effort (3) responds
to the need for support of the guidelines
and hazardous waste criteria mandated under
RCRA. This project has been designed to
provide the EPA with data to enable annual
revisions to the landfill liner design
manual now being prepared. Furthermore,
it will provide assistance in permit
application evaluation, assessment of
performance and expert testimony.
A sixth on-going effort (3) relates to
evaluatina long term performance of liner
systems. As liner technology advances and
new materials and design continue to be
developed, these developments will be made
available to liner users and permit writers
through incorporation of information into
the technical resource document, "Lining
of waste Impoundment and Disposal Facil-
ities".
A seventh effort (3), now on-going, is
primarily concerned with the evaluation of
specific technologies (chemical liners)
for the disposal and recovery of sludges
from the metal finishing industry.
An eighth effort (4), now completed,
relates to a laboratory study of agri-
cultural limestone and hydrous oxides of
Fe to evaluate their use as landfill
liner materials to minimize the migration
of metal contaminants. The results of
this effort are being published in a
report entitled "Liners of Natural Porous
Materials to Minimize Pollutant Migration".
Haste Modification (Chemical Stabilization)
Chemical stabilization is achieved by
incorporating the solid and liquid phases
of a waste into a relatively inert matrix
which exhibits increased physical strength
and protects the components of the waste
from dissolution by rainfall or by soil
water. If this slows the rate of release
of pollutants from the waste sufficiently
and no serious stresses are exerted on the
environment around the disposal site, then
the wastes have been rendered essentially
harmless and restrictions on siting will
be minimal.
The initial chemical fixation effort
(3) which is on-going relates to the
transformation of wastes into an insoluble
form to minimize leaching. The test pro-
gram consists of investigating ten indus-
trial waste streams in both the raw and
fixed states. The waste streams are
treated with at least one of seven separate
fixation processes and subject to leaching
and physical testing. The lab and field
studies have been completed. The results
have been compiled and discussed in a re-
port entitled "Pollutant Potential of Raw
and Chemically Fixed Hazardous Industrial
Wastes and Flue Gas Desulfurization Sludges-
Interim Report," EPA-600/2-76-182, July
1976. The final results of these studies
will be reported in three reports entitled
"Small Column Leaching Tests on FGD Sludges",
"Small Column Leaching Tests on Industrial
Sludges", and "Large Column Leaching Tests
on FGD Sludge".
-------
A second on-going effort (3) relates
to the co-disposal of chemically treated
and untreated industrial wastes in a munic-
ipal refuse environment. Large lysimeters
are being utilized to determine the differ-
ence in leachate quality.
In the third effort (3), a distilled
water shake test, the elutriate test, was
developed and tested to provide a fast,
simple procedure for predicting the escape
of pollutants from treated and untreated
sludges. The short-term elutriate test
results were compared to results of a long-
term leaching test using the same treated
and untreated sludges. The results of this
study are discussed in a report entitled
"Elutriate Test Evaluation of Chemically
Stabilized Waste Materials," EPA-600/2-79-
154, August 1979.
A fourth effort (1) relates to a
laboratory assessment of fixation and encap-
sulation processes for arsenic-laden wastes.
Three industrial solid wastes high in arse-
nic concentration were treated by generic
processes in laboratory and by proprietary
processes at vendors' facilities. Leaching
studies on treated wastes consisting of
Shake tests on pulverized samples and on in-
tact monolithic samples were performed to
assess the relative safety of each product
for disposal. The results of the completed
effort will be published in a report enti-
tled "Stabilization Testing and Disposal of
Arsenic Containing Wastes".
A fifth on-going effort (6) relates to
encapsulating processes for managing haz-
ordous wastes. Techniques for encapsulating
unconfined dry wastes are discussed in a
report entitled "Development of a Polymeric
Cementing and Encapsulating Process for
Managing Hazardous Wastes," EPA-600/2-77-
045, August 1977. Additional evaluations
are currently being performed whereby con-
tainers of hazardous waste (i.e., 55 gallon
drums) are placed in a fiber glass thermo-
setting resin casing and covered with a high
density polyethylene. The results of these
evaluations will be published in three
reports entitled "Study of Encapsulate Forma-
tion with Polyethylene Resin and Fiberglass
for Use in Stabilizing Containerized Haz-
ardous Wastes", "Study of Encapsulation for
Securing Corroding 55-Gallon Drums Holding
Hazardous Wastes by Welding Polyethylene
Resin", and "Study of Encapsulate Formation
with Resin by Spray and/or Brush Under
Atmospheric Conditions for Use in Management
of Containizered Hazardous Wastes".
A sixth on-going effort (3) is to
determine how potentially useful sulfur-
asphalt binder composites are for
stabilization/solidification and disposal
of toxic metal bearing wastes. Indica-
tions from work carried out on paving
material suggest that sulfur-asphalt blends
can furnish uniquely desirable properties
when used as binders for toxic waste
control.
LANDFILL ALTERNATIVES
Due to the concern for environmental
impact and economics, alternatives to
waste disposal in sanitary landfills and
by incineration have been proposed.
Alternatives currently being researched
by SHWRD include land treatment and
surface impoundments.
Land Treatment
The SHWRD program includes research
and evaluation of spreading (or incorpora-
ting) of industrial hazardous waste onto
the land for treatment by biological,
chemical, and physical processes resulting
from interactions of the soil (and its
constitutents) with the waste. The
technology is referred to as "land treat-
ment." Land treatment enhances degradation
by providing an aerobic soil medium capable
of supporting microorganisms and providing
exposure to sunrays. The land treatment
research program has three basic objectives.
These are a) to determine environmental
effects from hazardous waste land treatment
facilities and their monitoring requirements
b) to determine hazardous waste land treat-
ment facility closure requirements and
post-closure monitoring requirements c) to
develop basic knowledge required to fully
understand land treatment of hazardous
waste.
One effort (3) is designed to develop
a matrix of industrial organic and inorganic,
and municipal solid waste streams versus
operational parameters. The final matrix
of information will be sufficiently complete
to develop design and guideline criteria.
The results of this completed effort will
be published in a report entitled "Field
Verification of Land Cultivation/Refuse
Farming".
XV
-------
A second effort (3) was aimed at
optimizing the operational parameters for
the biodegradation of API oil-water sepa-
rator sludge on the germination and yield
of ryegrass in order to generate data
on acceptable sludge loading rates and to
elicit mechanisms which affect plant re-
sponses. Long-term yield reductions largely
resulted from impaired water, air and nutri-
ent relations associated with recalcitrant
hydrophobic hydrocarbons.
A third effort (6) was the development
of a best engineering judgement (BEJ),
Permit Writer's Guidance Manual which com-
piles the information necessary to develop
a land treatment facility. This project
also offered technical support for the
management of a land treatment facility.
This manual includes comprehensive back-
ground information and criteria for refer-
ence by designers, permitting officials,
and technical (field) personnel in the area
of land treatment. Illustrative case stud-
ies unique laboratory methods, and a land
treatment glossary are also included. The
manual is entitled "Design and Management
of Hazardous Waste Land Treatment Facili-
ties ".
A fourth on-going effort (6) evaluates
current land treatment data required to
support development of and revisions to the
best engineering judgement (BEJ), Permit
Writer's Guidance Manuals, required in
support of RCRA regulations. Research
under this agreement will involve a combi-
nation of evaluations of available data,
laboratory and greenhouse investigations,
on using the land as a treatment/management
medium for hazardous wastes. Studies will
involve determination of effects of food
and non-food chain crops, development of
crop response curves, determination of soil
effects on hazardous pollutant migration,
ion exchange capacity effects, soil adsorp-
tion and neutralization effects, feasibility
of mixing hazardous waste streams, waste
preconditioning, and numerous other factors
required to better understand land treatment
of hazardous waste.
A fifth on-going effort (3) identifies
and prioritizes hazardous waste streams
likely to be amenable to land treatment as
a management alternative, recommends labo-
ratory and field research to fill in sig-
nificant data gaps, and verifies the environ-
mental acceptability of land treatment as a
waste treatment/resource recovery method
for a significant fraction of the hazardous
wastes generated.
A sixth on-going effort (6) assesses
technical requirements for effective clo-
sure of hazardous waste land treatment
facilities. Effective closure is accom-
plished when the environment is completely
protected from potential adverse effects
of the land treatment facility. Although
petroleum land treatment facilities will
be used in cooperation with the American
Petroleum Institute, tests will be designed
so that information is applicable to other
wastes.
A seventh on-going effort (6) was
undertaken to determine the utility of
bioassays as a control/monitoring tech-
nique for evaluating the operation and
engineering performance of hazardous waste
land treatment sites. Results will be
used to determine hazardous waste land
treatment site engineering design and
control technology required to prevent
movement of potential mutagens from the
site.
An eighth on-going effort (6) charac-
terizes the technical and environmental
aspects of using land treatment technology
for treating tannery waste sludges. The
project is a field investigation and eval-
uation designed to verify information
available from laboratory and greenhouse
work, and to provide EPA with data on the
proper design and operation (including
closure) of similar hazardous waste land
treatment facilities.
A ninth on-going effort (6) studies
active petroleum land treatment sites
operating for five years or longer. Core
samples will be taken at selected locations,
segmented at different levels, and analyzed
for various constitutents. Information
generated will help evaluate the fate of
organics and other contaminants, and their
potential migration from hazardous waste
land treatment facilities.
Surface Impoundments
The temporary, or permanant retention
of liquid hazardous wastes in surface
impoundments (i.e., pits, ponds, or lagoons)
is an alternative to disposal in a sanitary
landfill, which SHWRD is researching. Sur-
face impoundments require special consider-
ation to control vaporization and volatili-
zation in the air.
-------
The initial effort (4) is providing
technical information necessary to evaluate
a landfill or surface impoundment design in
the form of a best engineering judgement
(BEJ) Permit Writer's Guidance Manual. The
manual is addressing movement of liquid
through the final cover and liner materials,
the efficiency of the leachate collection
system, and volumes of leachate collected.
The results of this effort will be published
in a manual entitled "Closure of Hazardous
Waste Surface Impoundments".
A second on-going effort (4) is inves-
tigating mechanisms of chemical movement of
pollutants from surface and near-surface
impoundments into the air. Methods for
estimating air emissions of hazardous chem-
icals from waste disposal sites are being
developed.
UNCONTROLLED SITES/REMEDIAL ACTION
An on-going study by OSW has identified
incidence of water well contamination due
to waste disposal sites. Seventy-five (75)
to eighty-five (85) percent of all sites
investigated are contaminating ground or
surface waters. In order to determine the
best practical technology and economical
corrective measures to remedy these landfill
leachate and qas pollution problems, a
research effort has been initiated (!) to
provide local municipalities and users with
the data necessary to make sound judgments
on the selection of viable, in-situ, reme-
dial action procedures and to give them an
indication of the cost associated with such
an action. This research effort is being
pursued at a municipal refuse landfill site
at Windham, Connecticut. This effort con-
sists of three phases. Phase I will be an
engineering feasibility study that will
determine on a site specific basis the best
practicable technology to be applied from
existing neutralization or confinement
techniques. Phase II will determine the
effectiveness, actual field verification,
of the recommendations/first phase study.
Phase III will provide a site remedial guide
to local municipalities and users. The
engineering feasibility study has been com-
pleted and a report entitled "Guidance
Manual for Minimizing Pollution from Waste
Disposal Sites," EPA-600/2-78-142, August
1978 has been published. This guidance
document emphasizes remedial schemes or
techniques for pollutant containment.
The scheme currently installed and
monitored at the Connecticut MSW site is
a surface capping technique which could be
followed by a leachate extraction scheme
if required.
A second on-going effort (1) is con-
cerned with the development of a manual on
remedial action for hazardous waste land-
fill sites. The methodology and documenta-
tion for this manual will be developed by
implementing a remedial program at a haz-
ardous waste landfill site in New Jersey.
This program will consist of a quantitative
site assessment, selection and design of a
remedial system and implementation of that
design system.
A third on-going effort (7) is a
technology analysis assessing the extent
of environmental damage associated with a
hazardous waste dump site, and planning
alternative programs for remedial action.
This effort relates solely to the hazard-
ous waste dump site in Conventry, Rhode
Island and is being performed in conjunc-
tion with the current effort undertaken
by the state of Rhode Island. An analyt-
ical report will be prepared which assesses
the situation at the Coventry site and
evaluates the engineering/technology
options for the abatement of the under-
ground chemical contamination at the site.
This report will be published and entitled
"Use of Remote Sensing Techniques in a
Systematic Investigation of an Uncontrolled
Hazardous Waste Site".
A fourth effort (7) provides for a
survey of on-going and completed remedial
action projects. Ten case study sites
were selected and these sites visited and
inspected for remedial action information.
The results of this study are published in
a report containing case study summary
statistics of each of the sites investi-
gated. The results of this completed
effort will be published in a final report
entitled "Survey of On-going and Completed
Remedial Actions".
A fifth on-going effort (7) is con-
cerned with the assessment of barrel and
drum reconditioning to determine environ-
mental impact and to make recommendations
for possible improvements in the recondi-
tioning processes. Based on industry
characterization and specific facility
identification, two facilities will be
chosen for sampling and analysis efforts,
XVll
-------
which will monitor for air, land and water
pollutant discharges. Should these efforts
show that the facilities tested are not
environmentally acceptable, a process de-
sign and optimization effort shall be
undertaken.
A sixth on-going effort (7) will eval-
uate and verify several selected concentra-
tion techniques for compatibility with the
ultimate disposal/detoxification of hazard-
ous materials. One concentration/pretreat-
ment scheme will be selected for further
development and subsequent field scale
verification with an actual waste stream
containing hazardous materials. The
results of this effort to date will be
published in a report entitled "Concentra-
tion Technologies for Aqueous Hazardous
Waste Treatment".
ECONOMIC ASSESSMENT
The use of market-oriented incentive
(disincentive) mechansims has received very
little consideration for pollution control
policy in the United States, particularly
in the area of hazardous waste management.
Economic theory suggests that incremental
pricing of waste collections and disposal
would reduce the waste generation rate,
enhance source separation of recyclable
materials, accelerate technological
innovation, and minimize total system cost.
One completed effort (8) examined the
applicability of cost-risk-benefit analysis
and related decision criteria to the man-
agement of hazardous wastes. The use of
cost-benefit techniques as tools for assist-
ing regulatory policy development was
addressed.
A second completed effort (8) has
resulted in a report entitled "Cost Compar-
ison of Treatment and Disposal Alternatives
for Hazardous Wastes" to be published soon.
Unit costs are estimated for sixteen (1(5)
treatment and five (5) disposal techniques
applicable to various hazardous wastes.
Life cycle average unit costs are presented
in both tabular and graohic form.
A third on-going effort (8) periodical-
ly updates unit operations cost data for long
term remedial action programs at uncontrol-
led hazardous waste sites.
CONCLUSIONS
The laboratory and field research
oroject efforts discussed here reflect
the overall SHWRD effort in hazardous
waste disposal research. The projects
will be discussed in detail in the follow-
ing papers. More information about a
specific project or study can be obtained
by contacting the project officer referenced
in the text. Inquiries can also be directed
to the Director, Solid and Hazardous Waste
Research Division, Municipal Environmental
Research Laboratory, U.S. Environmental
Protection Apency, 26 West St. Clair
Street, Cincinnati, Ohio 45268. Infor-
mation will be provided with the under-
standing that it is from research in
progress and the conclusions may change as
techniques are improved and more complete
data become available.
PROJECT OFFICERS
All the Project Officers can be con-
tacted through the Solid and Hazardous
Waste Research Division (SHWRD), whose
address is shown above.
1. Mr. Donald E. Sanning (SHWRD)
513/684-7871
2. Mr. Richard A. Carnes (IERL)
513/684-4303
3. Mr. Robert E. Landreth (SHWRD)
513/684-7871
4. Dr. Mike H. Roulier (SHWRD)
513/684-7871
5. Mr. Dirk R. Brunner (SHWRD)
513/684-7871
6. Mr. Carlton C. Wiles (SHWRD)
513/684-7871
7. Mr. Stephen C. James (SHWRD
513/684-7871
8. Mr. Oscar W. Albrecht (IERL)
513/684-4318
XVlll
-------
IMPROVED TECHNIQUES FOR FLOW OF LIQUIDS
THROUGH HAZARDOUS WASTE LANDFILLS
Elfatih Ali
Charles Moore
The Ohio State University
Columbus, Ohio 43210
ABSTRACT
This paper describes research performed to develop techniques for improving predictions
of flow of liquids through hazardous waste landfills. Statistical methods are used both
to describe the porous media and to solve the differential equations of flow. A practical
exampleJs presented.
Introduction
This paper presents a summary of the
work completed through December 1980 on
the Improved Assumptions for Flow of
Liquids in Landfills awarded by the Nego-
tiated Contracts Branch of the U.S. Envi-
ronmental Protection Agency under Contract
No. 68-03-2963.. The contract/award is
being monitored by the Ohio State Univer-
sity Research Foundation (OSURF) under the
direction of Dr. Charles A. Moore who
serves as principal investigator. The
OSURF designation for the project is
RF712979 U.S. EPA.
The contract is intended to provide an
improved state of the art approach for
determining the rate at which liquids move
through hazardous waste disposal land-
fills. These improved assumptions will
then be used to update the best engineer-
ing judgement Evaluation Procedure Manual
for Landfill and Surface Impoundment Con-
tainment prepared by Geotechnics, Inc.
Background
The present version of the Landfill and
Surface Impoundment Evaluation Procedures
manual uses highly simplified assumptions
regarding the rate at which liquid is
transmitted through the waste cell. Spe-
cifically, plug flow is assumed. Thus
liquid that enters the top of the waste
cell is presumed to appear instantaneously
at the bottom of the cell. This simplifi-
cation was predicated on the following:
(1) the state of the art for the flow
through extremely heterogeneous media is
not well understood,
(?) because hazardous waste disposal prac-
tice is still evolving, the manner in
which waste is emplaced in the cell is
variable, and
(3) innovative disposal schemes are likely
to be proposed in the near future.
Scope of the Research
This research project has a rather
broad scope -- spanning the gamut from the
development of fundamental scientific con-
cepts to the development of specific prac-
tical configurations for cell construc-
tion.
The first phase of the research consists
of developing statistical means for
describing the geometric arrangement of
wastes placed in disposal cells. The sta-
tistical formulation will be abstract in
that there will be parameters that can be
varied to allow for the simulation of
essentially any desired placement geometry
for the waste. For example, it will be
-------
possible to vary the parameters to simu-
late disposal in drums, disposal of shred-
ded dry material, disposal of baled dry
material, disposal of sludges, etc. The
philosophy upon which this first phase is
based has the important advantage of
allowing the same analytical scheme to be
used to describe the geometry of all waste
placement configurations whether they be
presently used or be proposed for use in
the future.
The second phase of the research con-
sists of developing the analytical proce-
dures to determine how liquids flow
through the waste. The flow processes to
be considered are complex and include:
(1) flow dominated by infiltration
wherein capillary forces draw liquid into
partially saturated fine grained mater-
ials,
(2) saturated Darcy flow wherein gravi-
tational forces predominate and flow is
macroscopically laminar, and
(3) flow in large openings where inertia
terms predominate.
Although continuum mechanics equations
will be developed to describe the flow
processes, the solution of the equations
will use statistical approaches such as
Monte Carlo or random walk schemes. The
statistical approaches for solving the
equations will be used because of their
analytical compatibility with the
statistical approaches being used to
describe the geometric properties of the
porous media.
Impact of Research on Future Disposal
Technology
The way in which statistical techniques
are used to solve the flow equations in
heterogeneous media may have an important
philosophical impact on future disposal
technology. Ordinarily, a configuration
for disposal is preconceived, and the
resulting flow characteristics are then
analysed. This approach can best be
termed "guess and then analyze". If the
analysis shows that the flow characteris-
tics do not meet requirements, the design
is altered by guessing at a configuration
that is likely to perform in a more desi-
rable manner.
The proposed approach to analysis con-
sists of transforming the actual geometri-
cal arrangement of the waste disposal cell
into a statistically equivalent but geome-
trically simpler geometric arrangement
called an image domain. The image domain
does not look like the landfill cell;
however, it has identical behavior from
the point of view of its ability to trans-
mit liquids. The flow analysis is actu-
ally performed on the image domain.
The philosophical impact of the statis-
tical approach is perhaps most significant
when viewed inversely. We can first
devise an image domain that meets the
desired flow requirements. Then we can
use inverse statistical mapping to create
one or more real physical geometrical
arrangements that have flow properties
identical to the image domain which, in
turn, is already known to have properties
that meet performance specifications. In
this manner, we are literally able to
create a variety of designs that meet per-
formance criteria. The choice of one par-
ticular design from among the alternatives
can be based on additional criteria such
as cost, preferred local practice, etc.
Results to Date
The results of practical importance
achieved to date are, of course, limited
because the project is in its early
stages. However, this section presents
some results that may be of general inter-
est. The configuration to be examined
consists of a hazardous waste landfill
cell (Figure 1) having dimensions X
(width) by Y (height). The landfill is
assumed to consist of two materials; waste
cells and earth fill. The permeabilities
of these two materials are KC and KF re-
spectively. The cells are assumed to be
uniformly distributed in the landfill, and
the earth cover is considered to have uni-
form thickness T all around the waste
cells.
Both the cells and earth cover mater-
ials were considered to be saturated, and
the liquid flow is assumed to follow
Darcy's Law. Only two dimensional con-
fined flow is considered in this example.
-------
n/
I nte r m •d I a t • cover
f
I waste
Figure 1. Geometry Assumed for Example Problem.
The flow was analyzed using a random
flow model for a deterministic medium.
Different landfills were simulated by var-
ying the geometric properties and the
permeabilities of the two materials:
R = T/X (ratio of earth cover thickness
to cell width)
RK = KC/KF (ratio of permeabilities of
cell and earth cover)
A = X/Y (ratio of width to height of cell)
Each landfill was studied and an aver-
age equivalent permeability K1 was com-
puted.
Initially a square cell geometry (A =
1) was investigated. Figure 2 shows the
variation of the normalized permeability
(K'/KF) with R for different values of RK.
The variation of K'/KF with RK for diffe-
rent R values is shown in Figure 3.
The effect of the shape of the cell
itself was then studied. Figure 4 shows
the variation of K'/KF with A for diffe-
rent values of RK. In this case the value
of R as kept constant at 0.05. (Other
values of R were also investigated and the
trends obtained were similar to those
shown in Figure 4; however, the numerical
values were different.)
Discussion of Results
Figure 2 indicates that K' decreases as
R increases. The reduction factor
obtained here for K1 ranged from 1/2 to
1/5. Although the value of the permeabil-
ity of the waste cell, K', is not known,
the permeability of the earth cover mater-
ials can be estimated. Hence a change in
the value of RK (= KC/KF) can be brought
about by changing the value of KF. From
Figure 3 we notice that K1 increases as RK
increases. The magnification factors
obtained range from 3 to 15.
The geometrical shape of the waste cell
has a .considerable effect on K1. If we
consider the square cell as the reference
shape, and assume that the landfill with
square cells has an average equivalent
permeability K", then Figure 5 can be
constructed. This figure plots a multi-
plication factor (M = K'/K") as a function
of A for different values of R. Two
separate regions are apparent in Figure 5:
(1) for A < 1 K' increases as A increases
but K' decreases as RK increases, and
-------
16 -
12
u. 8
X
•t
*
,RK= 100
.05
.1
.15
.2
.25
Figure 2. Ratio of Normalized Permeability (K'/KF) as a
Function of the Ratio of Cover Thickness to Cell
Width (R = T/X) for Different Values of RK
-------
16
R = .05
12
8
20
40
60
RK
Figure 3. Ratio of Normalized Permeability (K'/KF) as a
Function of the Ratio of Permeabilities of
Cell and Earth Cover (RK = KC/KF) for Different
Ratios of Cover Thickness to Cell Width (R =
-------
30
20
1 0
RK = 100
R = .05
1 .O
2.0
Figure 4. Ratio of Effective Permeability to Permeability
of Cover Material (K'/KF) as a Function of Width
to Height Ratio (A = X/Y) for Cell and for
Different Permeability Ratios (RK = KC/KF)
-------
1.5
M
.5
RK= 100
R-.05
Figure 5. Ratio of Effective Permeability of Square Cell to
Effective Permeability (M = K'/K") as a Function
of Ratio of Cover Thickness to Cell Width (R = T/X)
-------
(?) for A > 1
and RK i ncrease.
K' increases as both A
For the case shown In Figure 5, the
value of M varied between 0.33 and 2.0.
The results of this investigation can
be used to give guidelines for the design
of landfills. If the goal is to increase
the value of K' then the course of action
should be to take one or more of the fol-
lowing steps:
(1) decrease the value of R,
(2) increase the value of RK,
(3) increase the value of A.
If the goal is to reduce K' then the
opposite course is to be followed.
The results presented here apply to the
simple case of uniform rectangular cells
with uniform earth cover thickness. More
sophisticated cell geometries, non-uniform
cell dimensions and earth cover, and three
dimensional flows will be analyzed as the
study progresses.
Summary and Conclusions
The research project deals with the
development of analytical techniques to
improve the analysis and design processes
for the transmission of liquids through
hazardous waste landfills. The approach
being used is highly versatile, and allows
for the treatment of innovative as well as
conventional waste disposal configura-
tions.
Preliminary results to date indicate
that the approach can be used effectively
to optimize designs in the single configu-
ration (rectangular cells) described.
-------
DEVELOPMENT OF A SOLID WASTE LEACHING PROCEDURE AND MANUAL
Benjamin C. Garrett, Mary M. McKown, Marvin P. Miller,
Ralph M. Riggin, and J. Scott Warner
Battelle Columbus Laboratories
Columbus, Ohio 43201
ABSTRACT
The Resource Conservation and Recovery Act of 1976 requires the U.S. Environmental
Protection Agency to institute a national program to control hazardous waste. One aspect
of this program involves the issuing of permits for operating a hazardous waste disposal
facility. This paper describes the development of a solid waste leaching procedure
manual that could be used in the design of land disposal facilities for hazardous wastes.
The manual details laboratory procedures for extracting or leaching a sample of solid
waste so that the composition of the laboratory leachate is similar to the composition
of leachate from the waste under field conditions. The solid waste leaching procedure in
the manual is summarized in this paper, and the major assumptions and rationale critical
to the procedure are discussed.
BACKGROUND
The U.S. Environmental Protection
Agency will review applications for permits
to operate hazardous waste disposal facili-
ties that are subject to the regulations
set forth in the Resource Conservation and
Recovery Act (RCRA), Section 3004 (40 CFR
Part 264). Those individuals involved in
the permit writing and review processes
will need guidance to aid in evaluating the
performance of a disposal facility.
One aspect of this performance evalua-
tion is the determination of the nature of
the leachate percolating out of the hazar-
dous waste and into either the underlying
soil or a leachate treatment system. This
leachate arises from both the loss of li-
quid present in the waste and the infiltra-
tion of natural moisture into the waste.
Information on leachate composition will be
used in judging the adequacy of a leachate
treatment system. This information will
also be needed for predicting the impact
on groundwater resulting from any leachate
seepage from the bottom of the landfill.
The need for leachate composition data
has prompted a project that will develop a
solid waste leaching procedure and manual.
This paper presents a report on that solid
waste leaching procedure project and out-
lines the currently proposed procedure.
The objectives and assumptions of the pro-
ject are discussed, along with the schedule
for the preparation and revision of the
procedure. The direction and scope of fu-
ture work on this project are outlined.
OBJECTIVES
The primary objectives of this solid
waste leaching procedure project are:
1) The development of a solid waste
leaching procedure applicable to the ha-
zardous waste disposal facility process.
2) The preparation of a manual de-
tailing the routine use of the solid waste
leaching procedure.
3) The preparation of a technical
background document describing the scien-
tific basis and rationale for the steps in
the procedure.
The above objectives point out that the
project involves laboratory studies both to
select a suitable leaching procedure and to
incorporate this procedure into a standard
laboratory manual. This leaching procedure
is being developed for a different purpose
than the Extraction Procedure (EP)
-------
specified in RCRA Section 3001. Although
some of the techniques in the EP may be
used, the two procedures will differ sub-
stantially.
The schedule that has been set for
meeting these project objectives is given
in Table 1. The table shows that the solid
waste leaching procedure manual is sche-
duled to go through three revisions prior
to publication of the final document.
Having several revisions allows both tech-
nical peer review of and public comment on
the proposed manual, as has been done for
the other technical resource documents.
In this way, the manual will be improved to
reflect suggested changes in either the
solid waste leaching procedure itself or
the directions given for carrying out that
procedure.
TABLE 1. PROJECT SCHEDULE
September 17, 1980
Solid Waste Leaching Procedure Manual,
Draft I, Submitted
January 15, 1981
Manual, Draft II, Due
July 31, 1981
Manual, Draft III, Due
January 31, 1982
Manual, Final, Due
Technical Background Document, Due
The manual has already gone through
one cycle of publication and review(7).
This process has resulted in certain im-
provements in the procedure. Future cycles
of publication and review are expected to
involve a wider audience of qualified re-
viewers than did the initial cycle. The
product of this extensive reviewing and
revising process is expected to reflect
current knowledge of leaching behavior of
wastes plus best judgement on the art of
conducting leachate testing.
LEACHING PROCEDURE
Goals
The solid waste leaching procedure
included in the manual is intended to:
• Provide a realistic leachate pro-
file, showing the change in consti-
tuent concentration with amount of
leaching
• Be site specific
• Be applicable to a variety of
solid wastes
• Be usable by minimally trained
technicians
• Be understandable by and seem
reasonable to supervisory and
management personnel lacking a
scientific background in solid waste
leaching .
Each of these goals reflects the need to
have the manual itself be a working docu-
ment for those personnel involved in the
hazardous waste disposal facility design
and permit application and review pro-
cesses. These people need a laboratory
procedure that will generate a leachate
which accurately predicts the site leach-
ate composition during the life of the
waste, hence the goal of providing a
realistic leachate profile. Further, the
leaching procedure must be site specific
so that the procedure can be tailored to
reproduce the conditions associated with
some identified disposal site.
The manual is likely to be used for
testing of any sort of solid waste that
is being evaluated for landfill disposal.
Therefore, the leaching procedure must be
applicable to the wide variety of wastes
which might be encountered in such testing.
The manual has been developed to in-
struct minimally trained technicians.
Although such individuals can be assumed
to be knowledgeable about general labora-
tory practices, they may be unfamiliar
with techniques used in leachate testing.
Furthermore, even though the solid waste
leaching procedure might be carried out
by technicians, certain management per-
sonnel will be entrusted with the respon-
sibility for supervising the actual con-
duct of this procedure and for interpreta-
tion of the results. These personnel must
understand the rationale for the leaching
procedure specifications given in the man-
ual. Consequently, the manual contains
guidance to aid in identifying and under-
standing the critical aspects of the waste
leaching procedure.
10
-------
The experiences of the users of the
manual will indicate whether the goals of
the solid waste leaching procedure have
been met. The previously discussed sche-
dule for reviewing and revising the manual
has been set up to facilitate meeting these
goals.
Development
The development of a leaching proce-
dure for inclusion in the manual involved
two major stages prior to drafting a spe-
cific procedure. Initially the reported
leaching procedures as well as the general
knowledge of leachate behavior were re-
viewed. Next the assumptions and limita-
tions peculiar to the hazardous waste dis-
posal facility environment under considera-
tion were identified. Once these stages
had been completed, appropriate solid waste
leaching procedures were evaluated.
LEACHATE REVIEW
Considerable research into the chemi-
cal and physical nature of leachates, pri-
marily from municipal wastes, has taken
place within the past two decades, and nu-
merous leaching procedures have been re-
ported. Comprehensive reviews of both
leachate behavior and leaching procedures
are available(ll,14). Ideally, the
leaching medium and test conditions used
in a leaching test should reproduce the
actual leachate and conditions to be en-
countered at the field disposal site.
The leachate percolating through a
particular waste reflects the composition
of all the materials through which that
leachate has travelled and depends on such
site characteristics as annual rainfall
volume and composition, evapotranspiration,
biological activity, and the nature of the
surrounding soil and wastes(11). Because
of the diversity of the site conditions
that are encountered in the field, no sin-
gle leaching medium can completely repro-
duce the composition of the actual leachate
to be found in all situations. However,
the factors that influence leaching medium
composition have been enumerated (Table 2),
and their relative influence on leaching
activity can be estimated.
The predominant characteristics of a
leaching medium that directly relate to
leaching ability are the proton and elec-
tron environments(10,14,19), and the pre-
TABLE 2. CRITICAL FACTORS IN A
LEACHING PROCEDURE
I. Leaching Medium Composition
A. Proton and Electron Environment
1) pH
2) Redox potential (pe or Eh)
3) Ionic strength
4) Buffering capacity
B. Presence of Solubilizing Agents
1) Complexing and chelating agents
2) Colloidal constituents
3) Organic constituents
II. Leaching Test Conditions
A. Contact Area/Particle Size
B. Method of Mixing
C. Mixing Time
D. Temperature Control
E. Number of Leachings on the Same
Solid
F. Number of Leachings on the Same
Liquid
G. Solid to Liquid Ratio
sence of solubilizing agents(16,17). The
proton and electron environments have been
determined for natural environments and
landfill leachates by measuring the pH,
redox potential, ionic strength, and buf-
fering capacity(1,2,13). Solubilizing
agents include constituents such as com-
plexing and chelating agents (hydroxyl ion,
ammonia, EDTA), colloidal constituents
(micelles or surfactants), and organic con-
stituents (melanic materials, humic acids).
Some of these agents can have a profound
effect upon the mobility of inorganic and
organic constituents of the waste, even
when the agents are present at low concen-
trations. Although most analyses of field
leachates have failed to examine the possi-
ble presence of such agents, some more
recently developed leaching media have in-
corporated solubilizing agents(12,20).
Two groups of leachate behavior
reports were especially useful in selecting
candidate leaching procedures for further
evaluation. The report of Lowenbach pro-
vides a detailed summary of the test condi-
tions for over 30 proposed leaching proce-
11
-------
dures. These leaching tests were evaluated
by Lowenbach with regard to the critical
factors given in Table 2.
The studies of Ham and associates pro-
vide a comprehensive evaluation of the
effects produced by systematically varying
each of the leaching conditions given in
Table 2.
ASSUMPTIONS AND LIMITATIONS
Three assumptions have been made re-
garding the solid waste and its associated
disposal facility that are being evaluated
by means of this solid waste leaching pro-
cedure manual: (1) the disposal facility
is a monof ill, (2) the site is engineered to
be secure, and (3) information is available
on site conditions.
Monofill
The solid waste leaching procedure is
oriented toward a monofill situation in
which a single type of solid waste is dis-
posed at the site. Any waste type will be
compatible with the procedure, providing a
representative sample is obtained. The
extension of the procedure to mixed fills,
in which a second type of solid waste is
disposed on top of an existing waste, is
being studied.
Secure Site
A secure site is engineered to control
the movement of water around, into, through
and out of the site. The rate of movement
may be controlled depending on the extent
of protection that is desired. This as-
sumption of a secure site seems reasonable
and allows the expected infiltration of
moisture during operation and prior to
final capping of the fill to be predicted
reliably. What the site is designed to do
must be known in order to make reliable
predictions.
Site Data
The manual user is assumed to have
access to disposal site specific informa-
tion for the site where the solid waste
will be disposed. This information in-
cludes meteorological and engineering data.
Meteorological data involves average and
extreme monthly precipitation volumes and
precipitation composition. The disposal
site engineering data that are likely to be
needed and are assumed to be available in-
clude the expected surface area and depth
of the site, the anticipated schedule for
disposal activities, and the intermittent
soil cover to be used. These site data are
needed for adequate interpretation of the
test results.
METHOD
The solid waste leaching procedure
developed for use in the manual is shown
schematically in Figure 1. This procedure
closely resembles Procedure R of the stan-
dard leaching test developed by Ham, et al.
(10,11) and involves repetitively leaching
the same sample of solid waste, using a
fresh portion of leaching medium for each
repetition.
Procedure R was designed principally
to estimate the total amount of leachable
species to be released from a unit mass of
solid waste. The users of this manual for
testing solid wastes will be interested not
only in the total amount released but also
in the profiles of the leachate constitu-
ents. The profile for any leachate consti-
tuent will indicate the concentration or
mass of that constituent likely to be pre-
sent in the leachate and the time period,
in terms of total volume of leachate pro-
duced, when that constituent will be pre-
sent at any particular concentration or
mass.
The method of Procedure R allows this
sort of profile to be generated and, hence,
was taken for further development with the
manual. However, certain changes have been
made to tailor that method to fit the in-
tended uses of the manual. The product of
these changes is the leaching procedure
whose test conditions are described below.
TEST CONDITIONS
A representative sample of the solid
waste under consideration for landfill dis-
posal is used together with a sample of the
soil if any is to be used in the fill as an
intermittent cover.
A significant aspect to this procedure
is the requirement to include a sample of
the intermittent cover soil along with the
solid waste for testing. This soil should
be of the same type and be present in the
12
-------
SAMPLE OF SOLID WASTE AND
INTERMITTENT SOIL COVER
EXTRACTION 1
ADD LEACHING MEDIUM IN
RATIO OF 1 Gto 10 mL
LIQUID/FILTRATE
EXTRACTIONS 1-4
ANALYSIS
EXTRACTIONS 1-4
MIX 24 HOURS
EXTRACTIONS 2-4
EXTRACTIONS 1-4
SEPARATE SOLID FROM LIQUID
EXTRACTION 4
RETAIN PENDING
RESULTS
EXTRACTIONS 1-4
•INTERPRETATION
OF RESULTS
SOLID/FILTER CAKE
EXTRACTIONS 1-3
RETAIN FOR NEXT
EXTRACTION
EXTRACTIONS 2-4
ADD SAME VOLUME OF
LEACHING MEDIUM AS"
IN EXTRACTION 1
FIGURE 1. SOLID WASTE LEACHING PROCEDURE FLOW SCHEME
13
-------
same proportion as that which will be en-
countered at the disposal site. The abili-
ty of soils to alter the mobility of spe-
cies in leachates is well known and has
been studied extensively for various types
of soils and wastes. For example, the
extent to which trace inorganic constitu-
ents of leachates are attenuated is di-
rectly related to the amounts of clay and
iron and manganese hydrous oxides in the
soil(6,9); and attenuation of organic com-
pounds such as PCBs is directly related to
the organic carbon content of the soil(8).
Furthermore, the soil type influences the
complexing agents and colloidal consti-
tuents present in a leachate and soil mix-
ture. Because of the importance of the
amount and type of soil that is mixed with
a solid waste, this procedure specifies
including the soil with the waste in the
sample to be leached. A possible exten-
sion of this procedure is to use it with
various soil types in order to select the
best type and optimal amount for inclusion
with the solid waste. Obviously, where
the solid waste disposal protocol calls
for the use of no soil during active fil-
ling, then no soil is used with the solid
waste sample for leaching.
Leaching Medium
Distilled, deionized water is used
unless another medium can be justified on
a site specific basis.
The choice of water as the leaching
medium is consistent with the assumption
of a monofilled solid waste. Where envi-
ronmental conditions warrant, the use of
alternate media, such as one to duplicate
acid rain, might be justified.
Solid to Liquid Ratio
The solid to liquid ratio should be
one to ten (weight/volume on a wet weight
basis). The true solid to liquid ratio
that a solid waste will experience is
highly site dependent and very difficult
to forecast precisely. In most cases the
ratio will be one of a large amount of
solid per unit volume of leachate. The
specified ratio of 1:10 does not truly
reflect the likely field conditions; ra-
ther it is a workable amount that will
allow sufficient liquid for proper mixing
and constituent analysis.
Time Per Leaching
The approximate time per leaching is
24 hours. The time specified for each
leaching should ideally be sufficient to
allow equilibrium to be attained. However,
due to the diversity of constituents and
effects, no reasonable time per leaching
is likely to be satisfactory for all situa-
tions. Therefore, the specification of
leaching time has to be made out of consi-
deration of factors other than attainment
of equilibrium. A time of approximately
24 hours is normally convenient for labora-
tory scheduling and is consistent with the
time specified for other related leaching
procedures(11,18).
Numbers of Leachings
Four leachings should be sufficient to
allow definite trends in a leachate con-
stituent level (increasing, decreasing, or
no change) to be noticeable. However, the
solid waste plus soil sample used for the
four leachings is retained pending the
analyses of the filtrates and interpreta-
tion of the results. In the event the
results warrant further leachings, then
the leaching cycle can be started with the
saved sample and can be continued for as
many repetitions as desired.
Temperature
The temperature should be normal room/
laboratory temperature. The temperature
has a decided effect upon the solubility,
rate of reaction, and, perhaps, leaching
of most species. Although ambient tempera-
tures to be expected at land disposal sites
range from extremely cold (^-40°C) to very
hot(^45°C), the temperatures for the
leachates associated with these sites are
likely to be less varied. The overlaying
soil and waste layers with which the
leachate is associated have a dampening
effect on variations in temperature(S).
Consequently, the temperature for the
leachate emerging from the bottom of a
disposal site is likely to be that of the
soil at the same depth. The limits on
seasonal fluctuations in soil temperature
at various depths are probably obtainable
from disposal site data or can be mea-
sured during preliminary site investiga-
tions. If the expected temperatures dif-
fer substantially from the range of normal
laboratory temperatures, then the use of
other temperatures is justified. In any
event, the temperature used in the
14
-------
leaching procedure should be close to that
expected for the site leachate.
Mixir.g Technique
Any mixing device that will impart
sufficient agitation to the mixture such
that stratification of the leaching medium
-sample mixture is avoided and sample sur-
faces are continuously brought into con-
tact with the leaching medium.
The method of agitating or mixing the
leaching medium-sample mixture has been
studied to determine which method offers
the best combination of ease of operation
and extent of liquid-solid contact. Ham,
et al., investigated five methods of
mixing: mechanical and manual shaking;
mechanical stirring; swing shaking (180
degree swing); and variable pitch rotary
mixing. Their results showed comparable
levels for inorganic constituents using
the various methods. However, thier recom-
mendation was to use the rotary mixer be-
cause visual observations of the methods
showed that the other methods occasionally
failed to wet the waste uniformly(ll).
Epler, et al., of Oak Ridge National
Laboratories have studied mechanical
stirrers and have found that the major pro-
blems included: binding of the stirring
blade by solid particles; movement of the
stirred vessel; stalling of the stirring
motor; uneven blade alignment; and sample
grinding(3). They have designed a stirrer
that eliminates or reduces these problems
(4); this unit is to be evaluated for
suitability for use in the extraction pro-
cedure (EP) of the U.S. EPA Toxicity Test.
A recent study completed for the U.S.
EPA compares the operation of the stirrer
and rotary mixer (NBS-design tumbler).
The results show that the rotary mixer
yields better precision than the stirrer
(15).
The specification given in the solid
waste leaching procedure in the manual
follows that contained in the EP of the
U.S. EPA Toxicity Test(18). Currently
only the rotary mixer meets these criteria
for preventing stratification and ensuring
continuous liquid-solid contact; however,
other methods for mixing are being evalu-
ated for approval for use with this EP.
Sample Particle Size and
Surface Area
The surface area should be ^_3.1 cm2/g
or sized to pass through a 9.5 mm standard
sieve, unless it is a monolithic waste.
The sample particle size and surface
area to be used in this procedure is con-
trolled by requiring the sample to be re-
duced until it meets either the sieving
test or the specified surface area of 3.1
cm2/g or greater. This requirement is
designed to approach the conditions likely
to be encountered in the field disposal
environment due to mechanical filling
operations and weathering, and it is con-
sistent with the Federal EP. Some wastes
are naturally monolithic. These will not
have their particle size reduced as this
would cause them to be more leachable than
under field conditions. Any waste passing
the Structural Integrity Procedure (as
given in Reference 18) will be considered
to be monolithic and will be tested as a
whole rather than at a reduced particle
FUTURE WORK
Plans for future work on this project
include
1) Revising the drafts of the manual
in response to changes that are suggested
during the public and peer review periods.
2) Continuing laboratory studies to
provide the scientific basis and rationale
for the leaching procedure, especially
a) development of an internal
standard for users to check grasp
of technique
b) determining the significance of
layers of same waste but of dif-
ferent ages
c) studying the effect on leachate
composition of passage of water
through soil cover before contact
with the waste.
3) Collecting data from other field
and laboratory studies for validating the
procedure.
Changes to the proposed solid waste
leaching procedure result from both inter-
nal comments and external review. As our
own experience with solid waste leaching
tests increases, we can anticipate altera-
tions that might be suggested. Other
changes may come from the general scien-
tific community due to its growing experi-
ence with the required Federal extraction
procedure that is a part of RCRA. The
Federal EP and this solid waste leaching
procedure differ in certain significant
15
-------
aspects of both their intended use and
their experimental design. For example,
the EP is a fundamental part of the re-
quired process for determining whether a
waste is hazardous within the meaning of
RCRA. The EP was designed to consider co-
disposal situations and, hence, uses an
acidic leaching medium. In contrast, this
solid waste leaching procedure is intended
to be an adjunct to the hazardous waste
disposal facility permitting process and
is aimed toward assessing the nature of
the leachate to be produced from that dis-
posal site. Co-disposal with municipal
waste has been purposefully excluded from
the experimental design of this procedure.
However, both procedures involve similar
basic concepts regarding leachate produc-
tion. Because of these similarities,
other changes may come about as the gen-
eral scientific community gains experience
with the EP. Increased testing and evalu-
ating of that procedure when used on a
wide variety of solid wastes and by many
different kinds of users may point out
improvements that should be adopted not
only for the EP but also for this solid
waste leaching procedure.
A substantial area for future work
involves continuing the laboratory studies
that are investigating the validation and
verification of this procedure. In par-
ticular, the correlation between the com-
position of a laboratory-generated leach-
ate from a specific waste and that of a
leachate generated from the same waste
when it is leached under large-scale pilot
conditions or under field conditions is
being investigated.
As a part of this correlation study,
field samples from an existing solid waste
disposal site have been collected. The
samples collected included disposal site
leachate, solid waste, and groundwater
from perimeter wells. The solid waste has
been subjected to one solid waste leaching
procedure; the results are inconclusive.
Further work with these samples is neces-
sary prior to interpreting the results.
Samples from other solid waste sites are
being sought.
In addition, controlled column leach-
ing studies of metal finishing wastes are
to be initiated by lERL-Cincinnati in 1981.
Our participation in these studies will
consist of using the proposed solid waste
leaching procedure to generate leachate
constituent profiles which can be compared
to the profiles for the leachate collected
from the columns.
Positive correlation between labora-
tory leaching tests and field leaching is
an essential part of the development of
the leaching procedure. Such correlation
will allow those individuals involved in
the hazardous waste disposal facility
permitting process to have confidence that
the results of the laboratory test are
indicative and representative of what will
occur in the field.
ACKNOWLEDGEMENTS
The work which is reported in this
paper is being performed under Contract
Number 68-03-2970, "Leaching Procedure
Manual", with the U.S. Environmental Pro-
tection Agency, Municipal Environmental
Research Laboratory, Cincinnati, Ohio.
The authors gratefully acknowledge
the support and help of Dr. Mike H.
Roulier, Project Officer.
REFERENCES
1. Baas Beckling, L.G.M., I. R. Kaplan,
and D. Moore. 1960. Limits of the
natural environment in terms of pH
and oxidation-reduction potentials.
J. Geol. 68:243-284.
2. Chian, E.S.K. and F. B. deWalle. 1977.
Evaluation of leachate treatment.
Volume 1 Characterization of Leachate.
EPA-600/2-77-186a, U.S. Environmental
Protection Agency, Cincinnati, Ohio.
226 pp.
3. Epler, J. L., et al. 1979. Toxicity of
Leachates. IAG No. DOE-IAG-40-646-77/
EPA-IAG-78-D-X0372. Interim progress
report to U.S. Environmental Protection
Agency. Oak Ridge National Laboratory,
Oak Ridge, Tenn.
4. Epler, J. L., et al. 1980. Toxicity of
Leachates. EPA-600/2-80-057. U.S.
Environmental Protection Agency,
Cincinnati, Ohio. 142 pp.
5. Fluker, B. J. 1958. Soil temperature.
Soil Sci., 86:35-46.
6. Fuller, W. H. 1978. Investigation of
Landfill Leachate Pollutant Attenua-
tion by Soils. EPA-600/2-78-158.
U.S. Environmental Protection Agency,
16
-------
Cincinnati, Ohio. 239 pp.
7. Garrett, B. C. , M. M. McKown, M. P.
Miller, R. M. Riggin and J. S. Warner.
1980. Manual for Solid Waste Leaching
Procedure. Contract No. 68-03-2970.
Draft report to U.S. Environmental Pro-
tection Agency. Battelle-Columbus
Laboratories, Columbus, Ohio.
8. Griffin, R. A. and E.S.K. Chian. 1980.
Attenuation of Water-Soluble Polychlo-
rinated Biphenyls by Earth Materials.
EPA-600/2-80-027. U.S. Environmental
Protection Agency, Cincinnati, Ohio.
101 pp.
9. Griffin, R. A. and N. F. Shrimp. 1978.
Attenuation of Pollutants in Municipal
Landfill Leachate by Clay Minerals.
EPA-600/2-78-157. U.S. Environmental
Protection Agency, Cincinnati, Ohio.
157 pp.
10. Ham, R. K., M. A. Anderson, R. Stan-
forth and R. Stegmann. 1978. The de-
velopment of a leaching test for indus-
trial wastes. In: Land Disposal of
Hazardous Wastes, Proceedings of the
Fourth Annual Research Sumposium, San
Antonio, Texax, March 6-8, 1978. EPA-
600/9-78-016. U.S. Environmental Pro-
tection Agency, Cincinnati, Ohio.
pp 33-46.
11. Ham, R., M. A. Anderson, R. Stegmann
and R. Stanforth. 1979. Background
Study on the Development of a Standard
Leaching Test. EPA-600/2-79-109. U.S.
Environmental Protection Agency,
Cincinnati, Ohio. 274 pp.
12. Ham, R. K., M. A. Anderson, R. Steg-
mann and R. Stanforth. 1979. Compari-
son of Three Waste Leaching Tests.
EPA-600/2-79-071. U.S. Environmental
Protection Agency, Cincinnati, Ohio.
234 pp.
13. Johansen, 0. J. and D. A. Carlson.
1976. Characterization of sanitary
landfill leachates. Water Research,
10:1129-1134.
14. Lowenbach, W. 1978. Compilation and
Evaluation of Leaching Test Methods.
EPA-600/2-78-095. U.S. Environmental
Protection Agency, Cincinnati, Ohio.
Ill pp.
15. McKown, M. M., J. S. Warner, R. M.
Riggin, M. P. Miller, R. E. Heffelfinger
B. C. Garrett, G. A. Jungclaus and T. A.
Bishop. 1980. Development of Methodolo-
gy for the Analysis of Solid Wastes.
Contract No. 68-03-2552. Draft final
report to U.S. Environmental Protection
Agency. Battelle-Columbus Laboratories,
Columbus, Ohio.
16. Means, J. L., D. A. Crerar and J. 0.
Duguid. 1978. Migration of radioactive
wastes: radionuclide mobilization by
complexing agents. Science, 200:1477-
1481.
17. Means, J. L., T. Kucak and D. A. Crerar.
1980. Biodegradability of NTA, EDTA,
and DTPA and environmental implications.
Envir. Poll., Series B, 1:45-60.
18. Office of Solid Waste. 1980. Test
Methods for Evaluating Solid Wastes.
SW-846. U.S. Environmental Protection
Agency, Washington, D.C.
19. Rai, D., R. J. Serne and J. L. Swanson.
1980. Solution species of plutonium in
the environment. J. Environ. Qual.,
9:417-420.
20. Stanforth, R., R. Ham and M. Anderson.
1979. Development of a synthetic muni-
cipal landfill leachate. J. Water Poll.
Control Fed., 51:1965-1975.
17
-------
BEHAVIOR OF Cd, Ni, AND Zn IN SINGLE AND MIXED
COMBINATIONS IN LANDFILL LEACHATES
W.H. Fuller, A. Amoozegar-Fard, E. Niebla, and M. Boyle
Soils, Water and Engineering
University of Arizona
Tucson, Arizona 85721
ABSTRACT
This describes a segment of a research orogram designed to develop mathematical models
that can be used as field-oriented tools to nredict migration rates through soils of metals
contained in landfill-type leachates. In this phase, the concepts of the Lapidus and
Amundson model were used along with actual data collected from soil columns experiments
conducted under controlled laboratory conditions. Simple mathematical equations are pre-
sented for predicting migration rates of Cd, Ni, and Zn contained singly and in mixed com-
binations in landfill-type leachates through soils of variable physical and chemical char-
acteristics. The migration behavior of Cd, Ni, and Zn was not significantly different
whether contained singly or in mixed combinations in the leachates. Field use is dis-
cussed.
INTRODUCTION
The Lapidus and Amundson mathematical
model has shown promise as a concept in the
prediction of movement of Cd, Ni, and Zn
from landfill-type leachates through soils
(O'Donnell et al., 1977 and Fuller et al.,
1979). The cadmium prototype was presented
in an earlier paper (Fuller et al., 1979)
in the attempt to provide a user-oriented
predictive tool. Refinements in the math-
ematics of the multitude of factors influ-
encing metal migration rates were necessary
before the final presentation of equations
for Cd, Ni, and Zn could be made with con-
fidence. No two metals respond (attenuate)
identically to the same set of soil and
leachate characteristics. Moreover, the
question of Cd, Ni, and Zn interaction had
to be resolved. Although a great volume of
data relating to this question vias avail-
able earlier, computer testing was incom-
plete.
Soil and broad leachate properties
that most influence Cd, Ni, and Zn attenu-
ation have been quantitatively measured as
previously described in detail (O'Donnell
et al., 1977; Korte et al., 1976: and
Fuller, 1977, 1978). These measurements
are in the form of stochastic functions of
specific properties of soil - leachate-
metal systems. The plan is to match theo-
retical curves describing selected metal
movement in soil columns under controlled
conditions of saturated flow with actual
data. Such a match gives rise to specific
values for the theoretical parameters in
the Lapadus and Amundson model. These
parameters can be translated directly into
a migration rate for a particularelement
(metal).
Variables introduced into the multiple
regression analysis are: clay, silt, sand,
hydrous oxides of iron called "free iron
oxides" (FeO), Total Organic Carbon (TOC),
soluble salts (SALT) primarily the inor-
ganic salts of Ca, Mg, Na, K, Cl, and SO^,:
, silt2, (FeO)2, (TOC)S salts clay x
clay x FeO, clay x sand, clay x TOC,
clay'
silt,
silt x FeO, silt x TOC, silt x salt, FeO x
TOC, sand x silt, sand x FeO, sand x TOC,
sand x salt.
OBJECTIVE
The objective is to determine the
long-range, steady-state migration rates of
Cd, Ni, and Zn contained in landfill-type
leachates and to arrive at simple formulas
that will predict migration rates. This
18
-------
objective was extended to include single
and multiple metal enrichment systems for
Cd, Ni, and Zn.
APPROACH
Because of the great variabilities
associated with soil and geologic materials
under field conditions, the examination of
metal migration is difficult and time con-
suming. Useful data for screening purposes
may be obtained through the use of soil
columns in the laboratory (Fuller, 1977,
1978; Fuller et al., 1979), while avoid-
ing the multitude of problems of field
conditions. The user-oriented predictive
equations may then be tested by field
experimentation for further verification
and necessary refinements. Our approach,
therefore, is to use soil column studies
in association with mathematics as ap-
plied to the concepts of the Lapidus and
Amundson model.
Many models can be used to describe
metal transport through soil (Van
Genuchten, 1978; Van Genuchten and
Wierenga, 1976; and Fuller et al., 1979).
All have some unknown parameters which
must be evaluated from experimental data.
When the solution is complicated, tedio-us
computer work is required. The Lapidus
and Amundson model and its solution has
an advantage in simplicity relative to
others we have considered (O'Donnell et
al., 1977) and it has an analytical repre-
sentation in contrast to Davidson et al.
(1975), and Selim and Mansell (1976).
These latter authors considered solute flow
in finite columns where the solute and ad-
sorbate followed as instantaneous equilib-
rium relationship. Skopp and Warrick (1974)
considered the soil to be composed of a
mobile and stagnant region where the solute
adsorption is a diffusion controlled pro-
cess. Van Genuchten and Wierenga (1976)
considered the soil to be composed of
several regions. The solute adsorption,
however, was considered to be a function of
the concentration gradient between the
mobile and stagnant regions. Models by
Selim and Mansell (1976), and Davidson et
al. (1975), had numerical solutions.
The Lapidus-Amundson mathematical
model is:
where:
c = concentration in soil solution (M/L3)
v = convective (pore water) velocity (L/T)
D = diffusion coefficient (L2/T)
a = fractional pore volume of soil (L3/L3)
n = amount adsorbed per unit volume of soil
(M/L3)
z = vertical distance (L)
t = time (T)
The term 3n/3t describes a linear, non-
equilibrium adsorption and is assumed to be
3n/3t =
- K2n
(2)
where K, and K? are forward and backward
reaction^ rates (1/T), respectively.
c = c
The boundary and initial conditions
steady input concentration (c ) are:
> z = 0 , t > 0 (3a)
t =
The solution to Eqs. (1) and (2) sub-
ject to (3a) and (3b) is
c = CQexp(vz/2D)[F(t)
where
/tF(t)dt] (4)
l£ . 1 3n_ _ n 32c 3c
3t a 3t 3Z7 9z
(D
______
F(t) = [exp(-K t)] / {I [2 A,K93(t-B)/a]
? ° ' i
exp[- zV4D3 - 3d]-[z/2ADpT>d3
with d = v2/4D + K,/a - Ko. IQ is the modi-
fied Bessel function of the first kind of
order zero. The term c/c0 is referred to
as relative concentration, C, and is a
function of z and t, and contains the pa-
rameter v, a, D, K.J and K2<
The Lapidus-Amundson model was formu-
lated to describe the effect of longitudi-
nal diffusion in ion exchange and chromato-
graphic columns, where the chemical and
physical parameters are uniform and well
controlled; and the system is relatively
simple. In a soil-leachate system, the
number of parameters increases greatly,
19
-------
and their interrelationships are much more
complex. The advantage of the Lapidus-
Amundson model is that it combines all
unknown or immeasurable parameters into
three unknowns, D, K-,, l<2, which can be
determined from the experimental data as
described below. It would be impossible
to determine the spatial interrelation
between all the different physical and
chemical sites in the soil-leachate sys-
tem. By applying the Lapidus-Amundson
model, the net effect of the chemical na-
ture of the leachate on the forward and
backward reaction rates at a multitude of
different adsorption sites is combined in-
to an effective forward and backward
reaction rates for the particular conditions
of the soil-leachate system in the experi-
ment. Similarly, an effective diffusion co-
efficient is determined, or a diffusion-dis-
persion coefficient since the effects of dif-
fusion-dispersion combine in one variable.
Miscible displacement theories and
experimental data indicate the solute
profile is not one of piston displacement,
but rather is a smooth distribution of
concentrations (Nielsen and Biggar, 1962).
In addition, different relative concentra-
tions appear to travel at different rates
(Figure 1). Since the velocity of a par-
ticular concentration ratio is a function
of distance, it follows that the shape of
the experimental breakthrough curve will
also be a function of z. However, the
model predicts that for z > 10 cm the
velocity of a particular relative concen-
tration approaches an asymptotic value;
i.e., steady-state velocity for any rela-
tive concentration has been achieved by a
depth of 10 cm (Figure 1).
Eight soils representing five of
ten major soil orders were used to make
10-cm soil columns which were leached with
two municipal solid waste leachates
(Fuller, 1978) whose salt contents had been
altered to varying degrees.
'C/Co - C4
Steady-state velocities are
the slopes of these lines
(z > 10 cm)
C/Co = C5
T1ME" *1 '„
Figure I. Trajectories of different relative concentrations. (Plot of z = z(t),
where z = z(t) satisfies c(t,z(t)) = C., i = 1,2,3,4,5.
20
-------
In the application of the model, v is
estimated by vj/a where v 0, the following is
true:
C(z(t), t) = C'
Thus, for any time t > 0, z(t) gives the
location of the relative concentration C'.
The migration rate of C' would then be
given by dz/dt.
The function C(z(t), t) = C1 is comoli-
cated, therefore it is not possible to
solve for z(t) directly. However, one can
find the values of z vs. t for each c'=0.1,
0.2, ..., 0.9 by interpolation. A computer
program was developed to find the steady-
state migration rates of c'. Using Eq. 4,
an array of c' vs. -z was calculated for 40
different times from 0.1 to up to 200 days.
The spline function (Erh, 1972) was then
used to calculate the value of z vs. t for
each one of the c'=0.1, 0.2, ..., 0.9 giving
a series of points along the curve z = z(t),
Figure 1. Note that in general z = z(t) is
not a straight line, however, for depths
greater than about 10 cm, the curve becomes
linear for all practical purposes. That is,
it achieves a steady-state condition.
By estimating the steady-state velocity
using the "linear" portion of the curve,
we can determine the effect of change in
the soil-leachate system on migration rates
using multiple regression analysis. The
main characteristics of the soils and
leachates have been measured to identify
the effect of soil and leachate properties
on Cd, Ni, and Zn attenuation. The soil
parameters are, a) percent sand, silt, and
clay, and b) free iron oxides (FeO); for
leachate they are a) total organic carbon
(TOC), and b) total inorganic salts of Ca,
Mg, Na, and K.
All the migration rates were adjusted
to pore water velocity of 25 cm/day. The
average pore water velocity for the experi-
ment was about this value. It was observed,
although not in all cases that when calcu-
lating the migration rates for each element
by changing the pore water velocity, the mi-
gration rates for all relative concentrations
would change accordingly. In other words
most of the velocities follow the simple
rule of being halved when the leachate ve-
locity is halved. So we assume that the
velocity of a given concentration ratio is
proportional to the pore velocity of the
leachate and can therefore put the Cd, Ni,
or Zn velocity at a particular relative con-
centration C' and particular pore fluid
velocity, v
Vel(C',v) = v(Al-FACTR-l + A2-FACTR-2 _ ...
+AN-FACTR-N)/25 (5)
The factors FACTR-i, i-1, ..., N, in Eq.
(5) are the particular soil leachate proper-
ties under consideration.
21
-------
The application of this type of
equation is of course only valid if the
migration rate is proportional to the pore
velocity of the carrier fluid. Otherwise
multitude of additional experiments will
need to be undertaken with pore velocity
as an additional variable. But preliminary
experiments indicate that over at least a
limited range, this proportionality is
maintained.
It was determined in the course of our
experiments that it is the ratio of the
forward and backward reaction rates which
is the dominant factor in determining the
breakthrough curve, as opposed to the
absolute magnitude of K] or l<2. A plot of
the values of K-| and K2 which give an ac-
ceptable fit of the theoretical break-
through curve to the experimental curve
results in a straight line with a slope
equal to K-|/l<2. The line segments gener-
ated for each of the soils reported varied
within a narrow range if at all at differ-
ent fluid velocities. For example, when
flux was changed fourfold (i.e., 3 ml/hr to
12 ml/hr) the K-j/Ko ratios varied according
to a common trend but to a limited extent
for the soils involved (see Figure 2).
Such narrow ranges of differences in slope
24
20
16
12
a
4
0
Leachate Flux
1 14.6 cm/da
--<>•- 3.7 cm/da
Kalkaska s
(Michigan)
Wagram Is
(N. Carolina)
Anthony si
(Ariz.-Calif.)
Ava sicl
(Illinois)
04 8 12 16 20 24 28 32
Fanno c
(Ariz.-Calif.)
i i i i
04 8 12 16 20 24 28
K.
Figure 2. Line segments for six soils relating forward (K-|) and backward (Kg) reaction
coefficients as influenced by flux, where: Rate of Adsorption
3n/3t = K-C - Kn (Eq. 2).
22
-------
of line segments as a result of differences
in fluid velocities are not universal for
all soils. Nicholson sic and Molokai c are
notable exceptions according to soils from
our laboratory not reported here. Never-
the less, for the purpose of model develop-
ment we chose to assume flux is of much
less importance as a variable than other
parameters in the soils used in this re-
search program (Alesii et al., 1979).
The flux dependency of the diffusion
coefficient also needs to be considered.
A true diffusion coefficient should be inde-
pendent of pore velocity (Bresler, 1973).
But since the D term in the Lapidus-Amund-
son equation is actually a combination dif-
fusion-dispersion term, there will be some
flux dependence since the dispersion co-
efficient is a function of pore velocity.
The diffusion coefficient has been shown to
be independent of the salt concentration
and dependent on only the water concentra-
tion for all practical purposes in soil-
water systems (Porter et al., 1960; Kemper
and van Schaik, 1966). Kemper and van
Schaik (1966) propose a functional relation-
ship between the diffusion coefficient, D,
and the volumetric water content, 6, as
being Dp(6) = D ae^9; where a and b are
enpirical constants, and DQ is the diffusion
coefficient in a free-water system. This
relationship has been found to hold when
the range of water content cooresponds to
0.30 to 15 bars suction by 01 sen and Kemper
(1968), when b equals 10 and a ranges from
0.001 in clays to 0.005 in sandy loams.
These values for a and b were also success-
fully applied by Bresler (1973), and Melamed
et al. (1977).
Bresler (1973) reports that many in-
vestigators have shown a linear relation-
ship between the dispersion coefficient
D, and the average pore velocity v in the
form D|j(V) = A v| where X is an empirically
determined constant. Bresler (1973) re-
ports values for A of 0.28, 0.39, and 0.55;
while Melamed et al. (1977) use a value of
0.40.
The net result is that the diffusion-
dispersion term, D, is related to D and
D, as follows: p
D(9,V) = Dp(e) + Dh(V)
D(6,V) = Doae
b6
A
(6a)
(6b)
In the experiments which we did, the
values for v were on the order of 0.82
cm/hr in Davidson clay and 1.17 cm/hr in
Wagram loamy sand. Using these flux rates
and the D0 of chloride (0.04 cm2/hr) as a
reference point.it can be seen from the
relation above that Dn is more than an
order of magnitude greater than Dp (0.23
cm2/he vs 0.0048 cm2/hr in Davidson clay).
In the flux range in which the experiments
were run, the equation predicts that D.
describes 98% of D. By decreasing the flux
by an order of magnitude, the Dn term
should still describe 83% of D. We assume
therefore for the purposes of model develop-
ment that the relationship between the dif-
fusion-dispersion coefficient and flux a
remains linear over the range of interest.
Biggar and Nielsen (1976) present a rela-
tion between the apparent diffusion coef-
ficient, D, and pore water velocity v,
D = 0.6 + 2.93 v1'11 with an r value of
0.795.
We regress the soil and leachate prop-
erties having most effect on migration
rates with the migration rates in the fol-
lowing way: The regression equation is
written one variable at a time with all
variables available to bring into the
equation at each step. The first variable
in the equation is that one which does the
most to explain the variation in the migra-
tion rates. The second variable to enter
the equation is that one which does the
most to explain the variation in the migra-
tion rates not already explained by the
first variable. Variables continue to be
added in this way until either all vari-
ables are in the equation or until none of
the unused variables are statistically sig-
nificant in explaining the remaining vari-
ation in the migration rates.
INTERPRETATION OF RESULTS
Cd - 11!4 cases with Ni and Zn
V.I =2^-[32. 5177CLAY + 0.27825 x CLAY
+0.0100 x (CLAY + SILT)2 - 105.6737
SILT + 1.3051 x SAND + 4.5440 x (TOC +
SALT) - 97.2421]
R2 = 0.74
V.5 = £5[28.896/CLAY + 0.24376 x CLAY +
.008847 x (CLAY + SILT)2 - 92.12477
SILT +1.1468 x SAND + 4.20529 x
(TOC + SALT) - 85.6638]
R3 = 0.73
23
-------
V.9 = |5[25.428/CLAY + 0.21597 x CLAY +
.007807 x (CLAY + SILT)2 - 79.56717
SILT + 1.0046 x SAND + 3.8101 x (TOC
+ SALT) - 75.2849]
R2 = 0.73
Cd - 75 cases, no Ni or Zn
V.I = ^[37.2697/CLAY + 9.40178/FeO -
0.0825 x SAND + 0.017126 x (SILT x
FeO) + 21.2165 x SALT + 1.4111 x TOC
- 2.7050]
R2 = 0.82
V.5 = |-5[35.3135/CLAY + 8.0903 /FeO -
0.08001 x SAND + 0.01496 x (SILT x
FeO) + 18.7552 x SALT + 1.47171 x TOC
- 2.0804]
R2 = 0.82
V.9 = |5[33.3546/CLAY + 6.9232 /FeO -
0.076324 x SAND + 0.013806 x (SILT x
FeO) + 16.01493 x SALT + 1.40034 x
TOC - 1.52744]
R2 = 0.81
Ni - 72 cases with Cd and Zn
V.I = |5[115.0336/CLAY + 0.14880 x CLAY -
82.4514/SILT + 0.000966 x SAND2 -
6.9386 /Fe00+ 4.3635 x TOC - 3.5346"]
R2 = 0.81
v,
V.5 = ^[92.7205/CLAY + 0.12965 x CLAY -
66.9761/SILT + 0.0008 x SAND2 -
5.4369/FeO + 4.1174 x TOC - 3.0682]
R2 = 0.81
Ni - 72 cases with Cd and Zn (contd)
V.9 = |5[80.5799/CLAY + 0.11377 x CLAY -
57.5043/SILT + 0.00068 x SAND2 -
4.6977/FeO + 3.8282 x TOC - 2.7105]
R2 = 0.77
Ni - 44 cases, no Cd or Zn
V.I = ^[-1.9665 x SAND - 0.04175 x (CLAY
x SAND) - 0.07765 x (CLAY x SILT)
0.00771 x SILT2 + 52.8340 x SALT -
164.7779 x SALT^ + 200."7?1
R = 0.89
V.5 = -^[-1.60976 x SAND - 0.03404 x (CLAY
x SAND) - 0.06332 (CLAY x SILT) -
0.006397 x SILT2 + 44.1637 -
135.41869 x SALT^ + 163.9379]
R2 = 0.87
v.9 = |5[-1.3083 x SAND - 0.02824 x (CLAY
x SAND) - 0.05098 x (CLAY x SILT) -
0.005567 x SILT2 + 49.5652 x SALT -
156.0114 x SALT2 + 133.0813]
R2 = 0.78
Zn - 29 cases, no Cd or Ni
V.I = £5[-0.96122 x CLAY + 0.011347 x
CLAY2 + 0.4278 x (CLAY +9SILT) -
0.002640 x (CLAY + SILT)^ + 31.1308
x TOC2 + 11.9237 x SALT + 1.4004]
R2 = 0.75
V.5 = |5[-0.8990 x CLAY + 0.010598 x CLAY2
+ 0.4109 x (CLAY + SILT) - 0.00255 x
(CLAY + SILT)2 + 28.47561 x TOC2 +
10.8695 x SALT + 0.9522]
R2 = 0.75
V.I = |5[-0.89588 x CLAY + 0.010568 x
CLAY2 + 0.4225 x (CLAY +9SILT) -
0.002644 x (CLAY + SILTT + 25.8125
x9TOC2 + 9.7207 x SALT + 0.65902]
R^ = 0.73
Zn - 54 cases with Cd and Ni
V.I = |g-[14.6428/CLAY - 0.14959 x CLAY
- 23.3821/SILT + .0045 x (CLAY x FeO)
- 0.0035519 x (SAND x SILT) + 1.4833
x TOC + 8.1533]
R2 = 0.63
Zn - 54 cases with Cd and Ni (contd)
V.5 = |g{13.7075/CLAY - 0.11745 x CLAY
- 19.824/SILT + 0.003747 x (CLAY x
FeO) - 0.002812 x (SAND x SILT) +
0.76496 x TOC + 6.5722]
R2 = 0.62
V.9 = ^{13.1353/CLAY - 0.09602 x CLAY
- 17.8912/SILT + 0.003123 x (CLAY x
FeO) - 0.002267 x (SAND x SILT) +
0.4699 x TOC + 5.5240]
R2 = 0.61
To use these equations:
1) Estimate Clay, Silt, & Sand (%)
2) Estimate TOC of leachate (%)
3) Estimate total ion content of the
24
-------
leachate (%): This is the sum
of soluble salts and iron in
solution.
4) Estimate free iron oxide content
(?).
5) Estimate the pore velocity of the
leachate (cm/day). (The Darcian
velocity divided by the leachate
filled porosity.)
6) Determine the particular relative
concentration.
7) Substitute the quantities from
steps 1-5 into the equation de-
termined from steo 6.
For example, suppose the soil has the
characteristics:
Clay content = 15%, silt content = 20%,
free iron oxide content - 1.5%. Suppose
the leachate has total soluble ions 0.08%.
If the porosity of the soil was 0.38 and
the infiltration rate was 0.6 cm-Vcrrr/day,
then the pore velocity would be 0.6/.38=1.6
cm/day. Suppose the initial cadmium con-
centration in the leachate is 2 ppm and
the concentration limit was 1 ppm. Then
the relative concentration of interest is
0.5. Substitution into the equation for
V.5 yields
V.5 = [35.31/(15) + 8.09/(1.5) -
.08(65) + .015(20 x 1.5) +
18.76(.08) + 1.47 x TOC(l.l)
- 2.08]
= 0.258 cm/day Alone and 0.291 mixed
To summarize this point, we have
plotted the data from all the columns run
and have matched theoretical curves to the
observed curves to get individual values
for D, K-| and Ko for each column. At this
point the only data that have been used
are values of a, v, c0 and the experimental
breakthrough curve from the column. And
for selected relative concentrations we
have migration rates for each column.
To generalize these results from the
few columns run to a more general setting
and also to display the results in an
easily accessible form we make use of mul-
tiple regression equations. One regres-
sion equation for each of the selected
relative concentrations.
o
Data in Table 1 showing R values were
assembled to show that nearly all of the
variables influencing attenuation v;ere
identified. For the most Dart, 77-90%; 73-
TABLE 1. COMPARISON OF CORRELATION COEFFICIENTS OF VELOCITY (MIGRATION RATE) THROUGH
SOIL OF Cd, Ni, and Zn ALONE AND MIXED IN LANDFILL-TYPE LEACHATES
Velocity
V.I
V.2
V.3
V.4
V.5
V.6
V.7
V.8
V.9
Ni alone
R2
0.8941
0.8895
0.8909
0.8749
0.8684
0.8570
0.8457
0.8231
0.7813
Ni
Ni + Cd + Zn
R2
0.8115
0.8120
0.8137
0.8092
0.8081
0.8051
0.8014
0.7914
0.7743
Cd alone
R2
0.8249
0.8222
0.8227
0.8173
0.8218
0.8196
0.8170
0.8144
0.8125
Cd
Cd + Ni + Zn
R2
0.7377
0.7355
0.7354
0.7323
0.7348
0.7329
0.7314
0.7306
0.7274
Zn alone
R2
0.7512
0.7534
0.7526
0.7521
0.7504
0.7492
0.7429
0.7374
0.7259
Zn
Ni + Cd + Zn
R2
0.6350
0.6286
0.6252
0.6199
0.6164
0.6131
0.6132
0.6129
0.6081
25
-------
83%; and 60-75% of the total variation for
Cd, Ni and Zn, respectively, are described
by the combined influence of independent
variables. Considering the great number of
soil and leachate variables, this unex-
pectedly high achievement is fortunate and
lends confidence in this approach to eval-
uating metal attenuation by soil.
The velocities of Zn through soil
tended to be about the same whether con-
tained singly or in mixed combination in
landfill-type leachates at concentrations
up to 100 ppm of each metal. Velocities
of Cd were not significantly influenced
by Zn but were significantly influenced
by Ni at the 5% level.
Velocities of Ni, on the other hand,
were not significantly influenced by Cd
but were influenced at the 5% level by Zn.
In view of the lack of significance in
four out of six comparisons and the low
level of significance (0.05) in the other
two comparisons, it appears that attenua-
tion is not appreciably influenced by the
presence of these metals in mixed combina-
tions at concentrations up to 100 ppm of
each in the leachates. These findings
might appear to contradict the results re-
ported for the effects of conmon salts in
leachates. The apparent discrepancy is
resolved by virtue of the fact the common
salts in the leachates were compared at
much higher levels of concentration than
were provided for Cd, Ni, or Zn. Also we
have a limited number of cases for the Cd,
Ni, and Zn compared with those of common
salts. For example, the velocity of Cd
was found to increase through a given soil
from 0.345, 0.411 to 0.462 cm/day as the
salt content of a landfill-type leachate
increased from 0.02, 0.05, to 0.08%.
These equations also provide quanti-
tative data demonstrating the comparative
migration rates of Cd, Ni, and Zn under a
variety of environmental conditions. At
V.5, for example, by using the above soil
the relative rates of migration when the
metals are present singly in the solid
waste leachates were 0.462, 0.258, and
0.117 cm/day for Ni, Cd, and Zn, respec-
tively. When all three metals are present
together, the rate of movement of Ni, Cd,
and Zn are 0.442, 0.291, and 0.128 cm/day,
respectively.
LIMITATIONS AND ADVANTAGES OF THE LAPIDUS-
AMUNDSON MODEL APPROACH
Limitations: Caution must be ex-
ercised when using this approximation for
several reasons. First, these were based
on column experiments and not field condi-
tions. Second, although the regression of
velocity with both soil and leachate prop-
erties have an R^ value of high order, con-
siderable variation in the actual velocity
values and those predicted by the regres-
sion equations are still present. Here we
feel that caution should be exercised when
these regression equations are required
for highly accurate predictions of metal
movement. Rather they should be used to
gain some idea of the degree to which the
selected soil and leachate properties
either contribute to or detract from metal
transport velocity.
Refinements in the Model are needed ir
order of most increase in predictive capa-
bility per expenditure of a given amount
of development work. These include:
1) Tests with leachates radically differ-
ent from those used to date.
2) More analyses of sensitivity of output
to character of input so it can be
determined where to exercise most
care in collecting input data and
which experimental conditions to con-
trol most closely in developing the
Model, such as:
—importance of soil bulk density
and pore size distribution.
--nature of total organic carbon.
--identification of the effect of
other hydrous oxides such as Mn
and Al.
--importance of soil-leachate matrix
pH.
Advantages: The Model provides cer-
tain advantages over others:
1) Less complicated than many other
models.
2) Does not require infinite information
of attenuation mechanisms which
would need many years to discover
and evaluate.
3) D, K],and l<2 have been derived as real
values from research experimental
data making it possible to apply the
Model to actual landfill disposal
problems for attenuation of poten-
26
-------
tially hazardous metals. In short,
D, K-|, and l<2 represent real values
beyond armchair theory.
4) The Model is most likely to be applied
in its present form to the evalua-
tion of disposal sites because it is
--independent of computers, as far as
the ultimate user is concerned.
--much easier to use the products in
the form of mathematical equa-
tions or tables,
--less demanding for data than other
prominent simulation model
approaches,
--no more limited by the quality of
input data than other approaches,
and (for a given site) much less
limited by quantity of input
data than other approaches,
--most likely to be used in evaluat-
ing relative migration of a sin-
gle metal at several alternative
sites (because of extrapolation
problems, no claim of predicting
exact movement rate in undis-
turbed soil is made. However,
the Model is very likely to es-
timate accurately metal movement
rate in disturbed soils as might
be found when landfills are lined
with clay-soil mixtures or with
native soils).
5) Concerned with nonconservative solutes
(such as metals) and nonconventional
leachates as well as raw, unspiked
municipal solid waste leachates.
6) Not conflicting with the large hydrol-
ogy modeling programs of the USGS.
7) Conveniently and accurately related to
easily measurable physical and chem-
ical properties of the soil-leachate
system.
ACKNOWLEDGMENTS
This research was supported in part
by the U.S. Environmental Protection
Agency, Solid and Hazardous Waste Research
Division, Municipal Environmental Research
Laboratory, Cincinnati, OH. Grant No. R
805731-01 and the University of Arizona,
Soils, Water and Engineering Department.
Arizona Agricultural Experiment Station
Paper No.
The authors wish to thank Rodney
Campbell and Dan 0'Donne!1 for assistance
during the initial phases of the modeling
program and Bruno Alesii and Nic Korte in
the development of some of the Cd migration
data.
REFERENCES
1. Abramowitz, M., and I.A. Stegun, eds.
1970. Handbook of mathematical
functions. Ninth Printing. U.S.
Department of Commerce. National
Bureau of Standards. Applied Mathe-
matics Series 55, pp. 916, 378.
2. Alesii B.A., W.H. Fuller, and M.V.
Boyle. 1979. Effect of leachate
flow rate on metal migration through
soil. Az. Agr. Expt. Sta. (in press).
3. Alloway, W.H. 1968. Agronomic controls
over the environmental cycling of
trace elements. Adv. Agron 120:257-
274.
4. Biggar,_J.W., and D. Nielsen. 1976.
Spatial variability of the leaching
characteristics of a field soil.
Water Resources Res. 12:78-84.
5. Black, C.A. (ed.) 1965. Methods of
soil Analysis. Part 2. Chemical and
microbial properties. Agron. Monogr.
No. 9. Am. Soc. Agron., Madison, WI.
6. Bresler, Eshel. 1973. Simultaneous
transport of solutes and water under
transient unsaturated flow conditions.
Water Resources Res. 9:(4).
7. Davidon, W.C. 1959. Variable metric
method of minimization. ANL-5990
(Aragonne National Lab. U.S. AEC
Research and Develop. Rep.)
8. Davidson, J.M., D.R. Baker, and G.H.
Brusewitz. 1975. Simultaneous
transport of water and adsorbed sol-
utes through soil under transient
flow conditions. Arer. Soc. Agr.
Engr. Transactions. 18:535-C37.
9. Fuller, W.H. 1978. Investigation of
landfill leachate pollution attenua-
tion by soils. EPA-600/2-78-158.
U.S. Environmental Protection Agency,
Cincinnati, OH.
10. Fuller, W.H. 1977. Movement of selec-
ted metals, asbestos, and cyanide in
soil: Application to waste disposal
problems, 1977. Environ. Protec.
Tech. Ser. EPA-600/2-77-220. U.S.
EPA, Cincinnati, OH 45268.
27
-------
11. Jenne, E.A. 1968. Controls on Mn, Fe,
Co, Ni, Cu, and Zn concentrations in
soils and water: The significant
role of hydrous Mn and Fe oxides.
Adv. Chem. Ser. 73:337-387.
12. Kemper, W.D. and J.C. van Schaik. 1966.
Diffusion of salts in clay-water sys-
tems, Soil Sci. Soc. Amer. Proc. 30:
534-540.
13. Korte, N.E., J. Skopp, W.H. Fuller, E.
E. Niebla, and B.A. Alesii. 1976.
Trace element movement in soil : In-
fluence of soil physical and chemical
properties. Soil Sci. 122:350-359.
14. Korte, N.E., J. Skopp, E.E. Niebla, and
W.H. Fuller. 1975. A baseline study
on trace metal elution from diverse
soil types. Water, Air. Soil Pollut.
5:149-156.
15. Lapidus, 1. and N.R. Amundson. 1952.
Mathematics of absorption in beds.
VI. The effect of longitudinal dif-
fusion in ion exchange and chromato-
graphic columns. J. Phys. Chem. 56:
984-988.
16. Leeper, G.W. 1978. Managing the heavy
metals on land. Marcel Dekker, Inc.
New York. 121 pp.
17. Lisk, O.-J. 1972. Trace metals in
soils, plants, and animals. Adv.
Agron. 24:267-325.
13. Melamed, D., R.J. Hanks, and L.S.
Willardson. 1977. Model of salt flow
in soil with a source-sink term.
Soil Sci. Soc. Amer. J. 41 :29-33.
19. Nielsen, D., and J.W. Biggar. 1962.
Miscible displacement: III. Theoreti-
cal considerations. Soil Sci. Soc.
Am. Proc. 26 216-221.
20. O'Donnell, O.F., B.A. Alesii, J.
Artiola-Fortuny and W.H. Fuller.
1977. Predicting cadmium movement
through soil as influenced by
leachate properties. U.S. EPA
Report (in press).
21. Olsen, S.R. and W.D. Kemper. 1968.
Movement of nutrients to plant
roots. Advan. Agron. 30:91-151.
22. Porter, L.K., W.D. Kemper, R.D.
Jackson, and B.A. Stewart. 1960.
Chloride diffusion in soils as
influenced by moisture content.
Soil Sci. Soc. Amer. Proc. 24:460-
463.
23. Selim, H.M., J.M. Davidson, and R.S.
Mansell. 1976. Evaluation of a
two-site absorption-desorption
.model for describing solute trans-
port in soils. Summer computer
simulation conference.
24. Selim, H.M., and R.S. Mansell. 1976.
Analytical solution of the equation
for transport of reactive solutes
through soils. Water Resources
Research 12:528-532.
25. Skopp, J., and A.W. Warrick. 1974.
A two-phase model for miscible dis-
placement of reactive solutes in
soils. Soil Sci. Soc. Am. Proc. 38:
545-550.
26. USDA. 1954. Saline and alkali soils.
Agri. Handbook #60. U.S. Govt.
Print. Office, Wash. D.C.
27. Van Genuchten, M.T. 1978. Simulation
models and their application to
landfill disposal siting; A review
of current technology. _In_ Land
Disposal of Hazardous Wastes. Proc.
4th Ann. fles. Symp. Ed. D.W. Shultz.
EPA-600/9-78-016. U.S. Environment-
al Protection Agency, MERL,
Cincinnati, OH. 425 pp.
28. Van Genuchten, M.T., and P.J. Wierenga.
1976. Mass transfer studies in
sorting porous media: I. Analytical
solutions. Soil Sci. Soc. Amer. J.
40:473-480.
ALSO
Fuller, W.H., A. Amoozegar-Fard and G.E.
Carter. 1979. Predicting movement of se-
lected metals in soils: Application to
disposal problems. Proc. Fourth Ann. Res.
Symp., San Antonio, TX, March 1978. U.S.
Environ. Protect. Agency. EPA-600/9-79-023a.
Cincinnati, OH 45268. pp 358-374
Erh, K.T. 1972. Application of the spline
function to soil science. Soil Sci. 114:
333-338.
28
-------
AQUEOUS CHEMISTRY AND ADSORPTION OF
HEXACHLOROCYCLOPENTADIENE BY EARTH MATERIALS
Sheng-Fu J. Chou, Barry W. Fisher, and Robert A. Griffin
Illinois State Geological Survey
Champaign, IL 61820
ABSTRACT
The aqueous chemistry, adsorption, and mobility of hexachlorocyclopentadiene (C-56)
in soil materials were studied in the laboratory. The solubility of C-56 in waters, soil
extracts, and sanitary landfill leachates ranged from 1.03 to 1.25 ppm. Sodium hydroxide
and sodium chloride decreased the solubility of C-56 in water; sodium hypochloride
slightly increased its solubility. C-56 underwent rapid hydrolysis at pH 2.51 and slower
hydrolysis at pH 8.45 and 11.73. When exposed to sunlight, the half-life of C-56 was
less than 4 min in aqueous solution and less than 1.2 min in hexane solution. At least
four products of hydrolysis and photolysis were identified. Hexachlorocyclopentenone was
the major product in low pH water, and the diolefins ois- and trws-pentachlorobutadiene
were the major products in mineralized water or high pH water.
Freundlich adsorption isotherm plots of C-56 sorption on soils and clay minerals
yielded linear regression lines with coefficients of correlation (r2) of at least 0.98.
C-56 was found to be readily adsorbed by the soil materials. The adsorption capacity and
mobility of C-56 were highly correlated with the total organic carbon (TOC) content of
nine soil materials (r2 = 0.99) and appear to be predictable from the TOC content of
soils. A further conclusion was that products of the degradation of C-56 appear to
migrate through soils and might cause more of a problem than C-56 itself.
INTRODUCTION
Hexachlorocyclopentadiene (C-56 or
"hex" waste) is a highly toxic compound
that poses a potential occupational hazard
(Hoecker et al., 1977). Fathead minnows
were adversely affected at concentrations
of C-56 as low as 3.7 ppb, and significant
decreases in survival occurred at 7.3 ppb
(Spehar et al., 1979). Because C-56occurs
as a by-product from production of chemical
feed stocks, pesticides, adhesives, resins,
and other related products, large quanti-
ties are produced in the United States
each year. Even with governmental bans on
the use of certain chlorinated pesticides,
there is currently no documentation that
the production of C-56 has subsided. The
reason may be that C-56 is commercially
important as a chemical intermediate,
although it has no end uses of its own
(Bell, Ewing, and Lutz, 1978, and National
Academy of Sciences, 1978).
C-56 is a light yellow liquid. It is
extremely volatile and photoreactive in
sunlight with near-surface half-lives of
less than 10 min in aquatic systems
(Zepp et al., 1979). The empirical formula
of C-56 is C5C16 and its structure formula
is:
ci
ci
The present environmental concern is
the disposal of large quantities of C-56
that is produced annually as a by-product
of pesticide and other manufacturing proc-
esses. Therefore, the study of C-55 waste
29
-------
has broad application to many industries.
The limited amount of information presently
available indicates that C-56 has a fairly
strong affinity for soil. The mechanisms
of attenuation in soil materials are
unknown. Data on the factors affecting
C-56 attenuation by soil materials, its
solubility in waters and landfill leach-
ates, and its mobility in soils would
provide a useful basis for determining
waste treatment methods, for predicting
C-56 migration under landfills, and for
selecting and designing future disposal
sites.
The specific objectives of this study
were to determine the attenuation mecha-
nisms and capacity of selected earth mat-
erials for C-56, to determine the effects
of caustic soda-brine on the attenuation
and solubility of C-56, to develop a chem-
ical model to predict C-56 migration
through soil materials, and to identify
the major degradation products of C-56 in
the environment.
This is a report of work conducted as
part of Grant No. R806335-010 from the U.S.
Environmental Protection Agency, Cincinnati,
Ohio. The results should be considered pre-
liminary and subject to reinterpretation if
future results require it. Mention of
trade names does not constitute endorsement.
CURRENT RESEARCH MATERIALS AND METHODS
Reagents
Reagent grade C-56 was obtained from
Pfaltz and Bauer, Inc., Stamford, CT, and
was used without further purification. An
analytical standard of C-56 (lot #0213) was
obtained from U.S. EPA, Research Triangle
Park, NC. Both materials were identical
and the purity was approximately 98%. Ana-
lytical reagent grade sodium hydroxide,
sodium chloride, and sodium hypochloride
were used in the solubility studies.
Waters and Leachates
Distilled water, deionized water, tap
water, Sugar Creek water, soil extracts
(soil : tap water =1 : 3), Blackwell land-
fill leachate, and Du Page landfill leach-
ate were selected for use in the solubility
study. Leachates were centrifuged through
a continuous flow centrifugation apparatus
(Model JCF-Z, Beckman Instruments) at
approximately 17,000 rpm, prior to passing
through a 0.22 ym pore size Millipore®
cellulose acetate membrane. Tap water,
deionized water, Sugar Creek water, and
soil extracts were also passed through
0.22 pin membrane before use.
Adsorbents
Earth materials, representing a wide
range in characteristics, were selected as
adsorbents. They were: Ca-montmorillonite,
Ca-bentonite, illite, kaolinite, silica
sand, and five soils: Houghton muck, Catlin
sil, Flanagan sicl, Bloomfield Is, and Ava
sicl.
Solubility Studies
In the solubility studies, 350 mL of
water, soil extract, or landfill leachate
were placed into 500-mL amber glass bottles.
C-56 (0.5 mL) was added to each bottle, and
the bottle was capped with a teflon-lined
cap. The water and C-56 mixtures were
shaken in a constant temperature water bath
(22° ± 1°C) at moderate speed for 15 hours
and then left quiescent for 3 hours. Ali-
quots of the aqueous solutions were centri-
fuged in a constant temperature centrifuge
(Beckman Model J-21B at 22° ± 1°C) at
17,000 rpm for 30 min; then 5-mL aliquots
of the supernatant were carefully with-
drawn and placed in 15-mL screw-capped
vials. All samples were extracted for C-56
three times with 5-, 5-, and 5-mL portions
of water-saturated hexane. Each extraction
was vibrated on a vortex mixer at fast
speed for 1 min; the hexane extract was
then transferred to a graduated test tube
using a Pasteur pipette. Studies proved
that a shaking time of 12 hours was suffic-
ient to reach equilibrium conditions.
Therefore, all solubility studies were
shaken overnight (15 hours) to ensure
equilibrium and for convenience. Results
of preliminary studies showed that the
extraction technique used in this study
was much better than conventional separa-
tory funnel extraction. The extraction
efficiency was 95% or better in samples
spiked with C-56 at three different con-
centrations. The extracts were then
analyzed on a Hewlett-Packard (HP 5840A)
gas chromatograph or a Perkin-Elmer Sigma
I gas chromatograph (GC). Tribromobenzene
was used as the internal standard. Both
gas chromatographs were equipped with Ni63
electron capture detectors. A glass capil-
lary column was used to facilitate the
separation of the hydrolysis products of
C-56 from "pure" C-56 in distilled water
30
-------
TABLE 1. CONDITIONS FOR GAS CHROMATOGRAPHIC ANALYSIS
Conditions
HP 5840A and
P-E Sigma I GC
Sigma I GC
Column:
Injector temperature:
Column temperature:
Detector temperature:
Carrier gas:
6 ft x 2 mm I.D. glass
column, 5% OV-17 on
100/120 mesh chromosorb
WHP
250°C
150°C
300°C
methane/argon, flow
32 mL/min
28 m x 0.25 mm I.D. grade
AA glass capillary column
(SE-30), split ratio:
100 : 1
250°C
145°C
280° C
methane/argon, velocity:
28 cm/sec
and deionized water extracts. The condi-
tions for the gas chromatographic anal-
ysis for C-56 are shown in Table 1.
To determine the effect of caustic
soda-brines and soluble salts on the
aqueous solubility of C-56, several con-
centrations of soluble salts were added
to water and the changes in solubility of
C-56 were measured. The concentrations
and salts used were: 0.513, 1.026, 1.465,
and 2.051 M NaCl; 0.135, 0.270, 0.405,
and 0.541 M NaOCl; and 0.188, 0.375, 0.563,
and 0.750 M NaOH.
Identification of Degradation Products
During Hydrolysis and Photolysis
In the hydrolysis study, the pH of
tap water (8.45) was adjusted to 2.51,
5.80, and 11.73. Mixtures of C-56 and
water were shaken in a constant-temperature
water bath (22° ± 1°C) for 18 hours and
then allowed to sit quiescent for at least
3 hours. Fifty-mL aliquots of the C-56
water solutions were filtered through 0.22
ym Millipore®cellulose acetate membranes.
The membrances were presaturated with C-56
by soaking the membranes in C-56 saturated
water prior to filtration. The membranes
were then washed with deionized water and
200 mL of the C-56 saturated tap water was
flushed through the membrane.
In a separate study, 50 mL aliquots
of the filtered aqueous C-56 were placed
in 125-tnL amber serum bottles and sealed
with teflon-faced septum and aluminum
seals. All samples were kept in darkness
and two replicate samples were analyzed
at 0, 1, 2, 4, 7, and 11 days. The extrac-
tion was accomplished by placing 10-mL
aliquots of the filtrates into 15-mL vials
and then extracting as previously described.
In the photolysis study, 10-mL ali-
quots of the filtrate were placed into
15-mL vials and exposed to sunlight for
1, 3, 6, 12, and 22 min. The sunlight
intensity was measured by using a YSI-
Kettoring Model 65A Radiometer (Yellow
Springs Instrument Co., Inc., Yellow
Springs, OH). The average sunlight inten-
sity was 770 joules-sec/m2 (or watts/m2).
Samples were set in darkness for a few
minutes and then extracted as previously
described. In a separate study, 5 mL of
C-56 (2 ppm) in hexane were placed in 15-
mL vials and capped with teflon-lined caps,
then exposed to sunlight for 0, 1, 2, 4,
and 6 min. The extracts and hexane soluble
C-56 were analyzed by gas chromatography.
The half-life and dissipation rate were
also calculated. The major hydrolysis
and photolysis products were identified by
mass spectrometry using a Hewlett-Packard
5985 GC mass-spectrometer and data system.
Adsorption Studies
Adsorption studies were carried out
31
-------
by shaking known volumes of C-56 solution
with varying weights of soil materials at a
constant temperature of 22° ± 1°C in amber
glass serum bottles. Amber glass was used
to prevent photolysis of the C-56, and the
bottles were sealed with a teflon-faced
septum and an aluminum crimp-cap to prevent
volatilization. The rate of adsorption of
C-56 by soil materials was relatively rapid.
A study indicated a shaking time of 2 hours
was sufficient to reach constant concen-
trations of C-56 in solution. Therefore,
all samples were shaken 4 hours to ensure
equilibrium. The bottles were removed from
the shaker and placed directly into a Model
JS-7.5 rotor and centrifuged at a constant
temperature of 22° ± 1°C in a Beckman Model
J-21B centrifuge for 15 min at 4000 rpm.
The seals were then broken and 10-mL ali-
quots of the clear supernatants were pipet-
ted for C-56 analysis.
Blanks were carried through all exper-
iments to determine-the degree of adsorp-
tion of C-56 onto the surface of the
bottles. The amount of adsorption was
determined from the difference between the
initial concentration and the equilibrium
concentration multiplied by the volume of
solution. A blank was subtracted, and the
amount adsorbed by each test material was
computed on a unit basis by dividing by the
dry weight of the adsorbent.
Mobility Studies: Determination by Soil TLC
and Soil Column Leaching
The method used to prepare soil TLC
plates has been reported previously by
Griffin and his co-workers (1977). In
these studies, the C-56 was spotted 2 cm
from the base and leached 10.5 cm with tap
water, caustic soda-brine solution, land-
fill leachate, acetone/water (1 : 1), meth-
anol, acetone, or dioxane. The plates
were immersed in 0.5 cm of the respective
solvent in a closed glass chamber and were
removed when the wetting front reached the
10.5 cm line. The soil plates were removed
and the soil was immediately scraped off in
1 cm increments, starting 1.5 cm above the
base of the plate. The soils were placed
in glass test tubes and extracted with hex-
ane or other suitable organic solvents.
The concentration of C-56 in the extracts
was measured by GC analysis. In a soil
column leaching study, a sample of Bloom-
field loamy sand that had been heavily
spiked with C-56 to simulate a C-56 spill
was leached with 192 in. of tap water
(approximately 5 to 6 times the average
annual rainfall in Illinois).
RESULTS AND DISCUSSION
Solubility of C-56 in Waters and Leachates
The solubility of C-56 in tap water,
Sugar Creek water, soil extracts, and
landfill leachates varied from 1.03 in
Sugar Creek water to 1.25 ppm in Du Page
landfill leachate. The results are shown
in Table 2. An increase in total organic
carbon content in the test waters and
leachates slightly increased the solubility
of C-56. The solubility of C-56 in water
has previously been reported as about 0.8
ppm (Liu et al., 1975), 1.8 ppm (Zepp et
al., 1979), and 2 ppm (Atallah, Whitacre,
and Butz, 1980).
An experiment showed that C-56 was
not stable in distilled water or deionized
water. This experiment was repeated
several times. In all cases a degradation
product was formed that had a retention
time almost identical with that of the
parent C-56. A possible explanation of
this phenomenon could be the dissociation
of a chlorine ion from C-56 to make the
water acidic, which in turn could initiate
a hydrolysis reaction. This may occur to
a greater extent in purified water than
in tap water because distilled water and
deionized water have no buffering capacity
to neutralize the acid. A similar degra-
dation product was also observed when
the pH of tap water was adjusted to values
below 6.0. Representative capillary column
GC chromatograms of tap water and deionized
water-soluble C-56 are shown in Figure 1.
GC peaks for the degradation product and
for C-56 were resolved on a capillary
column (Fig. 1A). There is a possibility
that previous workers using a packed GC
column may have measured the degradation
product or a combination of the degradation
product and C-56 rather than "pure" C-56.
This may have led to the conclusion that
greater concentrations of C-56 were present
in the water samples than actually was
the case.
The effect of several soluble salts
on the solubility of C-56 were treated
by fitting the solubility data to the
Setschenow equation (1892):
log -£- = Km
32
-------
TABLE 2. SOLUBILITY OF C-56 at 22°C AFTER 15 HOURS EQUILIBRIUM IN WATERS
AND LANDFILL LEACHATES USING CENTRIFUGATION TECHNIQUE
Maters and leachates Concentration (mg/L)*
Tap water 1.08
Sugar Creek water 1.03
Blcornfield soil extract 1.06
(soil : tap water =1:3)
Cat!in soil extract 1.20
(soil : tap water =1:3)
Blackwell leachate 1.19
Du Page leachate 1.25
*Each value is a mean of 2 replicates except tap water which is 20 replicates.
TABLE 3. EFFECT OF DISSOLVED SALTS ON THE WATER SOLUBILITY OF
C-56 AT 22°C AND SETSCHENOW PARAMETER FOR C-56
Setschenow parameter m
Salts K r2 (mole/L)
NaOH +0.266 0.996 0.188
0.375
0.563
0.750
NaCl +0.161 0.970 0.513
1.026
1.465
2.051
NaOCl -0.058 0.889 0.135
0.270
0.405
0.541
Vs
1.099
1.239
1.396
1.563
1.306
1.622
1.683
2.239
0.998
0.956
0.942
0.940
33
-------
(A)
(B)
3.87
4.17
3.87
4.18
JL
10.79
10.96
C-56
10.97
C-56
Figure 1. GC chromatograms: (A) Deionized water soluble C-56, and (B) Tap water
soluble C-56 (28 m x 0.22 mm glass capillary column).
34
-------
The results are given in Table 3; listed
are the observed solubility ratios, S /S
where S is the solubility (ppm) of C-56
in tap water and S its solubility (ppm)
in a salt solution of concentration m
3 also gives the Setsche-
The best values of K
linear regression ana-
(mole/L). Table
now parameter, K.
were determined by
lysis of log -5
as a function of m; the
coefficient of correlation (r2) is also
given in the table.
Of the three salts studied, sodium
hydroxide and sodium chloride decrease
the solubility of C-56 in water (K = +) and
sodium hypochloride slightly increases its
solubility (K = -). A similar effect of
sodium chloride on water solubility of
lindane was reported elsewhere (Masterton
and Lee, 1972).
Hydrolysis and Photolysis
C-56 undergoes rapid hydrolysis at
pH 2.51 and slower hydrolysis at pH 8.45
and 11.73. In a preliminary study, we
found that C-56 is not stable in either
distilled water or deionized water. In
this study, we affirmed that low pH causes
hydrolysis. Figures 2A and 2B show gas
chromatograms obtained by using a packed
GC column. Peaks c, d, and the combined
peak for the hydrolysis product and C-56
decreased as the pH of tap water was
increased. However, the height of peaks
a and b (see Figures 2A-D) were slightly
increased as the pH was increased. Yu and
Atallah (1977) found C-56 was hydrolyzed
much faster at pH 12 than at 3 and 6. We
have repeated this experiment several times,
and it appears that the large peak close to
the C-56 peak was the major hydrolysis pro-
duct at low pH, but that it cannot be
resolved into a distinct peak using a
packed GC column as is illustrated in Fig-
ure 2.
In the photolysis study, we found peaks
c and d (see Figure 2A) were also the major
degradation products. Of interest was the
height of peak d which increased in tap
water when the photolysis time was
increased. However, peak d remained con-
stant in hexane soluble C-56 solutions
exposed to 1ight.
Values of the first-order rate con-
stant (K), the half-life (ti/2), and
regression coefficient (r2) for the dis-
sipation of C-56 in solutions exposed to
sunlight are given in Table 4. The dissi-
pation rate of C-56 in aqueous solution and
hexane follows a first-order reaction. C-56
dissipated much faster in hexane than in tap
water. These results also show that C-56
was dissipated in a closed vial when it
was exposed under sunlight; this indicates
that the dissipation of C-56 under the
tested conditions was mainly due to photo-
lysis and/or hydrolysis and was only par-
tially due to volatilization.
TABLE 4. FIRST-ORDER RATE CONSTANT (K), HALF-LIFE (tij, AND REGRESSION
COEFFICIENT (r2) FOR THE DISSIPATION OF C-56° IN TAP WATER AND HEXANE
UNDER SUNLIGHT AND DARKNESS
Treatment
Under sunlight
C-56 hexane in capped vial
C-56 water in capped vial
Under darkness
C-56 water in sealed bottle
0.58 min"1 1.19 min
0.19 min"1 3.71 min
0.04 day
-i
16.09 day
0.9996
0.9972
0.9243
35
-------
0.37
2.98
Hydrolysis
product
+C-56
d IS
0.37
| 2.88
Hydrolysis
product
+C-56
UJLjl
IS
0.37
2.87
:-56
(B)
(C)
0.37
0.51
IS
V
(D)
Figure 2. Packed column GC chromatograms of C-56 in tap water at various pHs adjusted with either HC1 or NaOH:
(A) 2.51 (B) 5.80 (C) 8.45 (original pH) (D) 11.73. IS = internal standard.
-------
Identification of Degradation Products of
C-56 in Water
The mass spectrum obtained from the
Hewlett-Packard 5985 GC/MS of the major
degradation product peak in distilled water
and deionized water is interpreted to con-
tain six chlorines and to have a molecular
weight of 286. The fragmentation patterns
show m/e, 258 (-CO); m/e, 251 (-C1); m/e,
223 (-CO, -CI); m/e, 216 (2-C1); m/e, 181
(-3C1); m/e, 188 (-CO, -2C1); m/e, 153
(-CO, -3C1); m/e, 118 (-CO, -4C1); and m/e,
111 (-5C1). We expect that hexachloro-2-
cyclopentenone (I) and hexachloro-3-
cyclopentenone (II) would have fragmenta-
tion patterns similar to the unknown degra-
dation product. Newcomer and McBee (1949)
reported that hexachlorocyclopentadiene can
be converted to octachlorocyclopentadiene
in the presence of C12 or A1C13 and further
converted to compound I and compound II.
They also noted that C-56 can be converted
to 2, 3, 4, 4, 5-pentachloro-2-
cyclopentenone (III) in sulfuric acid solu-
tion.
(Ill)
Simonov et al. (1975) also found these
three hydrolysis products of C-56 in water.
Compound (I) and (II) have GC retention
times very close to C-56; however, separa-
tion was achieved and their presence was
confirmed by using high resolution glass
capillary column GC techniques (see Figure
1).
In the mineralized water (tap water)
sample, the two major transformation pro-
ducts were different from those in the dis-
tilled water and deionized water samples.
They were isomers of C^HCls, and are
thought to be
and
ci
ci
ci
ci
In the hydrolysis and photolysis studies,
one of the degradation product peaks was
found to have the same GC retention time
as that of 2, 3, 4, 4, 5-pentachloro-2-
cyclopentenone (III). In acidified
aqueous extracts of C-56, one neak also
matched a second standard, pentachloro-
cis-2, 4-pentadienoic acid. Both compounds
have previously been reported by Russian
workers (Simonov et al., 1975) in a study
of C-56 hydrolysis promoted by various
metals. Zepp et al., (1979) suggested that
tetrachlorocyclopentadienone is a major
hydrolysis and photolysis product. How-
ever, we did not find this component present
in our samples.
Adsorption
The results for C-56 adsorption by
various soil materials used in this study
are shown in Figures 3 and 4. All data
were fitted by linear regression to the log
form of the empirical Freundlich adsorption
equation:
log £= log Kf + J- log C
Where X = yg of compound adsorbed; M =
weight of adsorbent (g); C = equilibrium
concentration of the solution (pg/mL);
Kf and 1/n are constants.
Values of Kf and 1/n were obtained
from the regression equations as the inter-
cepts at a concentration of 1 ppm and the
slope, respectively. These Freundlich
parameters and the regression coefficient
(r2) are shown in Table 5. All the regres-
sion lines generated had coefficients (r2)
of at least 0.98, which indicate an excel-
lent fit of the data by the Freundlich
equation. The molar K, (Kp) is also shown
in Table 5. It was calculated from mass K,
(Kf) by using the equation described by
Osgerby (1970):
KF (molar) = Kf
x Mo1 •
Mol. Wt.
1/n
37
-------
400
200-
.001
Figure 3.
.002
Equilibrium concentration of C-56 (ppm)
.6
SGS 1980
Freundlich adsorption isotherms of C-56 adsorption on four soil materials
from tap water at 22° ± 1°C.
Where Mol. Wt. is molecular weight of the
compound, K (mass) is the Freundlich Kf,
and 1/n is the slope of the isotherm
plotted according to the Freundlich equa-
tion. The adsorption of C-56 on these
four soils followed the series:
muck > muck + Catlin (!:!)>
Catlin > Flanagan > Ava
This suggests a relationship between the
organic carbon content of these soils and
their adsorption capacity for C-56. The
adsorption of C-56 on four clays followed
the series: Ca-bentonite > illite > mont-
morillonite > kaolinite. This again sug-
gests a similar relationship between the
organic carbon content of these clays and
their adsorption capacity for C-56.
The relationship between total organic
carbon content of the nine adsorbents and
adsorption of C-56 by the earth materials
was investigated. The molar K (Kp) plotted
as a function of TOC in percent is shown in
Figure 5. A high correlation was found
38
-------
with a linear regression relation of
KF = 8.92 (%TOC) + 0.78
(r2 = 0.99).
Similar results for PCBs, PBBs, and HCB were
reported elsewhere (Lee et al., 1979;
Griffin and Chou, 1980).
These results indicate that the adsorp-
tion properties of soil materials for C-56
can be predicted accurately when the TOC
content of the involved earth materials are
known. This information should be used
with caution, however, because only a limi-
ted number of soil materials were used to
develop the equation.
Mobility
The mobility of C-56 in soils
expressed as Rf values is summarized in
Table 6. The results showed that under
the conditions tested, C-56 stayed fairly
immobile in all soils when leached with tap
water and landfill leachate. However, C-56
was highly mobile when leached with an
acetone/water mixture, pure acetone,
methanol, and dioxane. In a separate
study, we also found that C-56 was immobile
in Catlin, Ava, and Bloomfield soils when
leached with caustic-soda brine solution.
It is interesting to note that mobility
of C-56 in soil was proportional to the
soil organic carbon content. C-56 was sig-
nificantly more mobile in the low organic
matter sandy soil than in the muck soil.
Clay content does not appear to be a sig-
nificant factor affecting mobility of C-56
except as it pertains to hydraulic con-
ductivity.
In a soil column leaching study, a
sample of Bloomfield loamy sand, which had
been heavily spiked with C-56 to simulate
a C-56 spill, was leached with 192 inches
of tap water. We found 0.0005% of the
C-56 was leached from this highly contami-
nated soil by the percolating water. How-
ever, caution should be taken because some
of the hydrolysis products of C-56
described in the previous sections appar-
ently have much higher solubility in water
20
10-
6-
2-
— i - 1 - 1 - 1
04 .06 .08 .1
Equilibrium concentration of C-56 (ppm)
r~
.2
—r"
.4
.01
02
6
ISGS 1980
Figure 4. Freundlich adsorption isotherms of C-56 adsorption on four clays from
tap water at 22° ± 1°C
39
-------
TABLE 5. FREUNDLICH Kf (ug/g), 1/n, REGRESSION COEFFICIENT (r2),
AND MOLAR KF FOR ADSORPTION OF C-56 BY VARIOUS ADSORBENTS
FROM TAP WATER AQUEOUS SOLUTION AT 22°C
Freundlich parameters
Adsorbents
Houghton muck
Muck + Catlin sil
(1:1)
Catlin sil
Flanagan sicl
Ava sicl
Ca-bentonite
Illite
Montmorillonite
Kaolinite
Kf
1045
284
54
28
13
32
24
18
4.4
1/n
0.79
0.76
0.94
0.92
1.18
0.63
0.55
0.60
0.58
r2
0.9854
0.9962
0.9971
0.9944
0.9940
0.9838
0.9991
0.9928
0.9924
KF
322
74
39
18
36
4.00
1.93
1.91
0.42
TABLE 6. MOBILITY OF HEXACHLOROCYCLOPENTADIENE IN
SEVERAL SOIL MATERIALS LEACHED WITH
VARIOUS SOLVENTS AS MEASURED BY SOIL TLC
Soil materials
Muck soil
Catlin sil
Flanagan sicl
Ava sicl
Bloomfield Is
Ottawa sand
Tap
water
0.002
0.002
0.001
0.002
0.002
ND**
Landfill
leachate
0.005
0.002
0.002
0.003
0.004
0.003
Rf
Acetone/
water (1:1)
0.093
0.793
0.841
0.871
0.869
0.903
value*
Methanol
0.053
0.879
0.893
0.891
0.972
ND
Acetone
0.137
0.972
0.994
0.990
0.981
ND
Dioxane
0.640
0.989
0.993
0.956
0.992
0.978
*Computed from statistical peak analysis of data by using values of 1st moment for
grouped data.
**ND = Not determined.
40
-------
325
300-
200-
100-
KF =8 92 (TOO +0.78
r2 = 0.99
%TOC
Figure 5.
KF vs TOC for C-56 adsorption
by nine soil materials.
than C-56 and are much more mobile in soil.
A general conclusion from this study is
that C-56 mobility appears to be predict-
able from the TOC content of soils and that
the secondary products of C-56 might tend
to migrate through soils and similar nat-
ural environments and may generate prob-
lems rather than C-56 itself.
Acknowledgments
The authors wish to acknowledge par-
tial support of this project under Grant
No. R806335-010 from the U.S. Environmen-
tal Protection Agency, Cincinnati, OH;
Dr. Michael Roulier, project officer.
The authors also wish to thank Dr.
R. Larsen of the University of Illinois
for aid in the mass-spectral analysis.
REFERENCES
1. Atallah, Y. H., D. M. Whitacre, and
R. G. Butz. 1980. Fate of hexachloro-
cyclopentadiene in the environment.
ACS preprints, 180th National Meeting,
San Francisco, California.
Bell, M. A., R. A. Ewing, and G. A.
Lutz. 1978. Review of the environmental
effects of mirex and kepone. Prepared
for the U.S. Environmental Protection
Agency. Office of Research and devel-
opment, by Battlle Columbus Laboratory.
EPA 600/1-78-013.
Griffin, R. A., F. B. DeWalle, E.S.K.
Chian, J. H. Kim, and A. K. Au. 1977.
Attenuation of PCBs by soil materials
and char wastes in "Management of
Gas and Leachate in Landfills." S. K.
Banerji [Ed.] U.S. Environmental Pro-
tection Agency, Cincinnati, OH 45268.
EPA 600/9-77-026. p. 208-217.
Griffin, R. A., and S.F.J. Chou. 1980.
Disposal and removal of halogenated
hydrocarbons in soils. Proceedings of
the 6th Annual Solid and Hazardous
Waste Research Symposium. U.S. Envi-
ronmental Protection Agency, EPA 600/
9-80-010. p. 82.
Hoecker, J. A., P. R. Durkin, A.
Hanchett, L. N. Davis, and W. M.
Meylan. 1977. Information profiles
of potential occupational hazards.
NTIS No. PB-276 687/OSL, 334 pp.
Lee, Me., R. A. Griffin, M. L. Miller,
and E.S.K. Chian. 1979. Adsorption of
water-soluble polychlorinated biphenyl
Aroclor 1242 and used capacitor fluid
by soil materials and coal chars. J.
of Environmental Science and Health
A14, p. 415-442.
Liu, P. Y., R. L. Metcalf, A. S. Hirwe,
J. W. Williams. 1975. Evaluation of
environmental distribution and fate
of hexachlorocyclopentadiene, chlor-
dane, heptachlor, and heptachlor epox-
ide in laboratory ecosystem. J. Agr.
Food Chem. 23(5):967-973.
Masterton, W. L., and T. P. Lee. 1972.
Effect of dissolved salts on water
solubility of lindane. Environmental
Science & Technology, 6(10), 919-921.
National Academy of Sciences. 1978.
Kepone/mi rex/hexachlorocyclopentadiene:
An environmental assessment. 78 pp.
41
-------
10. Newcomer, J. S. and E. T. McBee. 1949. 13.
The chemical behavior of hexachloro-
cyclopentadiene. I. Transformation
to octachloro-3a, 4, 7, 7a-tetra-
hydro-4, 7-methanoindene-l, 8-dione. 14.
JACS 71, p. 946.
11. Osgerby, J. M. 1970. Sorption and
transport processes in soils. S.C.I.
Monograph No. 37, Society of Chemical
Industry, London, S.W. 1 p. 63-78. 15.
12. Spehar, R. L., G. D. Veith, D. L.
DeFoe, and B. V. Bergstedt. 1979.
Toxicity and bioaccumulation of hexa- 16.
chlorocyclopentadiene, hexachloro-
norbornadiene and heptachloronor-
bornene in larval and early juvenile
fathead minnows, Pimephales promelas.
Bull. Environ. Contam. Toxicol. 21,
p. 576-583.
Setchenow, M., 1892. Action De L'Acide
Carbonique Sur Les Solutions Des Sels.
Ann. Chem. Phys. V. 25:226.
Simonov, V. D., et al. 1975. Hydroly-
sis of hexachlorocyclopentadiene.
Vsesoiuznoe Khimicheskoe Obshehestvo
Imeni D. I. Mendeleeva Zhurnal, v. 20,
p. 477.
Yu, C. C., and Y. H. Atallah. 1977.
Velsicol Chemical Corporation Project
482428 report Nos. 1, 2, and 3.
Zepp, R. G., N. L. Wolfe, and G. L.
Baughman, P. F. Schlotzhaner, and
J. N. MacAllister. 1979. Dynamics of
processes influencing the behavior of
hexachlorocyclopentadiene in the
aquatic environment, American Chemical
Society Preprints, 178th National
Meeting, Washington, D. C. Sept. 9-14.
42
-------
METHODS OF SOIL HYDRAULIC CONDUCTIVITY
DETERMINATION AND INTERPRETATION
David W. Roberts and Michael A. Nichols
ABCDirt
Soil Scientists
Seattle, WA 98103
ABSTRACT
Section 3004 of the Resource Conservation and Recovery Act mandates the
promulgation of standards for the location, design and construction of hazardous
waste treatment, disposal or storage facilities. Such legislation deals with the land
disposal of industrial or hazardous wastes and focuses on the potential water
pollution that can result from such practices because more than half the U. S.
relies on ground water for their drinking supply. This paper provides a look at the
categories of water movement that occur in connection with various types of land-
based disposal systems, some background information and general equations re-
lating to soil water movement, a brief summary of soil testing methods divided into
saturated and unsaturated sections with subsections of laboratory and field
techniques, a soil testing methods matrix that attempts to evaluate the applicabil-
ity of individual tests for solving various soil water movement problems, and some
remarks regarding the present status of available information, what is trying to be
obtained in the future, and how do we get there from here.
INTRODUCTION
As more and more waste is being applied to
land, the question of where it goes and how fast
it gets there needs to be answered better than
has been done in the past, as it is apparent that
some sites have not been well-designed in ac-
cordance with soil hydrological properties.
Considering the array of waste materials
and the different modes of application, there
appears to be three categories of water move-
ment that come into play. First of all is the
consideration of water movement within the
root zone when crops are grown as part of the
treatment-disposal system. Such waste streams
as municipal and industrial wastewaters,
sludges, and some hazardous wastes are mater-
ials that may be applied to the soil surface or
injected into the rootzone. One big advantage
of such systems is in the way that plants can
use or immobilize both water and contami-
nants/nutrients.
A second type of water movement relates
to use of compacted soil layers as liners for
landfills and surface impoundments. Since such
soil is disturbed by mechanical equipment it
will no longer behave the same hydraulically as
it did in its undisturbed state.
A third type of water movement related to
waste disposal sites is the unsaturated zone
between the root zone (or liner) and the ground
water table or bedrock as water movement will
be very different in subsoil materials than in
surface horizons.
In this paper we intend to discuss some of
the basic principles involved in water move-
ment in soil with a brief look at several types
of soil testing methods for both laboratory and
field varieties of saturated and unsaturated
hydraulic conductivity determination. After
this discussion the available information on
advantages, disadvantages, and limitations has
been compiled into a soil testing methods
43
-------
matrix that attempts to rate the applicability
of each individual test for the three types of
water movement, problems that may arise in
considerations of waste applications to land.
The final part of this paper is a look at what
can be expected in the future.
WATER MOVEMENT
Water movement in soils can be divided
into two types of flow systems for general
consideration: (1) saturated flow where all
pores are filled with water and (2) unsaturated
flow where both air and water are present in
the pores.
The basic equation for describing water
flow through porous materials, whether satu-
rated or unsaturated flow, is expressed by
Darcy's Law:
J = K • AH
where: J = flux or volume of liquid moving
thru a cross-sectional area of
soil per unit of time
K = hydraulic conductivity, and
A H = hydraulic gradient = -r— where
d H = difference in hydraulic
potential and dx = difference in
distance over which the differ-
ence in hydraulic potential is
measured.
The hydraulic conductivity, K, is not con-
stant for a given soil as it can vary with
changes in the water content as well as changes
in the pressure head. Therefore, the hydraulic
conductivity can either be expressed as a func-
tion of water content, K(6) or as a function of
pressure head, K(h). For determination of
either function, the soil moisture characteris-
tic/retentivity curve/water capacity function
(6(h)) usually must be known also. Figure 1
demonstrates these two different cases: A) Soil
moisture characteristics for four soils with the
respective hydraulic conductivity curves ex-
pressed as a function of pressure head and
B) Soil moisture characteristics for one soil and
its hydraulic conductivity relation expressed as
a function of water content.
Saturated Flow
The determination of saturated hydraulic
conductivity is physically easier and concep-
tually simpler than the determination of un-
saturated hydraulic conductivity because the
purpose is to find the hydraulic conductivity at
only one point, (h = 0, 6 = constant), rather
than over the range of pressure heads or range
of water contents necessary for unsaturated
flow determination. Therefore, at saturation it
is assumed that all the soil pores are constantly
and completely filled with water and that the
hydraulic conductivity is a constant. This is
shown graphically in the right part of Fig. 1(A)
where the saturated conductivity for each soil
is where their respective curves intercept the
y-axis.
However, in a given soil, the saturated hy-
draulic conductivity may not remain constant
over time due to processes such as particle
rearrangement within the soil matrix. Also,
depending on the mode of wetting, saturated
soils rarely have all soil pores completely filled
with water. It is not unusual to expect 2-12%
air remaining in the soil pores (Yong & v, arker:-
tin, 1975). This implies that attention should be
placed on the mode of saturation because if K
is determined for a situation of entrapped air,
it is basically an unsaturated flow situation, and
unsaturated flow will always be less than satu-
rated flow.
Unsaturated Flow
The situation for unsaturated hydraulic
conductivity determination is more complex
than for saturated flow because the purpose is
to determine a complete hydraulic conductivity
function over a range of pressure heads or
water contents rather than the one-point analy-
sis as for saturated flow.
Another concept that should be introduced
to the discussion of unsaturated flow is diffu-
sivity which can be expressed as:
K(6)
C(6)
D(0) = K(6) 4- =
where D(6) = soil water diffusivity as a
function of water content
K(8) = hydraulic conductivity as a
function of water content
C(6) = water capacity/retentivity/
soil moisture characteristics
as a function of water
content
44
-------
I ' I
F>pe I I sand I
T\pe II isands loam)
1 » I
SOIL MOLS I L Rl I h NSION i MB \R t
I!0 40 61) 80 100
SOIL MOISTURE TENSION (V1BAR)
DRYING »•
50 100 HO 200 2SO
SOIL WATER PRESSURE HEAD (CM)
0 24 0 28 0 32
SOIL WATER CONTENT (CM'/CM't
Figure 1. Water retention curves and hydraulic conductivity. (A) Water retention
curves and hydraulic conductivity as a function of pressure head (SSWMP, 1978) and
(B) water retention curve and hydraulic conductivity as a function of water content
(Flocker et al., 1968).
45
-------
--
= reciprocal of slope of C(9)
function at a particular water con-
tent.
Anisotrophy
Anisotrophy is the term used to describe
the fact that soil properties can vary with
depth as well as in horizontal direction. Such
differences are the result of weathering of
rocks or sedimentary bodies and the reorganiza-
tion, translocation, and concentration of the
more mobile constituents due to the effects of
the environment (Brewer, 1976). Unfortun-
ately, anisotrophy in soils is the rule rather
than the exception, which will affect soil test-
ing methods and their interpretation.
Soils are stratified formations consisting of
distinct layers (horizons) which have differen-
ces in particle size distributions, structure,
boundary conditions, and other properties which
affect the hydraulic properties of each horizon
and macroscopic level (soil series) as well as on
the microscopic level (soil fabric). Such differ-
ences will influence water movement in soils.
At the soil series level, anisotrophy is
manifested in two ways: (1) horizons will vary
in depth as well as thickness and can exhibit
anomalies such as vertical sand seams or clay
lenses and (2) boundary conditions between soil
horizons can sometimes influence overall hy-
draulic conductivity more than the properties
of the mass of the horizon itself. This is shown
graphically in Figure 2.
In Case A, water added to the sand will
flow through the sand quickly into the clay,
while in Case B, no water will enter the sand
layer until the clay layer is nearly saturated be-
cause the water is held by the finer pores in the
clay horizon. Therefore, the clay horizon in
Case B might hold up to two or three times as
much water as the upper clay horizon in Case
C.
Further, anisotrophic conditions can affect
the direction of water movement as there are
both Horizontal and vertical components to the
hydraulic conductivity function. Given the
Darcy equation:
J = K
. ,, ,. dH
L A = Is.-3—
dx
and knowing that the hydraulic potential H =
the sum of the gravitational (2) and pressure (h)
potentials, expressions can be derived for both
the horizontal and vertical components of flow.
for horizontal flow: J = K •-j—
dx
where
d(2+h)
J = flux
K = hydraulic conductivity
change in pressure potential
divided by change in distance
for the vertical flow: J = K
d(2+h)
d3
where J = flux
K = hydraulic conductivity
d(Z+h) . .
-j— = change in sum of gravita-
3 tional and pressure potentials
divided by change in gravita-
tional potential
A review of the literature for the ratio of
horizontal K to vertical K for fine-textured
soils reveals ratios of 0.7 to 40 with a common
range of approximately 1 to 4 (Roberts and
Nichols, 1980).
Figure 2. Influence of underlying layer on hydraulic conductivity.
46
-------
At the microscopic level, soil fabric will
also exhibit anisotrophy which can cause sub-
stantial hydraulic conductivity anisotrophy.
Mitchell (1976) states, "Of the properties of
importance in the analysis of geotechnical
problems of fine-grained soils, none is more
influenced by fabric than the hydraulic conduc-
tivity." Also, initial fabric anisotrophy may
cause mechanical property anisotrophy with
spatial and vertical differences being signifi-
cant.
Further, fabric anisotrophy may be devel-
oped in initially homogenous soils from shear or
compression forces. The amount of shear or
compression required for development of aniso-
trophic fabric varies for mineral soils, and
depends on such factors as soil mineralogy,
composition of pore fluid, and initial fabric
(Mitchell, 1976).
O Points during drying
A Points during welling
A practical ramification of soil fabric
anisotrophy is that the hydraulic conductivity
of a dry soil is dependent on the structure and
fabric to such an extent that analyses based on
properties determined from the same material
but with a different structure may be totally in
error. That is, where fabric and isotrophy can
affect a parameter being tested (such as the
hydraulic conductivity of clay soils), laboratory
results often may not reproduce field results
due to sampling and preparation procedures of
laboratory methods which will change soil
fabric. Therefore, extrapolation of laboratory
results to field conditions should only be used
with caution and supporting iustifications.
Hysteresis
Another phenomena that should be men-
tioned under the topic of water movement is
hysteresis which is a term describing the fact
that the water capacity/retentivity/soil mois-
ture characteristic is different for when the
soil is drying than when the soil is being wetted.
Said another way, it is harder to get water out
of the soil once it is in, than to get it back in
once it is out. Figure 3 depicts this phenome-
non and shows that the water content of a soil
at a given suction will be greater following de-
sorption (drying) than following adsorption
(wetting).
Figure 3. The relations between water content
(6) and suction (h) (Tzimas, 1979)
Since the moisture retention curves are
usually determined by a desorption process
starting with a saturated soil, values thus ob-
tained may not apply to water contents and
moisture potentials occurring in the field when
an initially dry soil is wetting up. However, a
physically correct characterization of the
extent of hysteresis on a given soil is a very
elaborate procedure (Gillham et al., 1976).
SOIL TESTING METHODS
Like the breakdown of water movement
into saturated and unsaturated flow, the soil
testing methods are also divided into these
same sections with further distinctions between
laboratory and field methods.
Saturated Flow — Laboratory Methods
The pressure cell technique is a method
where an undisturbed core, or disturbed volume
of soil, is placed in a metal pressure cell. After
the soil is initially saturated, it is connected to
a standpipe where a falling head procedure
allows water to move through the pressure cell
(Klute, 1965a).
47
-------
The compaction mold technique is a
method where disturbed soil is compacted in a
standard compaction mold and then placed in a
pan of water and allowed to saturate. A falling
head technique can be utilized or pressure can
be used to speed up testing time. The volume
of water is recorded over time (Matrecon,
1980).
There are two ways to use consolidation
cells to determine hydraulic conductivity. One
method employs direct measurement by means
of a falling head permeameter of special design
used in connection with the consolidation test.
A second method relies on the use of the
coefficient of compressibility, the coefficient
of consolidation, and the void ratio in consoli-
dation theory to calculate hydraulic conductiv-
ity (Means andParcher, 1963).
The triaxial apparatus technique is very
similar to the consolidation cell method as the
hydraulic conductivity can be determined di-
rectly from water flow measurements in the
triaxial apparatus, or the hydraulic conductivity
can be calculated from consolidation theory
(Matyas, 1967).
The micromorphometric data technique is
a calculation method which is very different
from the other procedures for determination of
saturated hydraulic conductivity. Rather than
using any measurement of flow or flux, charac-
terization of planer slits and tubular pores at a
microscopic level by freeze-drying techniques
for thin sections and scanning by a Qantimet
computer are used to predict saturated hydrau-
lic conductivity (Bouma et al., 1979).
Saturated Flow — Field Methods
The piezometer method is based on the
measurement of flow into an unlined cavity at
the lower end of a lined hole. Water entering
the unlined cavity and rising in the lined hole is
removed several times by pumping or bailing to
flush the soil pores along the cavity wall. After
flushing is completed, the water is allowed to
come to equilibrium with the water table
(Boersma, 1965a).
It should be noted here that there are
several varieties of 'piezometer' tests used,
depending on the geometry and materials lo-
cated at the point of measurement. Besides the
cylindrical cavity outlined in this procedure,
there are spherical cavities, sand-filled cavi-
ties, piezometer tips placed in sand-filled cav-
ities, or piezometer tips pushed directly into
the soil. This last method (Jezequel and
Miuessens, 1975) looks especially promising for
evaluation of the low conductivities of clay
soils.
The double ring infiltrometer technique is
a method where two open metal cylinders, one
inside the other, are driven into the soil (from a
few centimeters to several feet) and partially
filled with water that is maintained at a
constant level in both rings. The amount of
water added to maintain the constant water
level is the measure of the volume of water
that infiltrates the soil. This volume can be
measured over time and thus be expressed in
cm hour-1 (Bertrand, 1965; ASTM D 3385-75).
A modification of the double ring infiltro-
meter called the permeameter method uses the
sides of a hole to make the outer ring with a
smaller metal cylinder positioned in the center.
There is a constant water supply system that
maintains the water level in the cylinder and in
the surrounding hole. This method also employs
the use of tensiometers to indicate saturated
conditions (Boersma, 1965b).
Cased boreholes are performed by pumping
water into a borehole (infiltration test) or out
of a borehole (drawdown test) with or without
the use of inserts. These tests may be con-
ducted as a constant head test in which the rate
of pumping which is necessary to maintain a
constant water level in the borehole is mea-
sured. Alternatively, a variable head test may
be performed by observing the change in water
level in the borehole after inflow has stopped.
Fully or partially cased boreholes are usually
employed in these methods. The hydraulic
conductivity of a localized zone of soil sur-
rounding the base and uncased portions of the
borehole is computed by the application of
theoretically derived equations.
Well pumping tests are performed by
pumping water into or out of a screened well
embedded below the natural ground water
table. As pumping progresses, the resulting
change of ground water level is monitored in
surrounding observation wells. Generally, four
or more such observation wells are employed.
Pumping tests may be categorized as equilib-
rium (steady state) or non-equilibrium (tran-
sient flow) tests (Milligan, 1975).
In an equilibrium pumping test, the ground
water level measurements are recorded once a
constant water level has been attained in the
observation wells. In the non-equilibrium test,
the rate of pumping and the consequent rate of
48
-------
water level change in the observation wells are
recorded. The advantages of transient flow
testings are that testing time may be reduced
as compared with equilibrium testing, and
ground water drawdown may be controlled or
halted if necessary.
Unsaturated Flow — Laboratory Methods
Steady state techniques are methods where
the hydraulic conductivity is measured by
applying a constant hydraulic head difference
across the sample contained in a soil core and
measuring the resulting steady state flux of
water. The flow rate, Q, and tensiometric
metric potentials (pressure head) are recorded.
There are two varieties of steady state
tests often referred to as the short or long
column method. The short column method
utilizes columns of 10-15 cm while the long
column variety uses lengths of 50-200 cm. An
advantage of the long column technique over
the short column is that more than two tensio-
meters are encased along the side of the longer
column which will allow the determination of
the head difference across each interval and
thus the hydraulic conductivity function for
each section or layer (Klute, 1965b; 1972).
In the pressure outflow method (Klute,
1965c) the time dependence of the outflow of
water from a soil core on a porous plate or
membrane in a pressure cell is used to deter-
mine the soil-water diffusivity (D).
Hydraulic equilibrium is first established in
a layer of soil (usually 1-5 cm in depth), with
the gas phase pressure in the cell at a given
level. Beginning at or near saturation, step
increases of gas pressure are applied to produce
an outflow of water, and the volume of outflow
as a function of time is measured. Paired
values of diffusivity and water content are
obtained from each pressure increment and thus
the drainage function, D(9), can be plotted.
Since the total volume of outflow and the
pressure head increment corresponding to each
gas phase pressure increment are known, the
water capacity can be obtained and used to
calculate hydraulic conductivity from the rela-
tionship: K = DC where K = hydraulic conduc-
tivity (cm sec~l), D = diffusivity (cm^ sec~*)»
and C = water capacity (cm~l).
The movement through impeding layer
technique is a vertical infiltration method that
makes use of the hydraulic resistance of a
membrane or crust at the soil surface to pro-
duce negative pressure-head values in the soil
just below the impeding layer. The lower the
hydraulic conductivity of the crust, the more
negative the pressure head in the soil. To
obtain a weighted mean diffusivity, D, and
hydraulic conductivity, K, as a function of
water content, 6, a series of columns, each
capped by a crust or porous plate of different
hydraulic resistance is required. A constant
shallow head of water is then applied to the top
of the impeding layer, and the rate of infiltra-
tion is measured (Hillel and Gardner, 1970).
The instantaneous profile technique is a
method that can be used in the laboratory or in
the field. In the laboratory, a flow system
(usually a soil column) is established whereby
the inflow or outflow, water content, and pres-
sure head are measured or inferred (Watson,
1966).
Basically, there are three options with re-
spect to measurements made on the flow sys-
tem: (1) the water content and pressure head
distributions may both be measured, (2) the
water content distribution may be measured
and the pressure head inferred from water
retention data, or (3) the pressure head distri-
bution may be measured and the water content
inferred from water retention data.
Unsaturated Flow — Field Tests
The crust technique is a field adaptation of
the laboratory procedure for flow through an
impeding layer. The crust test can be used to
determine both saturated and Unsaturated con-
ductivity (SSWMP, 1978).
The crust test method requires carving out
a cylindrical pedestal of soil, whose upper sur-
face is at the depth where the hydraulic con-
ductivity is desired. A ring infiltrometer of the
same diameter as the pedestal is firmly affixed
to the top of the pedestal so that the soil's
horizontal surface is enclosed by the ring. A
paste consisting of gypsum and sand, mixed
with a small amount of water, is spread over
the entire horizontal surface and is allowed to
harden, forming a crust. When water is introd-
uced through the infiltrometer and maintained
at a constant head, flow into the soil is re-
stricted by the crust. A constant, steady state
flow rate is established, inducing a nearly uni-
form moisture potential (measured by tensio-
meters) and the steady state flow rate (meas-
ured at the infiltrometer with a burette)
49
-------
defines a plot of infiltration vs. soil moisture
tension (the K(h) curve). Crusts of different
resistances yield different points on the K(h)
curve. A series of crusts ranging from greater
to lower resistance assures that the data points
fall on the wetting curve.
The technique can be extended to include
measurement of saturated hydraulic conductiv-
ity which requires the addition of a barrier to
flow around the sides of the soil pedestal. This
is usually done with the soil pedestal thoroughly
wetted and with a slurry of dental grade plaster
applied to the sides. No crust is applied to the
surface in this case.
The traditional instantaneous profile tech-
nique in situ is a method where a nearly level
plot of soil is diked to pond 2-3 cm of water.
After addition of water the plot is covered to
prevent evaporation, and drainage occurs. At
frequent intervals both pressure head and water
content values are measured. From these
measurements, instantaneous values of the
pressure gradients and fluxes can be deter-
mined, which will also provide hydraulic con-
ductivity values (SSWMP, 1978).
The unit gradient-drainage method is a
type of instantaneous profile method where the
soil is wetted deeply and allowed to drain while
evaporation is prevented. In a uniformly drain-
ing soil profile, the hydraulic gradient is often
nearly unity, and the water content is a func-
tion of time and nearly independent of depth.
This relation can be expressed ast
where D = diffusivity (cm^ sec M
L = length (cm)
h = average hydraulic head (cm)
This method differs from the traditional
instantaneous profile method in that instead of
measuring both water content and pressure
head in the field, only pressure head is mea-
sured in the field, and water content from
laboratory-determined water retention curves
is used to calculate hydraulic conductivity from
the equation K = Dd6/dt. Description of the
unit gradient-drainage method has been report-
ed by: Black et al., 1969; Davidson et al., 1969;
Gardner, 1970; Klute, 1972; Nielsen et al.,
1973; and Simmons et al., 1979.
A third method called the Libardi tech-
nique (Libardi et al., 1980) is a very recent
attempt to simplify the field determination of
hydraulic conductivity by means of a modified
instantaneous profile method. Similar to the
unit-gradient drainage method, it measures only
one parameter rather than both water content
and pressure head. However, unlike the unit-
gradient drainage method, the other parameter
is not measured in the laboratory. In fact, in
the Libardi method only the neutron probe
device is used for water content measurement,
and hydraulic conductivity is determined with-
out any further information.
Because of various difficulties involved in
the direct measurement of the hydraulic con-
ductivity function, there has been considerable
interest in the potential of calculating the
conductivity from the other properties of the
medium that may be easier to measure such as
the pore size distribution or the water retention
curve. The basic concepts are given by Childs
and Collis-George (1950), Marshall (1958), and
Milligan-Quirk (1960, 1961, 1964). Since there
is quite a variety in the approaches taken and
lack of space here for elaboration, the reader is
referred to Roberts and Nichols (1980) for brief
summaries of a number of such research
efforts.
ADVANTAGES, DISADVANTAGES AND
LIMITATIONS OF TESTS
Laboratory Tests
The main advantages of laboratory testing
are economy and convenience. Laboratory
tests offer financial benefits accruing from the
fact that a large number of tests can be per-
formed routinely in a well-equipped soil testing
laboratory. Also, such testing can be perform-
ed by technicians, in contrast to the generally
higher level of skill required for the execution
and interpretation of field testings. Laboratory
testing offers protection from adverse weather
conditions. They can be performed at all times,
while field testing is necessarily restricted to
non-rainy, unfrozen seasons, which generally
means it must be performed during the summer.
However, such advantages are offset by an
array of disadvantages.
Limitations of Laboratory Tests
Olson and Daniel (1979) have listed poten-
tial sources of error in laboratory tests for both
saturated and unsaturated flow. These lists are
presented in Table 1.
50
-------
TABLE 1. SOURCES OF ERROR IN LABORATORY TESTING FOR
SATURATED AND UNSATURATED FLOW (Olson and Daniel, 1979)
Saturated
Unsaturated
(1) Non-representative samples (1)
(2) Voids formed during sample (Z)
preparation
(3) Smear zones (3)
(4) Alterations in clay chemistry (4)
(5) Air in sample (5)
(6) Growth of microorganisms (6)
(7) Meniscii problems in capillary (7)
tubes
(8) Use of excessive hydraulic gradients
(9) Temperature effects
(10) Volume change due to stress change
(11) Flow direction
Non-representative samples
Smear zones
Incorrect flow direction
Growth of microorganisms
Chemical effects
Temperature effects
Filter impedance
Of these possible sources of error, of
greatest concern is the use of samples that may
not be representative of field conditions. Dis-
turbed samples that are taken back to the
laboratory and repacked in the various types of
apparati will not have the same fabric, struc-
ture, or pore configuration as they did in their
undisturbed state. Additionally, the use of
small samples compounds the disturbed soils
problem by omitting macroscopic variations o-
ccurring at the field level such as sand lenses,
fissures and joints, and root holes. Therefore,
since laboratory tests are deficient with regard
to both microscopic and macroscopic features
of the soil in the field, extreme caution is urged
with regard to laboratory results' extrapolation
to the field scale level.
Comparison of Laboratory to Field Testing
Results
A comparison of corresponding laboratory
and field hydraulic conductivity test results will
show the field hydraulic conductivity to be
considerable, but unpredictably higher than
values measured in the laboratory. Olson and
Daniel (1979) compared field and laboratory
hydraulic conductivity values from the litera-
ture and found that the range of the ratios of
field K/laboratory K have been reported from
0.3 to 46,000 with 90% of the observations in
the range from 0.38 to 64 with field results
usually being higher than laboratory results.
They report the major causes of this effect as
being: (1) a tendency to run laboratory samples
on more clayey samples, (2) the presence of
sand seams, fissures, and other macrostructural
features in the field that are not represented
properly in laboratory tests, (3) the use of lab-
oratory K values backcalculated from consolid-
ation theory rather than directly measured
values, (4) the measurement of vertical flow K
in the laboratory and of horizontal K in the
field, (5) the use of distilled water in the lab-
oratory, and (6) air entrapment in laboratory
samples.
Field Tests
While field tests have some sources of
error similar to problems encountered in lab-
oratory testing, field testing has some unique
problems. One such difficulty arises from the
fact that most experience in in situ soil testing
has been on coarse-textured soils. Also, most
of the field-scale experience of soil engineers,
geohydrologists, and soil scientists has been
experimental rather than standardized and has
addressed problems such as ground water flow
and soil - water - plant relations, and this in-
formation is not always completely applicable
to the problems of hazardous waste landfills
and surface impoundments.
Another problem associated with field-
scale testing lies in the use of Darcy's Law in
the flow analysis, in the excessive attention
directed at the analysis rather than the value of
K. This is an unfortunate occurrence since no
other soil parameter is as likely to exhibit the
wide range of values (from 10~1 to 10~10 cm
sec~l) as does hydraulic conductivity. Impor-
tant considerations for quantification of field
level hydraulic conductivity are fabric and
spatial variability, as elaborated below.
51
-------
Fabric
Spatial Variation (Anisotrophy)
Of the parameters of significance in the
analysis of environmental problems of fine-
textured soils none is more influenced by fabric
than hydraulic conductivity (Mitchell, 1976). A
simple example of this phenomenon is provided
in a consideration of two equal volumes of soil
(A and B) with exactly the same properties with
the exception of one long continuous pore in
soil volume A and four discontinuous pores with
total equivalent pore volume equal to soil
volume A in soil volume B. Hydraulic conduc-
tivity will be dramatically different between
these two cases.
Therefore, it is of importance that the soil
testing required for site evaluation actually
test the soil fabric that will exist in the landfill
or surface impoundment. The tested soil fabric
will be of two separate types: soil fabric of the
liner and soil fabric of the unsaturated zone
between the liner and ground water table or
bedrock.
Soil Fabric and Soil Liners
The importance of soil fabric to considera-
tions of the hydraulic conductivity of soil liners
cannot be overstated. In a chapter of Fabric,
Structure, and Property Relationships, Mitchell
(1976) states that the main conclusion to be
drawn from the considerations of his chapter is
that the geotechnical properties of any given
soil are dependent on the structure to such an
extent that analyses based on properties deter-
mined from the same material but with a
different structure may be totally in error.
Therefore, considerations of use of a particular
soil for a liner must be based on soil testing
methods that simulate the field compactive
effort to the highest degree possible.
Soil Fabric in the Unsaturated Zone
As mentioned in "Limitations of Laboratory
Tests" one considerable failing of laboratory
tests is that the soil fabric of the tested sample
will not be the same as the fabric of the soil in
situ. Lambe and Whitman (1979) state that
because permeability depends very much on soil
fabric (both microstructural—the arrangement
of individual particles, and macrostructural
such as stratification) and because of the diffi-
culty of getting representative soil samples,
field determinations of hydraulic conductivity
are often required to get a good indication of
the average hydraulic conductivity.
The soil is not a static body, but rather a
dynamic set of processes always being affected
by the soil-forming factors of climate, parent
material, topography, vegetation, and time.
Therefore, a soil body will vary in both the
vertical and horizontal directions. For
example, vertical variation is manifested in
different hydraulic conductivities between suc-
cessive soil horizons, and horizontal variation is
manifested in differing thicknesses of soil
horizons between soil profiles of the same soil
series. Thus non-homogeneity (anisotrophy) in
soils is the rule rather than the exception, and
soil testing of homogenous or mixed samples for
average conditions may be totally in error in
predicting conditions that will exist in the field.
Additionally, during the site evaluation and
review process, attention must be devoted to a
qualification of errors involved in soil testing
results. That is, some determination of
whether the variation in soil testing results is a
consequence of the soil's spatial variability or a
variation inherent in the soil testing method
itself. At this time, excepting the information
included in Roberts and Nichols (1980) little
information is available to quantify the com-
ponents of variation in the results of many soil
testing methods.
SOIL TESTING MATRIX
The rating of the various soil testing
methods presented must be predicated upon the
objectives of the results' interpretation or ex-
trapolation. As mentioned earlier in this paper,
there are three situations where water move-
ment is important: 1) in the root zone if plants
are part of the receiver system, 2) around the
liner where natural soil is compacted, and
3) the unsaturated zone between root zone or
liner and ground water level or bedrock. There-
fore, soil testing methods should be chosen with
regard to which of the three types of water
movement will be most important in a proposed
waste disposal site.
Table 2 is an attempt to rate the applica-
bility of the various soil testing methods for
1) root zone movement with crops, 2) liner
evaluation, and 3) the unsaturated zone be-
tween waste and ground water.
52
-------
TABLE 2. SUMMARY OF TESTING METHODS
TYPE OF TEST
Root Unsat.
Zone Liner Zone
Comments
SATURATED-LABORATORY
pressure cell 3
compaction mold 3
consolidation cells 3
triaxial apparatus 3
micromorphometric data 2
SATURATED - FIELD
piezometer 3
double ring infiltrometer 2
per mea meter 2
cased boreholes 4
well pumping tests 4
UNSATURATED-LABORATORY
steady state (long column) 2
pressure outflow 3
impeding layer 2
instantaneous profile 2
UNSATURATED-FIELD
crust 1
traditional instantaneous 1
profile
unit gradient-drainage 2
instantaneous profile
Libardi instantaneous profile 2
3 4 small sample; only saturated K
1 4 simulates remolded soil liner and waste;
only saturated K
2 4 simulates remolded soil liner;
only saturated K
2 4 simulates remolded soil liner;
only saturated K
2 2 only experimental at this time; potential
value due to fabric quantification
4 3 generally of limited value; self-boring
piezometer has potential for clay soils
3 3 easy to perform; only saturated K
3 3 easy to perform; only saturated K
4 3 easy to perform; calculation of K though
questionable on silts,
not applicable to clays
4 3 expensive and requires presence of
ground water table; again questionable
for silts and not applicable to clays
2 2 larger sample than short column;
longer time period required;
can get K for layers
3 3 small sample; difficulties associated
with age of excessive pressure heads
2 2 potential exists for use of liner material
rather than crust
2 2 small sample
3 1 both unsaturated and saturated K;
potential to simulate liner in situ
3 1 measures water content and pressure head
so thatK(O) or K(h) can be determined
3 2 simplified instantaneous profile method;
less field time than traditional method
3 2 less field time than unit gradient-
drainage method; needs field verification
THE ABC EQUATION
Beyond the applicability of the individual
testing methods for the three types of water
movement of concern in land-based waste dis-
posal systems) some remarks can be made re-
garding the present status of available infor-
mation, what is the end point we are trying to
obtain, and how do we get there from here.
The end product of research on soil testing
methods for hydraulic conductivity determina-
tion could be a formula in the form of "a"
number of "b" type of soil tests per "c" unit of
land area, (abc equation) that could be used on a
uniform basis nationally for any proposed waste
disposal sites.
However, before discussing the abc equa-
tion, one other point regarding site assessment
and soil testing methods needs to be empha-
sized. Site assessment and soil testing methods
need to interact and be performed in conjunc-
tion with each other rather than be separate
and distinct segments of the longer study since
test result numbers are only as good as how
53
-------
well the whole site is characterized. For
problems of practical interest, it is clear that
permeability tests (laboratory or field) must be
performed with a great degree of care and
attention to detail. However, just performing
the tests properly still does not insure success-
ful results. Thorough field investigation to
identify zones of maximum and minimum per-
meability, and careful selection of samples or
layers for testing, are in some respects more
important than experimental technique. Even
with a comprehensive field investigation and
suitable experimental technique, some degree
of judgment must inevitably be exercised
before results are used for field predictions
(Olson and Daniel, 1979). The following
example can be used to illustrate this point.
The usual approach that does not integrate
the site assessment into the soil testing pro-
gram will take a site and place a grid over the
study are as shown below with a 180-acre site
having 1 test per 5 acres or 36 soil-testing
locations.
Figure 4. Site topo with testing grid.
One simple twist of the site assessment by
use of Soil Conservation Service Soil Surveys
which can give preliminary soils information
which may eliminate some of the areas from
the testing program. Show below is the soil
survey of the area shown above. From pre-
liminary study of the Soil Survey we discover
that only the soil series "a" and "b" appear
applicable for waste disposal. Gridding this
area using the same parameters as above, we
now need only 20 tests.
Figure 5. SCS Survey with testing grid.
However, SCS maps may lack site-specific
precision due to mapping scale, usually
1:20-24,000. Thus most county soil surveys can
only represent general soil bodies and not the
specific soil physical and hydrological proper-
ties needed to evaluate and design land-based
waste disposal systems. Since a "good" soil
survey is usually considered to be 70-80%
accurate with soil boundary lines within 200-
300 feet, such accuracy can be further im-
proved by the additional use of color aerial
photographs and/or on-ground, detailed soil
mapping or ground-truthing as shown in Figure
6. This figure shows the ABCDirt survey that
utilized aerial photographs coupled with de-
tailed soil mapping. This can be compared to
the SCS map in Figure 5.
Figure 6. Detailed soil survey.
54
-------
At this point it should be noted that the
detailed soil survey discovered a soil series
variant, "cl" which was not noted at all on the
SCS Soil Survey. Since this soil was prelimi-
narily appropriate for waste disposal, this area
was included in the area to be tested.
Going one step further beyond the detailed
soil survey could be the consideration of varia-
tion within each individual soil series which can
help determine how many tests should be per-
formed and where these tests should be located
(such a procedure will also be useful to the
monitoring system design). In this example
(Figure 7) soil series "cl" proved to be very
uniform and thus three testing sites were
equally spread over the area. However, in soil
series "b" the percent clay content in the
argillic horizon was found to significantly vary
in the northwesterly direction. Therefore, test-
ing sites were located at the two extremes and
one in the middle of this soil variation transect
meant to represent the average or modal loca-
tion for this soil property, clay percentage-
Extended out from the modal site are soil
property isotherms intended to represent simi-
lar properties to the modal site which will help
to statistically support the soil testing results
at the modal location. Likewise for soil series
"a" the thickness to limiting layer was found to
signficantly vary in a near northerly direction
and a soil variation transect and soil property
isotherm was established.
This example demonstrates how helpful the
integration of site assessment information into
the soil testing program can be and its impor-
tant practical significance, as the total number
of tests could be dramatically reduced (from 36
in Figure 4 to 13 tests in Figure 7) with the
location of the actual testing sites established
according to soil variation. Such a procedure
would be a vast improvement over trying to
grid a site to obtain an "average" hydraulic
conductivity for a whole site.
Returning to the abc equation and the
above discussion, it can be seen that using a soil
series basis for "c" unit of land areas would be
an improvement over use of some arbitrary land
area such as acre or hectare because such a
procedure would allow for less tests on a more
uniform site and require more tests on sites
that contain several soil series. Further, use of
the soil series basis will allow the utilization of
the vast amount of soils information that is
already available in this form from the USDA
which presently has data on approximately
12,000 soil series. Also, use of this type of
terminology will allow for national consistency
and also fit well with the approaches taken by
other nations.
Considering the "b" types of tests part of
the equation, it appears that attempts such as
the Soil Testing Methods Matrix are beginning
to provide answers to this problem even to the
extent of assessing which types of tests are
most applicable to the three categories of
water movement that arise from the various
types of land-based waste disposal.
Lastly, the "a" number of tests part of the
abc equation remains the most unclear due to
the paucity of available information regarding
inherent variation of the testing methods as
well as capabilities to quantitatively assess
field soil spatial variability of hydrologic prop-
erties important in waste disposal. However, if
done as suggested above with transects and
isotherms, soil testing information from exist-
ing and newly proposed sites can be a valuable
aid to research results that can accelerate the
pace toward resolving the "a" number of "b"
types of tests and establishing an abc equation
that could be used on a uniform basis nationally
for any type of land-based waste disposal site.
Figure 7. Detailed survey with transects and
isotherms.
55
-------
REFERENCES
1. ASTM Committee D-18. 1979- Standard
Test Method for Infiltration Rate of Soils
Using Double-Ring Infiltrometers, ASTM D
3385-75. Annual Book of ASTM Standards,
Part 19-
2. Bertrand, A. R. 1965. Rate of V/ater
Intake in the Field. In: Methods of Soil
Analysis. C. A. Black, Ed. Amer. Soc. of
Agronomy, Madison, Wisconsin. pp. 197-
208.
3. Black, T. A., W. R. Gardner, and G. W.
Thurtell. 1969. The Prediction of Evapor-
ation, Drainage, and Soil Water Storage for
a Bare Soil. Soil Sci. Soc. Amer. Proc.
33: 655-660.
4. Boersma, L. 1965a. Field Measurement of
Hydraulic Conductivity Below A Water
Table. In: Methods of Soil Analysis, C. A.
Black, Ed. Amer. Soc. of Agronomy,
Madison, Wisconsin, pp. 222-233.
5. Boersma, L. 1965b. Field Measurement of
Hydraulic Conductivity Above A Water
Table. In: Methods of Soil Analysis, C. A.
Black, Ed. Amer. Soc. of Agronomy,
Madison, Wisconsin, pp. 234-252.
6. Bouma, J., A. Jongerius, and D. Schoon-
derbeek. 1979- Calculation of Saturated
Hydraulic Conductivity of Some Pedal Clay
Soils Using Micromorphometric Data. Soil
Sci. Soc. Amer. J. 43: 261-264.
7. Brewer, R. 1976. Fabric and Mineral
Analysis of Soils. R. E. Krieger Publishing
Co., Huntington, N. Y. 482 pp.
8. Childs, E. C. and N. Collis-George. 1950.
The Permeability of Porous Materials.
Roy. Soc. (London), Proc., A. 201:
392-405.
9. Davidson, J. M., L. R. Stone, D. R. Nielsen,
and M. E. LaRue. 1969- Field Measure-
ment and Use of Soil-Water Properties.
Water Resources Res. 5: 1312-1321.
10. Flocker, W. J., M. Yamaguchi, and D. R.
Nielsen. 1968. Capillary Conductivity and
Soil Water Diffusivity Values from Vertical
Soil Columns., Agron. J. 60: 605-610.
11. Gardner, W. R. 1970. Field Measurement
of Soil Water Diffusivity. Soil Sci. Soc.
Amer. Proc. 34: 832-833.
12. Gillham, R. W., A. Klute, and D. F. Heer-
mann. 1976. Hydraulic Properties of a
Porous Medium: Measurement and Empiri-
cal Representation. Soil Sci. Soc. Amer.
J. 40: 203-207.
13. Hillel, D. and W. R. Gardner. 1970.
Measurement of Unsaturated Conductivity
and Diffusivity by Infiltration through an
Impeding Layer. Soil Sci. 109: 149-153.
14. Jezequel, J. F. and C. Mieussens. 1975. In
Situ Measurement of Coefficients of Per-
meability and Consolidation in Fine Soils.
In: Proceedings of the Conference on In
Situ Measurement of Soil Properties,
A. S. C. E., New York, N. Y., Vol. L,
pp. 208-224.
15. Klute, A. 1965a. Laboratory Measurement
of Hydraulic Conductivity of Saturated
Soil. In: Methods of Soil Analysis, C.A.
Black, Ed. Amer. Soc. of Agronomy,
Madison, Wisconsin, pp. 210-220.
16. Klute, A. 1965b. Laboratory Measurement
of Hydraulic Conductivity of Unsaturated
Soil. In: Methods of Soil Analysis. C. A.
Black, Ed. Amer. Soc. of Agronomy,
Madison, Wisconsin, pp. 253-261.
17. Klute, A. 1965c. Water Diffusivity. In:
Methods of Soil Analysis, C. A. Black, Ed.
Amer. Soc. of Agronomy, Madison, Wiscon-
sin, pp. 262-272.
18. Klute, A. 1972. The Determination of the
Hydraulic Conductivity of Unsaturated
Soils. Soil Sci. 113: 264-276.
19. Lambe, T. W. and R. V. Whitman. 1979.
Soil Mechanics, SI Version. John Wiley and
Sons, New York, N. Y. 553 pp.
20. Libardi, P. L., K. Reichardt, D. R. Nielsen,
and J. W. Biggar. 1980. Simple Field
Methods for Estimating Soil Hydraulic
Conductivity. Soil Sci. Soc. Amer. J. 44:
3-7.
21. Marshall, T. J. 1958. A Relation Between
Permeability and Size Distribution of
Pores. J. Soil Sci. 9: 1-8.
22. Matrecon, Inc. 1980. Lining of Waste
Impoundment and Disposal Facilities
Manual (Draft). U. S. Environmental Pro-
tection Agency, Cincinnati, Ohio. 595 pp.
56
-------
23. Matyas, E. L. 1967. Air and Water Per-
meability of Compacted Soils. In: Perme-
ability and Capillarity of Soils, ASTM STP
417, Amer. Soc. of Testing Materials.
pp. 160-175.
24. Means, R. E. and J. V. Parcher. 1963.
Physical Properties of Soils. Charles E.
Merrill Books Inc., Columbus, Ohio.
464 pp.
25. Milligan, V. 1975. Field Measurement of
Permeability in Soil and Rock. In: Pro-
ceedings of the Conference on In-Situ
Measurement of Soil Properties, Vol. IE,
Specialty Conference of the Geotechnical
Div., A.S.C.E., N. Carolina State
University, Raleigh, North Carolina, pp. 3-
36.
26. Millington, R. J. and J. P. Quirk. I960.
Transport in Porous Media. Trans, of 7th
Int. Congr. Soil Sci., Madison, Wisconsin, I:
97-106.
27. Millington, R. J. and J. P. Quirk. 1961.
Permeability of Porous Solids. Trans.
Faraday Soc. 57: 1200-1207.
28. Millington, R. J. and J. P. Quirk. 1964.
Formation Factor and Permeability Equa-
tions. Nature 202: 143-145.
29. Mitchell, J. K. 1976. Fundamentals of Soil
Behavior. John Wiley and Sons, Inc., New
York, N. Y. 422 pp.
30. Nielsen, D. R., J. M. Davidson, J. W.
Biggar, and R. J. Miller. 1962. Water
Movement through Panoche Clay Loam
Soil. Hilgardia35: 491-506.
31. Olson, R. E. and D. E. Daniel. 1979. Field
and Laboratory Measurement of the Per-
meability of Saturated and Partially Satu-
rated Fine-Grained Soils. Geotechnical
Eng. Rep. GE 79-1, Dept. of Civil Eng.,
Univ. of Texas, Austin. 78 pp.
32. Roberts, D.W. and M.A. Nichols. 1980.
Soil Properties, Classification, and
Hydraulic Conductivity Testing (Draft).
U.S. Environmental Protection Agency,
Cincinnati, OH. 235 pp.
33. Simmons, C. S., D. R. Nielsen, and J. W.
Biggar. 1979- Scaling of Field-Measured
Soil-Water Properties. I. Methodology, IE.
Hydraulic Conductivity and Flux. Hilgar-
dia47: 77-153.
34. Small Scale Waste Management Project
(SSWMP). 1978. Management of Small
Waste Flows. EPA-600/2-78-173, U. S.
Environmental Protection Agency, Cincin-
nati, Ohio. 764 pp.
35. Tzimas, E. 1979. The Measurement of
Soil-Water Hysteretic Relationships on a
Soil Monolith. J. Soil Sci. 30: 529-534.
36. Watson, K.K. 1966. An Instantaneous
Profile Method for Determining the Hy-
draulic Conductivity of Unsaturated Porous
Materials. Water Resources Res. 2: 709-
715.
37. Yong, R. N. and B. P. Warkentin. 1975.
Soil Properties and Behaviour. Elsevier
Scientific Publishing Co., New York, N.Y.
449 pp.
57
-------
II.
EVALUATION OF MOLECULAR MODELLING TECHNIQUES
TO ESTIMATE THE MOBILITY OF ORGANIC CHEMICALS IN SOILS:
Water Solubility and the Molecular Fragment Mobility Coefficient
James Dragun
U.S. Environmental Protection Agency
Washington, DC 20460
Charles S. Helling
U.S. Department of Agriculture
Beltsville, MD 20705
ABSTRACT
This is the second in a series of studies evaluating the use of various molecular
modelling techniques to estimate the mobility of organic chemicals in soils. The objec-
tives of this study were to (a) identify molecular fragments within organic chemicals that
may significantly enhance or retard mobility in soils, (b) identify fragments that do not
significantly enhance or retard mobility, (c) quantify the fragment effects and (d) incor-
porate these effects into the estimation method that predicted mobility from the logarithm
of the solubility. Soil thin-layer chromatography was used to quantify the mobility of 55
pesticides, having diverse chemical structures, in Hagerstown silty clay loam soil. Thir-
teen fragments were identified and divided into two groups. One group of fragments pos-
sessed an electronegative or electropositive character whose mobility enhancement or
retardation effects were not created by, or were negligibly affected by, aromatic ring
resonance. In the second group, ring resonance played an important role in establishing
mobility effects. Two least squares multiple regression models, incorporating water solu-
bility and quantifying the surface repulsion or attraction effect as a molecular fragment
mobility coefficient, successfully predicted Rf (R2 = 0.93 and 0.94). These two
models were superior techniques for predicting the mobility of diverse organic chemicals
compared to the estimation technique utilizing solubility as the only independent
variable.
INTRODUCTION
Knowledge concerning the potential ad-
sorption of organic chemicals to sediments
and soils is essential for assessing the
hazard to human health and the environ-
ment associated with various wastewater,
sludge and chemical use and disposal
methods. The degree of adsorption will
affect the mobility of a chemical as well
as other fate determining processes such as
volatilization, photolysis, hydrolysis and
biodegradation. Legislation such as the
Federal Water Pollution Control Act and its
Amendments (FWPCA), the Toxic Substances
Control Act (TSCA), the Federal Insecti-
cide, Fungicide, and Rodenticide Act
(FIFRA), the Resource Conservation and Re-
covery Act (RCRA) has created a need for
adsorption and mobility data.
The most widely used and accepted es-
timation techniques to predict the extent
of adsorption involve correlation of the
solid-to-solution distribution coefficient
(Kj) or the retention factor (Rf), de-
rived from soil thin-layer chromatography
(TLC), to some measure of the molecular
hydrophobicity. The measure usually chosen
is the logarithm of the solubility (log S)
[Karickhoff et al . (14); Dragun et al . (3)]
or the logarithm of the octanol:water par-
58
-------
tition coefficient (log P) [Briggs (2);
Dragun et al. (3); Karickhoff et al. (14)].
However, these estimation techniques
usually provide only general approxinations
of the mobility of a chemical in soil or
the potential for adsorption to sedinents.
For example, when Koc (i.e., the
adsorption coefficient divided by the
fractional organic carbon content present
in the soil or sediment) versus solubility
was plotted [Mill and Mabey (16)] using
data fron three sources, the predicted
values of Koc varied from observed
values by at least one order of magnitude
for 20 of the approximately 100 chemicals,
and by at least two orders of magnitude for
3 chemicals.
While organic chemical mobility is
mainly governed by its lipophilic attrac-
tion to soil organic matter, it is affect-
ed to some extent by electrical charge
phenomena. Recent reviews [Harter (8);
Theng (21)] have implicated the type and
amount of clay, the exchange capacity and
surface acidity as factors affecting the
extent of adsorption. Dragun et al . (3),
using molecular connectivity indices and
molecular charge transfer calculations,
concluded that size and shape of the mole-
cules and molecular polarization induced by
clay particle charge fields appear to be of
negligible importance compared to the
hydrophobic-hydrophilie balance of the
molecule for mobility prediction purposes.
However, this study also suggested that
future attempts to model chemical mobility
in soils should consider the interaction of
molecular fragments with colloid surfaces
bearing permanent and pH-dependent charge.
The attractive and repulsive interactions
of various molecular fragments with colloid
surfaces has been studied [Mortland (17,
18); Tahoun and Mortland (20); Theng (21)].
However, no attempt has been made to quan-
tify these molecular fragment interactions
into a mobility estimation technique. The
purposes of this study were to identify the
mobility enhancement and retardation
effects of various molecular fragments and
to quantitatively incorporate these effects
into an existing estimation technique based
on solubility. This report is the second
in a series of studies evaluating the use
of various modelling techniques to estimate
the mobility of organic chemicals in soils.
THEORETICAL DISCUSSION
Hammett (6) related the rate constant
of a meta or para substituted derivative of
C^Hj-x (with X as the reacting center)
to the parent structure using the general
relationship (the Hammett equation):
log (ks/k0)
= per
[1]
The quantity kQ represents the rate con-
stant of the unsubstituted parent structure
and ks that of the substituted deriva-
tive. The quantity a is a Gibbs energy re-
lated fragment or substituent constant that
reflects the ability of the fragment to
attract or repel electrons and p is a
constant characteristic of the type of
reaction. By analogy with the Hammett
equation, Hansch and Leo (7) described
1ipophilicity as:
log [PRX/PRH] = PT(X)
C2]
where PRH and ?RX represent the solvent-
solvent partition coefficients of the
organic molecule R containing H and X
fragments; TT(X) represents the hydrophobic
fragment constant, i.e., the contribution
of substituent X to the 1ipophilicity of
the structure RH when X replaces a hydrogen
aton in RH; and p reflects the characteris-
tics of the solvent pair used in deter-
mining the partition coefficient. The
constant TT is a Gibbs energy related
constant which measures the difference in
the Gibbs energy changes involved in moving
RH and RX from one solvent phase to
another. The practical applications of
Equation 2 are significant. For any
structure R, the total lipophilicity as
measured in an octanol:water system can be
construed by simple addition of the proper
IT values characterizing the fragments
[Hansch and Leo (7)]:
log PR =
[3]
where n is the total number of fragments
comprising the structure R. Dragun et al.
(3) showed that Rf values >0.1 can be
represented as a linear function of log S,
which in turn is linearly related to
log PR. Therefore, Rf can be defined
in terms of IT or log S for the structure R:
Rf = a
+ b
or alternatively by
Rf = c log S + e
[4]
[5]
59
-------
where a, b, c, and e are constants. By
analogy with a of Equation 1, the quantity
8 can be defined as the substituent or
fragment coefficient that reflects the
Gibhs energy effects caused by attraction
to -or repulsion from colloid surface
charges, thereby retarding or enhancing
nobility (i.e., the molecular fragment
mobility coefficient). Ue assumed that the
magnitude of a fragment's mobility en-
hancing or retarding effect was a constant,
independent of the structure of the parent
molecule. An analysis of the sediment ad-
sorption data of Hassett et al. (9) for the
primary amine fragment of various poly-
nuclear aromatic amines indicates that
these two assumptions appear reasonable.
We also assumed that, for an organic chemi-
cal RX having the fragment X that interacts
with soil colloid surfaces, the a values
are independent of, and additive with, the
TV values. Then, R^ can be defined in terms
of a and IT or log S:
Rf = a ETT(X J + P3 + b [6]
t n n
or
Rf = c log S + pa + e [7]
The quantity p is a constant charac-
teristic of the soil system. The objec-
tives of this study were to (a) identify
molecular fragments that may significantly
enhance or retard organic chemical nobility
in soils, (b) identify fragments that do
not significantly enhance or retard
mobility, (c) quantify 9 for fragments
identified in (a) above, and (d) incor-
porate 3 into the log S - R-r relationship
to provide an improved approach for esti-
mating the mobility of organic chemicals in
soils.
MATERIALS AND METHODS
Soil Mobility Determination
Soil TLC is a laboratory method using
soil as the adsorbent phase in a TLC sys-
tem. By its innate homogeneity, the ad-
sorbent in TLC represents the milieu of a
chemical moving through soil aggregates
rather than around them. This relatively
simple system yields quantitative data on
the mobility of organic chemicals that
correlate well with trends noted in the
literature on field tests and monitoring
studies. Soil TLC [Helling (10)] was used
as the basis for quantifying the mobility
of organic chemicals used in this study.
Rf values for most of these chemicals on
Hagerstown silty clay loam soil have been
previously reported [Helling (10); Helling
et al. (11, 12)]. Rf values for norflu-
razon and perfluidone were determined on
Magerstown silty clay loam soil by the same
method used for the other chemicals. A
standard development distance, 100 mm, was
used on all plates. The Rf is obtained
directly by measuring the distance the
chemical migrates in mm on the plate and
dividing this distance by 100 mm (e.g., 70
mm = Rf of 0.70). Table 1 contains the
common and chemical names and the Rf
values of the 55 organic chemicals used in
this study.
The soil contained 2.50% organic mat-
ter and 39.5% clay. The soil pH was 6.8.
Moisture capacity at a pressure of 33 kPa
(1/3 bar) was 34.1% for soil sieved to less
than 250 um. The dominant clay minerals
were vermiculite and kaolinite.
Solubility Determinations
The water solubilities of the 55
organic chemicals used herein were obtained
from the literature and are listed in
Table 1.
Molecular Fragment Selection and Analysis
The organic chemicals selected for
this study have diverse chemical struc-
tures, thereby allowing the analysis of the
mobility enhancement or retardation effects
of a large number of molecular fragments.
Most of these chemicals are polyfunctional ,
i.e., they contain more than one fragment
or substituent group. However, using the
assumptions and methods described in this
section, most chemicals were found to be
monofunctional with respect to mobility
enhancement or retardation, even though
they were polyfunctional in structure.
First, the least squares linear regression
of solubility on Rf was calculated:
Rf = 0.20 log S + 0.03; (RZ = 0.50) [8]
Then, all organic chemicals containing an
aromatic fragment were plotted in relation
to Equation 8 (in a manner analogous to
Figure 1). Aliphatic chemicals were not
treated in this analysis using Equation 8
for reasons that are discussed in the Re-
sults and Discussion section. Chemicals
containing a given fragment were identified
60
-------
TABLE 1. COMMON AND CHEMICAL NAMES, Rf AND THE LOG OF THE
SOLUBILITY IN PPM (log S) OF CHEMICALS USED IN THIS STUDY
No.
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
Common Name
Ametryne
Amiben
Amitrole
Atratone
Atraz ine
Aziprotryne
Bar ban
Bromacil
Captan
Carbaryl
CDAA
Chlor-
bromuron
Chloroxuron
Chlorpropham
(CIPC)
Cyclohexi-
mide
2,4-D
Dalapon
Dtcamba
Dichlobenil
Diphenamid
Diuron
Chemical Name
2-(ethylamino) -4-
( isopropylamino)-6-
(methylthio) -s-
triazine
3-amino-2, 5-
dichlorobenzoic acid
3-amino-s-triazole
2-ethylamino-4-
isopropylamino-6-
methoxy-s-triazine
2-chloro-4 -ethyl
amino-6-isopropyl-
amino-s-triazine
2-azido-4-( isopropyl-
amino) -6 -(me thy 1th io)-
s-triazine
4-chloro-2-butynyl jn-
chlorocarban ilate
5-bromo-3-sec-butyl-
6-methyluracil
N-[ ( trichloromethyl ) -
thio] -4-cyclohexene-
1, 2-dicarboximide
1-naphthyl
methylcarbamate
N,N-diallyl-2-
chloroacet amide
3-(4-bromo-3-
chlorophenyl ) -1-
methoxy-1-methylurea
3-[p-(p-chloro-
phe"noxy) -phenyl] -
1,1, -dime thy lure a
isopropyl m-chloro-
carbanilate
3-[2-(3,5-dimethyl-
2-oxocyclohexyl ) -
2-hydroxyethyl]
glatar imide
2, 4-dichlorophenoxy-
acetic acid
2, 2-dichloropro-
pionic acid
3,6-dichloro-g-
anisic acid
2, 6-dichlorobenzo-
nitr ile
N,N-dimethyl-2,2-
diphenyl ace t amide
3-( 3,4-dichloro-
phenyl) -1, 1-dimethyl-
urea
Rf
0.44
0.91
0.73
0.56
0.47
0.30
0.06
0.69
0.39
0.38
0.82
0.14
0.09
0.18
0.89
0.69
0.96
0.96
0.22
0.49
0.24
log S a
2.27
2.85
5.45
3.22
1.52
1.88
1.04
2.91
-0.30
K
1.60°
r»
4.30C
1.70
0.57
1.95
rt
4.32d
2.79
5.70
3.65
1.26
2.41
1.62
61
-------
No.
Common Name Chemical Name
log s
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
Fenac
Fenaminosulf
(Dexon)
Fenuron
Fluometuron
Formetanate
Hydrochlor ide
Linuron
MCPA
Metobromuron
Monuron
Neburon
Norflurazon
Nortron
Oxycarboxin
PCP
Perfluidone
Phenmed ipham
Picloram
PPG - 124
2,3,6-trichloro- 0.84
phenylacetic acid
p-(dimethylamino)- 0.88
"Benzenediazo
sodium sulfonate
l,l-dimethyl-3- 0.69
phenylurea
l,l-dimethyl-3- 0.50
(alpha,alpha,alpha-
tri fluoro-m-
tolyl)urea
m- [(dimethylamino)- 0.75
methylene]amino]-
phenyl methylcarbamate
hydrochloride
3-(3,4-dichloro- 0.17
phenyl)-1-methoxy-l-
methylurea
2-chloro-4-methyl 0.79
phenoxyacetic acid
3-(p-bromophenyl)- 0.31
1-methoxy-l-methyl-
urea
3-(p-chlorophenyl)- 0.48
1,1-dimethylurea
l-butyl-3-(3,4- 0.07
dichlorophenyl)-1-
methylurea
4-chloro-5-(methyl- 0.40
amino-2-(3-tri fluoro-
methyl-phenyl)pyrid-
azin-3-one
6-chloro-2-trifluoro- 0.65
methyl-3H-imidazo-
[4,5-b]pyridine]
5,6-dihydro-2- 0.73
methy1-1,4-oxathiin-
3-carboxanilide
4,4-dioxide
pentachlorophenol 0.40
1,1,1-trifluoro-N- 0.82
[2-methyl-4-(phenyl-
sulfonyl)-phenyl]-
methane-sulfonamide
methyl m-hydroxy- 0.17
carbanilate m-
me thyIcarbanilate
4-amino-3,5,6- 0.84
trichloropicolinic
acid
p-chlorophenyl 0.57
methylcarbamate
2.30
4.48
3.59
1.90
3.00
1.88
2.95
2.52
2.36
0.68
1.45
2.04
3.00
1.30fc
1.78C
1.00
2.63
2.76e
62
-------
No.
40.
41.
42.
43.
44.
45.
46.
47.
48.
49.
50.
51.
52.
53.
54.
55.
Common Name
Promecarb
Prometone
Prometryn
Propachlor
Propanil
Propazine
Prop ham
Pyrazon
Stdurbn
Simazine
Simetone
Stmetryn
Solan
2,4,5-T
TCA
Trietazine
Chemical Name
3-isopropyl-5-methyl
phenyl methylcar-
bamate
2,4-bis( isopropyl-
amino)-6-methoxy-s-
triazine
2,4 bis( isopropyl-
amino)-6-methylthio-
s-triazine
2-chloro-N-
isopropyl-acetanilide
31 ,4'-dichloro-
propionanilide
2-chloro-4 ,6-
bis( isopropyl-
amino)-s-triaztne
isopropyl carbani-
late
5-amino-4-chloro-2-
phenyl-3(2H)-
pyr idazinone
l-(2-methylcyclo-
hexyl ) -3-phenyl-
urea
2-chloro-4, 6-
bis(ethylamino)-s-
triazine
2,4-bis(ethylamino)-
6-methoxy-s-triazine
2, 4-bis(ethylamino)-
6- ( me thy 1th io ) -s-
triazine
3 '-chloro-2-methyl-
p-valerotoluidide
"2,4, 5-trichloro-
phenoxyacetic acid
trichloroacetic acid
2-chloro-4-diethyl-
amino-6-ethylamino-
^-triazine
Rf
0.34
0.60
0.25
0.63
0.24
0.41
0.51
0.44
0.30
0.45
0.45
0.33
0.08
0.54
0.96
0.36
log S a
1.96
2.88
1.68
2.85
2.35
0.93
2.10
2.60
1.26
0.70
3.51
2.65
0.95
2.38
6.00
1.30
a From the Pesticjde Manual, British Crop Protection Council
(1974) at 20-25 c, unless specified otherwise.
b Measured at 30°C.
c From Thomson WT, Agricultural Chemicals Series, 1979 Revision,
Thomson Publications, Fresno, CA.
d Measured at 2°C.
e From Herbicide Handbook, Weed Science Soc. America (1974).
63
-------
on the plot in relation to Equation 8. The
distribution of the fragment was then clas-
sified usinq inductive reasoninq as either
(a) mobility enhancinq (i.e., most chemi-
cals with a qiven fraqment, when plotted in
relation to Equation 8, appeared to the
left of reqression line), (b) mobility
retardinq (i.e., most chemicals appeared to
the riqht of the reqression line) or (c) no
observable effect. Karickhoff et al. (14)
showed that mono-, di- and polycyclic
aromatic hydrocarbon and chlorinated hydro-
carbon adsorption is hiqhly correlated to
solubility. This study assumed that Rf
is highly correlated with water solubility
for unsubstituted aliphatic and aromatic
hydrocarbons. Therefore, aliphatic and
aromatic hydrocarbon fragments were
classified as having a statistically
nonsignificant effect. An analysis of the
distribution of these fragments shows that
this classification, in general, was valid
for the chemicals used in this study. The
results of these fragment plots were quan-
tified and statistically analyzed by
developing the linear least squares fit
model that was analogous to Equation 7:
Rf = m log S +
—- + d3r
^ +
+ e
[9]
where
e, m = constants
S = water solubility in ppm
sl' 82> 8n = molecular fragment
mobility coefficients
b, c, d = unity, if the molecule
contains the fragment
that is classified as
mobility enhancing or
retarding
or b, c, d = zero, if the fragment is
not classified as
mobility enhancing or
retarding (i .e., frag-
ment has no observable
effect), or if the
fragment is not present
within the molecule.
The quantity p, derived in Equation 7, was
assigned a value of unity and is, there-
fore, not found in Equation 9.
Since the slope and intercept of
Equation 9 were somewhat different than
those of Equation 8, all organic chemicals
containing an aromatic fragment were again
plotted, but in relation to Equation 9 with
the values b, c and d equal to zero. Chem-
icals containinq a given fraqment were
again identified on the plot, but in rela-
tion to the modified version of Equation 9.
The fragment was then reclassified using
inductive reasoning as either mobility en-
hancing, mobility retarding or no effect.
Then, the 8 values were reestimated for
those fragments that either enhanced or
retarded mobility through the development
of a new linear least squares fit model
similar to Equation 9. This iterative re-
estimation of the 8 values was repeated
three times. On the third iteration, no
new fragments were identified and no frag-
ment changed its mobility class.
RESULTS AND DISCUSSION
Aliphatic Chemicals
Figure 1 shows the qraphical rela-
tionship between Rf and log S for the or-
ganic chemicals used in this study. Two
distinct relationships exist between Rf
and log S: one Rf - log S relationship
fits chemicals containing no aromatic
fragments (i.e., aliphatics) and one fits
chemicals containing at least one aromatic
fragment (i.e., aromatics). Since soil or-
ganic matter contains both aliphatic and
aromatic constituents, the strength of
chemical binding to soil may be influenced
by the inherent aliphatic or aromatic char-
acter of the chemical. Therefore, these
two distinct groups were treated separately
in this study. This observation also sug-
gests that future attempts to model mobil-
ity should distinguish aliphatic from aro-
matic chemicals.
Six organic chemicals comprise the
aliphatics group. Each of these contains
at least one of these molecular fragments
that possibly interact with colloid sur-
faces [Harter (8); Portland (17)]: the
carbonyl, the primary amine, and the car-
boxylic acid group. Three chemicals con-
tained one of these fragments (Mo. 11, 17,
54), two contained two molecular fragments
(No. 8, 9), and one contained three frag-
ments (No. 15). Since each of the chemi-
cals contained at least one fragment that
may interact with colloid surfaces, 9
values were not calculated.
Aromatic Chemicals
Table 2 identified the molecular frag-
ments that did not enhance or retard the
64
-------
"f
1.0-
09-
0.3 •
02
AROMATICS
R, = 0 03 + 0 20 Log S
R2 = 0 50
(Equation 8)
60 Log S. ppm
Figure 1. Linear relationships between the logarithm of water solubility (log S; ppm) and
the soil TLC Rf for 7 aliphatic and 48 aromatic organic chemicals.
mobility of these organic chemicals beyond
the limits of statistical nonsignificance
and the number of aromatic chemicals con-
taining the fragment. One classic view of
the adsorption of polar but nonionic or-
ganic molecules by clay minerals attributes
a major portion of the surface interaction
to functional groups containing nitrogen
and oxygen [Harter (8); Mortland (17);
Theng (21)]. In this study, however, most
of the fragments listed in Table 2 that
apparently did not enhance or retard mo-
bility contained either nitrogen or oxygen
or both. Apparently the interaction is in-
significant for these fragments compared to
the molecule's hydrophobicity, for soil mo-
bility prediction purposes. So far as we
are aware, a significant interaction of
chlorinated alkyl and aryl chemicals with
colloid surfaces, except for the findings
of Briggs (1), has not been reported. The
findings of this study, in general, do not
provide evidence for the existence of the
interaction.
Table 3 lists those molecular frag-
ments that either enhanced or retarded the
soil mobility of aromatic chemicals. Re-
gression Model I, based on Equation 9, was
developed:
Rf = 0.208 log S + 0.003 + 3
[10]
where 3i is the molecular fragment mobili-
ty coefficient (R2= 0.93). Model II was
then derived from Model I:
log [Rf/(l - Rf)] = 0.469 log S [11]
- 1.131 + 32
where 32 is the molecular fragment nobility
coefficient. It was fitted to the logit
transformation of the observed chemical mo-
bility in order to constrain the predicted
response to values between zero and unity
(Rz=0.94). The quantities 3i and 32 are
assigned values of zero for any fragment
that is not listed in Table 3. If the
fragment is listed, 3j or 32 is assigned
65
-------
TABLE 2. MOLECULAR FRAGMENTS THAT "EITHER
ENHANCED NOR RETARDED SOIL MOBILITY
OF VARIOUS AROMATIC CHEMICALS
Fragment
Amide, substituted
Amine, secondary
" tertiary*
Azide
Carbamate
Cyanide
Diazosulfonate
Ether
Halogen, alky!
aryl t
I mine
Pyridazinone
Sul fone
Urea
Number of
chemicals
5
14
5
1
8
1
1
5
4
?4
O
i.
2
1
10
*Includes the tertiary am'ne within the
pyridazinone fragment.
"^Exceptions noted in Table 3.
the correspondinq value from Table 3. Roth
models only accomodate one fragment per
molecule.
The molecular fragments listed in
Table 3 can be divided into two groups.
The first group is comprised of those frag-
ments that possess an electropositive or
electronegative character that is not
created by, or is negligibly affected by,
aromatic ring resonance. Since Hagerstown
soil colloid surfaces possess a net nega-
tive charge, fragments that possess an
electronegative character should he re-
pelled from these colloid surfaces
[Greenland (4); Mortland (17)]. Mobility
enhancement of the molecule in this soil
should result. The 3 values of Table 3
showed the the carboxyl, trifluoromethyl
and trifluoromethylsulfonyl fragments en-
hanced soil chemical mobility, in agreement
with this hypothesis. At soil pH 6.8, the
dissociated carboxylic acid fragment, pos-
sessing a negative charge, was assumed to
be minimally affected by resonance.
However, this assumption will not be valid
TABLE 3. MOLECULAR FRAGMENTS ENHANCING (+) OR RETARDING (-) MOBILITY OF AROMATIC
CHEMICALS, AND THEIR MOBILITY COEFFICIENTS (3)*
Fragment
Chemical
Acid, carboxylic
benzoic
benzylic
phenoxyacetic
Amine, aliphatic, primary
Benzene
jD-bromo
m-chloro
m,£-dihalo
hydroxy (dissociated)
0.20 0.80 18
0.36 0.77 22
0.10** 0.18** 16, 28, 53
-0.10** -0.19 47
-0.22 -0.40 29
-0.19 -0.50 7, 14
-0.17 -0.36 12, 21, 27, 31', 44
0.13** 0.35 35
Triazine
chloro
methoxy
methyl thio
Trifl uoromethyl
Tri f 1 uoromethyl sul fonyl
0.19
-0.13
-0.12
0.14
0.45
0.47
-0.31
-0.20
0.32
0.96
5, 45, 49, 55
4, 41, 50
1, 6, 42, 51
25, 32, 33
36
*Values are significant at or below 5% level, except those (**) at 10% level.
tlsed in Model I (Equation 10).
*Used in Model II (Equation 11).
66
-------
at low soil pH's. Fragments that possess a
permanent or pH-dependent electropositive
character should be attracted to Hagerstown
soil colloid surfaces [Mortland (17); Theng
(21)] and nobility retardation of the mole-
cule should result. Table 3 also showed
that this was the case for the aliphatic
primary amine fragment, which possesses a
pH-dependent electropositive character.
The second group of molecular frag-
ments listed in Table 3 includes those in
which resonance is significant. It con-
tains three subgroups. The j>-triazine sub-
group chemicals share a common ring struc-
ture and possess similar amino fragments.
The general properties and reactions of the
j-triazines, and hence the behavior of
these compounds at the clay surface, are
related to the nature and location of the
ring substituent groups. Weber (22) con-
cluded that the adsorption process is
primarily controlled by the nature of the
non-amino fragment. In this study, mobili-
ty enhancement or retardation effects were
primarily due to non-amino fragments. The
chloro fragment significantly enhanced
mobility, probably through surface
repulsion due to its mild electron with-
drawing effect, while the methoxy and
methylthio fragments retarded mobility.
These findings agree, in qeneral, with
other studies dealing with the relative
ranking of the clay adsorption potential of
similar _s-triazines [Theng (21)].
The halogen-substituted benzenes com-
prise the second subgroup within the reso-
nance group of molecular fragments. The
subgroup is composed of carbanilates, phen-
ylureas and an anilide. Significant
retardation of mobility was observed with
chlorine substitution in the 3 (meta) posi-
tion. Bromine, chlorine or methyl substi-
tution in the 4 (para) position did not
alter the retardation effect of 3-chloro
fragment; however, additional ring substi-
tutions, in general, destroyed the
retardation effect. Significant retarda-
tion of mobility was observed for bromine
substitution in the para position with only
one chemical (No. 29). Briggs (1) found
similar significant changes in adsorption
due to chlorination of rj-alkyl phenylureas
at the meta position and reported that the
best linear relationship with the adsorp-
tion distribution coefficient was obtained
by using the Hammett sigma constant for
meta and para substituents in the phenyl
ring. However, two assumptions made in
this study were that the magnitude of a
fragment's mobility-enhancing or -retarding
effect was (a) constant and (b) independent
of the structure of the parent molecule.
An examination of Figure 1 reveals that, in
general, this was the case for the halogen
substituted benzene subgroup (No's. 7, 12,
14, 21, 27, 29, 31, 44, 52). Some variance
in Figure 1 can be atributed to the solu-
bility data which were obtained from the
literature and were not necessarily
measured in the same laboratory under simi-
lar experimental conditions. Variance can
also be attributed to the fact that at
Rf <0.1, the Rf - log S relationship
is, in general, curvilinear. Due to these
sources of variance, the study of the
linear free energy relationships of this
subgroup, as applied to Models I and II,
appeared unwarranted and was therefore not
attempted.
A unique chemical within the halogen
substituted benzenes is PCP or pentachloro-
phenol (Mo. 35). Increasing the chlorine
substitution around an aromatic ring
results in decreasing pK,. Since the
pKa of phenol is 9.89 and the pKa of
trichlorophenol is 6.00, the pKa of PCP
should be well below 6.00. Therefore, PCP
is expected to exist as an anion with some
degree of mobility enhancement caused by
the dissociated hydroxyl group. The data
in Table 3 show that this may have been the
case. The use of these hydroxybenzene 3
values, however, should be applicable only
to this special case.
The aromatic ring substituted primary
amines are the third subgroup within the
resonance group of molecular fragments
(Nos. 2, 3, 38). Chemical 3's position in
Figure 1 indicates that the primary amine
fragment retards mobility. An analysis of
the sediment adsorption data of Hassett et
al. (9) for the primary amine fragment of
various polynuclear aromatic hydrocarbons
also indicates that the fragment should
retard mobility and enhance adsorption. On
the other hand, the position of chemicals
No. 2 and 38 in Fiqure 1 indicates that
this fragment enhances mobility. These two
chemicals are chlorinated and contain a
carboxylic acid fragment whose enhancement
effect cannot be separated from the primary
amine effect. Therefore, the 3 value for
the primary amine fragment cannot be
estimated with confidence using these data
and these three chemicals were not included
in the development of Models I and II.
67
-------
Comparison of Models
Table 4 lists the difference, ARf(8J,
between the observed Rf and the Rf Pre-
dicted by Equation 8 (the model which uti-
lized log S as the only independent vari-
able) for 46 of the chemicals used in this
study. Table 4 also lists the difference,
ARf(H)J between the observed Rf and the
Rf Predicted by Model I (the model which
utilized both log S and as independent
variables). Relative to ARf(8), the
absolute values of ARfUOj were lower for
32 chemicals, equal for 7 chemicals, and
somewhat higher for 7 chemicals. There-
fore, Model I was the superior model for
predicting the soil mobility of these
organic chemicals.
Soil adsorption and mobility estima-
tion techniques may be used to assign a
chemical to a general mobility class in
order to approximate the distance a sur-
face-applied chemical nay leach. Another
indication of the accuracy of a predictive
technique is its ability to place a chemi-
cal into the observed mobility class.
Helling and Turner (13) proposed five
mobility classes: Class 1, Rf 0-0.09
(immobile); Class 2, Rf 0.10-6.34 (slightly
mobile); Class 3, Rf 0.35-0.64 (inter-
mediate mobility); Class 4, Rf 0.65-0.80
(mobile); and Class 5, Rf 0.90-1.00 (very
mobile). Table 4 shows that the use of
Equation 8 places 31 chemicals into the
wrong mobility class. On the other hand,
the use of Model I places only 7 chemi-
cals in the wrong class, with 6 of these
chemicals being on the borderline of the
proper class. Compared to the estimation
technique utilizing solubility as the only
independent variable, we concluded that the
estimation technique represented by Models
I and II, utilizing both log S and the 8
molecular fragment mobility coefficient
as independent variables, is a superior
technique for predicting the mobility of
diverse organic chemicals in Hagerstown
silty clay soil.
TABLE 4. DIFFERENCE BETWEEN OBSERVED Rf AND THAT PREDICTED BY
EQUATIONS 8 [ARf(8J] OR 10 [ARf(l£)], FOR 46 ORGANIC CHEMICALS
Chemical
1
4
5
6
7
10
12
13
14
16
18
19
20
21
22
23
24
25
26
27
28
29
30
ARf(8)
-0.04
-0.11*
0.14*
-0.11*
-0.18*
0.03
-0.23*
-0.05*
-0.24*
0.10*
0.20*
-0.06
-0.02
-0.11*
0.35*
-0.05*
-0.06
0.09
0.12*
-0.24*
0.17*
-0.22*
-0.02
ARf(lp_)
0.08
0.02
-0.04
0.02
0.03
0.03
-0.05
-0.03*
-0.04
0.01
0.00
-0.05
-0.01
0.07
0.00
-0.05*
-0.06
-0.04
0.12*
-0.05
0.07
0.00
-0.01
Chemical
31
32
33
34
35
36
37
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
55
ARf(8)
-0.10*
0.08*
0.21*
0.10*
0.11*
0.43*
-0.06
-0.01
-0.08*
-0.01
-0.12*
0.03
-0.26*
0.19*
0.06
-0.11
0.02
0.28*
-0.28*
-0.23*
-0.14*
0.03
0.07*
ARf(:iOJ
0.10
-0.04
0.08*
0.10*
0.00
0.00
-0.04
-0.01
-0.07*
0.13
0.02
0.03
-0.08
0.02
0.07
0.00
0.03
0.10
-0.15
-0.10
0.12*
0.06
-0.10
*Predicted mobility class differed from the observed mobility class [classes
were defined by Helling and Turner (13)].
68
-------
DISCLAIMER
The contents and views expressed in
this article do not necessarily reflect the
views and policies of the U.S. Department
of Agriculture or the Environmental Protec-
tion Agency, nor does the mention of trade
names and commercial products constitute
endorsement or recommendation for use.
REFERENCES
1. Briggs, G. G. 1969. Molecular struc-
ture of herbicides and their sorption
by soils. Nature 223:1288.
2. Briggs, G. G. 1973. A simple rela-
tionship between soil adsorption of
organic chemicals and their octanol-
water partition coefficient. Proc.
7th Brit. Insectic. Fungic. Conf.
1:83-86.
3. Dragun, J., R. Potenzone, Jr., C. S.
Fowler, and C. S. Helling. 1980.
Evaluation of molecular modelling tech-
niques to estimate soil-chemical
mobility: 1. Molecular connectivity
and charge related indices. Proc. Res.
Symp., 53rd Ann. Mtg., Hater Pollut.
Contr. Fed., Las Vegas, NV.
4. Greenland, D. J. 1965. Interaction
between clays and organic compounds in
soils. Part 1. Mechanisms of interac-
tion between clays and defined organic
compounds. Soils Fert. 28:415-425.
5. Guth, J. A., N. Burkhard, and D. 0.
Eberle. 1976. Experimental models for
studying the persistence of pesticides
in soils. In: Proc. Brit. Crop
Protect. Coun. Symp.: Persistence of
Insecticides and Herbicides (K. I.
Beynon, ed.), Monogr. No. 17, Brit.
Crop Protect. Coun., pp. 137-157.
6. Hammett, L. P. 1940. Physical Organic
Chemistry. McGraw-Hill, New York.
404 pp.
7. Hansch, C., and A. Leo. 1979. Substi-
tuent Constants for Correlation Analy-
sis in Chemistry and Biology. Wiley,
New York. 336 pp.
8. Harter, R. D. 1977. Reactions of
minerals with organic compounds in the
soil. In: Minerals in Soil Environ-
ments (J. B. Dixon and S. B. Weed,
eds.). Soil Sci. Soc. Am., Inc.,
Madison, WI. pp. 709-739
9. Hassett, J. J., J. C. Means, W. L.
Banwart, and S. G. Wood. 1980. Sorp-
tion Properties of Sediments and
Energy-Related Pollutants. EPA 600/3-
80-041, U.S. Environmental Protection
Agency, Athens, GA. 133 pp.
10. Helling, C. S. 1971. Pesticide
mobility in soils: I. Parameters of
soil-thin layer chromatography. Soil
Sci. Soc. Am. Proc. 35:732-737.
11. Helling, C. S., D. G. Dennison, and
D. D. Kaufman. 1974. Fungicide move-
ment in soils. Phytopathology
64:1091-1100
12. Helling, C. S., D. D. Kaufman, and
C. T. Dieter. 1971. Algae bioassay
detection of pesticide mobility in
soils. Weed Sci. 19:685-690.
13. Helling, C. S., and B. C. Turner.
1968. Pesticide mobility: Determina-
tion by soil-thin layer chromatography
Science 162:562-563.
14. Karickhoff, S. W., D. S. Brown, and
T. A. Scott. 1979. Sorption of
hydrophobic pollutants on natural
sediments. Water Res. 13:241-248.
15. Meyers, H. E. 1937. Physico-chemical
reactions between organic and inor-
ganic soil colloids as related to
aggregate formation. Soil Sci.
44:331-357.
16. Mill, T., and W. R. Mabey. 1980.
Laboratory Protocols for Evaluating
the Fate of Organic Chemicals in Air
and Water. Final Rep., Contract No.
68-03-2227, U.S. Environmental Protec-
tion Agency, Washington, DC. 329 pp.
17. Mortland, M. M. 1970. Clay-organic
complexes and interactions. Adv.
Agron. 75-117.
18. Mortland, M. M. 1975. Clay-organic
complexes and interaction: aromatic
molecules. Proc. IUPAC Symp. Environ.
Qual. Safety 3:226-229.
69
-------
19. Rekker, R. F. 1977. The Hydrophobia
Fragments! Constant, Vol. 1.
Elsevier, New York. 389 pp.
20. Tahoun, S. A., and M. M. Mortland.
1966. Complexes of montmorillonite
with primary, secondary and tertiary
amides: 1. Protonation of amides on
the surface of montmorillonite. Soil
Sci. 102:248-254.
21. Theng, B. K. G. 1974. The Chemistry
of Clay-Organic Reactions. Adam
Hilger, London. 343 pp.
22. Weber, J. B. 1966. Molecular struc-
ture and pH effects on the adsorption
of 13 £-triazine compounds on nont-
morillonite clay. Am. Mineral.
51:1657-1670.
70
-------
PREDICTION OF LEACHATE PLUME MIGRATION
SHWRD Annual Research Symposium
Wayne A. Pettyjohn
Oklahoma State University
Stillwater, Oklahoma 74078
Douglas C. Kent
Oklahoma State University
Stillwater, Oklahoma 74078
Thomas A. Prickett
Consultant
Champaign, Illinois 61820
Harry E. LeGrand
Consultant
Raleigh, North Carolina 27609
ABSTRACT
The purpose of this investigation is to develop a manual for predicting leachate plume
migration and mixing in ground water that is typified by concepts and techniques that will
prove useful to a permit writer with little or no expertise in hydrology. The general
approach is the development of a series of techniques, ranging from very simple to rather
complex that take into account the chemical and physical factors that control or influence
chemical transport. The broad range of techniques consists of three phases—a numerical
rating system, analytical-graphical solutions, and computer models.
Introduction
The purpose of this study is to devel-
op a manual that will serve as a guide to
U.S. Environmental Protection Agency permit
writers in evaluating permit applications
dealing with hazardous waste storage, treat-
ment, and disposal facilities, as described
in Section 3004 of the Resource
Conservation and Recovery Act. This
investigation, only one of several under
development, is designed generally for
predicting leachate mixing and movement in
ground water and specifically addresses
three major questions.
1. How much will leachate contaminant
concentrations be reduced by mixing
with ground water?
2. In what direction will the leachate
plume travel and how will its shape
change along the travel path?
3. Which ground-water models are appropri-
ate for predicting movement of solutes
in ground water at a site with specific
hydrogeologic characteristics, assuming
that the solutes do not interact with
aquifer materials?
It is assumed that the permit appli-
cants will use a variety of methods for
predicting effects of a disposal site on
ground water and that the permit writer
will not have a strong background in
hydrology. Consequently, the manual will
provide a variety of predictive techniques
that range from exceedingly simple to
highly complex, all of which are based on
sound hydrogeological principles, but which
are presented in a simplified manner. The
project is conceived as consisting of two
separate parts continuing through a time
frame of three years. During the first
year predictive techniques will be developed
and tested using available data. During the
second two years the models will be applied
to several existing sites that are located
throughout the nation and represent the
broadest range of hydrogeologic environments
that can be obtained within the limits of
time and funding.
71
-------
Fundamentals of Ground-Water Flow
From a hydrogeologic point of view,
the subsurface is divided into two distinct
zones—(1) the unsaturated zone, which
extends from land surface to some depth
where the water table is encountered, and
(2) the saturated zone, which extends down-
ward from the water table. The water table,
although it may fluctuate within several
feet each year, is the boundary between
these two major zones. Void spaces are
filled largely with gases in the unsatur-
ated zone, while the in the zone of
saturation they are filled with water.
Commonly, although not in every case,
waste liquids must migrate through the
unsaturated zone in order to contaminate
ground water. The unsaturated zone is
exceedingly important because it is here
that many chemical and biological reactions
occur that attenuate wastes. Unfortunately,
undue reliance has been placed on the
"living filter" concept of the unsaturated
zone, with the erroneous assumption that
biological degradation and sorption
phenomena will attenuate wastes to such a
degree that ground water contamination is
not likely to occur. The fallacy of this
argument is well substantiated by the
thousands of ground-water contamination
cases that have been reported. Nonetheless,
the unsaturated zone is a major control on
ground-water contamination because it exerts
a strong influence on the movement of
biodegradable and nonconservative
substances. Obviously the longer it takes
a waste to migrate through the unsaturated
zone, the more time and mechanisms there
are to attenuate it. In addition to
thickness, the major controlling factors are
the permeability, grain size, and chemical
composition of the earth materials.
Generally speaking, those materials that
have a high ion-exchange capacity, a low
permeability, and consist of fine-grained
materials, will have the greatest attenua-
tion affects. Therefore, although not a
part of leachate plume mixing and migration
in ground water, the thickness and
composition of the unsaturated zone must be
examined in any waste disposal scheme.
In its simplest form, the movement of
ground water can be described by means of
Darcy's law, that is, Q = KIA, where Q =
the flow rate, K = permeability, I =
hydraulic gradient, and A = the cross
sectional area through which the flow
occurs.
Permeability (K) refers to the inter-
connection of pore spaces and the ease with
which water can move from one point to
another. Permeability ranges within wide
extremes being greater for coarse materials
than for fine.
Permeability is divided into two types:
primary and secondary. In the first case
the permeability came into being when the
strata were deposited, but secondary
permeability originated after the rocks
became compacted or lithified. Examples of
secondary permeability include fractures
and solution openings.
The hydraulic gradient is an expression
of the slope of the water table and the
difference in head or water level is a
function of the amount of energy used as
the water moves from one point to another.
A water-level map is a graphical represen-
tation of the hydraulic gradient. Although
hydraulic gradient is the driving mechanism
for horizontal laminar flow, in certain
cases there is a vertical component as well.
The vertical component of flow may be
either in a downward or upward direction in
response to recharge or discharge.
Recharge is the addition of water
to a ground-water reservoir. It is derived
from the infiltration of precipitation,
interaquifer leakage, inflow from streams
or lakes, or inadvertently, by leakage from
lagoons, leaky sewer lines, etc. Discharge
occurs where ground water flows to springs,
streams, swamps and lakes, or is removed by
evapotranspiration or pumping wells, among
others. Pumping wells, by far, has the
greatest impact on the hydraulic gradient,
locally steepening it several orders of
magnitude.
Ground-water velocity is very important
in pollution problems, because among other
things, it has a major influence on
dispersion. The ground-water velocity
equation can be derived from Darcy's law
and the basic velocity equation of
hydraulics. It has the following form; v =
Kl/a, where a = effective porosity, in
percent, and the other terms are as
previously defined. Ground-water velocities
are generally very small, ranging from a
few feet per day to a few feet per year.
The openings in a rock are referred to
as its porosity, which is expressed as a
percent. Porosity determines the maximum
amount of water a rock can contain when it
72
-------
is saturated. Porosity can be divided into
the part that will drain under the influ-
ence of gravity, which is called specific
yield or effective porosity, and the part
that is retained as a thin film of water on
rock surfaces, which is called specific
retention.
Fundamentals of Ground-Water Pollution
In any ground-water pollution study it
is essential to obtain the background
concentration of a wide variety of chemical
constituents, particularly those that might
be common both to ground water and a
leachate. Since shallow ground-water
quality may fluctuate within fairly wide
limits during short intervals, it is
essential to determine background
concentrations by means of several samples
taken at different times.
The severity of ground-water pollution
is related to the characteristics of the
waste or leachate. That is, its volume,
composition, concentration of the various
constituents, time rate of release of the
contaminants, the size of the area from
which the contaminants are derived, and the
density of the leachate, among others.
Data describing these parameters are
difficult to obtain.
Once the leachate is formed it begins
to slowly migrate through the unsaturated
zone where several physical, chemical, and
biological forces act upon it. Eventually,
however, the leachate may reach saturated
strata where it will flow in the direction
of the hydraulic gradient. From this point
on the leachate will become diluted due to
a number of phenomena including filtration,
sorption, chemical processes, microbial
degradation and dispersion. Dilution will
continue but the long-term effects are
related to time and distance of travel.
Filtration removes suspended particles
from the water mass, including particles of
iron and manganese or other precipitates
that may have been formed by chemical
reaction. Dilution brought about by
sorption of chemical compounds, is related
largely to clays, metal oxides and
hydroxides, and organic matter, all of
which function as sorptive material. The
amount of sorption depends on the type of
pollutant and the physical and chemical
properties of the solution and the sub-
surface material.
Chemical processes are important when
precipitation occurs as a result of excess
quantities of ions in solution. Chemical
processes also include volatilization, as
well as radioactive decey. In many
situations, particularly in the case of
synthetic organic compounds, microbiological
degradation effects are not well known. It
does appear, however, that a great deal of
degradation can occur if the system is not
overloaded and appropriate nutrients are
available.
Dispersion is a microscopic phenomenon
caused by a combination of molecular
diffusion, which is important only at very
low velocities, and hydrodynamic mixing,
which occurs with laminar flow through
porous media. In porous media, flow paths
have various lengths and leachate transport
along a shorter flow path would arrive at
an end point sooner than that part following
a longer path, thus resulting in hydro-
dynamic dispersion. Dispersion can be both
longitudinal and lateral and the net result
is a conic form downstream from a
continuous pollution source. The
concentration of a leachate is less at the
margins of the cone and increases toward
the source. Since dispersion is directly
related to ground-water velocity, a plume
or slug will tend to become larger with more
rapid flow.
The rate of advance of a contaminant
plume can be retarded if there is a
reaction between its components and ground-
water constituents or if adsorption occurs.
The plume, which tends to have a normal
(Gaussian) distribution both longitudinally
and transversally, will expand more slowly
and the concentration will be lower than
those of an equivalent non-reactive
leachate. Hydrodynamic dispersion affects
all solutes equally, while sorption and
chemical reactions can affect various
constituents at different rates. Thus, a
leachate plume consisting of a number of
solutes can have each moving at a different
rate due to these attenuation processes.
Monitoring Concerns
Not uncommonly, monitoring wells are
placed in a convenient location that may
not be the most appropriate from a hydro-
geologic perspective. Moreover, little
thought may be given to their design and
installation. A monitoring system should
be designed and constructed to achieve two
73
-------
purposes: (1) determination of hydraulic
gradients and (2) evaluation of quality
distribution, both in time and space.
Since the hydraulic head and ground-water
quality may vary vertically, horizontally,
and temporaly, the aquifer should be
sampled at a considerable number of times,
places and depths.
Traditionally monitoring schemes
consist of wells of different depths that
are distributed throughout the area of
concern but which are generally few in
number due to cost considerations. One
method used to sample the aquifer is to
install a single well and slot or screen it
at several different depths. In the first
case, there may be an insufficient number
of wells to adequately describe the
vertical and horizontal head distribution,
which may be significant in regions of low
gradients. In the second case, multiple
screened wells provide both composite water
levels and water samples, which are useful
for determining if a plume exists, but are
of limited value for locating the actual
position of a plume at depth or its
chemical distribution. In recent years it
has become a common practice to install a
nest of monitoring wells that are closely
spaced but of different depths to sample a
single interval. The wells may be
installed in a single large diameter hole
and each gravel packed and grouted at a
specific depth.
A number of investigators are beginning
to recognize the importance of accurate head
and quality measurements and, as a result,
a number of innovative monitoring
techniques are being developed. A contri-
bution to the wide array of systems is the
monitoring well described herein (Fig. 1).
The actual configuration of the
monitoring well, particularly slot size and
intervals, should be designed to best suit
the conditions that have been determined at
a pre-drilled site. The well consists of
two separate components—an outer slotted
casing and a moveable packer arrangement
that fits tightly within the casing. The
materials used to construct the system are
dictated by the type of samples to be
collected. For example, if the ground
water is contaminated by organic compounds,
the slotted part of the outer casing should
consist of Teflon and that part of the
packer assembly that will come in contact
with the water sample should be constructed
of stainless steel and Teflon. On the
other hand, if the contaminants consist
largely of inorganic compounds the entire
assembly can be constructed from a variety
of plastics or metals.
A prototype sampling system consists
of 5 cm diameter plastic pipe into the end
of which is cemented a cone-shaped plastic
plug. Vertical slots (3 mm x 7.6 cm) were
milled into the casing. The packer
assembly consists of threaded 1 m sections
of 12 mm diameter conduit. A grooved
circular plate, just slightly smaller than
the inside diameter of the casing, is
attached to the bottom section. The groove
is fitted with a rubber 0-ring. Approxi-
mately 2.5 cm above the plate is a 6 mm
diameter hole in the conduit. A second
0-ring fitted plate slips over the conduit;
it can be positioned any place on the shaft
by means of a set screw. The upper plate
is perforated by a 6 mm diameter, 3.8 cm
long, brass fitting onto which is attached
flexible tubing that is suffficiently long
to reach from the bottom of the well to the
land surface.
Once the casing is placed in a pre-
drilled hole, the upper packer is set at
the appropriate interval, for example, .3 m
above the base plate, to form a collection
chamber. The packer assembly is then
slowly forced to the bottom of the well.
Water is withdrawn from the well by one of
two means. If the water level lies less
than about 6 m below land surface, a
peristaltic pump may be attached to flexible
tubing that is connected to the packer
assemblage, which extends above the casing.
The water can then be lifted by vacuum
directly from the well. If the water level
lies at a greater depth, pumping by either
air or nitrogen gas is appropriate. The
flexible tubing, stretching from the upper
packer, is attached to the appropriate
pressure source, and gas is then slowly
allowed to flow into the sampling chamber
forcing water through the hole just above
the lower packer and then discharging
through the center of the pipe. The
positive pressure pumping method will allow
the collection of a water sample from a
considerable depth.
After the first sample is collected,
the packer assemblage is pulled up to the
next sampling interval and the process is
repeated until the entire saturated thick-
ness is sampled. It is important to remove
a sufficient volume of water to empty the
sampling tube and chamber in order to
74
-------
Casing -
Plug-
gas source
-Flexible tubing
tipper packer
0-ring
-Conduit samoling tube
Access opening
0-ring
Casing
~Lower packer
Packer Assembly
Figure 1. Schematic of monitoring well
collect uncontaminated samples from
succeeding intervals. Since the plunger
pipe is threaded at regular intervals,
sections can be removed as appropriate.
Water levels can be measured inside the
conduit by means of an electric tape at
each sampling interval prior to pumping.
When completed the casing should remain in
place for subsequent sampling, but the
packer assemblage can be used in other
wells after it has been thoroughly cleaned.
The monitoring well should be
installed in a pre-drilled hole. If the
hole is approximately the same diameter as
the casing it should be possible to force
the casing directly into the hole, which
should produce a sufficient seal between
the casing and annulus to prohibit vertical
leakage along the casing. In some cases
fine material may plug the slots and it
might be necessary to cement screeens over
them, although it might be possible to
clear them by injecting gas into the sample
chamber, or pumping the well prior to
insertion of the packer assembly. Another
method is to install this well in a larger
diameter hole. In this case the well would
be set on the bottom of the hole, gravel
poured into the annulus to a position above
the top of the bottom sampling interval,
and then bentonite or clay would be placed
in the annulus, reaching nearly to the next
set of sampling slots. The second set of
slots are then gravel packed and isolated
by another clay or bentonite seal. This is
probably the most appropriate procedure and
should allow adequate sampling without the
need for well development.
Development of Predictive Methods
Introduction
Although seemingly a simple task from
a theoretical point of view, the develop-
ment of ground-water transport models is
fraught with uncertainty. Existing models
are highly simplified and are not easily
adaptable to the complexities of the
subsurface environment, changing land use,
and density differences between a leachate
and ground water, to mention only a few.
For example, nearly all ground-water
systems include both aquifers and aquitards
(confining beds), which can vary widely in
thickness and permeability. Differences in
permeability between aquifers and aquitards
or even between zones of different
permeability within a single aquifer, cause
refraction of flow lines at their
boundaries. As flow lines move from
aquifers into and through aquitards, for
example, they are refracted toward a
direction that is perpendicular to the
boundary that produces the shortest path.
Consequently, in order to predict the path
that a leachate plume will follow, it is
essential to determine the location of all
units of different permeability since they
will cause a change in the flow direction,
both vertically and horizontally. This is
possible only in the broadest sense.
Models that adequately describe the
movement of plumes that are more dense or
less dense than water are simply not
available. In most cases it is assumed
that the leachate density is the same as
water density and that there is an even
distribution of the waste constituents
throughout the thickness of the aquifer.
Unfortunately this is not the case where
the leachate is less dense due to the
presence of light hydrocarbons, such as
gasoline, or denser because of highly
mineralized solutions.
Another externality that complicates
predictive models is the effect brought
about by changes in land use. For example,
a predicted flow path and velocity may be
completely negated if a new well is
subsequently turned on, or if an old well is
abandoned. In either case there will be a
major change in the hydraulic gradient in
the vicinity of the well and, therefore, a
change in flow path and velocity.
A particularly vexing problem in
predicting leachate mixing and migration is
75
-------
the time rate of release of contaminants to
ground-water reservoirs. Several
investigators have described or illustrated
the fact that the chemical quality of water
in many shallow and, particularly, surficial
aquifers exhibits cyclic fluctuations.
These fluctuations are caused by the inter-
mittent flushing of contaminants into the
ground during recharge (precipitation)
events. The contaminants may be naturally
occurring or reflect man's activities, such
as waste disposal. The greatest amount of
contaminant influx occurs in the spring,
but less concentrated and smaller masses
may be flushed to the water table at any
time. Because of the cyclic nature of
recontamination events, care and common
sense must be exercised in the extrapolation
of quality data, particularly in regard to
the prediction of flushing rates and
leachate plume mixing and migration.
Several coefficients are required in
chemical transport equations and for many
of these the data base is inadequate.
Examples include reliable field data for
dispersion, sorption, and microbiological
degradation.
Methods of Development
The philosophy of this investigation
is to develop several techniques that can
be used to approach the problem of leachate
plume mixing that range from broad
generalized predictions to much more
complicated procedures that require a
substantial data base and some expertise in
hydrogeology. This approach is based on
the premise that some wastes are so
innocuous that leachate is not likely to
form or be of any significance if it does,
and therefore, only a general, pessimistic
approach is required. Therefore, only
conservative pollutants will be considered.
An example is the disposal of demolition
wastes, such as fragments of concrete.
Some sites are so obviously unsuitable for
disposal that only a cursory examination is
necessary to bring its inadequacies to
focus.
Our predictive scheme is divided into
three phases, starting with the most
generalized and least quantitative method.
The phases are (1) a rating system,
(2) analytical-graphical solutions and
(3) computer models.
Phase I Rating System
A system for evaluating the contamina-
tion potential of waste disposal sites was
originally formulated by LeGrand (1964) and
has undergone a series of subsequent
refinements leading up to a widely used
numerical system (LeGrand, 1980). The
system focuses on weightening four key
geologic and hydrologic characterisitcs in
the vicinity of contamination sources. The
key parameters are: (1) distance to a water
supply, (2) depth to the water table,
(3) hydraulic gradient and (4) permeability-
sorption, as indicated by the geologic
setting. Four digits and four additional
letter codes concisely describe the
characteristics of a site for standardized
interpretation and indicate the relative
contamination potential.
Hydrogeologic characteristics at a site
are identified on a scale of various aquifer
sensitivities to contamination spread, and
the particular waste's toxicity, concentra-
tion, volume and persistence are identified
on a scale of severity of contaminants.
These identified positions are integrated
into a matrix in which the hazard potential
is graphically displayed. The identified
position in the matrix reveals the degree
of seriousness. The probability of
contamination can be determined by
correlating a sites numerical description
with a synthetic numerical standard in the
matrix. Acceptance or rejection standards
are suggested that are based first on the
natural setting and then on modifications
to the site by specific engineering
practices.
The system optimizes existing data and
provides a useful first round evaluation of
all types of leaks, spills and waste
disposal sites in relation to nearby ground-
water and surface water supplies. An
experienced evaluator using the system can
assess a potential pollution situation
quickly. The system represents a quantifi-
cation of parameters, appraised graphically
in a logical sequence, and the results are
indicated in a standardized form that is
meaningful to evaluators anywhere. A
sample worksheet for the rating system is
shown in Table I.
LeGrand's evaluation scheme is useful
for the "first—cut" type of waste disposal
sites. However, his scheme lacks certain
aspects, particularly as it relates to
amount and type of expected leachate
76
-------
Step 1
Determine the distance
on ground between con-
tamination source and
water suppl\
TABLE I. RATING SCHEME WORKSHEET
Poi_nt_yaiug___ 0
Distance in 2000+ 1000-2000 ^00-999 150-299 ~"S-H9 SO-^4 4549 20-34 15-19 0-11
meters
Distance in 6200-*- M 00-6200 1001 MOO 501 1000 251-500 161 250 101 -160 61-100 46-60 0--45
feet
Step 2
r-stimate the depth
to the water table
Below- base of con-
tamination source
more than 5"<> of the
year
Meters 60+ W-60 20-29 12-19 8-11 5 " *> 4 15-25 5-1 0
H-et 200+ 91-200 61-90 Wi-60 26 45 16 25 9-15 4-H 0 2 0
Step 3
tStimatc water
table gradient
from contamin-
ation site
Water table
gradient and
(low direction
Gradient Gradient
awavfrom almost
all water flat
supplies that
are c loser
than 1000
mete rs
Gradient
less than
2 percent
toward
water
supply hut
not the
anticipated
direction of
flow
Gradient less
than 2
percent
toward
w-ater suppK
and is the
anticipated
direction of
flow
Gradient
greater than
2 percent
toward
water supph
hut not the
anticipated
direction of
now
Gradient
greater than
2 percent
toward
water supply
and is the
anticipated
direction of
now
Step 4
* 31
tu -i t-
< la\ with
no more Sand with Sand with (lean
than 50% 15-40% less than ( lean fine gra\el or
sand ela\ 15",. i lay sand coarse sand
More than W
25 29
20-24
15 19
II) 1 i
\- 9
Less than 3
OA<->
I"1 11'"
OB 1<
0( 2(
(ID (, 6F
SH ->n
SI ^E
SJ 8D
SK 9<
KA
1 11
"1- 8h
'(, HF
"H Hd
-1 91)
^1 9E
^K 9F
9A
1 II
9C, 9M
911 9N
91 9O
9J 9P
9K 9Q
9L 9R
M»rt- than 95
^5 9^
6() ^ t
4659
2KH5
10 2^
Less than 10
Bedrock at land surface, I = 57. II = 97,
HA/ARI) K)T[-N MAI MATRIX
DEGREE OF SFRlOIISNtSS
Modrralcly HIJ^I
DEGREE OF SERIOl SNESS
Moderated Low
1)F(,RFF OF StRJOl'SNEVS
[>t<,R£fc OF SFRJOI SNF.SS
High
DFdRI-F OFSERJO
DE(.RF_F OF SFRJOI SI
DM,REE OF SFRJOI SNESS
Fxirtmek High
Dt(,RFF OF SFRIO1
High
[)fc(,RFF OFS1-HKK SNFVS
Mtxlcraith High
Record Description, Par \ .due and Rating
as noted in example
77
-------
relative to the flow and quality of the
naturally occurring ground waters (mixing
and resultant plume characteristics). As a
first step in developing a more comprehen-
sive set of rules on site selection, we
believed that it would be helpful to make
additions to LeGrand's scheme to account for
the leachate/native ground-water mixing
effects. This would be done in such a way
to be consistent with the qualitative
nature of his scheme and at the same time
expand its scope to include additional
important pollution considerations. The end
result of this first exercise would still be
a first cut type of evaluation but with an
important expanded scope. A total quantita-
tive evaluation would follow (as described
in the next sections of this report) to
fully evaluate pollution potential, if need
be. A further explanation of this approach
is given below and along with an outline as
to how it might be incorporated into
LeGrand's scheme.
The potential for causing ground-water
pollution in the vicinity of a landfill
depends not only on the hydrogeologic
setting and the characteristics of any
landfill contaminant, but also on the
quantities of ground water flowing in the
receiving aquifer compared with the leachate
rate from the landfill. A small amount of
contaminant entering a copious aquifer is
a much different situation than a large
amount of contaminant entering an aquifer
of small flow. LeGrand applies this concept
in his rating but does not quantify it.
Many landfills and waste disposal
scenarios may contribute to leachate having
a liquid density quite different than that
of the native ground waters. We believe it
is important to distinguish between possible
leachates being either heavier or lighter
than the native ground waters. Pollution
potential evaluation, at least in a general
sense, should include consideration of the
possibility of the contaminant being
vertically segregated from the main stream
of the flow in the aquifer. Although we
are recommending a leachate/aquifer-water
density consideration, it must be understood
that this is merely a contaminant
characteristic not particularly related,
pollution potential wise, to something good
or bad. The knowledge of density differ-
ences earmarks different analysis routes in
predicting plume movement, in both the
saturated and unsaturated zones.
Finally, a mixing characteristic
related to the leachate being either
miscible or immiscible in water is needed.
Petroleum based liquid wastes are the most
commonly occurring immiscible products
found in some landfills. The plume
characteristics, and pollution potential
from immiscible versus miscible liquids,
are different for these two distinctive
leachate types. If there are immiscible
products to be disposed of in a landfill,
one should be aware of them for the plume
movement analyses.
In summary, we feel that LeGrand's
rating scheme needs an additional three
descriptors involving (1) the process of
dilution of the flowing leachate mixing
with the volume rate of flowing native
ground waters, (2) the definition of
whether the leachate will likely be heavier,
the same as or lighter than the native
ground waters of the site, and (3) whether
any of the leachate liquids are immiscible
in water. In each of these cases, the
evaluation of pollution potential or the
characteristics of the pollution plume
movement is different. The procedure for
evaluating and characterizing the mixing
and the additional plume characteristics
follows.
A study is now underway to rank the
dilution ratios of the leachate by native
ground waters in such a way that will
produce numbers between 10 and 1. This
would fit LeGrand's general 10 to 1 point
system for each of his key hydrogeologic
parameters. The dilution ratio would be
calculated on the basis of dividing the
volume rate of flow of a pollutant by the
volume rate of flow of the effective under-
lying ground-water reservoir.
Another study is underway that is aimed
at ranking the seriousness of various
concentration levels of pollutant species.
This study will result in another 10 to 1
point value number, depending upon whether
the drinking water standards are exceeded
by factors of, for example, 100, 10, 2, 1,
.1, etc., times. (Possibly the point values
would involve negative numbers in the case
of the native ground waters being of poor
quality.)
A final dilution-ranked point value
would be calculated on the basis of
multiplying the above two numbers. The
resultant value would then be added to
LeGrand's point system. As a matter of
record, a note would be made of whether the
78
-------
leachate is miscible or not and whether it
is heavier, about the same, or lighter than
the native ground water.
The above procedure is not without some
misgivings. First, a sample and quality
analysis is needed of the native ground
waters. Secondly, some estimate of leachate
flow rate must be made (this might be the
ground-water recharge rate if no other
substantiating data are available). And
finally, LeGrand's Hazard Potential Matrices
may have to be expanded or readjusted to
account for high dilution ratios.
The addition of the dilution and plume
density and miscibility characterizations
would improve the first-cut evaluation
significantly.
Phase II Analytical-Graphical Techniques
One of the basic drawbacks to the
LeGrand system is that it is not very
quantitative. Many of the point system
parameters are given equal weight and others
are not. The rationale behind this point
system is outlined but the physics of the
processes (in terms of laboratory experi-
ments, field determined permeabilities, and
widely accepted mass transport equations,
for example) are neither documented nor
explained. Furthermore, his definitions of
such terms as "degree of seriousness, hazard
potential, probability of contamination, and
degree of acceptance," are opinionated.
This is not a criticism of LeGrand's method
for what it is as a first-cut evaluation.
However, there may be a need for a second-
cut evaluation in light of unclear or
questionable results of the first cut.
First, we will study and define a set
of widely accepted equations that govern
both flow and mass transport of ground water
and pollutants from a landfill source.
These equations would involve flow and mass
transport of pollutants starting from the
bottom of the landfill, going downward
through any unsaturated zone, and then
flowing and mixing horizontally through the
unsaturated zone. All of the processes of
dilution, dispersion, diffusion, ion
exchange, passing of time, radioactive and
biological decay would be considered. All
of the references contained in this report
include such equations of flow and mass
transport. There are many other equations
also to be considered.
Second, once that the equations are
agreed upon, there then exists a mechanism
for quantifying the evaluation of a land-
fill or waste disposal site. Each
parameter of the equations can be varied in
sequence to illustrate their relative
effects on the resulting concentration
plume. The weighting system and importance
of each parameter can thus be evaluated
directly without opinion (other than the
assumption upon which the equations were
derived). A normalized point system ranging
from 1 to 100 would then be devised to
display the range of parameter combinations
giving the best to the worst case plume
concentration distributions.
Third, based upon direct measurement of
parameter values or upon lacking data,
defining worst case values, solutions of the
equations are obtained for the specific site
under evaluation. The site specific land-
fill would then be compared to the above
devised point system and a number assigned
(or range of numbers depending on the
confidence of the input data).
The purpose of developing a more
quantitative evaluation method is for
clarification purposes. The aim of the
quantification process would be to weight
the major hydrogeologic parameters in the
landfill evaluation process by analyzing
known physical and chemical processes that
are involved. If we can do this, opinion
is minimized until the last step of
accepting or rejecting a site. The approach
of the quantifying procedure will be as
follows.
Finally, a judgement (preferably by an
EPA official) would be made as to how
serious a pollution potential is at the
particular site under evaluation based on a
single number ranking.
An example of a typical equation that
is under consideration for mass transport or
plume concentration distribution from a
continuous source in a uniform flow is as
follows for the saturated zone of flow (see
Wilson and Miller, 1978):
79
-------
- X
f exp -
3
4irm(D D )
x y
W(u,r/B)
where
- 2Dx/V
= 1+(2BX/V')
u = r /4yD't
) dG
(D
(2)
(3)
(4)
(5)
(6)
and
C = solute concentration
f = mass rate of pollutant
,Dy = longitudinal and transverse
dispersion coefficients
= a V, a V, respectively
x y
n = porosity
R = retardation coefficient due to
linear, equilibrium absorption
t = time
V = interstial velocity
V = V/R,
d
x,y = coordinates in space with origin
at pollutant source and uniform
ground-water velocity in x
direction
\ = radioactive or biological decay
constant
a ,a
x y
= longitudinal and transverse
dispersivities, respectively
The application of Equations 1 through
6 allows mapping the entire plume concentra-
tion distribution eminating from a pollution
source as a function of the x and y space
coordinates and with respect to time.
Despite the apparent complexity of the
equation, the solution can be obtained
easily with the aid of a pocket calculator.
The evaluation of W(u,r/B) is obtained by
an algorithm given by Wilson and Miller
(1978) in combination with a polynomial
approximation of an error function
subroutine. An example Radio Shack TRS-80
Basic language program code is given as an
example in the Appendix of this report.
Equation 1 allows quantification of the
sensitivity of varying dilution dispersive
mixing, and degradation parameters. As a
simple example, you can observe that if the
mass rate of pollutant (f) entering the
aquifer is doubled, then the concentration
at any point in the downstream plume is also
doubled. However, if the aquifer porosity
is doubled, the concentration is halved.
The other parameter change impacts are less
obvious. Consider what happens when the
velocity is doubled. Is this necessarily
a good situation since there is more water
flowing past the landfill with double the
velocity? What weight should you give to
absorption? The equation is there to
answer those questions.
There are many equations available for
evaluation of the unsaturated zone also
(see Enfield et al., 1980; Warrick, et al.,
1971; and Wilson and Gelhar, 1980). It is
likely that no more than four equations
would be necessary for both the saturated
and unsaturated system to quantify the
entire flow combinations.
The problem remains as to handling site
specific data deficiencies. For example, it
is unlikely that dispersivities, hydraulic
conductivity, retardation coefficients,
porosity, etc., would be available from
first- or second-cut type of study.
Our quantification scheme would thus
involve a study of the literature and a
cataloging of case histories of prior
pollution studies wherein these unknown
parameters were determined and correlated
with hydrogeologic evidence of the type
that is normally known. Some geologic logs
and soil samples are likely to exist and
regional hydrogeology is generally known.
Estimates could be made of the unknown
parameters via the correlation study
results. As an example, a description of
a geologic log of a test well may be
sufficient to estimate hydraulic conduc-
tivity (permeability to water). In the
event that a satisfactory correlation study
cannot reasonably substitute for missing
data, then either the permit applicant would
be required to initiate a field investiga-
tion or a worst case parameter would be
chosen for analysis. It is likely that the
retardation coefficient will be one of the
unknown parameters, in which case the worst
situation might be to route the pollutant
through the ground-water reservoir without
retardation and make a judgement accordingly
without observation. An attempt at
correlating dispersion, diffusion, porosity,
80
-------
and permeability coefficients with generally
known hydrogeologic description is now under
study.
Man-made modifications such as bottom
liners and compacted top covers can be
evaluated by the same set of equations
mentioned above.
If possible, the entire quantifying
approach will be simplified as much as
possible through the use of tables,
nomograms, or very simple calculator
routines. Presently, this simplification
process is uppermost in our work and we will
be diligent in keeping it this way as far
as possible.
Phase III Computer Models
There will undoubtedly be special cases
where neither a first- nor second-cut
evaluation of pollution potential from a
landfill will yield a satisfactory result
for either the permit writer or the landfill
owner. Unusual circumstances are likely to
occur when the hydrogeologic setting is
complicated (stratified, nonhomogeneous,
anisotropy, multilayered, or when irregular
landfill and external boundary conditions
exist, etc.) or when the geochemistry of
the pollutant transport is in question
(density differences, multiple species
involved with numerous possible retardation
coefficients, variable leaching rates
through unsaturated absorbing type rocks,
viscosity and temperature effects, etc.).
For these cases, a higher level of evalua-
tion may be in order.
The quantified system for evaluating
potential pollution hazards presented in
the last section of this report is based
upon analytical formulas and basic physical
and chemical process consideration. This
quantified approach is at a level of
sophistication far above an all qualitative
approach but lacks a comprehensive
capability to account for the above
mentioned nonhomogeneities, both in flow
and water quality aspects.
There are numerical models available
that can account for nonhomogeneities and
they may fit into the landfill evaluation
process when conditions warrant. Based
upon experience, we recommend that numerical
models be used with caution since most
models are capable of predicting far beyond
available data, thus creating a false sense
of security. Our approach is to use the
simplest model available to solve the
evaluation problem.
Several levels of numerical models will
be described in the final report. Presently,
we are assembling a selection of useable
models that can be applied directly or
indirectly to the landfill/waste disposal
site selection and evaluation process. A
review of the EPA report by Bachmat et al.,
(1978), (which presents a worldwide survey
of numerical ground-water models) reveals
the fact that less than 16 percent of the
flow models and less than 8 percent of the
mass transport models reported are useable.
Useable was defined as, at a minimum, the
existence of an available computer code and
a document explaining its application.
Based upon our experience, documentation of
ground-water computer models is not
ordinarily done. There have been numerous
models developed and described since 1978
and we are searching the recent literature
and contacting several researchers and
private firms for more useable codes.
What we want to assemble is a group of
numerical ground-water models that can be
applied to various aspects of the landfill
evaluation process. There is no one
numerical model that can solve all of the
problems. Particular problems that would
be addressed beyond the nonhomogeneities
mentioned above are (1) models that could
be generated in the vertical as well as the
plan view, (2) models that would be
applicable to the various geologic environ-
ments, such as fractured rocks, karst
regions, glacial till areas, alluvial
valleys, multiple layered scenarios, etc.,
(3) models that are simple to operate and
can be upgraded as needed to provide detail
only to a level appropriate with the problem
to be solved—as opposed to super models
including many complications beyond which
the problem may call for, and (4) models
that can handle not only the geochemistry
of leachate transport but the related
physical density, miscibility, and viscosity
problems that are likely to be involved in
operating a complex landfill.
The final report on computer models
will also contain descriptions of needed
computer facilities, the advantages and
disadvantages of the selected codes, where
help can be obtained in operating and
modifying the code for specific applications,
and how much time can be expected to be
consumed in solving typical problems.
81
-------
The same point system of the previously
described section will be applied to the
plume distributions determined from these
higher level models. The plume character-
izations generated by the higher level
computer models are expected to fit within
that general evaluation scheme. However,
one should expect that the higher level
models will be needed for pinpointing and
clarifying very specific problems rather
than replacing the first- and second-cut
type of evaluations.
References
1. Bachmat, Y., B. Andrews, D. Holtz, and
S. Sebastian. 1978. Utilization of
Numerical Groundwater Models for Water
Resource Management. U.S.E.P.A. Publi-
cation EPA-600/8-78-012, Robert S. Kerr
Environmental Research Laboratory, Ada,
Oklahoma, 177 pp.
2. Bresler, E. 1973. Simultaneous transport
of solutes and water under transient
unsaturated flow conditions. Water
Resources Research, 9 (4). pp. 975-986.
3. Enfield, C. G., R. F. Carsel, S. Z.
Cohen, T. Phan, and D. M. Walters. 1980.
Methods of approximating transport of
organic pollutants to groundwater.
Robert S. Kerr Environmental Research
Laboratory, Ada, Oklahoma, and Hazard
Evaluation Division, Office of Pesti-
cide Programs, Washington, D.C. (in
review process.)
4. Hunt, B. 1978. Dispersive sources in
uniform ground-water flow. Journal of
the Hydraulics Division, Am. Soc. of
Civil Eng., Paper Number 13467, HY1.
pp. 75-85.
5. LeGrand, H. E. 1964 System for evalua-
ting the contamination potential of some
waste sites. Am. Water Works Assn., 56
(8). pp. 959-974.
LeGrand, H. E. 1980. A Standardized
System for Evaluating Waste-Disposal
Sites. National Water Well Association,
Worthington, Ohio. 42 pp.
Warrick, A.W., J. W. Biggar, and D. R.
Nielsen. 1971. Simultaneous solute and
water transfer for an unsaturated soil.
Water Resources Research, 7(5).
pp. 1216-1225.
Wilson, J. L., and L. W. Gelhar. 1980.
Analysis of longitudinal dispersion in
unsaturated flow - Part 1. The analy-
tical method. Water Resources Research.
(in review process.)
Wilson, J. L., and P. J. Miller. 1978.
Two-dimensional plume in uniform ground-
water flow. Jour, of the Hydraulics Div.
Am. Soc. of Civil Eng., Paper Number
13665, HY4, pp. 503-514.
82
-------
APPENDIX Computer Code For Mapping The Concentration Distribution
Of A Plume Of Pollution From A Point Source
Of Variable Strength
10:
15:
20:
25:
30:
40:
50:
55:
57:
60:
65:
70:
80:
85:
90:
95:
97:
99:
101:
103:
105:
ji._ii
C =
BEEP
A =
IF H
A =
BEEP
BEEP
Z =
BEEP
B =
: BEEP 3: PRINT "CONTINUOUS POLLUTION"
0
1 : INPUT "UNITS? 1 FOR C--0 FOR G"; H
1
>0 THEN 50
7.48
1: INPUT "ENTER X VELOCITY = "; V
1: INPUT "ENTER RETARD. COEF. =" ; S
V/S
1: INPUT "ENTER X DISPERSIVITY ="; D
2*D
BEEP 1: INPUT "ENTER Y DISPERSIVITY ="• E
BEEP
K =
BEEP
L =
G =
BEEP
FOR
BEEP
BEEP
1: INPUT "ENTER POROSITY ="; P
4*Ti*P*V*^(D*E)*A
1: INPUT "ENTER DECAY CONST. ="; L
LN 2/L/365
1 + (2*B*L/Z)
1: INPUT "ENTER NUMBER OF RATES ="; M
W = 1 TO M
1: PRINT "PREPARE FOR RATE ="; W
1: INPUT "MASS RATE ="; F
107: A(27+W) = F
109: BEEP 1: INPUT "TIME="; F
111: A(37+W) = F
113: NEXT W
115: BEEP 1: PRINT "RATES DONE--GO AHEAD"
120: BEEP 1: INPUT "OB POINT X ="; X
125: 0 = EXP (X/B)
130: BEEP 1: INPUT "OB POINT Y ="; Y
140: R = /"((X*X + (D/E)*Y*Y)*G)
145: FOR II = 1 TO M
150: U = R~ 2/(4*G*D*Z*A(37+W))
160: Q = R/B
165: IF (U<1)*(Q<1) PRINT "W(U,R/B) IS INACCURATE"
170: GOSUB 220
180: C = C + ((A(27+W)-A(26+W))*0/K*N
183: BEEP 1: PAUSE "CONCENT. ="; C
185: NEXT W
190: BEEP 2: PRINT "CONCENT. ="; C
200: BEEP 1: PRINT "PRESS ENTER FOR NEW OBPT"
205: C=0
210: GO TO 120
220: J = (Q-2*U)/(2*/~U)
222: I = 0
224: IF (J>0) + (J=l) THEN 230
226: I = 1
228: J = -J
229: IF J >55 THEN 245
230: 1 = 1 + !/(!+.0705*0+.0423*0*0+.00927*0-3 + .00015*0-4
+.000277*0- 5+.000043*0-6} -16
235: IF I> = 1 THEN 245
240: N = /lTr/2/Q)*EXP (-Q)*(2-I)
243: RETURN
245: N = /~(ir/2/Q)*EXP (.-Q)*(I-1)
250: RETURN
83
-------
Appendix (Continued)
This program allows entering several sequencial mass flow rates (.up to
10) preceeding a chosen day of reckoning for which the concentration
distribution plume is desired. The groundwater flow is assumed
uniform in the x direction. The effects of dispersion and diffusion,
radioactive or biological decay, and ion exchange are allowed.
The theory for a constant source is from Wilson and Miller (1978) and
should be referred to for details. Superposition holds so that
multiple rates and boundary effects may be included if desired.
The variable mas-rate schedule is set up as follows:
fl
_fLJ
'3
<«
*
Day of
Reckoning
t.
ji__
A typical output can be mapped as illustrated in the following example
(one point at a time with the above program code.)
84
-------
MECHANISMS AND MODELS FOR PREDICTING THE DESORPTION OF
VOLATILE CHEMICALS FROM WASTEWATER
Charles Springer and Louis J. Thibodeaux
College of Engineering, University of Arkansas
Fayetteville, Arkansas 72701
ALSTRACT
Industrial wastewater contains a variety of volatile chemical constituents. The large ma-
jority of the volatiles are organics. The general mechanism and kinetics of desorption
are well established. Generic flux equations for obtaining emission rates into air from
water bodies in the case of flow and non-flow operations are available. What is not gen-
erally available are the necessary transport coefficients. Desorption of volatiles can be
mechanically influenced or nature driven. Surface aerators, submerged aeration, and ther-
mal evaporation are types of systems that have a dominant mechanical influence. Surface
winds, water temperature, and water circulation are important nature driven parameters
that regulate rates of desorption in some surface impoundments. The paper presents the
available literature information of transport coefficients and establishes conceptual mo-
dels for those not available. Directions for research on selected transport coefficients
are presented. Methods for controlling or reducing desorption from surface impoundments
are suggested.
INTRODUCTION
For many years it was generally assumed
that all the organic chemicals in industrial
wastewaters and processed in biochemical ox-
idation reactor were either mineralized or
converted to biosolids. Little or no thought
was given to the fraction that may be vola-
tile and hence transferred to the air phase
to result in air pollution. Thibodeaur and
Millican (14) developed a test method for
determining the fraction of the organic mat-
ter that is volatile and reported the vola-
tile fractions for several industrial waste-
water. Numerous investigators have detected
known volatile chemical species in waste-
waters. The fact that volatiles are enter-
ing surface impoundments with the waste-
water is well established. Air sampling
and analysis near surface impoundments and
bio-oxidation reactors have detected orga-
nics in concentrations above background
levels. Several investigations are on-
going to develop methodologies and measure
emission rates from these sources (6, 9).
In addition, models of a predictive nature
are needed.
Verified desorption models provide a
means of predicting emission rates from
surface impoundments. Field testing is ex-
pensive and models can provide an economical
alternative. In the case of proposed faci-
lities models provide the only means of as-
sessing the air-pollution impact of treat -
ing the wastewater or otherwise impounding
the water on the surface for storage, etc.
KINETICS OF DESORPTION FROM WATER
An evaporation process which occurs at
temperatures below the boiling point of the
liquid involved can be treated by conven-
tional mass transfer rate equations. Typi-
cally, the transfer rate is given byTreybal
(16).
NA =
(1)
where N^ is the molar mass-transfer flux of
chemical component A KA2 is the overall
mass transfer coefficient based on the gas
phase driving force, X^ is the mole fraction
in the liquid and X^ is the 1 '.quid r.ole
fraction of A in the surrounding air.
85
-------
The coefficient, KA2» can be expressed
in the so called "two film" theory or "two
resistance" theorv as:
1
Where now
I—
TkA2
1
(2)
and
are individual
transport coefficients for the liquid and
gas film respectively. The term HXA is
Henry's Law constant in mole fraction form:
IXAAA
(3)
Equation (2) is simply the sum of re-
sistances. If one of the coefficients is
very large compared to the other terms, the
other resistance is said to be "the con-
trolling resistance". Similarly, the gas
phase resistance, I/HXA. 1'Al can be control-
ling if the term HXA is very small. Also,
if the liquid is a pure material, there is
no liquid phase resistance (i.e., water
evaporation from a basin), so the gas phase
is controlling in such cases. Since HXA
for oxygen in water is very large the liquid
phase resistance, I/ kAo , controls. Often
both resistances are important for volatile
organics in wastewater.
Desorption rate equation for continuous
flow surface impoundments without biochemi-
cal oxidation have been developed (15). For
a basin in plug flow, the emission rate
through the air-water interface is:
WAO =
(A)
where WAQ is in moles A per time, Wg is the
water flow rate in moles per time, and A is
the water-air interfacial area. For com-
pletely mixed basins the emission rate is:
WAO = WBXA1{1-1/(1+1KA2A/WB)}
(5)
Equations (4) and (5) give the maximum emis-
sion rates from the basins since XAI is the
inlet water concentration. Equation (1) can
be used for calculating the flux in subre-
gions of the basin for which XA is the lo-
cal surface water concentration. In the
case of non-flow basins the emission rate
through the air-water interface is:
WAO = ^AZAXAO exP (-1KA2At/MB) (6)
where MB is the moles of water in the ba-
sin, t is time, and XAQ is the initial con-
centration of A in the water. Equation (6)
assumes the water is of uniform concentra-
tion and no loss of chemical A through bio-
chemical oxidation, absorption on the muds,
etc.
Equations (1) through (6) summarizes
the kinetics and flux equations for comput-
ing emission rates for both flow and non flow
basins. The critical parameter for obtain-
ing reliable estimates of emission rates is
KA2- This overall transport coefficient \s
inturn dependent upon -*-kA2 and 2kAi as shown
by Equation (2). In general good estimates
of these two transport coefficients for most
wastewater impoundments under all environ-
mental conditions are not available. The
following section summarizes the available
information on these two transport coeffi-
cients for surface water impoundments.
TRANSPORT MODELS AND CONCEPTS FOR SURFACE
IMPOUNDMENTS
Table 1 contains a classification of
liquid containing surface impoundments. Me-
chanically influenced impoundments are those
that contain some equipment that mixes or
adds heat to the liquid. This influence is
in addition to that which occurs naturally.
Natural processes or nature driven process
are the actions of wind, temperature, sun-
light, rain, etc. that influence the water
environment. The transport coefficients of
mechanical influence can be drastically dif-
ferent in magnitude from those of natural
processes. These differences must be quan-
tified by appropriate models if realistic
flux rate calculations are to be obtained.
Aerat&d Basins
Aerated Stabilization basins (15) and
Activated sludge wastewater treatment ba-
sins (3) have received considerable atten-
tion with respect to air-stripping of chemi-
cals. The surface of such basins have
been divided into two zones. One zone is
turbulent (or forced zone) and exists in
a region directly influenced by and imme-
diately surrounding each surface aerator.
The remain surface is virtually undisturbed
by the aerators and is called the natural
zone.. Two sets of transport coefficients
characterize these basins. The total flux
rate is the sum of both zones:
(f)
AQ
(XA-XA)Af+2K^) (XA-XA)An
(7)
where the superscripts denote forced and
natural zone coefficients and Af and An are
the surface areas of each. The individual
-------
Table 1
Types of Liquid Surface Impoundments
1. Mechanically influenced wastewater im-
poundments
a. Aerated basins-surface and/or dif-
fused aeration
b. Thermal evaporation ponds-heat added
c. Pan evaporation-heat added
2. Nature driven wastewater impoundments
a. Natural evaporation ponds
b. Stabilization basins
c. Water holding ponds
3. Solvent (non-aqueous) evaporation pits
9 (f)
gas phase coefficient for this zone K^ ,
is obtained from Reinhardt (10). The indi-
vidual liquid-phase coefficient for the
natural zone, KJ^jJ > is obtained from Cohen,
etal (?), and the individual gas-phase co-
efficient for the natural zone., ?K^?', is
obtained from Harbeck (5). ^K^ is obtain-
ed from the oxygen absorption coefficient
(15). The model and respective transport co-
efficients appear to give reasonable esti-
mates of the methanol emissions from aer-
ated stabilization basins treating pulp and
paper industry wastewaters. Table 2 con-
tains a comparison of field measurements
and calculated rates (9). A similar meth-
ology has been developed for diffused aer-
ation activated sludge system (4). Very
low stripping fractions (VL%) were predict-
ed and verified for acrilonitrile in a high
solids activated sludge laboratory scale
bio-oxidation reactor. Table 3 contains
some typical coefficients.
Pan Evaporation
Hazardous chemical emissions occur to
the atmosphere from wastewater produced by
the wood preserving industry when "treated"
by evaporation methods. The evaporation
process can be accelerated by directly ap-
plying heat to evaporate the water, as in a
thermal pan evaporator. Here the waste-
water is piped to an open vessel that uses
an external heat source. Merrill, etal,
(8) have performed laboratory pan evapora-
tion experiments with pentachlorobenzene
and naphthalene in which practically all
the chemical dissolved in the water vapor-
ized with the water.
Employing Merrill's data Thibodeaux
(13) demonstrated that the emission rate of
PCB and naphthalene in high temperature
water evaporation is directly related to
the water evaporation rate:
NA = NB(xA-xA)
(8)
where N-g is the water evaporation in moles
per area per time. A film theory model for
high flux rates was used to support Equa-
tion (1). For all pan evaporation experi-
ments combined, Ng = 380 moles/hr-m and
lKA2 = 360 moles/hr-m2.
Water Holding Ponds
Holding ponds are often-time used for
short or long-term wastewater storage.
These ponds are also used as water stabili-
zation basins, natural evaporation basins,
run-off catch basins, etc. Holdponds can
Table 2
Aerated Basin Physical, Chemical Data and Methanol Emission (9)
Mill-Visit H
XA
v.8
PA2
number
flux calc. flux
(Acres) (mi/hr) (mg/L) (°c) observations (ng/cm^.s) (np/cm^.s)
1-2
1-3
2-1
2-2
3-1
3-2
4-1
4-2
.24
.30
.29
.38
.39
.43
.43
.43
127
127
265
265
7
7
110
110
3.1
6.2
15.8
13.2
7.3
7.8
11.8
10.3
23.7
6.0
16.1
28.3
14.0
-
-
9.1
25.0
28.7
28.1
33.8
34.4
36.3
36.2
35.6
3
6
3
6
1
4
3
3
0.10-0.36
0.42-4.7
2.7 -4.1
.80-4.8
.43
.069-1.1
2.4 -3.3
.60-2.8
.63
.40
2.5
4.9
1.6
1.9
1.7
1.4
*Flux gas phase controlled and Harbeck's correlation used (5).
87
-------
Table 3
Typical Overall Liquid Phase Volatilization Coefficients
Type of Water Impoundment
TKA2 (mol/hr-m2)
Aerated Basins (3)
Acrylonitrile, calculated, turbulent zone
Acrylonitrile, calculated, natural zone
Pan Evaporation (13)
Laboratory test, PCB and Naphthalene
Holding Pond (7)
Tritium, outdoor basin, f'eld measurements
42500.
210.
360.
39.
be generally categorized as open air waste-
water vessels in which the chemical desorp-
tion forces are nature driven. When wind
is present, the correlations for the re-
spective gas phase and liquid phase trans-
port coefficients are the same as in the
case of the natural zones of aerated basins
and equations developed by Harbeck (5) and
Cohen, etal (2) can be used.
It would appear that the emission rate
of trace volatile chemicals from aqueous
water bodies is drastically reduced in the
absence of wind. As noted above, wind is
a prime mover of the air and water near the
interface and promotes relatively hif.h che-
mical exchange rates. Even without wind
there are still some subtle driving forces
which results in a non-zero emission rate.
The primary driving forces under a no wind
condition appears to be temperature differ-
ences across the interface and the direc-
tion of water vapor movement.
When the water surface is warmer than
the overlying air, an unstable condition is
present. The individual gas-phase trans-
port coefficient is affected by air bouyarry
which can be quantified by the Grashoff num-
ber for heat-transfer. This situation is
similar to heat transfer from a hot plate
facing upward which has been studied ex-
tensively. Using the analogy theories re-
liable estimates of the gas phase coeffi-
cient can be made. The following relation
is suggested:
ShA
= 0.14(GrScA1)
1/3
(9)
where Sh^ = 2kAiL/CiDA;L is the Sherwood
number, Gr = g61L3(Ti-T1)/V2 is the Grashoff
number for heat transfer and ScAi =
is the Schmidt number.
When the water surface and the air
above are at the same temperature but the
air is unsaturated there can be water evap-
oration. Water vapor has a lower molecular
weight than the surrounding air so that
when evaporation is occuring micro-instabil-
ity exist immediately above the water sur-
face to promote exchange. Thibodeaux (12)
demonstrates how to quantify the individual
gas-phase coefficient for this situation
based on the Grashoff number for water eva-
poration. The relation is the same as Equ-
ation (9) with Gr replaced by GrB1, where
(10)
In the case where the air is colder
than the water, a stable condition exists
and the transport coefficient is likely
less than that obtained from Equation (9).
When the air is super-saturated with water
vapor, there is a net condensation of water
at the surface and Equation (10) does not
apply. For these stable cases little is
known about the nature of the individual
gas-phase coefficients and no quantitative
expressions have been developed. What can
be said is that the process is molecular
diffusion controlled and much slower than
the unstable case.
All the above was concerned with the
individual gas phase transport coefficient.
The individual liquid phase transport co-
efficient also need be quantified if real-
istic emission rates are desired. If the
88
-------
gas phase transport is not controlling then
liquid mixing and renewal of surface layer
with water undepleted of the volatile chem-
ical becomes the controlling mechanism.
This mechanism is quantified by the magni-
tude of the individual liquid phase trans-
port coefficient. The direction of heat
transfer would appear to also play a sign-
ificant role in the magnitude of this co-
efficient.
the water impoundment is loosing
heat through the air-water interface to
the surrounding, the water column is un-
stable since the cooler surface water is
more dense than the underlying warm water.
The cold water tumbles down just as occurs
at lake surfaces during the fall of the
year. This plunging cold water is un-
doubtly effective in bringing chemical
species to the surface. No quantitative
expressions have been located to obtain es-
timates of this transport coefficient, how-
ever, some oxygen doefficients or heat tran-
sfer coefficients for lakes may be available
for use as estimates. Since the air layer
is also unstable when the water is loosing
heat, the combination should give the high-
est emission rate likely when no wind is
present .
Waen heat is mrving from the air to the
water, a stable water condition results
since warm water tends to float (.la thermal
stratification). Water mixing is slowed con-
siderbalJy. Horton, etal (7) report on the
rate of loss of tritium from a salt-strati-
fied surface impoundment. Based on their
field data, the coefficient shown in Table 3
was obtained. It appears that this omission
process is liquid-phase controlled- in this
case just as it was for pan-evaporation.
Solvent (no— aqueous) Evaporation Pits
Hopefully the storage of pure solvents
or solvent mixtures in open pits will be
discouraged since evaporation can be ex-
tremely high and the associated air pollu-
tion and odor problems significant. In the
case of pits containing pure solvent or a
solvent layer above water, evaporation is
gas-phase controlled and highly dependent
upon vapor pressure:
NA = 2KAl(PA°-PA)/PT
(ID
Where p^° is pure component vapor pressure,
PA is the vapor pressure of the volatile
species in air and pT is the total pressure.
In the case of mixtures the transport rate
may be dependent upon the individual liquid
phase coefficient. In general, the emission
mechanisms for solvent mixtures will be sim-
ilar to those from water. Techniques such
as those developed by Springer (11) , for
hydrazene and related compounds, can be us-
ed for estimating the emission rates from
non-aqueous evaporation pits.
CONTROL OF CHEMICAL DESORPTION
The following is a list of means that
can be used to control the emission of vol-
atile chemicals entering surface impound-
ments:
*maximize biochemical oxidation and
minimize desorption through high bio-
solids and/or pure oxygen,
*build wind fences to decrease the ef-
fect of the wind on the transport co-
efficients,
*cover water surface with a "membrane"
(i.e., oil film, ping pong balls, etc. )
*completely enclose the impoundment
and scrub or incinerate the exit air,
or
*employ alternative treatment or stor-
age which involves no air-water inter-•
face.
CONCLUSIONS
*Volat~iles can and do escape from sur-
face water impoundments and enter the
overlying air mass.
*The general principles and mechanism
for the desorption of volatiles from
surface impoundments are well estab-
lished.
*Accurate calculation of emission
rates is based on reliable informa-
tion about transport coefficients in
the gas phase and in the water phase.
*Some reliable transport coefficients
exist for aerated basins.
^Research is still needed on a variety
of surface impoundment transport co-
efficients, particularly pan or ther-
mal evaporation and the no-wind con-
dition.
*Research is needed in the area of
control measures.
89
-------
REFERENCES
1. Bird, R. B., W. E. Stewart, and E. N.
Lightfoot, Transport Phenomena, John Wiley,
•i.Y., 658-668.
2. Cohen, Y., W. Cocchio, and D. Mackay,
Laboratory Study of Liquid-Phase Controlled
Volatilization Rates in the Presence of Kind
Waves, Environ, Sci, Techno., 12(5):553-558.
3. Freeman, R. A., Stripping of Hazardous
Chemicals from Surface Aerated Waste Treat-
ment Basins. In: Proceedings of Air Pol.
Control Asso./Water Pol. Cont. Fed. Special-
ty Conf. on Control of Specific (Toxic) Pol-
lutants, Gainesville, Florida, February 1979.
4. Freeman, R. A., J. M. Schroy, J. R.
Klieve, and S. R. Archer, Experimental Stu-
dies of the Rate of Air-Stripping of Haz-
ardous Chemicals from Wastewater Treatment
Systems, Air Pol. Control Asso. Mtg. Mon-
treal, Canada, June 1980.
5. Harbeck, G. E., A Practical Field Tech-
niques for Measuring Reservoir Evaporation
Utilizing Mass-Transfer Theory, Geol. Sur-
vey Prof. Paper 272-E, U.S. Govt. Print.
Off., Washington, D.C., p. 101-105.
6. Harrison, P. R., R. M. Stockdale, and
L. F. Tischler, Development of Sampling
and Experimental Methodologies for the
Measurement of Hydrocarbon Emissions from
Petroleum Refinery Wastewater Systems, U.S.
EPA, IREL, Res. Triangle Park, N.C.
7. Horton, J. H., J. C. Corey and R. M.
Walace, Tritium .Loss from Water Exposed
to the Atmosphere,Environ, Sci. Techno.,
5(4): 338-343.
8. Merrill, R., Wood Treating Industry
Multimedia Emission Inventory, Acurex Corp.
Final Report (in preparation), U.S. EPA,
IREL, Cinn., Ohio.
9. Parker, D. and L. Thibodeaux, Measure-
ment of Volatile Chemical Emissions from
Wastewater Basins, Final report (in prepar-
ation), U.S. EPA, IREL, Cinn., Ohio.
10. Reinhardt, J. R. Gas-side Mass-trans-
fer Coefficient and Interfacial Phenomena
of Flat-bladed Surface Agitators, Ph.D.
Dissertation, University of Arkansas, Fay-
etteville, Arkansas.
11. Springer, C., Evaporation and Disper-
sion of Hazardous Materials, Final report,
Dept. Air Force, Air Force Office Scienti-
fic Research, June 1979.
12. Thibodeaux, L., Chemodynamics - Envi-
ronmental Movement of Chemicals in Air,
Water, and Soil, J. Wiley, New York, 159-
162.
13. Thibodeaux, L. and R. Merrill, Pen-
tachlorophenal and Naphthalene Emission to
Air During Thermal Evaporation of Wastewat-
er, in preparation.
14. Thibodeaux, L. and J. D. Millican, •
Quantity and Relative Desorption Rates of
the Air-strippable Organics in Industrial
Wastewater, Environ. Sci. Techno., 11(9):
879-882.
15. Thibodeaux, L. and D. Parker, Desorp-
tion of Selected Industrial Gases and Li-
quids from Aerated Basins, Amer. Inst. Chem.
Eng. Symp. Series 156-Air Pol. and Clean
Energy, p. 424-434.
16. Treybal, R. , Mass Transfer Operations
3 ed., McGraw Hill, New York, p. 109-111.
90
-------
MANAGEMENT OF HAZARDOUS WASTF BY UNIQUE ENCAPSULATION PROCESSES
H. R. Lubowitz
Environmental Protection Polymers
13^1^ Prairie Avenue
Hawthorne, CA. 90250
C. C. Wiles
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
ABSTRACT
The Solid and Hazardous Waste Research Division of the U.S. Environmental Protection
Agency has conducted research and evaluation of uniaue methods for encapsulating hazardous
wastes. These methods are intended for the treatment of extremely toxic wastes not readily
managed by other treatment processes. Three process schemes have been developed using
resins to encapsulate dry•unconfined wastes, sludges, and confined or containerized wastes.
The processes have been shown to be capable of producing modules of waste that are extreme-
ly resistant to physical and chemical stresses.
1
INTRODUCTION
This paper reviews several studies of
uniaue methods for encapsulation of haz-
ardous wastes. The studies, which were
supported by the Solid and Hazardous Waste
Research Division, Municipal Environmental
Research Laboratory, U.S. Environmental Pro-
tection Agency (EPA), grew out of recom-
mendations of earlier hazardous waste
studies that had been the basis for FPA's
197^ report to Congress on hazardous waste
These studies had clearly indicated that
the hazardous waste problem would reauire
new solutions to accommodate extremely
toxic residuals not readily manageable by
routine processes. Although available, no
detailed descriptions of the encapsulation
methods are given here; rather, concepts
are presented that offer uniaue opportu-
nities for processing hazardous materials
for long-term, safe retention.
The main objective of the hazardous
waste encapsulation program is to develop
and/or evaluate unique methods for encapsu-
lating extremely toxic wastes, particularly
those that could not be adequately managed
in an environmentally acceptable manner by
other means. The advantages sought are the
following:
— The process should yield high per-
formance products.
— The process should be simple and
products are readily reproducible.
— The process without adjustment
should be capable of managing wastes
with varied compositions and con-
sistencies.
— The process should be versatile
(i.e., adaptable to technical
advances, changes in wastes, etc.).
— The process should be reasonable in
cost (note that a measure of reason-
able expense would be relative,
depending on the degree of toxicity
or hazards associated with the waste
and the effectiveness of other
available alternatives for managing
that waste.).
91
-------
Ideally, the objective of any hazardous
waste management program must be to destroy
the hazardous waste completely, to detoxify
it, or to recover it for reuse. In reality,
however, toxic residues occur that require
some type of ultimate disposal. Encapsu-
lation offers an ultimate disposal option in
that its objective is to stabilize haz-
ardous material to resist its delocalization
by environmental stresses. High performance
encapsulation is required in view of the
trend to manage concentrated contaminants
in order to reduce waste volume in final
deposition.
Different ultimate disposal and/or
long-term retention options may be consid-
ered for encapsulated waste, depending on
its final characteristics and the control-
ling regulatory provisions. If technical
considerations alone were used, properly
processed encapsulated hazardous material
could be stored in salt mines, landfills, or
other storage facilities—or they could
even be disposed of in the ocean. Place-
ment of encapsulated hazardous waste in a
landfill may offer some advantages, such as
placement in specifically chosen locations
and provision of additional options for
controlling the waste (e.g., easy removal
of encapsulates if necessary). On the
other hand, it may be considered a dis-
advantage to leave such wastes in place for
the concern of future generations. But
regardless of which options are used, some
form of ultimate disposal is and will pro-
bably remain a needed hazardous waste
management technique, and the objective of
encapsulation must therefore remain as the
means to prevent, or at least markedly
reduce the rate of, entry of contaminants
into the environment.
ENCAPSULATION VERSUS CHEMICAL FIXATION
In proper fixation of a hazardous
waste, the contaminants must be dispersed
in the molecular and/or colloidal state
through the matrix of fixing materials.
The chemical affinity of the contaminants
and a suitable fixer will stabilize the
waste by yielding a composition capable of
resisting chemical stresses. Under real
conditions, however, this result is dif-
ficult to realize because of the many dif-
ferent and changing waste constituents, and
the requirement of contaminant solubility.
Adjusting the process to specific components
in mixed industrial waste gives rise to cost
incurring and probably technically diffi-
cult tasks.
Perhaps a more realistic goal would be
to microencapsulate the contaminant (i.e.,
to encapsulate each individual particulate
of contaminant) by localizing it in a
matrix that is chemically stable. The en-
cased contaminants could then exist as
particulates in sizes greater than the
molecular or colloidal size range and still
be adequately stabilized. Whether chemical
fixation or microencapsulation, the re-
sulting waste modules must be physically
stable and mechanically strong to assure
long-term containment of contaminants when
placed under stresses expected in manipu-
lation and final disposition. While true
chemical fixation appears difficult to
achieve, criteria to effect high perform-
ance microencapsulation are also exacting,
because chemical affinity forces do not
come into play to aid the stability of the
microencapsulated contaminant in the event
flaws or fractures occur in the module.
Making insoluble compounds of contaminater
(stabilization) prior to microencapsulation
must take into account the dispersing ef-
fect of aggressive waters as well as its
solubilizing effect.
POSSIBLE APPROACHES TO THE PROBLEM
Three general approaches are available
to solve the dilemma. These are:
• To tailor the fixation process to
the different waste compositions
by analyzing the waste and, based
on its chemical composition, making
appropriate chemical adjustments to
the process to handle each of the
many different hazardous components
or classes of components that might
be present. This procedure has
proven difficult to accomplish in
many cases. (In our opinion, en-
cased contaminant particulates, as
well as contaminant fixation,
characterized the products of many
chemical fixation processes.).
• To employ a generally applicable
microencapsulation process rather
than chemical fixation, but to
assure that each contaminant is
sealed within a matrix. This
approach is likely to prove dif-
ficult because of the exacting
92
-------
processing conditions that are ex-
pected to be reauired to assure
that all of the contaminant parti-
culates in a mixed stream are
suitably microencapsulated. High
performance microencapsulation with
minimal amounts of matrix material
distinguish such processes.
• To use a generally applicable micro-
encapsulation process and to com-
pensate for probable incomplete
contaminant isolation by macro-
encapsulating the matrix containing
the contaminant. Satisfactory
macroencapsulation requires use of
materials having engineering pro-
perties.
The last approach has been the major
direction of our hazardous waste encapsu-
lation program, and is judged to offer the
best potential for meeting the program
objectives. Although we have conducted
special evaluations of asphalts and cements,
our efforts have concentrated on unique
polymeric approaches.
POLYMERIC ENCAPSULATION SCHEMES FOR HAZ-
ARDOUS WASTES
Three different waste forms have been
considered:
• unconfined dry wastes,
• preconfined (containerized) wastes,
and
• sludges
Investigations of methods to encapsulate
these waste forms have led to three general
schemes: (1) Fusion of resin onto the sur-
face of microenc'apsulated waste, (2) fusion
or spray-on of resin onto preformed case-
ments holding containers of waste, and (3)
overpacking of waste containers with welded
plastic containers. (See Figure 1 and
Table 1.) Physical and mechanical require-
ments of encapsulates dictate use of macro-
molecular materials. Organic resins were
preferred to inorganic and metallic
materials due to greater chemical stability
of these materials in the low-in-cost range.
Scheme Description
In scheme I, the polyolefin, high-
density polyethylene (HDPE), is fused onto
the surface of a polybutadiene microencap-
sulated waste forming a seamless jacket
(macroencapsulation) around the composite
matrix. This scheme will accommodate dry
(dewatered), unconfined waste, or difficult
to dewater sludges made friable by use of a
dehydrating agent such as off-specification
Portland cement.
Scheme II provides the opportunity to
use the fusion techniques of scheme I to
encapsulate containers of hazardous waste
or to spray-on or brush-on a resin jacket
around containers of hazardous materials.
The resin jacket is applied to a fiber
glass/epoxide substrate encasing leaking
and corroding 55-gal drums. The fiber
glass encasement is made watertight by
fusing or spraying-on or brushing-on the
resin jacket. If required, further rein-
forcement may be obtained by filling voids
in the encapsulate with concrete, sand,
soil, or other suitable low-cost fillers.
HDPE resin fused onto fiber glass casements
provided mechanical strength similar to
that experienced when resin was fused onto
the surface of microencapsulated waste.
The sprayed-on resin jacket, however, did
not produce similar mechanical properties
without repeated applications.
Scheme III involved overpacking of 55-
gal waste containers with welded plastic
receivers. This scheme resulted from
efforts to provide a quick solution for
managing corroding and leaking 55-gal drums
of hazardous waste. To accomplish this
objective, techniaues used to weld large
plastic pipe were modified and applied to
encapsulation. Plastic, 85-gal receivers
were used to overpack 55-gal containers of
hazardous waste. A seamless container was
formed by butt-welding a piece of flat
plastic polyethylene (PE) stock to the roto-
molded PE 85-gal containers (Figure 2).
Information is available describing
studies to produce scheme I, scheme IT.
and scheme III encaDsulates.2:,3j5,5,o,7
Methods Used in the Resin Fusion Process
of Schemes I and~lT
Three methods can be used for encapsu-
lation by the resin fusion process of
schemes I and II (Figure 3): (1) Micro-
encapsulation of waste particulates with
polybutadiene and subsequent macroencaosu-
lation of the matrix with HDPK, (2) use of
low-cost cementing agents to form a waste
93
-------
4
( l
PLASTIC
JACKETS
INTERPACIAL
LAYER
REINFORCEMENT
(MICROENCAPSULATED
SCHEME 1 WASTE MODULE)
L
w
(FIBERGLASS)
SCHEME II
\
WELDED LID
•LOW-COST FILLER
•PLASTIC OVERPACK
SCHEME
Figure 1. Three•schemes for encapsulating hazardous waste: Scheme I, resin fused onto
surface of microencapsulated waste; scheme II, resin fused gnto or sprayed
onto fiber glass substrate holding containers of waste; scheme III, overpacking
of 55-gal waste containers with welded plastic containers.
-------
55-GAL DRUM
OP HAZARDOUS'
WASTE
^SEAMLESS
SEAL
-PE (1/4" THICK) ROTOMOLDED
CONTAINER (85-GAL)
Figure 2. Plastic overpack system for encapsulating 55-gal drums of hazardous wastes.
-------
HDPE SEAMLESS JACKET, 1/4" THICK
\\X\\\\\\
DEWATERED WASTE/
POLYBUTADIENE BINDER
WASTE SLUDGES/
PORTLAND CEMENT
\ \ \
x\ \ \
SPACE FOR FTBERGLASS/EPOXIDE
SECURING CASEMENT (1/50" THICK)
WASTE
METHOD 1
METHOD 2
METHOD 3
Figure 3- Three methods that can be used for encapsulation by the resin fusion process of
schemes I and II: Method 1, mlcroencapsulatlon of partlculated waste with poly-
butadiene and subsequent macroencapsulatlon with HDPE; Method 2, use of low-cost
cementing agents to form a waste block onto which the HDPE Jacket Is fused;
Method 3, use of fiber glass/epoxlde casement as substrate for HDPE Jacket.
-------
TABLF 1. FABRICATION TECHNIQUES AND PERFFRRED MATERIALS FOR ENCAPSULATION SCHEMES
Scheme
Waste
Process
Material
II
III
Unconfined
• dry
• sludges
Small Containers
• 55-gal drums
• other
55-Gal Drums
• other
Fusion
Fusion
Spray-on or Brush-on
Welding
Polyolefin
Polybutadiene
Polyolefin
Glass fiber
Glass binder (epoxy or
polyester)
Concrete
Polyolefin
block onto which the HDPF .jacket is fused,
and (3) use of a preformed fiber glass/
thermosetting resin casement as a substrate
for the HDPE jacket.
In the first method, 1,2-polybutadiene
is mixed with the particulated waste which,
after solvent evaporation, yields free
flowing dry resin coated particulates.
Further processing involves the formation
of a block of the polybutadiene/waste
mixture by the application of moderate heat
and pressures in a mold. The final encap-
sulated module is produced by fusing a
1/4-inch-thick HDPE jacket onto the block.
The second method involves the use of
low-cost cementing agents such as Portland
cement to form the waste block onto which
the HDPE Jacket is fused. This modifi-
cation provides the opportunity to process
sludges without having to form particulates
and the limited number of wastes that are
not compatible or will not agglomerate with
polybutadiene.
The third method employs a fiber glass
thermosetting resin matrix as a substrate
for the HDPE jacket. Powdered HDPE is
fused onto fiber glass which yields re-
ceivers to contain hazardous waste, which
is usually containerized. The receivers
are then sealed with additional fiber glass
(for the lid) and the HDPE.
POLYMER SELECTIONS
The selection of polymers for forming
the encapsulated modules was based on
factors consistent with the program objec-
tive—to provide a cost effective system
capable of managing extremely hazardous
waste. A rather Intensive screening of
available polymers resulted in the con-
clusion that the polyolefins, and particu-
larly PF, offered the best overall
advantages. PE is mass produced, well
characterized, relatively low in cost,
excellent in chemical stability, and
flexible with mechanical toughness. The
use of PE for macroencapsulation is common
to all three resin fusion methods discussed
and to the pipe-welding scheme.
Polymers Selected For Method 1 of Scheme I
Microencapsulated Matrix/Macroencapsulated
Module
To form the microencapsulated waste
block (the block onto which the macro-
encapsulate is formed by resin fusion), a
thermoset resin was produced from 1,2-poly-
butadiene and powdered PE resins. Crude
polybutadiene and scrap PE are acceptable
as well as virgin materials. The poly-
butadiene was thermoset by peroxides, as
employed in peroxide vulcanization of
rubber. By contrast, however, the 1,2-
configuration gives a high yield of chemical
cross-links in a fast chemical reaction.
The resulting polymers have singly bonded
carbon backbones, which provide resistance
to oxidative and hydrolytic degradations
and to permeation by water. Furthermore
the polybutadiene possessed characteristics
that aided processing and provided desir-
able chemical and mechanical properties.
The resin-treated waste particulates were
free-flowing and stable under atmospheric
conditions, provided they were sheltered
from the sunlight. This is an advantage
97
-------
not available with two-component resin
systems where reaction proceeds upon com-
bination of the two parts. This stability
permitted the subsequent microencapsulation
step to be scheduled as time permitted. In
addition, the chemical nature of the 1,2-
polybutadiene and the thermosetting reaction
assures reproducible encapsulation of waste
particulates, even when their chemical com-
positions and consistencies are mixed. In
the subsequent fusion step, powdered high-
density PE grafted chemically onto the
polymer backbones and provided the final
matrix with ductile Qualities. Various
combinations of the two resins (i.e., 1,2-
polybutadiene and PF) permitted tailoring
of the matrix mechanical properties without
reduction of system stability when exposed
to severe chemical stress.
The macroencapsulated module is seam
free, and observations showed the 1/^-inch-
thick HDPE jacket to be mechanically and
chemically locked to the surface of the
block of microencapsulated waste. If de-
sired, variations in processing conditions
will permit controlling the degree to which
the PE bond penetrates the surface of the
agglomerate, providing additional mechani-
cal and chemical resistance. This control
is an outstanding feature of the resin
fusion process. It also makes low tolerance
machined, simple molds suitable for fusing
resin jackets onto microencapsulated waste.
Polymers Selected for Method 3 of Scheme II
In the case of the spray or brush-on
system of scheme II, recent developments
with aliphatic polyurethanes suggested that
a formulation using a water-based poly-
urethane could be developed that would
yield tough, flexible coatings upon atmos-
pheric curing. As a class atmospheric
curing resins normally yield brittle
coatings. Additional studies produced a
formulation and a process that permitted
the brushing or spraying on of a tough,
flexible coating to a preformed fiber glass
epoxide casing (method 3)• Excellent ad-
hesion of the polyurethane overcoat to the
glass casing was attributed to hydroxyl
groups stemming from the epoxide and inter-
acting with the polyurethane in curing.
For field operations, the use of water
based polyurethane has advantage over
organic solvent entrained polyurethane.
Although the encapsulates were watertight,
the major problem with the system was that
it required the application of more than
one coat to provide the stringent mechanical
properties desired for manipulating the
ensemble. Ve are confident that additional
materials research would yield ways to over-
come this problem. At the present stage of
development, spray-up method can be used to
ston or prevent leaking of hazardous mate-
rials from malfunctioning containers.
Polymers Selected for Scheme III
For the 55-gal drum overpack system of
scheme III, rotomolded wide-mouth, medium-
density PF receivers, normally used as
liners for vessels, were selected. Flat
stock of high-density PE was selected for
welding a lid onto the wide-mouth receiver.
These items were commercially available and
required only slight modifications to accom-
modate our objectives. The welding appara-
tus, frame, and auxiliary eauipment were
either adapted from the plastic pipe welding
industry or fabricated to meet our specific
needs.
The use of dissimilar materials to
encapsulate 55-gal drums demonstrated that
materials need not be matched for making
proper closures. The materials employed
were those most readily available. In a
scale-up operation, materials and closure
design would be specified to facilitate
product fabrication and assure duality
control.
PERFORMANCE EVALUATIONS
Specimens of encapsulated modules have
been subjected to various chemical and
mechanical stresses. Tests were designed to
evaluate the modules' resistance and capa-
bility for isolating the encapsulated con-
taminants from the environment.
Resin Fusion Methods (Schemes I and II)
In laboratory tests of leaching re-
sistance, modules containing extremely
difficult to manage hazardous wastes were
encapsulated by each of the three resin
fusion methods (Table 2) and exposed to
rather harsh environments for up to 120
days. Solutions to which these modules
were exposed included, for example, 1.5N
HC1, NaOH, citric acid, (NHl^S, Dioxane,
NF[|OH, etc. Results indicated excellent
retention of contaminants, including dif-
ficult to contain arsenic-bearing salts
and hazardous sodium salts.
98
-------
TABLE 2. WASTES MANAGED IN TESTS OP RESIN FUSION METHODS
Method
Waste
Dewatered sludges:
Electroplating
Battery
SO scrubber
A
Pigment
Chlorine, calcium fluoride
MSMA (monosodium methane-
arsenate)
ASpCL (copper smelters)
As2S3
Aaueous solutions of salts
Cu, Cr, NJ, Cd, Zn, Pb, Sb,
As, Fg, Se
In tests conducted at the U.S. Army Corps of
Engineers Waterways Experiment Station, the
resin fusion process successfully managed
the wastes indicated in Table 3. These
tests involved exposure of encapsulated
modules placed in columns and exposed to q
leaching solutions over a 2-year period.">"
Specimens of encapsulated modules were
also evaluated for resistance to mechanical
and physical stresses, (i.e., compression,
freeze-thaw, impact strength, punctur-
ability, bulk density, drop test, etc.).
Results showed that the encapsulated modules
can be expected to withstand mechanical and
physical stresses greater than those ex-
pected during handling, transportation, and
disposal/storage. Details of the chemical
and_mechanical tests are available.^'3jO,9
Overpack Method (Scheme III)
The modules resulting from scheme III,
which involved waste containers overpacked
with welded plastic containers, did not
provide the structural stability of schemes
I and II, even though they were seamless.
This is as expected because the plastic
jackets were not reinforced as in schemes
I and II (see Figure 1). Additional
structural stability may be obtained by
filling the empty spaces between the over-
pack and the inner container with cement
or foamed resin. Chemical stability was
not tested but is expected to be similar.
COSTS
Some preliminary costs have been
estimated for the encapsulation processes
discussed in Table b. Commercial resin
costs were determined to account for more
than 50 percent of the production costs
for the resin fusion option. The use of
scrap materials should help to reduce
these costs significantly.
Cost can be a relative matter. For
example, the fact that these processes may
be the only existing techniaues that can
effectively manage certain hazardous wastes
must be factored in with other cost con-
siderations.
Other cost evaluations have indicated
that encapsulation techniques have re-
latively low life-cycle costs, indicating
that overall costs are favorable when
compared with other processes.10 But
detailed cost estimates for these pro-
cesses cannot be accurately developed with-
out additional pilot scale and field work.
99
-------
TABLE 3. ENCAPSULATED WASTE EVALUATED AT THE, U.S. ARMY
WATERWAYS EXPERIMENT STATION
Code No.
100
200
300
400
500
600
700
800
900
1000
Source of Waste
SO scrubber sludge, lime process, eastern
xcoal
Electroplating sludge
Nickel - cadmium battery production sludge
SO scrubber sludge, limestone process
eastern coal
SO scrubber sludge, double alkali process
eastern coal
SO scrubber sludge, limestone process,
western coal
Pigment production sludge
Chlorine production brine sludge
Calcium fluoride sludge
SO scrubber sludge, double alkali process,
western coal
Major Contaminants
p Q Qn ~~ /QCi
\sCL , ^^h / ^*- ~)
Cu, Cr, Zn
Ni, Cd
Cu SOi ~/SO ~
Na, Ca, SOj^/SO ~
Ca, S0i|"/S03~
Cr, Pe, CN
Na, Cl~, Kg
Ca, P~
Cu, Na, S0lt~/S03~
TABLE M. ESTIMATED COSTS OF ENCAPSULATION
Process Option Estimated Cost
Resin Fusion:
Unconfined waste $110/dry ton
55-Gallon drums *0.45/gal
Resin spray-on Not determined
Plastic Welding $253/ton = $63.Wdrum
(80,000 55-gal drums/year)
100
-------
Unique Properties and Advantages of the
Polymer Encapsulation Techniques
The polymeric encapsulation schemes
are considered uniaue. These systems
offer exceptionally high resistance to
rather severe chemical and mechanical
stresses. Unlike other polymeric systems,
the resin fusion process does not reauire
the mixing of 2 components, thus not con-
straining the process to timed operations.
This feature provides flexibility in pro-
cessing. In addition, the 1,2-poly-
butadiene/PE system has characteristics
that permit unusually high loading of the
filler (waste). The final encapsulated
product can contain approximately 9^ per-
cent waste by weight. This percentage mav
vary some from waste to waste. But such
high loading capacities make this polymer
system unique when compared with most
others, and they provide economic advan-
tages with regard to processing and
materials.
toother advantage of the polymeric
systems is that they can accept a wide
range of waste types without modifications.
The processes can, however, be rather
easily tailored for specific circumstances.
Although laboratory and pilot processing
steps and eauipment used may appear some-
what complicated and/or sophisticated,
they are not. Pull-scale systems and pro-
cesses are expected to be rather simple to
conduct.
SUMMARY AND CONCLUSIONS
This paper discusses concepts for
using polymers to encapsulate hazardous
wastes. Polyolefin (primarily high- and
medium-density polyethylene), polybuta-
diene, polyurethane, and fiber glass/
epoxides have been used to produce encap-
sulated modules of hazardous waste. Mod-
ifications of these polymer systems have
made it possible to produce encapsulated
modules capable of handling different
waste forms—dry unconfined wastes, sludges,
and containerized wastes. The encapsula-
tion systems are compatible with a wide
range of waste compositions, and the
modules they produce have been shown to
have high resistance to chemical and
mechanical stresses. They can be expected
to provide hazardous waste containment
over exceptionally long periods of time.
The nature of the systems is such that they
can be tailored to different waste
characteristics and containment objectives.
For example, the encapsulated waste can be
stored under strict control and removed if
reauired, or it can be disposed of under
various regimes.
Details of the processes discussed in
this paper are available as noted in the
references. Fach process discussed is in
a different stage of development. The
polybutadiene/PE resin fusion system for
dry unconfined waste is completed at the
laboratory stage and a full-scale engi-
neering process and eauipment design has
been set forth. The spray-on system re-
auires additional material R&D to produce
structurally acceptable coatings in one
application. The plastic weld overpack
system is essentially under full-scale
development because of its nature. Only
limited specimens, hov/ever, have been pro-
duced. Some additional pilot-scale work
Is reauired, followed by a field demon-
stration of techniaues.
Though uncertain, the preliminary
economics of these systems appear to be
acceptable. If future cost considerations
take performance capabilities into account,
these systems are expected to become more
economically viable.
REFERENCES
1. "Disposal of Hazardous Wastes," Report
to Congress, SW-115. U.S. Environ-
mental Protection Agency, 197^.
2. Lubowitz, F. R., et al. Development of
a Polymeric Cementing and Encapsulating
Process for Managing Hazardous Wastes.
EPA 600/2-77-OU5, U.S. Environmental
Protection Agency, Cincinnati, Ohio.
1977.
3. Lubowitz, P.P., and C. C. Wiles. En-
capsulation "Technique for Control of
Hazardous Wastes. In: Toxic and
Hazardous Waste Disposal, Volume I,
R. B. Pojasek, ed. Ann Arbor Science,
Ann Arbor, Michigan, 1979. pp. 1?9-
232.
4. Lubowitz, H. R., et al. Encapsulation
of 55-gal Drums Holding Hazardous
Wastes. In: Proceedings of the Sixth
Annual Research Symposium—Treatment of
Hazardous Waste, March 1980. FPA 600/
q-80-011, p. ^3.
101
-------
5. Lubowitz, H. R. , and R. W. Telles.
Study of Encapsulate Formation with
Polyethylene Resin and Fiber Glass for
Use in Stabilizing Containerized Haz-
ardous Waste. Draft Final Report,
Contract 68-03-2483, U.S. Environmental
Protection Agency, Cincinnati, Ohio.
1980.
6. Lubowitz, H. R., et al. Encapsulation
of 55-gal Drums of Hazardous Waste by
Welding Polyethylene Resin. Draft
Final Report, Contract 68-03-2483. U.S.
Environmental Protection Agency, Cin-
cinnati, Ohio. 1980.
7. Lubowitz, H. R., and R. W. Telles.
Encapsulating Hazardous Waste by Spray-
ing or Brushing On Atmospheric Curing
Resins. Draft Final Report, Contract
68-03-2183. U.S. Environmental Pro-
tection Agency, Cincinnati, Ohio. 1980.
8. Mahloch, J. L., D. E. Averett, and M. J.
Bartos, Jr. Pollutant Potential of Raw
and Chemically Fixed Hazardous Indus-
trial Wastes and Five Gas Desulfuri-
zation Sludges—Interim report. EPA
600/2-76-182, U.S. Environmental Pro-
tedtion Agency, Cincinnati, Ohio. 1976.
105 pp.
9. Bartos, J. J., Jr., and M. R. Palermo.
Physical and Engineering Properties of
Hazardous Industrial Wastes and Sludges.
EPA 600/2-77-139, U.S. Environmental
Protection Agency, Cincinnati, Ohio.
1977. 77 pp.
10. Hansen, W. G., and H. L. Rishel. Cost
Fffectiveness of Treatmeet and Disposal
Alternatives for Hazardous Waste,
Volume I. Draft Final Report, Contract
68-03-2754, U.S. Environmental Pro-
tection Agency. 1980.
102
-------
ESTIMATION OF POLLUTION POTENTIAL OF INDUSTRIAL WASTE
FROM SMALL-SCALE-COLUMN LEACHING STUDIES
Philip G. Malone
Larry W. Jones
Richard A. Shafer
Robert J. Larson
Environmental Engineering Division
U. S. Army Engineer Waterways Experiment Station
Vicksburg, MS 39180
ABSTRACT
Two industrial wastes, a nickel-cadmium battery sludge and a pigment production
sludge, and three solidified/stabilized products produced from them were subjected to
long-term leaching in small (10 cm diameter by 122 cm) plexiglas columns. Leachate from
untreated nickel-cadmium battery sludge contained large amounts of soluble salt (probably
mostly sodium nitrate) and exceeded drinking water standards for mercury, manganese and
selenium. The pigment production sludge produced leachates exceeding drinking water
standards for cadmium, chromium, mercury, manganese, lead, and sulfate. Of the three
solidification/stabilization processes evaluated, the process using soluble silicate and
cement additives gave the best containment of the heavy metals, but exceeded the un-
treated sludges in initial losses of soluble species. A flyash and lime additive process
produced a product which lowered the overall conductivity of the leachate but did not
significantly lower the losses of the heavy metals. A urea-formaldehyde treatment pro-
cess greatly increased the rate of loss of most contaminants to the leachate. The over-
all effect of the treatment processes on contaminant loss may be lessened because of the
relatively low loss of pollutants from the untreated sludges. After 2 years leaching
time, much less than 1 percent of all heavy metal pollutants had been leached from any of
the sludge samples except from the urea-formaldehyde treated sludge.
INTRODUCTION
The investigation reported here
assesses the pollutant potential of two
industrial wastes. Parts of this work
have previously been reported (1,2,3).
Also addressed is the effectiveness of
three sludge solidification/stabilization
processes in preventing the loss of con-
stituents from treated and untreated
sludge specimens under controlled long-
term, submerged conditions. Two of the
three processes are currently being used
to treat selected industrial wastes on a
commercial basis; the third (the urea-
formaldehyde resin) has been used exten-
sively in handling and transport of radio-
active waste for several years and has
been suggested for industrial waste
applications (1).
Two common, waste sludges which have
high concentrations of toxic heavy metals
are included in this study — a nickel-
cadmium (Ni-Cad) battery sludge and a
pigment production sludge. Ni-Cad bat-
tery sludges are produced during the
precipitation of nickel and cadmium from
their nitrate salts to form the battery
electrodes. The metals precipitate as
the solution pH is increased to 11 or 12
by the addition of sodium hydroxide. The
excess hydroxides and any other materials
remaining in the suspension of spent
salts are settled out to form the sludge.
This sludge typically has a high slurry
pH (12 to 12.5), low solids content (35
to 45 percent solids), and contains
greater than 1.0 percent dry weight of
calcium, nickel, chloride, and silicon
(Table 1).
103
-------
TABLE 1. SELECTED PHYSICAL PROPERTIES AND LEACHING COLUMN
LOADING OF TREATED AND UNTREATED SLUDGES
Parameter
Physical Properties:
% dry sludge solids
Slurry pH
Specific gravity
Median grain size (mm)
Permeability (10 cm/sec)
Ni-Cad Battery Sludge
(#300)
Untreated Process A Process B
45
12.3
3.96
0.044
5.70
21
12.7
2.71
(solid)
1.91
40
9.8
3.68
0.125
1.89
Pigment Prod. Sludge
(#700)
Untreated Process C
41
8.4
3.09
0.016
6.51
26
4.8
1.74
(solid)
160
Leaching Column Loading:
Final product in
column (Kg)
Dry sludge solids in
column (Kg)
13.8
6.15
8.5
1.78
5.1
3.26
11.7
4.83
5.6
1.14
Other ions usually present at greater than
1 g/kg dry solids are cadmium, iron, mag-
nesium, and zinc.
The paint pigment sludge is produced
by treating waste water with lime, sodium
sulfide to precipate the metals, and an
organic polymer to aid settling. Ferris
sulfate is then added to remove the excess
sulfides. This sludge is typically weakly
alkaline (pH 8 to 9), has low solid con-
tent (25 to 35 percent solids), and has
greater than 1.0 percent dry weight of
calcium, chromium, iron, magnesium, lead,
chloride, sulfate, and silicon. Other
metal ions present at greater than 1 g/kg
dry solids in the sludges used in this
study are cadmium, copper, magnesium,
nickel, and zinc. Both sludges present a
high probability of environmental pollu-
tion if disposed of by shallow land
burial.
In the leaching method used, treated
and untreated sludges are placed in small
(10 cm diameter) plexiglass leaching col-
umns. Water saturated with carbon dioxide
was flowed around or through the sludge
for periods of up to 2 years. The leach-
ate produced was collected at progress-
ively increasing time intervals and
analyzed for selected constituents. The
results of these analyses were used to
evaluate the pollutant potential of the
two industrial waste sludges and the ef-
fectiveness of the treatment processes to
contain the sludge constituents.
METHODS AND MATERIALS
Solidification/Stabilization Procedures
Sludges used in this study were mixed
thoroughly and divided into several sub-
samples. A subsample of the two sludges
were sent to three processors for prelimi-
nary evaluation and testing. Two of the
processors (A and B) declined to treat the
pigment production sludge but did submit
samples of treated Ni-Cad battery sludge
for leach testing. The third processor
(C) declined to treat the Ni-Cad battery
sludge but did submit treated samples of
the pigment production sludge for leach
testing. In each case, the effectiveness
of the treatment process is evaluated on
the basis of leaching columns containing
untreated samples of the same sludge type.
The solidification/stabilization
processes used in this study are proprie-
tory. Therefore details on the exact
104
-------
composition of additives and the exact
proportions used are not available. The
general nature of each process, however,
has been supplied by the processors (2).
The processors are identified here by
letter only and no further information can
be supplied.
Process A—
This patented process uses bituminous
flyash and a "lime additive" to produce a
pozzolan product with good structural in-
tegrity. The amount of bituminous flyash
added is related to the amount of total
solids in the waste being treated. The
sludge remains alkaline during the entire
process. The sludge was dewatered by
settling. Approximately 50 kg of the
solidified product was prepared and cast
into 7.6 cm x 40 cm cylinders using stan-
dard concrete-testing molds. The molds
were cured for 30 days under humid condi-
tions to prevent cracking. Specific addi-
tive to sludge ratios were not provided by
the processor, but the percent of dry
sludge solids in the final product was
21 percent. A solidified sample of the
product is shown in Figure 1.
Process B—
This process which is also patented,
uses a soluble silicate and a cement based
additive to produce a "soil-like" mate-
rial. The consistency of the final prod-
uct can be varied to suit the product
characteristics required, but for landfill
disposal, the "soil-like" material is the_
most economical to produce. Molds 120 cm
x 9 cm deep were covered with polyethylene
during the 12-day curing period. The
cured specimens were broken into irregular
chunks, 2 to 5 cm in dimension and placed
in leaching columns without compaction.
The final product contained 40 percent dry
sludge solids. Figure 1 shows samples of
the final product which were broken to fit
into the leaching column.
Process C—
This process uses a patented urea-
formaldehyde resin formulation (with
sodium bisulfate catalyst) to produce a
solid, rubber-like material. The polymer-
ization reaction requires an acid medium
which is provided by the addition of suf-
ficient sodium bisulfate solution to pro-
vide a pH below about three. The
sludge-resin mixture polymerizes rapidly
in the 7.6 x 40 cm molds. Since the re-
action is rapid, no curing period is nec-
essary. Any weep water produced was
allowed to evaporate on the surface of
the mold. The final product contained
20 percent dry sludge solids and is shown
along with the untreated, pigment produc-
tion sludge in Figure 2.
Column Leaching Procedure—Between 8
and 12 kg of each of the treated and un-
treated sludges were loaded into three
identical leaching columns (see Table 1).
Leachates from one column of each set of
three was randomly selected for analysis
with very sensitive techniques (high pre-
cision columns); samples from the remain-
ing leaching columns were analyzed using
conventional, flame-atomic-absorption
methods (low precision columns). Statis-
tical analyses were only carried out
where parameters of the leachate from all
three columns were above the higher de-
tection limits. Analyses from the high
precision columns alone are used for
those constituents present below the de-
tection limits of the higher precision
techniques. Detection limits of the two
analytical methods are presented in
Table 2.
The leaching columns were made from
1.5 m lengths of 10.16 cm ID plexiglass
pipe as shown in Figure 3. The bottom of
each column was closed with a perforated
plate that allowed the leachate to drain
into a 1.27 cm deep collecting well. The
leaching fluid was introduced through a
tube entering the column 19 cm from the
top of the column. The tops of the col-
umn were covered with loose-fitting
plastic lids to prevent dust from enter-
ing. The flow of leachate water through
each column was regulated with a Teflon
stop-cock. A 7.5 cm layer of polypropy-
lene beads (60 mm diameter) was packed in
the column to retard the movement of par-
ticulate material into the collecting
well. Leachate from each column was col-
lected in a 4.5-f polyethylene bottle lo-
cated below the column. A view of the
columns taken during the leaching experi-
ment is shown in Figure 4.
Untreated sludges were placed di-
rectly in the columns and back-flooded
with distilled water to remove all air
bubbles. The fixed sludge samples that
were not cast into cylinders were broken
105
-------
SLUDGE NO. 1300
Figure 1. Samples of treated and untreated Nickel-Cadmium battery sludge
SLUDGE NOJ 700
Figure 2. Sample of treated and untreated pigment production sludge
106
-------
152 4cm
Figure 3. Design details of leaching column
Figure 4. View of racks holding leaching columns
107
-------
TABLE 2. METHOD OF ANALYSIS AND LIMITS OF DETECTION
OF LOW AND HIGH LEVEL ANALYSIS OF CATIONS
Cation
As
Be
Ca
Cd
Cr
Cu
Fe
Pb
Mg
Mn
Ni
Se
Zn
Flame AA
Detection Limit (mg/jf)
2.0
0.05
0.2
0.05
0.5
0.2
0.3
1.0
0.02
0.1
0.3
1.0
*
Heated Graphite Atomizer AA
Detection Limit (mg/f )
0.005
0.005
*
0.003
0.003
0.003
0.003
0.002
*
0.002
0.005
0.005
0.014
* Not used.
into conveniently handled pieces and
placed in the column. The fixed sludges
that had been cast into cylinders were
placed in the columns and polypropylene
pellets were packed around the cylinders.
The variations in column loading proce-
dures were made to accommodate those pro-
cessors (A and C), who required their
samples to be tested as a monolithic mass
rather than as fragmented or ground mate-
rial. In each case, the processor felt
that the material, as packed in the col-
umns, reflected the way the material would
be placed in a landfill.
The leaching liquid used in this test
was deionized, distilled water saturated
with carbon dioxide. The pH of this solu-
tion is approximately 4.5. The liquid
flow through the columns was maintained in
a range which approximated the range of
percolation through the covering material
on a landfill in the eastern U. S.; 7 to
70 ml/day for these columns. This is
equivalent to an hydraulic conductivity of
10 to 10 cm/sec. In some cases the
natural flow rate was less than this
amount, and in other cases the flow rate
varied unpredictably. If the natural
flow rate was lower than desired, the
natural rate was allowed to regulate the
flow. If the flow was too rapid, the
stop-cock was used to control the flow.
If the flow rate was variable, leachate
was collected until 4.5 ( had passed
through the column; then, the flow of
leachate was shut off.
Samples of leachate from the columns
were preserved by adding nitric acid to
lower the pH to 2.0. A detailed discus-
sion of analytical procedures is pre-
sented in a previously published
report (2). All chemicals used were
reagent grade. An analytical quality
control program including blank correc-
tions, spiked samples and extreme stan-
dards was incorporated into all phases of
the experiment.
RESULTS
The complete data set from this ex-
periment consists of 16 leachate samples
from each of the 15 leaching columns.
Each leachate sample, if of sufficient
volume, was analyzed for up to 24 param-
eters including 13 metals, 6 anions, and
5 other parameters such as conductivity
and pH. These data are summarized and
discussed in three different ways. The
highest single value in any single sample
for each parameter is given as the
108
-------
worst-case estimate of the maximum level
of that constituent that might he encoun-
tered. A second viewpoint is presented by
the "overall" level of the parameter which
is calculated as if all leachate samples
from each column were pooled and a single
analysis performed (total ml leached/total
litres of leachate). This value gives an
estimate of the long-term effects of the
leachate on surrounding environment.
Neither of these calculations take
into account differences in the amount of
actual dry sludge solids loaded into each
column; and as can be seen in Table 1, a
considerable variation in the amount of
dry sludge solids in each column occurred.
Some solidification processes had a small
amount of dry sludge solids in their final
treated product. Therefore, a third pre-
sentation, the total amount and percent of
each constituent leached from each column,
is calculated. This calculation is made
only for the high-precision columns to
minimize the effects of the samples which
were below detection limits (values below
detection limits were set equal to zero).
Levels of constituents found in
nitric acid digests of the two sludges
used in this study are given in Table 3.
Both are low calcium sludge (less than
2 percent) and have relatively high levels
of several toxic metals. The low percent
recovery for the Ni-Cad battery sludge
constituents may reflect high levels of
sodium and nitrate, parameters which were
not included in the analysis.
Leaching Patterns
A typical difference between the
leaching patterns of treated and untreated
sludge columns is illustrated in Figure 5
for leachate conductivity. This is a rel-
ative measure of the ionic strength of the
leachate. Leachates from the treated
sludge columns show high initial values
which rapidly drop to lower, fairly con-
stant levels as the experiment progresses.
The highest conductivity is found in the
initial leachate sample from the column
containing Ni-Cad battery sludge treated
by Process B. Leachate conductivities
from Process A treated sludge show the
same general patterns but at lower con-
ductivity levels. The leachate conduc-
tivity from the untreated samples of the
same sludge begin and remain high and
relatively constant over the same time
span. The same overall pattern is also
evident in leachates from the pigment
production sludge. These high initial
values of conductivity may represent the
soluble materials from a surface wash of
the solidified/stabilized sludges, while
the untreated sludges continue to release
high concentrations of the soluble con-
stituents throughout the experiment.
The relatively low permeability of
the untreated sludges restricted the flow
of leachate through these columns so that
many of the leachate samples were too
small for a complete analysis to be made.
The treated sludges, even though having
similar low permeability, were surrounded
by polypropylene beads which allowed free
flow of liquid through them, larger leach-
ate samples, and consequently, more com-
plete analyses. This also may have only
allowed a surface wash of the treated
sludges where the leachate completely
permeated the untreated sludges.
The curve for magnesium concentra-
tion in leachate from pigment production
sludge treated by Process B (Figure 6) is
typical of the leaching pattern of most
constituents from the treated sludge
column—a high initial value followed by
a rapid decline to a rather constant
lower value. However, other constituents
in the same leachate samples from the
same column may have very different leach-
ing patterns. Such a case is illustrated
by the concentration curve for calcium
concentrations in the same leachate
samples also shown in Figure 6. Calcium
concentrations, in this case, start out
very low, slowly increase for several
months, and then slowly decline. The low
initial calcium values are probably
caused by common ion effects with mag-
nesium and monovalent cations, both of
which are present in high concentrations
in the early leachate sample. An even
more dramatic increase in calcium concen-
tration is seen in the curve for leachate
samples from the Ni-Cad battery sludge
treated by Process B. A similar cause
may be responsible in that the calcium
increase takes place during the rapid de-
crease in leachate conductivity seen for
this same sludge in Figure 5. Evidently,
the high levels of monovalent cations re-
duce the solubility of calcium salt in
early leachate samples. This kind of
interaction between leaching rates of dif-
ferent constituents presents a major
109
-------
30000r
_c
E
h-
o
SLUDGE 300 UNTREATED
O SLUDGE 300 TREATED BY PROCESS A
o SLUDGE 300 TREATED BY PROCESS B
180 200
DAYS LEACHING
Figure 5. Plot of conductivity of leachate from treated and untreated Ni-Cad battery
sludge
8 (1620)
70C
600
500
5! 400
0>
s
300
200-
100
LEGEND
V CA FOR 300 B
O CA FOR 700 C
• Mg FOR 700C
0 30
60
90
120
150
180
210
240 270 300
DAYS LEACHING
Figure 6. Concentration of calcium and magnesium by days of column leaching for Ni-Cad
battery sludge treated by process B (300B) and for pigment production sludge treated by
process C (700C)
110
-------
TABLE 3. CONSTITUENTS IN NITRIC ACID DIGEST OF NI-CAD
BATTERY AND PIGMENT PRODUCTION SLUDGES
Constituent
Mg/Kg in Dry Sludge Solids
Ni-Cad Battery
Sludge
Pigment Production
Sludge
As
Be
Ca
Cd
Cr
Cu
Fe
Hg
Mg
Mn
Ni
Pb
Zn
Cl
so4
Si
% recovery
BDL*
12.9
24,300
4,640
151
322
6,260
4.49
1,070
170
169,000
102
1,440
112,000
BDL
19,390
34
170
BDL
19,000
6,350
86,500
6,950
68,500
57.0
27,300
2,590
1,410
115,000
3,280
491,000
160,000
167,000
114
* BDL = below detection limits.
problem to the interpretation of short-
term elutriate testing methods.
Overall and Highest Leach Concentrations
The overall and highest levels of
those constituents which are generally
present in amounts above the high preci-
sion analyses detection limits are shown
for the Ni-Cad battery sludge in Table 4
and for pigment production sludge in
Table 5. Results from the three columns
used for each of the untreated and treated
sludge samples and the mean of the three
values are given. Those treatment means
which were found to be significantly
different from the untreated sludge column
(control) means at the 95 percent confi-
dence level are indicated by an asterisk
in the table. The statistical test used
was a one-way analysis of variance with
five degrees of freedom.
Both the highest and average volumes
of the leachate samples reflect the low
permeability of the untreated control
column; treated sludge columns produce
leachate samples that averaged 2.5 to 3
times larger than those control columns.
The largest leachate samples collected
from the untreated control columns were
well below the 4.5 I maximum sample size.
The pH of the leachate from the control
columns and Process B columns were not
significantly different, but those sam-
ples from Process A columns had signif-
icantly lower pH values (by 2.5 to 3
units).
The level of constituents lost to
the leachate from the untreated Ni-Cad
battery sludge columns is very low con-
sidering the nature of the sludge. The
overall and highest concentrations of
constituents in the leachate exceeds
drinking water standards (4) only for
mercury, manganese, and selenium. The
overall and highest concentrations of
111
-------
TABLE 4. OVERALL AND HIGHEST VALUES OF SELECTED PARAMETERS FROM ALL
LEACHING COLUMNS CONTAINING NICKEL-CADMIUM BATTERY SLUDGE
Untreated Sludge Columns
Leachate
Parameter
Vol
pH
COND
As
Ca
Cd
Cu
Mg
Hi
Cl
so4
Average
Largest
Overall
Highest
Overall
Highest
Overall
Highest
Overall
Highest
Overall
Highest
Overall
Highest
Overall
Highest
Overall
Highest
Average
Highest
Overall
Highest
Low
Precision
1.19
3.75
10.6
12.4
14600
27000
0.068
0.300
1.4
2.0
B
0.09
0.048
0.056
B
0.60
0.7
2.3
N
250
N
N
Low
Precision
1.21
2.98
11.4
12.4
19400
29000
0.009-
0.028
1.4
2.0
B
0.05
0.036
0.049
B
0.60
0.4
0.6
N
170
N
N
Hi
Precision
0.86
2.55
11.5
12.4
19700
24800
0.005
0.012
0.95
1.70
0.010
0.050
0.013
0.080
0.610
1.20
0.314
2.90
35.1
704
109
245
Mean
1.08
3.09
11.2
12.4
17900
26900
0.027
0.110
1.2
1.9
B
0.06
0.032
0.062
B
0.80
0.47
1.93
N
375
N
N
Process A Columns
Low
Precision
2.65
4.5
8.2
9.6
6560
10000
0.009
0.032
375
1300
0.09
0.15
N
0.060
2.25
7.70
B
0.4
110
400
1400
1675
Low
Precision
2.49
4.5
8.4
9.6
3000
9000
0.015
0.100
443
1490
0.19
0.57
N
0.058
2.35
7.10
0.3
1.3
139
562
1365
1780
Hi
Precision
2.62
4.5
8.8
10.1
2780
9000
0.010
0.024
302
1280
0.043
0.226
0.012
0.076
1.13
4.70
0.024
0.111
8.79
32.0
499
1300
Mean
2.58*
4.5*
8.5*
9.8*
4110*
9300*
0.011
0.052
373*
1360*
0.107
0.315
N
0.064
1.91
6.50*
0.162
0.60
86
331
1090
1535
Process B Columns
Low
Precision
3.14
4.5
10.5
12.9
5940
28900
0.005
0.012
248
464
B
B
N
0.024
B
0.10
B
B
47
102
N
9.0
Low
Precision
3.02
4.5
10.9
12.7
7180
28000
N
0.008
317
579
B
B
N
0.050
B
0.10
B
0.35
119
459
11
21
Hi
Precision
3.24
4.5
11.0
12.6
7450
30000
BDL
0.003
225
618
0.002
0.008
0.004
0.025
0.022
0.100
0.012
0.130
13.3
87.0
4.69
15.0
Mean
3.13*
4.5*
10.8
12.7
6860*
20600
N
0.008
263*
554*
B
B
N
0.033*
B
0.10*
B
0.24
60
216
7.8
15
NOTE: All values in mg/f except volume (in t ) and conductivity (in mhos/cm) and pH; B = below detection limits; N
*Mean is significantly different from untreated control columns at 5% confidence level (see text).
insufficient data.
-------
TABLE 5. OVERALL AND HIGHEST VALUES OF SELECTED PARAMETERS FROM
ALL LEACHING COLUMNS CONTAINING PIGMENT PRODUCTION SLUDGE
Untreated Sludge Columns
Leachate
Parameter
Vol(O
PH
COND
Ca
Cd
Cr
Cu
Mg
Mn
Ni
Pb
Zn
Cl
SO
4
COD
Overall
Highest
Overall
Highest
Overall
Highest
Overall
Highest
Overall
Highest
Overall
Highest
Overall
Highest
Overall
Highest
Overall
Highest
Overall
Highest
Overall
Highest
Overall
Highest
Overall
Highest
Overall
Highest
Overall
Highest
Low
Precision
0.88
2.75
8.0
8.4
8,510
9,600
517
731
0.75
1.50
0.3
1.2
0.5
0.7
638
1,012
1.6
2.5
0.3
0.8
1.2
2.3
B
B
N
260
6,930
11,000
N
N
Low
Precision
0.66
2.55
7.6
8.5
8,380
10,000
515
759
0.82
1.29
B
1.0
0.8
1.4
736
937
1.9
2.9
0.4
0.8
1.3
2.0
B
B
N
260
4,430
7,000
N
N
Hi
Precision
0.92
3.02
8.0
8.4
8,340
10,200
492
599
0.914
1.430
0. 116
0.999
0.528
0.700
720
1,050
1.83
2.80
0.310
1.210
0.872
2.300
0.050
0.092
211
266
8,330
14,200
N
N
Mean
0.82
2.77
7.9
8.5
8,400
9,900
516
696
0.83
1.41
0.2
1.1
0.6
0.9
698
1,000
1.8
2.7
0.3
0.9
1.3
2.2
B
B
N
262
6,563
10,733
N
N
Low
Precision
2.17
3.08
6.1
7.2
N
N
348
499
31.6
57.9
0.5
1.3
0.6
1.6
731
3,390
26.1
38.1
1.6
3.4
2.1
3.5
12
40
31
51
4,660
12,890
5,610
6,822
Process
Low
Precision
2.52
4.5
4.9
5.8
4,000
8,500
235
398
36.5
54.0
21.6
80.0
4.6
9.5
267
545
26.9
85.0
2.4
4.0
1.9
3.8
11
20
20
51
3,380
4,225
4,225
5,234
C Columns
Hi
Precision
2.21
3.37
5.5
6.4
5,530
14,700
309
450
40.10
96.9
5.47
20.0
2.62
16.00
394
1,620
22.5
77.0
1.96
5.70
1.038
4.200
7.370
30.00
17.
51.
3,670
15,500
N
6,900
Mean
2.33*
3.75
5.5*
6.5*
4,760
11,600
297.*
449.*
36.0*
69.6*
9.19
33.7
2.6
9.0
464
1,851
25.2*
66.7*
2.0*
4.3*
1.7
3.8*
10.
30.
23
51*
3,903
10,871
4,918
6,319
NOTE: All values in mg/f except volume (I), conductivity (mhos/cm), and pH;
B = below detection limits; N = insufficient data.
* Mean is significantly different from untreated control at 95% confidence level (see text).
-------
calcium, chloride, and sulfate are also
extremely low. Other constituents for
which only the highest concentrations ex-
ceeded drinking water standards were
arsenic and cadmium. The high conduc-
tivity seen for these leachates must be
indicative of large amounts of sodium and
nitrate which is present in the sludge.
The leachate from untreated pigment pro-
duction sludge, on the other hand, exceeds
the drinking water standards for both
overall and highest concentrations of cad-
mium, chromium, mercury, manganese, lead,
and sulfate (Table 5). These leachate
samples also showed high levels of cal-
cium, magnesium and sulfate. The rela-
tively low loss rates from the untreated
sludges probably reflects their low perme-
ability and low pH. The bulk of the heavy
metal sulfides and hydroxides have very
low solubilities at high pH's.
The solidification/stabilization pro-
cess used to treat the Ni-Cad battery
sludge significantly effected only the
levels of calcium, the highest values of
magnesium, and in one case, the highest
concentration of copper found (Table 4).
The levels of calcium in the leachate from
the processed sludges are over 200 times
higher than calcium levels in the un-
treated sludge leachates. These high cal-
cium levels must be caused by calcium
added in the treatment reagents (e.g.,
lime and cement) since the original Ni-Cad
sludge is relatively low in calcium con-
tent. Appreciable amounts of magnesium
must be present in flyash or "lime addi-
tive" used by Process A; but absent in the
reagents used by Process B since leachates
from Process A-treated sludges had signif-
icantly higher, and from Process B treated
sludges significantly lower, magnesium
levels. Process B also significantly
lowered the amount of copper in leachate
from this sludge.
Other trends found in the leaching
data from treated and untreated Ni-Cad
battery sludges are higher losses of cad-
mium and sulfate from Process A, and lower
levels of nickel and sulfate from Pro-
cess B-treated sludges. Several differ-
ences also occurred in leachate losses of
other parameters which are present in
concentrations below the low precision
detection limits. The overall and high-
est leachate concentrations of these con-
stituents as determined for the high
precision columns only are given in
Table 6. Leachate from the column con-
taining Ni-Cad battery sludge treated by
Process A had higher overall and highest
concentrations than did leachate in the
control columns for beryllium, chromium,
manganese, selenium, zinc, and cyanide.
Of these, only chromium and zinc were
above the drinking water standards. Pro-
cess B produced leachates with higher
chromium, lead, and cyanide, with only
lead exceeding drinking water standards.
Treatment of pigment production
sludge by Process C (Table 5) signifi-
cantly increased the rate of loss of
cadmium (by 40-50 times), manganese (by
14-25 times), and nickel (by 5-7 times).
Although not significant statistically,
higher rates of loss from the treated
sludge columns also were found for chrom-
ium (by 30-45 times), copper (by 4-10
times), and zinc (by at least 100 times).
Significantly lower amounts of calcium
(about 1/2) and chloride (about 1/5) were
also lost from Process C-treated pigment
production sludges. Sludge treated by
Process C had higher overall concentra-
tions of seven, and larger "highest"
values of nine, of the parameters
tested.
Percent of Major Constituents Leached
Calculation of the percent of each
constituent leached from the columns over
the length of the study gives an estimate
of constituent loss rates which takes
into account the different amounts of dry
sludge solids loaded into each column.
If the evaluation was based only on con-
centrations, those processes which added
the largest amount of treatment reagents,
and therefore the least sludge solids,
would be given an advantage over pro-
cesses with high proportions of dry
sludge solids in their final product.
The weight of dry sludge solids loaded
into each column was given in Table 1.
Percent of each constituent leached is
calculated by summing the product of the
volume of each sample (I) times the con-
centration of that constituent in the
sample (mg/f) to give the total amount
leached (mg); this value is then divided
by the amount of the constituent in the
dry sludge solids which were loaded into
the column (using sludge composition data
in Table 3) and multiplied by 100 to
give a percentage value. The results of
these calculations are given for Ni-Cad
114
-------
TABLE 6. OVERALL AND HIGHEST VALUES OF CONSTITUENTS OF LOW CONCENTRATION
FROM HIGH PRECISION COLUMNS CONTAINING NI-CAD BATTERY SLUDGE
Untreated Sludge
Columns
Constituent
Be
Cr
Hg
Mn
Pb
Se
Zn
Cn
Overall
N*
0.001
0.0110
0.006
0.003
0.006
0.006
0.002
Highest
0.0002
0.004
0.0570
0.122
0.019
0.020
0.050
0.010
Process A Column
Overall
0.0040
0.018
N
0.061
0.003
0.010
0.134
0.015
Highest
0.0470
0.099
0.0003
0.922
0.026
0.073
1.640
0.110
Process
Overall
N
0.010
N
0.002
0.052
N
0.009
0.024
B Column
Highest
0.0004
0.047
0.0038
0.016
0.899
0.003
0.070
0.140
* N = no or insufficient data.
battery sludge in Table 7 and pigment pro-
duction sludge in Table 8.
Less than 1 percent of all constitu-
ents for which analyses were made were
lost from the columns containing untreated
Ni-Cad battery sludge. A very small frac-
tion of the dry sludge solids were removed
from the column over the period of the
study. Both treatment processes are seen
to have similar but generally higher per-
cent loss values reflecting the lower
amount of sludge solids in the columns.
Notable exceptions are calcium from both
processed-sludge columns and magnesium
from the Process A columns. These higher
losses probably reflect the addition of
calcium (and perhaps magnesium) in the
treatment reagents which are intentionally
not included in these calculations. The
"treatment" actually increased the total
amount of calcium in the waste. The col-
umn containing the Ni-Cad sludge treated
by Process A lost a higher percentage of
all constituents analyzed except arsenic
and nickel (which were below detection
limits), and mercury. The column contain-
ing sludge processes by B lost higher pro-
portions of all constituents analyzed
except arsenic, cadmium, mercury, mag-
nesium, nickel, selenium, and sulfate.
The total loss of selected constit-
uents from the column containing untreated
pigment production sludge was also less
than 1 percent except for calcium, mag-
nesium and sulfate (Table 8). About
7.5 percent of both calcium and mag-
nesium, and about 15 percent of the sul-
fate in the sludge were lost from these
columns. The higher rate of loss of
these constituents from the pigment pro-
duction sludge may reflect their higher
solubility in the lower pH of this leach-
ate (8.0) than in the untreated Ni-Cad
battery sludge leachate (11.5), and the
lower amount interfering, monovalent ions
present in the sludge. Process C was un-
successful in lowering the percent of any
of the cations lost to leaching waters in
this experiment—cadmium, chromium, cop-
per, manganese, nickel, and zinc being
lost over 30 times the proportion lost by
the control sludges. Between two and
three times the percentage of sulfate was
lost from the treated sludge, but only
about half the percentage of chloride.
DISCUSSION AND CONCLUSIONS
The industrial waste sludges used in
this study have the potential of being
serious threats to surface and ground-
waters. Although both sludges lost heavy
metals to leaching waters in appreciable
amounts, the rate of metal release from
the sludge appears to be limited by their
solubility at the basic pH of the sludges
and the low permeability of the untreated
115
-------
TABLE 7. TOTAL AMOUNT AND PERCENT OF EACH CONSTITUENT LEACHED FROM
HIGH PRECISION COLUMNS CONTAINING NI-CAD BATTERY SLUDGE
Untreated Sludge
Column
Constituent
As
Be
Ca
Cd
Cr
Cu
Hg
Mg
Mn
Ni
Pb
Se
Total mg
Leached
2.1
<0.001
159.
0.10
0.013
0.137
0.12
6.3
0.061
3.20
0.029
0.061
%
Leached
B*
<0.001
0.11
<0.001
0.001
0.007
0.42
0.096
0.006
<0.001
0.005
N*
Process A Column
Total mg
Leached
0.28
0.16
11,800
1.7
0.724
0.48
0.002
44.0
2.4
0.93
0.118
0.372
%
Leached
B
0.69
27.3
0.02
0.27
0.083
0.025
2.3
0.79
<0.001
0.066
N
Process B Column
Total mg
Leached
0.008
0.001
10,700
0.089
0.47
0.197
0.021
1.1
0.091
0.56
2.5
0.008
%
Leached
B
0.002
13.5
<0.001
0.096
0.019
0.143
0.031
0.016
<0.001
0.74
N
Zn
Cl
CN
SO
Avg. pH
0.060 0.001
363
0.02
42,200
33
0.051
N
N
N
5.2
351
0.58
19,500
10
0.20
0.17
N
N
N
11.5
8.9
0.43
633
1.15
223
94
11.0
0.009
0.16
N
N
N
B = constituent below detection limits in sludge analysis;
N = constituent not determined in sludge analysis.
sludges. Over the 2 year leaching period
of this study, less than 1 percent of the
heavy metals was found in the leachate.
The loss of toxic metal apparently would
continue at these levels for many years.
The Ni-Cad battery sludge also appears to
be releasing high concentrations of sol-
uble constituents which also continued
over the length of the experiment.
The solidification/stabilization
techniques used in this experiment
had variable effects upon pollutant
containment for the sludges. Process B,
the soluble silicate and cement additive
procedure, reduced the amount of metals in
the leachate to the greatest degree.
Sludge treated by this process, however,
had the highest initial conductivity
levels indicating a large immediate loss
of soluble constituents. Process A, using
flyash and lime, on the other hand had
very little effect upon the rate of loss
of metals but did appreciably lower the
initial and long-term loss of soluble con-
stituents as indicated by lower conductiv-
ity levels. The loss of soluble constitu-
ents parallels the amount of surface area
exposed—the Process B product being
broken into chunks with relatively higher
surface area (and conductivity values)
than the monolithic Process A product.
Process B, however, even with a greater
surface leaching area, must have a much
greater containment ability due to its
higher pH and silicate additives.
Process C, the urea-formaldehyde
resin product, appears to be counter pro-
ductive to the containment of the metal
species in the sludge. The acidified so-
lution necessary for the polymerization
116
-------
TABLE 8. TOTAL AMOUNT AND PERCENT OF SELECTED CONSTITUENTS LEACHED
FROM HIGH PRECISION COLUMNS CONTAINING PIGMENT PRODUCTION SLUDGE
Untreated Sludge Column
Constituent
As
Be
Ca
Cd
Cr
Cu
Hg
Mg
Mn
Ni
Pb
Se
Total rag
Leached
0.080
0.070
6,800
12.6
1.5
7.3
0.062
9,960
25.4
4.3
12.6
0.15
%
Leached
0.010
B*
7.4
0.041
<0.001
0.022
0.023
7.5
0.20
0.063
0.002
N*
Process C Column
Total mg
Leached
0.049
1.2
7,514
974
133.
63.
0.17
9,600
546
47.5
25.2
N
%
Leached
0.020
B
27.4
10.6
0.12
0.64
0.21
24.4
14.6
2.3
0.015
N
Zn
Cl
so4
Avg. pH
0.70
2,900
115,000
8.0
0.004
0.12
14.9
179.
415
89,100
5.6
3.8
0.059
38.6
B = constituent below detection limits in sludge analysis;
N = constituent not determined in sludge analysis.
reaction apparently serves to solubilize
many of the contained metals increasing
their rate of loss from the treated
sludge. The smaller surface area of the
treated sludge does, however, lower the
long-term loss of soluble constituents as
measured by conductivity.
Calculation of the percent of each
constituent lost from the column shows
that a very low proportion of most con-
stituents were lost from the leaching col-
umns over the length of this study.
Sludges treated by Processes A and B lost
a greater proportion of all constituents
measured (except mercury and nickel for
both and magnesium for Process B). Pro-
cess C treated sludges lost much greater
percentages of all constituents measured
except chloride. Thus, when calculated on
a sludge basis, the small benefits of
solidification/stabilization largely dis-
appeared in this experiment.
Assessment of contaminate loss from
treated and untreated sludges using small-
column, long-term, submerged leaching pro-
cedures confirms the pollutant potential
of these sludges. The treatment processes
used in this study did not appear to in-
crease the containment of toxic metals.
Further studies using other experimental
methods should be undertaken.
ACKNOWLEDGEMENT
This study was part of a major re-
search program on the chemical fixation
technology, which is now being conducted
by the U. S. Army Engineer Waterways Ex-
periment Station and funded by the Envi-
ronmental Protection Agency, Municipal
Environmental Research Laboratory, Solid
and Hazardous Waste Research Division,
Cincinnati, Ohio under Interagency Agree-
ment, EPA-IAG-D4-0569. Robert E. Landreth
is the EPA Program Manager for this re-
search area.
117
-------
References
1. Bartos, M. J., Jr., and M. R. Palermo.
1977. Physical and Engineering Prop-
erties of Hazardous Industrial Wastes
and Sludges. EPA-600/2-77-139. U. S.
Environmental Protection Agency, Cin-
cinnati, Ohio. 89 pp.
2. Maloch, J. L., D. E. Averett, and
M. J. Bartos, Jr. 1976. Pollutant
Potential of Raw and Chemically Fixed
Hazardous Industrial Waste and Flue
Gas Desulfurization Sludges. EPA-600/
2-76-182. U. S. Environmental Pro-
tection Agency, Cincinnati, Ohio.
117 PP.
3. Malone, P. C., R. B. Mercer, and D. W.
Thompson. 1978. The Effectiveness of
Fixation Techniques in Preventing the
Loss of Contaminants from Electroplat-
ing Wastes. In: First Annual
Conference on Advanced Pollution Con-
trol for the Metal Finishing Indus-
try. EPA-600/8-78-010. Industrial
Environmental Research Laboratory,
U. S. Environmental Protection
Agency, Cincinnati, Ohio.
pp 130-143.
4. National Academy of Sciences, National
Academy of Engineering, Water Quality
Criteria, 1972. A Report of the Com-
mitte on Water Quality Criteria, EPA-
R-73-033, U. S. Environmental Protec-
tion Agency, Washington, D. C. 1973.
594 pp.
5. Thompson, D. W., P. G. Malone, and
L. W. Jones. 1979. Survey of Avail-
able Stabilization Technology. In:
Toxic and Hazardous Waste Disposal,
Vol 1. R. B. Pojasek, Ed. Ann Arbor
Science Publishers, Inc., Ann Arbor,
Mich. pp 9-22.
118
-------
ORGANIC LEACHATE EFFECTS ON THE PERMEABILITY OF CLAY LINERS
David Anderson and K. W. Brown
Texas Agricultural Experiment Station
Texas A&M University
Soil & Crop Sciences Department
College Station, TX 77843
ABSTRACT
Permeability remains the primary criterion for evaluating the suitability of clay
soils for the lining of hazardous waste landfills. An evaluation of the physical classes
of hazardous waste and the leachates they release indicates that clay liners could be
exposed to a wide range of organic fluids. Procedures are outlined for rapidly testing
the effect of organic chemicals on the permeability of clay liners. Large permeability
increases were observed in two smectite clays after the soils were permeated with neutral-
polar, neutral-nonpolar, and basic organic fluids.
INTRODUCTION
In the last decade, several hazardous
waste disposal facilities have been
identified as the source of contaminants
found in nearby groundwater supplies. As
a result, many states have adopted regula-
tions that require hazardous waste land-
fills and surface impoundments to have
clay liners with permeabilities not to
exceed 10~' cm/sec.
Test procedures for determining the
permeability of clay liners usually are
performed using a standard permeant (0.01 N
CaCl£ or CaSO,) which may bear little
resemblance to the leachate released from
disposed liquids or sludges. An evaluation
of the physical classes of industrial waste
and the leachates they generate has
indicated that clay liners could be exposed
to a wide range of organic chemicals.
Procedures are given below for
evaluating the permeability of compacted
clay soils to leachates released from
disposed liquids and sludges. The proce-
dures are followed by a discussion of the
results obtained from the testing of two
smectite clay soils permeated with organic
fluids.
A. PHYSICAL CLASSES OF WASTES
Wastes found in industrial disposal
facilities fall into four physical classes;
aqueous inorganic, aqueous organic,organic,
and sludges (EPA, 197A)(Figure 1).
Cheremisonoff et_ al. (1979) estimated
that 90%, by weight, of industrial hazard-
ous wastes are produced as liquids. These
liquids are further estimated to contain
solutes in the ratio of 40% inorganic to
60% organic.
"Aqueous-inorganic" is the class of
wastes in which water is the solvent
(dominant fluid), and the solutes are
mostly inorganic. Examples of these
solutes are inorganic salts, metals dis-
solved in inorganic acids, and basic
materials such as caustic soda. Examples
of wastes in this category are brines,
electroplating waste, metal etching wastes,
caustic rinse solutions, and spent metal
catalysts.
FIGURE 1: PHYSICAL CLASSES OF WASTE
* Aqueous
* Aqueous
* Organic
* Sludges
- Inorganic
- Organic
119
-------
"Aqueous-organic"is the class of
wastes in which water is the solvent and
the solutes are predominantly organic.
Examples of these solutes are organic
chemicals that are polar or charged, as
is inferred by their water-solubility.
Examples of this class of wastes, are wood
preserving wastes, water base dye wastes,
pesticide container rinse water, and
ethylene glycol production wastes.
"Organic" is that class of waste in
which an organic fluid is the solvent and
the solutes are the other organic chemi-
cals dissolved in the organic solvent.
Examples of this class of wastes are oil
base paint waste, pesticide manufacturing
wastes, spent motor oil, and spent clean-
ing solvents.
"Sludges" represent the fourth class
of wastes. They are generated when a
waste stream is dewatered, filtered, or
treated for solvent recovery. Sludges
are characterized by high solids content
such as that found in settled matter or
filter cakes and consists largely of clay
minerals, silt, precipitates, fine solids,
and high molecular weight hydrocarbons.
Examples of this waste are API separator
sludge, storage tank bottoms, treatment
plant sludge, or any filterable solid
from any production or pollution control
process.
Both economic and pollution control
pressures continue to mandate solvent
recovery and other reductions in the
volume of liquid waste streams. These
factors have and will continue to make
sludges the fastest growing class of
wastes. After placement of sludges in a
waste disposal facility, leachates migrate
out of the sludge due to gravitational
forces, overburden pressures, and hydrau-
lic gradients. These leachates are
similar in physical form to the other
three classes of waste given in Figure 1.
B. LEACHATES GENERATED BY INDUSTRIAL
WASTE
To determine the effect of a
specific waste on the permeability of a
specific clay liner, two unique leachates
must be investigated. The unique
leachates are the flowable constituents of
the waste and the flowables generated
from percolating water leaching through
the waste (Figure 2).
Flowable constituents of a waste,
hereafter referred to as the primary
leachate, include both fluids in the waste,
the solvent, and all components dissolved
in these fluids (solutes). The nature of
the primary leachate solvent phase depends
on the composition of the waste and may
be predominantly water or any organic
fluid discarded by industries. Flowables
generated from percolating water are
composed of water (the solvent) and all
components dissolved in this water
(solutes). This water based mixture is
hereafter referred to as the secondary
leachate.
The predominant fluid or solvent
phase of a leachate may be water or any
organic fluid (Figure 3). Corresponding
solutes in a leachate are any chemicals
that dissolve in the solvent. As in the
waste classification, both primary and
secondary leachates are divided into a
solvent, the dominant fluid component, and
a solute, components dissolved in the
solvent (Figure 4 and 5). While the
solutes in a leachate can affect the
permeability of a clay liner, the solvent
phase of a leachate will usually exert a
dominating influence on the liner's
permeability.
Essentially all available literature
on the behavior of leachates in soil
relate to systems where water is the
solvent phase (Goring et _al., 1972).
Water is viewed as the carrier fluid and
the organic chemicals are present in only
trace quantities. However, solvent
phases of a landfill leachate may be water
or any of the organic fluids present in
industrial waste (Figure 6). Organic
fluids placed in industrial waste disposal
facilities cover the spectrum of chemical
species such as acidic, basic, neutral-
polar and neutral-nonpolar compounds.
Organic acids are any organic fluid
that has acid functional groups. The
group has the potential to be very
reactive with and mobile in clay liners.
Organic bases are any organic fluid
capable of accepting a proton to become
an ionized cation. Since these fluids
are positively charged, they will adsorb
strongly to clay surfaces. By adsorbing
120
-------
HYDRAULIC AND BEARING
PRESSURE
FLUID PORTION OF
THE WASTE
1
PRIMARY
LEACHATE
INFILTRATION OF
OUTSIDE WATER
1
WASTE WATER SOLUBLE PORTION
OF THE WASTE
1
CLAY SECONDARY
LINER LEACHATE
UNDERLYING STRATA
FIGURE 2. SOURCES OF LEACHATE
THAT MAY COME IN CONTACT WITH CLAY LINERS
LEACHATE
SOLVENT PHASE
SOLUTE PHASE
ORGANIC FLUIDS H20 ORGANIC INORGANIC
FIGURE 3. COMPOSITION OF THE LEACHATE OF
A WASTE
to the clays, these fluids have the
potential for causing volume changes in
clays by changing interlayer spacings, as
well as by dissolving certain constituents
of the clay minerals.
Neutral-nonpolar compounds are those
organic fluids that have no charge and a
small, if any, dipole moment. This group
of fluids is further divided into aliphatic
and aromatic hydrocarbons. With no charge
and little dipole moment, these chemicals
have the potential for moving through
clay liners rapidly and eroding the pores
through which they pass, thus increasing
permeability. These chemicals may also
increase permeability of a clay liner by
displacing water from the clay liner
which may in turn cause shrinkage.
Neutral-polar compounds have no
charge but exhibit stronger dipole moments
than the "neutral-nonpolar" organics.
This group of fluids can be further
divided to represent its functional group
types: alcohols, aldehydes, glycols,
alkyl halides and ketones. Their effect
on permeability may be related to their
effect on the interlayer spacing of a
clay and on the surface tension of the
pore water.
Water is a unique solvent in several
ways. It has a strong dipole moment and
a high dielectric constant. Water may
cause a clay liner to shrink, swell, heave,
crack or pipe. Water may also act to
increase the hydraulic gradient that moves
fluids in soil. While water might not be
the solvent in the primary leachate of a
waste, it is usually the solvent in the
secondary leachate.
C. TEST METHOD FOR DETERMINING THE EFFECTS
OF WASTE LEACHATE ON THE PERMEABILITY
OF COMPACTED CLAY SOILS (CONSTANT
ELEVATED PRESSURE METHOD)
To assess the suitability of com-
pacted clay soils for the lining of waste
disposal facilities, the primary labora-
tory measurement made is saturated
conductivity or permeability. Permeability
measurements should be made on a core of
the clay soil that has been compacted at
its optimum moisture content to achieve
the maximum density possible for a given
compactive effort. The procedure for
determining soil moisture-density relations
is given in the 1978 Annual Book of ASTM
Standards Part 19, Test Method Number D
698-70.
121
-------
PRIMARY
LEACHATE
FLUID PHASE
\
DISSOLVED PHASE
PREDOMINANTLY
ORGANIC
PREDOMINANTLY
AQUEOUS
INORGANIC
COMPONENTS
\
ORGANIC
COMPONENTS
FIGURE 4. PRIMARY LEACHATE GENERATED AT DISPOSAL SITES
SECONDARY
LEACHATE
FLUID PHASE
AQUEOUS
DISSOLVED PHASE
INORGANIC
COMPONENTS
ORGANIC
COMPONENTS
FIGURE 5. SECONDARY LEACHATES GENERATED AT DISPOSAL SITES
SOLVENT PHASE
ACIDIC
BASIC
NEUTRAL-
POLAR
NEUTRAL-
NONPGLAR
FIGURE 6. SOLVENT PHASE OF THE LEACHATE OF A WASTE
122
-------
Clays compacted at optimum moisture
content may have permeability values as
low as 10~" cm/sec. In addition, it may
be necessary to pass one or more pore
volumes of a standard leachate, .01N CaSO,/,,
to obtain a stable permeability value for
the clay. For these reasons, it may be
necessary to use high pressure to increase
the hydraulic gradient and reduce the
time length of the test (Bennett, 1966).
Trapped air is a common cause for
artificially low permeability values
(Christiansen et al., 1946). An increase
of pressure reduces air trapped in the
core by increasing the weight of gas
that will dissolve in water flowing
through the core (Jones, 1960). The
higher pressures will also reduce the
volume of the remaining air pockets.
For a given waste-clay liner combina-
tion, there are two fluids that may alter
the liner's permeability, the fluids
present in the waste (primary leachate)
and the fluids generated by water percola-
ting through the waste(secondary leachate).
The primary leachate of a waste may be
obtained for permeability testing by
decanting the liquids present in the
waste or by collecting the filtrate from
vacuum filtration of the waste. The
filter cake obtained from filtration of
the waste should then be shaken with an
equal weight of water and refiltered.
The resulting second filtrate can then be
used as the secondary leachate of the
waste. Enough waste should be processed
in this manner to yield at least four
pore volumes of primary and secondary
leachate to use in the permeability tests.
Two liters of each leachate should be
enough to run duplicate cores with each
leachate in the test procedures discussed
below. Since these leachates may cover
the spectrum of organic fluids, the
permeameter used for testing these fluids
must be capable of safely operating with
industrial solvents, volatile compounds,
corrosive acids and strong bases.
Tests on soils of low permeability
must be carefully performed if they are
to be accurate. Leaks, volatile losses,
or channel flow along the interface of
the permeameter and soil will greatly
affect permeability values (Bowles, 1968)•
To avoid channel formation, the clay
is seated at low pressure. By letting
10 mm of standard leachate (.01 N
stand on the core for 24 hours, an
effective seat is obtained for the top
few millimeters of the clay core. This
thin layer will prevent bulk flow and
the rest of the core should adequately
seal at elevated pressures.
In order to facilitate ease of
duplication of the test apparatus, the
permeameter is based on readily available
and easily modified components. Figure 7
illustrates the modified compaction
permeameter which is based on the standard
compaction permeameter that is available
through most soil testing supply houses.
All components are in common with the
standard permeameter except for the
enlarged fluid chamber, extended studs
and high pressure fittings.
For use with fluids other than water,
all gaskets should be teflon. To avoid
leakage around the gaskets, all metal
surfaces against which the gaskets are
seated should be wiped clean of grit. All
components should withstand continuous
operation at pressures up to 4 atmospheres.
To limit the volume for diffusive
mixing of leachate samples after they have
passed through the clay core, the fluid
outlet should be fitted with an adapter
to 1/8 inch inside diameter teflon tubing.
Using translucent teflon at the per-
meameter 's outlet provides a convenient
window with which to monitor the explu-
sion of entrapped air. Standard leachate
should be passed through the permeameter
until there are no air bubbles visible
in the outlet tubing. If soil piping
occurs, eluted soil clays will be visible
either clinging to the inside walls of
the outlet tubing or as a suspension in
the collected flow samples.
Volatile losses may occur during
sample delivery from the outlet tubing
to the sample bottles in the fraction
collector. To limit these volatile
losses, the top of each sample bottle is
fitted with a long stem funnel and the
entire fraction collector is placed in
an air tight cooled compartment.
The entire test apparatus should
then be fitted into a vented hood.
(Figure 8) . This extra precaution is
insurance against worker injury in case
123
-------
— Pressure Intake
— Pressure Release
ZZI- Top Plate
\
1
Fluid
Chamber
\
X
x
X
\
b t
\
\
\
Soil
Chamber
X
\
X
x
&
Gasket
— Extended Stud
-Base Plate
*— Porous stone insert
— Va inch Teflon tubing
Outlet to fraction collector
FIGURE 7. SCHEMATIC OF THE COMPACTION
PERMEAMETER
of a gasket blow out. If multiple permea-
meters are to be used, each should have
its own pressure cut-off valve to prevent
the complete shut down of a test to remove
a single permeameter.
With several permeameters producing
leachate, it may be desirable to have an
automatic leachate fraction collector.
This is especially useful with long term
tests.
Equipment requirements for the simul-
taneous testing of two wastes ( 2 wastes -a
2 leachates (primary and secondary) x 2
replications + control permeameters = 10
permeameters) are as follows:
1. Soil crusher (C-2 Laboratory
Crusher)1
2. Soil grinder (Hewitt Soil Grinder)'
3. 2 mm sieve (CB-810 Brass Sieves)1
4. 10 moisture cans
(LT-30 Tin Sample Boxes)
5. Balance capable of weighing 20
kg (L-500 Heavy Duty Balance)
6. 105° C drying oven to determine
water content of soil samples
7. 10 compaction molds (CN-405
Standard Compaction Mold)
8. Compaction hammer (CN-4230 Mec-
hanical Compactor)1
9. Steel straight edge
10. 10 permeameter bases and top
plates (K-611 Permeameter
Adapter)1
11. A source of compressed air with
a water trap, regulator and
pressure meter
12. An air manifold to permit simul-
taneous testing of 10 core samples
13. A fraction collector with auto-
matic timer for collection of
samples over time from 10 per-
meameters (Brinkmann Linear II)
14. An air tight, cooled chamber to
limit volatile loss of samples
during and after sampling
15. A vented hood to hold the compac-
tion permeameters and chamber
containing the fraction collector
Procedures are given below for the
10 permeameter test system using the
equipment described above. Approximately
50 kg of the air dried clay soil will be
required for each test incorporating 10
p ermeame t e r s.
1. Obtain at least 100 kg of the
clay soil (CS) to be tested.
Break the CS down to golf ball
size clods and lay them out to
air dry.
Soil Test Inc.
2Robert B. Hewitt Welding and Repair
(815-269-2030). Danforth, Illinois 60930.
124
-------
COMPACTION
PERMEAMETER
AIR TIGHT COOLED CHAMBER
FIGURE 8. SCHEMATIC OF THE COMPACTION PERMEAMETER TEST APPARATUS IN A VENTED HOOD
2. Grind the air dried CS and pass it
through a 2 mm sieve.
3. Mix the sieved CS thoroughly and divi-
de it.into two lots. Each lot of CS
should be placed in air tight con-
tainers and stored at room tempera-
ture until their time of use. Each
lot should be enough CS to prepare
10 compaction molds. (Fifty kilo-
grams will provide for some spillage
losses assuming a mold volume of
about 1,000 cm3).
A. Use one lot of the CS to determine
the moisture density relations of
the CS by following the ASTM Method
D-698-70*. This will determine the
CS's optimum moisture content to
obtain the maximum density at the
compactive effort to be used.
5. Use the second lot of CS to prepare
10 compaction molds at optimum
moisture content.
6, Fit a valve on top of the permeameter
top plate with pressure fittings and
connect it to a source of air pres-
sure via copper tubing. Place a water
trap, pressure regulator and pressure
gauge in line between the air pressure
source and the permeameter. The water
trap should be located between the
pressure source and regulator to
prevent build up of debris on the
membrane in the regulator. The
pressure gauge should be located
between the regulator and a pressure
manifold to the 10 permeameters so
that the hydraulic head being exerted
on the clay cores may be monitored.
7. Pass at least one pore volume of the
standard leachate (.01 N CaSO, or
through the clay cores. After
This method can be found in the 1978
Annual Book of ASTM Standards Part 19.
8.
9.
the permeability values are stable and
below 10~~7 cm/sec , release the
pressure, disassemble the permeameter
and examine the core.
If the clay core has shrunk, it is
unsuitable as a clay liner.
If the clay has expanded out of the
mold, remove the excess with a
125
-------
straight edge while trying not to
smear the clay's surface. Reweigh
the core to determine its density
and then remount it on the permeame-
ter.
10. If the clay has not changed its
volume, remount it on the permeameter.
11. Repressurize the permeameter and
pass standard leachate until the
permeability value stabilizes below
10~7 cm/sec.
12. Remove the remaining standard leach-
ate from eight of the ten fluid
chambers and replace it on four of
the cores with duplicates of the
primary leachate from the two wastes
to be tested. The other four cores
should be tested using the filtrate
from well shaken 1 to 1 mixtures
of water and the wastes to be
tested. Thus, one test run (10 cores)
is sufficient to maintain two
control cores while evaluating the
permeability of duplicate cores
treated with the primary and secon-
dary leachate of two different
wastes.
13. If after passage of one pore volume
of the various leachates, the
permeability values of the cores
are still below 10~? cm/sec
disassemble the permeameter and
reexamine the cores.
14. If the clay core has shrunk, it is
unsuitable as a clay liner for that
waste.
15. If the clay has not changed volume
or has expanded, remount it on the
permeameter.
16. Repressurize the permeameter and
pass volume of the standard leachate.
If its permeability has climbed
above 10"? cm/sec , the clay is
not suitable for containing the
waste. If the permeability values
measured on a waste's primary and
secondary leachate have consistently
stayed below 10~7 cm/sec , proceed
to step 17.
17. Examine the translucent teflon out-
let tube for signs of soil particle
migration out of the core. If there
is evidence of soil migration, pass
at least one more pore volume to
observe if this internal erosion of
the core continues. If it does
continue after the two pore volumes
of standard leachate have passed, the
clay is unsuitable for that waste.
If the soil migration stops, at least
one pore volume of the standard
leachate should be passed to assure
that the core stabilization is
permanent and then proceed to step 18.
18. If there are no signs of soil migra-
tion, depressurize the system and
extrude the clay cores from their
molds to examine them carefully for
signs of cracking, internal erosion,
soil piping, clay dissolution,
structural changes, or any other
difference from the control cores
(those having received only standard
leachate).
If there are no signs the cores
have deteriorated, the clay should be
suitable for lining the disposal
facility to contain that waste.
Permeability (K) of the clay soils
can be calculated using the following
formula:
K =
V L
At (L + H)
K = permeability(cm/sec)
V = volume of flow (cm3) in time (t)
A = cross-sectional area of flow
(cm2)
t = time (sec.)
L = length of soil core (cm)
H = pressure (cm of 1^0)
The sign is negative to indicate the
direction of flow.
D. PERMEABILITY OF TWO SMECTITE CLAY
SOILS PERMEATED WITH ORGANIC FLUIDS
Two smectite clay soils that are
typical of those used to construct clay
liners were selected for use in the
evaluation of the developed test proce-
dures. Table 1 lists physical and
chemical properties for those soils.
Smectite clay minerals are composed of
expandable lattices and therefore exhibit
a large capacity for shrinking and swelling.
126
-------
TABLE 1. PHYSICAL AND CHEMICAL PROPERTIES OF TWO SMECTITIC CLAY SOILS
Soil Name
Clay
%
O.M.
CEC Exc. Ca Exc. Na
meq/lOOg
Lufkin
48
24
18
Houston
Black
56
36
saturated
<1
Of the two clay soils, Houston Black
has the highest amount of clay sized par-
ticles, organic matter, and cation
exchange capacity. Houston Black, which
developed over a limestone bedrock, is
approximately 33% by weight calcium
carbonate. Consequently its cation ex-
change sites are essentially saturated
with calcium.
Lufkin clay has a substantially
larger percentage of its cation exchange
sites occupied by sodium ions than does
the Houston Black clay. Consequently, of
the two clay soils studied, the Lufkin
clay has the larger potential for shrink-
ing and swelling.
Table 2 lists physical and chemical
properties of the permeants used in this
study. Fluid types were selected from
each of the four classes of organic
leachates described in Section B and
generated by the "organic" and "sludge"
waste classes described in Section A.
All test procedures were performed as
described in Section C. In the place of
actual leachates, reagent grade organic
fluids were used as the test permeants.
Results of these permeability tests are
given in Figures 9-11.
Acetic acid (Figure 9) caused
decreases in the permeability with both
Houston Black and Lufkin. However, there
was a significant amount of soil piping
occuring in the two acid permeated cores
as was evidenced by the presence of soil
particles in the leachate. Both acid
dissolution and soil piping have been
discussed by these authors as possible
failure mechanisms for clay liners
(Brown _et _al. , 1980).
Aniline (Figure 9) treated cores both
showed substantial permeability increases.
Permeability increased nearly 100 fold for
Lufkin clay over that obtained with water.
Permeability of the Houston Black core
increased approximately 10 fold.
Acetone (Figure 10) treated cores
showed an initial decrease followed by
large increases in permeability. Lufkin
and Houston Black clays exhibited 1,000
fold and 50 fold increases in permeability
respectively.
Ethylene glycol (Figure 10) cores
showed 100 fold and 4 fold increases in
permeability with the Lufkin and Houston
Black clays respectively. Interestingly,
the ethylene glycol treated Houston Black
core showed a steady decrease in perme-
ability following the initial permeability
increase.
Both xylene and heptane (Figure 11)
treated cores showed substantial perme-
ability increases. Xylene treated cores
gave the larger permeability increases
with greater than 1,000 fold and 100 fold
increases for the Lufkin and Houston Black
clays, respectively.
While the initial permeability values
obtained with water(0.01N CaSO,) were
lower for the Lufkin clay, this clay usual-
ly ended up with the largest permeability
values after treatment with the organic
fluids. This may have been due to the
larger shrink-swell capacity of the Lufkin
clay.
Significant increases in permeability
were obtained with basic, neutral-polar
and neutral-nonpolar organic fluids over
those values obtained with water. This
indicates the need to test the permeability
of clay liners with the actual leachates
generated by a disposed waste especially
where organic fluids may be present in the
waste.
127
-------
TABLE 2 PHYSICAL AND CHEMICAL PROPERTIES OF THE PERMEANTS TESTED
PERMEANTS _ ,Wf'?j
Solubility
(gm/1)
Organic Fluids Name @ 20° C
Acid, Carboxy- Acetic Acid c°
lie
Base, Aromatic Aniline 34.0
Amine
Neutral-Polar, Acetone «,
Acetone
Neutral-Polar, Ethylene ™
Glycol Glycol
Neutral-Non Polar Heptane 0.003
Aliphatic
Neutral-Non Polar Xylene 0.20
Aromatic
Water
Dielectric
Constant
II 20°C (1 25
6.1
6.9
38.66
1.0
2.5
80.4
20.7
78.5
Dipole Temp. Range Viscosity Density Molecular
Moment of the Liquid (Centipoise) (g/cm3) Weight
(debyes) State (°C) 1? @ @
°C Freezing Boiling 20°C 25°C 20°C
1.74 17 118 1.28 1.16 1.05 60
1.55 -6 184 4.40 3.71 1.02 93
2.90 -95 56 0.33 0.32 0.79 58
2.28 -13 198 21.0 17.3 1.11 62
0.00 -91 98 0.41 0.39 0.68 100
0.40 -47 137 0.81 0.62 0.87 106
1.83 0 100 1.0 .98 18
10-5-
"
io-(
—
G 10"'
1
e
|
1
'
ID"8
10-'
.,„ -*,»- Or,.nlc F1.1J.
_„-**
/
/
/
/
^y
ACID
^^^^^^p'"'" • y ACIT:C ACie
! BLACK
^S^ "^ ; — —
\i
\i
\i
! \* .
j » h 1 1 1-
_- — • ""
BASE
ANILINE
,
_,.'''
WATER
^^^^^
|
lO-V
'
10~°
^
|
£ i»";
3
1
10"
/
(
/
/
a 1 m 1
acer ~ l~ Organl. Tiuids j
1
1
1
1
1
1
1
1
1
1
1
\
1 /
1 /
//
I"-"
If
//
// 1 / ~___
/'/ I/
/ • jlf
1 : /]
. ^ •_!/ 1 H^.r^-PoUr
j^^t-.. // / *"""' Es;''""
/* / Bl.ck —
' yv / / L"'k"
/%..; /
^*A /
\ /
\/
^ . ~i 1 1
" (t ', l' 0 Li 20 2.5 J
1.0 1.5 20 2.5 3. I
PORE VOLUMES
PUItF VOLUMES
JTCURE 9. PERMEABILITY OF HOUSTON BLACK
AND LUFKIN CLAY SOILS TO
ACETIC ACID AND ANILINE.
FIGURE 10. PERMEABILITY OF HOUSTON BLACK
AND LUFKIN CLAY SOILS TO ACE-
TONE AND ETHYLENE GLYCOL.
128
-------
Houston
Black
Lufkin
NEUTRAL-NOWOLAR
H.pt«n
X7lBn.
WATER
—
1 1 I 1 1 1
0,5 0.0 0.5 1.0 l.S 2.0 2.5 3.0
PORE VOLUMES
FIGURE 11. PERMEABILITY OF HOUSTON BLACK
AND LUFKIN CLAY SOILS TO
HEPTANE AND XYLENE.
CONCLUSIONS
An evaluation of the physical classes
of hazardous waste indicated that many
contain organic fluids. These fluids may
in turn leach from the disposed waste and
enter the clay liner. Permeability tests
using two clay soils typical of those
used to construct clay liners showed
significant increases in permeability
with several of the organic fluids
studied. These results point to the need
to test the permeability of clay liners
with the actual leachates generated by
the disposed wastes.
Ongoing tests are evaluating the
effect organic fluids have on the perme-
ability of kaolinitic and illitic clay
soils. In addition, Phase II tests are
planned for evaluating the effects of
organic and aqueous based fluid mixtures
and waste leachates on the permeability
of kaolinitic, illitic, and commercially
prepared clay liner products (Table 3).
REFERENCES
1. Bennett, J. P. 1966. "Permeability
of soils at elevated permeant
pressures,," Masters Thesis at
Colorado State University, Fort
Collins, Colorado.
2. Bowles, J.E. 1968."Coefficient of
permeability-Falling Head Method"
pp. 105-110. J[n Eno_. Prop, of _S_oil_s
and their -Measurement. McGraw Hill
Book Co., N.Y., N.Y.
3. Brown, K. W. and David Anderson.
1980. "Effect of organic chemicals
on clay liner permeability." In;
Proceedings of 6th Annual Research
Symposium. EPA-600/9-80-010. U.S.
Environmental Protection Agency,
Cincinnati, Ohio. pp. 123-134. PB
80-175086.
4. Cheremisonoff, N. P., P. N.
Cheremisonoff, F. Ellerbusch, and
A. J. Perna. 1979. Industrial and
Hazardous Wastes Impoundments. Ann
Arbor Sci. Publ. Inc., Ann Arbor,
Michigan.
5. Christiansen, J. E., M. Fireman, and
L. E. Allison. 1946. "Displacement
of Soil-Air by C02 for Permeability
Tests." Soil Science. Vol. 61 pp.
355-365.
6. EPA. 1974. Report to Congress:
Disposal of Hazardous Waste. USEPA
//SW 115. Washington, D.C.
7. Goring, C.A.I, and John W. Hamaker.
1972. Organic Chemicals in the Soil
Environment, Vol. I & II. Marcel
Dekker, Inc., N.Y., N.Y.
8. Jones, C. W. 1960. "Permeability
tests with the permeant water under
pressure". Earth Laboratory Report
//EM-559. Division of Engineering
Laboratories, Commissioner's Office,
Denver, Colorado.
9. Sherard, J. L. and R. S. Decker.
1977. "Summary-evaluation of the
symposium on dispersive clays."
Dispersive Clays, Related Piping and
Erosion in Geotechnical Projects.
ASTM STP 623. J. L. Sherard and
R. S. Decker (eds.).
129
-------
TABLE 3. FLUIDS AND CLAY LINERS TO BE EVALUATED IN THE PHASE II LABORATORY
STUDY
CLAY LINERS
COMMERCIALLY PREPARED
NATIVE CLAY SOILS
1. Sodium Saturated Smectite
2. Organic Polymer Treated
Smectite
1. Mixed Cation Smectite
2. Calcium Saturated Smectite
3. Mixed Cation Illite
4. Mixed Cation Kaolinite
ORGANIC BASED
MIXTURES
FLUIDS
AQUEOUS BASED
MIXTURES
LIQUID INDUSTRIAL
WASTES
Various mixtures of
the organic fluids
used in Phase I
(eg. 50% xylene in
acetone, etc.)
Various mixtures of
miscible organic
fluids in a water base
(eg. 10:10:80 acetone:
ethylene glycol:water,
etc.).
1. Spent solvent
wastes
2. Off specification
liquid products
130
-------
MEMBRANE LINER SYSTEMS FOR HAZARDOUS WASTE LANDFILLS
Robert C. Gunkel
U. S. Army Engineer Waterways Experiment Station
Vicksburg, Mississippi 39180
ABSTRACT
A series of tests were conducted to determine an economical method of protecting flexible
membranes from damage during the construction of landfills. Subgrade soils were selected
that were considered to be representative of those which would be found in areas where
landfills are constructed. Four membranes were tested. The membranes were placed on top
of the subgrades and covered with various thicknesses of a sand material. The test items
were trafficked using three different vehicles that were representative of the type of
loadings that could be applied during landfill construction. Performance of the mem-
branes was judged by their resistance to puncture and wear.
Introduction
Background
Today, many industrial wastes may be
highly toxic to the environment if their
disposal is not properly controlled. A
common method for disposal is the use of
landfills, but improperly designed land-
fills could result in contamination of
ground and surface water by toxic waste
material. This condition exists because
of various physical, chemical, and biolog-
ical processes occurring when water or
fluids percolate through the wastes, re-
sulting in a leachate which contaminates
the soil and ground water. The placement
of impervious flexible membrane over the
subgrade material in hazardous waste land-
fills could be one solution to controlling
the leachate. One problem associated with
this approach is potential damage to the
membrane caused by earthmoving equipment
during construction. Not only does the
heavy equipment tend to damage the liner
material, but puncture of the flexible
membrane by underlying angular rock and
soil particles in the natural soil also
presents a source of damage. To study the
problem, the Environmental Protection
Agency (EPA) requested the U. S. Army
Engineer Waterways Experiment Station
(WES) to conduct an investigation to de-
termine the requirements needed to pro-
tect flexible membrane from damage.
Objective
The objective of this study was to
determine an economical method to protect
flexible membranes from damage during the
construction of landfills.
Scope
The performance of flexible membranes
used as liners in landfills was investi-
gated. A test section containing 12 test
items was constructed and subjected to
three types of vehicle traffic (tracked,
pneumatic-tired, and cleated). During
this study four flexible membranes, six
selected subgrades, three thicknesses of
a sand protective layer, and bedding mate-
rials were investigated. The materials
used, the construction of the test sec-
tion, the tests that were conducted, and
the results including some conclusions
and recommendations are described below.
Construction and Materials
Construction
A test section was constructed under
shelter at the WES. The test section was
16 ft wide, 240 ft long, and consisted of
12 test items, each 20 ft long and 16 ft
wide. A plan and profile of the test sec-
tion are shown in Figures 1 and 2, respec-
tively. Construction was started by
131
-------
ITEM 1
240'
10 11 12
SHOULDER<<
I I
LANE - 1 D - 7 BULLDOZER TRACK
LANE - 2 PNEUMATIC - TIRED TRACTOR
16'
LANE-3 CLEATED LANDFILL COMPACTOR
LANE - 4 SAMPLE MEMBRANE DURING CONSTRUCTION
I i I i I i i
Figure 1. Plan "iew of test section
"SHOULDER
ITEM
SHOULDER
PROTECTIVE LAYER OF SAND - 6-18"
_l_
SAND
r 7
CRUSHED-^
GRAVEL GRAVELLY
CLAYEY
SAND
Figure 2. Profile of test section
excavating an area of the subgrade floor
of the shelter, where the test section had
been staked out, to a depth of 6 in. and
a width of 16 ft (Fxgure 3). The last
40 ft, at the north end of the test sec-
tion, was excavated to a depth of 12 in.
to accommodate a 6-in. layer of coarse
gravel that was overlaid with 6 in. of
sandy silt. This fine grained sandy
silt was used to act as a bedding
Figure 3. Test section prior to placement
of selected subgrades
material to protect the flexible mem-
branes from puncture during traffic tests.
The remainder of the test section was
backfilled with the selected subgrades.
The subgrades were compacted with
pneumatic-tired and vibratory rollers.
Figure 4 shows the test section with the
selected subgrades in place. After the
six subgrades had been placed, each of
the 12 test items in the test section
was covered with flexible membranes.
Shoulders were constructed on both sides
of the test sections using material that
had been excavated previously from the
shelter floor. The height of the shoul-
ders depended on the thickness of the
protective layers of sand that were
placed over the membrane liners for each
test program. For the first program,
the shoulders were 6 in. in height so as
to contain that depth of sand. The sand
was dumped between the shoulders at each
end of the test section. Then a bulldozer
pushed the sand toward the center of the
test section. Care was taken during
placement operations to maintain at least
6 in. of sand between the dozer's tracks
and the flexible membranes. Figure 5
shows the completed test section with a
protective sand layer of 6 in.
132
-------
It should be noted that the sand in items
5 and 6 was the same type of sand that was
used for the protective layers. The sand
was a local (Vicksburg, Mississippi) sand
usually used normally as the fine aggre-
gate in concrete.
Membranes
Figure 4. Overall view of test section
with subgrades in place
Figure 5. Completed test section with
6 in. of protective sand
Subgrade Material
Six subgrade materials were selected
and used for the 12 items of the test sec-
tion. These materials were classified ac-
cording to the Unified Soil Classification
System (USCS) as follows:
Item No.
1-2
3-4
5-6
7-8
9-10
11-12
Classification
Crushed gravel (GP)
Gravelly clayey sand (SP-SC)
Sand (SP)
Gravelly sand (SP)
Coarse gravel (GP)
Sandy silt (ML)
The four flexible membranes and one
fabric used as a bedding material were as
follows:
Thick-
Desig- ness
nation* mils Type
M-l 20 Elasticized polyolefin
(3110)
M-2 20 Polyvinyl chloride (PVC)
M-3 30 Chlorinated polyethylene
(CPE)
M-4 36 Reinforced chlorosulfated
polyethylene (CSPE)
F-l 30 Nonwoven polypropylene
and nylon
* The membranes and fabric material are
referred to hereafter by the designation
symbol assigned above.
Traffic Vehicles
The vehicles used to apply traffic to
the various test programs were: A D-7
bulldozer equipped with 22-in.-wide
tracks, weighing approximately 44,000 Ib,
and having a contact pressure of 9 psi; a
pneumatic-tired tractor weighing 37,190 Ib
and equipped with two 29.5x29, 22-ply
tires (each tire had a contact area of
574 sq in. which produced a contact pres-
sure of 32 psi); and a model 816 landfill
compactor weighing 40,900 Ib equipped with
four cleated steel wheels having a contact
pressure of 18 psi. The traffic vehicles
are shown in Figures 6-8.
Tests, Data Collection, and Failure
Criteria
Traffic Pattern
Traffic tests were conducted on each
test item to simulate actual heavy equip-
ment operations during the construction of
landfills. Traffic was applied with both
the "tracted dozer" and "cleated landfill
compactor" in the same manner. These ve-
hicles were operated in one direction
133
-------
Figure 6. D-7 bulldozer
Figure 7. Pneumatic-tired tractor
Figure 8. Cleated landfill compactor
134
-------
until they had traveled the entire length
of the section where they were stopped
and then returned in the same track or
wheel path in reverse. The pneumatic-
tired tractor was operated in the same
direction and in the same wheel path as
it had traveled in the preceding pass.
One pass of the dozer and pneumatic-tired
tractor resulted in one coverage within
their respective traffic lanes while one
pass of the landfill compactor resulted
in two coverages within the traffic lane.
Soil Data
Except for the crushed and coarse
gravel material, laboratory compaction
tests and unsoaked CBR's were performed on
the selected subgrades. Field tests to
determine moisture content, density, and
CBR value on the in-place material of the
test section were also performed. The
results of the soil data obtained from
these tests are presented in Table 1.
Membrane Evaluation
In evaluating flexible membrane per-
formance, only the after-traffic condition
was considered. After 10 passes of the
traffic vehicles, a trench was excavated
across each traffic lane in all 12 items.
A sample of the membrane was removed from
each lane in each item, marked for identi-
fication, and inspected. After patching
the membrane and replacing the protective
layer of sand in the trenches, traffic was
continued on the items in which the mem-
branes showed only a few or no punctures.
After 30 passes, traffic was stopped and a
final inspection was made.
Failure Criteria
Each sample of membrane was placed
over a light table and inspected for punc-
tures. A 5-sq-ft area, within the wheel
path, was marked on the membranes and the
number of punctures in this area was re-
corded. A membrane was considered
"failed" if any punctures were noted.
Traffic Testing
Test Program 1
In test program 1, two types of mem-
branes (M-l and M-2) were placed over the
six subgrade materials. The M-l and M-2
membranes were placed on the odd- and
even-numbered items, respectively. A
6-in.-thick sand cover was placed over
the membranes to act as a protective
layer. Traffic was applied to the test
section with the D-7 tracked bulldozer
and the pneumatic-tired tractor. After
10 passes, traffic was stopped and
trenches were dug across the traffic lanes
in each of the 10 items. Samples of the
membrane from each traffic lane and each
item were removed for inspection. Fig-
ures 9 and 10 show a typical inspection
trench and the condition of the membrane
sample after removal. After patching the
void in the membrane and replacing the
protective layer of sand in the trenches,
traffic was continued on items 5-6 and
11-12 which showed only a few or no punc-
tures in the membranes. After 30 passes
were completed, traffic was again stopped
and a final inspection was made. The
final results for test program 1 are
graphically displayed in Figure 11.
TABLE 1. SOILS DATA
Item
No.
Field tests
Laboratory tests
Soil Dry density Moisture Max density* Optimum Unso'aked
classification pcf content CBR pcf moisture CBR
3-4 Gravelly clayey
sand
9.8
7.0
131.8
7.4
20.4
5-6
7-8
11-12
Sand
Gravelly sand
Sandy silt
104. 1
118.6
98.5
8.1
6 0
18.7
4,
8.
7.
.0
.3
.0
105,
126.
105.
.6
.6
.5
14.
8.
14.
.0
.0
8
30
21.
24.
.5
.3
.4
* CE-12 compactive effort.
135
-------
Figure 9. Typical view of inspection
after removal of membrane
Figure 10. View of holes in membrane
after 10 passes of the pneumatic-tired
tractor
30
LJ 20
m
2
6-IN PROTECTIVE SAND COVER
10 PASSES
TIRE
I
30 PASSES
_l_
I H
,
J
ITEM 1 - 2 | ITEM 3-4 | ITEM 5-6 | ITEM 7-8 | ITEM 9-1O | ITEM 11-12 | ITEM 5-6 | ITEM 11-12 |
SAND GRAVELLY COARSE SANDY SILT SAND SANDY SILT
SAND GRAVEL OVER COARSE OVER COARSE
GRAVEL GRAVEL
CRUSHED GRAVELLY
GRAVEL CLAYEY
SAND
Figure 11. Results of test program 1
136
-------
Test Program 2
Test program 2 was identical to the
first test program with the exception of
the thickness of the sand protective layer
which was increased from 6 to 18 in. In
trafficking test program 2, three types of
vehicles were used (track-tire-cleated).
The cleated vehicle was a "Model 816 Land-
fill Compactor" which was owned by Bolivar
County, Mississippi, and was operated at
one of their landfill sites. The 816
landfill compactor was leased for testing
of test program 2. After traffic, the
sampling of the membrane was accomplished
in the same manner as in test program 1.
Typical rutting that occurred in the sand
protective layer during traffic operations
is shown in Figure 12. Final results of
test program 2 are shown in Figure 13.
Test Program 3
After trafficking and obtaining the
final data in test program 2, the sand
protective layer and membranes were re-
moved and preparation of test program 3
was started. Test program 3 contained the
same selected subgrades, and 18 in. of
sand was placed as the protective layer.
However, the existing 20-ft-long items
were subdivided into two 10-ft-long sub-
Figure 12. Lane 1 (track), lane 2 (tire)
rutting after 30 passes
items which resulted in four separate test
items in each of the six types of subgrade
materials. The four test subitems per
subgrade material were overlaid with mem-
branes M-l, M-2, M-3, and M-4, respec-
tively. A nonwoven fabric material (F-l),
used as a separation barrier or a bedding
material, was placed between the M-l and
M-2 membranes and the subgrade. Test re-
sults for test program 3 after trafficking
with the D-7 bulldozer and the pneumatic-
tired tractor are presented in Figure 14.
18-IN PROTECTIVE SAND COVER
50
40
i/l
UJ
£ 30
O
2
[L
U.
O
tr
Hi 20
ID
Z
10
10 PASSES
_^^^_ TIRE
^^IMMI M-2
IWAVWI M-1
^._____ TRACK
"
r
ii
TEM 1 -
n
-------
18-IN PROTECTIVE SAND COVER
10 PASSES
nvvvxi M-3 TIRE
U":-'.:-:'^\ M-4
I^VV^M M-1/F-1
' ' ' 1 TRACK
iimiiiiniin M-S
;
ill
in ' ••
\]\KHl fir! n flfi 1 ill
i re §
ITEM 1-2 ITEM 3-4 | ITEM 5-6 | ITEM 7-8 ITEM 9-10 ITEM 11-12
30 PASSES
n (i fl
ITEM 5-6 | ITEM 11-12
CRUSHED GRAVELLY SAND GRAVELLY COARSE SANDY SILT SAND SANDY SILT
GRAVEL CLAYEY SAND GRAVEL OVER COARSE OVER COARSE
SAND GRAVEL GRAVEL
Figure 14. Results of test program 3
Test Program 4
After completion of test program 3,
6 in. of the protective sand layer and
6 in. of the shoulders were removed from
the entire test section. With the pro-
tective layer of sand now 12 in. in thick-
ness, another series of traffic tests was
performed using the same two vehicles.
During this test program, traffic was ap-
plied to the various test items by moving
the test vehicles to lanes 3 and 4 of the
test section. Results from this series of
tests are shown in Figure 15.
20
12-IN PROTECTIVE SAND COVER
10 PASSES
L
30 PASSES
TEM 1-2 ITEM 3-4 ITEM J-6 ITEM 7-8 ITEM 9-10 ITEM 11-12 ITEM 5-6 ITEM 11-12
CRUSHED
GRAVEL
GRAVELLY
CLAYEY
SAND
GRAVELLY
SAND
COARSE
GRAVEL
SANDY SILT
OVER COARSE
GRAVEL
SANO
SANDY SILT
OVER COARSE
GRAVEL
Figure 15. Results of test program 4
138
-------
Discussion of Results
The large number of punctures that
were noted in the four flexible membranes
tested was attributed to the selected sub-
grades used in the construction of the
test section. Five of the six subgrades
contained large percentages of sand and
gravel with only a small amount of fines.
The one remaining soil that was used in
the test section was classified as a sandy
silt. This fine-grained sandy silt soil
was used as a bedding material, and 6 in.
of it were placed over the coarse gravel
in items 11 and 12. The purpose of the
bedding was to act as a protection barrier
between the coarse gravel and the flexible
membrane. For comparison purposes, items
9 and 10 contained the same coarse gravel
subgrade but were not covered with the
sandy silt bedding material. Final re-
sults were quite evident that the bedding
material used in items 11 and 12 aided in
the protection of the flexible membrane by
reducing the number of punctures. During
test programs 3 and 4, another type of
bedding material was used. A nonwoven
polypropylene and nylon-type material
(F-l) was placed under membranes M-l and
M-2 during traffic testing. After final
inspection of the membranes and a compari-
son of results from test program 2, no
significant reduction in the number of
punctures was noted.
It was also observed during the in-
spection of the trafficked membrane that
most of the punctures detected occurred
form the bottom in an upward direction.
Because of these observations, it is as-
sumed that subgrades containing angular
gravel and coarse soil particles require
a bedding and/or cushioning material to
prevent punctures.
The four membranes investigated during
this study received numerous punctures
when subjected to the subgrades containing
gravel-size material. However, a consid-
erable decrease in the number of punctures
was observed when the membranes were
trafficked on the items containing the
sand and sandy silt subgrades. After
completion of some traffic operations
on these test items, no punctures
were detected in several of the
membranes.
It was also observed that the three
types of vehicle loadings (track, tire,
and cleated) used to apply traffic to the
membranes produced similar amounts of
damage.
Conclusions
Based on the results of the traffic
tests performed and the observations made
during the inspection of the flexible mem-
branes, the following conclusions are
presented:
The three traffic vehicles produced
similar amounts of damage to each
membrane.
All membranes showed less damage when
trafficked on the sand and sandy silt
subgrades than on the subgrades con-
taining gravel.
The 6 in. of bedding material placed
in items 11 and 12 reduced the number
of punctures in the membranes.
From information obtained from this
study, it was assumed that both a bed-
ding and protective cover will be nec-
essary to protect a flexible membrane
from puncture during the construction
of landfills.
The fabric material, used as a bed-
ding, did not perform satisfactorily
as it permitted punctures to the
liners.
Because of the limited data, thick-
ness requirements for the bedding
and protective soil layer could not
be determined.
139
-------
DURABILITY OF LINER MATERIALS FOR HAZARDOUS WASTE DISPOSAL FACILITIES
H. E. Haxo, Jr.
Matrecon, Inc.
Oakland, California 94623
ABSTRACT
The results of several exposures of a variety of samples of soil, admix, sprayed-on,
and polymeric membrane liners to a selected group of real hazardous wastes and three test
fluids, i.e. distilled water, 5% aqueous solution of sodium chloride, and a saturated
solution of tributyl phosphate are discussed. The exposures included the primary test of
12 liners in contact with 6 wastes for three years, immersion tests of 12 polymeric
membranes in 9 wastes for up to 2.25 years, and roof exposure of polymeric membranes and
wastes. Most of these results have been reported at previous research symposia and in
interim reports.
The effects of the exposures vary greatly with liner materials and with wastes. The
effect of minor amounts of organic constituents in a waste can have significant effects
on liner materials on prolonged exposure. The degree of swelling of polymeric membranes
in contact with a waste is a measure of the compatibility of the liner with the waste.
The greater the swelling the less the compatibility. Swelling of a membrane results in
the loss of physical properties such as tensile strength, elongation at break, and tear
strength. Major factors in the swelling of a membrane in a given waste are its solu-
bility parameter with respect to that of the waste, its degree of crosslinking and
crystal 1inity, and the presence of water soluble constituents in the membrane compound.
No single liner material in the test program appears to satisfy all of the re-
quirements for all wastes. The complexity of the various waste streams requires that
compatibility tests must be run to demonstrate the durability of the lining material
under consideration with the waste which is to be impounded. Liners based on polymers
with crystal 1 inity appear to be more resistant to water and chemicals than the other
polymeric materials.
INTRODUCTION
This research study of liners was
started in 1975 with the primary purpose
of assessing interaction of a broad range
of lining materials and a selected,
representative series of hazardous wastes.
The linings considered for this project
included all major types of man-made
linings, i.e., compacted native soils and
clays, admixed liner materials, spray-on
membranes, and flexible polymeric membrane
materials.
The primary thrust of the project was
to expose two specimens each of one
compacted native soil, three admixes, one
sprayed-on membrane, and eight polymeric
membranes to wastes under conditions which
simulated those that might be encountered
in actual service. The specimens were of
sufficient size so that, when they were
recovered from the exposure cells, tests
could be made of some of their physical
properties. One square foot surface of
each material was exposed under one foot
of depth to the selected wastes.
In addition, supplemental work was
done to measure overall characteristics of
140
-------
lining materials and to develop test
methods which might be used in the selec-
tion of specific materials for a given
waste disposal facility. Interim data
have been presented at three previous
annual research symposia (Haxo, 1976;
Haxo, 1978; Haxo, 1980a) and one interim
report has been published (Haxo et al.,
1977). At the Sixth Annual Conference in
March of 1980, the results of three years
of exposure to the wastes were reported.
These results and those relating to liners
for MSW landfills are discussed in the
Technical Resource Document, "Lining of
Waste Impoundment and Disposal Facilities"
(Matrecon, Inc.), which EPA made available
for public review in December 1980.
Though new data have been obtained
and are presented, the principal objective
of this paper is to discuss the results
that have been obtained with particular
reference to the durability of lining
materials that have been exposed to the
various wastes on pilot, bench, and
laboratory scales which simulate service
conditions.
LINING MATERIALS
The purpose of lining a waste dis-
posal site is to prevent the potentially
polluting constituents of the waste from
leaving the site and entering the ground-
water or surface water system in the
proximity of the site. The pollutants
include organic and inorganic materials,
solids, liquids, gases, and bacterio-
logical species. In their performance
liners function by two mechanisms:
a. To impede the flow (flux) of the
pollutant and pollutant carrier,
usually water, into the subsoil
and thence into the groundwater.
This requires a construction
material having low permeability.
b. To absorb or attenuate suspended
or dissolved pollutants, whether
organic or inorganic, in order to
reduce the concentrations so that
they fall within the ranges set
by the EPA for groundwater. This
absorptive or attenuative cap-
ability is dependent largely upon
the chemical composition of
the liner material and its mass.
Most liner materials function by both
mechanisms but to different degrees
depending on the type of liner material
and the waste fluid and constituents.
Polymeric membrane liners are the most
impermeable of the liner materials, but
have little capacity to absorb materials
from the waste. They can absorb a rela-
tively small amount of organic material
but, due to their small mass, their total
absorption is small. Soils can have a
large capacity to absorb materials of
different types, but they are considerably
more permeable than a polymeric membrane.
However, the greater thickness of the soil
can result in low flux through the liner.
The choice of a particular liner material
for a given site will depend upon many
factors.
We consider a liner to be a material
constructed or fabricated by man. Such a
definition includes soils and clays having
low permeability which are either brought
to a site or available on the site
and compacted or remolded to reduce
permeability and increase strength.
Liners can be classified in a variety
of ways, such as by construction method,
physical properties, permeability, com-
position, or type of service. Some
of these classifications are presented in
more detail in Table 1. In this project,
the types of liners that were studied
are:
- Admixes
- Soils and clays
- Sprayed-on membranes
- Polymeric membrane liners
These materials had been used in
water containment and conveyance and
appeared to be promising for use in the
impoundment of hazardous wastes.
The specific types of materials, with
their respective thicknesses, selected for
the exposure testing are:
Admixes:
Hydraulic asphalt concrete (2.5 in.)
Modified bentonite and sand (5.0 in.)
Soil cement, with and without surface
seal (4.0 in.)
Soil:
Compacted native fine-grain soil
(12.0 in.)
141
-------
TABLE 1. CLASSIFICATION OF LINERS FDR WASTE DISPOSAL FACILITIES
A. BY CONSTRUCTION:
- Onsite construction:
- Raw materials brought to site and liner constructed on site.
- Compacted soi1.
- Mixed on site or brought to site mixed.
- Sprayed-on liner.
- Prefabricated:
- Drop-in polymeric membrane liner.
- Partially prefabricated:
- Panels brought to site and assembled on prepared site.
B. BY STRUCTURE:
- Rigid (some with structural strength)
- Soil
- Soil cement
- Semirigid
- Asphalt concrete
- Flexible (no structural strength)
- Polymeric membranes
- Sprayed-on membranes
C. BY MATERIALS AND METHOD OF APPLICATION:
- Compacted soils and clays.
- Admixes, e.g. asphalt concrete, soil cement.
- Polymeric membranes, e.g. rubber and plastic sheetings.
- Sprayed-on membrane linings.
- Soil sealants.
- Chemisorptive liners.
Sprayed-on:
Emulsified asphalt on a nonwoven fabric
(0.3 in.)
Polymeric membranes:
Butyl rubber, fabric reinforced
(34 mils)
Chlorinated polyethylene (32 mils)
Chlorosulfonated polyethylene, fabric
reinforced (34 mils)
Elasticized polyolefin (20 mils)
Ethylene propylene rubber (50 mils)
Neoprene, fabric reinforced (32 mils)
Polyester elastomer, experimental
(8 mils)
Polyvinyl chloride (30 mils).
EXPOSURE CONDITIONS OF LINERS IN SERVICE
IN HAZARDOUS WASTE IMPOUNDMENTS
Field Service Conditions
To estimate the durability or the
service life of a product, the environ-
mental conditions in which it serves must
be known and understood. The service
conditions encountered by a liner for a
hazardous waste impoundment contrast
greatly with those encountered by a
liner in a landfill. In the latter case,
the usual environment of a liner is
anaerobic, cool, dark, and continually
moist (Haxo, 1976). Also, the waste fluid
or leachate, though variable from landfill
to landfill and even within a landfill,
will not be as variable or as highly
142
-------
concentrated in pollutants or in con-
stituents that are aggressive to liner
materials as is encountered in a waste
disposal facility that contains indus-
trial, hazardous, and toxic wastes.
Some of the conditions that could be
encountered by liners in hazardous waste
impoundments are:
- Exposure to a vast array of dif-
ferent materials in direct contact
or under soil cover.
- Exposure to weather, e.g. sunlight,
rain, wind, ozone, and heat.
- Wave action of the fluid in the
pond.
- Intermittent exposure to both waste
fluids and weather.
- Low and high temperatures.
- Burrowing and hoofed animals.
- Ground movement.
- Irregularity of the soil beneath a
1iner.
- Changing temperatures.
These conditions must be recognized
in evaluating materials in the laboratory
or on a pilot scale.
Simulating Test Conditions
In conducting this research, our
basic approach was to expose specimens of
various commercial lining materials under
conditions which simulated real service,
to use actual wastes, to measure seepage
through the specimens, and to measure
effects of exposure by following changes
in important physical properties of the
respective lining materials.
All lining materials were subjected
to two types of exposure testing:
- Bench screening tests in which
small specimens were immersed in
wastes.
In addition to these two types of
tests, selected polymeric membranes were
subjected to the following additional
tests which simulate different aspects
of actual field service:
- Immersion in wastes.
- Roof exposure on rack.
- Outdoor exposure of small membrane
lined tubs containing wastes.
- Water absorption.
- Membrane pouches containing wastes,
placed in deionized water.
Bench Scale Screening Test
All of the different types of liner
materials were subjected to a screening
test to select materials for the long-term
testing. Strips of the membranes were
immersed in the waste fluids and those
materials that swelled inordinately in a
matter of a few weeks were eliminated from
further consideration with that particular
waste. In the case of the admix liners,
small cylindrical specimens were sealed in
short sections of glass pipes and covered
with six inches of waste. If the liner
leaked in a short time, i.e. about one
month, it was eliminated from testing with
that particular waste.
Exposure of Primary Liner Specimens
The exposure cells for the primary
specimens are shown schematically in
Figures 1 and 2. Figure 1 shows the cell
for the thick admix and soil liners and
Figure 2 shows the cell for the membrane
specimens. Each membrane specimen was
prepared with a field-type seam across the
center made according to the recommended
practice of the respective supplier of the
membrane. Two specimens of each liner
material were placed in separate cells
which were loaded with portions of the
same waste. Two sets of cells were placed
in exposure. Cells were dismantled at two
times and the specimens were recovered,
analyzed, and their physical properties
measured.
Immersion Tests
- One-side exposure to
primary exposure cells.
wastes in
Concurrent with the exposure of the
primary liners in the bases of the cells,
143
-------
Waste Column
lO"xl5"»IZ''High
Flonged Steel ^^.
Spocer
Figure 1. Exposure cell for thick liners.
!
t
Waste i
'Outlet lube with
Epoxy-coated
Diaphragm
-Collection
Bag
Figure 2. Exposure cells for membrane
liners. Dimensions of the
steel tank are 10 x 15 x 13
inches in width, length, and
height.
supplemental polymeric membrane liner
specimens were hung in the wastes. The
effects of exposure were measured by
determining the increase in weight,
analyzing the exposed specimens, and
measuring selected physical properties.
Details of the procedure followed in
conducting the immersion tests are pre-
sented in the Manual (Matrecon, Inc.,
1980). Detailed results of the test
have been reported (Haxo, 1978; Haxo,
1980a; Haxo, 1981b; Matrecon, Inc., 1980).
Roof Exposure of Specimens Mounted on Rack
These specimens were six-inch squares
cut from the polymeric sheetings and
mounted at 45° on a rack facing southward
on the roof of our laboratory in Oakland,
California. On removal from the rack, the
samples were cleaned, weighed, and their
dimensions measured, after which selected
physical properties were measured.
Such a test can only be comparative
of the specimens under test. Each test
location is different. No mechanical
stress is applied to the specimens, which
lie on a flat surface. The specimens are
exposed in a static condition to ultra-
violet and heat from sunlight, and to wind
and rain.
Roof Exposure of Liners in Tubs Filled
with Wastes
In addition to the small slabs of
liners which were exposed to the weather
on the roof rack, 12 small tubs were lined
with various polymeric membranes and
filled with wastes. In this test, which
is continuing, nine different lining
materials were exposed to four different
wastes. Each test specimen was greater
than one square yard, had a field seam
through the middle, and many folds of
different severity.
This type of test simulates exposed
liners in open ponds. The liner contacts
the waste at various depths. Thus, if the
waste is multi-layered, the liner will
contact all layers. Also, the liner
simultaneously contacts the air as well as
the waste. A zone of the liner is inter-
mittently exposed to air, then waste,
etc. Another variable in this test is the
direction faced by the sheeting; the
exposure to sunlight and wind varies for
the different sides of the tub.
Water Absorption
As most wastes contain water which
most polymeric materials absorb to varying
degrees, the water absorption of a series
of liners is being measured for extended
periods of time in accordance with ASTM
D570. In this test, small, 1x3 inch
specimens are immersed in water. In
addition to water absorption, the loss of
soluble fraction can be assessed at the
end of the exposure. The test specimens
are too small to test properties; gen-
erally, one tensile measurement can be
made.
Pouch Test
In this test, small pouches are
fabricated of the membranes to be tested
and are filled with wastes or other test
144
-------
fluids, such as salt water, sealed, and
immersed in deionized water. The permea-
bilities of the membranes to water and to
pollutants are determined by observing,
respectively, the change in weights of the
bags and the measurements of pH and
electrical conductivity of the deionized
water. Due to osmosis, water should enter
the pouch, and ions and dissolved con-
stituents should leave the bag. Details
of the test procedure are presented in the
Technical Resource Document (Matrecon,
Inc., 1980). A schematic representation
of the pouch test showing the movement of
the various constituents in shown in
Figure 3.
CONDUCTIVITY!
oH
We
DEIONIZED
—WATER—
JEST BAGS
—WALL^
c:.:;::::::;~
WASTE
FLUID
.*-
WATER
SALTS/
IONS
DRGANIC
^^*-
—i— !?•==:
^
jffitWAT'ER—
*****?******•?****•!•?•:
OUTER BAG
Figure 3.
Schematic representation of the
movements of the mobile con-
stituents in the pouch test of
membrane liner materials.
The initial tests were made with
thermoplastic materials because they could
be fabricated into pouches with relative
ease by heat sealing. Some of these
pouches have now been exposed more than
1000 days.
Bags containing the wastes actually
increased in weight, indicating the flow
of water into the bags through osmosis.
The long-term tests now show that some
ionic material is diffusing through the
liners into the deionized water.
WASTES AND WASTE FLUIDS
Not only do the waste fluids contain
pollutants which could contaminate ground-
water, but these fluids can also contain
agents that are aggressive to various
liners. The principal constituent of
most wastes is water, which in itself
can be an aggressive agent to many ma-
terials. In addition, waste fluids
generally contain dissolved organic
and inorganic components that can cause
deterioration depending upon the type of
liner and the amount and type of the
agent. Organic constituents are partic-
ularly aggressive to most of the liners
that are considered for lining waste
impoundments. Even in minor amounts in
the fluid, these organic constituents can
swell or dissolve the organic types of
lining materials. Furthermore, these
agents can affect some soils. The
dissolved salts, although not generally
aggressive toward the membrane-type
liners, can be aggressive toward some
soils. Strong acids and strong bases can
affect most lining materials. The composi-
tion of the waste fluids that were used in
this study are presented Table 2.
After the primary exposure tests,
specimens of these wastes were collected
and are being re-analyzed to determine the
effects of change in composition during
exposure periods. Some of the liner
specimens were sufficiently massive, e.g.
the soil, that they might have absorbed
components from the fluid above them and
thus reduced the concentration of pol-
lutants and aggressive agents within the
waste fluids. Another factor relating
to the waste fluid that was apparent in
the work was the inhomogeneity of some of
the wastes. Several of the wastes were
single-component solutions; others were
mixtures of oils and water, each with
dissolved constituents. Consequently, in
several cases the wastes that were in the
cells were in two layers with the oily
components on the top. This inhomogeneity
resulted in different effects upon the
liners, depending upon the depth at which
the liner was tested.
In addition, to the real wastes that
were used in the exposure tests, we added
three additional tests fluids, i.e.
distilled water, a 5% solution of Na Cl in
water, and a saturated solution of tri-
butyl phosphate in water. Distilled
water and brine solutions are commonly
used as test fluids for assessing rubber
and plastic materials. In many situa-
tions, the distilled water causes more
swelling 'and loss of properties than do
solutions containing dissolved salts. The
saturated aqueous solution of tributyl
phosphate was selected as a test fluid to
145
-------
TABLE 2. COMPOSITION OF HAZARDOUS WASTES IN EXPOSURE TESTS
Type
Acidic
n
Alkaline "
ii
Lead
II
Oily
11
ii
Pesticide "
aSolid phase
of waste
HFL"
HN03, HF, HOAC"
Slopwater"
Spent caustic"
Low lead gas washing"
Gasoline washwater"
Aromatic oil"
Oil pond 104"a
Weed oil"
Weed killer"
was 11%.
PH,
water
phase
4.8
1.5
12.0
11.3
7.2
7.9
• . .
7.5
2.7
Solids, %
Total
2.48
0.77
22.43
22.07
1.52
0.32
~36
1.81
0.78
Volatile
0.9
0.12
5.09
1.61
0.53
0.17
• • •
~31
1.00
0.46
Organic
1 i quid
phase,
%
0
0
0
0
104
1.5
100
89
21
0
simulate the presence of dissolved organic
material in a waste fluid.
RESULTS OF EXPOSURE TO HAZARDOUS WASTES
Incompatibility of some of the liners
with particular wastes was shown in the
screening tests and consequently several
liner-waste combinations were deleted.
The only admix liner material that was
placed below the acid waste was the
hydraulic asphalt concrete. Neither of
the two oily wastes was placed on the
asphaltic liners; however, the lead waste,
which contained a light, oily fraction,
was placed on these liners. The per-
formances of the individual liners are
discussed below.
Compacted Fine-Grain Soil
All of the wastes, except the nitric
acid waste, were placed above the
compacted fine-grain soil liner. Seepage
below all of the liners took place. The
amount of seepage was measured and the
respective pH, conductivity, and percent
total solids were determined. The fol-
lowing observations are made with respect
to the seepages through the soil liners:
a. The rate of seepage is 10~8 to
1O'^cm•sec~1 which compares
favorably with the permeability
of the soil measured in the
laboratory permeameter. There is
some variation in the amount of
seepage collected below the liner
which may reflect permeability
differences, perhaps due to
density of the soil.
b. The fluids being collected after
more than three years of exposure
still continue to be essentially
neutral and to have high solids
content (mostly salt) and elec-
trical conductivity.
c. There is a downward trend in
solids content of the seepages
collected under the pesticide
and lead wastes, but the seepage
under the spent caustic waste
continues to be 23% solids.
One set of the soil liners was
removed and tested. The permeability of a
specimen taken from the cell containing
the soil and the aromatic oil waste was
determined using a "back-pressure" permea-
meter. The sample was collected from a
depth of seven to ten inches below the
surface of the soil, i.e. from that part
of the soil which was not penetrated by
the oil. The three consecutive values
obtained were: 1.83 x 1Q-8, 2.43 x 10'8,
146
-------
and 2.60 x 10"^ cm-sec'1. These
figures indicate the low permeability of
the soil, which had a bulk density of
1.318 g cm~3 and a saturation degree of
101%.
Analyses for trace metals were made
of the soils which were below the lead
waste, Oil Pond 104, and the aromatic oil.
The testing included determination of pH
and heavy metal content (cadmium, chro-
mium, copper, magnesium, nickel, and lead)
on samples collected at different depths
in the eel Is.
With the exception of the liner
exposed to spent caustic, the pH of the
soil liner was not significantly altered
by the wastes. The pH of these samples
was in the range of 7.0 to 7.6; the ratio,
soil rsolution, was 1:2 with 0.01N CaCl2
being the equilibration solution.
In the case of the spent caustic, the
pH values were around 9.0 for samples
collected in the first two to three
centimeters, which concurs with our
previous findings that, over the exposure
period of 30 months, the wetting front of
the wastes penetrated the soil to a depth
of only three to five centimeters.
The heavy metals distribution, as
indicated by the analysis, shows, in the
case of the lead, only a shallow contami-
nation of the soil. Similar results were
obtained on all six heavy metals in the
case of the soil below the Oil Pond 104
waste.
Admixes
Soil Cement
All of the wastes except the acid
waste were placed on the soil cement
liner. No seepage occurred through the
liner during the 30 months of exposure.
One set of the soil cement lining
materials was recovered after 625 days of
exposure to the various wastes and the
individual linings were cored and tested
for compressive strength. In all cases,
compressive strength of the exposed soil
cement was greater than that of the
unexposed material. There was some
blistering of the epoxy asphalt coating
which was applied to one-half the surface
of each specimen.
Modified Bentonite and Sand
Two types of modified bentonites were
used in admix liners in ten cells. One
type allowed somewhat less seepage than
the other. There was measurable seepage
in seven of the ten cells and one failed
allowing the waste (Oil Pond 104) to come
through the liner.
Irrespective of the type of waste
above the liner, the quality of the
seepage was not greatly different among
the samples collected. The seepages
collected below the pesticide waste on
both types of modified bentonite liners
were similar.
When the spacers containing the
bentonite-sand were sampled, it was found
that there had been considerable chan-
neling of the wastes into these liners.
There was no channeling at the walls of
the spacers, which indicates that the
epoxy seal at the top of the liner was
effective in preventing wall effects and
by-passing of the liner.
The bentonite admixes are probably
not satisfactory for the types of waste
that were used in this study. The use of
a cover on the bentonite layer to produce
an overburden would probably reduce the
channeling effect.
Hydraulic Asphalt Concrete
Liner specimens of hydraulic asphalt
concrete were placed under four non-oily
wastes.
This lining material functioned
satisfactorily under the pesticide and
spent caustic wastes, but failed beneath
the nitric acid waste. The failure arose
primarily from the failure of the aggre-
gate which contained calcium carbonate;
also, the asphalt hardened considerably.
In the case of the lead waste, the
asphalt absorbed much of the oily con-
stituents of the waste and became "mushy".
Some staining of the gravel below the
asphalt liner occurred.
Duplicate cells containing the
hydraulic asphalt concrete and the lead
waste are still functioning without
seepage.
147
-------
Sprayed-Un Membrane Liner Based on
Emulsified Asphalt and Nonwoven Fabric
This membrane was placed under only
three of the six wastes: pesticide, spent
caustic, and lead. The acid waste was
excluded because it caused severe hard-
ening of the asphalt, and the oil wastes
were excluded because of the high mu-
tual solubility of the asphalt and the
wastes.
The asphalt membrane functioned
satisfactorily with the pesticide and
spent caustic wastes; however, when the
cell containing the lead waste was dis-
mantled, the gravel below the liner was
wet and stained brown. This result
indicates that some seepage took place.
Polymeric Membrane Liners
All of the flexible polymeric mem-
branes survived the exposure testing,
except the polyester elastomer which
failed on exposure to the nitric acid
waste. This material completely lost its
elongation and cracked in the cell
at the edges of the liner. The nitric
acid waste caused several of the cells
to leak due to corrosion of the steel
walls through pinholes in the epoxy
coating, particularly at the welds between
the walls and the flanges.
The membrane liners that were removed
from the cells were photographed, in-
spected visually, and subjected to the
following tests:
- Determination
extractables.
of vol ati1es and
- Tensile properties in machine and
transverse directions.
- Hardness.
- Tear strength in machine and
transverse directions.
- Puncture resistance.
- Seam strength.
The tests of the membranes were
performed soon after their removal from
the wastes and while they were still
swollen.
Determining the volatiles and ex-
tractables of the exposed liner supplies
information with respect to the amount of
waste that is absorbed by the liner and
the amount of the original compound that
is leached out during the exposure. The
volatiles were run first by determining
the loss in weight of a two inch diameter
disk heated for two hours at 105°C. The
extractables were determined on specimens
that had been devolatil ized by exposure
to circulating air at room temperature and
heating at 105°C. The appropriate solvent
was used in order to extract the nonpoly-
meric fraction without dissolving the
polymer.
The volatiles test was also used to
verify the direction of the grain of the
membrane which was introduced during
manufacture. (On heating, polymeric
liners shrink more in the machine direc-
tion.) Some physical properties are
significantly affected by grain direction,
a factor which was not considered in the
preparation of some of the early specimens
for exposure.
The tensile properties that are of
interest in the assessment of the effects
of exposure are:
- Tensile strength at break.
- Elongation at break.
- Stresses at 100% and 200% elong-
ation.
- Tensile set.
Because of the grain introduced in
most membrane liners during manufacture
tensile properties must be measured
in both machine and transverse directions.
Tensile strength is an important
quality of a polymer and is often a good
measure of the quality of the compound of
the polymer. Tensile per sj? does not
necessarily reflect on performance when
various materials are compared. Many
compound ingredients, e.g. fillers and
oils reduce tensile; also, on aging,
tensile strength generally drops.
Elongation, which is an important
property in the functioning of many rubber
and plastic products, appears to be an
important property for liners. The loss
of elongation in service either by loss of
plasticizer or by excessive swelling could
result in breakage, since the membrane
148
-------
may not tolerate an extension beyond which
it was designed.
The stresses at 100% and 200% elonga-
tion are measures of the stiffness or
modulus of a rubber or plastic compound.
It is affected by the amount of swelling
that takes place and by crosslinking or
oxidation that might take place in the
rubber. This property is related to
hardness.
Tensile set is the stretch that is
retained in a specimen after it has been
tensile tested. The value gives an
indication of the creep that a material
might take during service to accommodate
stress.
Tear strength and puncture resistance
are two properties of membranes that are
important in their installation in the
field. Tear resistance can also be
important in exposed liners, as can
puncture strength, particularly on
prolonged exposure of liners under a load
or on a soil surface.
Seams, particularly those made in
the field, can be a source of problems
during installation and service. Speci-
mens of the recommended seams should be
subjected to exposure testing similar to
that given the membrane and preferably
under mechanical stress conditions.
Results of the Primary Exposure Tests
All of the tests discussed above were
performed on the primary liner specimens.
Of particular interest are the volatiles,
extractables, and retention of ultimate
elongation after exposure to the various
wastes. The results for the tests of
these properties are presented in Tables
3-5, respectively.
All of the materials that were
exposed absorbed volatiles, primarily
water, and, in the case of the lead waste,
some low molecular weight hydrocarbons.
The pesticide waste had a low ion concen-
tration which resulted in high absorption
of water. The lead waste contained minor
amounts of oily material which was absorb-
ed by the neoprene. The two liners that
absorbed the least amount of volatile
materials were the elasticized polyolefin
and the polyvinyl chloride membranes. The
neoprene, chlorinated polyethylene, and
chlorosulfonated polyethylene membranes
absorbed the most and sustained the
greatest changes in properties.
In most cases, the "extractables" of
the exposed liners were higher than the
extractables of the respective unexposed
liner materials. Exceptions were the
polyvinyl chloride, the chlorinated
polyethylene, and, to a minor extent, the
neoprene, all of which had been immersed
in lead waste. The plasticizer in these
TABLE 3. VOLATILES3 OF PRIMARY POLYMERIC MEMBRANE LINER SPECIMENS AFTER EXPOSURE TO SELECTED WASTES
Volatiles, %
Waste and exposure time in days
Liner
Polymer
Butyl
CPE
CSPE
ELPOe
EPDM
Neoprene
Polyester
PVC
data
Number0
57R
77
6R
36
26
43
75
59
Compound data
TypeC
VZ
TP
TP
TP
VZ
VZ
TP
TP
Extractablesd, %
6.4
9 1
3.8
5.5
18.2
13.9
2 7
35.9
Unexposed
0.29
0.00
0.29
0.15
0.50
0.45
0.26
0.26
Pesticide
1260 d
4.8
7.9
9.7
(f)
6.3
13 6
2.9
3.6
HN03
1220 d
11.5
13 2
7.2
5.3
12 09
7.4h
Spent
caust ic
1250 d
1.4
2.8
5.8
(f)
1.3
5.7
0.9
1.8
Lead
1340 d
3 5
19 2
11.4
1.5
5 3
17 5
1.7
4 4
Oil
Pond
104
1360 d
10 1
10 3
5 1
21.3
2.6
4.2
Aromat ic
oil
(f)
(f)
(f)
^Percent weight loss after 2 h at 105'C.
°Serial number of liner set by Matrecon, R = Fabric reinforced.
cType = Vulcanized (VZ) or thermoplastic (TP).
^After volatiles were removed.
eELPO = Elasticized polyolefin
^Specimens still under exposure to waste.
SExposure time = 1150 days.
"Exposure time = 509 days.
149
-------
TABLE 4. EXTRACTABLES3 OF PRIMARY POLYMERIC MEMBRANE LINER SPECIMENS AFTER EXPOSURE TO SELECTED WASTES
Extractables, %
Waste and exposure time in days
Liner
Polymer
Butyl
CPE
CSPE
ELPOd
EPDM
Neoprene
Polyester
PVC
data
Number*1
57R
77
6R
36
26
43
75
59
Compound data
Type0 Extractables3, %
VZ
TP
TP
TP
VZ
VZ
TP
TP
6
9
3.
5
18.
13.
2
35
4
1
.8
.5
.2
.9
7
.9
Pesticide
1260 d
7 6
9.4
5.4
(e)
25 2
16.1
5.8
33 4
HN03
1220 d
8.7
10.6
4.6
7.1
22. 8f
13 59
(e)
Spent
caust ic
1250 d
7.9
9.1
3 8
(e)
24.0
13 7
3.3
35 6
Lead
1.340 d
7.
7
6
8
26
12
5
22
.9
2
0
1
0
.2
.4
.5
Oil
Pond
104
1360 d
17.0
9 5
20.7
15 9
7 3
30.0
Aroma t i
01 I
(e)
(e)
(e)
aAfter volatiles were removed
^Serial number of liner set by Matrecon; R = Fabric reinforced.
cType = Vulcanized (VZ) or thermoplastic (TP).
dELPO = Elasticized polyolefin.
eSpecimens still under exposure to waste.
'Exposure time = 1150 days.
9Exposure time = 509 days.
TABLE 5. RETENTION OF ULTIMATE ELONGATION3 OF PRIMARY POLYMERIC LINER SPECIMENS AFTER EXPOSURE TO SELECTED WASTES
Ultimate elongation, % retention
Waste and exposure time in days
Liner
Polymer
Butyl
CPE
CSPE
ELPOf
EPDM
Neoprene
Polyester
PVC
data
Number^
57R
77
6R
36
26
43
75
59
Compound data
Typec
VZ
TP
TP
TP
VZ
VZ
TP
TP
Extractablesd, %
6.
9.
3
5
18.
13
2
35
,4
.1
8
.5
.2
9
.7
.9
Or i gi nal
elongation6, %
70
405
235
665
450
320
575
385
Pesticide
1260 d
77
90
88
(g)
103
84
87
93
HN03
1220 d
419
83
83
96
79n
-------
materials leached into the wastes result-
ing in a lower extractables content.
The elongation retention data pre-
sented in Table 5 show almost complete
loss of elongation by the polyester
elastomer in the nitric acid waste. The
other lining materials retained most of
their respective elongations. The CSPE
showed the greatest loss other than the
polyester elastomer,
original elongation.
retaining 68% of its
Results of Immersion Testing
Tables 6 and 7 present results of the
immersion of 12 polymeric liner materials
in 8 wastes. Five of the liners duplicate
the primary liners and 6 of the wastes are
the same as those used in the primary
exposure test. The data on the weight
increases of the liners during immersion
are presented in Table 6. Table 7 pre-
sents the data on the changes in extract-
ables of the immersed specimens during the
exposure period. The extractables are
determined on the liners after the vola-
tiles are removed and reflect the amount
of nonvolatile material that is either
gained or lost by the immersed specimens.
Most specimens absorbed oil when immersed
in Oil Pond 104, while the PVC liners lost
plasticizer to this waste.
A supplemental study was performed on
three polymers that are partially crystal-
line, i.e. high-density polyethylene,
low-density polyethylene, and polypropy-
lene (PP), and one composition based on
vulcanized CPE. None of the crystalline
sheets are specifically liner materials.
They were selected for this test because
membranes based on these polymers were
potentially suitable for lining waste
impoundments, and because they have
comparable thicknesses to most of the
membrane materials being tested. LDPE in
the 10-mil version has been used for
lining canals and water impoundments;
recent'ly, HOPE was introduced as a com-
mercial lining material for chemical
waste impoundments. The vulcanized CPE is
not a commercial material either, and, if
used as a liner, would require the devel-
opment of an adhesive system. Results of
immersion tests of these materials for 99
days are reported in Table 8.
The absorption of wastes by these
four materials is significantly less than
observed with the 12 membranes discussed
above. Minor changes in physical propert-
ies were also observed, thus indicating
their good chemical resistance. Addition-
al specimens of these four materials
continue to be exposed and will be tested
later.
TABLE 6.
ABSORPTION OF WASTE BY POLYMERIC MEMBRANE ON IMMERSION IN SELECTED WASTES
(Data in change in weight, percent)
Waste and immersion time in days
Compound data
Liner
Polymer
Butyl
CPE
CSPE
CSPE
ELPOf
EPOM
EPOM
Neoprene
Polyester
PVC
PVC
PVC
data
Numberd
44
77
6R
55
36
83R
91
90
75
11
59
88
Type^
VZ
TP
TP
TP
TP
TP
VZ
VZ
TP
TP
TP
TP
Extractables0, *
11.8
9.1
3.8
4.1
5.5
18.2
23.6
21.5
2.7
33.9
35.9
33.9
Pesticide
807 d
1.6
12.7
17.3
15.7
0.5
4.5
20.4
11.4
4.2
5.1
1.0
1.6
HN03
751 d
3.8
19.9
10.0
10.9
7.6
4.2
50.9
17.4
6.4
22.1
-6.1
28.2
HF
761 d
3.7
12.9
9 0
7.7
1.1
3.1
23.9
12.0
2.0
18.1
0.9
14.3
Spent
caustic
780 d
0.8
1.1
4.3
3.3
0 6
1.6
1.3
1.5
1.5
0.4
-0.9
1 1
Lead
786 d
28.7
118.9
120.7
116.2
17.0
24.8
34.7
59.1
7.4
-1.5
7 4
-5.2
Oil
Pond
104
752 d
103.9
36.9
49.5
55.0
28.9
26.5
84.7
26.3
8.5
-10.4
-0.5
-9.8
Aromatic
oil
761 d
31.2
226. 4d
105.2
110.5
29.4
19.8
34.2
142.6
16.6
18.5
28.9
14.1
Weed
oil
809 d
64.2
ND«
368.4
• 347.5
38.1
84.4
76.2
89.3
14.7
15.3
24.7
25.2
^Serial number of liner set by Matrecon; R = Fabric reinforced.
bType = Vulcanized (VZ) or thermoplastic (TP).
cAfter volatiles were removed.
'•Specimen partially dissolved in the waste; value reported is an approximation.
eHO = Specimen was lost; there is some indication that it dissolved in the waste.
fELPO = Elasticized Polyolefin.
151
-------
TABLE 7 CHANGE IN EXTRACTABLES OF FLEXIBLE MEMBRANE LINERS DURING EXPOSURE IN WASTES
Waste and immersion time in days
Liner
Polymer
Butyl
CPE
CSPE
CSPE
ELPOf
EPDM
EPDM
Neoprene
Polyester
PVC
PVC
PVC
data
Number3
44
77
6R
55
36
83R
91
90
75
11
59
88
Compound data
Typeb
VZ
TP
TP
TP
TP
TP
VZ
VZ
TP
TP
TP
TP
Pesticide
Extractablesc, % 807 d
11.8
9.1
3.8
4.1
5.5
18.2
23.6
21.5
2.7
33.9 +0.7
35.9 -0.5
33.9 -1.1
HNOj
751 d
-0.6
+2.5
0
-0 1
+0 7
-0 5
+1 2
-1.7
+7 2
-0 8
-8.0
-0.3
HF
761 d
-1 1
+ 1.0
0
-0 5
+0 4
-0.9
-0 8
-1.4
+0.6
-0.6
-0 9
-1.5
Spent
caust ic
780 d
-0.8
+0.1
0
+0 1
0
-0.2
-0.5
-0.5
0
0
0
-0.7
Oi
Po
d Aromatic
Lead 104 oil
786 d 752 d 761 d
+4 3 +15
+6.1 +8
+0 8 +10
-0 4 +9
+1 3 +10
+2.2 +2
+3 9 +8
-1 7 +1
+0.6 +3
-12.3 -11
-4 5 -4
-11.7 -14
5 +10.7
0 NDd
4 +22.0
8 +21 0
0 +13.4
9 +6.0
6 +74
7 +9 4
5 +11 5
6 +26
4 +1.8
8 +1.9
Weed
oil
809 d
+2 8
N0e
+ 12 2
+10 0
+2.4
+0.4
+1.2
-12.6
+3.2
-8.7
-3 1
-9 6
aSenaI number of liner set by Matrecon; R = Fabric reinforced
bType = Vulcanized (VZ) or thermoplastic (TP).
cAfter volatiles werre removed.
dND = Specimen partially dissolved in the waste.
eND = Specimen was lost, there is some indication that it dissolved in the waste.
fELPO = Elasticized Polyolefin.
TABLE 8. ABSORPTION OF WASTE BY THREE CRYSTALLINE POLYMERS AND
A VULCANIZED CPE COMPOSITION
(Data in change in weight, percent)
Waste and immersion time
Polymer
CPE
HOPE
LDPE
ppC
Liner data
Number d
100
105
108
106
Typeb
VZ
XL
XL
XL
HN03
99d
+0.1
-0.3
+0.1
HF
99d
+4.2
+0.05
-3.3
+0.05
Spent
caustic
99d
+0.2
+0.1
+0.2
Aromatic
oil
99 d
+11.9
+4.4
+8.5
+0.4
in days
Weed
oil
99d
+6.4
+10.7
+1.4
Slopwater
99 d
-0.5
+0.2
+3.5
+0.1
aSerial number of liner set by Matrecon.
bType = Vulcanized (VZ) or crystalline (XL).
CPP = Polypropylene.
152
-------
SWELLING UF POLYMERIC MEMBRANES
The polymeric compounds that have
been studied in this project can be
classified in the following four types:
-Vulcanized elastomers, e.g.
butyl rubber, neoprene, EPDM,
CSPE, CPE, ECO, nitrile rubber,
and blends.
- Thermoplastic elastomers (TPE),
e.g. CSPE, CPE, polyolefins,
and blends.
- Thermoplastics, e.g. plasticized
PVC, and PVC blends with se-
lected elastomers.
- Crystalline
plastic), e
polymers (thermo-
,g. LDPE and HOPE.
Some of the polymers are used in both
crosslinked and thermoplastic versions.
For example, lining materials of CSPE and
EPDM have been manufactured in both
crosslinked and thermoplastic versions.
On exposure to fluids, most polymeric
materials, whether they are crosslinked
or not, will tend to swell and change in
some properties. The major factors
involved in the swelling of polymeric
materials are:
- Solubility parameter.
- Crosslinking of polymer.
- Crystal 1inity of polymer.
- Chemical stability.
- Soluble constituents in compound.
The solubility parameter (Hildebrand
and Scott, 1950) is used by polymer
scientists to measure the similarity in
chemical characteristics of polymers
such as are used in the lining materials.
This parameter is related to the polarity
of materials. A polymer which has a
solubility parameter value close to that
of the fluid medium in which it is placed,
will swell and sometimes dissolve. For
example, a hydrocarbon rubber like natural
rubber will swell and dissolve in a
hydrocarbon such as gasoline. On the
other hand, a highly polar polymer, such
as polyvinyl chloride or nitrile rubber,
will not dissolve in gasoline.
Crosslinking a polymer or a rubber
limits its swelling in solvents. Polymer
scientists use the swelling of a cross-
linked rubber as a measure of the degree
of Crosslinking - the greater the cross-
linking, the less the swelling. This
effect is pronounced in such rubbers as
CSPE and CPE, liners of
able in both vulcanized
versions.
which are avail -
and unvulcanized
Crystal 1inity in a polymer acts much
like Crosslinking to reduce the ability
of a polymer to dissolve. Highly crystal-
line polymers, such as high-density
polyethylene, will not dissolve in gaso-
line, even though they are basically
similar in chemical composition. Such
high density polymers are finding con-
siderable use in containers for a wide
range of solvents-and chemicals.
The soluble constituents of a rubber
compound have a strong bearing on the
ability of that material to swell. Most
commercial polymers contain minor amounts
of soluoles, e.g. salt, which are used in
the manufacture of the polymer. Soluble
constituents can also arise from ingre-
dients which are used in the compounding.
Swelling is a result of the diffusion
of water into the compound by osmosis.
The effects that the first three
factors have on the swelling of liner
materials is illustrated in Figure 4.
The swelling of the thermoplastic type
of material in a fluid with which it is
somewhat compatible is represented in
Curve A, which indicates that the material
will continue to swell with time and that
no real plateau is reached.
The swelling
is represented in
swelling reaches
of crosslinked material
Curve B, in which the
a plateau and changes
TYPES OF SWELLING OF POLYMERIC MEMBRANE LINERS
Figure 4. Types of swelling of polymeric
membranes.
153
-------
only slightly with time. The level of the
plateau is determined by the degree of
slinking and by the solubility parameters
of the waste fluid and the polymer.
Curve C represents the case of a
plasticized thermoplastic or an oil-
extended rubber in which the plasticizer
is leached from the polymer. In this
case, there is an initial swelling fol-
lowed by a reduction in swelling. In some
cases, there can be a shrinkage of the
liner due to the loss of plasticizer or
oil.
In selecting polymer and rubber
compounds for service in a liquid medium,
a designer generally selects materials
which have low or negligible swell.
Swelling of a compound usually has adverse
effects which will make the product
unserviceable. Some of the major effects
of swelling generally are:
- Softening and possibly dissolving.
- Loss of tensile and mechanical
strength.
- Loss in elongation.
- Increased permeability.
- Increased potential of creep.
- Greater susceptibility to polymer
degradation.
TABLE 9.
EFFECTS ON PROPERTIES OF LINER MATERIALS3 ON IMMERSION
(0.« TBP)
All of these effects are adverse with
respect to liner performance. Swelling,
therefore, is a valuable indicator of the
compatibility of a liner to a waste.
Shrinkage can also be a measure of
compatibility for liner compositions that
are highly plasticized. For example, in
the case of highly plasticized PVC com-
pounds, the plasticizer can leach and
diffuse out of the polymer compound,
leaving it stiff and brittle.
IMMERSION OF POLYMERIC MEMBRANES IN AN
AQUEOUS SOLUTION OF TRIBUTYL PHOSPHATE
The immersion of a series of widely
different polymeric membrane liners in a
saturated aqueous solution of tributyl
phosphate demonstrates the effect of a
very low (0.40%) concentration of organic
chemical constituents on the swelling and
properties of a membrane liner during
prolonged exposure. Tributyl phosphate
was selected for this experiment because
of its slight solubility in water, its low
volatility, and its phosphorus content
which can be used to follow the migration
of the plasticizer. Results of exposing
specimens of 12 polymeric membrane lin-
ings, all without fabric reinforcement,
are presented in Table 9. The sheetings
include thermoplastic, vulcanized, and
crystalline types of polymeric membranes.
IN A SATURATED SOLUTION OF TRIBUm PHOSPHATE IN UEIONIZEU HATER
FOR 522 DAIS
Polymer
Type of compound
Matrecon number
Initial thickness, mils
Analytical properties.
Height gain, X
Extractables , original, X
Extractables, exposed, X
Gain in vol at i les, %
Physical properties^:
Final thickness, mils
Tensile strength, X retention
Elongation at break, X retention
Stress at 100X elongation.
X retention
Tear resistance, X retention
Hardness change, points
Puncture test:
Stress, X retention
Elongation, X retention
Butyl
VZ
44
63
21 9
11 8
Z 2
64 3
107
115
74
-2
73
126
CPE
TP
77
30
107. Z
9 1
42 2
47.9
10
155
6
14
-60
20
127
CPE
VZ
100
36
34 4
19 7
40.9
63
79
46
29
-20
85
125
CSPE
TP
55
33
31 6
4 1
20 7
37.9
48
79
80
39
-14
112
131
ELPOb
TP
36
23
6.7
5 5
2 1
20.2
48
74
70
79
-3
101
142
EPDM
VZ
91
32
5.2
23 6
4 3
36.8
93
106
81
105
-1
107
111
Neoprene
VZ
90
35
41 4
21 5
24.8
43 5
49
82
15
34
-25
66
119
PESEC
TP
75
7
4 5
2 7
4.1
6.6
89
101
90
90
-10e
77
100
PVC
TP
59
33
46 2
35 9
11 9
36 4
31
89
28
23
-33
48
133
HOPE
XL
105
3?
0 56
0
31 5
88
101
92
77
-1
101
107
LOPE
XL
108
31
0 51
0
30.9
104
102
103
81
-20
92
108
PP
XL
106
33
0 IB
0
32 2
97
105
110
270
-2
94
42
aAlI materials are unreinforced.
bElasticized polyolefin
CPESE = Polyester elastomer
d|Jata for tensile, elongation, S-10Q, and tear are the averages of measurements made in both machine and transverse directions
eDue to curling of sample, magnitude of change as reported is probably too great.
154
-------
The crystalline materials are not commer-
cial liner sheeting, but two, HOPE and
LPDE, are available as liner membranes.
These three sheetings were used in this
exposure test in order to be able to make
comparisons at the same thickness with
most of the other polymeric materials.
The immersion exposure was for 522 days
and the following parameters were
measured: weight change, extractables,
volatiles (water), tensile strength,
elongation at break, modulus, hardness,
tear, and puncture resistance.
The thermoplastics, PVC, CPE, and
CSPE, increased significantly in weight
and lost considerably in tensile strength
and hardness. The vulcanized CPE sus-
tained less change than did the thermo-
plastic CPE composition. The neoprene
liner, though vulcanized, swelled con-
siderably and lost in tensile strength and
hardness. The butyl and EPDM liners,
also vulcanized, did swell to some
extent and sustained only modest losses
in properties.
Of particular interest are the low
swelling and loss in properties incurred
by the crystalline polymers, thus showing
the good chemical resistance of these
materials.
Immersion of the second set of
specimens in this test is continuing
and a full analysis is being made of the
swollen exposed materials to determine the
composition of the extractables.
DISCUSSION
Results generated in the project
indicate that, within present liner
technology, materials are available
which are suitable to use to line impound-
ments of a great range of hazardous
wastes. Certainly, materials are avail-
able to handle the wastes that were used
in this project. On the other hand, no
single available lining material appears
to be suitable for long-term impoundment
of all wastes. There is considerable
variation in the compatibility of wastes
and liners. Generally, wastes that are
highly ionic, contain salts, strong acids,
or strong bases, can be aggressive to
soils, soil cement, aggregate, hydraulic
asphalt concrete, and to some of the
membranes, particularly those containing
plasticizers. Wastes which have organic
components can be aggressive toward the
membrane liners by swelling the polymer
compounds and toward asphaltic materials
by dissolving them. In particular, the
oily wastes cause swelling and loss of
physical properties in the membrane liners
and the asphaltic materials.
It is quite apparent, therefore, that
compatibility studies must be performed
prior to the selection of the liners for
handling specific wastes and that, once
the selection has been made and the site
lined, the waste that is being impounded
should be monitored closely to insure that
only proper waste is added. In making
the compatibility studies, contact of
samples of the liner materials under
consideration and the waste to be im-
pounded should be as long as possible.
Our tests show that, in many cases, no
leveling off in the effects of the waste
on the membrane occurs with time. Such
behavior means that, for long-term ex-
posure, there can be a continuing change
of the liner with time.
In the case of the polymeric membrane
liners, some of the more useful tests and
properties which appear to be promising
for assessing the suitability of a given
liner for use with a given waste are:
1. The measurement of swelling of
the membrane materials in the
wastes or in the leachate of the
wastes.
2. The loss of elongation, with
time, of the polymeric membranes
when in contact with the waste or
its leachate.
3. The use of the pouch test for
measuring the compatibility of
the liners and the waste and the
permeability of the liner to the
constituents of the waste.
The laboratory tests that are pro-
posed need continuous comparison with
long-term exposures in order that they can
be used to assess liners and predict their
long-term performance in service. Such
correlation requires observation of liners
in actual service over extended periods
of time.
The analyses and characterization of
wastes with respect to those constituents
that are aggressive to lining materials
will be useful in the selection of liners.
155
-------
In developing a new liner material
for use in the lining of waste impound-
ments, it is quite obvious that the new
material under development should be
tested under a variety of environmental
conditions. Furthermore, it would appear
desirable to consider composite types of
liners involving two or more materials for
use in impounding the more aggressive and
polluting types of wastes.
ACKNOWLEDGMENT
The research which is reported in
this paper is being performed under
Contract 68-03-2173, "Evaluation of
Liner Materials Exposed to Hazardous
and Toxic Sludges", with the Environmental
Protection Agency, Municipal Environmental
Research Laboratory, Cincinnati, Ohio.
The support and guidance of Mr. R. E.
Landreth, Project Officer, is gratefully
acknowledged.
REFERENCES
Fong, M. A., and H. E. Haxo. 1981.
Assessment of Liner Materials for Munici-
pal Solid Waste Landfills. Seventh Annual
Research Symposium, Philadelphia, PA. (See
these proceedings).
Haxo, H. E. 1976. Evaluation of
Selected Liners when Exposed to Hazardous
Wastes. In: Proceedings of the Hazardous
Waste Research Symposium, Residual Manage-
ment by Land Disposal. EPA-600/9-76-015.
U. S. Environmental Protection Agency,
Cincinnati, Ohio.
Haxo, H. E. 1978. Interaction of
Selected Lining Materials with Various
Hazardous Wastes. In: Proceedings of the
Fourth Annual Research Symposium. EPA-600/
9-78-016. U. S. Environmental Protection
Agency. Cincinnati, Ohio.
Haxo, H. E. 1980a. Interaction of
Selected Liner Materials with Various
Hazardous Wastes. In: Disposal of
Hazardous Waste, Proceedings of the Sixth
Annual Research Symposium. EPA-600/9/80-
010. U. S. Environmental Protection
Agency. Cincinnati, Ohio.
Haxo, H. E. 1980b. Laboratory Evalu-
ation of Flexible Membrane Liners for
Waste Disposal Sites. Presented at the
117th Meeting of the Rubber Division.
American Chemical Society. Las Vegas,
Nevada.
Haxo, H. E., R. S. Haxo, and R. M.
White. 1977. Liner Materials Exposed to
Hazardous and Toxic Sludges. First Interim
Report. EPA-600-2-77-081. U. S. Environ-
mental Protection Agency. Cincinnati,
Ohio.
Haxo, H. E. 1981a. Testing of Ma-
terials for Use in Lining Waste Disposal
Facilities. Presented at ASTM Symposium on
Hazardous Solid Waste Testing. Ft.
Lauderdale, FL.
Haxo, H. E., et al . 1981b. Liner
Materials Exposed to Hazardous and Toxic
Sludges. Final Report. EPA-68-03-2173. In
preparation.
Hildebrand, J. H., and R. L. Scott.
1964. The Solubility of Nonelectrolytes.
Third Edition. Dover Publications, Inc.,
New York. 488 pp.
Matrecon, Inc. 1980. Lining of Waste
Impoundment and Disposal Facilities.
SW-870. U. S. Environmental Protection
Agency. Washington, DC. 385 pp.
156
-------
INSTALLATION PRACTICES FOR LINERS
David W. Shultz
Michael P. Miklas, Jr.
Southwest Research Institute
6220 Culebra Road
San Antonio, Texas 78284
ABSTRACT
Southwest Research Institute is presently conducting a study to identify current methods
and equipment used to (1) prepare supporting subgrade and to (2) install liners at various
waste disposal facilities in the United States. Information obtained during the site
visits includes:
(1) Methods and equipment used to prepare the subgrade;
(2) Methods and equipment used to place the liner material;
(3) Design and construction considerations; and
(4) Any special problems encountered during installation and subsequent solutions.
The subgrade preparation and installation of three generic types of liner materials are
discussed in this paper. Photographs depicting various construction and placement activi-
ties are presented.
INTRODUCTION
The use of surface impoundments and
landfills to store, treat and/or dispose of
unwanted materials has been and continues
to be common practice for industry and
municipal agencies since these types of
facilities have proven to be cost effective
solutions to treatment and disposal prob-
lems. Recent studies have shown that the
use of such facilities can result in sub-
surface migration of hazardous materials
into groundwater resources.
The Resource Conservation and Recovery
Act of 1976 and proposed EPA regulations
will require positive control of subsurface
migration of contamination from many of
these facilities where hazardous materials
are stored. Future treatment and/or dis-
posal sites will likely have to be designed
to prevent groundwater contamination. It
is likely that impermeable liners with high
integrity will find increased usage in the
future.
The proper planning, design and con-
struction of surface impoundments and/or
landfills designed to contain hazardous
wastes involve numerous steps, including
the following: (1) defining facility func-
tion and geometry; (2) selection of a liner
material which is compatible with the mate-
rial to be stored or treated; (3) planning
suitable subgrade preparation, proper liner
installation, adequate seepage monitoring
and collection provisions; and (4) develop-
ing appropriate post-installation operation
and maintenance planning.
This study, which addresses subgrade
preparation and liner placement, has the
following objectives:
(1) To identify current subgrade
preparation procedures and equip-
ment used to build surface im-
poundments and landfills;
(2) To identify methods and equipment
utilized to install various liner
157
-------
materials; and
(3) To identify special problems
which should be considered during
the planning, construction, and
operation of such facilities.
The following generic types of liners
are intended to be included in this study:
(1) compacted native soils (clays); (2)
admixes (asphalt, concrete and soil cement);
(3) polymeric membranes (rubber and plastic
sheeting); (4) sprayed-on linings; (5) soil
sealants; and (6) chemical absorbants.
This paper discusses three of these types,
i.e., asphalt, soil sealant, and sprayed-on
liners.
SITE 1 - ASPHALT LINED LANDFILL
Background
Site 1 is a municipal landfill begun
approximately 5 years ago. Presently, the
landfill covers 20 acres with an average
compacted fill depth of 100 feet. The site
is situated in a ravine. The ravine is 300-
600 feet across at the top of the hill and
it gradually widens toward the bottom. Con-
struction was initiated at the upper end and
is progressing downward. Future plans call
for the filling of the entire ravine, re-
sulting in a fill covering 170 acres.
The landfill is lined with a road
grade asphalt, approximately 2 inches thick.
The liner has been expanded in 5 acre sec-
tions as the landfill construction pro-
gressed. Standard roadway paving equipment
has been used to install the liner. The de-
sign of the liner provides for containment
of any leachate produced by the fill. Sub-
grade preparation and liner placement are
discussed in the following sections.
Subgrade Preparation
Adequate compaction of subgrade after
grading is critical to the integrity of any
liner. Sufficient compaction insures that
the liner will not be damaged as a result of
differential settling. The site was graded
to conform to the slope of the ravine.
Figure 1 shows the drainage provided at the
edge of the compacted subgrade. Note the
slope of the subgrade off to the left. The
subgrade gradually slopes from this side of
the ravine to the other side, where a curb
allows any leachate produced to be effec-
tively collected. Surface runoff into the
land-ill is controlled by the drainage
channels surrounding the site. These chan-
nels prevent overland flow from entering
the fill. Surface drainage from the chan-
nels is directed to an erosion control
basin. After final grading, 12 inches of
specially selected material were recom-
pacted in 6 inch lifts to achieve an opti-
mal subgrade consistency. Standard vibra-
tory rollers were used to produce a sub-
grade with a modified Proctor density of
90 percent.
Liner Placement
The liner at this site consists of 2
inches of road grade asphalt applied using
a standard paving machine, as shown in
Figure 2.
The roller shown in Figure 3 was used to
compact the asphalt to the desired density.
After the surface cooled, a seal coat was
hydraulically applied. The seal coat is in-
tended to protect the asphalt from potential
degradation which might occur as a result of
exposure to leachate. An asphalt curb was
constructed at the edge of the liner to
direct leachate to a perforated PVC pipe
located inside the curb upon the normal
asphalt surface. The PVC leachate collec-
tion pipes from each 5 acre section of
asphalt are linked together via a manifold
system. As new acreage is added, the re-
quired new piping is connected to the mani-
fold system. A 20,000 gallon tank stores
any leachate produced. Leachate can either
be hauled off site or recycled back into the
fill. To date, after 5 years of operation,
about 13,000 gallons of leachate have been
hauled off site.
Cost estimates for the purchase and
placement of the asphalt are $15,000 per
acre, or $0.35 per square foot. Costs for
subgrade preparation are not available.
SITE 2 - SOIL SEALANT LINER
Background
Site 2 is a 42 acre diked wastewater
storage impoundment. Excavated material
from the impoundment floor was used to con-
struct the dikes. The floor is approxi-
mately 40 feet below the elevation of the
dikes. Side slopes are three (horizontal)
to one (vertical). A Wyoming bentonite clay
modified with a cross-linked polymer was
158
-------
selected as the soil sealant. The native
soil is a glacial silt having an unaccept-
able permeability for the intended use of
the impoundment.
Subgrade Preparation
After the site had been graded and re-
compacted according to design specifica-
tions (90% Proctor) the floor of the im-
poundment was disced to a depth of 4 inches
using the dozer drawn disc shown in Figure
4. This was necessary to provide the loose
topsoil for blending with the soil sealant.
Only one pass of the disc was required.
The next step was the application of the
bentonite.
Liner Placement
The modified bentonite soil sealant
was applied pneumatically by using two air
bottle transport trucks equipped with hard-
ware specifically designed for this appli-
cation. The trucks each carried an air
compressor, two storage tanks and associ-
ated hardware. Two important additions
were a 6 foot wide material distribution
boot and an outrigger. These are shown in
Figure 5. The boot was attached to the
rear of the truck. Soil sealant was forced
out through the top of the boot. Internal
veins running from the inlet to the bottom
allowed an even distribution of material.
The bottom skirt helped minimize the loss
of material during application. The out-
rigger shown in Figure 5 allowed the opera-
tor immediate and safe access to the mate-
rial control valves on the side of the
truck.
Laboratory tests indicated the desired
application rate of soil sealant to be 34
tons/acre of 1.6 pounds/square foot. The
required flow rate of soil sealant from the
truck as well as the rate of travel were
established. Prior to initiating applica-
tion of the sealant, distance markers were
set every 100 feet along the toe of the
dike. These markers helped the application
crew determine the rate of forward movement
of the truck. A strict control of the
forward rate of travel was necessary to es-
tablish the proper application rate of
sealant.
As soon as the marker stakes were set,
the application crew began spreading the
sealant, as shown in Figure 6. The sealant
can be seen to the right of the truck.
This operation requires at least two people,
a driver and a person to control the appli-
cation of sealant. To allow continuous
spreading during daylight hours, two trucks
were utilized. While one truck was spread-
ing the sealant, the second truck was being
loaded with sealant stored in six 25-ton
capacity silos located on site. The stor-
age capacity in the silos equaled the
quantity of material which could be applied
in one day by two trucks.
The tarp shown in Figure 7 was used by the
on-site quality control engineer to assure
the proper quantity of sealant was being
applied. The tarp was weighed after the
truck passed over it to determine the
quantity of sealant material applied in one
pass. This test was performed periodically
throughout the application procedure.
The disc previously shown in Figure 4
followed the application crew. Approxi-
mately eight passes in a crisscross pattern
were made to blend the sealant with the
soil. The decision as to whether adequate
blending had been achieved was made by the
application foreman and the on-site techni-
cal staff representing the owner of the
facility. Adequate blending is necessary
for the sealant to reduce permeability to
the desired level. The crisscross pattern
helped break up clods. Figure 8 shows the
disc making a second pass over the soil
after sealant had been applied. Freshly
applied sealant can be seen to the right in
this figure.
Figure 9 shows an overview of sealant ap-
plication and discing under way. The seal-
ant truck is to the left. The disc can be
seen turning around to make a new pass at
the foot of the dike near the end of the
impoundment.
On the second day of application, the
installation contractor located an 8 foot
wide rototiller. This unit was used to
complete the blending for the remaining 39
acres. It performed better than the disc,
producing a more uniform mixture of soil
and bentonite.
Soil sealant was applied to the side
slopes by mounting the spreader boot on the
blade of a bulldozer. The bulldozer trav-
elled the side slopes horizontally while
tethered to the sealant supply truck by a
hose. The supply truck was driven around
the top of the dike during application on
the upper half of the slope, and around the
159
-------
bottom during application on the lower half
of the slope.
Upon completion of the soil and seal-
ant blending operation, water was added to
the soil on the floor of the impoundment in
sufficient quantity to allow compaction to
95 percent of maximum Proctor density.
Sufficient moisture was present in the soil
on the side slopes to allow the desired
compaction without additional water. Due
to their steepness, the side slopes were
compacted with a rubber tired roller which
could traverse the existing grade. The
level floor was compacted with a steel rol-
ler. To protect against erosion, slag was
placed along the side slopes.
SITE 3 - SPRAYED-ON LINER
Located at Site 3 was a small test
evaporation pond lined with a urethane
modified asphalt liner system. The system
is comprised of a polypropylene mat coated
with the urethane asphalt material. Loose
desertic soil with high permeability was
present at the site.
To prepare the urethane asphalt for
spray application, a premix was combined
with an activator at a volume ratio of nine
to one. Mixing was accomplished in 5 gal-
lon buckets. The mixture was applied using
a specialized spray application unit. This
unit attached to a 5 1/2 gallon container.
A compressor/generator unit was used to
provide necessary power for the spray unit.
The asphalt was then applied using the hand
held spray gun. Sufficient asphalt was
applied to achieve saturation of the fabric
mat. One coat was applied at the rate of 2
gallons per minute, covering about 8 square
yards per minute. This application rate
produced the desired membrane thickness of
50 mil. Figure 12 shows the pump unit (to
the right) and the spray application using
the hand held spray gun. Note that the
horizontal seam on the fabric mat behind
the applicator has been pre-soaked with the
asphalt material. This was done for all
fabric seams to assure penetration of mate-
rial at the critical seam overlaps. In
addition, the fabric mat extending into
the anchor trenches was coated with asphalt.
The installer recommends that the membrane
be allowed to cure for 24 hours before be-
ing put into service.
Subgrade Preparation
This impoundment was totally excavated
below the surrounding grade using a small
bulldozer. Final grade of the side slopes
was approximately three (horizontal) to one
(vertical). The surface was smooth, com-
pacted, and free of holes or rocks.
Liner Placement
As previously noted, the liner con-
sisted of a fabric mat to serve a support-
ing medium and the asphalt material applied
to the mat. Figure 10 shows the fabric mat
positioned over the subgrade. A 6 inch
overlap was maintained along all joints.
Four inch staples were used to temporarily
secure the mat to the soil until the as-
phalt could be applied.
The mat was extended over the top of the
side slopes into an anchor trench 1 foot
deep around the perimeter of the impound-
ment, as shown in Figure 11. The installer
was careful to limit the thickness of the
mat to no more than three layers.
SUMMARY
Subgrade preparation requirements and
liner placement procedures for three differ-
ing processes observed at three installation
sites have been presented in this paper.
The following table presents a comparactive
ranking of various subgrade and placement
variables for the three systems. This
ranking is based upon personal on-site
observations of the authors and available
information from manufacturers. Ranking
should be utilized for relative/subjective
comparisons only.
ACKNOWLEDGEMENTS
This stuuy is being supported by the
Solid and Hazardous Waste Research Division,
Municipal Environmental Research Laboratory,
U.S. Environmental Protection Agency,
Cincinnati, Ohio (Grant R806645010) and
Southwest Research Institute, San Antonio,
Texas. The authors want to thank Robert E.
Landreth, Project Officer, for his guidance
and support.
160
-------
COMPARATIVE RANKING* OF SELECTED SUBGRADE AND PLACEMENT VARIABLES
FOR THREE LINER SYSTEMS
Liner System
Urethane Asphalt
Asphalt Over Fabric Mat
Degree of Weather Limitations
Level of Crew Experience Needed
Intensity of Subgrade Compaction
Necessary
Availability of Materials
Availability of Installation Crews
Intensity of Quality Control Required
During Installation
Cost of Materials and Installation
2
3
1
1
1
3
2
3
1
2
3
3
1
1
Soil Sealant
1
2
3
2
2
2
3
*1 = highest
3 = lowest
Figure 1. Drainage channel to divert surface water away
from an asphalt lined landfill.
161
-------
Figure 2. Asphalt paving machine installing an asphalt
liner at a landfill site.
Figure 3. Finishing roller at work on an asphalt liner.
162
-------
Figure 4. Disc used to prepare soil for the application
of a polymer modified Wyoming bentonite.
Figure 5. Air bottle transport truck equipped with a
special distribution boot attached to the
rear and an outrigger on the left.
163
-------
Figure 6. Application of modified bentonite soil sealant
using an air bottle transport truck.
Figure 7. A piece of tarp is used to assure the proper
quantity of soil sealant is applied to the
soil.
164
-------
Ji' ;* «* * *• /--•TW*"'**-^ 5*W
V P «"V * K-- * , ^fe? %*•»" * *- "W^
Figure 8. A disc is used to blend the modified bentonite
into the soil.
Figure 9. An overview of sealant application and blending
activities.
165
-------
i/
Figure 10. A fabric mat anchored in position which will
be covered with a spray-on liner system.
Figure 11. The fabric mat extends into the anchor trench
around the perimeter of the impoundment.
166
-------
•Jfc
Figure 12. A workman applies a methane asphalt to a
fabric mat using a hand held spray applicator.
167
-------
IDENTIFICATION OF HAZARDOUS WASTE FOR LAND TREATMENT RESEARCH
Joan B. Berkowitz, Judith C. Harris, and Bruce Goodwin
Arthur D. Little, Inc.
Cambridge, Massachusetts 02140
ABSTRACT
A methodology is discussed for identifying and prioritizing hazardous waste streams,
potentially amenable to land treatment as a waste management option. The methodology is
being developed to suggest promising candidates for land treatment research, field testing,
and verification.
INTRODUCTION
Under contract to the U. S. Environ-
mental Protection Agency, Cincinnati,
Ohio, Arthur D. Little, Inc. is currently
working on a project to identify potential
industrial waste candidates for land treat-
ment techniques. The objectives of the
project are: (1) to identify and priori-
tize hazardous waste streams likely to be
amenable to landtreatment as a management al-
ternative and (2) to recommend laboratory
and field research to fill in significant
data gaps and to verify the environmental
acceptability of landtreatment as a waste
treatment/resource recovery method for a
significant fraction of the hazardous
wastes generated.
The scope of work includes:
• Development of prioritization cri-
teria for ranking hazardous wastes
potentially suitable for treatment
and/or utilization in a land treat-
ment facility;
• Application of the criteria to the
list of wastes promulgated and pro-
posed by the U.S. EPA to be regu-
lated as hazardous under RCRA,
Subtitle C;
• Identification of research require-
ments to verify the suitability of
land treatment for the more promis-
ing hazardous waste streams;
• Recommendations for future research
programs.
The RCRA regulations published in the
Federal Register of May 19, 1980 (Vol. 45,
No. 98, p. 33205 ff) define a land treat-
ment facility as "that part of a facility
at which hazardous waste is applied onto or
incorporated into the soil surface." Land
treatment practice generally involves ap-
plication of waste to the soil surface, or
injection of waste immediately below the
surface, and incorporation of the waste in-
to the upper 6-8 inches of the soil by
standard farming techniques such as plow
harrowing or disc burrowing. In contrast
to hazardous waste landfills, which are de-
signed for long term storage of wastes,
land treatment facilities are designed to
use the soil system for physical/chemical/bio-
logical treatment of hazardous waste con-
stituents.
Land treatment has been used for oily
wastes from the petroleum refining industry
for more than 25 years. To our knowledge,
none of these land treatment facilities,
which are located in many different parts
of the country, have given rise to human
health or environmental damages incidents.
Furthermore, extensive monitoring of field
sites has demonstrated that certain types
of oily wastes are transformed in an aero-
bic soil environment to a less hazardous
or non-hazardous composition.
Land treatment is not used extensively
168
-------
for hazardous wastes other than petroleum
refining wastes, but the U.S. EPA has
stated that it "may be feasible for other
types of wastes" (Fed. Reg. Vol. 45, No.
98, May 19, 1980, p. 33205). In order to
use land treatment for the management of
a hazardous waste, one must demonstrate
that the waste can be made less hazardous
or non-hazardous by biological degradation
or chemical reactions occurring in or on
the soil. Demonstration of the feasibil-
ity of land treatment for any particular
hazardous waste will generally involve a
three step process. The first step con-
sists of analysis of the waste, and review
of available information on the biological
and chemical action of natural soil treat-
ment processes on each of the waste con-
stituents in order to establish whether
or not land treatment appears to be prom-
ising. The next step would involve labora-
tory and greenhouse studies under simu-
lated field conditions to obtain basic data
on degradability, sorption and mobility,
volatilization, and toxicity, suitable for
developing a preliminary land treatment
facility design and operating plan. Final-
ly, since water-soil interactions are dif-
ficult to predict, and not easily simula-
ted in the laboratory, field pilot
studies will generally be carried out to
verify the feasibility and environmental
acceptability of the preliminary design
and operating plan. Each step requires an
increased commitment of resources. Hence,
it is important to develop decision cri-
teria for proceeding from the initial
assessment, to laboratory studies, and
ultimately to field testing. The purpose
of this paper is to suggest a highly sim-
plified approach to deciding whether land
treatment appears to be sufficiently
promising for wastes listed as hazardous
under RCRA to warrant laboratory testing
and/or field demonstration.
METHODOLOGY
The approach being used to meet the objec-
tives of the program is based on the Stage
I land limiting constituent analysis pro-
posed by Overcash.d) This analysis fo-
cuses primarily on plant-soil assimilative
capacity, and Overcash's "basic nondegra-
dation constraint." The latter is stated
by Overcash as follows:
"The industrial waste, when con-
sidered on a constituent-by-consti-
tuent basis, shall be applied to
the plant-soil system at such
rates or over such limited time
spans that no land is irreversibly
removed from some other potential
societal usage."
Since the present program is concerned
with land treatment as a hazardous waste
management option, the emphasis here is on
soil assimilative capacity. Plants may not
be grown, and generally will not be grown,
on industrial hazardous waste land treat-
ment sites. However, following closure,
the site should conform to the basic non-
degradation constraint.
For purposes of ranking wastes listed
as hazardous under RCRA with respect to
potential land treatment effectiveness, a
series of three screens are being developed-
(1) a heavy metal screen; (2) an organics
screen; and (3) a screen for mobile con-
servative species. The data requirements
for application of each screen include quan-
titative information on waste generation
and composition, and general information
on soil assimilative capacity for a wide
range of hazardous waste constituents.
The screening methodology is based on
average annual rates of waste generation
by individual manufacturers in each waste
generating industry of interest. The
average was calculated by dividing the
annual rate of waste generation for the
industry segment by the number of waste
generating plants in the industry segment.
It has been assumed that land treatment
would be given low priority by a waste
generator, if a land area greater than
100 acres (40 km) would be required to
assimilate the wastes from an individual
plant for a period of 10 years. This
figure is somewhat arbitrary, but does
correspond to a relatively large soil-
based waste management facility in current
operating practice.
Assembly of Basic Data
For each waste listed as hazardous in
the proposed RCRA regulation of December 18,
1978 (Fed. Reg. December 18, 1978, Vol. 43,
No. 243), the following data were tabulated
to the extent that they were available:
• Annual rate of waste generated
(kg/yr) = R;
• Number of waste generating plants =
N;
169
-------
• Average rate of waste generation
per plant (kg/yr/plant) = R/N;
• Waste composition (chemical com-
ponents of the waste and the frac-
tion of the waste accounted for by
each chemical component);
• Annual rate of generation of each
component in the waste stream from
an average plant (kg/yr/plant).
The primary sources of data were the
background documents prepared by the EPA
to justify listing of the waste streams
as hazardous; a series of EPA studies on
hazardous waste management in a number nf
major waste generating industries (2)-(l4))
and other available published and unpub-
lished literature.
Data on the composition of hazardous
wastes is very limited, and has focused
largely on major constituents and hazardous
components. Available data only permit a
very rough screening of listed wastes as
primary potential candidates for land
treatment research and demonstration.
Chemical analyses are typically insufficient
for a mass balance. Metals analyses are
typically not available for waste streams
that are primarily organic chemical in
nature. For waste streams that are primar-
ily inorganic in nature, analyses are often
limited to toxic heavy metals, with little
data on anonic constituents or other metals
such as sodium, calcium, magnesium, etc.
which also impact on land treatment feasi-
bility.
Metals Screen
Based only on available metals analysis
data for each waste stream, area require-
ments were calculated for assimilation of
each analyzed metal, if all of the waste
from an average plant were applied for a
period of ten years. From this analysis,
the land limiting constituent was identified
(i.e., the constituent requiring the
largest land area for assimilation). With
the assumption that the soil pH would be
maintained between 6 and 7.5, cumulative
limits for metals are primarily a function
of soil cation exchange capacity (CEC).
Four prioritization categories for further
research were defined as follows:
Prioritization Area Require-
Category ment for LLC CEC
High <_ 40 ha < 5
Moderately High <_ 40 ha 5-15
Moderate <_ 40 ha > 15
Low > 40 ha < 15
The rationale for the first three cate-
gories is that if a site smaller than 100
acres will suffice in soil of low CEC, then
soils of higher CEC will also be acceptable.
Hence, there would be a wider choice of
sites than if a CEC of 15 or more were
required.
Allowable cumulative limits used in
the analysis are given ir Table 1.
TABLE 1. ALLOWABLE CUMULATIVE LIMITS (kg/haj
CEC
Metal
<5
5-15
Cd
Cu
Pb
Ni
Zn
5
125
500
50
250
10
250
1000
100
500
20
500
2000
200
1000
Other metals (data not available as a
function CEC:
Mn 500 - 1000 kg/ha
Hg 8 - 10 kg/ha at 500 kg/ha/yr of N
Sn 2000 Kg/ha
Ti very high
Sb <800 kg/ha
Cr 500 kg/ha (ADL estimate based on a
similarity to lead)
Source: Reference (1)
Results are given in Table 2. In
addition to the key information under the
column headings "Area Required for Land-
treatment" and "Prioritization Category,
a column titled, "Excess Area Requirement
Factor" gives the ratio of the total cal-
culated land area requirement divided by
40 ha. The latter quantity provides a
a rough measure of the extent to which a
given waste stream exceeds the 40 ha
maximum land excess area constraint.
With the single exception of "spent H2S04
Pickle Liquor from Steel Finishing Industry'
waste, all candidate waste streams were
found to be unsuitable for landtreatment
using this approach due to high concentra-
170
-------
Table 2. Heavy Metals Screen of Hazardous Waste Streams for Land Treatment Suitability
Listed Waste Stream
Primary Copper Smelting
and Refining - Slowdown
Slurry/Sludge
Acid Plant
Primary Lead Smelting -
Dredged Solids from Liquid
Waste Impoundments
Plant A
Plant B
Secondary Lead Smelting -
Emission Control Dusts/
Sludges
Sludges
Heavy
Metal
Cd
Pb
Hg
Cr
Se
Cd
Pb
Cd
Pb
Pb
Cd
Cr
Pb
Average Quantity
Generated per Area Required for Prioriti- Excess Area
Plant Land Treatment zation Requirement
(kg/10 yrs) (ha) Category Factor
3.9X101* kg 7.8xl03(C.E.C. <5)
3.9xl03(C.E.C. 5-15)
2.0xl03(C.E.C. >15)
6.0xl05 kg 1.2xl03(C.E.C. <5)
6.0xl02(C.E.C. 5-15)
3.0xl02(C.E.C. >15)
60 kg 7.5-6.0
3.8xl03 kg 7.6-1.0
2.3xl03 kg ?
l.SxlO1* kg 3.0xl03(C.E.C. <5)
1.5xl03(C.E.C. 5-15)
7.5xl02(C.E.C. >15)
2.9xl05 kg 5.8xl02(C.E.C. <5)
2.9xl02(C.E.C. 5-15)
1.5xl02(C.E.C. >15)
5.6X101* kg l.lxlO'1(C.E.C. <5)
5.6xl03(C.E.C. 5-15)
2;8xl03(C.E.C. >15)
S^xlO1* kg 5.8xl02(C.E.C. <5)
2.9xl02(C.E.C. 5-15)
1.5xl02(C.E.C. >15)
6.8xl010-7.6xl010 kg 1.4-1.5xiae(O.E.C. <5)
6.8-7.bxl07(C.E.C. 5-15)
3.4-3.8xl07(C.E.C. >15)
4.4xl08-4.9xl08 kg 8.8-9.8xl07 (C.E.C. <5)
4.4-4.9xl07(C.E.C. 5-15)
2.2-2.5xl07(C.E.C >15)
3.8xl01(-4.3xl07 kg (7.6-8.6X1011) -
(9.5xl03-l.lxlO't)
low
low
low
high
high
high
low
mod .high
moderate
high
low
low
low
low
low
low
low
low
low
low
low
low
low
9.2xl09-1.0xl010 kg 1.8-20xl07(C.E.C. <5) low
9. 2xl06-1.0xl07 (C.E.C. 5-15) low
4.6-5.0xl06(C.E.C.>15) low
2.0xl02
98
50
30
15
7.5
—
—
?
75
38
19
15
7.3
3.8
2.8xl02
1.4xl02
70
15
7.3
3.8
3.8xl06
1.9xl06
9.5xl05
2.5xl05
1.2xl05
6.3xl05
2.2xl03
S.OxlO5
2.5xl05
1.3xl05
Comments
Cd and Cr concentrations in leach-
pit given in background document,
but no annual volume given
Quantity generated based on amount
disposed; remainder is recycled
Concentration and generation rates
for both Cd and Pb given in back-
background document, but the numbers
do not agree; generation rates (kg/yr)
given in background document were
used for this calculation.
Concentration data given in back-
ground document may not apply to
waste from plants for which gemera-
tion rates are given.
1980 estimate for total quantity of
waste generated was used for this
calculation.
-------
Heavy Metals Screen of Hazardous Waste Streams for Land Treatment Suitability (continued)
Spent 'Pickle Liquors froi
Steel Finishing Industry
and Treatment Sludge
Dusts
(cont'd)
Sludges
HjSOi, liquor
HC1 liquor
Mixed acids
.on
il Dust
Average Quantity
Generated per
Heavy Plant
Metal (kg/10 yrs)
Cd 6.9xl07-7.7xl07 kg
Cr I.lxl07-1.3xl07 kg
Pb 2.4xl06 kg
Cr 4.7xl06 kg
Cr 2.1xl03-2.1xl01' kg
Pb 0-1.6xl02 kg
Cr 2.5xl02-4.6xl03 kg
Pb 2.5xl02-1.9xl05 kg
Cr 3.9xl01'-5.0xlO'' kg
Pb 11-47 kg
Mn 3.8xl06 kg
Ni 3.5x105 kg
Pb 8.3xl01(-1.4xl06 kg
Area Required for
Land Treatment
(ha)
1.4-1.5xl07(C.E.C. <5)
6.9-7.7xl06(C.E.C. 5-15)
3.5-3.9xl06(C.E.C. >15)
(2.2-2.6x10'')-
(2.8-3.3xl03)
4.8xl03(C.F..C. <5)
2.4xl03(C.E.C. 5-15)
1.2xl03(C.E.C. >15)
9.4-1.2xl03
(4.2-42)-(.53-5.3)
3.2xlO~l(C.Z.C. <5)
1.6x10 ^C.E.C. 5-15)
8.0x10 2(C.E.C. >15)
Prioriti-
zation
Category
low
low
low
low
low
low
low
low
high
high
(0.5-9.2)-(.06-1.2) high
5.0 xlO~|-3.8x!02 low-high
2.5 xlO_2-t.9x!02 low-mod. high
1.3 xlO -95 low-moderate
(78-100)-9.8-13)
2.2xlO~2-9.4xlO-2
I.lxl0~2-4.7xl0~2
5.5xlO-3-2.4xlO-2
7.6-3.8xl03
7.0xl03(C.E.C. <5)
3.5xl03(C.E.C. 5-15)
1.8xl03(C.E.C. >15)
1.7xl02-2.8xl03(C.E.C. <:
83-1. 4xl03(r. E.G. 5-15)
42-7. Oxl02(C. E.G. >5)
low-high
high
low
low
low
low
5) low
low
low
Excess Area
Requirements
Factor
3.8xl05
1.9xl05
9.5X1011
83
1.2xl02
60
30
2.4xl02
1.1
none
none
9.5
4.8
2.4
2.5
none
1.9xl02
l.SxlO2
88
45
70
35
18
Comments
Conens. and volume generat ion rates
given in background document for 3
acid waste streams; distribution of
those streams assumed to be the
same among all 240 plants .
Electric Furnace Production
of Steel—Emission Control Dust Mn 3.8xl06 kg 7.6-3.8xl03 low 1.9xl02 Number of plants and component genera-
Dusts/Sludges tion rates calculated from data in
background documentation totaj U.S.
steel production, average plant
capacity, and amount of waste/unit
volume of product.
l.lxlO6 kg 2.2xl03-2.8xl02 low 55 Concentrations of Pb and Cr in
leachate given in background document ,
but no corresponding annual genera-
tion rate.
-------
Heavy Metals Screen of Hazardous Waste Streams for Land Treatment Suitability (continued)
Listed Waste Stream
Electric furnace production
of steel—emission control
dusts/sludges
Primary zinc smelting and
refining - Process waste-
water treatment sludge and/
or acid blowdown
Primary zinc smelting and
refining process waste-
water treatment sludge
and/or acid blowdown
Heavy
Metal
Zn
Dust
(cont'd)
Cu
Sludge Pb
Cr
Anode Pb
Slimes/
Sludges
Cd
Cadmium Pb
plant
residues
Cd
Pyrometal- Cd
lurgical
process acid
plant blow-
down sludge-
Plant A
Pb
Pyrometal- Cd
lurgical
process acid
plant blow-
down sludge- Pb
Plant B
Average Quantity
Generated per Area Required for Prioriti- Excess Area
Plant Land Treatment zation Requirements
(kg/10 yrs) (ha) Category Factor
2.3xl05 kg 9.2xl02(C.E.C. <5)
4.6xl02(C.E.C. 5-15)
2.3xl02(C.E.C. >15)
5.7X101* kg A.6xl02(C.E.C. <5)
2.3xl02(C.E.C. 5-15)
l.lx!02(C.E.C. >15)
2.4x10'' kg 48 (C.E.C. <5)
24 (C.E.C. 5-15) mod
12 (C.E.C. >15)
S.LxlO3 kg 16-2
2.1xl06-4.0xl06 kg 4.2-8.0xl03(C.E.C. <5)
2.1-4.0xl03(C.E.C. 5-15)
l.l-2.0x!03(C.E.C. >15)
280-3.3x10'* kg 56-6.6xl03(C.E.C. <5)
28-3.3xl03(C.E.C. 5-15)
14-1.7xl03(C.E.C. >15)
3.9xl05 kg 7.8xl02(C.E.C. <5)
3.9xl02(C.E.C. 5-15)
2.0xl02(C.E.C. >15)
low
low
low
low
low
low
low
. high
high
low
low
low
low
low
low
low
low
low
23
12
5.8
12
5.8
2.8
1.2
none
none
S.bclO2 kg 1.0xl02(e.E.C. <5) low
51 (C.E.C. 5-15) low
26 (C.E.C. >15) moderate
5.7xlO*-1.9xl05 kg 1.1-3. 6x10" (C.E.C. <5)
5.7xl03-1.8xlO'*(C.E.C.5-]5)
2/Oxl03-9.0xl03(C.E.C.>15)
3.8xl05 kg 7.6xl02(C.E.C. <5)
3.8xl02(C.E.C. 5-15)
1.9xl02(C.E.C. >15)
1.3xl03-4.0xl03 kg 2.6-8.0xl02(C.E.C. <5)
1.3-4.0xl02(C.E.C. 5-15)
65-2. Oxl02(C. B.C. >15)
8.6xl03 kg 19 (C.E.C. <5)
8.6 (C.E.C. 5-15)
4.3 (C.E.C. >15)
low
low
low
low
low
low
low
low
low
high
9.0xl02
4.5xl02
2.3xl02
19
9.5
4.8
20
10
5
none
Total annual quantities anode slimes/
sludges and cadmium plant residues
used for this calculation, since
number of plants generating these
kinds of wastes is not clear from
background document; number of plants
would have to be ^50 for anode slimes /
sludges and -5 for cadmium plant resi-
dues to satisfy per plant area con-
straint (i.e., ha) for land treatment
suitability.
Total quantity of acid plant blowdown
sludge and to contribution of 2 plants
to total quantity given in background
document used for this calculation;
proportion of total quantity of waste
accounted for by lime treatment sludger
and electrolyte process Held plnnt
blowdown sludges not clear from back-
ground document and assumed for pur-
poses of this calculation to be =0.
-------
Heavy Metals Screen of Hazardous Waste Streams for Land Treatment Suitability (continued)
Average Quantity
Generated per
Heavy Plant
Listed Waste Stream Metal (kg/ 10 yrs)
Electroplating and metal
finishing operations —
sludge Cr, Cd , Ni ?
tiludge from production Chrone yellow
of pigments and orange Cr l.lxlO1*-!.
6X101*
Pb 4.5xl05-5.6xl05
Molybdate Pb 3.0xl05-3.
orange
Cr l.lxlO1*-!.
Zinc yellow Zn 1.3xl05-l.
Cr 2.3xl05-ll
Chrome green Pb 3.6xl01|-4.
Cr 9.0xl03-l.
Chrome oxide
green Cr 2.3xl05-2.
7xl05
SxlO1*
6xl05
.4xl05
4x10"
1x10"
9xl05
kg
kg
kg
kg
kg
kg
kg
kg
kg
Area Required for Prioriti- Excess Area
Land Treatment zation Requirements
(ha) Category Factor
(260-320)-(33-40)
9.0xl02-l.lxl03
4.5-5.6xl02
2.3-2.8xl02
6.0-7.4xl02
3.0-3.7xl02
1.5-1.9xl02
(260-320)-(33-40)
5.2,6.4xl02(C.E.C
2.6-3.2xl02(C.E.C
1.3-1 .6xl02(C.E.C
1.3xl03-1.6xl03
1.6xl02 -2.0x102
72-88
36-44
18-22
(18-22) - (2-28)
(1.3xl03-1.6xl03)
low-high
low
low
low
low
low
low
high
. <5) low
. 5-15) low
. >15) low
low
low
low
mod. high- low
moderate
high
-
none
28
14
7
19
9.
4.
none
16
8
'<
40
5.
2.
1.
3
7
0
2
1
none
none
No data for quantity of waste
generated.
Quantity of waste generated given
in background document is for
entire chromium pigment industry
(70-73) low 40
Oven residue from
production of chrome Cr 1.2xl05-l.4xl05 kg (2.4-2.8xl02) - Quantity of waste generated given
oxide green pigments (30-35) low 7 in background documents is for
entire chromium pigment industry
-------
tions or large total amounts of one or more
waste constituents. Inspection of the "Ex-
cess Area Requirement Factors" column shows
that for most wastes the calculated land
area requirements are substantially in
excess of our 40 ha maximum criterion, with
factors ranging from one to several orders
of magnitude. Had these factors been on
the order of 2 - 3 or less, we may reason-
ably have concluded that waste pretreatment
or other pre-waste application alternatives
could be considered which might improve the
landtreatment potential for certain wastes.
Given the large volumes of most waste
streams and the substantial amount of pre-
treatment that would apparently be required,
these alternatives do not appear to be
economically justifiable and have not been
considered further.
Organics Screen
Over 70 organic compounds have been iden-
tified as major constituents among the waste
streams listed as hazardous by the EPA.
The most important criterion for the po-
tential effectiveness of land treatment
for organic chemical wastes is environmental
persistence. For screening purposes,
waste streams that contain one or more con-
stituents with half-lives in soil of more
than nine months were considered of low
priority for future research. The nine-
month cut-off point was derived as the
half-life necessary to rid a landtreatment
area of 99.99% of its waste application in
10 years so that applications in year one
would be degraded by the beginning of a 20-
year post-closure period. Waste streams
are scored for persistence as follows:
Priority for Research Half-Life in Soil
TABLE 3. PERSISTENCE OF ORGANIC HAZARDOUS
WASTE CONSTITUENTS IN SOIL
High
Moderately high
Moderate
Moderately low
Low
less than 7 days
1-4 weeks
1-5 months
5-9 months
greater than 9 months
Persistence ratings for a number of
hazardous waste constituents are given in
Table 3. These were derived from published
biodegradation data including half-lives,
hydrolysis, photolysis and biodegradation
rate constants, Tabak static culture flask
numbers, BOD5 data and BOD/COD ratios.
Organic Component
Acetonitrile
Acrolein
Acrylamide
Acrylonitrile
Benzal chloride
Benzanthracene
Benzofluoranthene
Benzopyrene
Benzotrichloride
Benzyl chloride
Carbon tetrachloride
Chlordane
Chloroacetaldehyde
Chlorobenzene
Chloroethers
Chloroform
Creosote
Dichloroethane
o-Dichlorobenzene
2,4-Dichlorophenol
2,6-Dichlorophenol
Dichloropropanol
m-Dinitrobenzene
2,4-Dinitrotoluene
Epichlorohydrin
Ethylchloride
Formaldehyde
Formic acid
Heptachlor
Hexachlorobenzene
Persistence
H
H
H
MH
H
L
L
L
H
H
L
L
M
H
H
MH
H
M
H
H
H
H
M
M
H
MH
MH
H
L
L
Hexachlorobutadiene MH
Hexachlorocyclopentadiene MH
Hexachloroethane ML
Hydrocyanic acid L.
Maleic anhydride H
Methylchloride MH
Methylene chloride MH
Napthaquinones H
Nitrobenzene M
Paraldehyde M
Phenol MH
Phorate M
Phosphorodithioic acid M
Phosphorodithioic esters M
Phosphorothioic exters M
175
-------
TABLE 3. PERSISTENCE OF ORGANIC HAZARDOUS
WASTE CONSTITUENTS IN SOIL
(concluded)
Organic Components Persistence
Picoline H
Phythalic anhydride M
Pyridine H
Quinones H
Tetrachloroethanes L
Tetrachloroethylene MH
Toluene H
Toluene diamines MH
Toluene diisocyanates H
Toxaphene L
Trichloroethanes L
Trichloroethylene H
o,o,o-Triethyle ester, M
Phosphorodithioic acid
2,4,6-Trichlorophenol MH
Trichloropropane H
Vinyl chloride H
Vinylidene chloride MH
The persistency rating of a waste stream was
assigned a low value if any constituent
with a half-life greater than nine months
was present. Streams that do not contain
constituents with tl/2 greater than 9
months receive a persistence score obtained
by weighted averaging the persistence scores
of all the stream constituents.
REFERENCES
1. Battelle-Columbus Laboratories,
"Assessment of Industrial Hazardous
Waste Practices: Electroplating and
Metal Finishing Industries—Job Shops,"
U.S. Environmental Protection Agency,
in preparation, to be distributed by
the National Technical Information
Service.
2. WAPORA, Inc., "Assessment of Industrial
Hazardous Waste Practices: Paint and
Allied Products Industry Contact
Solvent Reclaiming Operations, and
Factory Application of Coatings,"
Environmental Protection Publication
SW-119c, U.S. Environmental Protection
Agency, PB-251 669 (1976).
10.
WAPORA, Inc., "Assessment of Industrial
Hazardous Waste Practices—Special
Machines Manufacturing Industries,"
U.S. Environmental Protection Agency,
in preparation, to be distributed by
the National Technical Information
Service.
Jacobs Engineering Company, "Assessment
of Industrial Hazardous Waste Practices
in the Petroleum Refining Industry,"
U.S. Environmental Protection Agency,
in preparation, to be distributed by
the National Technical Information
Service.
Gruber, G. I., "Assessment of Industrial
Hazardous Waste Practices, Organic
Chemicals, Pesticides, and Explosives
Industries," Environmental Protection
Publication SW-118c, U.S. Environmental
Protection Agency, PB-251 307
(April 1975).
Versar, Incorporated, "Assessment of
Industrial Hazardous Waste Practices,
Storage and Primary Batteries Industries,"
Environmental Protection Agency, PB-241
204 (January 1975).
Calspan Corporation, "Assessment of
Industrial Hazardous Waste Practices
in the Metal Smelting and Refining
Industry," U.S. Environmental Pro-
tection Agency, in preparation, to be
distributed by the National Technical
Information Service.
Versar, Incorporated, "Assessment of
Industrial Hazardous Waste Practices,
Textiles Industry," U.S. Environmental
Protection Agency, in preparation, to
be distributed by the National Technical
Information Service.
Foster D. Snell, Inc., "Assessment of
Industrial Hazardous Waste Practices,
Rubber and Plastics Industry," U.S.
Environmental Protection Agency, in
preparation, to be distributed by the
National Technical Information Service.
Shaver, R. G., et al., "Assessment of
Industrial Hazardous Waste Practices:
Inorganic Chemicals Industry," Environ-
mental Protection Publication SW-104c,
U.S. Environmental Protection Agency
PB-244 832 (March 1975).
176
-------
11. SCS Engineers, Inc., "Assessment of
Industrial Hazardous Waste Practices—
Leather Tanning and Finishing Industry,"
U.S. Environmental Protection Agency, in
preparation, to be distributed by the
National Technical Information Service.
12. Arthur D. Little, Inc., "Pharmaceutical
Industry: Hazardous Waste Generation,
Treatment, and Disposal," Environmental
Protection Publication SW-508, U.S.
Environmental Protection Agency, 1976.
13. Swain, John W., Jr., "Assessment of
Industrial Hazardous Waste Management
Practices: Petroleum Rerefining
Industry," U.S. Environmental Protection
Agency, in preparation, to be distribut-
ed by the National Technical Information
Service.
14. WAPORA, Inc., "Assessment of Industrial
Hazardous Waste Practices—Electronic
Components Manufacturing Industry," U.S.
Environmental Protection Agency, in
preparation, to be distributed by the
National Technical Information Service.
177
-------
STATISTICAL ANALYSIS OF TRACE METAL
CONCENTRATIONS IN SOILS AT SELECTED LAND TREATMENT SITES
Michael A. Grossman
Bruce A. Goodwin
Paul M. Brenner
Arthur D. Little, Inc.
Cambridge, Massachusetts 02140
ABSTRACT
The problem of determining the sample size necessary to determine if a significant increase
in trace metal concentrations in waste application area soils relative to control area
soils has occurred is addressed. The paper discusses the sampling of two land treatment
sites, sampling plans, and hypothesis testing and sample size considerations. The use of
statistical techniques are demonstrated using data obtained from the sampling and chemical
analysis of soils from the two land treatment sites. The two sample t-test is recommended
as the most appropriate statistical test. Setting the minimum significant difference
equal to the sample standard deviation is shown to yield a useful criterion for determin-
ing the necessary number of soil samples. A modification of two stage sampling to the
periodic sampling of land treatment sites is also recommended.
INTRODUCTION
The goal of the analysis presented in
this paper is to determine if the concen-
tration of specific chemicals in the soil
has changed significantly from background
levels at land treatment sites. Geologic-
al, chemical and statistical analysis all
play an important role in determing if a
change has occurred. Geological analysis
must determine the homogeneity of the soil
conditions and how well each soil sample
represents the surrounding land area.
Chemical analysis of the soil samples must
measure the concentration of the chemicals
of interest precisely and determine the
tolerance limits of these measurements.
The statistical analysis must determine if
a significant change has occurred based on
the data provided by the prior two analy-
ses. This paper discusses the sampling of
two land treatment sites, sampling plans,
and hypothesis testing and sample size
considerations. The use of statistical
techniques are demonstrated using data ob-
tained from the sampling of the two land
treatment sites. Conclusions and recom-
mendations are made.
FIELD SAMPLING OF TWO LAND TREATMENT SITES
Two active industrial land treatment
operations were studied. The minimum cri-
teria for site selection were that the
sites should have identifiable and appropri-
ate control and waste application areas, and
that the wastes should be representative of
the range of waste compositions encountered
in actual practice. Control areas had to
be at least one acre in area to permit col-
lection of the desired minimum number of
samples for statistical analysis. Other-
wise, they were as nearly identical as pos-
sible to the waste application areas -ex-
cept that no waste had been applied.
The same minimum area requirement was
applied to waste application areas. In ad-
dition, reasonably accurate descriptions of
the types and amounts of wastes applied
were obtained from personnel familiar with
the site to assure that the area had not
been subject to other activities which
might influence the analysis. Candidate
waste types were not limited to wastes
which could be classified as hazardous ac-
cording to proposed federal regulations,
178
-------
since we wished to test the hypothesis that
assessment of the variability of waste con-
stituent concentrations does not necessari-
ly require measurement of a particular
species. Rather, information was obtained
on a group of chemical characteristics
which included species which were of par-
ticular interest because they were con-
sidered likely to serve as indicators of
waste constituent accumulation and/or mi-
gration.
The sites actually chosen for this
study satisfied all these criteria. Site
A was a plastics manufacturing facility
which landfarmed a centrifuged sludge from
an activated sludge wastewater treatment
system. The waste is known on the basis of
chemical analysis to contain percentage
concentrations of sodium. Waste composi-
tion was reported to be relatively constant
over time. Waste was applied by subsurface
injection and mixed further with the soil
by ordinary farm cultivation methods. Site
B^ was a petroleum refinery which landfarmed
sludge collected from wastewater holding
ponds. Considerable variation in the com-
position of the waste over time was con-
sidered likely by site personnel. Chem-
ical analysis of selected waste samples in-
dicated the percentage concentrations of
chromium were typically present in the
landfarmed material. Waste was applied by
gravity flow from vacuum transport trucks
and mixed further with the soil by tractor-
drawn rototiller.
Soil and core samples were collected
from both an application and a control area
and from two depths:
Depth 1: 0-6 inches below grade
Depth 2: 6-12 inches below grade
At Site A, the control area had 20 equally
spaced sampling locations taken approxi-
mately 45 feet apart from one acre; and the
application area had 60 sampling locations
taken approximately 45 feet apart from
three acres. At Site B, 20 equally spaced
soil samples were taken from both the ap-
plication and control areas. Sampling lo-
cations were approximately 45 feet apart
on one acre sampling areas. The patterns
of the sampling site locations for each
area were:
1. A triangular pattern for the con-
trol area of Site A;
2. A rectangular pattern for the three
other areas sampled.
A total of 240 soil core samples col-
lected from Sites A and B were analyzed and
results for the following seven constitu-
ents are reported later in the paper:
1. Chromium
2. Copper
3. Lead
4. Nickel
5. Sodium
6. Zinc
7. Conductivity
The six metals—chromium, copper, lead,
nickel, sodium, and zinc—were determined
in nitric-perchloric acid digests of all
soil samples using plasma emission spectro-
scopy. Conductivity was determined in a
1:1 soil-water extract using a portable
conductivity meter. Conductivities are re-
ported in units of micromhos per centimeter.
All other results are expressed as parts
per million in air-dry soil.
The results of all chemical analyses
and statistical calculations are available
from the authors upon request.
SAMPLING PLANS
In each land treatment site, there are
two areas of interest to be sampled: the
application (treatment) area and the con-
trol area. The object of the sampling plan
is to furnish soil samples representative
of each area in sufficient number to deter-
mine if a significant difference in concen-
tration between the two areas exists. Sev-
eral operations will be discussed and some
suggestions made for selecting from among
the sampling plans.
It is most important to remember that
the sampling plans and sample sizes dis-
cussed in this paper apply to each area
(application and control) of a land treat-
ment site. They are not designed for any
arbitrary land areas such as an acre. As
such, the land area they apply to is as-
sumed to be uniform in its geological and
chemical properties. If this assumption is
violated, the plan must be revised to pro-
vide a separate set of soil samples for
each homogeneous sub-area within the appli-
cation and control areas. It is important
to note that as long as geological and
chemical judgements show the soil condi-
tions to be uniform (homogeneous), the num-
ber of soil samples required will be a
179
-------
function of the statistical and chemical
concentrations independent of the size of
the area studied; i.e., the number of soil
samples for a homogeneous area is not a
function of land size (area) .
A large number of standard statistical
sampling techniques exist for selecting the
soil samples. A few of the most common
will be discussed briefly. Any statistic-
ally sound sampling plan which produces an
accurate estimate of chemical concentra-
tions in the soil within the required tol-
erances is valid.
Three of the most common plans which
are almost exclusively used in sampling
are:
1. Random Sampling - Sample sites are
selected from the area of interest
with equal probability. No strat-
ifications or subareas are crea-
ted.
2. Systematic Sampling - If n soil
samples are needed from an area,
the area is uniformly subdivided
and n equally spaced soil samples
are taken. This corresponds to
subdividing the area into n sub-
areas and taking one sample per
subarea. The size of the subarea
would be that of the total area
divided by n. Systematic sampling
is generally used for its ease and
facility of choosing sample mem-
bers when its use is appropriate.
3. Stratified Random and Stratified
Systematic Sampling - Two or more
subareas or strata are defined for
an area to be sampled. Within
each stratum the sample members
are then selected randomly if
stratified systematic sampling is
used.
Most soil sampling has been performed
as a single-stage sample. That is, all the
soil samples are taken at one time or with-
in a short period of time. A second meth-
od that may be advantageous when the chem-
ical concentrations and their variances are
unknown is two-stage sampling. With this
method, the sampling is conducted in two
parts. In the first stage a small pilot
study is conducted to obtain estimates of
the means and variances of the chemicals of
interest. Then, when a clear idea of the
sample size is known, a second, larger,
full-scale sample is conducted based on
revisions from the first stage. The re-
sults of both stages are pooled together
for the analysis.
A pilot study is useful when the mean
and variance of the quantity being measured
is unknown. With estimates from the pilot
study, a precise calculation of the sample
size can be attempted. Furthermore, a
pilot test gives the sampler a chance to
test out his sampling methods and make any
changes necessary to improve the ease and
efficiency of his sampling.
HYPOTHESIS TESTING AND SAMPLE SIZE CONSID-
ERATIONS
After a sampling method has been
chosen and the results of the pilot study
analyzed, the next step in sampling is to
determine the efficient (proper) sample
size. To answer this question, it is nec-
essary to study the relationship of the
sample size to the hypothesis being tested
and to decide (specify) the desired cer-
tainty of the answer. In the land treat-
ment site monitoring context, the hypothe-
sis tested by the monitoring program is:
the concentrations of the chemical species
in the soil are the same in the application
and control areas, against the alternative
that they are unequal. We wish to have
the probability of reaching the wrong con-
clusion about our hypothesis or its alter-
native to be very small. The probability
of incorrectly determining that the chem-
ical concentrations of the two areas are
unequal is denoted by the a-level; and the
probability of incorrectly determining that
the chemical concentrations are equal is
denoted by the g-level. The first kind of
error is known as Type I Error and the
second as Type II Error. If we denote our
hypothesis by HQ and the alternative by H-^,
Table 1 depicts the possibilities of the
outcome of the analysis.
In choosing the sample size, protec-
tion against both Type I and Type II Error
is sought. A common trap that must be
avoided is committing a Type II Error when
no difference between application and con-
trol areas has been found. This can occur
when too small a sample size has been cho-
sen; i.e., the sample size is too small to
show the significant difference that does
exist. In committing a Type II Error be-
cause of a small sample size, the experi-
180
-------
TABLE 1. POSSIBLE OUTCOMES OF STATISTICAL ANALYSIS
True Situation
Conclusion
of Analysis
HQ True
H False
H False
H1 True
H0 True
H- False
H False
H True
Conclusion Correct
Conclusion False
(Type I Error)
Conclusion False
(Type II Error)
Conclusion Correct
menter observes no significant difference
between application and control areas and,
therefore, concludes that the chemical con-
centrations in the two soils are equal al-
though the probability of that conclusion
being incorrect is quite large.
In choosing the a and g levels, one
would wish them to be very small. Typical-
ly, the a-level, the probability of commit-
ting a Type I Error (concluding the two
areas are different when, in fact, they are
the same) is chosen to be .05. One would
want the same level of protection against
the probability of committing a Type II Er-
ror (concluding the two areas are the same
when, in fact, they are different) and it
is reasonable to also set the 6-level equal
to .05. These a and 3 levels offer the ex-
perimenter adequate protection against er-
ror without being overly restrictive.
The next step in choosing the sample
size is to inspect the statistical test
used to differentiate between the two areas
and observe.what other quantities must be
specified or estimated from the data.
Since two areas are being compared for
their chemical concentrations in soil, the
statistical test used should be the two-
sample t-test.
The test statistic for the t-test is:
(1)
where: X = average concentration for the
application area
X = average concentration for the
control area
S 2 = the variance of the concentra-
tion in the application area
S 2 = the variance of the concentra-
tion in the control area
n = the number of soil samples
from the application area
n = the number of soil samples
from the control area
The problem at hand is to determine
and n
^.
The variances
and
S^2 are
calculated (estimated from the sample data
as are XA and X,,) .
The last parameter that must be spec-
ified before the final sample size can be
determined is_the minimum difference be-
tween XA and XQ which is to be considered
significant. In other words, how small a
difference between the chemical concentra-
tions in each area do we 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._ If the minimum
difference between X^ and X^ is specified,
the sample size for each homogeneous area
can be estimated using the formula:
™ p ~r
(2)
181
-------
where Un
and U-,
are the stan-
'1 - 8 """ °1 - a/2
dard normal deviates* for the indicated
probability levels and a is the standard
deviation. If a is unknown it may be es-
timated from the sample data.
Once the a and B levels have been spec
ified, the sample size is a function of the
right-hand factor in equation (2), i.e.,
(a \^. If this difference is de-
— _ — I noted_by d, where d =
A C / XA - XQ the sample size n
is now seen as a function of the ratio o/d.
Since we are dealing with variances and
means that are not known before the sam-
pling is conducted, it is simpler and more
sensible to state the minimum difference we
wish to be able to detect as a percentage
of the standard deviation, a.
Once the minimum detectable difference
is stated in terms of a percentage of the
standard deviation, the o/d ratio is de-
fined and the sample size is a straightfor-
ward calculation.
For example, suppose one wanted to be
able to detect a difference equal to
one standard deviation. The a and B levels
have already been specified as .05, so the
probability of committing a Type I or Type
II Error is set at .05. Then, the sample
size is found to be:
- B + U
= (1.645 + 1.96)
= 12.99
(1)'
Therefore, the sample size, n, is
rounded up to the nearest integer and is
chosen to be 13. Similarly, if one wanted
to be able to detect a difference equal to
half of the standard deviation, the sample
size would be four times greater or 52.
One easily sees that overly restrictive
specifications of the a/d ratio can lead to
huge sample sizes. If one wants to detect
a very small difference and 0 is large rel-
ative to d, then a large number of soil
samples will be required.
*Values from the Normal Probability Tables.
For example, U q7C. = 1.96 and U qc. =
1.645.
Common sense should be used in speci-
fying the 0/d ratio desired or d if an
estimate of 0 is known apriori, as
from a pilot test. For example, if one
wishes to set d equal to the chemical de-
tection limit (which is quite small),
while a relatively high concentration of the
chemical is present in the soil and the var-
iance is large, then astronomical sample
sizes will be necessary to detect minute
differences. Data from the two land treat-
ment sites discussed in the next section
will be used to illustrate the point.
The estimated detection limit for
chromium is 0.5 ppm. The estimated average
concentration of chromium at Site B was
264.65 ppm and the estimated standard devi-
ation was 334.63. Substituting these val-
ues into the sample size formula with a and
6 set to .05 yields a necessary sample size
in excess of 6.9 million soil samples per
homogeneous area. In addition to being
impractical, it is unnecessary to be inter-
ested in a difference of .5 ppm when the
average is over 250 ppm. As a matter of
fact, a highly significant difference be-
tween application and control area was dem-
onstrated with 20 soil samples per area. A
more sensible formulation would be to ex-
press the minimum observable difference as
a percentage of the mean or in terms of the
mean to standard deviation ratio.
For example, we might want the minimum
detectable difference to be 10% of the con-
trol area mean. This assumes that esti-
mates of the variance and application and
control area mean exists. Another specifi-
cation might be for the 0/d variance to
minimum significant difference ratio to be
specified as one or one-half. The decision
to specify the sample size in either of the
above two ways does not exempt the chemist
and the geologist from telling the statis-
tician how large a difference is signifi-
cant in physical terms.
Table 2 lists and Figure 1 displays
the sample size, n, as a function of the
0/d ratio for three combinations of protec-
tion from Type I and Type II Error (a and
B levels):
1. a = B = -05 (the suggested
specifications)
2. a
= .01
3. a = .05; B = .10
182
-------
3.00
0 .$ /.
Figure 1. Sample Size as a Fraction of the a/d Ratio for Three Combinations of a and
Levels
183
-------
TABLE 2. SAMPLE SIZE AS A FUNCTION OF THE a/d RATIO FOR
THREE COMBINATIONS OF a AND B LEVELS
1,
2,
a = [
n
6
11
24
54
96
151
217
603
868
542
410
3 = .
0
0
1
1
2
2
3
5
6
8
10
01
a
d
.5
.67
.0
.5
.0
.5
.0
.0
.0
.0
.0
a = (.
n
3
6
13
30
52
82
117
325
468
832
1,300
5 = .05
0
0
1
1
2
2
3
5
6
8
10
a
d
.5
.67
.0
.5
.0
.5
.0
.0
.0
.0
.0
a =
6 =
n
3
5
11
24
42
66
95
265
382
678
1,060
.05
.10
a
d
0.
0.
1.
1.
2.
2.
3.
5.
6.
8.
10.
5
67
0
5
0
5
0
0
0
0
0
The sample size has been calculated as if
there were a single estimate of a2 although
indications from the two land treatment
sites measured are that a^ and a^2 are un-
known before the pilot test and unequal.
This is known as the Aspin-Welch problem.
A solution is to calculate the estimated
sample size for each area, n, using a
pooled estimate of a, double the sample
size estimate to 2n and adjust the individ-
ual area sample sizes based on the ratio of
their standard deviations. The sample size
obtained in this manner will satisfy the
requirements for the specified a and 6 lev-
els and any error will be on the conserva-
tive side further reducing the probability
of Type I and Type II Error.
A final note on sample size. If a
two-stage sampling plan is used, the first
stage may be taken with n^ -tip. The sec-
ond stage may then be adjusted to yield the
required number of soil samples and the
proper ratio of control to application soil
samples. A second stage may not always be
necessary, but must be based on the results
of the first stage of the sampling plan.
STATISTICAL ANALYSIS OF SOIL SAMPLE RESULTS
The original analysis plan called for
performing a nested analysis of variance
(ANOVA) for each site to calculate the ef-
fects of both treatment and depth nested
within treatment. The descriptive statis-
tics, however, showed the cell standard
deviation to be highly variant. Therefore,
combining the variance estimates in an
ANOVA would violate the hypothesis of equal
cell variances and cause heteroscedasticity.
Rather than weighting the ANOVA to adjust
the variances, a two sample t-test was per-
formed comparing the average concentrations
between treatment and control for each
depth. The mean concentrations and conduc-
tivity for each depth at sites A and B are
displayed in Table 3.
The results of the t-tests are shown
in Table 4. It is interesting to note that
almost all of the application/control dif-
ferences were significant at the .05 level,
showing an almost complete separation be-
tween application and control areas. The
chemicals for which the t-tests were not
significant are underlined in Table 4. Any
differences between the variances for the
application and control areas was automat-
ically accounted for by the denominator of
the t-statistic. This was one of the rea-
sons that it was the ideal statistical test
to compare the mean concentration in the
application and control areas at land
treatment sites.
To obtain a better idea of the varia-
bility of the data relative to the mean
concentration of the chemicals, the coef-
ficient of variation ([standard deviation/
mean] x 100) was calculated for each chem-
ical concentration and conductivity at both
sites, at both depths, and for application
184
-------
TABLE 3. AVERAGE CONCENTRATIONS (PPM) FOR APPLICATION AND CONTROL AREAS
Copper
Nickel
Lead
Zinc
Sodium
Conductivity
Chromium
Application
Control
Application
Control
Application
Control
Application
Control
Application
Control
Application
Control
Application
Control
Site
Depth 1
8.98
11.26
10.17
9.49
9.78
8.51
52.79
48.49
340.85
252.20
535.42
212.50
18.27
17.53
1
Depth 2
10.54
9.61
10.37
10.99
9.36
8.12
53.19
50.61
347.32
283.55
432.50
216.15
21.69
20.80
Site
Depth 1
57.67
10.11
20.34
15.09
58.04
10.59
296.30
59.75
814.50
330.50
2581.00
243.25
757.35
47.50
2
Depth 2
22.93
10.57
17.98
17.61
20.12
10.57
105.33
59.52
936.65
530.45
2098.50
239.75
198.87
54.90
TABLE 4. t-TEST VALUES FOR TESTS OF
DIFFERENCES BETWEEN APPLICATION AND CONTROL AREAS
Chemical
Copper
Nickel
Lead
Zinc
Sodium
Conductivity
Chromium
Chemical
Copper
Nickel
Lead
Zinc
Sodium
Conductivity
Chromium
Site 2
Depth 1
0-6 Inches
9.91
6.79
6.34
11.67
5.62
8.52
11.18
Site 1
Depth 1
0-6 Inches
-.91
1.31
4.00
2.31
6.37
15.81
.34
Depth 1
6-12 Inches
4.43
.32
3.46
3.34
1.98
4.35
3.78
Depth 2
6-12 Inches
.76
-1.36
2.32
1.04
3.23
11.75
.36
NOTE; Positive values indicate higher application area concentrations than control areas.
All differences are significant at .05 level except those underlined.
185
-------
TABLE 5. COEFFICIENTS OF VARIATION EXPRESSED AS A PERCENTAGE; - x 100
A
Copper
Nickel
Lead
Zinc
Sodium
Conductivity
Chromium
Application
Control
Application
Control
Application
Control
Application
Control
Application
Control
Application
Control
Application
Control
Site
Depth 1
16
99
15
22
19
11
20
12
19
20
18
34
35
51
1
Depth 2
76
30
21
15
39
13
31
12
18
28
23
27
31
49
Site
Depth 1
37
8
14
12
58
10
30
22
40
61
47
59
37
14
2
Depth 2
54
11
24
16
61
12
55
33
87
79
91
41
85
20
and control areas. The coefficients of
variation are shown in Table 5.
No pattern for variability of the data
emerges. The question of whether variation
was proportional to the concentration was
raised; however, the entries in the table
are not consistent. The coefficient of
variation did not vary proportionally with
the chemical concentration. Furthermore,
there was no trend in the data by applica-
tion versus control, depth, or site. The
range for the coefficients of variation was
from 8 percent up to 99 percent. In con-
clusion, the coefficients of variation did
not shed any light on the diffusion of
chemicals in the soil or soil variability.
It also was assumed in the previous
analyses that samples taken anywhere in a
given plot at a given depth are all from
the same statistical population. However,
it could be postulated that the internal
area of a test plot may have more of less
concentration of chemicals, or a higher or
lower variation in concentration. This
possibility has been tested using the
same two sample t-test which studies dif-
ferences between application and control
areas and between depths.
The results for Site A and Site B are
displayed in Table 6. Each value is an ap-
proximate t statistic at a given depth for
a given test area. Positive values indi-
cate that perimeter concentrations averaged
higher than interior concentrations. At
the . 0_5 level, the null hypothesis*
(Xp = X^) was accepted for nearly all tests
performed, since the test statistics were
nearly all less than 2 in absolute value.
The instances where the differences
were apparently significant are infrequent
enough to fall within the bounds of random
fluctuation. In particular, a total of 56
t-tests were performed for the two sites
with 6 tests showing apparently significant
differences. This is only slightly above
the expected rate of false significance if
there really was no difference between
perimeter and interior samples. However,
there is evidence that lead and zinc con-
centrations in the interior of the Site 1
application area were higher than perimeter
concentrations.
*H,.
H:
perimeter concentration,_X =
interior concentration, X.^
xp y x±
186
-------
TABLE 6. t-TEST VALUES FOR TESTS OF
DIFFERENCES BETWEEN PERIMETER AND INTERIOR SAMPLES
Site 1
Depth 1 Depth 2
Site 2
Depth 1 Depth 2
Copper
Nickel
Lead
Zinc
Sodium
Conductivity
Chromium
Application
Control
Application
Control
Application
Control
Application
Control
Application
Control
Application
Control
Application
Control
-3.61
1.05
-1.43
1.20
-2.63
.65
-3.21
.52
1.13
.39
.74
-.04
-.16
1.21
-1.86
1.53
-1.00
.12
-2.58
.85
-2.37
.43
-.80
.49
.07
.62
.20
2.41
1.05
-.38
-.27
-.50
-.40
-.59
.60
-1.37
1.52
-.27
1.01
1.17
.71
-1.07
.62
-1.21
-.40
-.85
.31
-1.25
.06
-.38
1.65
-.21
1.18
-1.23
.78
-2.27
NOTE: Positive values indicate higher perimiter concentrations than in-
terior concentrations. Significant differences are underlined.
A visual inspection of the perimeter
versus the interior standard deviations
also indicate no systematic differences.
Apparently, little or none of the variation
in chemical concentration can be attributed
to differences between perimeter and in-
terior areas.
CONCLUSIONS AND RECOMMENDATIONS
1. Based on analysis of soil samples from
the two land farm sites, it is possible
to test for and measure significant
differences in chemical concentrations
between application and control areas
of each land farm.
2. It is possible to guard against com-
mitting the serious error of finding
no significant difference between ap-
plication and control areas when in
fact, one exists.
3. The proper statistical test to use is
a two sample t-test.
4. The sample size, the number of soil
samples from each area, can easily be
determined.
The crucial decision in choosing the
size (number of soil samples) is how
big a difference in concentrations
between the areas must be found before
it is considered significant. Of the
several possibilities available, such
as the detection limit, a percentage
of the mean, an arbitrary fixed amount,
or a percentage of the standard devia-
tion, choosing one standard deviation
as the minimum significant difference
is the most stable, simple to'use, and
efficient criterion. In all cases for
the two land treatment sites tested,
the one standard deviation criterion
demonstrates a significant difference
where one exists.
A modification of a systematic two-
stage sample appears to be the most ef-
ficient and economical method of sam-
pling. Starting with a sample size of
10 to 13 soil samples, the sample size
can be adjusted to the necessary number
of soil samples for the next round of
sampling.
187
-------
FACTORS INFLUENCING THE BIODEGRADATION
OF API SEPARATOR SLUDGES APPLIED TO SOILS
K. W. Brown, K. C. Donnelly, J. C. Thomas and L. E. Deuel, Jr.
Texas Agricultural Experiment Station
Soil and Crop Sciences Department
Texas A&M University
College Station, Texas 77843
ABSTRACT
Biodegradation of the organic components of hazardous industrial wastes is
essential for complete disposal. Information on the rates of degradation as a function of
edaphic and management factors is necessary in order to properly design a land treatment
system and to estimate the level of management and the length of time to achieve closure
following the last application.
The influence of various environmental parameters on biodegradation of a refinery
and a petrochemical waste was evaluated using a continuous flow soil respirometer.
Biodegradation rates of the two wastes were measured by collecting C02 evolved and
residual hydrocarbon analysis. The microbial population was determined six months after
incubation had begun. The environmental parameters studied included soil texture, soil
moisture, mineral nutrient amendments, application rates, application frequency and
temperature. Maximum degradation rates were achieved with the Norwood sandy clay at a
temperature of 30°C. The half life of the refinery sludge was 141 days, while that of
the petrochemical waste was 692 days. The addition of mineral nutrients was not
effective in increasing the rate of biodegradation of the refinery waste, but did
produce a small increase in the biodegradation rate of the petrochemical waste. Degrada-
tion was generally optimum at field capacity and decreased when the soil moisture content
was greater or lower than this level.
INTRODUCTION
Soil is increasingly being utilized
as a medium for the disposal of municipal
and industrial wastes. The objectives of
such disposal include the recycling of
water and nutrients, the decomposition of
organic fractions and the adsorption and
inactivation of environmentally undesir-
able constituents. While the effects of
addition of water, salts, and nutrients
to the soil is well understood, and there
is an increasing body of information on
the fate of metals applied to soils, much
less information is available on the
decomposition rate of added organics, or
the factors which influence these rates.
By optimizing application rates and
the controllable environmental parameters,
biodegradation of hydrocarbons by soil
microorganisms may be maximized, thus
making land treatment an environmentally
acceptable alternative for disposal of
oily sludge. Previous studies by Francke
and Clark (1974) reported degradation
rates up to 100 bbl./ac./mo. when used
crankcase oil was applied to fertilized
soils; however, a study by Raymond et al.
(1976) found rates of oil biodegradation
to be no greater than 20 bbl./ac./mo.
The rate and extent of the biodegrad-
ability of an oil sludge is dependent on
188
-------
the relative proportions of susceptible
and recalcitrant hydrocarbons (Bartha and
Atlas, 1973). The biodegradation rate of
an oily waste in soils was doubled with
the addition of nitrogen and phosphorus
(Kincannon, 1972) , while a separate study
by Raymond et al. (1976) showed that the
addition of fertilizer only increased oil
losses following prolonged incubation as
compared to unfertilized cases. In
addition, Dibble and Bartha (1979a) found
that higher sludge application rates favor
the biodegradation of the more recalci-
trant aromatic and asphaltic fractions.
The discrepancies between degradation
rates in the various studies is considered
to be a function of the varying composi-
tion of the oily sludge and the properties
of the receiving soil. In an attempt to
rectify these discrepancies, we have
utilized both a refinery and petrochemical
sludge to study the affect of several
environmental parameters on degradation of
oily sludges in four soils. The parameters
studied were soil texture, soil moisture,
incubation temperature, application rate,
application frequency and nutrient
additions.
EXPERIMENTAL DESIGN
API separator waste was collected
from a petroleum refinery and a petro-
chemical plant for use in a laboratory
biodegradation study. Separate experi-
ments were conducted by mixing each waste
with four different soils. The soils were
selected to represent a range of physical
characteristics. The soil-waste mixture
was placed in a controlled temperature
incubation chamber and incubated for
periods of 180 and 360 days.
Biodegradation rates of the two wastes
were studied at three temperatures,
three soil moisture contents, and with
varying amounts of mineral nutrients. In
addition, application rates and frequency
were varied to determine their effect on
the biodegradation of the oily waste.
Throughout the study period, CO,,
evolved from the waste amended soil was
adsorbed in columns of NaOH. Cumulative
C0~ evolution was determined at frequent
intervals by titration of the NaOH
solutions. At the end of each study, soil
was removed and analyzed for total carbon
or residual hydrocarbon. Selected soils
were also analyzed for microbial popula-
tions.
EXPERIMENTAL PROCEDURES
Soil
Four soils were selected to represent
a range of soil textures. They were
Norwood sandy clay (Typic Udifluvent),
Nacogdoches clay (Rhodic Paleudalf);
Lakeland sandy loam (Typic Quartzipsamment)
and Bastrop clay (Udic Paleustalf).
Characteristics of each soil are given in
Table 1. Detailed studies were conducted
using the Norwood sandy clay soil using
all levels of all the parameters consider-
ed. The other three soils were initiated
only at the optimum level of each control-
led parameter.
Two API separator sludges, one from
petroleum refinery and one from petro-
chemical plant were selected for this study.
Total carbon of the two wastes was
determined by wetcombustion
to be eleven and 35%,respectively. The
nitrogen content of each waste was
determined by theKjeldahl method described
by Bremner (1965). The refinery waste con-
tained 0.09% N while the petrochemical
waste had 0.11%. The refinery waste
contained 21.4% decantable liquid, while
the petrochemical waste was 12.5% decant-
able liquid. To facilitate handling and
assure that representative amounts of
these fractions of the waste were applied
to each treatment, the liquid phase was
decanted from each waste and each
component was measured separately, and
recombined prior to application.
Water content was determined by
distillation, and the solvent extractables
were quantified gravimetrically following
Soxhlet extraction with the respective
solvents. Pentane, benzene and dichloro-
methane were used to extract the saturates,
aromatics and asphaltics respectively.
Residues were heated at 750°C for ash
content. Wet combustion with potassium
dichromate and l^SO.tH-PO, yielded total C
by gravimetric analysis of C02 evolution.
The constituents of both wastes are given
in Table 2.
Selected soil samples were analyzed
for residual hydrocarbon content following
189
-------
TABLE 1. CHARACTERISTICS OF SOILS UTILIZED IN BIODEGRADATION EXPERIMENTS
8
Soil Series
Norwood
Nacogdoches
Lakeland
Bastrop
Sand
48.
41.
81.
60.
2
8
1
3
Silt
15.2
12.9
4.5
10.0
Clay
36.6
45.3
14.4
29.7
OM
1.4
1.3
0.7
1.0
CEC pH
19.6 7.69
17.2 7.95
0.3 6.45
27.4 6.86
N
Low
Low
Low
Low
P
94
3
20
25
K
— ppm -
312
164
200
200
Ca
4000
1280
600
2920
% Moisture
Mg WP FC SAT
485 12 18 33
400 8 20 36
100 1 5 23
385 6 22 25
-------
TABLE 2. CHARACTERISTICS OF THE SLUDGES
Extractable CLASS % OF TOTAL HYDROCARBON
Oil Carbon Ash Water Pentane Benzene Dichloro-
Sludge g C/lOOg methane
Refinery 10 41 46 72 22
Petrochemical 62 25 12 63 36
incubation in the soil respirometer. A
25 g sample was mixed in a Waring blender
with 25 g anhydrous Na-SO,. To this,
150 ml of diethyl ether was added followed
by blending for one minute. The ether was
decanted and filtered using a Buchner
funnel and Whatman No. 42 filter paper.
This extraction was repeated twice with
50 ml diethyl ether and transferred to
another Buchner funnel. The extract was
refiltered until all soil particles were
removed, and then transferred with rinsing
into a tared round-bottom flask. The
solvent was evaporated in a Brinkman-Buchi
Model Roto-Vapor-R* and the extractable
hydrocarbons determined gravimetrically.
Biodegradation
The rates of biodegradation were
determined in respirometers in which the
soils were incubated at controlled
temperature. A continuous stream of C02 -
free air passed over the soil in the
incubation flasks, and the evolved CO,, was
collected from the air leaving each flask
in columns containing 0.1N NaOH. A schema-
tic diagram of the apparatus is shown in
Figure 1. Generally, each treatment was
replicated.
Apparatus consisting of concentrated
H2SO, , 4N NaOH and C02~ free water was
used to remove CO- and other carbon com-
pounds and to provide a water saturated
air stream for each incubation flask
(Stotzky, 19bO; Stotzky, 1965). Twelve
flasks were placed in each controlled
tenparature incubation chamber, 18 x 34 in.
constructed from 3/4" styrofoam. A coil
of copper tubing connected to a water bath
was used to provide temperature control.
For experiments run at 10°C, sand was used
Trap cone HJ.SQ, Trap
Trop
Mention of brand name does not consti-
tute endorsement.
Soil and Oil OIN NgOH
Incubation Chambtr
Figure 1. Schematic diagram of a
respirometer.
in the bottom of the chamber to provide
thermal contact with the soils.
The equivalent of 100 g of dry soil
was crushed on a glass plate. A 20-25 g
subsample of the soil was placed in an
evaporating dish and the waste added on a
percent weight basis. After thorough
mixing, the total soil sample was mixed
and placed in a pre-weighed 500 ml
Erlenmeyer flask. The exhaust air passed
through an NaOH trap. The accumulated
C0? was determined by titration (Stotzky,
1965).
Soil Moisture
Soil samples above field capacity
were dried to reach field capacity, while
samples collected below field capacity
were wetted with the required amount of
distilled water to reach field capacity
prior to waste application. Field
capacity for the four soils was 18%, 5%,
22% and 20% for the Norwood, Lakeland,
Bastrop and Nacogdoches, respectively. In
addition, samples of the Norwood soil were
tested at moisture contents of 33% and
12%. The incubation flasks were reweighed
191
-------
periodically, and moisture contents were
adjusted as needed.
lenperature Effects
The effect of temperature on degra-
dation of the two sludges was studied by
incubation of the soil-sludge mixtures at
10, 30 and 40°C. Each waste was studied
at five loading rates at each temperature.
Application Rates
Application of oil sludge to soils was
studied at treatment rates of 0.1,0.5,1.0,5,
and 10 grams of sludge per 100 grams dry
weight of soil, equivalent to a hydrocarbon
loading of 11, 55, 110, 550 and 1100 mg
refinery or 35, 175, 350, 1750 and 3500 mg
petrochemical waste. Control soils to
which no sludge had been applied were
included in each study to monitor the
respiration rate of the indigenous soil
microorganisms.
Application Frequency
Two experiments were conducted to
determine the effect of repeated applica-
tions of waste sludge on the degradation
rates. In the first experiment, reapplica-
tions of the petrochemical waste were made
following 180 days of decomposition,
corresponding to the initial application
rates of 0.1, 0.5, 1 and 5 g per 100 g soil.
In the second experiment, reapplications
were made at a rate of 5 g per 100 g
soil 45, 90 and 135 days after the initial
application to two flasks of each waste and
90 days after the initial application to
one flask of each waste. One flask of each
waste served as a control receiving 5 g per
100 g soil at the beginning of the experi-
ment .
Mineral Nutrients
The C:N ratio of the refinery and
petrochemical wastes was determined to be
110:1 and 350:1 respectively. In an
attempt to improve the microbial activity,
mineral nutrients were added in the form of
Ca(N03)2, KC1 and Ca3(P04)2.These were mixed
with the soil prior co addition of the
wastes to achieve the desired treatment
ratios. The refinery wastes were tested
at C:N ratios of 100:1, 50:1 and 10:^and
C:K and C:P ratios were 50:1. In addition,
one flask was supplied with N, P, and K at
ratios of 50:1. The C:N ratios of the
petrochemical wastes studied were at 300:1,
150:1 and 20:1. The C:P and C:K ratios of
the soils in two flasks were adjusted to
150:1 and the soils in another flask were
adjusted to 150:1 for all their nutrients.
RESULTS AND DISCUSSION
Initial results indicated that a flow
rate of at least 20 ml of air per minute
per 100 grams of soil was required to
prevent depletion of oxygen in the incuba-
tion flasks. Reproducibility was increased
when each flask was removed at 45 day
intervals and the soil moisture content
determined by weighing. When the moisture
content of the soil fell below 10% or
increased to above 35%, the respiration
rate deviated from replications in which
the moisture content had not changed, and
such data were rejected.
Soils Comparison
Four soils were studied to determine
their utility as a site for the landfarm-
ing of oily sludges. Each was incubated
at 30°C at field capacity moisture
content. The results are presented in
Figures 2 and 3 and in Table 3. For both
wastes, the greatest amount of degradation
occurred in Norwood soil, while degrada-
tion in the Nacogdoches and Lakeland soils
was intermediate, with the least amount
occurring in the Bastrop soil. The amount
of carbon evolved as CC^ from petrochemical
,»*
Q 30 CO 9O 120 «0 *0
Tlme(Days)
Figure 2. The effect of soil texture on C07
evolution from soils amended with 52
(vrt./wt.) refinery waste and incubated
for 180 days at field capacity and
30°C.
sludge-amended Bastrop clay was only 7 mg
greater than from the unamended soil, this
is equivalent to a decrease of only 0.4%
total carbon. Twenty-five percent of the
refinery, and 5% of the petrochemical
waste decomposed from the Lakeland soil
192
-------
ilia:
T«ne(Devs>
Figure 3. The effect of soil texture on CO
evolution from satis amended witl 51
(vt./vt.) petrochemical waste and
Incubated for 180 days at field capacity
at 30°C.
during the 180-day incubation period;
while 29% and 10% of the refinery and petro-
chemical sludges, respectively, were
decomposed in the Nacogdoches soil. The
order of decomposition was not influenced
by the pH of the soils. This may have
been due to the random range of pll of the
soils.
Soil Moisture
The influence of soil moisture on the
average biodegradation rates at 30 and
40°C is given in Figures 4 and 5. For
both wastes, the moisture content had a
8'
0>
E 3
a 33
• e
4 12
0 50 tX) 150 200
Time (Days)
Figure 4. The influence of soil moisture on
CCL evolution from Norwood soil
amended with 1% (wt./wt.)
refinery waste and incubated for
180 days at 35°C.
<5
o
• .
.533
•»
0 50 WO 150 200
Time (Days)
Figure 5. The influence of soil moisture on CO.
evolution from Norwood soil amended
with 1% (wt./wt.) petrochemical waste
and incubated for 180 days at 35°C.
TABLE 3. EFFECT OF SOILS ON THE BIODEGRADATION OF BOTH WASTES
APPLIED AT 5% AT FIELD CAPACITY, INCUBATED FOR 180
DAYS AT 30°C
Soil
Refinery
Norwood
Nacogdoches
Lakeland
Bastrop
Petrochemical
Norwood
Nacogdoches
Lakeland
Bastrop
Carbon
Applied
(mg)
550
550
550
550
1750
1750
1750
1750
Determined by (XL Evolution
Carbon
Degraded
(mg)
218
158
136
134
293
173
92
7
Carbon
Degraded Percent
of Waste Applied
40
29
25
24
17
10
5
0.4
193
-------
greater relative influence on the rate of
biodegradation at 10°C than it did at the
higher temperatures. At both temperature
regimes, the refinery waste exhibited
maximum degradation at 18% moisture which
represented field capacity. The biodegra-
dation rates were less at both wetter and
drier moisture contents. For the petro-
chemical waste, the maximum biodegradation
rates occurred at the lowest moisture
content at the low temperature and at the
highest moisture content for the high
temperature. Thus, while field capacity
appeared to be the optimum soil moisture
for the refinery waste, no such optimum
was found for the petrochemical waste. For
both wastes, there appeared to be a rather
broad range of soil moisture contents
including conditions both wetter and drier
than field capacity which had minimal
influence on the waste decomposition. Only
at excessively wet or dry conditions did
the moisture content become a dominant
factor.
Temperature
The results of experiments to deter-
mine the effect of temperature on the
biodegradation of wastes applied at a rate
of 5% are shown in Figures 6 and 7. The
evolution of carbon dioxide from soils
treated with both wastes increased with
temperature through 40 C. Increases in the
respiration rate of the unamended soils
between 30 and 40°C were greater than the
increase in the rate of the soil to which
waste had been added. Therefore, a com-
parison of the percentages of waste
degraded at the various temperatures
reveals a reduction in percentage of carbon
degraded from the waste which was incubated
at 400C. Biodegradation was calculated
both from COj evolution and residual carbon
measured by wet combustion. The wet com-
bustion data will be used here as a basis
for discussion. At 30°C, 45% of the
refinery waste, and 31% of the petrochemi-
cal waste were degraded, while at 40°C
only 36 and 23% of the refinery and petro-
chemical wastes respectively were degraded.
At 10°C, respired CO,, was equivalent to
only 18 and 14% above the unamended soils
for the refinery and petrochemical waste
respectively.
The biodegradation rate generally
doubled between 10 and 30°C but decreased
again at 40°C indicating some inhibition.
These results are in general agreement with
la
C
o
2 2
I
30°C
• e°c
100
Time (Days)
Figure 6. Effect of temperature on CO evolution
from Norwood soil amended with 5% (wt./wt.)
refinery waste and incubated for 180 days.
4 40°C
4 30"C
• 10°C
Time (Days)
Figure 7. Effect of temperature on CO- evolution
from Norwood soil amended with 57. (wt./wt.)
petrochemical waste and incubated for 180
days.
those reported by Dibble and Bartha (1979a),
who reported the optimum temperature for
degradation, of an oily sludge to be about
20°C.
Application Rates
The evolution of (X>2 from soils
amended with either waste increased with
increasing application rates shown in
Figures 8 and 9 and Table 4. Since data
were available from both 30 and 40°C and
differed only a little, the data averaged
over these two treatments are shown here.
Biodegradation of both wastes
increased with increasing application. Of
the two wastes investigated, a much larger
fraction of the carbon applied in the
refinery waste was released as CX^. An
average of 33.4 mg of the carbon applied
in the refinery waste was respired, while
only 13.4 mg of the petrochemical waste
carbon was released.
When comparisons are made strictly on
the basis of the amount of waste applied,
194
-------
• OJ <» lowing
O OS •
Time (Days)
Figure 8. The Influence of loading rate on the
biodegradation of refinery vaate applied
to Norwood soil and Incubated for 180 days
at 30°C and 18Z laoisture.
ftdudge/IOOq sol
• at one tooting
o 05 •• "
a 10 " "
a so ••
A»O "
A 200 four barings
Tine (Days)
Figure 9. The Influence of loading rate on the
biodegradation of petrochemical waste
applied to Norwood soil and incubated
for 180 days at 30°C and 18Z moisture.
TABLE 4. THE INFLUENCE OF LOADING RATE ON THE BIODEGRADATION OF WASTE
APPLIED TO NORWOOD SOIL INCUBATED AT 35°C AND 18% MOISTURE
FOR 180 DAYS
Sludge
Applied
C
Applied
(mg)
C0?
Evolution
(mg)
Refinery Sludge
1 Loading
Petrochemical Sludge
0.1
0.5
1.0
5.0
10.0
10
55
110
550
1,100
as shown in Figure 10, it is evident that
a slightly greater percentage of the petro-
chemical waste was degraded, particularly
at the greater application rates. The
respiration is nearly linear at low rates,
but appears to be approaching a plateau
at the greater rates. This indicates
that inhibiting mechanisms decreased the
fraction of the applied carbon which was
respired.
The greatest fraction of the refinery
waste respired when 0.5% was applied to
the soil, while the greatest fraction of
the petrochemical waste was released as
C02 at the 1% application rate. The
fractions respired do not decrease
10
56
34
40
27
1 Loading
u
M
II
0.1
0.5
1.0
5.0
10.0
40
180
350
1,750
3,500
12
10
19
17
9
10. Th* affect of application
of nflaary and i
195
-------
greatly at 5% of each waste, but fall at
10% ; therefore, applications of between
1 and 5% appear to encompass the maximum
degradation rates.
Mineral Nutrients
The influence of nutrient addition on
the respiration rates is shown in
Figures 11 and 12. The two wastes differ-
ed greatly in their C:i2 evolution (Table 5).
The addition of nitrogen at a ratio of
10:1, C:H, increased the rate of biodegra-
dation for the refinery waste by 6%. The
addition of all three mineral nutrients at
a rate of 150:1 C:NPK increased the rate
of biodegradation for the petrochemical
waste by 5%.
Thus , the decomposition rate of the
refinery waste was increased slightly by
adjusting the C:N ratio from 110:1 to 10:1.
The nitrogen deficient petro-
chemical waste exhibited an increase in
decomposition when the C:N ratio was
adjusted from 350:1 to 150:1, but did not
differ from the unfertilized rate when
additional nitrogen was added to lower the
C:N ratio to 20:1. The addition of P and
K alone or with nitrogen did not increase
the respiration rates from soils amended
with either waste. The lack of increased
biodegradation of the petrochemical waste
at C:N ratios narrower than 150:1 indicates
that factors other than nitrogen were
limiting breakdown of the waste.
O
(J
0,3
o Control (soil only)
a Soil, Waste, N IC« 350-t)
A NPK (ISO 1)
A P (150 1)
• K (ISO I)
• Soil, Waste, N 1C N 601)
x Sal.Waste.N [CN 201)
Time (Days)
Figure 12. The influence of nutrient additions on the
cumulative respiration of Norwood soil
amended with 52 (wt./wt.) petrochemical
waste and incubated for 180 days at 30°C
and i8% moisture.
Other researchers have also reported
that the addition of mineral nutrients to
waste amended soil offset nutritional
imbalances, thus providing a more suitable
environment for the microbial utilization
of petroleum compounds. Increases in the
rate of waste biodegradation have been
reported by Cook and Westlake (1974),
Dotson et al. (1971), Maunder and Ward
(1975), and Dibble and Bartha (1979b).When
several different oily wastes were applied
to soils, Kincannon (1972) found that the
rate of biodegradation was doubled when
nitrogen and phosphorus were added to the
soil with the waste. In a field study on
waste degradation in soil, Raymond et al.
(1976) did not find an increase in the rate
of biodegradation in fertilized plots until
196
-------
TABLE 5. THE EFFECT OF MINERAL NUTRIENTS ON RESIDUAL CARBON IN WASTE
AMENDED SOIL AFTER 180 DAYS INCUBATION
Mineral Nutrients Residual Carbon
Ether Extract
Refinery Sludge: Total carbon applied
None
C:P 50:1
C:NPK 50:1
C:N 10:1
(gm)
0.55 gm
0.13
0.12
0.14
0.10
Carbon
Degraded
(%)
76
78
75
82
Petrochemical Sludge: Total carbon applied 1.75
C:P
None
50:1
C:NPK150:1
C:N
20:1
0.
0.
0.
0.
88
88
78
88
50
50
55
50
a year after waste application. Dibble and
Bartha (1979a) reported that excessive
rates of mineral nutrient addition did not
favor waste biodegradation. Atlas (1977)
suggests that the best fertilizer for soil
application is a form of readily usable
nitrogen and phosphorus along with a form
of slow release fertilizer to prevent
leaching losses.
Thus the present data and that
reported in the literature indicates that
additions of the proper amounts of select-
ed nutrients can increase biodegradation,
while excessive amounts provide no benefit.
Most of the results have indicated that
nitrogen is the most needed nutrient and
that additions are dependent on the waste
characteristics. Dibble and Bartha (1976b)
found an optimum C:N ratio of 60:1 for
their oily waste, while one of those consi-
dered here exhibited an optimum of 10:1 and
the other at 150:1. It would appear that
optimum nutrient additions will need to be
determined for each waste. Care must also
be taken, however, to avoid additions of
excess nitrogen fertilizer which could
contribute to leaching of nitrates.
CONCLUSIONS
The rates of biodegradation of both
wastes depended greatly on the soil to
which they had been applied. Maximum bio-
degradation was found in the Norwood sandy
clay soil, while the minimum was found
in the Bastrop clay soil, when both
were maintained at field capacity moisture
content. The soils had a minimum range of
pH values, and respiration rates were not
correlated with this parameter. Nutri-
tional differences could have been a
factor, but fertilizer additions were not
evaluated on all soils.
The soil moisture content influenced
the biodegradation rates of both wastes.
Field capacity was the optimum moisture
content for the petrochemical waste ,depend-
ent to some extent on the incubation
temperature. For both wastes, there
appeared to be a broad range of moisture
contents including conditions both wetter
and drier than field capacity which had a
minimum influence on the rate of biode-
gradation. The rates decreased when the
soil become either too wet or too dry.
The biodegradation of both wastes
increased with application rates.
Inhibitory mechanisms however resulted in
smaller increases in biodegradatior. per
amount of waste applied between the 5 and
10% application rates.Biodegradation of the
applied wastes increased through 30°C and
decreased slightly at 40°C. The biodegra-
dation rate at 10°C was about half of the
maximum found at 30°C.
The rates of C02 evolution were
similar for both wastes at the same appli-
cation rates, but since the petrochemical
waste contained almost four times more
carbon than the refinery waste, the frac-
tion of applied carbon respired was much
less for the petrochemical waste.
197
-------
Respiration rates for both wastes applied
to the soil increased with application
rate, but the increases were less than
proportional to the amounts applied at the
5 and 10% application rates.
Small frequent applications of both
wastes increased respiration rates per unit
applied over that obtained on a single
equivalent application. While repeated
applications resulted in a greater total
respiration, the fraction of the applied
carbon which was released as COo decreased
at increasing application rates. Thus,
biodegradation rates of subsequent applica-
tions per amount applied were less than
those of the first applications.
Adjustments in the C:1I ratio by the
addition of fertilizer nitrogen increased
the rate of biodegradation. The petro-
chemical waste, which contained a large
fraction of carbon, exhibited the greatest
degradation rate at a C:N ratio of 150:1;
while the refinery waste, which contained
less carbon and more nitrogen, exhibited
a small increase in biodegradation rate
when the C:N ratio was adjusted from 111:1,
the original, to 10:1. Additions of P and
K either alone or with nitrogen did not
increase the rates of biodegradation of
either waste.
Microbial analysis revealed that both
hydrocarbon utilizing bacteria and fungi
are present in the unamended soils.
Additions of either waste resulted in
increased populations of hydrocarbon
utilizing bacteria and fungi. The great-
est populations of both were found in
soil to which 1% by weight of the sludge
had been added. Populations were
diminished at greater application rates,
and at the 10% level, the population
generally decreased below levels found in
unamended soils.
REFERENCES
1. Atlas, R. M. 1977. Simulated Petro-
leum Biodegradation. Grit. Rev. Micro-
biol. 5: 271-386.
2. Bartha, R. and R. M. Atlas. 1973.
Degradation of oil in seawater: Limit-
ting factors and artificial stimulation
in the microbial degradation of oil
pollutants. D. G. Ahearn and S.P.
Meyers (ed.). Center for Wetland
10.
11.
12.
Resources, Louisiana State Univ.,
Baton Rouge, LA. Publ. //LSU-S. G-73.01.
Bremner, J. M. 1965. Total Nitrogen.
In Methods of Soil Analysis, Chemical
and Microbiological Properties, Part 2.
C. A. Black (ed.) pp. 1149-1176.
American Society of Agronomy.
Cook, F. D., and D. W. S. Westlake.
1974. Microbial Degradation of North-
ern Crude Oils. Information Canada
Cat. No. R72-12774.
Dibble, J. T. and R0 Bartha. 1979a.
Effect of environmental parameters on
biodegradation of oil sludge. Appl.
Environ. Microbiol. 37: 729-738.
Dibble, J, T. and R. Bartha. 1979b.
Leaching aspects of oil sludge
biodegradation in soil. Soil Sci. 127:
365-370.
Dotson, G. K., R. B. Dean, W. B. Cooke
and B^ A. Kenner. 1971. Land spread-
ing, a conserving and non-polluting
method of disposing of oily wastes.
In Proc. 5th Int. Water Pollution
Research Conf. Pergamore Press,
New York.
Francke, H.C. and F. E. Clark. 1974.
Disposal of oily waste by microbial
assimilation. U.S. Atomic Energy
Report Y-1934.
Kincannon, C. B. 1972. Oily Waste
Disposal by Soil Cultivation Process.
Report EPA-R2-72-100. U.S. Environ-
mental Protection Agency, Washington
D.C.
Maunder, B. R. and J. S. Ward.. 1975.
Disposal of Waste Oil by Landspreading.
In 3rd Int. Biodegradation Symposium.
August 17-23, 1975. Univ. of Rhode
Island, Kingston, TS XXV-5.
Raymond, R. L., J. 0. Hudson, and
V. W. Jamison. 1976. Oil Degradation
in Soil. Appl. Environ. Microbiol. 30:
522.
Stotzky, G. 1960. A Simple Method
for the Determination of the Respira-
tory Quotient of Soils. Can. J.
Microbiol. 6: 439-4520
198
-------
13o Stotzky, G. 19650 Microbial
Respiration. In Methods of Soil
Analysis, Chemical and Microbiological
Properties. C. A. Black (ed.). Part 2,
pp. 1550-1572. American Society of
Agronomy.
199
-------
REVIEW AND PRELIMINARY STUDIES OF INDUSTRIAL LAND TREATMENT PRACTICES
Rufus L. Chaney
Sharon B. Hornick
Lawrence J. Sikora
USDA-SEA-AR
Biological Waste Management & Organic Resources Laboratory
Beltsville, Maryland 20705
ABSTRACT
This paper discusses the planned efforts of the Biological Waste Management and
Organic Resources Laboratory to conduct research on the land treatment of various
hazardous wastes through an Interagency Agreement with the U.S.E.P.A. The effort began
with a critical review of information relevant to land treatment, with the first draft
now completed. The overall plans of the Agreement includes research on the fate and
effect of organic and inorganic waste constituents in the environment with studies in
the greenhouse, in the laboratory, on new field plots, and on field plots at existing
land treatment sites.
A review of natural controls which limit the adverse effects of heavy metals in
wastes is also presented. For most elements, the "Soil-Plant Barrier" protects the food
chain. However, humans or livestock can suffer health effects from mismanaged wastes
containing elevated levels of Cd, Se, Mo, Pb, and Co. Lastly, as an example, a
discussion of Cr in hazardous wastes, soils, plants, and animals is presented along with
the specific consideration of land treatment of leather manufacturing wastes.
Introduction
In March 1980, our laboratory began a
5-year scientific study concerning the
land treatment of hazardous wastes under
an Interagency Agreement (IAG) with the
U.S.E.P.A. Since 1971, our laboratory
has studied various aspects involving the
land application of sewage sludge. Some
of these areas are: trenching; com-
posting; the utilization of municipal
wastes on cropland, gardens, and
pastures; and the fate of nutrients,
pathogens, toxic organics, and heavy
metals in the environment. The aerated
pile method of controlled composting was
also developed in our group. Our inter-
disciplinary research program involved
engineers, microbiologists, and soil and
plant scientists.
Because of this research experience,
EPA decided that our interdisciplinary
approach in the study of land treatment
of wastes could assist them in the
development of land treatment technology
and regulations. Although pesticide,
petroleum, and other "hazardous" wastes
have been land-applied for over 25 years,
little scientific data is available to
assess the efficacy and environmental
acceptability of land treatment.
Scope of Research Program
In order to perform cooperative
research with our laboratory, an IAG
which included 6 Tasks was developed.
The Tasks which involved many hazardous
constituents were as follows:
200
-------
Task 1: Compile, review, and criti-
cally evaluate available data on land
treatment of wastes for application to
hazardous wastes. A 450 page draft
report (20) was submitted to EPA in
January 1981. The report reviewed
processes in waste treatment, soils,
plants, and animals which can be affected
by certain constituents or waste types
which are land spread. Management and
research problems were also addressed for
each waste.
Task 2: Evaluate treatment of selec-
ted hazardous wastes in soils, and
effects on soils and plants in the green-
house and laboratory. This task will
allow our team to study the fate and
effect of both organic and inorganic
hazardous constituents, the transfor-
mations of bulk organic matter and plant
nutrients, the relative tolerance of
crops to heavy metals and toxic organics,
and bioassay for site management.
Hazardous wastes to be studied will
depend on EPA's agreement with our con-
clusions in our Task 1 report, and on the
recommendations from a report from A. D.
Little which evaluated the land-treata-
bility of identified hazardous wastes.
Task 3: Evaluate land treatment of
selected hazardous wastes in field
studies to verify the greenhouse and
laboratory investigations. Research of
this task will allow the field verifi-
cation of results from Task 2. Here
wastes which have a high probability of
being land-treatable and/or have a high
need for an alternative disposal method
can be studied.
Task 4: Evaluate crop species and
cultivars to exclude or accumulate
pollutants from hazardous waste-amended
soils. Some crops tolerate soil condi-
tions where organic matter is rapidly
oxidized, or soils rich in toxic organics
or heavy metals. Other crops can hyper-
accumulate metals, and possibly have the
ash of the crop biomass sold as an ore.
Still other crops may be shown to
tolerate the land treatment of wastes and
thus be utilized as a cover crop during
the operational period.
Task 5: Evaluate crops on existing
active and closed land treatment sites.
Some treatment processes are time depen-
dent; others depend on level of organic
matter in the soil. Thus, examining
sites used a long time ago, or over a
long term, provides information unavail-
able through new experimental plots.
Task 6: Determine the feasibility of
combining waste streams, and processing
wastes (e.g., by composting) to improve
land treatment of hazardous wastes.
Wastes with a high C:N ratio are dif-
ficult to manage on land. Some wastes
contain hazardous constituents which are
volatile, or are very weakly adsorbed and
thus Teachable. Pretreatment by com-
posting may prevent many management
problems, and the residue may be applied
on hazardous waste land treatment sites
(32), or on RCRA 4004 solid waste treat-
ment sites (30) if the hazardous
constituent is sufficiently degraded.
Task 1 Report
The Task 1 report (20) covers many
items in detail, and should help land
treatment site managers and planners,
researchers, and regulatory officials to
understand the scientific principles upon
which land treatment is based. It is
impossible to present a brief summary of
the whole report in this paper.
On the other hand, if one selected
waste were discussed, the reader would
have an idea of the approach of our
report and know what information to
expect when the final report becomes
available. We have, therefore, selected
Cr and leather manufacturing wastes for
discussion in this paper as an example of
our approach. The literature has pro-
vided considerable information, and
allowed us to identify the remaining
research needed to support the land
treatment of Cr-rich wastes.
Routes for Environmental Impacts
Research has shown that the
food-chain can be influenced by a
waste-applied hazardous constituent
through: 1) direct ingestion of the
waste, 2) ingestion of spray-applied
waste adhering to crops, 3) ingestion of
the waste lying on the soil surface or
waste-amended surface soil, 4) ingestion
of crops which accumulated the hazardous
constituent from soil, 5) ingestion of
animal tissues enriched in the hazardous
constituent after the animal consumed the
201
-------
crops or wastes, or 6) ingestion of
groundwater enriched in the hazardous
constituent through leaching from the
waste amended soil.
A committee from the National
Research Council (61) recently the
literature on mineral tolerance of
domestic animals (Table 1). If land
application of wastes could cause an
element to be higher in forage crops than
tolerated by livestock, or higher in
vegetable crops than tolerated by humans,
a potential would exist for adverse
environmental impacts. Table 1 also
shows the level of an element present in
a crop when the crop is substantially or
visibly injured (25% yield reduction) by
the element (phytotoxicity). If soil and
plant processes keep the level of the
element in forages below that injurious
to animals, then a "Soil-Plant Barrier"
has protected the food chain (19,20).
For nearly all elements, the food
chain is protected under worst case con-
ditions. For a few more elements, one
must rely on dietary interactions which
increase tolerance in order to assert
that the food chain is protected. But
for Cd (in humans), Pb (in children and
livestock which consume soil), and Cd,
TABLE 1. MAXIMUM TOLERABLE LEVELS OF DIETARY MINERALS FOR DOMESTIC LIVESTOCK
IN COMPARISON WITH LEVELS IN FORAGES
Element "Soil-Plant Level in Plant Foliage^ Maximum Levels chronically tolerated^/
Barrier" NormalPhytotoxfc CattleSheepSwineCTi
—mg/kg dry foliage—
Sheep Swine
--mg/kg dry diet-
thicken
As
B
Cd ±'
Cr 3+
Co
Cu
F
Fe
Mn
Mo
Ni
Pb2/
Se
V
Zn
yes
yes
Fail
yes
Fail?
yes
yes?
yes
?
Fail
yes
yes
Fail
yes?
yes
0.01-1
7-75
0.1-1
0.1-1
0.01-0.3
3-20
1-5
30-300
15-150
0.1-3.0
0.1-5
2-5
0.1-2
0.1-1
15-150
3-10
75
5-700
20
25-100
25-40
_
-
400-2000
100
50-100
-
100
10
500-1500
50.
150.
0.5
(3000.)
10.
100.
40.
1000.
1000.
10.
50.
30.
(2.)
50.
500.
50.
(150.)
0.5
(3000.)
10.
25.
60.
500.
1000.
10.
(50.)
30.
(2.)
50.
300.
50.
(150.)
0.5
(3000.)
10.
250.
150.
3000.
400.
20.
(100.)
30.
2.
(10.)
1000.
50.
(150.)
0.5
3000.
10.
300.
200.
1000.
2000.
100.
(300.)
30.
2.
10.
1000.
_]_/ Based on literature summarized in Chaney et al. (20).
21 Based on (61). Continuous long-term feeding of minerals at the maximum tolerable
levels may cause adverse effects. Levels in parentheses were derived by inter-
specific extrapolation by NRC.
3/ Maximum levels tolerated based on human food residue consideration.
202
-------
Mo, Se, and Co (in livestock and
wildlife), the "Soil-Plant Barrier" does
not adequately protect the food chain.
However, as is discussed below, the food
chain is adequately protected from Cr in
wastes applied to land.
Chromium in Hazardous Wastes
Many wastes contain Cr. EPA identi-
fied electroplating sludges (F006), Cr
pigment wastes (K002-008), petroleum
sludges (K048-051), tannery wastes
(K053-058), and a few other wastes (K061,
062, 063, 069, and 074} as hazardous
because they contained Cr (31). Only
tannery wastes (K053-058) and processed
titanium ore wastes (K074) were subse-
quently delisted (34) when the EP test
was changed to reflect chromate rather
than total Cr (33). The Cr in petroleum
sludges and many others began as chromate
used in cooling towers as a corrosion
inhibitor, but was subsequently reduced
to an insoluble chromic form in the
anaerobic organic wastes. As will be
discussed below, chromate smelting wastes
can be high in chromate, as well as a few
wastes not listed by EPA.
Chromate in wastes is a risk to
humans since chromate has been found to
be carcinogenic to industrially exposed
workers. Thus, the fate of Cr in wastes,
soils, plants, and animals has to be
considered in planning land treatment
sites for Cr-rich wastes. Transforming
any chromate in a waste to chromic is a
necessary part of the waste handling
program if land treatment is planned.
Chromium in Soils, Plants, and Animals
Chromium is not essential for plants
(42), but chromium is essential for
animals (59, 61, 81). An organic
chelated Cr+3 compound (78) is a
cofactor in insulin hormone response
controlling carbohydrate metabolism.
Human diets are often deficient in this
Cr compound, and some older humans
presently experience Cr deficiency (61,
81).
Chromium exists in two redox forms in
nature: chromic (Cr3+) and chromate
(Cr6+). Bartlett and Kimble (12) show
an Eh-pH diagram for Cr in water.
Chromium in soil solution is further
decreased since chromic is strongly
adsorbed and chelated by soils at all
practical pH levels (12, 18, 38).
Chromate is rapidly reduced to chromic in
soils; reduction is more rapid in acidic
soils (13, 18, 38).
Bartlett and James (11) discovered
that soluble chromic could be oxidized to
chromate in soils and that air drying
soils prevented this reaction. Although
oxidation may be important in plant
uptake of Cr, it is not yet clear that
chromic oxidation is significant in the
natural environment since chromic is so
insoluble and strongly sorbed. Recent
research has shown that freshly precipi-
tated (less crystalline) Mn02 oxidizes
the chromic, producing chromate (3, 10,
11, 75).
Plant uptake of Cr to plant shoots is
generally very limited. Even when
soluble chromate is supplied, it is
reduced to chromic in plant roots and
kept there as chelates, precipitates, or
adsorbed Cr. Chromate can be quite
phytotoxic (16, 21, 36, 60, 75, 80).
However, even under conditions of
chromate phytotoxicity, plants contain
less than 10 ppm Cr (normal plant shoots
are lower than 2 ppm Cr).
Because humans are often deficient in
Cr, research has been conducted on
methods to increase Cr in plants.
Although chromate existence is more
favorable at alkaline pH, adjusting soil
pH had little effect. Repeated
sub-phytotoxic additions of chromate to
soils, or adding 1% Cr to soil as freshly
precipitated Cr(OH)3, did increase Cr
somewhat in plant leaves, but not in
grain or fruits (17, 18). Heating soils
to 300°C to simulate effects of forest
fires caused soil chromate to increase;
corn grown in heated soil had greater Cr
in roots than plants grown on unseated
soils, but not in shoots (40). Most
plants growing on naturally high Cr
serpentine soils have only normal levels
of Cr (4, 18, 65). However, some plants
occurring naturally on these soils are
tolerant of the low Ca status and high Ni
and Cr. Both endemic and ecotypic ser-
pentine tolerant plant species include
some Cr accumulators (50, 51, 53, 70, 85).
Lyon, Peterson, and Brooks (52) found
that Leptospermum scoparium (a Cr
accumulator) translocated Cr in xylem as
chromate. However, Cr in roots and
203
-------
shoots was chromic (49); the major iden-
tifiable soluble Cr anionic chelate was
trisoxalatochromiumflll) in all tissues
except xylem exudate. Peterson et al.
(69, 74) examined Cr uptake and translo-
cation by barley. In contrast to the
results with L_. scoparium, chromate
remained chromate in the roots; further,
some chromic was oxidized to chromate by
the roots, and the trisoxalato-
chromiumflll) was not found in roots or
shoots. Chromate may limit plant growth
on high Cr serpentine soils. Based on
the study of Anderson, Meyer, and Mayer
(4), chromate in the displaced soil
solution was thought to be too low to be
toxic to any plants. However, this work
used air-dried soils, and air-drying
temporarily inhibits a soil's ability to
oxidize chromic to chromate (11). Thus,
the whole subject must be re-evaluated
with high Cr soils not air-dried before
study.
The food chain is well protected from
excess Cr by the "Soil-Plant Barrier."
Animals tolerate high levels of insoluble
Cr components in their diet (61).
Chromic oxide has actually been fed as a
dietary marker for animals (61, 66, 81)
and even humans (43, 84). Recently,
McLellan et al. (58) showed that
51cr3+ mixed in a diet is an
effective label to indicate when all
(99.9% of ingested SlCr by 16 days) of
a meal has finally transited the diges-
tive system and been excreted. NRC (61)
indicated that 3000 ppm Cr as insoluble
Cr20s or 1000 ppm Cr as other chromic
salts were tolerated by domestic
animals. Because crop plants have such
low Cr levels even when grown on soils
very high in Cr, the food chain is
protected against excess plant absorbed
Cr. Animal ingestion of organic wastes
rich in Cr appears to offer a similarly
low hazard based on sewage sludge (45)
and tannery sludge (23, 46, 83) feeding
studies. Hydrolyzed leather meal (not
over 2.75% Cr) is an approved feed
ingredient (6).
Chromate has been found to be
mutagenic (63) and carcinogenic, par-
ticularly if inhaled; humans and experi-
mental animals exposed to chromate dusts
and mists are quite susceptible to nasal
cancer (61, 79). Chromate in feed has
not caused similar cancer in experimental
animals, and chromic in feed was not
carcinogenic or mutagenic. Chromate in
land-applied organic wastes is rapidly
converted to chromic, greatly reducing
the chance for a carcinogenic risk. In
contrast, when chromate-rich inorganic
wastes are placed on the land, chromate
may remain for a long period (16, 36).
Land application of refuse compost,
sewage sludge, tannery sludge, cooling
tower blowdown, and sewage effluent can
cause Cr in the surface soil to be
greatly enriched. However, these
practices have not yet been shown to
affect Cr in soil below the tilled zone,
and do not increase plant Cr after the
first crop year (22, 24, 25, 26, 28, 38,
44, 60, 73, 76). Gemmell (37) and Breeze
(16) found that sewage sludge and refuse
compost were very effective in reducing
chromate toxicity in smelter wastes.
In summary, Cr in wastes is very
unlikely to cause food chain problems
because animals tolerate high levels of
insoluble Cr compounds in diets, and
plants do not absorb and translocate high
levels of Cr to edible plant tissues even
when soils are greatly enriched in Cr.
Leather Manufacturing Wastes
Leather is stabilized collagen
prepared from the hide of livestock.
Hides are transformed into leather by two
major methods: 1) chrome tanning, and
2) vegetable (polyphenol) tanning.
Wastes from this industry were listed as
hazardous wastes in May, 1980 (31)
because of Cr, Pb, and sulfide present in
the wastes. Upon reevaluation, these
wastes were removed from the listed
hazardous wastes (34) based on the
presence of chromic rather than chromate
in the wastes. The EP test was changed
to consider chromate rather than total Cr
(33).
This section discusses sources and
nature of wastes from the chrome tanning
industry, results from feeding these
wastes to livestock or applying them to
land, and research needs for land treat-
ment of tannery wastes. Much of the
discussion of Cr is relevant to other
waste streams declared hazardous because
of Cr.
204
-------
Sources of Wastes
EPA and its contractors have
described tannery operations, waste
streams and waste disposal practices (14,
29, 68). Flesh and hair are removed in
mechanical and chemical operations which
use lime and sodium sulfide. Manganous
salts are added to catalyze oxidation of
the sulfide in dehairing wastewater (8).
This Mn could also contribute to oxida-
tion of sludge chromic to chromate (see
below). It seems unwise to mix Mn-
treated wastewater or sludges from
dehairing processes with Cr-rich sludges
from waste tanning solution or tanned
hide rinse water. Segregation of these
waste streams would reduce the potential
for chromate production from tannery
sludges. The organic matter plus lime
sludge from treatment of dehairing waste-
water is similar to limed raw sewage
sludge, but lower in heavy metals other
than Cr and Mn. This lime-organic waste
sludge could be applied to cropland as a
low grade limestone with low grade
fertilizer value. If farmland is near
the tannery, utilization of the liquid
sludge on farmland should be less expen-
sive than depositing it in a landfill.
Fleshing wastes are often sold as a
by-product for rendering or feed; these
could be used as organic fertilizer on
cropland if applied using technology to
avoid malodors.
After removal of flesh from the
hides, they are acidified and treated
with the chrome tanning solution. The
solution contains chromic sulfate at low
pH. The Cr3+ reacts with the collagen
and stabilizes it against deterioration.
The waste chrome tanning solution is
recycled directly in some tanneries; in
others, the 03+ is precipitated by pH
adjustment, collected, and reused (41,
54, 64). In others, it is discharged to
the wastewater treatment system or
publicly owned treatment works (POTW).
The tanned hide is rinsed, and then
cut or split into grain (outer) side
(used to make finished leather products),
and split side (inner, or flesh side)
which is used to make rougher leather
products.
The grain side is "shaved" to a
uniform thickness, generating a coarse
particulate leather shavings waste.
Shaved tanned grain sides are then
"buffed" or sanded to create a very
smooth surface. This generates a fine
particulate leather buffing dust. Pieces
of tanned leather are trimmed away to
achieve the final leather shapes
generating leather trimmings.
During other finishing steps dyes and
colors are applied, the leather is lubri-
cated, retanned with organic chemicals,
and otherwise processed to make finished
colored leather products (clothing, auto-
mobile and furniture coverings, etc.).
Some waste dyes, organic solvents, etc.,
enter the waste stream at this point.
Most tanneries have ceased use of Pb
pigments. This has sharply lowered Pb
levels in their wastes.
Nature of Leather Manufacturing Wastes
The Cr, Pb, and sulfide levels in the
many possible leather manufacturing
wastes have been described (29, 35, 68).
Chromium ranged from 2,200 to 21,000 ppm
wet weight, and Pb from 40 to 724 ppm.
During a site visit we obtained samples
of solid wastes from a modern tannery.
Analyses showed low Cd, Pb, In, Cu, and
Ni in all wastes, as was reported by
Thorstenen and Shah (77). Segregation of
Cr-rich wastewaters for separate treat-
ment has kept the Cr level in waste
streams other than Cr recycle sludge
relatively low. The wastewater treatment
sludge was high on Mn from use of Mn?+
to catalyze sulfide oxidation, and this
may merit study.
Uses of Leather Wastes
Technology has been developed to
utilize some tannery wastes. Leather
trimming, shaving, and buffing wastes
(collectively called leather tankage) are
essentially pieces of Cr tanned leather,
or fairly pure protein containing about
11% N and 1% Cr. Because of the
protein-N, these wastes are now used as a
feed ingredient or organic-N fertilizer.
Use as Feed --In response to a
petition, FDA developed rules for use of
hydrolyzed leather meal as an approved
feed ingredient for swine (6). It may
contain no more than 2.75% Cr, and may
not comprise more than ]% of the mixed
feed. Other feed uses will require new
petitions.
205
-------
Results of feeding trials with
leather tankage have been reported by
numerous researchers (23, 39, 46, 62, 67,
83). In general, they found that
steaming or acid hydrolysis raises its
protein equivalent value, and that the N
is used in a manner similar to urea by
sheep. No toxicity has been reported.
Reduced rates of gain may result if the
hydrolyzed leather meal is used to
replace too high a proportion of the
total protein in a feed.
Use as a Fertilizer --Because tannery
wastes contain organic N in the form of
the protein collagen, they have been
tested for use as fertilizers. Leather
tankage normally contains about 11% N. A
process was developed to react leather
tankage with urea and formaldehyde to
produce a higher N analysis fertilizer.
The product "Organiform", contains 22-24%
N (9). The manufacturer has reported N
release curves for this product in
comparison with other commerical ferti-
lizers. It contains no more than 2% Cr,
and low levels of other heavy metals.
Use as Soil Conditioner/Fertilizer —
Many studies have been conducted on
different leather manufacturing solid
wastes and wastewater treatment sludges.
Because research on land application of
sewage sludges has pointed out several
experimental difficulties in studying
wastes, it would be useful to reevaluate
the literature on land application of
leather wastes.
When high application rates of
unstabilized organic wastes are land-
applied, the excessive BOD can cause
anaerobic soil conditions and increase
soluble organic compounds in the soil.
Some of these are phytotoxic. Also,
soluble salts can be phytotoxic until
leached from the surface soil. Some
soluble Cr may be found during anaerobic
decomposition, but it is Cr3+ chelates.
Studies using soluble metal salt addi-
tions to soils can not be used to esti-
mate the effects on plants or ground
water of applying wastes containing
precipitated metals. This last point is
especially relevant to Cr, since Cr3+
compounds and complexes are kinetically
quite inert (1). When sludge Cr3+
precipitates equilibrate at pH 6 to 8, a
quite stable insoluble Cr sludge forms.
Study of soluble salts additions can lead
to serious errors in estimating risks of
Cr in leather manufacturing wastes.
Kick and Braun (44) reported studies
on tannery sludge which was actually a Cr
recycle sludge containing only 0.08% N
and 32% Cr (dry weight). Research on
this low volume inorganic waste is not
useful toward understanding the potential
effects on the environment of bulk
organic tannery wastes containing only
0.3-3.0% Cr.
A number of conclusions can be drawn
from the whole body of research on
Cr-rich leather wastes, including study
of tannery combined wastewater treatment
sludges (47, 55, 56, 57, 82), POTW
sludges produced largely from tannery
wastewaters (2, 22, 24, 25, 26, 27, 38),
and leather tankage (3, 5, 15, 48, 71,
72, 73).
1. The organic-N is apparently
released more slowly than from other
proteinaceous wastes because the Cr
stabilizes the protein. Volk (82)
estimates that about 10% of tannery
wastewater treatment sludge N becomes
crop available during the year of
application.
2. These wastes seldom caused any
phytotoxicity, especially when rates were
limited to the N requirement of crops.
Crops were somewhat enriched in Cr during
the first crop year.
Leaching Chromate from Leather Waste
Amended Soils —The critical remaining
issue is leaching of Cr to groundwater.
Chromate in groundwater may not exceed 50
ug/L (31, 33). If the immobile Cr3+
applied to soils in tannery wastes could
be oxidized to the mobile chromate by
soil Mn02 (11), it seems possible that
this could leach and contaminate ground
water (35). However, Amacher and Baker
(3) found that a soluble Cr-fulvic acid
extracted from ground leather was not
significantly oxidized by Mn02-
Chromate is rapidly reduced by soils to
form insoluble chromic (13, 18, 76). Low
concentrations of Cr are rapidly sorbed
by soils (7). Soil water at 1 m depth
was not increased in Cr by sewage sludge
rich in tannery waste borne-Cr (24, 25).
Water leaching from pots treated with
ground leather at N-fertilizer rates was
also not enriched in Cr (73).
206
-------
Together, these studies support a
model in which precipitated or chelated
Cr is not readily oxidized by Mn02, and
that ground water will not be polluted by
Cr from land-applied leather wastes if
reasonable management is practiced
regarding soil pH and rate of applica-
tion. This hypothesis has not been
proven conclusively by the present data
since the lower redox potentials
resulting from sludge application would
more rapidly reduce any chromate formed
back to chromic. Soils treated with
leather wastes long ago should be
evaluated since these are redox
stabilized and would allow testing of
worst case conditions. Field-moist,
never-dried soils must be studied to
avoid the errors identified by Bartlett
and James (11).
In summary, most leather tanning in
the U.S. uses chromium based processes.
When tannery waste streams are segre-
gated, only a few are high in Cr: the
precipitated Cr(OH)3 removed from used
tanning solutions, and the leather
trimmings, shavings, and buffing dust
generated during transformation of tanned
hides into finished leather. The precip-
itated Cr3+ waste has no practical use
except to be recycled, and should not be
considered for land application.
Leather wastes (tankage) have been
converted to steamed leather meal, and
are an approved swine feed ingredient.
These wastes have been experimentally fed
to swine, poultry, cattle, and sheep with
no injury to the livestock. The protein
is of low nutritional value (collagen).
Leather wastes (115UN) have also been
treated by a urea-formaldehyde process to
increase N content for use as a
fertilizer. This is a commercial
practice producing the 22-24% N product
"Organiform." Land application of ground
leather products supplies slow release
nitrogen.
Chromium is somewhat increased in
crops, at least during the first crop
year after application of Cr-rich organic
wastes. Chromium phytotoxicity has not
resulted from land application of
precipitated Cr3+. Some other leather
manufacturing wastewater treatment
sludges have caused crop problems, but
not demonstrably due to Cr. Many
problems occur in pot studies with wastes
because soluble salts are in excess, or
inadequate fertilizers are applied for
pot studies, etc.
A small amount of research has been
conducted on leaching of Cr from soils
treated with sewage sludge, tannery
sludge, or chromate. Chromate in soil
water at 1 m depth has not exceeded the
50 ug/L groundwater standard even when
large amounts of Cr3+ bearing wastes
are added. Often Cr in water at 1 m was
unchanged by the surface applied Cr.
Although soil MnO? can oxidize soluble
Cr3+ to chromate, this appears to be
environmentally irrelevant because Cr3+
has very low solubility at practical soil
pH levels.
Research Needs for Land Treatment of
Leather Manufacturing Wastes
Potential oxidation of chromic to
chromate is important in environments
rich in Cr3+. More basic research is
needed to assure that opportunities for
environmental pollution by Cr are not
ignored. The mechanisms of these
reactions must be better understood.
Chromium rich soils should be amended
with Mn02 to see whether environ-
mentally stable chromic can be converted
to chromate under some worst case
conditions.
Land treatment sites freshly amended
with tannery sludges will have lower
oxidation potential than long-term well
oxidized sites. These latter sites
should be studied to see whether Cr3+
application can ever pollute groundwater
with chromate. Economics of segregated
treatment of tannery waste streams should
be explored. Chromate leaching'should be
evaluated in relation to soil MnOp,
etc., since it is likely that most
chromate leaching past the waste-amended
surface soil will be reduced back to
chromic and precipitated.
References
1. Allaway, W. H. 1977a. Soil and
plant aspects of the cycling of
chromium, molybdenum, and selenium.
Proc. Intern. Conf. Heavy Metals in
the Environment 1:35-47.
207
-------
2. Alther, E. W. 1975.
Chromium-containing sludge and
chromium uptake by forage plants.
Leder 26:175-178.
3,'Amacher, M. C., and D. E. Baker.
1980. Kinetics of Cr oxidation by
hydrous manganese oxides. Agron.
Abstr. 1980:136.
4. Anderson, A. J., D. R. Meyer, and F.
K. Mayer. 1973. Heavy metal
toxicities: levels of nickel,
cobalt, and chromium in the soil and
plants associated with visual
symptoms and variation in growth of
an oat crop. Aust. 0. Agric. Res.
24:557-571.
5. Andrzejewski, J. 1971. Effect of
chromium fertilization on the yield
of several plant species and on the
chromium content of the soil (in
Polish). Roczn. Nauk Rolnickzych
A97:75-97,
6. Anonymous. 1967. Food additives
permitted in the feed and drinking
water of animals or for the treatment
of food-producing animals. Fed. Reg.
32(157):11734.
7. Article, J., and W. H. Fuller.
1979. Effect of crushed limestone
barriers on chromium attenuation in
soils. J. Environ. Qual. 8:503-510.
8. Bailey, D. A., and F. E. Humphreys.
1967. The removal of sulphide from
limeyard wastes by aeration. J. Soc.
Leather Trades Chemists 51:154-172.
9. Baker, H. J., and Bros., Inc. 1972.
Organiform LT. Product brochure. H.
J. Baker Co., 360 Lexington Ave., New
York, NY 10017
10. Bartlett, R. J., 1979.
Oxidation-reduction status of aerated
soils. Agron. Abstr. 1979:144.
11. Bartlett, R. J., and B. James.
1979. Behavior of chromium in soils:
III. Oxidation. J. Environ. Qual.
8:31-35.
12. Bartlett, R. J., and J. M. Kimble.
1976a. Behavior of chromium in
soils: I. Trivalent forms. J.
Environ. Qual. 5:379-383.
13. Bartlett, R. J., and J. M. Kimble.
1976b. Behavior of chromium in
soils: II. Hexavalent forms. J.
Environ. Qual. 5:383-386.
14. Bauer, 0. H., E. T. Conrad, and R. J.
Lofy. 1977. Solid waste management
practices in the leather tanning
industry. J. Am. Leather Chem.
Assoc. 72:270-280.
15. Berkowitz, J. B., S. E. Bysshe, B. E.
Goodwin, J. C. Harris, D. B. Land, G.
Leonardos, and S. A. Johnson. 1980.
Field verification of land
cultivation/refuse farming, p.
260-273. _In_ D. Schulz (ed.) Proc.
Sixth Ann. Res. Symp. on Disposal of
Hazardous Waste. EPA-600-9-80-010.
16. Breeze, V. G. 1973. Land
reclamation and river pollution
problems in the Croal Valley caused
by waste from chromate manufacture.
J. Appl. Ecol. 10:513-525.
17. Gary, E. E., W. H. Allaway, and 0. E.
Olson. 1977a. Control of chromium
concentration in food plants. 1.
Absorption and translocation of
chromium by plants. J. Agric. Food
Chem. 25:300-304.
18. Gary, E. E., W. H. Allaway, and 0. E.
Olson. 1977b. Control of chromium
concentration in food plants. 2.
Chemistry of chromium in soils and
its availability to plants. J. Agr.
Food Chem. 25:305-309.
19. Chaney, R. L. 1980. Health risks
associated with toxic metals in
municipal sludge, pp. 59-83. In G.
Bitton et al. (eds.) Sludge -- Health
Risks of Land Application. Ann Arbor
Science Publishers, Inc., Ann Arbor,
MI.
20. Chaney, R. L., D. D. Kaufman, S. B.
Hornick, J. F. Parr, L. J. Sikora, W.
D. Burge, P. B. Marsh, G. B. Willson,
and R. H. Fisher. 1981. Review of
information relevant to land
treatment of hazardous wastes. Draft
Report to EPA-SHWRD.
21. Chapman, H. D. 1966. Diagnostic
criteria for plants and soils. Univ.
Calif., Division Agr. Sci.,
Riverside. 793pp.
208
-------
22. Cunningham, J. D., D. R. Keeney, and
J. A. Ryan. 1975. Yield and metal
composition of corn and rye grown on
sewage sludge-amended soil. J.
Environ. Qual. 4:448-454.
23. Dilworth, B. C., and E. J. Day.
1970. Hydrolyzed leather meal in
chick diets. Poult. Sci.
49:1090-1093.
24. Dowdy, R. H., C. E. Clapp, W. E.
Larson, and D. R. Duncomb. 1979.
Runoff and soil water quality as
influenced by five years of sludge
applications on a terraced
watershed. Agron. Abstr. 1979:28.
25. Dowdy, R. H., C. E. Clapp, D. R.
Duncomb, and W. E. Larson. 1980.
Water quality of snowmelt runoff from
sloping land receiving annual sewage
sludge applications, pp. 11-15 IJT_
Proc. Nat. Conf. Municipal and
Industrial Sludge Utilization and
Disposal. Information Transfer,
Inc., Silver Spring, MD.
26. Dowdy, R. H., and G. E. Ham. 1977.
Soybean growth and elemental content
as influenced by soil amendments of
sewage sludge and heavy metals:
Seedling studies. Agron. J.
69:300-303.
27. Dowdy, R. H., and W. E. Larson.
1975. The availability of
sludge-borne metals to various
vegetable crops. J. Environ. Qual.
4:278-282.
28. El-Bassam, N., P. Poelstra, and M. J.
Frissel. 1975. Chromium and mercury
in a soil after 80 years of treatment
with urban sewage water. (In
German). Z. Pflanzenern. Bodenk.
1975:309-316.
29. EPA. 1974. Development document for
effluent limitations guidelines and
new source performance standards for
the leather tanning and finishing
point source category.
EPA-440/l-74-016-a. 158 pp.
30. EPA. 1979. Criteria for
classification of solid waste
disposal facilities and practices.
Federal Register 44(179):53438-53464.
31. EPA. 1980a. Hazardous waste
management system: Identification
and listing of hazardous waste.
Federal Register 95(98):33084-33133.
32. EPA. 1980b. Hazardous waste
management system: Standards for
owners and operators of hazardous
waste treatment, storage, and
disposal facilities. Federal
Register 45(98):33154-33258.
33. EPA. 1980c. Hazardous waste
management system: Identification
and listing of hazardous wastes.
Fed. Reg. 45(212):72029-72034.
34. EPA. 1980d. Hazardous waste
management system: Identification
and listing of hazardous wastes.
Fed. Reg. 45(212):72037-72039.
35. EPA. 1980e. Listing background
document: Leather tanning and
finishing industry. Supports
Appendix 7 in EPA (1980a). On file
in Docket 3001 at EPA.
36. Gemmel, R. P. 1973. Revegetation of
derelict land polluted by a chromate
smelter. 1: Chemical factors causing
substrate toxicity in chromate
smelter waste. Environ. Pollut.
5:181-197.
37. Gemmel1, R. P. 1974. Revegetation
of derelict land polluted by a
chromate smelter. 2: Techniques of
revegetation of chromate smelter
waste. Environ. Pollut. 6:31-37.
38. Grove, J. H., and B. G. Ellis.
1980. Extractable chromium as
related to soil pH and applied
chromium. Soil Sci. Soc. Am. J.
44:238-242.
39. Gruhn, K., and H. Luedke. 1972. Use
of hydrolyzed chrome leather waste in
pig feeding and the content of
chromium in flesh, liver, kidneys,
and hair (in German). Arch.
Tierernaehr. 22:113-124.
40. Hafez, A. A. R., H. M. Reisenauer,
and P. R. Stout. 1979. The
solubility and plant uptake of
chromium from heated soils. Commun.
Soil Sci. Plant Anal. 10:1261-1270.
209
-------
41. Hauck, R. A. 1972. Report on
methods of chromium recovery and
reuse from spent chrome tan liquor.
J. Am. Leather Chem. Assoc.
67:722-430.
42. Huffman, E. W. D., Jr. and W. H.
Allaway. 1973. Growth of plants in
solution culture containing low
levels of chromium. Plant Physiol.
52:72-75.
43. Irwin, M. I., and E. W. Crampton.
1951. The use of chromic oxide as an
index material in digestion trials
with human subjects. J. Nutr.
43:77-85.
44. Kick, H., and B. Braun. 1977. The
effect of chromium containing tannery
sludges on the growth and the uptake
of chromium by different crops (in
German). Landwirtsch. Forsch.
30:160-173.
45. Kienholz, E. W. 1980. Effect of
toxic chemicals present in sewage
sludge on animal health, pp.
153-171. Jji G. Bitton et al. (eds.)
Sludge—Health Risks of Land
Application. Ann Arbor Science
Publishers, Inc., Ann Arbor, MI.
46. Knowlton, P. H., W. H. Hoover, C. J.
Sniffen, C. S. Thompson, and P. C.
Belyea. 1976. Hydrolyzed leather
scrap as a protein source for
ruminants. J. Anim. Sci.
43:1095-1103.
47. Koc, J. 1979. Effect of the
temperature, moisture content in
soil, and additions of fertilizers on
the decomposition of tannery sludges
(in Polish). Rocz. Glebozn. 30:73-83.
48. Kravchenko, S. N. 1979. Effect of
leather waste on the yield and
quality of some vegetable crops (in
Russian). Nauchnye Trudy Ukr. S.-Kr.
Akad. 228:62-65.
49. Lyon, G. L., R. R. Brooks, and P. 0.
Peterson. 1969. Chromium-51
distribution in tissues and extracts
of Leptospermum scoparium J. R. et G.
Forst. (Myrtaceae). Planta
88:282-287.
50. Lyon, G. L., R. R. Brooks, P. J.
Peterson, and G. W. Butler. 1968.
Trace elements in a New Zealand
serpentine flora. Plant Soil
29:225-240.
51. Lyon, G. L., R. R. Brooks, P. J.
Peterson, and G. W. Butler. 1970.
Some trace elements in plants from
serpentine soils. N. Z. J. Sci.
13:133-139.
52. Lyon, G. L., P. J. Peterson, and R.
R. Brooks. 1969. Chromium-51
transport in the xylem sap of
Leptospermum scoparium J. R. et G.
Forst. (Myrtaceae).N. Z. J. Sci.
12:541-545.
53. Lyon, G. L., P. J. Peterson, R. R.
Brooks, and G. W. Butler. 1971.
Calcium, magnesium, and trace
elements in a New Zealand serpentine
flora. J. Ecol. 59:421-429.
54. Maire, M. S. 1977. A comparison of
tannery chrome recovery systems. J.
Am. Leather Chem. Assoc. 72:404-418.
55. Mazur, T., and J. Koc. 1976a. The
fertilizing value of tannery sludge.
II. Effect of fertilization with
tannery sludges on the crop (in
Polish). Rocz. Glebozn. 27:113-122.
56. Mazur, T., and J. Koc. 19765. The
fertilizing value of tannery sludge.
III. Effect of fertilization with
tannery sludges on the chemical
composition of plants (in Polish).
Rocz. Glebozn. 27:123-135.
57. Mazur, T., and J. Koc. 1980. The
fertilizing value of tannery
sludges, p. 328-338. lr± Handbook of
organic waste conversions. Van
Nostrand-Reinhold, New York.
58. McLellan, J. S., P. R. Flanagan, M.
J. Chamberlain, and L. S. Valberg.
1978. Measurement of dietary cadmium
absorption in humans. J. Toxicol.
Environ. Health 4:131-138.
59. Mertz, W. 1969. Chromium occurrence
and functions in biological systems.
Physiol. Rev. 49:165-239.
210
-------
60. Mortvedt, 0. J., and P. M. Giordano.
1975. Response of corn to zinc and
chromium 1n municipal wastes applied
to soil. J. Environ. Qual. 4:170-174.
61. NRC. 1980. Mineral Tolerance of
Domestic Animals. National Academy
of Sciences, Washington, O.C. 577pp.
62. Otevrelova, K., and M. Mackova.
1977. Tanning waste 1n feeds for fur
animals. 1. Hydrolytlc treatment of
different types of waste to increase
digestibility (In Czech.). Sbornik
Vedeckych Praci Vyzkumneho Ustavu
Vyzivy Zvirat, Pohorellce 11:41-48.
63. Petrllli, F. L., and S. deFlora.
1977. Toxlcity and mutagenldty of
hexavalent chromium on Salmonella
typhimurium. Appl. Environ.
Microbiol. 33:805-809.
64. Pierce, R., and T. C. Thornstenen.
1976. Recycling of chrome tanning
liquors. 0. Am. Leather Chem. Assoc.
71:161-164.
65. Proctor, 0., and S. R. J. Woodell.
1975. The ecology of serpentine
soils. Adv. Ecol. Res. 9:255-366.
66. Raleigh, R. J., R. J. Kartchner, and
L. R. Rittenhouse. 1980. Chromic
oxide in range nutrition studies.
Oregon State Univ. Agr. Exp. Sta.
Bull. 641:1-41.
67. Salimbaev, S. 1979. Use of tannery
shavings for fattening pigs (in
Russian). Zhivotnovodstvo 8:60-61.
68. SCS Engineers, Inc. 1976.
Assessment of industrial hazardous
waste practices - Leather tanning and
finishing industry. NTIS I PB
261-018.
69. Shewry, P. R., and P. 0. Peterson.
1974. The uptake and transport of
chromium by barley seedlings (Hordeum
vulgare L.). J. Exp. Bot. 25:785-797.
70. Shewry, P. R., and P. J. Peterson.
1976. Distribution of chronium and
nickel in plants and soil from
serpentine and other sites. J. Ecol.
64:195-212.
71. Shivas, S. A. J. 1980a. Factors
affecting the oxidation state of
chromium disposed in tannery wastes.
J. Am. Leather Chem. Assoc. 75:42-48.
72. Shivas, S. A. J. 1980b. The effects
of trlvalent chromium from tannery
wastes on plants. J. Am. Leather
Chem. Assoc. 75:288-299.
73. Silva, S., and B. Beghi. 1979.
Problems Inherent in the use of
organic fertilizers containing
chromium. Ann. Delia Fac. Agr. Univ.
Cath. Sacro Cuore 19(l):31-47.
74. Skeffington, R. A., P. R. Shewry, and
P. J. Peterson. 1976. Chromium
uptake and transport in barley
seedlings (Hordeum vulgare L.).
Planta 132:209-214.
75. Slater, J. P., and H. M. Reisenauer.
1979. Toxlcity of Cr(III) and Cr(VI)
added to soils. Agron. Abstr.
1979:38.
76. Taylor, F. 6., Jr. 1980. Chromated
cooling tower drift and the
terrestrial environment: A review.
Nuclear Safety 21:495-508.
77. Thorstenen, T. C., and M. Shah.
1979. Technical and economic aspects
of tannery sludge as a fertilizer.
J. Am. Leather Chem. Assoc. 74:14-23.
78. Toepfer, E. W., W. Mertz, M. M.
Polansky, E. E. Roginski, and W. R.
wolf. 1977. Preparation of
chromium-containing material of
glucose tolerance factor activity
from brewer's yeast extracts and by
synthesis. J. Agr. Food Chem.
25:162-166.
79. Towill, L. E., C. R. Shriner, J. S.
Drury, A. S. Mammons, J. W. Holleman,
and J. 0. Pierce. 1978, Reviews of
the environmental effects of
pollutants: III. Chromium.
EPA-600/1-78-023.
80. Turner, M. A. and R. H. Rust. 1971.
Effects of chromium on growth and
mineral nutrition of soybeans. Soil
Sci. Soc. Am. Proc. 35:755-758.
211
-------
Hydrolyzed leather meal in broiler
81. Underwood, E. J. 1977. Trace diets. Poult. Sci. 49:1259-1264.
Elements in Human and Animal
Nutrition. 4th Ed. Academic Press, 84. Whitby, L. G., and D. Lang. 1960.
New York. 545pp. Experience with the chromic oxide
method of fecal marking in metabolic
82. Volk, V. V. 1980. Department of balance investigations on humans. J.
Agronomy, Oregon State University, Clin. Invest. 39:854.
Corvallis, OR. Personal
communication. 85. Wild, H. 1974. Indigenous plants
and chromium in Rhodesia. Kirkia
83. Waldrup, P. W., C. M. Hillard, W. W. 9:233-241.
Abbott, and L. W. Luther. 1970.
212
-------
ASSESSMENT OF HYDROCARBON EMISSIONS FROM
LANDTREATMENT OF OILY SLUDGES
Robert G. Wetherold
Radian Corporation
8501 Mo-Pac Boulevard
Austin, Texas 78758
Donald D. Rosebrook
Radian Corporation
8501 Mo-Pac Boulevard
Austin, Texas 78758
E. W. Cunningham
Standard Oil Company (Ohio)
Midland Building
Cleveland, Ohio 44115
The first phase of a program to assess the atmospheric emissions from the landfarming
of refinery oily sludges has been recently completed. This study is jointly funded by the
American Petroleum Institute and the U.S. Environmental Protection Agency. The first phase
of this study was intended to characterize refinery oily sludges and to examine the effects
of a number of parameters on the mass and rate of fugitive hydrocarbon emissions from the
landfarming of refinery oily sludges. A small laboratory apparatus was developed to simu-
late some landfarming operations. All of the studies reported in this paper were conducted
in this enclosed landfarm simulator in which the major important variables could be con-
trolled.
The variables which were studied included sludge type, sludge volatility, soil mois-
ture content, wind speed, relative humidity, air temperature, soil temperature, sludge
loading on soil, and sludge application technique.
Important parameters affecting emission rates and total hydrocarbon mass emissions
were found to include sludge volatility, sludge loading on soil, application method
(surface or subsurface), and soil moisture content (and/or porosity).
Introduction
Oily sludges, which are generated in
various petroleum refinery operations, must
be disposed of in an environmentally accept-
able manner. Provisions of the Resource
Conservation and Recovery Act of 1976
restrict the available options for disposal
of this waste. One accepted method of
disposal for refinery oily sludge that
appears to be economically attractive is
landfarming. This technique has been prac-
ticed in some refineries for several years,
but it has attracted much more interest in
recent years. Along with this growing
interest, several studies have been
initiated to investigate landfarming
techniques.
In 1980, the American Petroleum
Institute (API) and the Environmental
Protection Agency (EPA) jointly funded a
study to assess the atmospheric emissions
from the landfarming of refinery oil
213
-------
sludges. The program was carried out by
Radian Corporation.
The objectives of the study were:
• to determine the magnitude of hydro-
carbon emissions from landfarming
operations with refinery oil
sludges,
• to identify those sludge character-
istics and landfarming parameters
which have the greatest influence
on the magnitude of hydrocarbon
emissions,
• to develop methods of characterizing
sludges, and
• to develop a method to predict the
emissions from landfarming opera-
tions from sludge characteristics
and operating parameters.
The first phase of this program has recently
been completed. Sludge characterization
tests and laboratory simulations of land-
farming operations were carried out in this
phase. This paper presents some of the more
important results obtained in the study.
Experimental Procedures
Sludges were characterized by both
physical and chemical testing, and a strip-
ping test was developed to provide a measure
of the sludge volatility. A landfarming
simulation system was developed in the
laboratory to evaluate the emissions under
varying landfarming conditions.
The oily sludges were subjected to the
following physical tests:
• total solids,
• total volatile solids,
• total suspended solids,
• total volatile suspended solids,
• weight loss during heating, and
• centrifugation.
The chemical characterization tests
performed on the sludge included:
• oil and grease content of the
sludges, and
• gas chromatography-mass spectro-
metry for volatile organic species
in the sludges and soils.
The oil and grease content of the
sludges was estimated by two separate
methods. In the gravimetric method, oil
and grease is extracted from the sludge
sample with Freon. The solvent is then
evaporated. The remaining residue is con-
sidered to be oil and grease.
An infrared procedure was also used to
estimate the oil and grease content of the
sludges. The sludge is extracted with
Freon, but the Freon is not evaporated.
Instead, infrared absorption measurements
are made which correspond to the concentra-
tions of the hydrocarbons present in the
solvent. The absorbance values are com-
pared to those obtained with a reference
oil composed of isooctane, hexadecane, and
benzene.
Sludge samples were subjected to a
stripping test in an effort to obtain a
quantitative measure of their relative
volatility. In this test a Bellar tube
and a Byron Instruments Model 301C Total
Hydrocarbon Analyzer (THC) were used. A
schematic diagram of the apparatus is shown
in Figure 1. The Bellar tube was filled
with 25 ml of water and a small volume of
sludge was added. Nitrogen was purged at
a constant measured rate through the water-
sludge mixture. The nitrogen stream from
the Bellar tube was analyzed for hydro-
carbons with the Byron 301C THC.
The THC can perform an analysis every
four minutes. The stripping was continued
for at least 80 minutes.
The landfarming simulation device
developed in this study is shown in- Figure
2.
The soil chamber provides one square
foot of soil surface area. Two different
plexiglass soil chambers were used which
provided for either 4 inch or 14 inch soil
depths. Each chamber had 4 inches of head
space above the soil. Incoming air was
heated as desired by a 1,650 watt heater
placed in front of the air intake to the
chamber. Air temperature and relative
humidity were monitored in the exit duct.
214
-------
Calibration
Gas In
Air Air Air H2
to to to to
Valves Carrier FID. FID.
Strip Chart
Recorder
Signal
Byron 301CTHC
Sample
Out
Fritted
Disc
For
Dispersion
Figure 1. Stripping Test Apparatus
Brush Mode! FS01W6D
Strip Chart Recorder
with Disc. Integrator
TLV Hydrocarbon
Analyzer
Air
Inlet
Space
Heater X
Position
4"
Buried Heater
W^A^'^*y*v*
Soil
Kurtz
Air Velocity Meter
Air Temperature &
Relative Humidity Monitor
12"
12"
Air Exit
To Blower
10" Diameter
12"-
Figure 2. Landfarm Simulation Apparatus
215
-------
Soil temperature was varied by means
of a heater buried one inch below the soil
surface and controlled with a feedback temp-
erature controller. Soil moisture and
temperature were varied manually and
monitored with a Soiltest MC300B analyzer
The air velocity could be controlled
with a. Dayton 2C939A, 2000 CFM exhaust
blower in the exit duct. A series of holes
in this duct could be covered or uncovered
to achieve the desired velocity. Air
velocity was measured with a Kurtz (hot
wire) air velocity meter.
The hydrocarbon concentration in the
air leaving the soil chamber was monitored
with a Bacharach TLV Sniffer. The Sniffer
probe was inserted into the exit ducting at
a point approximately six inches downstream
from the soil chamber.
Samples of the sludge to be studied
were preweighed into one-gallon jugs.
Attached to the jug was a piece of Teflon
tubing leading to a 12 inch x 1/2 inch
piece of plexiglass tube which served as
a spreader bar. The plexiglass tube had a
number of 1/8 inch holes drilled in it to
allow spreading of the sludge. Subsurface
injection was simulated by repeated inser-
tion of the wand from a hand pressurized
garden sprayer.
Tilling was accomplished manually with
a small, short-handled four-tined garden
hoe.
A typical simulation run using the
above apparatus was conducted as follows:
1. The heaters (air and soil) were
turned on and set for the desired
temperatures.
2. The exhaust blower was started and
the flow rate through the duct was
adjusted to attain the desired
velocity.
3. The Bacharach TLV Sniffer was
turned on and the instrument and
recorder were zeroed on ambient
air in what was a hydrocarbon-free
environment. Instrument response
was checked with methane (533 ppm).
4. After recording a background hydro-
carbon level, the preweighed amount
of sludge was spread over the soil
surface (or injected, as the case
may be). This operation required
somewhat less than 60 seconds.
5. Hydrocarbon concentration levels in
the air (leaving the soil chamber)
and disc-integrated peak areas were
monitored continuously from the
start of spreading.
Variables that were considered in this
phase of the study included:
• soil type,
• air/soil temperature,
• moisture content of soil,
• relative humidity,
• sludge volatility,
• spreading technique,
• wind speed,
• soil loading, and
• sludge temperature during spreading.
In landfarming simulation experiments,
each parameter was varied independently
from a set of base conditions to a higher
or lower value. In some cases, additional
values were tested to more clearly define
the effects of a particular parameter.
A total of 24 experiments were run in
which sludges were applied to soils in the
laboratory simulation of landfarming. The
matrix of experiments and experimental
conditions is presented in Table 1. Two
soils, Soil S and Soil T, were used in the
simulations. These soils were taken from
operating refinery landfarms.
Three oily sludges were obtained from
refineries engaged in landfarming. These
sludges are identified as:
SI - storage tank bottoms
S3 - API separator sludge
T - API separator sludge
It was necessary to calibrate the TLV
Sniffer disc integrator to get quantitative
results from the landfarming simulation
experiments. This was done by fitting the
soil box with a false bottom at the normal
soil level. The modified soil box was
placed on a Mettler P4400 balance. Sludge
SI samples were placed on the false bottom
of the soil box. The conditions were
216
-------
TABLE 1. MATRIX OF EXPERIMENTAL CONDITIONS FOR LAHDFARM SIMULATION RUNS
«un
Ho.
1
2
3
4
5
6
7
>
9
10
U
12
13
14
15
16
11
19
20
21
22
23
24
23
26
27
28
29
Mute
Material
Hexanc
Hexane
Gasoline
GasoHne
SI
SI
SI
SI
SI
SI
SI
SI
SI
SI
T
SI
SI
SI
S3
S3
S3
S3
S3
S3
S3
S3
SI
SI
Waste
Loading
Ub/ft1)
1.0
1.0
1.0
1.0
2.5
2.5
2.5
2.5
2.5
1.0
1.0
2.5
2.5
2.5
2.5
2.5
0.5
2.5
1.0
2.5
4.0
1.0
4.0
2.5+
2.5+
2.5+
2.5
2.5
Soil
Type
None
S
Done
S
Hone
S
g
S
S
S
S
T
S
S
S
g
S
S
S
S
S
T
T
S
S
S
S
S
ion
Moisture
(»tJ)
MA
10.7
NA
10.7
NA
10.7
10.7
10.7
10.7
10.7
10.7
4.7
20.7
10.7
10.7
10 7
10. 7
10.7
10.7
10.7
10.7
10.7
4.7
4.7
10.7
10.7
10.7
10.7
10.7
Soil/Air Air
Temperature Velocity
( P/°D («ph)
HA/100 3
120/100 3
NA/100 3
120/100 3
NA/100 3
120/100 3
120/100 3
120/100 1
100/80 3
120/100 3
120/100 3
120/100 3
120/100 3
120/100 3
120/100 3
120/100 3
120/100 3
120/100 3
120/100 3
120/100 3
120/100 3
120/100 3
120/100 3
120/100 3
120/100 3
120/100 3
120/100 3
120/100 3
lelatlve
Humidity
«)
tab*
Amb«
/tab*
Amb*
Amb*
Amb*
Amb*
40
Amb*
A»b*
Amb*
Amb*
75
Amb*
Amb*
Amb*
Amb*
Amb*
Amb*
Amb*
A«b*
ABO*
Amb*
Amb*
Anb*
Amb*
Amb*
Spreading
Mechanism
Pour
Surface
Pour
Surface
Pour
Surface
Surface
Surface
Surface
Surface
Surface
Surface
Surface
Surface
Surface
Surface
Surface
Surface
Surface
Surface
Surface
Subsurface
Subsurface
Remarks
Pure hexane - no soil
Pure hexane with soil
Pure gasoline-no soil
Pure gasoline with soil
Sludge 1 with no soil
Base case
Reduced air velocity
Reduced solVair temperature
Reduced soil loading
Repeat of Run 10
Used sludge T
soil box
Lowest soil loading
Heat sludge to 140 F
before application
Sludge S3 at low loading
Sludge S3 at base case loading
Sludge S3 at increased loading
Sludge S3 with aoll T-low loading
Sludge S3 with soil T-high loadiog
Spike sludge with 0.5 Ib of gas
Spike sludge with 1.5 Ib of gas
Spike sludge with 1.5 Ib of 8,0
Till soil after spreading
use deeper Injection
blent humidity was typically 501 at room t«
•rature but measured «30I at 100 f.
adjusted to the base case conditions (Runs
6 and 7). Then the weight loss due to
vaporization was measured as a function of
time. The integrated area under the TLV
Sniffer recorder curve was also monitored
as a function of time. From these values,
a calibration factor which relates weight
loss to integrated area can be estimated.
Results
The three sludge samples were quite
different from one another in their physical
characteristics. Sludge SI contained
approximately 92 percent oil (by weight).
Sludge S3 contained approximately 20 percent
oil, while only 10 percent of Sludge T con-
sisted of oil.
There was also a marked difference in
their apparent volatility. Results of
stripping tests on samples of the three
sludges are shown in Figure 3. Sludge SI
is the most volatile while Sludge T is the
least.
In the course of the landfarming simu-
lation experiments, a concentration-time
plot, such as that shown in Figure 4, was
generated for the hydrocarbon concentration
in the air flowing across the sludge-laden
soil. In almost every case there was a
very rapid rise in hydrocarbon concentra-
tion, usually reaching a maximum within a
minute or two of the start of the spread-
5 21
STRIPPING TIME (MINUTES)
Figure 3. Concentration of Hydrocarbons in
Purge Gas During Stripping Tests
ing operation. An equally precipitous
decline from the maximum and a slow
approach to zero (laboratory ambient air)
usually followed. Generally, the concen-
tration declined to about 1-2 ppm in about
30-60 minutes from the time of spreading.
217
-------
Figure 4. Typical Chart Trace from THC Monitor
The results of the 29 landfarming
simulation experiments are presented in
Table 2. The maximum hydrocarbon concen-
tration, and the integrated area under the
concentration-time trace at six minutes and
30 minutes from spreading are given for
each run. The area is expressed in
arbitrary (but consistent) area units.
The results of the calibration of the
landfarming simulation apparatus are shown
in Figure 5. The calibration factor of
9.85 area units/gram was used to estimate
the weight loss of hydrocarbon during the
landfarming simulations. The weight losses
were calculated from the 30-minute inte-
grated time-hydrocarbon concentration curve,
with the following assumptions:
• The response of the TLV Sniffer to
sludge SI vapors during the simula-
tion tests was linear and equal to
that determined in the calibration.
TABLE 2. RESULTS OF SIMULATION RUUF
Yari«blc'
Maxima
Hydrocarbon
(pp.)
Integrated
Area In
6 Kin.
Integrated
Area In
30 «ln.
Pure hexane (falae bottoaO
Pure hexane
Gaaollne (falae botton)
Gaaollne
SI (falae botto.)
Baae caae
Baae caae (repeat)
Air velocity decreaaed to 1 arth
Alr/aoll temperature decreaaed
10 toadlng decreaaed
11 Loading decreaaed (repeat of 10)
12 Soil I uaed
13 Holature in toll increased
14 Holature in air increaaed
IS Haate T uaed
16 Baae Caae in a deeper box
17 Loading increaaed
18 Loading decreaaed draatlcally
19 Uaata heated before apreadlng
20 waete S3 at low loading
21 Uaate S3 uaed
22 Waate S3 at high loading
23 Uaate S3 low load on T aoll
24 Waate S3 high load on T aoll
15 Uaate S3 Kith lew gee apike
26 Uaate S3 with high gaa aplke
27 Uaate S3 cut with water
21 Subaurface injection
2SA Tilling of 28
29 Subaurface deep injection
29A
29B
10,000
10,000
5,000
10,000
400
215
200
310
260
325
200
485
500
500
35
350
425
250
350
23
15
25
17
34
7,000
> 10,000
25
300
375
0
225
50
2,460
3,900
690
1,750
43
21
22
57
23
49
31
86
85
68
1
27
33
31
45
4
1.5
4
2
4
480
760
5
17
29
0
15
3
7,980
5,500
1,810
2,470
342
78
62
231
53
143
136
230
263
116
1
71
81
61
74
10
5
16
5
12
760
1,360
114
68
0
15
3
8580 area at dryneaa
2400 area at tero hydroearboM
Tilled 80 mil. after epreadiae.
Tilled 4 daya after apreadiBg
Tilled 6 day. after apre.diag
'variable clUMM treat tke »a*e cm coWltloae of lieu t m4 7. All etkar Tariaklee reMlaiil aa H
t and 7.
218
-------
2000-1
1500-
Ul
cc
D
U)
cc 1000
UJ
z
o
500-
Run 8 (0.25 Ib.)
Run 7 (0.50 Ib.)
Slope = 9.85 area units
gram
Figure 5.
Run 10 (0.25 Ib.)
Run 9 (0.25 Ib.)
—r~
50
—I—
100
—1—
150
—I
200
SLUDGE S1 WEIGHT LOSS (GRAMS)
Calibration Factor for Sludge
SI Vaporization in Landfarming
simulator
• The response of the TLV Sniffer
to sludges S3 and T vapors is the
same as that for sludge SI vapors.
The results are shown in Table 3, and they
indicate that from 0.5 to 3.2 percent of
the total applied sludge SI is lost in the
first 30 minutes after application.
Slightly larger percent losses relative
to the total applied oil are incurred.
Only 0.07 to 0.22 percent of the total
weight of sludge S3 is emitted in 30
minutes, but this represents from 0.22 to
1.1 percent of the oil applied. Sludge T
loses only 0.09 percent of the applied oil
and an order-of-magnitude less of the
total applied weight in 30 minutes. Thus,
the most volatile sludge, sludge SI, as
determined from the stripping test, emits
the most hydrocarbons. Conversely, the
least volatile sludge, sludge T, emits the
least material.
The effect of volatility on emissions
is further illustrated in Figure 6. The
emissions are proportional to the relative
response of the TLV Sniffer. The relative
responses at the maximum concentration and
for the 6-minute and 30 minuted cumulative
response (area) are shown. The behavior of
the three sludges (SI, S3, and T) during
the landfarming simulation indicates the
TABLE 3. ESTIMATED EMISSIONS FROM SLUDGE APPLIED TO
SOIL IN LANDFARMING SIMULATION TESTS
Run/Sludge
6/S1
7/S1
9/S1
10/S1
11/S1
12/S1
13/S1
14/S1
15 /T
16 /SI
17/S1
18/S1
19/S1
20/S3
21/S3
22/S3
23/S3
24/S3
27/S3
Area
30
78
62
53
143
136
230
263
116
1
71
81
61
74
10
5
16
5
12
8
Grains
Volatilized*
7.9
6.3
5.4
14.5
13.8
23.3
26.7
11.8
0.10
7.2
8.2
6.2
7.5
1.0
0.51
1.6
.51
1.2
.81
Weight
Sludge
Applied, g
1135
1135
1135
454
454
1135
1135
1135
1135
1135
1816
227
1135
454
1135
1816
454
1816
1135
X Total
Sludge
Volatilized
0.70
0.56
0.48
3.2
3.0
2.1
2.4
1.0
0.009
0.63
0.45
2.7
0.66
0.22
0.045
0.088
.11
0.066
0.071
Weight
Oil
Applied
1044
1044
1044
418
1044
1044
1044
1044
114
1044
1671
209
1044
91
227
363
91
363
227
X
Volatilized
0.75
0.60
0.52
3.5
1.3
2.2
2.6
1.1
0.088
0.69
0.49
3.0
0.72
1.1
0.22
0.44
0.56
0.33
0.36
*Based on response factor given in Figure 5
219
-------
higher emissions of the more volatile
sludge (SI).
As a further test of the volatility
effect, 0.5 and 1.5 pounds (Ibs) of gasoline
were added to 2.5 Ibs of sludge S3 and the
mixtures were applied to the soil in the
simulator. The results of these tests are
also shown in Figure 6. The emission rates
of the mixtures were much higher than that
of the unadulterated sludge. In addition,
the 6-minute and 30 minute cumulative
responses (areas) indicated an increase in
emissions with increased amounts of added
gasoline.
S K«H
I
I i«H
MAXIMUM
HYDROCARBON
CONCENTRATION
(PPM)
8 MINUTE
INTEGRATED
AREA
30 MINUTE
INTEGRATED
AREA
Figure 6. Effect of Volatility
In another test, 1.5 Ibs of water were
added to 2.5 Ibs of sludge, and the mixture
applied to the soil surface in the simula-
tor. As shown in Figure 6, there was
virtually no effect of the additional
water on the emission rate.
The amount of sludge applied to the
soil surface (loading) was varied to dis-
cern any effect on the emissions rate.
Variable loadings were examined only for
sludges SI and S3. The emissions from
sludge T were so low at a loading of 2.5
lb/ft2 that any effect of loading would
very likely be undetectable.
The effects of loading on the cumula-
tive response after 30 minutes are
presented in Figure 7. There appears to be
a worst-case loading in terms of pounds of
sludge per square foot of soil particularly
for sludge SI (all other variables being
constant). Within the loading ranges
studied, this same phenomena was not as
evident for sludge S3. This may : be due to
the effect of the water in the sludges.
Sludge SI contains virtually no water;
sludge S3 is primarily water.
10
LOADING (LBS/FT2)
Figure 7.
Hydrocarbon Emissions as a
Function of Soil Loading
Two subsurface injection experiments
were carried out, and they produced quite
different results. In the first case (Run
28) injection was performed at a shallow
depth (~ 3-4 inches) and some sludge
actually bubbled to the surface. The
emissions were relatively high during the
initial application and during the tilling
which was done 80 minutes after applica-
tion (see Table 2).
A second subsurface injection experi-
ment was performed in which the sludge was
injected at a deeper level (~6 inches) in
the soil. No sludge appeared on the sur-
face of the soil. There were no detectable
emissions during or after the sludge injec-
tion. The soil was tilled after four days
and again after six days. Some hydrocar-
bons were detected in the air at the time
220
-------
400-1
Tlllad 4 daya attar Injection
Tlllad 6 daya attar ln|»ctlon
Initial Subaurfaca Infection (Hydrocarbon Laval - 0)
I
9
TIME (MINUTES)
I
12
IS
I
17
Figure 8.
Effect of Resting Time Before Tilling on
Emissions After Subsurface Injection.
of tilling. Figure 8 shows the chart
tracings for these three measurements.
There can be little doubt from these experi-
ments that a substantial overall emission
reduction can be achieved by subsurface
injection and a reasonable waiting period
before tilling.
Figure 9 shows the effects of various
other parameters on the relative magnitude
of emissions from the landfarming simula-
tion. All the runs shown in the figure
were made with sludge SI and a consistent
loading of 2.5 lb/ft2. In all except one
run, the moisture content of the soil was
10.7 percent by weight. In the experiment
with 20.7 percent soil moisture, there was
a large increase in the amount of fugitive
hydrocarbon loss. This effect may be due
to the wetter soil being much less perme-
able to the liquids in the sludge. Pools
of liquid were observed on the soil sur-
face. The residence time of these liquid
pools was much longer than that observed
with the lower moisture soils.
In one run, the sludge was heated to
140°F before applying it to the soil sur-
face. There was no impact of this higher
temperature on the 30-minute integrated
area, probably because of relatively rapid
cooling. The maximum concentration and the
6-minute integrated area were somewhat
higher than the base case.
The effect of soil type on the emis-
sion rate was not adequately defined in
this study. When sludge S3 and soil T were
used in test runs, the emissions were lower
than found with soil S. This is consis-
tent with the lower moisture content of
soil T. However, emissions were higher
with sludge SI and soil T than with sludge
SI and soil S. At this point, the anomaly
cannot be explained.
221
-------
29 30
TIME (MINUTES)
Figure 9.
Effect of Moisture, Loading
and Air Flow on Emissions
Figure 9 indicates that emissions
increase dramatically when the air velocity
is reduced from 3 mph to 1 mph. However,
it should be remembered that the cumulative
integrator counts are equivalent to inte-
grated concentrations in the exiting air.
Thus, a threefold increase in integrated
concentration was detected for a three-
fold decrease in air flow rate. This is
equivalent to the same mass emission for
both wind speeds.
All but one experiment were run at
ambient humidity (~ 50%). One run was made
at approximately 75 percent relative
humidity. As shown in Figure 9, there
appears to be a pronounced effect on the
emission rate. The cumulative emissions
at 30 minutes from spreading are about
twice as great at the elevated humidity
level.
One test run was made with the air
temperature reduced from 100°F to 80°F and
the soil temperature reduced from 120°F to
100°F. The hydrocarbon emissions for the
first 30 minutes following spreading were
reduced by about 25 percent for the test
run at lower temperatures.
Much of the variability in the data
and results could not be attributed solely
to the independent effects of the variables
which were studied. This suggests the
possibility of important interactions
between these variables. Interactions were
not adequately studied in this program
because of constraints on the number of
tests which could be made,
Conclusions
The results obtained in this study
support the following conclusions:
• The laboratory landfarming simula-
tion device developed during this
study can provide very repro-
ducible results for sludge
emission studies.
• The volatility of the sludge is
very important, perhaps the most
important, parameter in estimating
the emissions from landfarming of
sludge.
• The sludge stripping test devel-
oped in this study gives a
quantitative measure of the
volatility of a sludge. This
volatility measure appears to
correlate with the landfarming
emission rates.
• The emission rates are affected by
the sludge loading on the soil.
There may exist some low soil
loading ranges which are less
desirable with respect to atmos-
pheric emissions.
• Within the range studied, air
velocity does not affect the
quantity of hydrocarbons emitted.
• Subsurface injection of sludges
appears to be the preferred
spreading technique for minimizing
atmospheric emissions from land-
farming operations.
• There may be important inter-
actions between some of the
variables studied in this project.
These could not be investigated
within the scope of this study.
• From 0.01 to 3.2 weight percent of
the sludges applied to the soil in
the landfarming simulations was
vaporized. This is equivalent to
222
-------
0.1 to 3.5 percent of the oil in the Acknowledgments
applied sample sludges.
The authors would like to thank the
The highest emission rates occur American Petroleum Institute and the U. S.
within the first 30 minutes after Environmental Protection Agency for their
application of the sludge. support and guidance in this work. We
would also like to acknowledge the contri-
butions of the Radian technical staff,
notably Mr. L. P. Provost, Mr. J. Randall
and Dr. R. Minear who has reviewed reports
and manuscripts.
223
-------
THE DEVELOPMENT OF LABORATORY AND FIELD STUDIES TO DETERMINE
THE FATE OF MUTAGENIC COMPOUNDS FROM LAND APPLIED HAZARDOUS WASTE
K. C. Donnelly and K. W. Brown
Texas Agricultural Experiment Station
Soil and Crop Sciences Department
Texas A&M University
College Station, Texas 77843
ABSTRACT
The future health of our citizens may well depend upon the proper disposal of the
more than 40 million metric tons of hazardous waste which are produced annually. The
hazards associated with improper disposal include both toxic and mutagenic responses.
While toxic responses are manifested soon after exposure, mutagenic responses are much
more subtle and may require decades before they are fully manifested. There are a
large number of plant, microbial, and mammalian cell culture assays which have been
developed to indicate the presence of compounds which cause genetic damage. In the
present study, two biological systems were evaluated as part of a battery of bioassays
to be used in determining the fate of mutagenic constituents from land applied hazardous
waste. The Salmonella/microsome assay and Bacillus subtilis DNA repair assay were used
to evaluate the genetic toxicity of a refinery and a petrochemical sludge. The two
wastes were separated into acid, base, and neutral fractions by liquid-liquid extraction.
All three separates exhibited mutagenic activity. The degradability of the petrochemical
waste was evaluated in a laboratory study. The waste was mixed with a Norwood sandy
clay and incubated in a soil respirometer for 180 days at 30°C. Following the incubation
period, the residual hydrocarbons were extracted from the soil for analysis by the
biological systems. The mutagenic effects of the saturate and aromatic fractions appear
to be reduced by soil incubation, while no such reduction was observed in the condensed
ring fraction. Additional information was collected on the mutagenicity of leachate and
runoff water collected from field lysimeters amended with the two wastes. Only a few
samples of the leachate contained low levels of activity, while the mutagenic activity
of the runoff decreased with time following application. These results indicate that
biological analysis can be useful to evaluate the fate and mobility of the mutagenic
constituents of wastes disposed of by land treatment.
INTRODUCTION
A by-product of the rapid expansion
of our industrial activities in recent
years has been a subsequent increase in
the rate of generation of hazardous waste.
The EPA (1979) estimated that more than
40 million metric tons of hazardous waste
would be generated in 1980. The most
common methods used to dispose of this
waste include landfills, deep well injec-
tion, ocean dumping, incineration, and
land treatment. Of these methods, only
land treatment and incineration provide
for the partial or complete destruction
of toxic organic waste constituents.
The use of non-destructive methods
of waste disposal on the land results in
dedicated sites which are removed from
productivity forever. Land treatment
may, however, result in sites which can be
reclaimed, provided it can be demonstrated
that hazardous constituents will be
retained or degraded within the immediate
environment of a land treatment facility.
It is essential that hazardous wastes be
managed so that the public is protected
224
-------
from the effects of waste constituents
which can cause genetic damage. Genotoxic
compounds in a hazardous waste must be
monitored in order to control accidental
exposure to mutagenic, carcinogenic, or
teratogenic agents, and to prevent trans-
mission of related defects to future
generations. Genetic toxicity should be
determined using a series of biological
systems which will provide an accurate
prediction of the potential of waste
constituents to cause gene mutations and
other types of genetic damage. Biological
tests provide the most accurate prediction
of the genetic toxicity of a sample, while
chemical analysis of the components of
a complex mixture may fail to account for
transformations in the soil or the
various interactions which can occur
between chemicals in a complex mixture.
There are four groups of biological test
methods which may be used to evaluate
the genetic toxicity of a sample. These
methods include epidemiologic, long-term
animal, long-term plant and animal cell
culture, and simple microbial, plant or
animal cell culture assays. The only
one of these methods which is not restric-
ted by time and cost is the short-term
assays using plant, microbial or
mammalian cell cultures. A list of some
of the possible short-term systems and
the genetic events which they can detect
is given in Table 1.
A complete battery of test systems
should be capable of detecting gene
mutations, compounds which cause various
types of chromosome damage, and those
which inhibit DNA repair. This project
used several test systems to respond
to the types of genetic damage previously
described, to detect the anticipated
compounds of the waste, and which can
incorporate metabolic activation into
their testing protocol. All of these
systems include provisions for sol-
vent controls and positive controls
which demonstrate the sensitivity of
the test system, the functioning of the
metabolic activation system, and act as
an internal control for the biological
system. Thus far, we have examined the
use of two biological test systems to
detect compounds which can cause point
mutations, and those which can produce
increased lethal damage in DNA repair
deficient organisms. The microbial
mutagenicity assay using Salmonella
typhimurium as an indicator organism has
been shown to be 80-95% efficient for
detecting certain classes of organic
chemical carcinogens (McCann et al., 1975;
Purchase et^ _al., 1976). Categories of
carcinogens which are poorly detected by
Salmonella assay include chlorinated
hydrocarbons, chemicals which are bac-
tericidal or volatile, and those which
crosslink DNA (Rinkus and Legator, 1979).
The Bacillus DNA repair assay has been
shown to be sensitive to pesticides
(Shiau jH^ jil., 1980) and bactericides
(Brown and Donnelly, 1981) which fail to
induce a clear response in Salmonella.
The two biological test systems were
selected to efficiently detect mutagens
and potential carcinogens, with the DNA
repair assay serving as a compliment to
enhance the efficiency of the mutagenicity
assay.
The complete battery of test systems
used in this project also includes
biological systems utilizing Aspergillus
nidulans (Scott et al., 1980); Glycine
max (Vig, 1975); and Tradescantia
(Sparrow et_ al., 1974) as indicator
organisms. The various types of chromo-
some damage which can be indicated by
these test systems include non-
disjunction, mitotic crossing over, major
chromosomal damage, recessive lethals,
translocation, and duplication (Scott and
Kafer, 1980; Nilan and Vig, 1976).
MATERIALS AND METHODS
For the present purpose, two
hazardous wastes were collected for study.
These were characterized and then extracts
were subjected to bioassays to detect
mutagenic activity. The wastes were also
incubated with soil to evaluate the fate
of the mutagenic compounds when exposed
to the soil environment. The same
wastes were applied to the large undis-
turbed lysimeters from which leachate
and runoff were collected. These samples
were concentrated on the XAD resin.
The organic residue was also evaluated
for mutagenic activity.
Waste
The two wastes used in this study
were collected from API oil-water
separators,,one from a petroleum refinery
and one from a petrochemical plant. Two
225
-------
TABLE 1« BIOLOGICAL SYSTEMS WHICH HAY BE USED TO DETECT GENETIC TOXICITY OF A HAZARDOUS HASTE
Organism
PROURYOTES
Bacillus subtilis
Escherichta coll
Salmonella typhimurtum
Streptomyces coelicolor
EUKAKYORS
Asperqlllus nidulans
menrompora craaaa
Saccharomyces cervisiae
Schisosaccharomyces pombe
PLAIITS
Tradmscsntia
Arabtdopsis thaliana
Bordeum vulqare
Pisum sativum
Triticum
Glvctne max
Vtcia faba
Allium cepa
Gene Mutation
Forward , reverse
Forward, reverse
Forward , reverse
Forward
Forward , reverse
Forward
Forward
Forward
Forward
Chlorophyll mutation
Chlorophyll mutations
Chlorophyll mutations
Morphological mutation
Chlorophyll mutation
Morphological mutation
Morphological mutation
Genetic Event Detected
Other Types of
Genetic Damage
DNA repair
DBA repair
DNA repair
DNA repair
DNA repair.
Chromosome aberrations
Not developed
Mitotic gene conversion
Mitotic gene conversion
Chromosome aberrations
Chromosome aberrations
Chromosome aberrations
Chromosome aberrations
Chromosome aberrations
Chromoaome aberrations
Chromosome aberrations
Chromosome aberrations
Metabolic
Activation
Mammalian
Mammalian,
plant
Mammalian,
plant
Not Developed
Mammalian,
plant
Mammalian
Mammalian
Mammalian
Plant
Plant
Plant
Plant
Plant
Plant
Plant
Plant
References
Felkner, Hoffman and Wells, 1979; Kada, Morija,
and Tanooka, 1977; Tanooka, Munakata and
Kitahara, 1978.
Green, Muriel and Bridges, 1976; Mohn, Eltenberger
and McGregor, 1974; Slater, Anderson and Rosen-
kranz, 1971; Speck, Soutella and Rosenkrans,
1978; Scott et aj. . , 1978.
Ames, McCaim and Yamasaki , 1975; Plewa and Gentile
1976; Skopek et al . , 1978.
Carere et al., 1975.
Bigani et al., 1974; Roper, 1971; Scott et al..
1978; Scott et al. , 1980.
DeSerres and Mailing, 1971; Cr.q, 1978; Tomlinson,
1980.
Brusick, 1972; Loprieno et al., 1974; Mortimer and
Manney, 1971; Parry, 15T7.
Brusick, 1972; Loprieno et al., 1974; Mortimer and
Manney, 1971; Parry, 1577.
Hauman, Sparrow and Schairer, 1976; Underbrink,
Schairer and Sparrow, 1973.
Redei, 1975.
Kumar and Chauham, 1979; Nicoloff, Gecheff and
Stoilov, 1979.
Ehrenburg, 1971.
Ehrenberg, 1971.
Vig, 1975.
Kihlman, 1977.
Harlmuthu, Sparrow, and Schairer, 1970.
-------
c
c
1
r-
r-
(7t
~*
5
A
c
fl
01
n
U
0
g
> a
i
^
S
3
2
i
i
i
i
1
I
i
r-
£ S
•H » r*
01 ffi
U -H
U Rt
0» Z
C S» •,
C Q r-j
0> fl
W -O
C 4-1
•n |
c
1 A
-------
TABLE 2. CHARACTERISTICS OF THE SLUDGES
CLASS % OF TOTAL HYDROCARBON
Oil Sludge
Refinery
Petrochemical
Extractable
Carbon
g C/100g
10
62
Ash
41
25
Water
46
12
Pentane
72
63
Benzene
22
36
Dichlorome thane
6
1
CRUDE SAMPLE
EXTRACT WITH
DICHLOROMETHANE IN NAOH
ORGANIC
PPT
AQUEOUS
BASE AND NEUTRALS
ACID
AQUEOUS
ORGANIC
ORGANIC
AQUEOUS
Figure 1. Schematic Diagram for Extraction Procedure Used to Separate Hazardous Waste
Fractions.
one lysimeter to which no waste was appli-
ed served as a control. Five monoliths
of each soil were encased in the 150 cm
deep lysimeters and brought to a central
location. A detailed description for the
installation of lysimeters is given by
Brown et^ _al. (1974). The leachate was
collected from porous ceramic suction cups
(Coors Type # 7001-P-6-C), which were
installed in the soil at the bottom of
the profile. The suction cups were
connected by nylon tubing to a 20 liter
glass bottle which was connected to a
vacuum manifold. Leachate samples were
collected once a week unless the rate of
water flow required a more frequent sampl-
ing schedule.
Runoff samples were collected by
means of a peristaltic pump. Glass be-
akers were placed in a small hole in the
corner of each lysimeter; the soil surface
in the lysimeter was sloped so that
runoff water could collect in the beaker.
A rainfall of greater than 1.27 cm
activated a switch which supplied power
to the pumps. The collected runoff was
pumped through nylon tubing to a 20 liter
glass bottle where the water was stored
in an air conditioned shed until collected.
Water samples were collected in amber
glass bottles and stored at 4°C until
processed.
Two methods were employed for the
extraction of water-bound organic com-
po.unds. Initially, a 20 cm3 bed was made of
non-polar XAD-2 resin (Applied Science
Laboratory, State College, PA) following
the methods of Hooper, et al..(1978).
Later in the study, it was decided,due to
the complex nature of the samples being
228
-------
OJ
S
2
•H
a
to
a)
E
n)
W)
nt
•o
u
aj
o
co
analyzed, to supplement the XAD-2 with a
moderately polar resin. Thus, samples
which were extracted in the latter portion
of the study were passed through a mixed
bed of 4.0 g of XAD-2 and 6.3 g XAD-7 or
approximately 20 cm^ of each resin as
suggested by Rappaport et^ _al. (1979).
The resins were washed prior to use by
swirling and decanting three times with
ten volumes each of acetone, methanol,
and distilled water. Washed resins were
stored at 4°C prior to use. Glass econo-
columns (Bio-Rad, Richmond, CA)* 1.5 x 50
cm^ were packed with 20 cm3 of XAD-2 resin
followed by 20 cm3 of XAD-7 resin. A
glass wool plug was placed above the resin
in order to trap soil particles. The
columns were flushed with 1,200 ml of
distilled water before loading the water
sample.
Mention of brand name does not constitute
endorsement.
Leachate or runoff water was placed in a
reservoir and allowed to pass through the
column by gravity flow at about 50 ml per
minute. After loading the water sample,
dry nitrogen was introduced into the column
to remove the residual aqueous phase, and
the column washed with 120 ml of distilled
water to remove residual histidine. The
adsorbed organic compounds were then
eluted with 160 ml of acetone. The acetone
extract was filtered through about 30 g of
anhydrous ^280)4 and Whatman No. 42 filter
paper into a flat bottom flask. Extracts
were reduced to less than ten milliliters
on a Brinkman-Bucci Model R roto-evapora-
tor, and taken to dryness in a screw-capped
glass culture tube under a stream of
nitrogen. Dimethyl sulfoxide (Sigma) was
added to the dried extract at a rate of
0.5 ml DMSO/1 of unconcentrated water, and
the resultant solution passed through a
0.2 urn average pore diameter Teflon filter
(Millipore-Fluoropore, Bedford, Mass.).
Samples were stored at 4°C prior to use.
Bioassay
The genetic toxicity of the extracted
samples was measured with two microbial
systems capable of detecting compounds
which produce point mutations or increased
lethal damage in DNA repair deficient
strains* The Salmonella/microsome assay
as described by Ames et al. (1975) was
used to measure the ability of a sample
to revert strains of bacteria to histidine
prototrophy. Samples were tested in two
strains at a minimum of three dose levels
with and without microsome activation
from Aroclor-1254 induced rat liver from
Litton Bionetics (Kensington, M.D.). The
procedures for the S-9 mix and the plate
incorporation assay were the same as
Ames et al. (1975), except that overnight
cultures were grown in 5 ml of Oxoid Broth
in a 125 ml Erlenmeyer flask. Positive
controls as well as solvent and sterility
controls were run. The control which was
used to verify metabolic activation was
10 ug 2-acetylaminofluorene, while
positive controls for TA 98 and TA 100
were 25 ug 2-nitrofluorine and 2 ug N-
methyl-N'-Nitro-N-nitrosoguanidine,
respectively. The genetic characteristics
of the Salmonella and Bacillus strains
used in this study are given in Table 4.
The Bacillus subtilis DNA repair
assay was used to evaluate samples which
229
-------
TABLE 3. CHARACTERISTICS OF SOILS UTILIZED IN BIODEGRADATION EXPERIMENTS
Norwood
Nacogdoches
Lakeland
Bastrop
sana
48.2
41.8
81.1
60.3
bllt
%
15.2
12.9
4.5
10.0
Clay O.M. CEC pH
meq/lOOe
36.6 1.4
45.3 1.3
14.4 0.7
29.7 1.0
19.6
17.2
0.3
27.4
7.69
5.95
6.45
6.86
N
70
47
42
160
P
94
3
20
25
K
— ppm
312
164
200
200
Ca
4,000
1,280
600
2,920
Mg
485
400
100
385
WP
12
8
1
6
FC SAT
18 33
20 36
5 23
22 25
Tsble 4. Microbial Systems Used to Datect Genetic Barrage
Bacteria
S. typhLiuriiEi
Strain
TA 1538
TA 98
TA 1535
TA 100
Genetic Event
Detected
fraircshift mutation
franiashift nutation
base-pair mutation
base-pair mutation
Mutation which Enhanca
Sensitivity*
hisD3052, rfa, AuvrB
hisD2052, rfa, AuvrB, pkM.01
hisG46, rfa, AurvB
hisG46, rfa, AurvB, pklflLOl
Reference
Ames, et al., 1975
B. subtilis
168 wt
inhibition of DNA
rec A8 '
rec E4
mc-1
hcr-9
fh2006-7
air wild type re^-et KL. di. , «/?
rec~, DRC~, trans"
rec~
mitomycin C sensitive
AuvrB
AuvrB, rec"
'hisD3052, hisG46 - mutation in histidine operon
rfa-deep rough characteristic-nutants lack lippcpolysachari.de cell wall, enhances peroeability
AuvrB-deficient in excision repair; sensitive to ultraviolet light
pkMlOl- contain R factor, plasmid which enhances sensitivity
rec" - deficient in reconbinant repair
DRC~ - blocked in donor-recipient complex formation
trans" - deficient in transformation and transduction
produced a toxic response in Salmonella.
The toxicity of the samples was compared
in six strains of Bacillus subtilis
(Table 4). The procedures used were the
same as in Felkner e_t^ _al. (1979). Over-
night cultures grown in Brain Heart
Infusion Broth (Difco) were streaked
radially from a centered sensitivity disk.
and the nutrient agar plates were grown
for 24 hours at 37°C. After the incuba-
tion period, the toxicity of the sample
was evaluated by comparing the zone of
inhibition produced.Controls included U.V.
sensitivity, 2 yg mitomycin C, and 2 ul
methylmethane sulfonate.
RESULTS AND DISCUSSION
Waste Characterization
Distribution of the mutagenic
activity in the acid, base, and neutral
fractions of the refinery and petrochemical
waste is given in Table 5. Additional
information is presented on the activity
of crude oil (Epler et^ al., 1978) and
cigarette smoke condensate (Kier et al.,
1976) for comparison purposes. Cumulative
mutagenic activity of both the refinery
and the petrochemical waste appear to be
greater than for unrefined crude oil
230
-------
TABLE 5. DISTRIBUTION OF MUTAGENIC ACTIVITY IN FRACTIONS OF SEVERAL COMPLEX MIXTURES
Revertants/mg
Sample
Refinery Waste
Petrochemical Waste
Crude Oil
Cigarette Smoke
Acid
813
498
10
5,450
Base
2,308
278
277
430
Neutral
33
246
150
1,940
Reference
Epler et_al. , 1978
Kier et^ al. , 1976
(437 revertants/mg), but considerably
less than for cigarette smoke condensate
(7,820 revertant/mg). For the two wastes
studied, the basic fraction of the
refinery waste induced the greatest
response, generating more than 2,000
revertants/mg of residue. The mutagenic
activity of the petrochemical waste was
distributed more evenly among the acid,
base and neutral fractions, with the
greatest activity detected in the acid
fraction (498 revertants/mg).
Biodegradation
The mutagenic activity in unamended
soil, soil with 5% waste which had not
been incubated, soil with 52 waste and
soil with 20% waste both which had been
incubated for 180 days at 30°C are given
in Table 6. No activity was evident in
any of the fractions from the unamended
soils. At the highest application rate,
the number of revertants induced by the
saturate and aromatic fractions was
reduced by 21 and 26% respectively. The
toxic potential of the saturate and
aromatic fractions was reduced to a
greater extent at the lowest application
rate, although the dichloromethane frac-
tion representing condensed ring compounds
failed to exhibit a similar reduction at
either application rate. Other research-
ers (Dibble and Bartha, 1979) have
reported an increase in the concentration
of residual condensed ring fractions after
soil incubation. This increase may be due
to the accumulation of metabolites from
other fractions, and could explain the
increase in observed mutagenic activity
in this fraction. With adequate time,
the condensed ring fractions should also
degrade, reducing the activity of the
residual materials.
When the mutagenic activity of incubated
and non-incubated soil is expressed as
revertants per.gram of soil, the reduction
produced by chemical and biological
activity becomes more evident. The results
in Table 7 show a significant reduction in
the mutagenic activity of both the saturate
and aromatic fractions. As discussed
earlier, the dichloromethane extract did
not show a reduction in mutagenic activity
after six months of incubation.
The extraction efficiency of these
procedures for hydrocarbons and mutagenic
activity from soil are given in Table 8.
These results indicate that less than 5%
of the applied hydrocarbons remained in the
soil after solvent extraction. The
efficiency of the recovery of mutagenic
activity from soil was significantly less
than for the hydrocarbons. The dichloro-
methane extract of the 'soil-waste mixture
recovered only 67% of the mutagenic
activity extracted from the waste alone
with dichloromethane. It is unknown at
this time whether the reduction was a
result of soil binding, or an artifact
created by the solvent extraction. Work
conducted in the future will attempt to
better define the cause of this deficiency
so that the extraction efficiency can be
improved.
Leachate and Runoff Water Samples
The mutation frequency of the leachate
and runoff water collected from field
lysimeters was calculated using the method
of Commoner (1976) in which a mutation
frequency greater than one indicates a
doubling of revertant colonies. The
231
-------
TABLE 6. EFFECT OF SOIL INCUBATION ON MUTAGENIC ACTIVITY OF PETROCHEMICAL WASTE
Sample
Revertant/mjJ Solvent Extract
Petroleum Petroleum Ether: Dichloro-
Ether Extract Dichloromethane methane
Soil
5% Waste and Soil
0
0
0
(Not incubated)
5% Waste and Soil
(Incubated)
20% Waste and Soil
(Incubated)
33
24
26
479
107
354
100
112
111
TABLE 7. MUTAGENIC ACTIVITY EXTRACTED FROM SOIL AND WASTE AMENDED SOIL
Revertant/g soil
Sample Petroleum Petroleum Ether: Dichloro-
Ether Extract Dichloromethane methane
Soil
5% Waste and Soil
0
1.3
0
1.9
0
3.8
(Not incubated)
5% Waste and Soil
(Incubated)
0.3
1.8
3.9
TABLE 8. EFFICIENCY OF EXTRACTION OF HYDROCARBONS AND MUTAGENIC ACTIVITY FROM SOIL
EXTRACTION EFFICIENCY (%)
Sample
Hydrocarbon
Mutagenlc Activity
Petroleum
Ether Extract
99
85
Petroleum Ether :
Dichloromethane
98
95
Dichloro-
methane
95
67
mutation frequency was calculated as:
E-C
MF =
Ave.
where E = number of revertant colonies on
an experimental plate with the
residue from 200 ml of water
C = number of revertant colonies on
Ave.
the solvent control plates from
the same day with the same tissue
preparation
overall average of revertant
colonies for solvent control
plates for the whole period of
study.
Using this scheme, a sample which
232
-------
induces a mutation frequency between 1
and 1.5 is considered to have a border-
line mutagenic potential, while a com-
pound which induces a mutation frequency
greater than 1.5 is considered to be a
positive mutagen.
The analysis of water collected from
the refinery waste-amended soil (Table 9)
indicates that the leachate water samples
with the greatest mutagenic activity were
collected within two months after the
first waste application. This data
represents a summary of the mutation
frequencies calculated from dose response
curves where the leachate or runoff water
was tested at a minimum of three dose
levels. The first leachate water collect-
ed from the Bastrop clay on the 47th day
after application and from the Norwood
soil on the 17th day after application,
induced mutation frequencies of 6.7 and
5.1 respectively. All but one of the
subsequent leachate samples collected
induced a mutation frequency equal to or
less than 1.5. None of the water samples
collected from unamended soils induced a
significant increase in the number of
revertants at the dose levels of water
concentrate tested.
While the runoff water samples
collected from the Bastrop lysimeters did
exhibit a reduction in toxic potential
with time, the potential remained signi-
ficantly above background levels. Runoff
samples collected from the Bastrop soil
40 and 145 days following the second
waste application induced mutation frequen-
cies of 9.0 and 4.2 respectively.
After the second waste application on
Lakeland sand, eight runoff samples were
collected over a period of 138 days.
Five of the first six samples produced a
clear positive response in the mutageni-
city assay. The dose-response curves
for these samples, presented in Figure 3,
indicate that samples collected 131 and
138 days after waste application failed
to induce a doubling of revertant colonies.
Thus, the toxic potential of runoff water
from refinery waste-amended Lakeland sand
appears to have returned to background
levels within five months after the
second waste application. The dose-
response curves for runoff water collect-
ed from unamended soil (Figure 4) indicate
that even at higher water concentrate
dose levels, there was no mutagenic
response.
Analysis of the leachate and runoff
water collected from petrochemical waste-
amended soils are presented in Table 10.
Only three of the leachate samples
collected induced a positive response
in the mutagenicity assay. Two of these
were collected from the Norwood soil 38
and 52 days after the first and third
waste application,respectively. One
sample collected more than six months
after the second waste application from
the Nacogdoches soils induced a mutation
frequency of 3.3. None of the leachate
samples collected from the Bastrop soil
induced a mutation frequency greater than
1.5.
The analysis of runoff water from
petrochemical waste-amended soil indicates
that large quantities of mutagenic com-
pounds are being released at the soil
surface. Runoff water collected from the
Bastrop soil 38 days after the first waste
application, and 113 days after the second
waste application induced mutation
frequencies of 14.3 and 8.4 respectively.
One runoff sample from the Norwood soil
and two from the Lakeland sand, all of
which were collected during the winter,
had mutation frequencies greater than 1.5.
The results of the analysis of
leachate and runoff water by the Bacillus
subtilis DNA repair assay are given in
Tables 11 and 12 . These results indicate
that there is a negative correlation
between mutagenicity in Salmonella and
toxicity to Bacillus subtilis. None of
the runoff samples which were mutagenic
were also toxic in the DNA repair assay,
while most of the samples which were toxic
to j^. typhimurium at higher dose levels
produced increased lethal damage in repair
deficient strains of 15. subtilis. A
runoff sample from petrochemical waste
amended soil collected 130 days after the
second waste application from the Lakeland
sand produced a clear positive response
in the DNA repair assay and was not
mutagenic in the Salmonella assay. Thus,
it is evident from these results that the
DNA repair assay serves as a compliment
to the mutagenicity assay by responding to
mutagenic compounds not detected by a
Salmonella assay and by detecting a second
type of genetic damage.
233
-------
TABLE 9 . MUTATION FREQUENCY OF
0 17 47 67
LEAQJAIE
RUNOFF
LEACHATE
HHIOFF AND TJi.vrt.ATg WATER COLLECTED PROM I
0, 202 and 403. TOE STMBOL
155 228 250 260 275 293 318
IEFIKERY WASTE AMENDED SOILS. WASTE WAS APPLIED ON DAY
(-) MEANS NO SAMPLES WAS AVAILABLE
TIME (DAYS)
322 330 335 345 355 375 388 395 410 420
465 475
ided soil - - 0.5 0 0.1
RUHOFF
Itmended soil - - - 0 0.3
0.1 - 0.8 0.1
- 0.2
A 58 day* after 2nd application
Q 67 day* after 2nd application
Q 70 days aftar 2nd application
A 8 day* aftar 3rd application
30
20
200
400
Volume (ml.)
600
800
VokmM)
Wgure 3, Hitagenlcity. as measured with S. cypnimim-.
strain IA98 without enzyK activation of runoff
water fro. refinery waste amended Lakeland sand
Figure 4. Mutagenicity, as measured with
S_. typhimurium TA 98, of run-
off water from unamended soils.
Line of significance ( ) is
equal to two times the number
of spontaneous revertants.
CONCLUSIONS
The results obtained from preliminary
testing with two petroleum sludges
indicate that short-term bioassays can be
a valuable tool for evaluating the muta-
genic activity of a hazardous waste. The
analysis of the refinery and petrochemical
waste indicates that they are more muta-
genic than crude oil" and that they
possess less mutagenic activity on a per
weight basis than cigarette smoke conden-
sate.
These procedures also seem adaptable
for the determination of the treatment
potential of a hazardous waste. Although
it has been determined that mutagenic
materials can be extracted from soil,these
methods will require further development
to determine the extraction efficiency of
various solvents. The determination of
the extent of reduction of the hazardous
characteristics of a land applied waste is
an essential element to the success of the
land treatment method of waste disposal.
Incubation studies revealed that the
mutagenic activity of the waste added to
soils decreased over a six month period.
Bioassays have also been shown to be
effective for evaluating the level of
234
-------
TABLE 10. MUTATION FREQUENCY OF LEACHATE AND RUNOFF WATER COLLECTED FROM PETROCHEMICAL WASTE AMENDED SOILS
WASTE WAS APPLIED ON DAY 0. 202 AND 403. THE SYMBOL ( - ) MEANS NO SAMPLE WAS AVAILABLE
Bastrop
Nacogdoches
Norwood
RUNOFF
TIME (DAYS)
° 38 S3 60 70 148 153 183 190 202 245 255 263 268 280
LEACHATE
0.7 1.2 0 0.3
1.1
1.2 1.3
Bastrop
Norwood
Lakeland
TABLE 1Q( CONTINUED) .
- 14.3
- 0.5 - ------ 1.1 -
--- - - - - - i.o - -
11 - i ^
1.3
0.7
295 305 315
325
TIME (DAYS)
332 340 345
383 410 420 430 455 468 495
LEACHATE
Bastrop
Nacogdoches
Norwood
RUNOFF
Bastrop
Nacogdoches
Norwood
LEACHATE
Unamended soil
RUNOFF
Unamended soil
-
-
-
_
0.8 -
_
_
- 3.3
0.6
0.5
0.9
-
-
1.0
0.1
-
0.7
3.1
-
-
1.5
-
-
0.5
0.7
1.8
8.4
1.5
4.2
0.6
0.7
2.8
1.6
0.6
0.7
0.9
0.9
0.8 -
0.1
0.2 0.
mutagenic activity in the effluent from a
land treatment facility. Only low levels
of mutagenic activity were detected in a
few samples of leachate water collected
shortly after waste application. This
indicates that the soil retained the
majority of the mutagenic constituents
and prevented them from entering the
groundwater. The runoff water collected
during the year after waste application
exhibited decreasing amounts of activity.
The most active sample of runoff water
was 14.3 times the background level. At
this level, 3 liters of water could
contain the activity of one cigarette.
As these systems are further
developed, and the testing protocol better
defined, short-term bioassays will
increase in their utility for the detec-
tion and characterization of mutagens and
potential carcinogens. Test systems
which are used as part of a monitoring
system should be selected with regards to
the types of compounds in the waste, and
the types of damage caused by these
compounds. Additional work to be performed
as part of this project will include the
complete characterization of three addi-
tional hazardous wastes, along with a
comprehensive greenhouse and lysimeter
study to determine the treatment potential
and the environmental fate of hazardous
waste constituents. Although additional
work is needed to more clearly define the
advantages and limitations of short-term
testing, the preliminary results indicate
that short-term tests can be a valuable
tool for use in defining the mutagenic
characteristics of a waste, for evaluating
their treatment potential, and for evalua-
ting their fate and mobility from land
treatment disposal sites.
235
-------
TABLE 11. COMPARISON OF LETHAL EFFECT OF WATER EXTRACTS ON DNA REPAIR DEFICIENT
AND PROFICIENT STRAINS OF B. subtilis
SAMPLE
INHIBITION RADIUS (mm)
168 wt recA8 mc-1 recEA hcr-9 fh2006-7 Response
RUNOFF REFINERY
Soil
Bastrop
Bastrop
Bastrop
Lakeland
Lakeland
Lakeland
Volume
400 ml eq
400 ml eq
400 ml eq
400 ml eq
400 ml eq
400 ml eq
Day
335
345
355
275
293
318
RUNOFF CONTROL
Lakeland 400 ml eq 315
POSITIVE CONTROL
Me thy line thane
sulfonate 2 ul
33
26
0
3
3
21
17
2U
TABLE 12 . COMPARISON OF LETHAL EFFECT OF HATER EXTRACTS ON DNA REPAIR DEFICIENT AND PROFICIENT
STRAINS OF B. subtilis
SAMPLE
INHIBITION RADIUS (mm)
168 wt recAS mc-1
recEA
hcr-9
fh2006-7 Response
RUNOFF PETROCHEMICAL
Soil
Bastrop
Bastrop
Bastrop
Lakeland
Lakeland
Lakeland
Lakeland
Lakeland
Lakeland
Norwood
Norwood
Norwood
Norwood
Norwood
Norwood
Norwood
Volume
400 ml eq
400 ml eq
400 ml eq
400 ml eq
400 ml eq
400 ml eq
400 ml eq
400 ml eq
400 ml eq
400 ml eq
400 ml eq
400 ml eq
400 ml eq
400 ml eq
400 ml eq
400 ml eq
Day
255
315
430
280
295
315
325
332
340
280
295
315
325
332
345
352
6
8
9
2
5
4
5
3
1
2
4
4
4
7
5
4
10
7
5
3
6
4
6
4
1
5
5
7
4
7
3
3
9
10
5
2
4
3
5
4
1
5
4
5
3
6
4
2
-
-
8
5
-
4
5
5
-
5
5
5
8
8
5
~
7
9
9
4
5
4
5
8
1
3
5
4
5
6
8
4
9
5
10
4
5
4
7
10
1
3
4
4
6
9
10
3
+
+
-
+
-
-
-
+ +
-
+
-
+
+
-
+ +
"
REFERENCES
1. Ames, B. N., J. McCann, and E.
Yamasaki. 1975. Methods for detect-
ing carcinogens and mutagens with the
Salmonella/mammalian microsome muta-
genicity test. Mutat. Res. 31:347.
2. Anderson, D. 1975. The selection and
induction of 5-iodo-2 deoxyuridine and
thymidine variants of P 388 mouse
lymphoma cells with agents which are
used for selection. Mutat. Res. 33:
399.
236
-------
3. Arlett, C. F. 1977. Mutagenicity
testing with V79 Chinese hamster cells.
JLn B. J. Kilbey, M. Legator, W.Nichols
and C. Ramel (ed.) Handbook of muta-
genicity test procedures, p. 175-191.
Elsevier Biomedical Press, Amsterdam.
4. Beek, B., G. Klein, and G. Obe. 1980.
The fate of chromosomal aberrations
in a proliferating cell system. Biol.
Zentralbl. 99(1):73-84.
5. Bigani, M., G. Morpurgo, R. Pagliani,
A. Carare, G. Conte, and G. DiGuideppe.
1974. Non-disjunction and crossing
over induced by pharmaceutical drugs
in Aspergillis nidulans. Mutat. Res.
26: 159.
6. Brown, K. W., C. J. Gerard, B. W. Hipp
and J. T. Ritchie. 1974. A procedure
for placing large undisturbed monoliths
in lysimeters. Soil Sci. Soc. of
Amer. 38: 981-983.
7. Brown, K. W., and K. C. Donnelly.
1981. Genetic toxicity of water
concentrates from the effluent of a
waste oil storage pond. (In Review).
8. Brusick, D. J. 1972. Induction of
cyclohexamide resistant mutants in
Saccharomyces cerevisiae with N-
methyl-N-nitro-N-nitrosoguanine and
ICR-170. J. Bacteriol. 109: 1134.
9. Carere, A., G. Morpurgo, G. Cardamone,
M. Bignami, F. Aulicino, G. DiGuisepie,
and C. Conti. 1975. Point mutations
induced by pharamaceutical drugs.
Mutat. Res. 29: 235.
10. Clive, D. W., W. G. Flamm, M. R.
Machesko, and N. H. Bernheim. 1972.
Mutational assay system using the
thymidine kinase locus in mouse
lymphoma cells. Mutat. Res. 16:77.
11. Clive, D. W. and J.F.S. Spector. 1975.
Laboratory procedure for assessing
specific locus mutations at the TK
locus in cultured L5178Y mouse lympho-
ma cells. Mutat. Res. 31: 17.
12. Dean, B. J. and K. R. Senner. 1977.
Detection of chemically induced
mutation in Chinese hamsters. Mutat.
Res. 46: 403.
13. Deserres, F. J. and H. V. Mailing.
1971. Measurement of recessive
lethal damage over the entire genome
and at two specific loci in the ad-3
region of a two component heterokar-
yon of Neurosporo crassa. 2 : 311-341,,
In A Hollaender (ed.). Chemical
mutagens, principles and methods for
their detection. Plenum Press. New
York.
14. Ehrenburg, L. 1971. Higher plants.
2: 365-386. Iri A. Hollaender (ed.).
Chemical mutagens, principles and
methods for their detection. Phenum
Press, New York.
15. EPA. 1979. Subtitle C. Resource and
conservation recovery act of 1976;
Draft environmental impact statement.
Office of Solid Waste, EPA, Washing-
ton, D.C.
16. Epler, J. L., Jennifer A. Young,
A. A. Hardigree, T. K. Rao,
M. R. Guerin, I. B. Rubin, C. H. Ho,
and B. R. Clark. 1978. Analytical
and biological analyses of test
materials from the synthetic fuel
technologies. Mutat. Res. 57: 265-
276.
17. Felkner, I.C., K. M. Hoffman, and
B.C. Wells. 1979. DNA - damaging
and mutagenic effect of 1,2- Dimethy-
Ihydrazine on bacillus subtilis
repair-deficient mutants. Mutat. Res.
28: 31-40.
18. Green, M. H. L., W. J. Muriel, and
B. A. Bridges. 1976. Use of a
simplified fluctuation test to detect
low levels of mutagens. Mutat. Res.
38:33.
19. Hooper, K., C. Gold, and B. N. Ames.
1978. Development of methods for
mutagenicity testing of wastewater
and drinking water samples. Report
to Water Resources Control Board,
State of California, Sacramento,
California.
20. Kada, T., M. Morija, and Y. Shirasu.
1974. Screening of pesticides for
DNA interactions by REC-assay and
mutagenic testing and frameshift
mutagens detected. Mutat. Res.26:243.
237
-------
21. Kier, L. D., E. Yamasaki, and B.N.
Ames. 1974. Detection of mutagenic
activity in cigarette smoke condensa-
tes. Proc. Nat. Acad. Sci. USA, 71
(10) 4159-4163.
22. Kihlman, B. A. 1977. Root tips of
Vicia fava for the study of the
induction of chromosomal aberrations.
p. 389-410. In B. J. Kilbey, M.
Legator, W. Nichols, and C. Ramel
(ed.). Handbook of mutagenicity test
procedures. Elsevier Biomedical Press,
Amsterdam.
23. Loprieno, N., R. Barale, C. Bauer,
S. Baroncelli, G. Bronzetti, A.
Cammellini, A. Cinci, G. Corsi,
C. Leporini, R. Nieri, M. Mozzolini,
and C. Serfa. 1974. The use of
different test systems with yeasts for
the evaluation of chemically induced
gene conversions and gene mutations.
Mutat. Res. 25: 197.
24. Marimuthu, K. M., A. H. Sparrow, and
L. A. Schairer. 1970. The cytologi-
cal effects of space flight factors,
vibration, clinostate, and radiation
on root tip cells of Tradescantia
and Allium cepa. Radiation Res. 42:
105.
25. Mohn, G. , J. Eltenberger, and D. B.
McGregor. 1974. Development of
mutagenicity tests using Escherichia
coli K-12 as an indicator organism.
Mutat. Res. 23:187.
26. Mortimer, R. K. and T. R. Manney.
1971. Mutation induction in yeast.
1:289-310. In A. Hollaender (ed.).
Chemical mutagens, principles and
methods for their detection. Plenum
Press, New York.
27. Nauman, C. H., A. H. Sparrow, and
L. A. Schairer. 1976. Comparative
effects of ionizing radiation and two
gaseous chemical mutagens on somatic
mutation induction in one mutable and
two non-mutable clones of Tradescantia.
Mutat. Res. 38: 53-70.
28. Neill, J. P., P. A. Brimer, R.
Machanoff, G. P. Hirseh, and A. W.
Hsie. 1977. A quantitative assay of
mutation induction at the HGPRT locus
in Chinese hamster ovary cells:
development and definition of the
system. Mutat. Res. 45:91.
29. Nilan, R. A. and B. K. Vig. 1976.
Plant test systems for detection of
chemical mutagens. p. 143-170.
In A. Hollaender (ed.). Chemical
Mutagens. Vol. 4. Plenum Publishing
Co., New York.
30. Ong, T. N. 1978. Use of the spot,
plate and suspension test systems
for the detection of the mutagenicity
of environmental agents and chemical
carcinogens in Neurospora crassa.
Mutat. Res. 54:121-129.
31. Parry, J. M. 1977. The use of
yeast cultures for the detection of
environmental mutagens using a
fluctuation test. Mutat. Res. 46:
165.
32. Perry, P. and H. J. Evans. 1975.
Cytological detection of mutagen-
carcinogen exposure by sister chroma-
tid exchange. Nature (London). 258:
121.
33. Plewa, M. J. and J. M. Gentile. 1976.
The mutagenicity of atrazine: a
maize-microbe bioassay. Mutat. Res.
38:287-292.
34. Rappaport, S.M., M. G. Richard, M.C.
Hollstein, and R. E. Talcott. 1979.
Mutagenic activity in organic waste-
water concentrates. Envir. Sci. and
Technol. 13(8) 957-961.
35. Redei, G. P. 1975. Arabadopsis as
a genetic tool. Ann. Rev. Genet. 9:
111-125.
36. Rinkus, S. J., and M. S. Legator.
1979. Chemical characterization of
465 known or suspected carcinogens
and their mutagenic activity in the
Salmonella typhimurium system.
Cancer Research. Volume 39: 9.
p. 3289-3804.
37. Roper, J. A. 1971. Aspergillus.
I_n A. Hollaender (ed.). Chemical
mutagens, principles and methods for
their detection. Plenum Press, New
York.
38. Scott, B. R., A. H. Sparrow, S. S.
Lamm, and L. Schairer. 1978. Plant
238
-------
metabolic activation of EDB to a
mutagen of greater potency. Mutat.
Res. 49: 203-212.
39. Scott, B. R., E. Kafer, G. L. Dorn,
and R. Stafford. 1980. Asperigillus
nidulans: systems and results of
test for induction of mutation and
mitotic segregation. Mutat. Res.
(In press).
40. Shiau, S.Y., R. A. Huff, B. C. Wells,
and I.C. Felkner. 1980. Mutageni-
city and DNA-Damaglng activity for
several pesticides tested with Baci-
llus mutants. Mut. Res. 71: 169-179.
41. Skopek, T. R., J. L. Liber, J. J.
Krowleski, and W. G. Thilly. 1978.
Quantitative forward mutation assay
in Salmonella typhimurium using 8-
azaguanine resistance as a genetic
marker. Proc. Nat'l. Acad. Sci.
75:410.
42. Slater, E., M. D. Anderson, and
H. S. Rosenkranz. 1971. Rapid
detection of mutagens and carcinogens.
Cancer Res. 31: 970.
43. Soderberg, K., J. T. Mascarello,
G. Breen, and I. E. Scheffler. 1979.
Respiration deficient Chinese
hamster cell mutants genetic charac-
terization. Somatic Cell Genet. 5(2)
225-240.
44. Sparrow, A. H., L. A. Schairer and
R. Villalobos-Pietrini. 1974.
Comparison of somatic mutation rates
induced in Tradescantia by chemical
and physical mutagens. Mutation Res.
26 : 265-276.
45. Speck, W. T., R. M. Santella, and
H. S. Rosenkranz. 1978. An evalua-
tion of the prophage induction
(inductest) for the detection of
potential carcinogens. Mutat. Res.
54: 101.
and X-radiation in spores and
vegetative cells of Bacillus subti-
llis. Mutat. Res. 49: 179-186.
48. Thilly, W. G., J. G. DeLuca,
I.V.H. Hoppe, and B. W. Penmann.
1976. Mutation of human lymphoblasts
by methylnitrosourea. Chem. Biol.
Interact. 15:33.
49. Tomlinson, C.R. 1980. Effects of
pH on the mutagenicity of sodium
azide in neurospora-crassa and
Salmonella-typhimurium. Mutat. Res.
70(2): 179-192.
50. Underbrink, A.G., L. A. Schairer,
and A. H. Sparrow. 1973. Tradescan-
tia stamen hairs. A radiobiological
test system applicable to chemical
mutagenesis. 3: 171-207. Ln A.
Hollaender (ed.). Chemical mutagens,
principles and methods for their
detection. Plenum Press, New York.
51. Vig, B. K. 1975. Soybean (Glycine
max): a new test system for study of
genetic parameters as affected by
environmental mutagens. Mutat.
Res. 31: 49-56.
52. Von Borstel, R. C. and R. H. Smith.
1977. Measuring dominant lethality
in Habrobracon. p. 375-387. In
B. J. Kilbey, M. Legator, W. Nichols
and C. Ramel (ed.). Handbook of
mutagenicity testing procedures.
Elsevier Biomed. Press, Amsterdam.
53. Wurgler, F. E. and E. Vogel. 1977.
Drosophila as an assay system for
detecting genetic changes, p. 335-
373. In B. J. Kilbey, M. Legator,
W. Nichols, and C. Ramel (ed.).
Handbook of mutagenicity test
procedures. Elsevier Biomed. Press,
Amsterdam.
46. Stetka, D. G. and S. Wolff. 1976.
Sister chromatid exchange as an
assay for genetic damage induced by
mutagen-carcinogens. Mutat. Res.
41: 333.
47. Tanooka, H., N. Munakata, S.Kitahara.
1978. Mutation induction with UV-
239
-------
CLOSURE TECHNIQUES AT A PETROLEUM LAND TREATMENT SITE
John E. Matthews
Fred M. Pfeffer
Lawrence A. Weiner
U.S. Environmental Protection Agency
ABSTRACT
Land treatment is noted in the hazardous waste regulations as one of the technologies for
the treatment and disposal of hazardous wastes. Interim status regulations on land treat-
ment were published in the Federal Register, May 19, 1980 (Vol. 45 FR 33153). These
regulations require that the owner or operator of a hazardous waste land treatment facil-
ity develop a closure plan that will control the release of contaminants from the site.
Two objectives of the EPA land treatment research program are: (1) determine the extent
of migration of pollutants through the soil at closed land treatment facilities, and (2)
compare the efficacy of different closure techniques. Two projects that have recently
been funded will attempt to satisfy these objectives.
The first project will be conducted by the University of Oklahoma. Three closed oily
waste land treatment sites will be selected to provide information on the extent of con-
taminant migration in the soil. The second study, a cooperative agreement between EPA and
the American Petroleum Institute, will compare the efficacy of four closure techniques
over a three-year period—(1) revegetating the site, (2) closing the site like a landfill,
(3) removing the soil, and (4) leaving the site as is.
INTRODUCTION
Land treatment is one of the land dis-
posal technologies for the management of
hazardous wastes. The practice involves
the application of waste to the soil fol-
lowed, in most cases, by incorporation of
the waste into the soil by conventional
farm equipment (e.g., disc, rototiller).
The purpose of land treatment is the de-
gradation and attenuation of wastes by
biological, chemical, and physical reac-
tions occurring in or on the soil. The
Agency considers land treatment to be
especially effective for hazardous wastes
that are high in degradable organics and
relatively low in heavy metals.
Since the general objective of land
treatment is the microbial degradation of
organic waste constituents, this waste
management practice has been used primarily
to treat oily waste. The petroleum in-
dustry has been utilizing the land treat-
ment concept for residual disposal for at
least 25 years [Berkowitz, (1) Huddleston
(3)]. Several existing petroleum refin-
eries use some modification of this treat-
ment method [CONCAWE (2)]. Examples of the
application of land treatment by U.S. re-
fineries are given in Table 1.
Interim status regulations on land
treatment of hazardous wastes were published
by EPA in the Federal Register, May 19, 1980,
(45 FR 33153). The standards address all
aspects of land treatment including: con-
trol of surface water run-on and contaminated
run-off, waste analysis, monitoring, and
closure/post-closure care. The standards
for closure are couched in a framework
that permits the manager or operator of a
particular site to develop and implement a
closure plan that best suits the site. The
standards provide direction in the form of
general human health and environmental ob-
jectives which must be addressed by the
closure plan. In addition, certain factors
240
-------
TABLE 1. EXAMPLES OF THE APPLICATION OF LAND TREATMENT BY U.S. REFINERIES
Refinery
1
2
3
Type of Sludge
Separator
sludge
Separator
sludge; tank
bottoms; stable
emulsions.
Oily sludge
from lagoons.
Farming
Technique
Spread by
bulldozer;
Disc into
soil after
drying for
1 month.
4 sections
worked in
rotation &
mixed with
bulldozer
2-4 times
per month.
Single ap-
plication
of 10-yr.
accumula-
tion from
lagoons ;
mixed by
discing.
Application Treatment
Rate Experiences Soil Type Land Area
10-12 cm More than Clay 75-100 acres
depth of sludge 20 years
15 cm of sludge Since 1961 Clay 7 acres
mixed with 15 cm
of soil.
10-12 cm of Since 1969 Silty loam
sludge mixed surface; clay
with soil to subsoil.
depth of 45 cm.
Desalter &
tank cleanings,
separator bot-
toms; spill
cleanings; bio-
sludge; filter
clays.
Land-spread- 8-15 cm depth
ing on quad- of sludge mixed
rants during with soil to
9 months of depth of 15-20
year; mixed cm.
by discing
once a week.
Experimental
scale 1972.
Full scale since
mid 1973.
8 acres
Separator sludge; Spreading
biosludge.
8-10 cm depth
followed by of sludge.
disc-harrowing
when partially
dried.
-------
are identified which must be considered in
the closure plan when addressing the above
objectives.
The use of performance standards for
closure was preferred because the dearth of
information on closing land treatment
facilities precluded setting design and
operating standards. Since the land treat-
ment process is relatively new, virtually
no experience with closure has been gained.
The projects described below are expected
to provide information on the fate of con-
taminants in closed land treatment facil-
ities and the efficacy of potential closure
techniques.
CLOSURE OPERATIONS
The objective of a closure operation
is to prevent the occurrence of any future
environmental or health hazards as mandated
by RCRA.
Under the interim status regulations,
the owner or operator of a land treatment
facility must develop and implement a
facility closure plan, the terms of which
are enforceable against him. The plan must
address four objectives:
(1) controlling migration of hazard-
ous waste and hazardous waste constituents
into groundwater.
(2) controlling release of contami-
nated runoff to surface water.
(3) controlling release of airborne
particulate contaminants.
(4) compliance with standards
established for food crops.
The owner or operator must consider a
range of factors affecting the ability of
the facilities to meet these objectives.
These factors include: the waste, the
climate, the site location, the soil, and
the depth of contaminant migration. He
must also consider the applicability of
various closure methods, including removal
of the soil, runoff collection and treat-
ment, use of cover materials, diversion
structures, and additional monitoring.
Two objectives of the EPA land treat-
ment research program are: (1) determine
the extent of migration of pollutants through
the soil at closed land treatment facilities,
and (2) compare the efficacy of different
closure techniques. In order to satisfy
these objectives and support the regulatory
efforts of the Office of Solid Waste, EPA
has recently funded two projects to satisfy
these objectives. Following is a discussion
of the proposed experimental designs for
these two projects.
EVALUATION OF CLOSURE TECHNIQUES AT OILY
WASTE LAND TREATMENT FACILITIES
The objective of this study is to obtain
information, over a 3-year period, on the
efficacy of four potential closure techniques
for oily waste land treatment facilities.
The project is being performed under a
cooperative agreement between the American
Petroleum Institute (API) and EPA.
The four closure techniques to be in-
vestigated include: (1) establish a vege-
tative cover on the site, (2) close the
site as a landfill, (3) remove the contami-
nated soil, and (4) leave the site as is.
Establishing a vegetative cover may be
the most economical method if it will pre-
vent erosion of the soil that was used as a
treatment medium. The effects a vegetative
cover will have on runoff quality, waste
migration, and degradation on any remaining
waste residuals will be investigated. Be-
cause of the inhibitory effect a vegetative
cover has on waste degradation, it may be
advantageous to postpone covering the site
until most of the waste residuals have
degraded.
Closing the site as a landfill will re-
quire the placement of a natural or synthetic
cap over the contaminated soil. Although
capping a land treatment facility may retard
infiltration of rainwater and prevent ero-
sion, it will also alter or even curtail the
biological degradation of remaining waste
residuals. In addition, the previously
aerobic zone of incorporation will become
anaerobic. Under anaerobic conditions
biological degradation is slow and incom-
plete. The oxidation state of any metals
bound to the soil will change, and the pH of
the soil will slowly drop as organic acids
are produced. These phenomena and the
effects they will have on contaminant mi-
gration will be investigated.
Removing the soil from a closed land
treatment facility requires only engineer-
ing and economic calculations. The costs
242
-------
for soil removal and the landfilling of
the soil can be estimated from existing
construction practices. After the soil is
removed, new soil will be brought in, and
the site will be regraded and a vegetative
cover established. A comparison will be
made of the cost-effectiveness of soil
removal versus a waste-in-place approach
with subsequent post-closure monitoring.
The fourth closure technique to be
evaluated is leaving the site as is. Ef-
forts that might be employed to retard con-
taminant release include: tilling the
site and adding fertilizer to foster fur-
ther degradation of any waste residuals;
maintaining the soil pH at 6.5 or greater
to immobilize metals.
Work Plan
Work for this project will be conduct-
ed in two phases—(1) development of the
site and (2) experimental work.
Phase 1 - Development of the Site
An active site, or portion thereof,
will be selected to be used in the study.
The experimental site will be of sufficient
size to allow at least six smaller plots to
be identified. These smaller plots will be
isolated from each other by barriers. The
barriers will not be made of any materials
which might add chemical species to the
soil or water and will not protrude above
the ground so as to interfere with or
change the normal wind direction.
A control plot will be set up on
similar soils nearby that have not received
waste. Background levels of contaminants
in the soil and runoff water will be
determined and compared with the levels
from the experimental plots.
Four of the six experimental plots
will have the following conditions:
(1) left in the natural state, (2) vege-
tative cover, (3) removal of contaminated
soil, placement of new soil, regraded, and
vegetative cover established, and (4) capped
like a landfill. The two remaining
experimental plots will be tilled at regu-
lar intervals, and one plot will receive
fertilizer.
Rainfall infiltrometers will be in-
stalled for use in determining the soil loss
for each plot. Lysimeters will be used to
'determine the vertical transport of contami-
uants in the soil-pore water.
Phase 2 - Experimental Work
Soil loss and infiltration of water
will be studied using either natural or arti-
ficial rainfall. In either or both cases,
the procedures to be used and the advantages
and disadvantages of these techniques will
be described.
Runoff and infiltration water and soil
losses will be analyzed for residual contami-
nants left in the soil. Individual classes
of compounds will be determined. Soil cores
will be taken from each plot and analyzed in
a manner so as to determine the amount of
residual inorganic and organic contaminants
left in the soil and any vertical migration
that might have occurred.
It is expected that a matrix of in-
formation will come from this experiment on
the effects of surface treatments on run-
off and infiltration waters and the poten-
tial for vertical migration of waste
residuals.
PROPOSED CLOSURE EVALUATION FOR PETROLEUM
RESIDUE LAND TREATMENT
The major objective of this study, con-
ducted by the University of Oklahoma, is to
determine the extent of migration of pollut-
ants through the soil at closed oily waste
land treatment facilities. In addition, the
effect of land treatment on runoff quality,
vegetation, and soil structure will be
studied.
The contractor will develop an unsatura—
ted zone monitoring plan in order to satisfy
the objective of the study. The plan will
include the use of soil cores and-lysimeters,
and based on site and waste specific condi-
tions, will specify the number of soil cores
and lysimeters to be used. The monitoring
plan developed will represent the measure-
ment tool for determining if contaminants
in the soil will pose any future health or
environmental problems. The study will
attempt to resolve whether monitoring
results serve as reliable indicators of
potential environmental or human health
problems.
Work for this project will be con-
ducted in three phases: (1) site selection,
243
-------
(2) monitoring program, and (3) correlation
studies.
Phase 1 - Site Selection
The first phase of the project will be
to secure three closed landfarming sites
for the subsequent evaluations. Criteria
for site selection will include:
(1) Time expired since site closure
(a) recent exposure
(b) minimum of 2 years
(c) 5 years or longer
(2) Information available on site
operation history; i.e., loading rates,
characterization of applied residues, fre-
quency of application, methods of opera-
tion, and other background information.
(3) Site and soil characteristics.
(4) Existence of similar soils repre-
senting background conditions.
Once the project study sites are se-
lected and investigation arrangements are
completed, a preliminary survey of the
selected sites will be conducted to gather
and organize pertinent information.
Specific information desired includes:
(1) Time periods of site operation.
(2) Source, quantity, and character-
ization of residues applied on the site.
(3) Background information on. soil
characterization.
(4) Site operation procedures; i.e.,
frequency of residue application, manage-
ment practices, fertilization practices,
and monitoring practices.
(5) Historical site monitoring data.
(6) Use of site since closure.
(7) Adjacent land-use practices and
planned site uses.
(8) Site map.
This information will provide a framework
from which to evaluate subsequent site data
to be collected.
Phase 2 - Monitoring Program
The contractor will develop an un-
saturated zone monitoring plan for each
site. The plan will be based on specific
site criteria, waste types (if known),
historical information on the site (i.e.—
waste application method, application rate,
frequency of tilling, etc.) and preliminary
sampling by the contractor.
After the plan is established for each
site, the contractor will take soil cores
and install soil lysimeters. Soil cores
will also be taken from nearby control plots
which have not received wastes.
The constituents to be analyzed include
polynuclear aromatic (PNA) compounds and
heavy metals on the EPA list of hazardous
compounds which potentially could be found
in petroleum refinery sludges. In addi-
tion, total-cyanides and total-phenols
concentrations will be measured. The soil
core constituent analysis parameters are
listed in Table 2.
TABLE 2. SOIL CORE CONSTITUENT ANALYSIS
PARAMETERS
Antimony
Arsenic
Benzene
Beryllium
Cadmium
Chromium
Copper
T-Cyanides
Fluoranthene
Lead
Mercury
Nickel
Benzanthracenes
Benzopyrenes
. Benzofluoranthene
Crysene
Dibenzanthracenes
Indopyrenes
Selenium
Silver
Zinc
T-Phenols
The results of sample analyses from the
closed sites and their respective controls
will be compared to determine what effects
waste application has had on the quality of
the soil and soil pore water. The contractor
will statistically determine if significant
differences exist between the experimental
and control plots.
Phase 3 - Correlation Studies
Current closure regulations for land
treatment facilities recognize that the dis-
244
-------
posal of a hazardous waste is a "crucial
environmental and health problem which must
be controlled." The monitoring approach
mandated in the regulations is to provide
a quantitative measure by which to evaluate
realization of the basic objective. The
proposed study tasks in Phase 2 will yield
valuable information to evaluate the
appropriateness of the monitoring approach.
The contractor will attempt to resolve
whether acceptable monitoring results can
be correlated with adequate control of
potential environmental and health prob-
lems and conversely. Correlation studies
will be conducted to address this question.
The following correlation studies are
proposed to be completed at each of the
three sites:
(1) Statistical variation study -
Determine statistical reliability of the
monitoring plan developed for each site in
describing site hazardous constituent
concentrations.
(2) Runoff simulation - Simulate
rainfall on an isolated area of each site
to determine selected hazardous waste con-
stituent concentrations in runoff samples.
(3) Agricultural impacts - Determine
whether effects of land cultivation prac-
tices on closed sites will inhibit plant
growth and result in significant heavy
metal uptake.
(4) Soil tests - Evaluate alteration
of physicochemical properties of the test
site soils by the land treatment process.
(5) Deep core and soil pore water
analysis - Determine whether vertical mi-
gration of hazardous constituents takes
place.
The correlation studies are expected
to yield quantitative answers to the follow-
ing types of questions for closed sites:
(1) Do results from the monitoring
plan developed for each site provide a
statistically reliable measure of hazardous
constituent concentrations and differences
between background levels?
(2) Will plant growth on the sites
be inhibited and result in high levels of
uptake of hazardous constituents?
(3) Will runoff from the sites con-
tain potentially hazardous concentrations
of toxic constituents?
(A) Has the soil structure been so
significantly affected by the land treatment
process as to destroy its future usefulness?
(5) Do acceptable levels of hazardous
constituents in the surface zone of the
site insure safe levels of the same con-
stituents in the deeper soil zones and
groundwater at time of closure?
Answers to these types of questions
will provide an indication as to whether
"acceptable" monitoring results as defined
in the regulations are in fact acceptable
from the standpoint of human health and the
environment.
REFERENCES
1. Berkowitz, Joan B., et al. Field
Verification of Land Cultivation/Refuse
Farming. Disposal of Hazardous Waste,
In: Proceedings of the Sixth Annual
Research Symposium, U.S. Environmental
Agency, EPA-600/9-80=010, Cincinnati,
Ohio 45268. pp. 260-273.
2. CONCAWE 1980. Sludge Farming: A
Technique for the Disposal of Oily
Refinery Wastes. Report No. 3/80.
The Hague, Netherlands. 94 pp.
3. Huddleston, R. L. 1979. Solid-waste
Disposal: Landfarming. Chemical
Engineering, February 26, 1979.
pp. 119-124.
245
-------
LAND TREATING TANNERY SLUDGES
INITIATION OF A FIVE YEAR INVESTIGATION
Robert M
Tanners ' Council
Un iver s ity of
Cincinnati,
Lollar
ot America
Cincinnati
Ohio ^.5221
(14)
ABSTRACT
The main objective of this investigation is the characterization of
the technical and the environmental aspects of the utilization of land
treatment technology for treating tannery sludges. This five year
project will assess the environmental impact of the application of two
tannery wastewater sludges when applied to soil plots. The two sludges
are the proteinaceous/lime rich unhairing (beamhouse) sludges and the
trivalent-chroraium containing sludges from the other portions of a full
chrome-tanned cattlehide tannery.
The two sludges and a mixture thereof will be added to appropriate
test plots annually at the end of each dry season. The grass grown on
the test plots, the soil, the ground water and any leachate will be exam-
ined by analysis for various parameters but especially considering chro-
mium partition. The site will be maintaiaed under regular surveillance
during the five-year period. The data will be assessed and appropriate
site closure procedures will be developed. An overall goal is the
development of data relevant to Best Engineering Judgement documents
for land treatment facilities for tanning industry wastewater sludges.
A site has been leased in Santa Cruz County; its soil characteris-
tics, geology and hydrology have been evaluated. A site plan for the
specific location of the eight 0.2 hectare experimental plots has been
developed, and run-off control facilities designed. These eight test
plots, equipped with wells and other devices necessary to monitor the
eight test plots, are being installed and will be given their initial
sludge loading followed by grass seeding dn December, 1980.
INTRODUCTION
The overall objective of
this project is to determine
whether land treatment technology
is an environmentally sound
method for the disposal of tann-
ery wastewater sludges. Land
treatment has been practiced for
many years in many countries.
Though the U.S. EPA/Corval1is has
recently conducted greenhouse
work on this topic, there has not
been a definitive field study in
accordance with applicable reg-
ulations. This project proposes
a long-term program designed to
provide the necessary data.
Specific objectives of the five
year project are:
1. To assess potential adverse
impacts of land treatment on
various environmental sectors.
2. To measure sludge degradation
rates of the sludges added.
3. To measure the accumulation
and migration of contaminants
in soil.
4. To optimize site design and op-
eration (including closure).
5. To estimate the costs involved in
a full-scale land treatment
practice.
246
-------
Since the project has the assistance
of the following parties as subcontrac-
tors:
1. SCS Engineers (SCS): Engin-
eering design, soil and plant
monitoring.
2. University of California at
Santa Cruz (UCSC): Surface
and subsurface waters monitor-
ing.
3. Salz Leathers, Inc. (Salz):
Sludge delivery, contribution
to project funds and technical
liaison. Salz Leathers, Inc.,
is located in Santa Cruz,
California. It has been in
operation for 113 years and
produces about 5 to 6 percent
of the total yield of United
States leather. Salz has
also retained Weber and Assoc-
iates as consultants; and
Weber are especially involved
in site select-ion and charac-
terization.
EXPERIMENTAL PLAN
Wastewater Sludges
Wastewater sludge is generated at
the Salz tannery at an overall level of
approximately 50 wet metric tons (at 25
to 30 percent solids) per day, or 15,000
wet metric tons per year. These sludges
are currently disposed of at a specially
designed municipal landfill site, re-
flecting State of California require-
ments. Tannery Wastewater sludges have
been delisted under RCRA (1).
The tannery Wastewater pretreat-
ment system produces two different pre-
treatment sludges:
1. The dehairing Wastewater
sludge which results from the
hair pulping (hair-burn) pro-
cesses. It is rich in lime
and proteinaceous residues
resulting from the hair des-
truction process.
2. The tanning process wastewater
sludges generated by Salz.
This sludge is characterized
especially by its content of
trivalent'chromium resulting
from the tannage of the cattle-
hides in their conversion to
leather.
Table 1 presents a representative analysis
of these two sludges.
Although these wastewater sludges
are generated separately at the Salz
tannery, in many tanneries the waste-
water pretreatment sludges originate
from the overall wastewater flow. The
experimental plan is designed to evaluate
the disposal of the individual sludges
and their mixture in the ratio generated
in the Salz tannery.
Each of three sludges will be applied
to individual test plots at two differ-
ent sludge levels each successive year.
Two sludge loading rates are planned,
one at 336 Kg N per hectare (300 Ib
N per acre), an optimum level, and the
other at 672 Kg N per hectare (600 Ib N
per acre), an excessive level.
Since there are three sludge types
each applied at two different levels,
six test plots are required. In addi-
tion, a control plot and an additional
plot spiked with additional trivalent
chromium are included. Hence, the
test site will include eight experiment-
al plots.
The total chromium loadings during
the four years of annual application in
the fall (late in the dry season) will
be 1880 Kg per hectare (762 Kg per acre)
in the plot to which the tanning process
wastewater sludge is applied at the lower
level. The plot to which the tanning
process wastewater sludge is applied at
the higher level will receive during the
four years a total of 3760 Kg per hectare
(1524 Kg per acre). These levels are
below the limit of 4940 Kg per hextare
(2000 Kg per acre) for a soil with a
CEC greater than 15 meq per g (calculated
from cumulative irrigation water
standards for 10 year periods adopted
by the FWPCA).
Site Plan
An 8 hectare (20 acre) site which
lies within the Scott Creek valley in
western Santa Cruz County, California,
has been leased for the project by Salz
247
-------
TABLE 1. CHEMICAL ANALYSIS OF SALZ SLUDGE**
Constituent
Moisture
pH in Water
pH in Q.Q1M CaCl
Electrical Conductivity
Sodium Adsorption Ratio
Total Organic Carbon*
Total Kjeldahl Nitrogen*
Nitrate Nitrogen*
Ammonia Nitrogen*
Chromium*
Lead*
Dehairing Sludge
(Hair-burn)
73.2%
9.5
9.8
5 . 5 mmhos/cm
7.7
24.4%
3.8%
18.0 mg/kg
123 mg/kg
49 mg/kg
13 mg/kg
Tanning Sludge
(Chrome)
73.1%
8.0
8.1
7 . 1 mmhos/cm
3.8
22.8%
4.0%
3.5 mg/kg
106 mg/kg
52,500 mg/kg
209 mg/kg
* Concentration data expressed on dry weight basis.
** Data supplied by SCS Engineers.
248
-------
Leathers, Inc. The site is on a small,
almost flat marine terrace remnant
which lies between 107 m (350 ft) and
122 m (400 ft) in elevation; it is between
91m and 107 m above the floor of Scott
Creek. The slopes at the site are
generally 5 to 8 percent, increasing
in some portions to 12 percent.
The site lies on a late Pleisto-
cene marine terrace that has been
formed in hard, poorly permeable,
tertiary, mudstones. The test site is
underlain by 9 to 14 meters (30 to 45
feet) of poorly consolidated sands
and silts. Hand-auger core-hole logs
indicate moderate-to-strong argilic
"B" horizon soils present over the
entire site. The site geology and
hydrology is being evaluated in detail
for later reference.
The six primary two tenths hect-
acre (0.5 acres) plots are placed on 5
to 10 percent slopes. The control plot,
and the plot for the spiked trivalent
chromium addition are adjacent to the six
primary plots. Each of the eight plots
is surrounded by a ditch-berm system to
catch the surface runoff on the plots.
The collected surface runoff is moni-
tored and managed by clarifier-silta-
tion pond energy dissipation systems.
Lysimeters, and shallow and deep wells
are installed to permit subsurface
water monitoring.
Sludge Application and Seeding
The soil, after sampling, will be
plowed or disced. The appropriate
sludge application, augmented by potash
and phosphate fertilizer nutrients as
indicated, will then be placed on the
plots and incorporated uniformly into
the top soil. Native grass species
(salt-tolerant and having high nitro-
gen requirement) will be seeded on the
plots after sludge incorporation.
Monitoring
Monitoring will be maintained
throuthout the contract period. Sludge
analyses, soil and leachate analyses,
plant analyses and ground water
analyses will be accomplished.
Sludge sample analysis will include
moisture, pH, electrical conductivity,
and total organic carbon. The total
nitrogen and the ammonia-N and nitrate-N
distribution will be determined. Sodium
adsorption ratio, total chromium and
lead will also be measured.
The soil samples will be analyzed
similarly. Furthermore, cation exchange
capacity will be determined. The distri-
bution of the chromium between the tri-
valent and hexavalent oxidation states
will also be monitored.
The primary purposes of plant moni-
toring are to determine:
1. The growth response to tannery
sludge
2. The nitrogen removal capacity of
the sludges
The uptake and phyto-toxicity of chromium
and salts are of secondary importance.
Shortly after seeding, the percen-
tage and rate of seedling emergence from
each plot will be observed during the
emergence period. Differences in the
plant density among the plots will be
visually compared and photographically
recorded. Plant tissues will be sampled
and analyzed for total nitrogen, total
chromium and lead.
The hydrological monitoring will be
especially directed toward the possible
leachate of nitrate nitrogen and total
chromium into the ground water. The
potential for appearance of hexavalent
chromium in the leachate following
organic matter decomposition will also
be evaluated. Lead, electrical conduc-
tivity and pH of the ground water will
also be monitored.
Part XI. Environmental Protection
Agency-Hazardous Waste Management System,
Federal Register, Vol. 45 No. 212, Oct.
30, 1980.
249
-------
INORGANIC HAZARDOUS WASTE TREATMENT. II
Warren J. Lyman
Arthur D. Little, Inc.
Cambridge, Mass. 02140
Gayaneh Contos
Versar, Inc.
Springfield, Va. 22151
ABSTRACT
This report describes an ongoing program to investigate selected processes for the
treatment of inorganic hazardous wastes. A primary focus of the program is on wastes con-
taining heavy metals; a secondary focus is on wastes containing organic as well as inor-
ganic components. The report contains preliminary results from laboratory and small-scale
pilot treatability studies on: (1) mixed acid wastes containing high levels of heavy
metals plus some organic matter; (2) landfill leachate containing some industrial pollu-
tants (heavy metals and organics); and (3) a heavy-metal containing sludge which is
treated by high-gradient magnetic separation.
INTRODUCTION
Arthur D. Little, Inc., along with
Versar, Inc. as a subcontractor, is cur-
rently in the final phase (Phase III) of an
inorganic hazardous waste treatability
study that was initiated in 1978 by the
Solid and Hazardous Waste Division of MERL.
A particular focus was on hazardous wastes
disposed of in the municipal sector (i.e.,
off-site). Such wastes are often complex
mixtures of wastes from various processes
(and sometimes different plants) and are not
usually amenable to material recovery. In
Phase I a survey of potential treatment
technologies was conducted. A total of 21
unit processes were eventually identified
as being applicable to the treatment of in-
organic hazardous wastes. In Phase II a
screening process was used to narrow down
the candidate list of treatment processes
with the intent to eventually select four
for in-depth study. Simultaneously we en-
gaged in a search for available waste
streams that could be used in our labora-
tory and pilot plant studies. The results
of our Phase I and Phase II studies are
described in more detail in a previous
report (Lyman and Contos (3)).
The treatment processes and wastes se-
lected for demonstration tests in Phase III
were: (1) the treatment of mixed acid
wastes, primarily from electroplating and
metal finishing industries, by a dual-pre-
cipitation (Ca(OH2) plus Na2S), filtration,
and adsorption (powdered carbon) process;
(2) the treatment of landfill leachates,
containing some industrial contamination,
by a dual-precipitation (Ca(OH)2 plus Na2S),
filtration, and adsorption (powdered -carbon)
process; and (3) the treatment of a heavy-
metal-contaminated sludge (primarily com-
prised of filter aid) by high gradient mag-
netic separation.
TREATMENT OF MIXED ACID WASTES
Waste Characteristics and Sources
The mixed acid wastes selected for
study were obtained from a licensed hauler
of hazardous wastes in N^w England. This
hauler collects wastes from many of the
small industries in this area and trans-
ports the wastes to treatment and disposal
sites in other states. Aqueous wastes col-
lected from electroplating and metal
250
-------
finishing plants are segregated into the
following eight categories: (1) ammonium
persulfate; (2) sodium persulfate; (3)
electroless copper; (4) aqueous alkali
(without NH3); (5) aqueous alkali (with
NH3); (6) acid bath (without NH3) ; (7) acid
bath (with NH3) ; and (8) cyanide bath.
Wastes in categories 6 and 7 were chosen
for study in this program, in part because
they form a significant fraction of the
wastes collected from these industries.
When a sufficient quantity of wastes in
one category are on hand - typically about
19,000 liters (5,000 gal) - the hauler com-
bines them (from individual drums) into
one tank for transport to a treatment and
disposal site. Two such batches - desig-
nated samples A and C - were sampled in
Phase II and subjected to laboratory analy-
sis and laboratory treatability studies.
Four additional batches - designated samples
F, G, H and I - were sampled (in 55-gallon
drum quantities) for the Phase III labora-
tory and pilot-scale treatability studies.
The principal, non-aqueous component
of these two wastes is a mixture of hydro-
chloric, sulfuric, nitric, and other acids.
The acid wastes with and without NtU differ
little except in their NH3 content (about
10,000 mg/L (as N) vs 100 rag/L, respective-
ly) . The principal characteristics of the
acids wastes are as follows:
• pH ^ 0.5 to 1
• Heavy metal content 'v 1,000 to 9,000
mg/L
• Principal heavy metals (in order) :
Ni, Cu, Cr, M, Fe, Mg, Sn, Pb,
Zn, Mn, Co
• Dissolved solids ^ 50,000 to 140,000
mg/L
• Suspended solids % 30 to 10,000 mg/L
• NH3 (as N) ^ 120 to 17,000 mg/L
• Total organic carbon ^ 3,900 to
5,300 mg/L
• Chemical oxygen demand ^ 10,000 to
17,000 mg/L
• Principal anions : CiL~ , SO^", F~,
• Toxic organics and chelating agents:
present
• Color: green
Treatment Objectives
The primary treatment objective for
these wastes is the removal of all hazar-
dous components to the extent that the
treated aqueous effluent could be safely
discharged to a publicly-owned treatment
works (POTW). Specific treatment objec-
tives for each pollutant are difficult to
define since federal regulations for hazar-
dous waste treatment operations have not
been set, and dilution ratios (at the point
of discharge to the sewer) and local sewer
ordinances (designed to protect the speci-
fic treatment operations at the POTW) may
vary significantly from site to site. In
spite of the above, we will be comparing
the quality of the treated effluent with
specific guidelines, including:
• Recommendations of the National Com-
mission on Water Quality (NCWQ (4))
for pretreatment standards for elec-
troplaters discharging to POTWs;
• U.S. EPA effluent limitations for
existing sources in the electropla-
ting point source category discharg-
ing to the POTWs (EPA (1));
• U.S. EPA options for effluent limita-
tions for the metal finishing point
source category (EPA (2)); and
• Other studies indicating what levels
of toxic chemicals should not be
discharged to POTWs.
These guidelines tend to limit the effluent
concentrations of various heavy metals to
about 1 mg/L each, with most limits being
in the range of 0.6 to 4 mg/L. A total
heavy metal limitation is specified in one
case (EPA (1)); this total ranges from about
7 to 20 mg/L depending on the volume of
waste generated, the type of sample col-
lected, and the type of treatment used.
Because of the complex and variable nature
of the acid wastes chosen for this study —
and because of their relatively small volute
compared to total POTW influent flows — we
do not expect that the cited guidelines
should or could be used as an absolute scale
for determining treatment effectiveness.
Secondary treatment objectives for this
study included: (1) the development of a
process that was not only simple and econo-
mical to operate, but also highly effective
and capable of handling a highly variable
input waste; and (2) an investigation of
the potential for separating and recovering
selected heavy metals.
251
-------
The Treatment Process
The treatment process used for the
mixed acid wastes was designed to neutra-
lize the acids, and to effect a high re-
moval of heavy metals, toxic organics and
some anions. The key unit operations in
this design are: (1) heavy metal precipi-
tation via addition of Ca(OH)2 and NagS;
(2) adsorption of organics with powdered
activated carbon; and (3) filtration for
the removal of both the metal precipitates
and powdered carbon. In addition, the pro-
cess was designed to (hopefully) overcome
a number of potential problems, including:
• The fact that the minimum solubili-
ties for the heavy metals present
are at different pH values, which
values are difficult to predict
because of the complexity and vari-
ability of the solution. Thus, the
selection of just one "optimum" pH
for removal of precipitated metals -
an optimum that would be used for
all waste batches - would be unlike-
ly to effect a consistently high
removal for all heavy metals.
• The fact that the high ionic
strength of the waste (>_1) was like-
ly to result in soluble ion pairs,
thus further complicating the pre-
cipitation of heavy metals.
• The fact that metal complexing
agents (NH3, organic acids, chelates
(e.g., EDTA)) were likely to be pre-
sent, also interfering with metal
precipitation.
• The possibility that too rapid a
rise in the pH (during the neutrali-
zation process) could lead, tempo-
rarily, to a highly supersaturated
solution which, in turn, could lead
to the formation of very fine pre-
cipitates that would be difficult
to filter out.
• The fact that some organics present
were acidic in nature and were not
likely to strongly sorb to carbon at
a pH where a significant fraction
was in the dissociated form.
The basic step
cess are:
in the treatment pro-
(1) The slow addition of (powdered) Ca(OH)2
to the mixed waste, and periodic filter-
ing, as the pH is raised from its ini-
tial value of 0.5-1 to about pH 8. (In
laboratory tests, the reaction mixture
was filtered after each unit increase
in pH, in subsequent pilot plant tests,
the filtration was typically once every
two pH units.) The waste was, at times,
filtered before any Ca(OH)2 was added.
(2) The addition of powdered activated car-
bon at two different points in the fil-
tration process within the pH range of
6 to 10. (A typical carbon dosage was
3 g/L (total) for a waste.)
(3) The slow addition of Na2S within the
pH range of 8 to 10, again with peri-
odic filtering. (The amount of Na2S
used generally resulted in a further
pH rise of 1/2 to 3 pH units.)
(4) The slow addition of Ca(OH)2, again
with periodic filtering, until a pH of
about 12 is attained.
(5) The back neutralization of the waste
with HCS, from pH 12 to a value in the
range or 7 to 10.
Steps 1, 3 and 4 are designed to maximize
the removal of heavy metals. The slow addi-
tion of reagents should prevent excessive
supersaturation. The periodic filtering
should take advantage of the minimum solu-
bility of each heavy metal, wherever it
lies within the pH range of 1 to 12. And
the addition of the sulfide reagent (Na2S)
is designed to effect a higher degreee of
heavy metal removal than is possible by hy-
droxide precipitation alone. In addition,
this reagent should reduce any hexavalent
chromium present and allow it to be preci-
pitated. Finally, in step 4, the raising
of the pH to nearly 12 was expected to help
break certain heavy metal complexes which
would be stable (and soluble) at lower pH
values .
The use of powdered activated carbon
(step 2) was designed to avoid the common
problems and associated expenses of fixed-
bed granular carbon systems. These prob-
lems include: (1) the necessity for more
complex systems (additional tanks, filter
beds, pumps, etc.); (2) filter bed clogging
and/or the necessity for frequent backwash;
(3) higher costs of granular carbon and/or
252
-------
the necessity to regenerate the carbon. The
powdered carbon in our process is not in-
tended to be recovered for regeneration.
Laboratory Treatment Tests
One of the more important laboratory
tests conducted was a trial run of the step-
wise precipitation/filtration process de-
scribed in the previous subsection. Five
acid waste batches were tested in this fash-
ion; for sample C a25-mLtest volume was used
(obtained from a 1-liter sample provided by
the waste hauler); for samples F, G, H and
I a 1.5-L test volume was used (obtained
by sampling a 55-gal. drum of the waste
provided by the hauler).
Figure 1 shows the amount (dry weight)
of precipitate removed by filtration during
this treatment process. Note that the only
common feature of the individual figures is
the high solids removal between pH 1 and 2.
These solids are made up primarily of cal-
cium and sodium salts; crystals of
10H20 were identified in one sample from a
measurement of the crystal's refractive in-
dex. The total amount of solids removed
during the treatment process ranged from
60-80 g/L for the three acid wastes "with-
out" ammonia (C, G, I), and from 130-160 g/L
for the two acid wastes "with" ammonia (F,
H).
Figure 2_ provides neutralization curves
for these same five wastes. The amount
of base (and acid) used is given in units of
gram-equivalents used per liter of waste.
(One gram equivalent of a base will release
one mole of OH~ ions.) These curves point
out additional differences between the acid
wastes "with" (F, H) and "without" (C, G, I)
ammonia. The latter only require about one
gram equivalent of base for the treatment
process while the former require as much as
three to five gram equivalents. In addi-
tion, the two wastes "with" ammonia both
show a sharply rising curve in the region of
Sjmpi.H
Initial pH • 0 5
Total jolids in raw sample = 119g/L
Total amount of precipitate
removed - 133 g/L
2 3 4 5 6 7 8 9 10 11 12
pH
25
20
10
5
0
Ini
To
Sample 1
tat pH - 0 89
al solids in raw sample * 86 5 g/ L
Total amount of precipitate
— 1 1 ,
m
'/////
/////
m
. —
1
2 345
10 11 12
j j
PH
Precipitation via addition of CafOHJj
Precipuation via addition of Na«S
•
—
InitialpH-OS
Total solids in raw sampla - 78.7 g/L
Total amount of pracipitata
(orimd-61.1 g/L
i
] i i i i j.i • • i . iii i i .1
50
3 30
S 0
•5.01234587
10 11 12
Sampla F
Initial pH - 0.95
Total solids in raw sampla - 87 4 g/L
Total amount of precipiuta
formad- 164.5 g/L
in
w,
u_r-
10 11 12
Santpto Q
Initial pH -085
Total solids in raw sampla • 65.1 g/L
Total amount of prtcipitata
formad - 80.7 g/L
01234587
PH
10 11 12
FIGURE 1 PRECIPITATE REMOVAL VS. pH FOR ACID PLATING WASTES
253
-------
Amount of Acid Added
(gram-equivalents/ liter I
0
Amount of Acid Added
(gram eguivaients/literl
0 0 1
Ammjm of B«M Addwt
-^ PnacMtaM rwnowd by ftltntion
Amount at B«M Added lflranv«qui*il«ntt/linrl
Amount of Actd Added
(grem-equivalents/liter)
0 01
02 04 06 08 10
Amount of Base Added (oram-equivatants/literl
FIGURE 2 ACID AND BASE REQUIREMENTS FOR TREATMENT OF ACID WASTES
pH 2 to 8, a factor which makes pH control
in this region difficult. These wastes
also showed - when the results of the
laboratory and pilot plant tests are com-
pared - some anomolous behavior at low pH.
The initial pH of the pilot plant waste
was about 2 pH units higher than that of
the laboratory sample taken from the same
drum a few weeks earlier; and the pH of the
pilot plant samples decreased with the addi-
tion of base until about one gram equiva-
lent of base had been added. Sample hand-
ling procedures, leading to higher losses
of NHa in the laboratory samples, may be
involved. Overall though, the neutraliza-
tion curves from laboratory and pilot plant
runs agree relatively well; this implies
that laboratory neutralization curves can
be used as a fairly accurate predictor for
the (batch) treatment of much larger volumes
of waste.
Figures J3 and 4^ show the removal of the
most prevalent heavy metals as a function of
pH in two laboratory tests. Both show alum-
inum and iron being removed at relatively
low pH (1-5) and most of the chromium coming
out at intermediate values (3-7). The major
difference lies in the behavior of copper
and nickel. In waste F (Figure 3), which
contains high levels of ammonia, the two
metals act similarly; small amounts are
254
-------
01 2 3 4 5 a 7 8 9 10 11 12
1 2 3 4 5 6 7 8 9 10 It 12
FI6UHE 3 REMOVAL OF METALS VS pH FOR ACID WASTE f (LABORATORY TEST)
removed in the 1-7 pH range and a peak is
seen in the pH 8-9 increment. In waste G
(Figure 4), which contains relatively
little ammonia, the copper comes out
earlier than the nickel. For this waste
there would appear to be some potential for
the separation (and recovery) of nickel;
other metals are concentrated in the preci-
pitates formed below pH 6 while the nickel
is concentrated in the precipitates formed
above pH 6.
Table _1 shows the results of laboratory
carbon adsorption tests. The data show
that even high doses of the carbon, up to
8.5 g/L, could remove only about 30% of the
total organic carbon (TOG) and that the pH
at which the carbon was added was not very
important (at least for TOG removal).
Fairly high removals of oil and grease were
seen under all conditions.
The treatment effectiveness (for heavy
metals removal) seen in these laboratory
tests is discussed later on in conjunction
with the pilot plant treatment tests.
FIGURE 4 REMOVAL OFMETALSVS pH FOR ACID WASTE G (LABORATORYTESTI
Pilot Plant Treatment Tests
Small-scale pilot plant treatment tests
were carried out with four acid wastes
(samples F, G, H and I) using the treatment
TABLE I. LABORATORY TESTS WITH POWDERED ACTIVATED CARBON
Sample
F
G
p
Adjusted
pH
I
5.0
I
f
5.0
I.
5.0
followed by
8.0
Carbon
Dose
(g/L)
0
1.7
8.5
0
1.7
8.5
5.1 1
\
3.4
Analysis of
TOC Oil
(mg/L)
3700
4100
2960
38CO
3500
2530
2700
Filtrate
and Grease
(mg/L)
93
< 1.
1.
62
2.
2.
2
5.0
followed by
8.0
255
-------
process described previously. A schematic
diagram of the treatment system is shown in
Figure 5. A 266-liter (70 gal.) plastic
barrel was used as the reaction tank. In
each, test about 200 liters of waste (the
complete contents of the 55-gal. drum ob-
tained from the waste hauler) was used.
The filters used were non-woven fabric fil-
ter bags (nominal pore diameters of 5 to 10
microns (20 microns for sample F)) held in
vertical cylinders. The original plan
called for recirculation of the waste
through these filters and back to the re-
action tank. However, the large volume of
solids (precipitates, carbon and filter
aid) frequently built up a substantial back
pressure in the filters (15-30 psi) result-
ing in recirculation flow rates (%2-15
liters/min.) that were too low to be effec-
tive. Thus, when filtration was desired -
about every 2 pH units - the complete con-
tents of the reaction tank was pumped
through the filters to a collection barrel;
the filtrate was subsequently pumped back
to the reaction tank for further treatment.
Filter aid (diatomaceous earth ) was used
for all filtration steps except those
during the treatment of sample F. Some
problems with solids retention had been
noticed in this initial run; the use of fil-
ter aid and more closely woven filter bags
nearly eliminated the problem in subsequent
runs.
The specific treatment regime given
each of the four acid wastes is shown
schematically in Figure 6_. Samples of the
raw and treated waste (for pollutant analy-
ses) were collected according to a standard
protocol just prior to, and just after, the
pilot treatment run. In some cases, inter-
mediate samples (for metals analyses) were
collected after each filtration step.
The chemical usage associated with these
pilot treatment runs is shown in Table 2.;
the values are in units of grams of chemical
used (mL for HCJl) per liter of waste. These
numbers are important since chemical costs
are expected to be a significant fraction
of the total treatment costs.
The amount of sludge generated during
this treatment process is significant.
Prntura Gauge
Arrows show direction of flow during
treatment while recycle loop is operitlng.
Only one filter will be used at any given
time.
Reversible
Pump
(max. 20 GPM)
Break-away
Joints for
Filter Change
FIGURE 5 SCHEMATIC DIAGRAM OF PILOT TREATMENT SYSTEM
256
-------
TABLE 2. CHEMICAL USAGE ASSOCIATED WITH PILOT PLANT TREATMENT RUNS
Acid
Waste
F (with NH3 )
H (with NH3)
G (no NH3 ;
I (no NH3)
Ca(OH)2
(g/L)
124
97.8
36.4
31.9
NaaS
(g/L)
69.4
16.1
3.6
3.6
HC£ (-37%)
(mL/L)
(42)a
46.2
4.3
8.4
Powdered
Carbon
(g/L)
2.6
4.0
3.1
3.5
Filter
Aid
(g/D
Not Used
11
10
10
a. Calculated from laboratory data.
TABLE 3. ESTIMATES OF SLUDGE VOLUME GENERATED DURING TREATMENT
0 II) 1 I 17 tO fl G
s K, r
n 1
^ampl" G
"L__
Soni,.). 1
* '
1
1
— r -
—
"~
f
1
- "
1,
„ 1&
r
li
/
1
r
i
'i, i.
r
1
..
t T
1
<| |
1 1
^|_
— T-T--I — T-T—
ansly^t
|l
| f
< 1 1 1 1
ri i 7 i
10 II 17 10 H 6
r - 1 !,,„,.,„„ i - 1 N,,,,,.I,,» ......
l— - ! wlltiri(OM).. t - 1 wll> Mr I
FHtUllF 9 SIIMMAIIY SrllPMAIIC Or rlLUI PI AN I IlirA IMP NT Of Ml XED ACID PI ATINO WASIFS
Table 2 provides estimates (the exact
amounts were not measured) of both the
weight and volume of sludge generated. The
acid wastes with ammonia (F, H) provide
the worst case. About 150-170 g (dry wt.)
of sludge is produced for each liter of
waste treated. If this sludge was only
dewatered to 25 wt.% solids then 510-580cm3
of sludge are generated per liter of waste,
i.e., the volume of hazardous wastes has
only been reduced by less than one half!
Dewatering to 50 wt.% solids produces 220-
240 cm3 of sludge per liter of waste. The
acid wastes without ammonia generate about
one-half the sludge generated by the wastes
with ammonia.
Acid
Waste la
C
F
G
H
I
a Values
these
Amount of
precipitate
b. teats >g/L)a
61.1
164
30.7
133
62.0
are grams (dry
materials assum
In pilot tests added*
Powdered
carbon
(g/Ua
c
2.6
3.1
4 0
3.5
Filter
aid
(g/Ua
c
d
10
11
10
weight) per liter of waste
ed to be (1)
precipitates
Sludge volume (cmJ/L)
if sludge contains. ^
50 wt. %
water
89
240
140
220
110
Densities
, 2 2 g/cm].
75 wt. %
water
210
580
330
510
260
> of
(2) powdered carbon, rfec densicy of 1 6 g/cm , (3) filcer aid, wee
density of 1.5 g/cra3.
i are cm3 of (wee) sludge generated per licer of raw waste
No filter aid used with sample F
Preliminary Results
Pre- and post-treatment pilot plant
samples were analyzed for: (1) over 20
metals; (2) specific organics (GC or GC-MS);
(3) anions; and (4) conventional parameters
At present we can only report some of the
metal data; other items will be discussed
qualitatively.
The degree of metals removal achieved
in the pilot runs is shown by the data in
Table .4. In general, very high removal
efficiencies were obtained in spite of the
potential problems mentioned previously.
A comparison of the Final concentrations
with various EPA and NCWQ limitations for
discharges to POTWs indicates relatively
few problem areas, although it is apparent
that these limitations could not be strict-
ly met. Maximum Final concentrations -
some of which exceed EPA limitations for
257
-------
TABLE 4. CONCENTRATION OF HEAVY METALS (mg/L) TN INITIAL AND FINAL SAMPLES FROM PILOT
PLANT TREATMENT OF ACID HASTES
• "— — -
Metal
A*
Asa
Ba
Be
Cd
Co
Cr
Cu
Fe
Hg3
Mg
Mn
Mo
Ni
p
Pb
Sea
Sn
Sr
Ti
Zn
Zr
Waste
Initial
166
260a
1.7
0.011
< 1
14.3
3.39
2000.
72
210. a
48.4
2.2
< 1.
6100.
313.
60.
ioa
190
0.39
3.39
3.90
2.2
F
Final
4.0
6a
0.5
< 0.005
< 1
< 0.5
0.08
1.73
2
2.4a
< 0.1
< 0.1
< 1.
8.7
< 1.
< 0.5
10a
2.8
0.68
< 0.05
0.35
< 0.5
Waste
Initial
88
780a
1.2
0.28
3.5
0.6
93.5
19.0
42
510a
36.3
1.9
< 3.
1960
1760
132
lla
89
0.12
< 0.1
26.5
0.5
H
Final
11
12a
2.9
< 0.01
< 0.7
< 0.5
< 0.08
0.09
< 2
0.4a
1.3
< 0.1
< 3.
1.3
< 5
0.15
3.6a
0.5
0.53
< 0.1
< 0.2
< 0.5
Waste
Initial
128
6.
0.
0.
0.
1.
689
1970
165
0.
100
1.
< 0.
3180.
2350
8.
4900.
130.
0.
0.
5.
< 0.
a
8
172
7
4
2a
75
6
00
a
52
28
89
1
G
Waste
Final
< 0.
1.
0.
< 0.
0.
< 0.
4.
2 .
2.
0.
0.
0.
< 0.
0.
364.
< 0.
< 2.
1.
0.
< 0.
0.
< 0.
7
a
8
003
2
1
96
04
4
4a
8
23
6
61
20
a
52
03
25
1
Initial
6.
610.
0.
0.
0.
O
40.
1050
262
0.
4.
1.
< 0.
5620.
1170.
18.
48a
12.
0.
0.
13.
0.
9
a
4
010
9
6
9
4a
3
85
6
5
03
37
7
3
I
Final
< 0.7
14. a
1.1
< 0.003
0.3
< 0.1
1.28
2.07
3.0
O.la
1.5
0.22
< 0.6
3.39
871.
< 0.2
19a
0.2
0.47
< 0.03
< 0.05
< 0.1
a. Values for As, Hg and Se are in ug/L.
electroplating/metal finishing shops - for
a few metals were: A5,, 11 mg/L; Cr, 4.96
mg/L; Cu, 2.07 mg/L; Fe, 3 mg/L; Ni, 8.7
mg/L; Pb, 90%), pood
removal of S0tt= O90%) , and some removal
(^30%) of N03~. Levels of CU'are increased
due to the addition of HC£. Preliminary
258
-------
TABLE 5. COMPARISON OF METALS REMOVAL IN LABORATORY AND PILOT PLANT RUNS
Total Initial Metals;3
Total Final Metals:
% Removal of Metals
Waste F
Lab Pilot
7567 8478
<21.2 <2K3
>99.7 >99.7
Waste H
Lab Pilot
2407
<26.4
>98.9
Waste
Lab
7618
<9.9
>99.9
G
Pilot
6881
<13.7
>99.8
Haste I
Lab Pilot
7022
<14.0
>99.8
a. Values in mg/L. Total is for Al, Cd, Co, Cr, Cu, Fe, Mg, Mn, Mo, Ni, Pb, Sr, Ti, Zn,
Zr. Values for As', Ba, Hg, Se, and Sn excluded since no analyses were done on samples
from laboratory tests. P and Be also excluded.
data from our organics analyses (primarily
for volatile compounds) have shown high
removal efficiencies for the hydrophobic
compounds (e.g., the volatile priority pol-
lutants) and moderate to low removal effi-
ciencies for the hydrophylic compounds
(e.g., alcohols, esters, ketones, acids).
The aqueous effluent from this treat-
ment process is essentially a concentrated
salt solution comprised principally of Na+,
Ca , K+, Ci~, N03~, and S0k= ions, and some
soluble organic solvents (alcohols, ketones,
esters, etc.) The range of values found
(for samples F, G, H, I) for dissolved
solids (DS), total organic carbon (TOC),
chemical oxygen demand (COD), and 5-day
biochemical oxygen demand (BOD) were as
follows:
Parameter
DS
TOC
COD
BOD
Effluent
Cone.
(mg/L)
60,000-120,000
1,400-3,300
4,000-12,000
500-4,600
Average %
Reduction
30
40
10
Increase
The four samples showed the following
ranges for the principal metal cations:
Na, 11,000-30,000 mg/L; Ca, 2,800-4,000
mg/L; and K, 4-310 mg/L. The principal in-
organic anion is C£~. The increase in the
BOD noted above is presumably due to the
removal (in the treatment process) of toxic
pollutants that inhibited biochemical acti-
vity in the raw waste.'
TREATMENT OF LEACHATES
Waste Characteristics and Sources
Four different leachates were investi-
gated in this program. The four sites are
briefly described below. Diagrams are pro-
vided in Figure _7 •
Site A_ - A municipal landfill for an old
city in New England currently accepting
150-200 tons of refuse per day. Industrial
hazardous wastes may have been placed in
this site in the past; current rules have
excluded such wastes for several years. A
sample of leachate (diluted with ground-
water) was obtained from a sump in a nearby
municipal building.
Site 15 - A municipal landfill in New England
currently accepting about 150 tons of refuse
per week. No hazardous wastes have ever
knowingly been disposed of at this site.
The site is about five years old and has
built up a refuse mound about ten feet deep.
Cones in the base of the landfill allowed
pure leachate (undiluted by groundwater) to
be collected.
Site C. - A municipal landfill for a New
England city with a population of nearly
40,000. Landfilling at this site started
about 1960; about 32,000 tons/yr. of refuse
are currently disposed here. Hazardous
industrial wastes are not allowed. Samples
were collected from the surface of a sump
(serving as a lift station) in a leachate
removal system. The samples are diluted by
groundwater and may be more representative
of "old" leachate (overlying material > 5
years old) than "new" leachate.
259
-------
Area in Use for Waste
Disposal at Time of
Sampling (July, 19801
Perforated Pipe
Buried in Trench
Below Natural
Ground Water
Level
SITEC
Compacted Refuse
FIGURE 7 SCHEMATIC OF LEACHATE COLLECTION SITES
Site D^ - An unauthorized chemical dump site
in New England. The site was an old sand
and gravel pit into which construction and
demolition debris, and chemicals (some in
drums), were placed. Over 1300 drums
(almost all 55-gal.) have been removed from
the surface of the site. Samples were col-
lected from a monitoring well about 50 m-
west of the dump site.
About 230 liters of leachate was col-
lected form both sites A and C; and 460
liters from both sites B and D. The wastes
were collected and held (several weeks) in
polyethylene-lined 55-gal. drums. The gen-
eral characteristics of these leachates are
described by the data in Table 6_. The lea-
chates from Sites B and D had very strong
odors.
Treatment Objectives
The primary treatment objective for
these leachates is the removal of all haz-
ardous components to the extent that the
treated aqueous effluent could be safely dis-
charged to a POTW or subjected to biologi-
cal treatment in other ways. In particular
we wanted to effect a high degree of heavy
metals and toxic organics removal. Since
our primary interest was in leachates con-
taminated with industrial hazardous wastes,
portions of two of the leachates that were
from relatively clean municipal landfills
(Sites B and C) were spiked with known quan-
tities of seven heavy metals (Co, Cu, Ni,
Zn, Pb, As and Cr) and seven organics (ben-
zene, chlorobenzene, phenol, p-chlorophenol,
toluene, naphthalene, and trichloroethylene)
TABLE 6. GENERAL CHARACTERISTICS OF THE LEACHATES USED3
pH (units)
Color
Suspended Solids
Dissolved Solids
Ammonia Nitrogen as N
Total Organic Carbon
Chemical Oxygen Demand
Biochemical Oxygen Demand
Oil and Grease
Conductivity (micromho/cm)
Site A
7.2
rusty
70
465
2.4
L8
80
25
-------
in order to simulate a more polluted lea-
chate. Unspiked portions were also treated.
The Treatment Process
The treatment process used for the
leachates - like that used for the acid
wastes - was designed to effect a high re-
moval of heavy metals, toxic organics and
some anions. The key unit operations were,
as Before: (1) heavy metal precipitation
via addition of Ca(OH)2 and Na2S; (2) ad-
sorption of organics with powdered activated
carbon; and (3) filtration for the removal
of both metal precipitates and carbon.
No biological treatment (for the re-
moval of BOD and COD) was employed. If
the treated leachate from the chosen pro-
cess could be easily discharged to a POTW
(by sewer connection or truck transport),
then this would be one way to economically
add biological treatment. Alternatively,
leachate recycle (via spray irrigation)
through the landfill could be used to
effect biological treatment at many sites.
It was not practical in this program to
simulate either of these alternatives in
our treatment regime.
The chosen treatment process was chosen,
in part, to allow relatively simple and
economical, on-site (pre) treatment of lea-
chate. It would hopefully overcome any
metal-complexation problems and be gen-
erally applicable to a variety of leachates.
The basic steps in the treatment pro-
cess are:
(1)
The addition of HC5, to acidify the
leachate to pH 3-5 and the subsequent
addition of powdered activated carbon
at this pH. The carbon is removed
by filtration. (The initial pH low-
ering is intended to reduce the degree
of dissociation of organic acids and
thus increase the amounts that would
be removed by the carbon.) This step
was not used with the Site D leachate.
(2)
The addition of
the mixed waste
about 8; one or
used over this
dose of carbon
moved with the
step. The Site
the sequential
of three doses
(powdered) Ca(OH)2 to
raising the pH to
two filtrations are
pH range. An additional
is added and then re-
first filtration in this
D leachate started with
addition and filtration
of the powdered carbon.
(3) The addition of a small amount of Na2S
near pH 8, raising the pH from 0.5-1
units (i.e., to 8.5-9).
(A) The slow addition of additional Ca(OH)2
with one or two additional filtration
steps, until pH 11-12 is reached (pH 10
for the Site D leachate).
(5) Back neutralization with HC£ to about
ph 7.
The rationale for steps 2-5 is similar
to that previously described for the acid
wastes.
Laboratory Treatment Tests
Trial runs of the step-wise precipita-
tion/filtration process were carried out
in the laboratory for each of the four lea-
chates. Sample volumes were 1.5 liters in
each case. Figure 8 shows the amount (dry
weight) of precipitate removed at each fil-
tration step. Note that none of the dia-
grams show any similarity to each other.
The neutralization curves for these lea-
chates (not shown) also indicate a signi-
ficant difference in their behavior during
the treatment process.
Laboratory treatment tests with the
powdered activated carbon, using TOC and
oil and grease analyses, were inconclusive.
As with the acid wastes, good removal of
oil and grease but poor removal of TOC was
apparent irrespective of the pH. The TOC
in the Site D leachate was, for example,
reduced from about 2300 mg/L to only 1600
mg/L after the addition of 10 g/L of carboa
Essentially the same results were obtained
at pH 6.2 and 4.0
Pilot Plant Treatment Tests
Small-scale pilot treatment tests were
carried out with the four leachates using
the process described above. The treatment
system was the same as was previously de-
scribed for the acid wastes (Figure 5).
About 200 liters of waste were used in each
run. Filter aid (diatomaceous earth) was
used in all filtration tests. The filter
bags were eigher 5 or 10 micron nominal
pore diameter.
Two leachates (Sites B and C) were
treated in duplicate runs, one with and one
without the heavy metal and organics spikes
mentioned previously. The specific treat-
261
-------
S>» A
Leachate
5 6 7
•Hal pH of leacHaw « 6 40
, Suspended solids removed
from caw i«achaw - 0 062 g/L
Total amount of precipitate
'ormad " 0 38 g/L
9 10 IT
mttal oH of 'eacnate * 5 25
5uso«naea solids r
amount of arecipiiate
3 52g/L
05
04
03
02
01
Laacnati
.
..
—
1 i ,-r^-
Initial pH of laacnau
• 8 7 Aodlfitd tp DH 4 0
|
1 Sujpanotd lolidi removad
|
from raw t«acna»
•0 124 g/L
Total amount of ortciDitat
removtd • 1 08 g/ L
0.3
0.2
01
0
S.ta 0
Leacnat
1 1
e
'///*
|
|
•nitiai oH ot eacnate
•SQ7
Total amount or
areei pnate removeo
- I 25 g/L
a audition of CalOHU
FIGURES PRECIPITATE REMOVAL VS pri FOR LEACHATE SAMPLES AFTER ACIDI FICATION TO pH • 5
ment regime for each of these wastes is
shown schematically in Figure 9_. Samples
of the raw and treated wastes (for pollu-
tant analyses) were collected according to
a standard protocol just prior to, and just
after, the pilot treatment runs. The chem-
ical usage with these runs (important in
the estimation of treatment costs) is shown
in Table _7. The estimates of the amounts
of sludge generated during the treatment
process are shown in Table J3. The dry
weight of the metal precipitates formed is
less than the weight of either the filter
aid or carbon added, and the total solids
generated is in the range of 8-20 (dry
wt.)/liter. Dewatering to 50 wt. % solids
would produce between 13 and 31 cm3 of
sludge per liter of waste.
Preliminary Results
Pre-and post-treatment pilot samples
were analyzed for: (1) over 20 metals;
(2) specific organics (GC or GC-MS); (3)
anions; and (4) conventional parameters.
At present we can only report some of the
metal data.
The data in Table 9_ show that heavy
metal concentrations in the four leachates
were generally quite low except for Fe,
Mg, and Mn. It would be difficult to de-
termine from these data if industrial
heavy metal wastes had ever been disposed
of at these sites; only the Ni content at
Site D appears to be suggestive of indus-
trial contamination. The treatment pro-
cess gave good removals for all of these
metals. This is seen especially in the
spiked pilot plant runs (Sites B and C)
where known quantities of Co, Cr, Cu, Ni,
Pb, Zn and As were added. The removal
efficiencies for most of these metals in
the spiked runs was over 98%. However,
the removal efficiency for copper was only
about 90% and final values of about 1 mg/L
resulted. Complexation with naturally oc-
curing organics may have been involved.
Sludge Treatment -Via High Gradient Magnetic
Separation
Currently, the typical treatment of in-
dustrial wastewaters that contain toxic
metals includes the chemical precipitation
262
-------
Trtatment Procws
TABLE 3. ESTIMATES OF SLUDGE VOLUME GENERATED DURING TREATMENT
PH
86 42 4 6 8 10 t
C
»k .•-.-. •••-.
i
c
SH. B j.
C
Sill B- j
C
ShiC 1
C
SiMC* i
SinD
• t i
•u
tsmmmte®
i i
c
*l 1 1
m$mmv®m&
f I t
c
ti 1 i
«N^4j*N
•f i 11
c
*l 1 1
mmmim-M
\ \ \
c
U n
t \ • f
i n
c
\, ,
•teto;
i i
10 8 e
IV---:--.----.-JT
.••.••.•.•.••:.JT
.-.-.•: •••.•.••:.|T
I- ••:••.••.•:. -|T
I-':". ::;.. -|T
86 4246 8 10 12 10 8 6
pH
R- Riwwaiu 1 . p,itr«tion f:-S:'S:-| Treatment
T- Treated*.,,. | «-*->»* C.IOHI,
£
« Addition of powdered ^_^_^ t^^ir
activated carbon | .'.- | Treatment V///A TrMtr™nt
•- Sample ,p,ked with "'* Hcl ^^ wlth Nl2S
metals and organic*
FIGURE 9 SUMMARY SCHEMATIC OF PILOT TREATMENT OF LEACHATES
TABLE 7. CHEMICAL USAGE ASSOCIATED WITH PILOT PLANT TREATMENT RONS
Powdered Filter
.eachate Ca(OH)2 Na2S HC« (~37X) Carbon Aid
Waste (g/L) (g/1) (nl/L) (g/L) (g/L)
,ite A 0.453 0.059 1.47 5.7 1.7
>ite B 4.81 0.52 8.87 5.7 6.6
ite B (spike) 4.99 0.49 10.8 5.6 9.4
>ite C 1.78 0.26 2.33 5.2 5.4
ite C (spike) 2.21 0.27 4.47 5.3 5.6
ite D 0.430 0.058 0.29 10.7 7.2
to reduce effluent discharges. Such treat-
ment generates sludges or solid residues
that contain these metals in various chemi-
cal forms and concentrations. The objec-
tive of this program is to test processing
techniques which mitigate the hazardous
nature of these wastes. One such process
is the separation of toxic metal precipi-
tates from the waste via high gradient
magnetic separation (HGMS).
Amount of
from formed tn
Sice lab. testa (g/L)
A 0.38
B 3 52
C 1.08
D 1.25
carbon, wet dcnsicy of 1.
g/cmi.
In pilot cests added:
carbon
<«/L>*
5 7
5.7
10.7
(1) precipil
.6 g/cm3. (3)
aid
(g/U1
1.7
6.6
5.4
7.2
rates. 2.2
filter aid
Sludge volume (cm-/L)
50 we.
water
13
25
19
31
Z 75 we. Z
wacer
28
57
70
g/cm3; (2) powdered
, wee density or I 5
HGMS is a filtration process which uti-
lizes a filtering matrix of flux conducting
material. Because of the geometry of the
matrix, intense magnetic field gradients
can be developed in the matrix which can
capture weakly magnetic material. The HGMS
process has been tested on a synthetic
waste comprised of nickel hydroxide preci-
pitation from a nitrate solution in a
slurry of water and diatomaceous earth.
The preliminary data obtained from a
pilot plant study indicate the HGMS can
remove the nickel hydroxide, a weakly
magnetic metal, although whether or not the
performance of the filter is sufficient to
make it a viable treatment process has not
yet been determined.
REFERENCES
1. EPA. 1980. Electroplating Point Source
Category Effluent Guidelines and
Standards, Pretreatment Standards for
Existing Sources. Federal Register,
44(130):45322-45328 (July 3, 1980).
2. EPA. 1980. Development Document for
Effluent Limitations Guidelines and
Standards for the Metal Finishing Point
Source Category (Draft). EPA-440-1-80-
091-A, U.S. Environmental Protection
Agency, Washington, D.C.
3. Lyman, W.J. and G. Contos. 1980. In-
organic Hazardous Waste Treatment. In:
Treatment of Hazardous Wastes,
Proceedings of the Sixth Annual Research
Symposium, EPA-600/9-80-011, U.S.
Environmental Protection Agency,
Cincinnati, Ohio. pp. 62-71.
4. NCWQ. 1975. Water Pollution Abatement
Technology: Capabilities and Costs —
Metal Finishing Industry. National
Commission on Water Quality, Washington,
D.C.
263
-------
TABLE 9. CONCENTRATION (mg/L) OF SELECTED HEAVY METALS IN PRE- AND*POST-TREATMENT LEACHATE SAMPLES
Metal
Fe
Mg
Mn
Co
Cr
Cu
Ni
Pb
Zn
As
Site
Initial
27.5
11.2
5.54
<0.05
0.016
0.045
0.12
<0.05
0.072
e
Aa
Final
1.1
1.2
0.02
<0.05
0.008
<0.005
0.09
<0.05
0.049
e
Site B
Initial1" Final
741 3
88.6 2.8
4.91 0.27
1.0 <0.1
(<.05, <.l)
8.01 0.04
(.028, <.02)
8.34 1.15
(.094, 39)
9.33 0.13
(<.01, <.03)
7.40 <0.2
10.0 < 0.05
(.16, .14)
9.0 0.033
(-, .008)
Site C
Initlalb
53.4
44.9
2.59
1.2
(<.5, .3)
9.02
(<.08, <.02)
8.86
(.09, .16)
9.12
(<.l, <.03)
8.10
9.65
(<.2, .35)
6.5
(-, .003)
Final
3.0
1.3
0.27
<0.1
0.06
0.60
0.07
<0.20
0.08
0.008
Site
Initial
322
51.9
85.1
<0.5
<0.08
0.14
1.1
<0.5
<0.2
0.77
DC
Final
<2
42d
4.2
<0.5
<0.08
<0.08
0.1
<0.5
<0.2
0.008
a. From laboratory tests.
b. Underlined values are artifically high due to addition of spikes in pilot runs; values in parentheses
underneath are from unspiked laboratory and pilot plant runs, respectively. Were spike was added,
"final" value is from the spike run.
c. From pilot plant run.
d. Questionable value; was <0.4 in final lab sample.
e. Not available.
264
-------
EMERGING TECHNOLOGIES FOR THE DESTRUCTION OF HAZARDOUS WASTE
ULTRAVIOLET/OZONE DESTRUCTION
Barbara H. Edwards, John N. Paullin, Kathleen Coghlan-Jordan
Ebon Research Systems
Washington, D.C. 20011
ABSTRACT
Ebon Research Systems has investigated new technologies for the disposal of hazardous
wastes. These methods are not state-of-the-art, but involve new technologies or a recent
variation of an established technology. Information was acquired by computerized search,
library searching, and personal contacts. A list of the emerging technologies includes:
molten salt combustion, fluidized bed combustion, high energy electron bombardment, the
catalyzed wet oxidation of toxic chemicals, dehalogenation of compounds by treatment with
ultraviolet light and hydrogen, UV-chlorinolysis of organics in aqueous solution, the
catalytic hydrogenation dechlorination of polychlorinated biphenyls, and Ultraviolet/Ozone
destruction.
Ultraviolet/Ozone destruction (UV/Ozonation) is discussed in detail. Theory, specfic
wastes treated, and economics are reviewed. Hazardous wastes are destroyed by treatment
with a combination of UV light and ozonation. Wastes streams of up to 1% toxic contaminant
can be treated. The blending of ozone's powerful oxidizing capacity with UV light's
additional energy produces free radicals and excited-state species to accomplish the
effective destruction of many hazardous organic chemicals. Among the wastes treated by
UV/Ozonation are: PCBs, chlorodioxins such as TCDD, various hydrazines, nitrobenzene, and
components of a copper process waste stream. Because ozone is a powerful oxidant, it
cannot be shipped and must be generated on site. Waste stream effluent must be monitored
for the possible generation of toxic intermediates during treatment.
Introduction
The quantity of hazardous wastes
generated in the United States will
exceed the capacity of landfills. The
ultimate fate of many of these wastes in
landfills is unknown. The argument can be
espoused: Why inflict upon the next gen-
eration the waste disposal problems of this
generation?
Conventional incineration has been a
popular alternative to landfills, however
there can be a problem with particulate
emissions. Moreover, most municipal and
industrial incinerators do not reach
sufficiently high temperatures to com-
pletely destroy halogenated organics such
as PCBs.
Ebon Research Systems has researched
emerging technologies for hazardous waste
disposal. These methods are not state-
of-the-art, but involve a new technique.
UV/Ozonation can effectively destroy toxic
organics such as PCBs.
Ozone
Ozone (02) is a three atom allotrope
of oxygen. It is second only to fluorine
in electronegative potential and has long
been recognized as a powerful disinfectant
and oxidant of both organic and inorganic
substances. Generated by solar energy,
ozone is a natural ingredient of the
earth's atmosphere. It is also generated
by atmospheric oxygen with energy from
lightning, and is associated with the
operation of most electrical equipment.
Under ambient conditions, ozone is a gas.
Unreacted ozone decomposes in a matter of
hours to simple, molecular oxygen. Low
levels of ozone can damage delicate nasal,
265
-------
bronchial, and pulmonary membranes.
Mechanisms for UV/Ozonation Reactions
with Organic Compounds
When ozone is used to oxidize organic
compounds in aqueous solutions, gaseous
phase ozone must dissolve in water in
order to react with liquid phase organics.
It has been reported that conditions
favoring organic ozidation, alkaline pH,
and elevated temperature also favor ozone
decomposition. Elevated temperature
reduces ozone solubility in water and
causes inefficient ozone utilization
even as organic oxidation increases.
Ozone is relatively soluble in acid solu-
tion, but it decomposes rapidly with
increasing alkalinity (Lee 4). However,
temperature control is often uneconom-
ical, and ozonation at high pH followed
by neutralization will also increase
costs.
In many cases, oxidation of organic
different pH levels and varying tempera-
tures with ozone alone results in the
formation of highly refractory organic
intermediates. If ozone treatment is
combined with UV radiation, the UV
activates the initial refractory compound
and all subsequent species, permitting
oxidation to carbon dioxide, water, and
other elementary species (Prengle and
Mauk 9).
The following groups are especially
vulnerable to attack by ozone enhanced
UV absorption:
• exposed halogen atoms
• unsaturated resonant carbon ring
structures
• readily accessible multi-bonded
carbon atoms
• alcohol and ether linkages
Shielded multi-bonded carbon atoms, sulfur,
and phosphorous, are much less vulnerable.
There is a substantial difference in
the chemistry involved and results
achieved when the UV/Ozone process for
treating hazardous waste is compared to
straight ozonation. Ozonation of dilute
solutions of organic species produces
organic oxidation and hydroxyl radical
formation. Bridging of carbon-carbon
double bonds by ozone forms unstable
ozonide intermediates that decompose
into smaller oxidation species. This
process continues until the ultimate final
products, carbon dioxide and water or rela-
tively stable refractory compounds (such as
acetic acids or oxalic acids are formed
(Prengle and Mauk 9).
When ultraviolet radiation, in the
180-400 ran range, is added to the process,
72-155 kcals/mole of additional energy is
provided, and the 03 molecule breaks down
in oxidizing 0 species. This is ample
energy for producing substantially more
free radicals and excited state species for
the initial compound and subsequent
oxidation species than those produced by
ozonation alone. As a result, excited
atomic species (0), hydroxy (OH), and
hydroperoxy (H02 ) radicals, and excited-
state species (M), are produced from
reactant molecules. These all greatly
enhance the overall oxidation rate, prevent
plateauing, and permit complete oxidation.
The following is a simplified representa-
tion of how these excited species are
probably formed:
M + hv
R, I + H
Overall:
R \+ OH, HO , 0,
-2 -3
(Yi H (i (M QO DT"1
2' 2 ' 4 ' 4
where M is the pollutant species being
oxidized, R is a free radical species, and
I is an intermediate molecular species
(Prengle and Mauk 9).
Elevated UV light also accelerates
ozone decomposition, thus using UV
excessively results in less favorable
economics (Prengle et al. 8). However,
no generalization regarding requirements
can be made because of the wide variation
in the reactivity of species. Free cyanide
is oxidized by ozone without elevated
temperature or UV, but iron-complexed
cyanides have maximum practical limits for
266
-------
temperature and UV requirements. It is
better to treat acetic acid at naturally
occurring pH rather than neutralize, yet
phenol destruction is not strongly pH
dependent (Prengle et al. 8).
To summarize, the overall mechanism of
UV/Ozone photo-oxidation of the pollutant
species in aqueous solution occurs by a
combination of the following:
• 03 photolysis to produce
oxidizing 0 species
• H20 photolysis and reaction with
O to produce OH
*
• M photolysis to produce M and
free radicals
The free radicals participate in a
sequence of oxidation reactions leading to
the final oxidation products. For M-
species that are low UV absorbers, the
rate controlling mechanism is photolysis
oxygen species oxidation, but for high
UV absorbers, M-species photolysis is
rate controlling.
Mathematical models have been
developed to determine the overall rate
equation for the UV/Ozone process. A rate
equation was proposed to describe the
reaction in the intermediate stage. The
rate during this period was described as
a simple function of superficial gas
velocity, ozone partial pressure, and the
contaminant concentration. It was con-
cluded, the destruction rate for toxic
organic chemicals is directly proportional
to the toxic organic chemical concentration
and to the 1.5th power of the ozone feed
rate. These models may provide a useful
tool for design and scale-up of ozone
contactors (Lee 4).
Ozone Generators
Unlike most chemicals, there is no
controlled available natural source for
ozone, and it is not practical to ship it
or store it in containers. Consequently,
ozone is generated on-site by the user.
Ozone is generated when an oxygen molecule
is sufficiently excited to disassociate
into atomic oxygen; further collisions with
oxygen molecules then cause the formation
of the ozone molecule. Excitation energy
can be supplied by ultraviolet light or
high voltage (Klein et al. 3).
Oxygen is converted into ozone as follows:
20.,
Most ozone generators employ high
voltage, and produce a corona discharge
(also know as an silent arc discharge or
brush discharge). The following condi-
tions are necessary to produce a corona
discharge:
• two electrodes separated by a gap
• a gas in the gap
• sufficient voltage potential
between the two electrodes to
cause current flow through the
dielectric and gas (Ozone Research
and Equipment Corporation 7)
The electrodes can be flat, tubular,
or any configuration that allows the
opposing surfaces to be parallel. The
distance between the parallel surfaces of
the electrodes should be large enough to
ensure uniform current flow. A gap
smaller than certain critical size may
restrict air flow to a point of excessive
pressure drop through the unit, while a
large gap increases voltage requirements
(Klein et al. 3).
The two basic types of commercial
ozone generators in use are the concentric
tube generator, and the parallel plate
generator. The concentric tube system,
first devised by Siemens, can be modified
to operate at higher pressures than the
plate design (Klein ^t a^. 3). In a con-
centric tube generator, a glass electrode
coated with a metal is surrounded by a
second electrode that is a metal tube. In
some designs, there is a stainless steel
center electrode, with the glass dielectric
tube concentric on the outside. The glass
tube is coated with a metal that serves as
the second electrode. Glass functions as
a dielectric to increase the voltage gap
between the electrodes without sparking.
Ozone is formed from the oxygen in the gap
between the tubes.
The more complex Otto system uses high
voltage parallel plates that alternate
with ones that are water cooled. Gas flows
in the air gaps between the plates. There
is more generating surface for a given
volume with the parallel plate design when
compared to the concentric tube design.
Air cooling may replace water cooling, be-
cause water cooling is more costly.
267
-------
Yet, many years of experience with water
systems have developed relatively trouble
free techniques (Klein, et al. 3).
Dielectric
The operational efficiency of an ozone
generator is dependent on design and
engineering considerations. The dielectric
material selected should have both a high
electrical resistance and high thermal
conductivity. Since these properties
rarely occur together in a material, the
dielectric is usually chosen for its high
electrical resistance, and used at a
minimum thickness to overcome heat transfer
deficiencies. Glass or other ceramic
materials are usually used.
Pressure
The operating pressure should provide
sufficient force to deliver ozonized gas to
the contacting vessel. The pressure may
have to overcome a back pressure of water
at the reactor. As pressure affects the
electrical impedance of the system, the
operating pressure must be correlated with
the air gap and preferred voltage (Klein et
al. 3).
Flow Rate
The time that the feed gas is in the
electrical discharge determines the con-
centration of ozone in the effluent. This
concentration can vary from 0.5-10 wt% in
a well designed, efficient generator
operating at ideal conditions. Klein notes
that the power efficiency drops rapidly as
the ozone concentration increases.
Temperature
Approximately 90% of the energy
applied to the ozonator is lost as heat.
As the ozone decomposition rate increases
with increasing temperature, provisions
should be made for the rapid removal of
excess heat. As ozone generator
temperature increases, the dielectric
material changes thermal characteristics
and is subject to rupture. Because
critical generator dielectric temperatures
are relatively low (120°-130°C), the
economics of refrigeration or coolant
quantity supplied to the system must be
evaluated versus the temperature of the
ozone stream (Ozone Research and Equipment
Corporation 7). Water is usually
considered a more efficient medium for
heat transfer than air. Water used for
cooling ozonators need not be of high
quality, and when it passes through the
generator it is not degraded or changed.
Feed Gas
Ozone generation using oxygen is
approximately twice as efficient as air
(ozone generated from air requires at least
twice the power of an equal amount of ozone
generated from pure oxygen). However, the
cost of pure oxygen is more than twice the
cost of air. Unless oxygen is recovered
via a closed system, it is cheaper,
especially when ozonate is generated at
less than 225 kg/day, to generate ozone
from air. A recent development is a
technique that uses a swing cycle to
concentrate oxygen from air by alternate
passage through molecular sieves (Derrick
and Perrich 2).
Power
The major operating cost for ozone
manufacturing is the cost of electric
power. Of the total power utilized in
ozone generation, approximately 10% is used
to actually produce ozone. The remainder
is consumed in air handling, air prepara-
tion, and waste heat. The use of higher
frequencies increases ozone generation
efficiency, yet the electrical cost of the
high frequencies may offset the gain. This
situation may be altered with new advances
in solid state technology, according to
Klein.
Destruct Systems
The reactor is a key element in the
efficient utilization of ozone. The
reactor should be designed for optimum
ozone utilization, following established
principles of mass transfer and reaction
kinetics (Derrick and Perrich 2). In some
cases, ozone not consumed in the primary
reactor system can be reused or applied
beneficially to another system. However,
it may be necessary to provide a destruct
system to guarantee removal of any unused
ozone from the system before the gas stream
is discharged to the atmosphere. There are
three types of destruct systems: thermal,
catalytic, and combination. Thermal
systems heat the entire gas stream to a
high temperature for a specific period of
time (600°C for 1 second). Catalytic
268
-------
systems heat the gas to 250°C, and then
pass the gas stream through a solid phase
catalyst bed for ozone destruction. The
catalysts are proprietary. Combination
systems represent a trade-off between
heating costs and catalyst costs. In all
cases, heat recovery should be considered
(Derrick and Perrich 2).
Hazardous wastes treated include PCBs,
dioxins, hydrazines, nitrobenzene, and a
copper process waste stream.
TCDD
2,3,7,8-tetrachlorod ibenzo-p-d iox in
was treated with an UV/Ozone system by
researchers at California Analytical
Laboratories and the Carborundum Company
(Wong, et al. 11). Two types of ozonation
systems were used: (1) purified oxygen
to generate ozone. (2) a commercial unit
was used. One ppb of TCDD was completely
degraded. OCDD, a common dioxin impurity
in the wood treatment agent PCP (penta-
chlorophenol) was also successfully
degraded. These preliminary studies
indicate the feasibity of UV/Ozone
degradation systems for the treatment of
dioxins in a waste stream (Vtong et al. 11).
Hydrazine fuels
The hydrazine family of fuels includes
hydrazine, monomethylhydrazine, and unsym-
metrical hydrazine as well as mixtures of
these compounds. A study for the United
States Air Force investigated the effect of
UV/Ozonation on these compounds. The
following conclusions resulted:
• The presence of UV light in thee
reactor reduces half-life valuess
(t 1/2), and increases reaction
rates for the ozonation of the
hydrazine fuels over ozone
reactions in the absence of UV
light.
• An increase in solution pH, in-
creases the oxidation rate of the
hydrazine fuels.
• The quantity of methanol produced
from ozone oxidation of monomethyl-
hydrazine is proportional to the
species concentration.
• Monomethylhydrazine decomposition
during oxygen or air sparging is
greatly enhanced by UV light.
• Increasing ozone partial pressure
decreased half-life values, but
ozone utilization efficiency is
reduced.
• An increase in species concentra-
tion produced an increase in the
required retention time to achieve
the proper effluent levels (Sierka
and Cowen 10).
Nitrobenzene
Investigators at the Vfestgate Research
Corporation studied the effects of ozone
and UV/Ozonation on nitrobenzene. The
authors concluded that the UV/Ozonation of
organic compounds in water at levels of
about 0.8 mM is probably not a free radical
phenomenon. The intermediates oxalic and
formic acids, although stable in the
absence of UV light, were readily oxidized
to carbon dioxide by the UV/Ozone system.
(Leitis, et al. 5).
Copper process waste stream
A copper process waste stream
containing primarily sodium sulfate, formic
acid, an anionic surfactant, and EDTA was
treated with UV/Ozonation by scientists at
the Systems Products Division of IBM. In
order to discharge the treated effluent
into the plant clarification system, the
final concentration of EDTA was limited to
5 ppm (Macur, et al. 6). The following
conclusions were reached:
• Waste solution pH should be
maintained in the 4-6 range.
• Iron (III) levels should be below
5 ppm.
• Copper (II) levels should be less
than 1 ppm.
PCBs
A pilot plant was set up at a General
Electric Company plant to demonstrate the
efficiency and cost-effectiveness of a
commercial ULTROX UV/Ozone system to
destroy PCBs. PCB levels in the range of
5-40 ppm were studied. Many of the runs
had no PCB residue in the effluent.
269
-------
An economic projection for a 150,000
gallons per day ULTROX treatment plant
designed to attain 1 ppm or lower PCB in
the effluent (assuming a 50 ppm feed)
varied from $300,000-$350,000. Design and
economic data for a 40,000 gallons per day
(151,400 liters per day) plant are sum-
marized below in Tables 1 and 2 (Arisman
and Musick 1).
TABLE 1. DESIGN DATA FOR A 40,000 GPD
(151,400 LITERS/DAY) ULTROX PLANT
Reactor
Dimensions,
Meters (LxWxH) 2.5 x 4.9 x 1.5
Wet Volume, Liters 14,951
UV Lamps
Number of 65 watt lamps 378
Total Power, KW 25
Ozone Generator
Dimensions,
Meters (LxWxH) 1.7 x 1.8 x 1.2
gms Ozone/minute 5.3
kg Ozone/day 7.7
Total Power, KW 7.0
Total Energy reguired 768
(KWH/day)
TABLE 2. EQUIPMENT PLUS OPERATING AND
MAINTENANCE COSTS 40,000 GPD ULTROX PLANT
Reactor $94,500
Generator 30,000
$124,500
O & M Costs/Day
Ozone generator power $4.25
UV lamp power 15.00
Maintenance 27.00
(Lamp Replacement)
Equipment Amortization
(10 years @ 10%) 41.90
Monitoring labor 85.71
TOTAL/DAY $173.86
Cost per 1,000 gals
(3,785 liters) with
monitoring labor $4.35
Cost per 1,000 gals
without monitoring labor $2.20
Source: Arisman and Musick, General Elec-
tric Co., Hudson Falls, NY, Zeff and Crase
Westgate Research Corp., West L.A., Calif.
Advantages of UV/Ozonation
• Aqueous or gaseous waste streams
can be treated.
• Capital and operating costs are not
excessive as compared to incinera-
tion.
• The system is readily adaptable to
the on-site treatment of the haz-
ardous waste.
• UV/Ozonation can be used as a
final treatment for certain wastes.
• It can be used as a preliminary
treatment for certain wastes.
• Effluent discharge standards can
be met.
• Modern systems are usually auto-
mated, thereby reducing labor
requirements.
Disadvantages of UV/Ozonation
• Ozone is a non-selective oxidant;
the waste stream should contain
primarily the waste of interest.
• Certain compounds because of their
structure are not amenable to
UV/Ozonation (high refractory
index).
• UV/Ozone systems are generally
restricted to 1% or lower levels
of toxic compounds. The system
is not amenable to bulky wastes.
• Toxic intermediates may persist
in the waste stream effluent.
• Ozone deconposed rapidly with in-
creasing temperature. Excess heat
must be removed rapidly.
Conclusions
UV/Ozonation is a viable emerging
technology for the treatment of hazardous
wastes when compared to the problems of
landfills and conventional incineration.
UV/Ozonation is not cost-prohibitive from
the standpoint of capital plus operating
and maintenance costs (0 & M). The process
is particularly suitable for the on-site
treatment of hazardous wastes.
270
-------
REFERENCES
1. Arisman, R.K. and R.C. Musick. 1980.
Experience in operation of a UV-Ozone
ULTROX pilot plant for destroying PCBs
in industrial waste effluent. Pre-
sented at the 35th Annual Purdue
Industrial Waste Conference, May, 1980.
2. Derrick, O.K. and J.R. Perrich. 1979.
Guide to ozone equipment selection.
Pollution Engineering, 11:42-44.
3. Klein, M.J. et al. 1973. Generation of
ozone. Presented before the First
International Symposium on Ozone for
Water and Wastewater Treatment.
4. Lee, M.K. 1980. Study of UV-ozone
reaction with organic compounds in
water. Presented before the Division
of Environmental Chemistry, American
Chemical Society, Houston, Texas.
5. Leitis, E. et al. 1979. An investi-
gation into the chemistry of the UV/
Ozone purification process. Presented
at the 4th World Ozone Congress,
Houston, Texas.
6. Macur, G.J. et al. 1980. Oxidation of
organic compounds in concentrated
industrial waste water with ozone and
ultraviolet light. Presented at the
35th Annual Purdue Industrial Waste
Conference, Layfayette, Indiana.
7. Ozone Research and Equipment Corpora-
tion, Phoenix, Arizona. Ozone Tech-
nology. Brochure 124.
8. Prengle, H.W. et al. 1975. Ozone/UV
process effective wastewater treat-
ment. Environmental Management,
54:82-87.
9. Prengle, H.W. and C.E. Mauk. 1978. New
technology: ozone/UV chemical oxidation
wastewater process for metal complexes,
organic species and disinfection. The
American Institute of Chemical
Engineers Symposium Series, 74:228-243.
10. Sierka, R.A. and W.F. Cowen. 1980. The
catalytic ozone oxidation of aqueous
solutions of hydrazine, monomethyl-
hydrazine, and unsymmetrical
dimethylhydrazine. Presented at the
35th Annual Purdue Industrial Waste
Treatment Conference, Layafette, Ind.
11. Wong, A.S. et al. 1979. Ozonation of
2,3,7,8-tetrachlorod ibenzo-p-d iox in.
Presented at the Symposium on the
Chemistry of Chlorinated Dibenzo-
dioxins and Dibenzofurans, American
Chemical Society National Meeting,
Washington, D.C.
* ULTROX is a registered trademark.
271
-------
EVALUATION OF CATALYZED WET OXIDATION
FOR TREATING HAZARDOUS WASTE
Richard A. Miller
Robert D. Fox
David M. Pitts
IT Enviroscience
Knoxville, Tennessee 37923
ABSTRACT
This paper describes the evaluation of a unique, patented catalyst system
for treating toxic and hazardous waste from active and abandoned chemical
landfills. The catalyst system uses an acidic solution of bromide,
nitrate, and manganese ions to destroy organic residues and aqueous
wastes. Fifteen compounds were selected to represent the wide variety
of wastes which could be found in a chemical landfill. Batch oxidations
of these compounds were performed in a 1-liter stirred autoclave to
determine the destruction rate and by-products of the process. While
detailed analysis of the results has not been completed, sufficient data
are available to provide insight into effectiveness and applicability of
the process. Preliminary designs of treatment processes have been made
based on the demonstrated destruction rates. Cost estimates for these
processes are provided to permit assessment of the technology for treating
hazardous wastes.
INTRODUCTION
The program described in this
paper had as its goal the applica-
tion of a new and novel catalytic
wet oxidation technology for the
treatment of toxic and hazardous
waste materials. On the basis of
preliminary range-finding testing
and economic evaluation, the tech-
nology appears to have wide ranging
applicability and potential for
alleviating significant waste
disposal problems facing the United
States. Candidate wastes which are
potentially treatable by this
technology include aqueous wastes,
organic liquids, sludges, and
solids. Of particular interest to
this project is the development and
application of new technology to
the problems associated with aban-
doned chemical waste disposal sites.
The Love Canal and the Valley of
the Drums are two among many
already unearthed. The estimate
of the number of such sites in the
United States and the cost to
clean them up is staggering.
Typical of the disposal problems
of these sites are treatment of
leachate and groundwater around
the site and the ultimate disposal
of the sludges, soils, liquids,
solids, and containers on the site.
The most desirable technology
for dealing with the problem of
chemical disposal sites would be
a process that will 1) achieve
complete destruction of toxic and
hazardous chemicals and minimize
the by-products and chemical
remnants requiring disposal;
2) minimize the energy required
for disposal, especially supple-
mental fuels; 3) have low chemical
usage; 4) have few unit operations;
272
-------
5) have low volume effluents, such
that polishing treatment to achieve
total containment can be readily
implemented; 6) be easily pilot
planted; and 7) be transportable
to the landfill sites.
Incineration systems are
generally regarded as capable of
meeting the first objective, but
run into difficulty in the other
areas. Conventional wet oxidation
systems have the capability to
meet the last six criteria, but
are limited in the first. Other
destructive processes such as
biological treatment or chemical
oxidation are either ineffective
or very expensive. IT Enviro-
science believes that the unique
features of its catalyzed wet
oxidation process provide the
potential to overcome the short-
comings of other destruction
processes. It approaches the
omnivorous and total destruction
characteristics of incineration
while retaining the advantages of
wet oxidation systems in other
areas.
CATALYZED WET OXIDATION PROCESS
CONCEPT
The catalyzed wet oxidation
process is based on U.S. Patent
3,984,311, originally assigned to
The Dow Chemical Company and now
assigned to IT Enviroscience for
development and commercialization.
The patent teaches the use of a co-
catalyst system consisting of
bromide and nitrate anions in
an acidic, aqueous solution.
Continued research with the co-
catalyst system has led to the
development of a new catalyst
mixture which is more effective
for oxidizing insoluble organics.
The new catalyst system consists
of bromide, nitrate, and manganese
ions in acidic solution. A patent
for the new catalyst has been
allowed and assigned to IT Enviro-
science. The destruction of most
organics by these catalyst systems
is rapid and essentially complete.
The keys to the operation of
this catalyst system are the
mechanism of oxygen fixation and
the fact that it is a water-
soluble, single-phase catalyst
system--a homogeneous catalyst
as contrasted to a heterogeneous
catalyst. In conventional wet
oxidation, heat and pressure are
used to drive the dissolution of
oxygen from air and the reaction
with dissolved organics in aqueous
solution. In the bromide-nitrate-
manganese catalyst system the
transfer of oxygen to the dissolved
state is speeded up by using very
rapid gas and liquid reactions
associated with the catalyst
components. The importance of
the enhanced oxygen transfer is
the ability to oxidize organics
at much lower temperatures than
uncatalyzed wet oxidation,
165—200°C versus 250—325°C. The
lower operating temperatures also
mean lower operating pressures
which not only reduces capital
cost but operational problems.
The second important aspect
of the catalyst system is its
homogeneous nature which permits
application to the destruction
of toxic or hazardous organic
residues, such as still bottoms
or other organic wastes. The
advantages of a homogeneous
catalyst are best utilized by
using a reactor design which is
different from conventional wet
oxidation processes. In simplest
form the reactor, a continuously
stirred tank reactor (CSTR)
contains the catalyst solution.
Air and the waste are continuously
pumped into the reactor and the
organics are oxidized with the heat
of reaction driving off water. The
only materials to leave the reactor
are CO^, N2, water vapor, any
volatile organics and inorganic
solids formed. Water is condensed
and returned to the reactor, if
necessary, as are condensable
organics. Any inorganic salts or
acids which may be formed have to
be removed by treatment of a closed
loop stream of catalyst solution.
Such treatment is individually
designed utilizing conventional
technologies, such as filtration or
distillation. The vent gases from
273
-------
the reactor are low in volume and
may, if necessary, be treated by
conventional techniques, such as
absorption, adsorption, or
scrubbing. The most important
features of this process concept
are that nonvolatile organics
remain in the reactor until
destroyed, and that there is no
aqueous bottoms product.
EXPERIMENTAL
Compound Selection
The study of a process
associated with chemical wastes
needs to evaluate a wide range of
chemical compounds to evaluate
the applicability of the process.
In order to meet this requirement
fifteen compounds were selected
from various chemical groupings
of the EPA priority pollutant list.
The specific compounds chosen from
each grouping were made based on
the likelihood of finding the
chemical in industrial waste and
chemical disposal sites. In
certain cases a lower toxicity
compound was substituted for a
highly toxic compound, e.g.,
diphenylhydrazine substituted for
benzidine, when it presented an
unacceptably high risk to labora-
tory personnel. The chemical
groupings and the compounds studied
were:
- Halogenated hydrocarbons
Ethylene dibromide
Hexachlorobutadiene
Trichloropropane
- Pesticides
Atrazine
DDT
Malathion
Mirex
- Phenols
Pentachlorophenol
- Phthalates
Di-n-butyl phthalate
- Polynuclear aromatics
Chloroanthracene
- Miscellaneous
Acetonitrile
Chloroaniline
Diphenyl hydrazine
Nitrobenzene
o-Xylene
Experimental Method
The experimental evaluation
of catalyzed wet oxidation was
conducted in a 1-liter agitated
titanium autoclave. The destruc-
tion rate of the compounds was
measured at various operating
conditions with batch reactions.
Typical operating conditions were
165—250°C, 0.5% Br, 5.0% NO3,
0.25% Mn, 30—120 minutes reaction
time, and 20% excess oxygen over
stoichiometric. The procedure
for conducting the oxidation
reactions consisted of loading
the reactor with the desired
quantity of deionized water, HBr,
and the organic to be oxidized.
The reactor was sealed, purged
with oxygen, and pressurized with
sufficient oxygen to totally
oxidize the organic. The reactor
was then heated to the desired
operating temperature and the
remaining catalysts, HNO3 and
MnSO4, were added to the reactor
with oxygen pressure. The reaction
was run for the desired time
period. At the termination of the
reaction, the reactor was cooled
to room temperature with cooling
water. The pressure was then
vented and any free bromine was
reduced to bromide by the addition
of 10 ml of sodium bisulfite
solution to prevent operator
exposure to toxic bromine vapors.
The reactor was opened and the
aqueous contents were aspirated
from the reactor. All internal
surfaces of the reactor were rinsed
twice with solvent, methylene
chloride. The solvent rinsings
were aspirated from the reactor
and combined with the aqueous'
effluent. The aqueous phase was
extracted with the solvent rinse
and two successive 50-ml portions
of solvent to recover unreacted
organics and reaction by-products.
The combined solvent extracts were
analyzed by gas chromatography and
gas chromatography/mass spectro-
metry (GC/MS). The destruction
rate of the organic was also
measured by the formation of CC>2,
by gas chromatography, and the
formation of Cl~ (from chlorinated
organics), by ion chromatography.
274
-------
These measurements were used to
determine the completeness of
the organic destruction.
Experimental Results
The compilation and inter-
pretation of the experimental
results has not been completed
at this time. Initial examination
of the data reveals that catalyzed
wet oxidation was effective in
destroying most of the compounds
tested. The oxidation of atrazine,
butyl phthalate, chloroaniline,
diphenyl hydrazine, ethylene
dibromide, malathion, pentachloro-
phenol and xylene was rapid (<60
minutes) and essentially complete
(>60% destruction to CC^) under
mild oxidation conditions (165—
200°C). Some of the compounds,
acetonitrile, chloroanthracene,
DDT, hexachlorobutadiene, nitro-
benzene and trichloropropane
required higher oxidation tempera-
tures (200 250°C), longer
reaction times (up to 120 minutes),
and were only partially oxidized
to C02 (<40% destruction to C02).
Mirex was the most resistant to
oxidation by the catalyst system
with <10% destruction even at
very high temperatures, 275°C.
While these results are very
general, the ability of the
bromide-nitrate-manganese
catalyst system to destroy a
wide range of organics was demon-
strated.
PROCESS DESIGN AND ECONOMICS
In applying this technology
to destroying the contents of a
chemical landfill a "worst case"
approach was used to design a
treatment process and estimate
the operating costs. This general
basis approach should be used
because the contents of a landfill
will be an unknown mixture of
organics and it is safest to
assume a slow destruction rate
for the mixture. Two different
processes were designed to provide
the basis of estimating destruction
costs. The equipment basis for
both processes is a 1000-gallon
reactor system which can be
transported to the landfill site.
The first process is designed
to continuously oxidize organic
residues with moderate-slow
destruction rates. Under continu-
ous feed conditions, organic
residues are added to the reactor
at a rate that represents the
steady-state destruction rate
and the organics are consumed
at the rate they are added; a
constant "heel" of organics is
maintained in the reactor. At
some point, when the catalyst
solution may need to be disposed
of, the reactor is batch-operated
to destroy the heel of organics
remaining in the reactor. It is
at this point that total destruc-
tion of the organics is achieved
and concentrations are reduced
to ppm or ppb levels. It is
estimated that the operating,
(without equipment depreciation)
cost for destroying organic
residues with catalyzed wet
oxidation will be 15 to 30 cents
per pound of organic.
The second process is designed
to treat an aqueous waste with 4%
organic. This process differs
from the first in the quantity
of water that must be handled.
The process handles the additional
water by utilizing the heat of
reaction of the organics to drive
the water overhead for removal
with a condenser. For dilute
aqueous wastes, where insufficient
energy is released by the oxi-
dation of the organics to evap-
orate all of the incoming and
formed water, it is necessary to
provide an evaporator to maintain
the catalyst concentration • in the
reactor. Energy for the evapora-
tion of the dilute catalyst
solution in the reactor comes
from the reactor vent gas. The
operating cost for this process,
less depreciation of equipment,
is estimated to be 27 cents per
gallon of aqueous waste.
CONCLUSIONS
The catalyzed wet oxidation
process has successfully destroyed
a wide range of organics in
275
-------
laboratory batch tests. The
strong oxidizing power of the
homogeneous catalyst system
coupled with a process design
which retains unreacted organics
makes this process a viable
technology for solving the problems
associated with chemical disposal
sites. Other aspects of the
catalyzed wet oxidation process
such as energy efficiency, low
chemical usage, low volume
effluents, and portability also
make the process attractive for
solving industrial waste problems.
This research was funded by
U.S. Environmental Protection
Agency Contract No. 68-03-2568,
Work Directive T-7016.
276
-------
REGISTRATION REQUIREMENTS FOR PESTICIDE LABELING - LAND DISPOSAL
Dimitrios Kollias and Janet Brambley
Systems Research Company
Philadelphia, PA 1910U
ABSTRACT
An outline is presented of the information necessary to initiate the development of
data requirements and testing protocols to evaluate label statements on the land disposal
of pesticides, in compliance with the proposed guidelines for registering pesticides in
the United States, ^0-CFR-l62. The information surveyed includes soil studies, physico-
chemical degradation (hydrolysis and photolysis) of pesticides, soil-pesticide inter-
actions (adsorption, volatility and leaching), metabolism (soil-microorganism relation-
ships, aerobic and anaerobic metabolism), effects on highter plants and animals, which
apply to land disposal methods. The lack of studies carried out with pesticides in
amounts compatible with the actual disposal leve7s, as opposed to use levels, is evident,
as in the absence of data on pesticide mixtures, pesticide interactions, etc.
INTRODUCTION
Pesticide containers are required by
law to carry on the label instructions
for the disposal and storage of excess
pesticides and pesticide containers. The
intelligibility and reliability of these
statements has been questioned in the
past (l).
The USEPA under J*0-CFR-l62 has pro-
posed guidelines for registering pesti-
cides in the United States, specifying
the data required to support the regis-
tration of a product, describing the
standards for acceptable testing, pro-
viding references to test protocols in
the literature, and outlining the format
for data reporting. Under CFR Section
162-62-13, environmental chemistry data
are required to support pesticide label
statements on disposal and storage of
registered manufactured use products and
all formulated products.
The object of this contract is to
generate data requirements and testing
protocols for better evaluation of spe-
cific pesticide disposal and storage re-
commendations, with special emphasis on
the development of improved label direc-
tions for proper storage and disposal of
excess pesticides and pesticide contain-
ers .
The chemical characteristics of pes-
ticides and their diversity of use may
preclude the development of universally
acceptable environmental chemistry data
requirements. Nevertheless, the scope of
work requires the development, as nearly
as possible, of universally acceptable
testing protocols to support label state-
ments on pesticide disposal and storage.
A recent survey (2) of manufacturers'
labels revealed that land disposal was
the most frequently recommended method for
the disposal of excess pesticides.
Several other studies (3-6) conclud-
ed that land disposal could conceivably
be an effective method for the disposal
of at least some of the individual pesti-
cides currently in use. Actual evidence
in this respect, especially field studies,
is scarce o- non-existent.
DATA REQUIREMENTS AND TESTING
General
Land disposal of waste pesticides
encompasses all methods by which waste is
deposited in or on land, including burial,
soil injection, soil mounds, soil pits,
infiltration/evaporation basins, soil in
277
-------
corporation, etc. The waste pesticides
that might be disposed in this manner are
primarily those generated in agricultural
operations, including obsolete and ex-
pired pesticides, unused spray mixes,
rinses, etc. Practically all research
has been conducted with use strengths and
dilute solutions of pesticides. The ef-
fectiveness of each method depends on the
nature and amount of pesticide disposed
and the soil and environmental conditions
prevailing at the disposal site.
The tests proposed below must be
conducted for all land disposal methods,
regardless of their nature, although spe-
cific methods might require additional
tests which are not outlined herein.
Since the disposal of pesticides in the
soil has features in common with their
actual use, some of the testing require-
ments will duplicate those for use, found
in Uo CFR Subpart D, Environmental Chem-
istry, 163.62-1 to 163.62-11.
Soil Studies
Studies of the physical properties
of the soil (7-11) must be conducted to
determine relative particle size, clay,
silt, sand, mineral and organic composi-
tion, salinity, and acidity, all of which
affect the absorption and adsorption
characteristics and the permeability of
the soil to pesticides. The geology of
the area must be investigated to deter-
mine the location of groundwater and the
presence and position of impervious
strata. A topographical survey may also
be required to determine surface features,
flood and erosion potential, and the oper-
ation of wells and other structures.
Soils to be tested must include the spe-
cific soil of-the disposal site, if known,
soils typical of the area where the pes-
ticide is to be disposed, or a standard
soil sample of known properties.
Physico-chemical Degradation
Pesticide molecules are altered in
the soil by physical and chemical pro-
cesses. The amount of degradation of
pesticides by hydrolytic (12-13) and
photolytic (lU-18) processes must be de-
termined, and the decomposition products
identified. Studies must include the
concentrated pesticide.
Soi]-Pesticide Interactions
Small scale studies must be con-
ducted to determine the effect of the
soil on the pesticide. The physical com-
position of the soil and the physical
properties of the pesticide will deter-
mine the adsorption (19-25),volatiliza-
tion (26-29), leaching (12,23,30-32) and
water dispersal of the pesticide, which,
in turn, determine the mobility of the
pesticide in soil. Even pesticides of
the same chemical family have widely dif-
ferent soil retention characteristics.
These factors will also be affected by
the climatic characteristics of the area,
so that the tests should be conducted at
as close to natural environmental condi-
tions prevailing at the time of disposal
as possible. Tests will be conducted with
the commercial pesticide formulations, use
strength and dilute solutions.
Metabolism
When a pesticide is applied to a
soil , it immediately has an effect on the
living organisms of the soil and may be
affected by them. Accordingly, determina-
tion of the effects of the microorganisms
on pesticides is essential (33-38). The
pesticide may be available for metabolism
under aerobic or anaerobic conditions, and
it may have an effect on the microbial
population of the soil (31,^9-53). Pes-
ticides such as fungicides and nematicides
are aimed at soil populations; the effect
on nontarget organisms must be determined
(37,5^-62).
Effects on Higher Plants and Animals
The effects of the pesticide on the
higher plants and animals living in the
disposal area, or that could be expected
to come into contact with the pesticide
must be determined. The effects of her-
bicides on plant life and of insecticides
on insects are known, but each class of
pesticides has the potential for affecting
nontarget organisms. These organisms will
include plants, amphibians, birds and
mammals.
The tests must include investigations
of accumulation and acute and chronic tox-
icity and, for mammals, investigations of
the possible mutagenicity, teratogenicity
and carcinogenicity. All these tests are
278
-------
given in Uo CFR 163-70-1 to 163.72-6.
DISCUSSION
A reasonable evaluation of the effec-
tiveness of pesticide disposal can be
made on the basis of the following cri-
teria: no leaching of pesticide into
ground or surface waters:, absence of air
pollution due to the volatilization of
pesticide; destruction of more than
99-999% of the pesticide within the use-
ful life of the disposal site; lack of
long-term alterations in the populations
of microorganisms and higher organisms
living on or near the site; rate of pes-
ticide leaching held at leas than 0.1
ft/year; and site location away from an
environmentally sensitive area.
Evaluation of pesticide detoxifica-
tion, however, is hampered by the fact
that the available information is over-
whelmingly based on research performed
with use strength and dilute solutions of
pesticides. Furthermore, since the opti-
mum degradation conditions for pesticides
in soils are largely unknown, more re-
search is required to determine the load-
ing rates and lifetime of a given dis-
posal site.
Moverover, disposal research has been
concentrated on single, pure compounds.
Accordingly, the land disposal of pesti-
cide mixtures, as well as the effect of
solvents, two conditions currently en-
countered in actual practice, are in need
of exhaustive scrutiny.
There is a definite lack of informa-
tion regarding the detoxification mecha-
nisms and ultimate fate of pesticides in
the soil, volatility, leaching, migration
and residual toxicity of pesticides at
disposal rather than use levels. The
tests and data requirements outlined
above are based primarily on studies
carried out with pesticides in use
strength, or 1-10 Ib/acre. Accordingly,
it is only reasonable to assume that the
results yielded by disposal levels, which
can be as high as 10,000 Ib/acre, or even
more, may vary to different degrees.
REFERENCES
1. Klee, A.K. 1980. Current Research on
Hazardous Waste Assessment and Con-
10.
trol. In: Treatment of Hazardous
Waste. Proceedings of the Sixth
Annual Research Symposium, D. Schultz
ed. EPA-600/9-80-011, U.S. Environ-
mental Protection Agency, Cincinnati,
OH. pp. 1-7-
Brambley, J., and D. Kollias. 1980.
In: Treatment of Hazardous Waste.
Proceedings of the Sixth Annual Re-
search Symposium, D. Schultz ed.
EPA-600/9-80-011, U.S. Environmental
Protection Agency, Cincinnati, OH.
pp. 101-111.
Lawless, E.W., T.L. Ferguson, and A.
F. Meiners. 1975- Guidelines for the
Disposal of Small Quantities of Pes-
ticides. EPA-670/2-75-057, U.S. En-
vironmental Protection Agency,
Cincinnati, OH. 331 pp.
Sanborn, J.R., B.M. Francis, and R.
L. Metcalf. 1977. The Degradation of
Selected Pesticides in Soil: A Re-
view of the Published Literature.
EPA-600/9-77-022, U.S. Environmental
Protection Agency, Cincinnati, OH.
617 pp.
Marsh, J.R., H.-T. Phung, and D.E.
Ross. 1978. Disposal of Dilute Pes-
ticide Solutions. SCS Engineers,
Long Beach, CA. Contract No. 68-01-
^729, U.S. Environmental Protection
Agency, Washington, D.C. 109 PP-
Shin, C.C., and D.F. Dal Porto. 1975-
Handbook for Pesticide Disposal by
Common Chemical Methods. Final Re-
port , EPA-530/SW-112C, U.S. Environ-
mental Protection Agency, Washington,
D.C. 109 pp.
Allison, L.E. 1975- Organic Carbon.
In: Methods of Soil Analysis, C.A.
Black, D.D. Evans, J.L. White, L.E.
Ensminger, and F.E. Clark (eds).
Amer. Soc. Agron. Publ. No. 9•
Madison, WI. pp. 1367-1378.
Broadbent, F.E. 1965. Organic Matter.
In: op. cit. pp. 1397-1^00.
Chapman, H.D. 1965- Cation-Exchange
Capacity. In: op. cit. pp. 891-901.
Day, P.R. 1965. Particle Fractiona-
tion and Particle-Size Analysis. In:
279
-------
12.
13.
lU.
15.
16.
op cit. pp. 5^5-567-
Peech , M. 1965- Hydrogen-Ion Activity.
In: op. oit. pp. 91^-926.
Krzeminski, S.F., C.K. Brackett , and
J.D. Fisher. 1975- Fate of Microbial
3-Isothiazolone Compounds in the En-
vironment : Modes and Rates of Dis-
sipation. J. Agr. Food Chem. 23:10-
60-1068.
Gomaa, H.M. , I.H. Suf f et , and S.D.
Faust. 1969. Kinetics of Hydrolysis
of Diazinon and Diazcxon. Residue
Rev. 29: 1T1-180.
Koshy, K.T., A.R. Friedman, A.L.
VanDerSlik, and D.R. Graber. 1975.
Photolysis of Benzoic Acid 2-(2,U,
6-Trichlorophenyl )-hydrazide . J. Agr.
Food Chem. 23:108l4-1088.
Nakagawa, M. , and D.G. Crosby. 197^-
Photodecoir.position of Nitrofen. J.
Food Chem. 22:8^9-853.
Nilles, G.?., and M.J. Zabik. 1975.
Photochemistry of Bioactive Compounds.
Multiphase Photodegradation and Mass
Spectral Analysis of Basagran. J. Agr.
Food Chem. 23:^10-1*15.
17. Hewsom, H.C., and W.G. Woods. 1973.
Photolysis of the Herbicide Dinitra-
mine (N-,H3-Diethyl-2,U-dinitro-6-
trifluoromethyl-m-phenylenediamine).
J. Agr. Food Chem. 21:598-601.
18. Su, G.C.C., and M.J. Zabik. 1972.
Photochemistry of Bioactive Compounds.
Photolysis of Arylamidine Derivatives
in Water. J. Agr. Food Chem. 20:320-
325-
19. Aharonson, N., and U. Kafkafi. 1975-
Adsorption, Mobility and Persistence
of Thiabendazole and Methyl 2-Benzi-
midazolecarbamate in Soils. J. Agr.
Food Chem. 23:720-72U.
20. Dao, T.H., and T.L. Lavy. 1975- Ex-
traction of Soil Solution Using a
Simple Centrifugation Method for
Pesticide Adsorption-Desorption
Studies. Soil Sci. Soc. Am. Proc.
1+2:375-377-
21. Harvey, R.G. 1971*. Soil Adsorption
22.
23.
2k.
25.
26.
27.
28.
29.
30.
31.
and Volatility of Dinitroaniline
Herbicides. Weed Sci. 22:120-121*.
Murray, D.S., P.W. Santelmann and
J.M. Davidson. 1975- Comparative
Adsoption, Desorption, and Mobility
of Dipropetryn and Prometryn in Soil.
J. Food Agr. Chem. 23:578-582.
Rao, P.C.S., and J.M. Davidson.
1979- Adsorption and Movement of
Selected Pesticides at High Con-
centrations in Soils. Water Res.
13:375-380.
Saltzman, S., L. Kliger and B. Yaron .
1972. Adsorption Desorption of Para-
thion as Affected by Soil Organic
Matter. J. Agr. Food Chem. 20:1221*-
1226.
Wu, C.H., N. Buehring, J.M. Davidson,
and P.W. Santelmann. 1975- Napropa-
mide Adsorption, Desorption, and
Movement in Soils. Weed Sci. 23:1*5^-
Harper, L.A., A.W. White, Jr., R.R.
Bruce, A.W. Thomas, and R.A. Leonard.
1976. Soil and Microclimate Effects
on Trifluralin Volatilization. J.
Environ. Qual . 5:236-21*2.
Jackson, M.D., and R.G. Lewis. 1978.
Volatilization of Two Methyl Para-
thion Formulations from Treated
Fields. Bull. Environ. Cont . Tox.
20:793-796.
Parmele, J.H. , E.R. Lemon, and A.W.
Taylor. 1972. Micrometeorological
Measurement of Pesticide Vapor Flux
from Bare Soil and Corn under Field
Conditions. Water, Air, and Soil
Pollut. 1:1*33-1*51.
Soderquist, C.J., D.G. Crosby, K.W.
Moilanen, and J.E. Woodrow. 1975-
Occurrence of Trifluralin and Its
Photoproducts in Air. J. Agr. Food
Chem. 23:30l*-309.
Helling, C.S. 1971- Pesticide Mobi-
lity in Soils I. Parameters of Soil
Thin-Layer Chromatography. Soil
Sci. Soc. Amer. Proc. 35:732-737.
Lichtenstein, E.P. , T.T. Liang, and
T.W. Fuhremann. 1978. A Compart-
280
-------
mentalized Microcosm for Studying the
Fate of Chemicals in the Environment.
J. Agr. Food Chem.. 26:9^8-953.
32. Lichtenstein, E.P., K.P. Schultz, and
T.W. Fuhremann. 1972. Movement and
Fate of Dyfonate in Soils under
Leaching and Nonleaching Conditions.
J. Agr. Food Chem. 20:831-838.
33. Benezet, H.J., and F. Matsumura. 197**.
Factors Influencing the Metabolism of
Mexacarbate by Microorganisms. J. Agr.
Food Chem. 22:1*27-1*30.
3k. Furukawa, K., and F. Matsumura. 1976.
Microbial Metabolism of Polychlorinat-
ed Biphenyls. Studies on the Relative
Degradability of Polychlorinated Bi-
phenyl Components by Alkaligenes sp.
J. Agr. Food Chem. 2k=251-256.
35- Ou, L.-T., D.F. Rothwell, W.B.
Wheeler, and J.M. Davidson. 1978. The
Effect of High 2,1*-D Concentrations
on Degradation and Carbon Dioxide
Evolution in Soils. J. Environ. Qual.
7:2Ul-2l*6.
36. Sikka, H.C., and J. Saxena. 1973.
Metabolism of Endothall by Aquatic
Organisms. J. Agr. Food Chem. 21:1*02-
1*06.
37. Stojanovic, B.J., M.V. Kennedy, and
F.L. Shuman, Jr. 1972. Edaphic Aspects
of the Disposal of Unused Pesticide
Wastes and Pesticide Containers. J.
Environ. Qual. 1:5^-62.
38. Van Alfen, N.K., and T. Kosuge. 197**.
Microbial Metabolism of the Fungicide
2,6-Dichloro-l*-nitroaniline. J. Agr.
Food Chem. 22:221-221*.
39. Bartha, R., and D. Pramer. 1965.
Features of Flask and Method for Mea-
suring the Persistence and Biological
Effects of Pesticides in Soil. Soil
Sci. 100:68-70.
1*0. Betts».P.M., C.W. Griddings, and J.R.
Fleeker. 1976. Degradation of U-Amino-
pyridine in Soil. J. Agr. Food Chem.
21* :571-57l*.
111. Fleeker, J.R., H.M. Lacy, I.R. Schultz
and E.C. Houkom. 1971*. Persistence
and Metabolism of Thiophanate-methyl
in Soil. J. Agr. Food Chem. 23:592-
595-
1*2. Gingerich, L.L., and R.L. Zimdahl.
1976. Soil Persistence of Isopro-
palin and Oryzalin. Weed Sci. 2h:
1*31-1*31*.
1*3. Gowda, J.K.S., and N. Sethunathan.
1976. Persistence of Endrin in
Indian Rice Soils under Flooded Con-
ditions. J. Agr. Food Chem. 2l* = 750-
753.
1*1*. Kazano, J., P.C. Kearney, and D.D.
Kaufman. 1972. Metabolism of Methyl-
carbamate Insecticides in Soils. J.
Agr. Food Chem. 20:975-979.
1*5. Marvel, J.T., B.B. Brightwell, J.M.
Malik, M.L. Sutherland, and M.L.
Rueppel. 1978. A Simple Apparatus
and Quantitative Method for Determi-
ning the Persistence of Pesticides
in Soils. J. Agr. Food Chem. 26:1116-
1120.
1*6. Messersmith, C.G., O.C. Burnside,
and T.L. Lavy. 1971. Biological and
Nonbiological Dissipation of Tri-
fluralin from Soil. Weed Sci. 19:
285-290.
1*7. Parr, F.F. , and S. Smith. 1969. A
Multipurpose Manifold Assembly: Use
in Evaluating Microbiological Effects
of Pesticides. Soil Sci. 107:271-276.
1*8. Parr, J.F., and S. Smith. 1973. De-
gradation of Trifluralin under Labo-
atory Conditions and Soil Anaerobio-
sis. Soil Sci. 115:55-63.
1*9. Schooley, D.A. , K.N. Cresswell, L.E.
Starger, and G.B. Quislad. 1975.
Environmental Degradation of the
Insect Growth Regulator Isopropyl
(2E, l*-E)-ll-methoxy-3,7,ll-tri-
methyl-2,l*-dodecadienoate (Metho-
prene). IV. Soil Metabolism. J. Agr.
Food Chem. 23:369-373.
50. Smith, R.A., W.S. Belles, K.W. Shen,
and W.G. Woods^ 1973. The Degrada-
tion of (N3, N -Diethyl 2,l*-dinitro-
6-trifluoromethyl-m-phenylenediamine)
in Soil. Pest. Biochem. Physiol.
3:278-288.
281
-------
51. Stevenson, P.J. 1965- Gross Chemical
Fractionation of Organic Matter. In:
Methods of Soil Analysis. C.A. Black,
D.D. Evans, J.L. White, L.E.
Ensminger, and F.E. Clark (ed). Amer.
Soc. Agron. Publ. No. 9- Madison, WI.
pp. 11+09-1^21.
52. Anderson, J.R. 1973. A System for
Evaluating Effects of Pesticides on
Soil Microbial Activity. Bull. Ecol.
Res. Comm. 17:^70-^72. (How: Ecol.
Bull., Swed. Nat. Res. Council).
53. Balandreau, J., and Y. Dommergues.
1973. Assaying Nitrogenase (09^2)
Activity in the field. Bull. Ecol.
Res. Comm. 17:2l+7-251*. (Now: Ecol.
Bull., Swed. Nat. Res. Council).
5U. Bohool,B.B., and E. L. Schmidt. 1973.
A Fluorescent Antibody Technique
Determination of Growth Rates of
Bacteria in Soil. Bull. Ecol. Res.
Comm. 17:336-338. (Now: Ecol. Bull.
Swed. Nat. Res. Council).
55. Casida, L.E., Jr. 1968. Methods for
the Isolation and Estimation of Soil
Bacteria. In: The Ecology of Soil
Bacteria. T.R.G. Gray and P.
Parkinsen (eds). University of Toronto
Press, Toronto, Canada, pp. 97-122.
56. Curl, E.A., and R. Rodriguez-Kabana.
1971• Microbial Interactions. In:
Research Methods in Weed Science, R.E.
Wilkinson ed. S. Weed. Sci. Soc. PCP
Enterprises, Inc. Atlanta, GA. pp.
163-19^.
57. Gray, T.R.G., P. Baxby, I.E. Hill,
and M. Goodfellov. 1969- Direct Obser-
vations of Bacteria in Soil. In: The
Ecology of Soil Bacteria, T.R.G. Gray
and J. Parkinsen (eds). University of
Toronto Press, Toronto, Canada.
58. Jensen, V. 1969. The Plate Count Tech-
nique. In: The Ecology of Soil Bacte-
ria, T.R.G. Gray and P. Parkinsen
(eds). University of Toronto Press,
Toronto, Canada.
59. Knowles, R.R., R. Brouzes, and P.O.
Toole. 1973. Kinetics of Nitrogen
Fixation and Acetylene Reduction, and
Effects of Oxygen and of Acetylene on
These Processes, in a Soil System.
Bull. Ecol. Res. Comm. 17:255-262.
(Now: Ecol. Bull., Swed. Nat. Res.
Council).
60. Matulewich, V.A., P.F. Strom, and M.
S. Finstein. 1975- Length of Incu-
bation for Enumerating Nitrifying
Bacteria in Various Environments.
App. Microbiol. 29:265-268.
6l. Ou, L.-T, D.F. Rothwell, W.B.
Wheeler, and J.M. Davidson. 1978.
The Effect of High 2,U-D Concentra-
tions on Degradation and Carbon Dio-
xide Evolution in Soils. J. Environ.
Qual. 7:21+1-21+6.
62. Pochon, J., and P. Tardieux. 1968.
Enrichment Techniques and the Elec-
tive Cultivation of Bacteria. In:
The Ecology of Soil Bacteria, T.R.G.
Gray and D. Parkinsen (eds). Univer-
sity of Toronto Press, Toronto,
Canada, pp. 123-137-
282
-------
PRELIMINARY STUDIES EVALUATING COMPOSTING AS A MEANS
FOR PESTICIDE DISPOSAL
Donald E. Mullins, Department of Entomology
Julie A. Petruska, Agricultural Engineering
Robert W. Nicholson, Environmental Sciences and Engineering
Eldridge R. Collins, Jr., Agricultural Engineering
Rodney W. Young, Biochemistry and Nutrition
Virginia Polytechnic Institute and State University
Blacksburg, Virginia 24061
ABSTRACT
A benchtop composter has been designed and constructed to provide laboratory data on the
fate of hazardous wastes subjected to composting conditions. Preliminary studies using
unlabeled pesticides have indicated this process might be quite useful in disposing of
certain hazardous materials. For example, 79% of Diazinon was lost when composted in a
150 g mixture of dairy cattle manure and sawdust for 10 days and 68% chlordane was lost
in 16 days when composted in 10 g composting media.
Introduction
A major problem often associated with
pesticide usage is the suitable disposal
of unused concentrated and dilute formu-
lations of pesticides and pesticide-
contaminated products. Since pesticides
are used extensively in U.S. agriculture,
there is need for abatement and control
of pesticide buildup in the environment.
Various methods of hazardous waste disposal
currently being researched include incine-
ration, biological treatment, physical-
chemical conversion and land disposal
methods (Wilkinson et al (12)). Some recent
reports on biological disposal methods
using disposal pits (Johnson and Baker (6),
Egg et al (3)) and agricultural waste
lagoons (Collins et al (1)) indicate these
processes might be useful under certain
conditions. However, information regarding
the effectiveness of biological methods on
a full scale basis is insufficient. There
are several pesticide classes which have
not been studied to determine their
potential susceptibility to biological
disposal (Wilkinson et al (12)).
It is conceivable that a technology
involving microbial degradation could
prove to be an effective, low cost means
for disposing of unwanted pesticide
materials and residues. Biological
degradation of hazardous materials is not
a new concept. It is well known that given
time and appropriate conditions, most
organic compounds will be metabolized by
living systems. Microbes have assumed
ever-increasing importance in studies
concerned with pesticide degradation; and
systems capable of supporting large
microbial populations such as sod.1, water,
and sewage, have long been thought to be
ideal sites for pesticide degradation
(Johnsen (4)). Unfortunately, the
volatilization rate and extent of pesticide
degradation for complex biological systems
is not well understood.
It has been pointed out that compost-
ing of pesticide wastes might be quite
effective, but has not been studied in
sufficient detail (Wilkinson, et al (12)).
A major problem with research on composting
of pesticides involves technical difficul-
ties associated with analysis and identi-
fication of the degradation products. As
283
-------
a consequence, the environmental impact
involving subsequent disposal of poten-
tially toxic residual materials produced
from composting is unknown.
Although composting is an ancient
practice, it is increasing in importance
as a means for solving some of our munic-
ipal waste problems (Poincelot (9); Parr
et al (8)). The development of large scale
sewage sludge composting technology was
facilitated by using bench model composters
in which composting conditions could be
controlled and studied in detail.
Pioneering research on the effect of
the compost process relating to pesticide
disposal was done in the late 1960's using
four pesticides (Diazinon, D^T, dieldrin
and parathion) mixed with rice hulls and
fruit and vegetable wastes (Rose and Mercer
(10)). Although jow concentrations of
pesticides were used, significant loss of
the parent materials was observed in about
50 days. Since that time, other studies
concerned w:th hazardous substance degra-
dation have been conducted using animal
wastes, and sewage sludge alone or mixed
with soil (Johnsen et al (5), Doyle et al
(2); Tucker et al (11)) .
Specific Objectives
The purpose of this research project
is to develop detailed information regard-
ing the fate of hazardous materials under
composting conditions. The specific
objectives include: 1) to develop a pro-
cedure to evaluate pesticide degradation
in bench top composters, 2) to study degra-
dation of selected hazardous materials
under controlled conditions, to determine
not only the disappearance of the parent
materials but also the formation of major
degradation products, and 3) by means of
laboratory and field studies, to evaluate
the possibility of using composting as a
method for disposing of small quantities
of pesticides (less than 2 kg or less than
20 liters).
Approach
Presently, the major thrust of this
project is centered around the development
of benchtop composters, and experimental
and analytical methods designed to examine
the effectiveness of microbes in composting
media in degrading hazardous materials.
Techniques are now available which may
allow for more rapid and thorough studies
of the composting process on hazardous
waste degradation. These are: 1) the use
of benchtop composters which allow control
of composting conditions on a small scale
(maintenance of high temperatures, moisture
and optimal air flow), and 2) the use of
radioisotope methodology which allows
measurement of small amounts of materials
including virtually all carbon-containing
compounds (primary, secondary, tertiary
metabolites and 002)• We are in the
process of applying these combined tech-
niques, representing a new approach for
examining the fate of pesticides in compost.
The requirement for only small amounts of
radiolabeled pesticide materials and the
potential for quantitative recovery and
identification of metabolites and end
products may make this approach quite use-
ful in facilitating the development of
this aspect of compost technology.
As indicated above, the experimental
methodology to be used in these studies
calls for the use of benchtop composters
and laboratory procedures which will
provide as complete information on the
degradation process as possible. The
general scheme of the composting procedure
is illustrated in Figure 1. Agricultural
wastes such as dairy cattle manure will
be mixed with sawdust along with specific
^C-labeled compounds prior to placement
CCWOST MIXTURE CONTAINING A SPECIFIC
I4C MATERIAL
[COMPOSTING PROCESS)
(SCRUBBING)
T 1
CO, ORGANIC VOLATILES
SOLVENT EXTRACTION
SOLUBLE MATERIALS
RESIDUAL MATERIALS
BOUND RESIDUES COMBUSTION THIN LAYER
CHROMATOGRAPHY, ETC.
ALKALINE EXTRACTION
ALKALINE SOLUBLE RESIDUAL MATERIALS
MATERIALS (COMBUSTION)
Figure 1. Compost Analysis Procedure
284
-------
into the composting apparatus. During
composting, 14C02 and -^C-volatiles will
be collected, and after incubation, the
composted mixtures will be exhaustively
extracted using conventional extraction
techniques. Where feasible, identification
or characterization of the reaction pred-
ucts will be made employing thin-layer, gas
and liquid chromatographic analytical
techniques. The following discussion out-
lines progress made in fulfilling the
objectives of this research.
Results
Benchtop Composter
A series of benchtop composters modi-
fied from those described by Willson (13)
have been assembled into one unit. Each
unit consists of 6 sealed glass vials which
act as the compost incubation chambers.
Each vial is partially immersed in water
contained in a 4 liter vessel. The tem-
perature in the vessel is controlled by a
circulating water bath which distributes
heated water through Tygon tubing encir-
cling the vessel (Figure 2). The assembly
is housed in a specially constructed styro-
foam container for insulation purposes.
The individual compost incubation
chamber consists of a glass vial (50 ml
capacity) with a stainless steel grid (16
mesh) held 1 cm above the bottom of the
vial (Figure 3). Air is drawn into the
bottom of the vial through the side arm,
passes through the compost, and exits
through the port at the top of the vial.
The air then goes to volatile gas scrubbers
which are polyurethane plugs for trapping
organic volatiles and KOH scrubbers for
trapping C02-
The system is designed in such a
manner that up to 6 vials may be run at
one time under constant temperature
AIR EXHAUST
r
POLYURETHANE
FOAM PLUG " <-
STAINLESS -"-~L-
STEEL SCREEN
1 1
^
7— J
)
-^
I
J
AIR INLET
Figure 3. Compost Vial
conditions. The intake air is humidified
and heated to the same temperature as the
vial to reduce the effects of cooling or
dehydration of the composting media.
The experimental conditions for
composting in the benchtop composting
apparatus are as follows: Equivalent
(w/w) dairy cattle manure is mixed with
sawdust (8-18 mesh) to achieve a moisture
content of 60-65%. A 5 to 150 gram
aliquot of this mixture is placed into
the composting vial. The initial compost-
ing temperature is maintained at 35°C for
24 hours and is raised about 5°C each day
until a final incubation temperature of
about 65°C is reached. During the incu-
bation, heated humidified air is drawn
through the vial and compost for 1 minute
at 10 minute intervals (50 ml/min).
The rate of C02 released from the
incubation mixture is determined by
titration of the KOH scrubbing solution
(containing trapped €62) to a phenol-
phthalein end point, and can be used as
an indicator of microbial activity.
Figure 4 compares data of C02 released from
5, 10, and 15 gram compost mixtures.
EXHAUST PORT
(TO SCRUBBERS)
Figure 2.
TYGON TUBING—-*
Benchtop Composter Unit
INSULATING MATERIAL
285
-------
tt: 30.0
o
I
I- 24.0
in
o
Q.
§,8.0
a
QL
O
U
in
U
5 GRRM SflMPLE
10 GRRM SBMPLE
IS GRHM SBMPLE
Figure A.
0 60 120 ISO 240
TIME ELRPSED ( HOURS )
Comparison of CC>2 Release Rates
from Compost.
Composting of Non-Radiolabeled Pesticides
Preliminary experiments designed to
examine non-radiolabeled pesticide loss
from compost were conducted using Diazinon
and chlordane. The results of two of these
experiments are presented here. Figure 5
illustrates results when about 100 ppm
Diazinon was incubated in a 150 g sample
of compost material for 10 days. It can
be seen that the Diazinon level is reduced
to 21 ppm after 10 days. Included in this
figure is the rate of CC>2 released during
the experiment. The rate decreased from
about 1A VJTTiole/g/h (AO h) to A ymole/g/h.
Figure 6 illustrates the loss of chlordane
(80 ppm) from 10 g compost material incu-
bated for 16 days. The final concentration
of chlordane was 30 ppm representing a loss
of 68% of the parent compound. The rate
of C02 release decreased from 21 umole/g/h
(2A h) to about 2.A ymole/g/h (382 h).
Evaluation of Losses "ue to Abiotic Factors
Since some of the losses observed
might have been due to volatilization or
chemical degradation, studies of the
effects of abiotic factors such as tempera-
ture and humidity on pesticide loss have
also been conducted in our laboratory. A
series of experiments using Diazinon
adsorbed onto glass or organic media
(sawdust, cellulose), contained in sealed
or open vials, in dry or humid environ-
ments at various temperatures have been
performed. Table 1 presents data obtained
when Diazinon was adsorbed onto a glass
surface, allowed an avenue of escape (open
versus closed vials) and incubated at
various temperatures for 36 hours. The
humidity was not controlled. It can be
seen that the Diazinon level was stable
in sealed or open vials when held at 0 or
2A°C. However, at 65°C Diazinon loss from
the glass surface was high in the open vial
(2% recovery) and to a lesser extent in
the sealed vial (7A% recovery). Since the
humidity was uncontrolled, and loss was
0 60 120 160 240
ELHPSED TIME ( HOURS )
0 100 200 300 400
ELflPSED TIME I HOURS )
Figure 5. Diazinon in Composting Dairy
Cattle T.'aste
Figure b. Chlordane in Composting Dairy
Cattle Waste
286
-------
TABLE 1. THE EFFECT OF TEMPERATURE ON DIAZINON LOSS
FROM A GLASS SURFACE IN OPEN AND SEALED VIALS*
Percent Recovery"
Time
elapsed
hours
12
24
36
0°
open
94
92
88
C
sealed
92
87
95
24°C
open
96
90
93
65°C
sealed
92
92
94
open
40
18
2
sealed
94
83
74
Open and sealed vials placed in a humid environment.
Initial concentration = 100 yg Diazinon.
observed in the sealed vial held at 65 C,
it was hypothesized that the combined
effects of temperature and humidity could
have produced the observed Diazinon loss.
Support for this hypothesis was obtained in
an experiment designed to examine the
effects of humidity on Diazinon adsorbed on-
to glass and incubated at 62 C for 36 hours.
The results are presented in Table 2. The
information presented in Tables 1 and 2
indicates that Diazinon loss from glass
surfaces can be quite significant when held
at temperatures above 62°C. These losses
may be due primarily to volatilization, but
in humid environments it is likely that some
loss may be due to degradation (i.e. chemi-
cal hydrolysis).
The process of adsorption onto and
absorption into organic media are involved
when pesticides are applied to compost
material. Therefore, the effects of temper-
ature and humidity on the loss of Diazinon
in the absence of significant microbial
activity wereexamined in order to evaluate
the Diazinon loss from organic media due
to abiotic factors. The results of one
such experiment are presented in Table 3.
It can be seen that increases in temperature
together with the presence of moisture
result in loss of Diazinon from organic
media.
The vials containing moisture (open
and sealed) showed the greatest amount of
Diazinon loss which suggests that the com-
bined effects of increased temperature and
humidity may result in Diazinon loss pri-
marily by degradation.
The effects of heat and humidity on
Diazinon in the composting apparatus with
and without aeration were studied in order
to further evaluate losses which might be
TABLE 2. THE EFFECT OF HUMIDITY ON DIAZINON LOSS FROM A GLASS SURFACE
IN OPEN AND SEALED VIALS INCUBATED AT 62°C FOR 36 HOURS
Replicate
Percent Recovery''
Open
Dessicated Env. Humid env.
Sealed
Dessicated env.
1 <5
2 <5
3 <5
<5
<5
<5
100
102
101
Initial concentration = 81 yg Diazinon
287
-------
TABLE 3. THE EFFECT OF TEMPERATURE AND HUMIDITY ON DIAZINON LOSS
FROM AN ORGANIC MEDIA IN OPEN AND SEALED VIALS INCUBATED FOR 36 HOURS
Treatment
25°C
Percent Recovery1
45°C
65°C
Sealed vial; desiccated env. 96+4
Open vial; desiccated env. 103+6
Sealed vial; humid env."1" 96+7
Open vial; humid env. 91+3
86+8
95+10
82+3
89+5
92+2
91+4
60+5
61+1
Diazinon applied to 1 g dry chromatographic grade cellulose.
Initial concentration = 82 ppm Diazinon, mean + standard deviation based on 3
replicates.
Water raturated cotton plug suspended in vial.
incurred due to abiotic factors. A series
of composting vials were set up in such a
manner that dry and wet organic media was
exposed to aerated (50 ml/min, 1 min. out
of 10) and non-aerated conditions for 36
hours. The results are shown in Table 4.
It can be seen that Diazinon recovery for
dry, aerated and non-aerated media was
higher than for the wet, aerated and non-
aerated. Furthermore, losses in the wet,
aerated media were greater than in the wet,
non-aerated media. The information from
this experiment indicates that humidity
does result in increased Diazinon loss and
that aeration in a humid environment results
in greater loss. Experiments designed to
further evaluate the effects of temperature,
volatility and hydrolysis will be conducted
using radiolabeled Diazinon.
TABLE 4. THE EFFECT OF AERATION AND HUMIDITY ON DIAZINON LOSS FROM
ORGANIC MEDIA INCUBATED AT 65°C IN BENCH TOP COMFOSTERS FOR 36 HOURS
Treatment*
Dry;
Dry;
Wet;
Wet;
aerated
non-aerated
aerated
non-aerated
Percent Recovery''
81+5
81+3
46+4
57+0
Dry treatment: 5 grams dry chromatographic grade cellulose, aerated with
desiccated air.
Wet treatment: 12.5 grams wet chrcmatographic grade cellulose (60% ^0), aerated
with humid air.
Non-aerated units were sealed.
Initial concentration: Dry treatment 92 ppm Diazinon
Wet treatment 96 ppm Diazinon
Mean + standard deviation based on 3 replicates.
288
-------
Studies on
Recovery
14,
Studies on recovery of CO 2 released
in the composting apparatus have been done
in order to determine the efficiency of
trapping this metabolite from composting
material. Carbon dioxide is trapped in 5
N KOH and subsamples are distilled into
ethanolamine and scintillation cocktail
prior to scintillation counting as described
by Mullins and Eaton (7). Results of these
studies are presented in Table 5. Recovery
of 1^C02 resulting from release, capture,
and subsequent distillation into the scin-
tillation fluid was quite good. The
presence of polyurethane plugs placed on
line in front of the KOH scrubbers (used
in trapping organic volatiles) does not
interfere with the -*-4C02 trapping efficien-
cy . It can also be seen from this table
that C02-saturated KOH (saturated during
the composting process) is also quite
efficient in trapping ^C02- However, we
have found that the scrubbing solutions
must be changed frequently enough (every
24 hours) during a compost experiment to
avoid over saturation which will result in
decreased trapping efficiency.
Trapping of ^C-Containing Volatiles on
Polyurethane Plugs
trapping efficiencies in KOH, a preliminary
study on trapping efficiencies of poly-
urethane plugs has been conducted. We have
found that when ^AC-chlordane is completely
volatilized off of a glass surface it is
trapped (100% recovery) in polyurethane
plugs contained in the compost vial.
Additional studies of the polyurethane
trapping efficiencies under composting
conditions are in progress.
Conclusions
A benchtop composting apparatus has
been designed and constructed which allows
for composting of small samples of organic
material. Preliminary studies indicate
that when two pesticides, Diazinon and
chlordane, are incorporated into composting
dairy cattle manure, the pesticide level
decreases. It does not appear that these
losses are due primarily to volatility.
However, humidity and temperature (abiotic
factors) may contribute to the losses of
the parent materials placed in a composting
environment. The use of radiolabeled pes-
ticide materials and the benchtop composting
apparatus should provide useful ii.formation
on the overall degradation process,
including the formation and release of C02
and other metabolites.
In addition to examination of C02
TABLE 5. RECOVERY OF 14C02 RELEASED FROM NA24C03
IN THE COMPOSTING APPARATUS
Treatment
Number of
Replicates
Percent Recovery
No compost present
polyurethane plugs
not on line
polyurethane plugs
on line
Compost present"
C02 - saturated KOH"1
unsaturated KOH
88.9 + 3.9 s.d.
96.5 + 1.0 s.d.
95.0 + 0.8 s.d.
88.0 + 1.0 s.d.
Initial activity Na214C03 = .23 yCi/replicate, scrubbed for 1 hour.
if
10 grams dairy cattle manure; sawdust mixture
1.35 mmoles C02 reacted with KOH
289
-------
References
1.
3.
4.
Collins, E. R., Jr., J. A. Petruska,
D. E. Mullins and R. W. Young. 1980.
Animal waste lagoon systems for dis-
posal of pesticide wastes. Paper No.
80-2112. Amer. Soc. Agri. Engin., St.
Joseph, Michigan. 22 pp.
Doyle, R. C., D. D. Kaufman and G. W.
Burt. 1978. Effect of dairy manure
and sewage sludge on 14-C pesticide
degradation in soil. J. Agr. Food
Chem. 26:987-989.
Egg, R., R. L. Reddell and R. Avant.
1980. Disposal of waste pesticides
in roofed and unroofed evaporation
pits. Paper No. 80-2111. Amer. Soc.
Agri. Engin. St. Joseph, Michigan.
15 pp.
Johnsen, R. E. 1976.
in microbial systems.
61:1-28.
DDT metabolism
Residue Rev.
11. Tucker, E. S., V. W. Saeger and 0.
Hicks. 1975. Activated sludge pri-
mary biodegradation of polychlorinated
biphenyls. Bull. Env. Cont. Toxicol.
14:705-713.
12. Wilkinson, R. R., G. C. Kelso, and
F. C. Hopkins. 1978. State of the
Art Report: Pesticide Disposal
Research. USEPA Document. EPA-6001/
2-78-183. August 1978. 225 pp.
13. Willson, G. B. 1971. Composting
dairy cow waste. Proc. Internat. Symp.
Livestock Wastes. Published by the
Amer. Soc. Agri. Engin., St. Joseph,
Michigan, pp. 163-165.
5. Johnsen, R. E., C. S. Lin, K. J.
Collyard. 1971. Influence of soil
amendments on the metabolism of DDT
in the soil. Proc. IUPAC 2nd Internat.
Congr. Pest. Chem., Israel 6:139-156.
6. Johnson, L. A. and J. L. Baker. 1980.
Water and pesticide volatilization from
pits used for disposal of dilute waste
pesticides. Paper No. 80-2110. Amer.
Soc. Agri. Engin. St. Joseph, Michigan.
27 pp.
7. Mullins, D. E. and J. L. Eaton. 1977.
Labeled ^C02 production in response
to light stimuli: a method and a pre-
liminary investigation. Comp . Biochem.
Physiol. 586:365-375.
8. Parr, J. F., E. Epstein and G. B.
Willson. 1978. Composting sewage
sludge for land application. Agri.
Environ. 4:123-137.
9. Poincelot, R. P. 1975. The Biochem-
istry and Methodology of Composting
Bull. 754. Connecticut Agr. Exp.
Station, New Haven, CT.
10. Rose, W. W. and W. A. Mercer. 1968.
Fate of insecticides in composted
agricultural wastes. Progress Report
Part 1. National Canners Association,
Washington, D.C. D-3049. 27 pp.
290
-------
TOP SEALING TO MINIMIZE LGACHATE GENERATION - STATUS REPORT
Grover H. Emrich and William W. Beck, Jr.
SMC-MARTIN
King of Prussia, PA 19406
ABSTRACT
Remedial actions consisting of regrading and the placement of a 20 mil
PVC membrane seal over a 10 hectare (25-acre) landfill was conducted
at Windham, CT Landfill during the fall of 1979. A bimonthly monitoring
program is being conducted to establish the effectiveness of the top
seal for mitigation of the existing ground-water pollution. Investi-
gations of the structural integrity of the top seal were conducted in
April and May of 1980. These investigations indicated a minimal
number of small punctures occurred in the top seal membrane. Spring
recharge was effectively intercepted by the membrane and diverted from
the landfill. Pan lysimeters installed below the membrane confirmed
the effectiveness of the membrane for intercepting recharge from
precipitation. Water samples from monitoring wells within the landfill
and surrounding it indicate that water quality is improving and the
plume of contamination is receding toward the landfill. It is concluded
that it is possible to place a 20 mil PVC top seal over a 10 hectare
(25 acre) landfill as a remedial measure to mitigate contamination of
ground water.
INTRODUCTION
The development and implementation
of a remedial action project de-
signed to abate pollution from a
landfill requires: 1) knowledge of
the potential remedial actions,
2) site selection and documentation
of the problems associated with the
landfill, 3) design and construc-
tion of the remedial actions, and
4) monitoring to determine the
effectiveness of the actions im-
plemented. Knowledge of potential
remedial actions was addressed and
resulted in the publication of
"Guidance Manual for Minimizing
Pollution from Waste Disposal
Sites" by A. L. Tolman, A. P.
Ballestero, W. W. Beck and G. H.
Emrich. The site selection process
for the landfill resulted from the
review of over 400 sites, field
inspection of 50 sites, a detailed
analysis of five sites, drilling
and geophysical work at two and
the final selection of the Town
of Windham, CT Landfill. The
Windham, CT Landfill located in
east central Connecticut, is
approximately 10 hectares
(25 acres) in size and consists
of two sections (see Figure 1).
The landfill is located in an
area of sand and gravel immedi-
ately adjacent to the City of
Willimantic water supply reser-
voir and a Corps of Engineers
flood control dam. The eastern
(old) half of the landfill is
approximately 4 hectare (10 acres)
with its lower portion in the
ground water. The western (new)
landfill is approximately 6 hec-
tare (15 acres) and is above the
ground water. Local sands and
gravels were used for cover
material. The landfill operated
291
-------
oVERNON
STORRS
o
O MANCHESTER
HARTFORD
GLASTONBURY
CONNECTICUT
TOWN OFWINDHAM
Figure 1. LocatJon of the Windham Landfill, Windham, Connecticut.
from approximately 1945 to 1978. A
series of monitoring points em-
placed through and around the
landfill has defined the area of
ground-water pollution and its
movement through time (see Fig-
ure 2).
Remedial actions proposed to abate
and prevent pollution from the
landfill included: 1) regrading of
the landfill to maximize surface
water runoff and minimize infiltra-
tion, 2) the placement of a 20-mil
PVC top seal, 3) covering the top
seal with approximately 18 inches
of final cover, and 4) revegetation
(see Figure 3). The remedial
actions were designed to be passive
to insure minimum future mainte-
nance. From the late summer to the
early winter of 1979, these remedi-
al actions were implemented at the
site and final revegetation was
accomplished in the spring of
1979.n '
Before the remedial actions began,
a detailed monitoring network was
established to develop informa-
tion pertinent to the reaction of
the landfill to precipitation
events prior to the remedial
actions and then after completion
of the remedial actions to evalu-
ate their effectiveness. This
monitoring network included:
1) the emplacement of suction and
pan lysimeters in the refuse
above the water table and in the
surrounding soils to determine
the reaction (infiltration) of
the landfill to precipitation
events, 2) the emplacement of
piezometers into the water table
below and around the landfill to
determine the quantity and qual-
ity of leachate being generated
and its movement through the
subsurface, 3) the establishment
of surface water sampling points,
4) the placement of suction
lysimeters to obtain leachate in
the refuse above the water table,
and 5) the implementation of a
bimonthly sampling program to
determine changes in the geometry
of the pollutional plume with time.
292
-------
O34
KEY
O 1-23,34-37-MONITORING WELLS
O 24-29-SUCTION LYSIMETERS
D 30-35- PAN LYSIMETERS
Figure 2. Location of monitoring points at the Windham Landfill,
Windham, Connecticut.
r-VEGETATED FINAL COVER
-PVC MEMBRANE SEAL
REGRADED LANDFILL
Figure 3. Typical section through the Windham Landfill.
293
-------
TOP SEAL INTEGRITY
The 20 mil PVC membrane seal with
solvent welded seams was placed
over the 10 hectare (25 acre)
Windham Landfill in the fall of
1979. Immediately beneath the
membrane are 10 cm (4 in) of com-
pacted fine grain sand washings.
The membrane was covered with an
average of 50 cm (18 in) of local
fine sand and gravel which was
spread in 15 cm (6 in) layers and
revegetated. In April 1980 an
investigation of selected areas was
conducted at the site to determine
the membrane's integrity. Seven
excavations (test pits) were made
by hand through the cover material
to the membrane in both the old and
the new sections of the landfill
(see Figure 4). It was determined
that the membrane maintained its
integrity during the placement of
the cover material. Although there
were numerous indentations up to
4 cm (1-1/2 in) deep in the mem-
brane, there was no evidence of
puncture. These indentations were
caused by the final cover material
which consisted of both sand and
gravel.
The thickness of the cover material
on the eastern (old) landfill
varied from 10 to 46 cm (8 to
18 in). On the western section of
the landfill (new) final cover
varied from 38 to 46 cm (15 to
18 in). During the placement of
the cover material over the seal,
truck traffic patterns had to be
established to assure a logical and
systematic completion of the pro-
ject. Several areas of high truck
traffic were selected for investi-
gation (see Figure 4). Test
Pits #1 and 12 were located in the
PONDS
CC
5
>
a:
UJ
CO
ui
en
OLD LANDFILL
O GAS VENTS
TEST PITS-4/80
D TEST PITS- 5/80
Figure 4. Location of test pits for Top Seal Integrity Tests.
294
-------
area of heavy truck traffic and it
was expected that these would be
the most likely areas for puncture
or tearing of the membrane. Test
Pit #5 was dug in an area in which
there was significant evidence on
the surface of heavy truck traffic.
The results of this investigation
showed no visible punctures in the
seal.
In the week previous to this in-
vestigation, there had been a
13 cm (5 in) rain over a five to
six hour period, which caused
local flooding. The rain eroded
sections of the cover material in
the swale between the two sec-
tions of the landfill as well as
on the northwest side of the new
landfill. Exposed areas were up
to 2 m (6 ft) wide and 9 m (30 ft)
long. Careful examination of
these indicated no loss of integ-
rity of the membrane.
In May 1980, the integrity of the
membrane was again inspected by
staff members from EPA and the
U. S. Corps of Engineers Waterways
Experimental Station, Vicksburg,
Mississippi. Three additional
test pits were dug on the two sec-
tions of the landfill. Test
Pit #1 was approximately at the
same location as Test Pit #1 dug
in April 1980 (see Figure 4). At
this pit there was 35 to 38 cm
(14 to 15 in) of cover material
and numerous indentations in the
top of the membrane. Test Pit #2
was located in the same approxi-
mate area as Test Pit #3 from the
April 1980 investigation. Cover
consisted of sand with some
gravel and was 23 cm (9 in)
thick. Test Pit #3 was dug on
the new landfill to the east of
Test Pit #6 dug in April 1980.
Cover material consisted of sand
with some gravel and was 30 cm
(12 in) thick.
Sections of the membrane seal
were removed for testing by the
Waterways Experimental Station
from Test Pits #2 and #3. The
membrane was patched using similar
material and solvent welds.
Several breaks in'the membrane
recovered from Test Pit #3 were
observed when it was held up to
the light with the maximum break
being one-quarter of an inch in
length. All of the breaks were
produced by punctures from the
top down. It was noted that
water was seeping into Test
Pit #3. This was from spring
recharge (rain) which moved down to
the membrane and then moved hor-
izontally along the membrane. The
membrane removed from Test Pit #2
contained no punctures. The mem-
brane could not be removed from
Test Pit #1 as it was covered by
water. The water was from 1 to
2 cm (0.5 to 0.75 in) deep result-
ing from obvious seepage of infil-
trated water down through the cover
material to the membrane with
movement along the membrane-cover
material interface. The membrane
was carefully observed in this test
pit for three hours. There was no
evidence of air bubbling through
the water which would indicate the
breaks in the membrane.
It was concluded from these studies
that even in areas of heaviest
truck traffic there were only a few
breaks in the membrane and these
breaks did not significantly reduce
the effectiveness of the membrane
to stop the movement of water in
the cover material. Therefore, it
was concluded that the majority of
the infiltration was moving down to
the membrane and then moved hori-
zontally across the surface of the
membrane discharging at the edges
of the refuse into a lined drainage
swale.
MONITORING
In the fall of 1978, a monitoring
program was implemented to deter-
mine the impact of the refuse on
the underlying ground water from
precipitation events. Suction
lysimeters were placed in the
refuse to determine the passage of
moisture (leachate) through the
landfill. In order to determine
the quantity of moisture moving
through the landfill, four pan
lysimeters were emplaced, three
with collection areas of 0.4 m
295
-------
(4 square feet) and the fourth with
a collection area of 2.3 m
(25 square feet) in order to pro-
vide cross correlation of data
between the smaller and larger
collection areas. Wells were
installed in and below the landfill
to further assess leachate quality
and its impact on water quality.
Wells were also located upgradi-
ent and downgradient from the
landfill to determine the movement
contamination from the landfill.
Adjacent ponds were sampled as
well as the adjacent water supply
reservoir to determine the effect
of the landfill on these bodies
of surface water (see Figure 2).
Between February and June 1979,
wells, ponds and lysimeters were
sampled on a monthly basis to
determine water quality. The
water levels in the'wells and
ponds were measured weekly and
the volume of water collected in
the pan lysimeters was measured.
This program indicated leachate
of moderate strength was generated
by the old fill whereas leachate
of somewhat higher strength was
generated by the newer portion of
the landfill. Leachate was found
to effect the ground water down-
gradient of the fill. There was
significant contrast between the
background water quality measured
upgradient and that of contami-
nated ground water found down-
gradient of the landfill. Anal-
ysis of the results of the weekly
monitoring of ground water and
surface water elevations and
daily precipitation data verified
significant infiltration was
taking place in to the landfill.
As much as 85 percent of the
water passing through the refuse
was derived from infiltration of
precipitation.
Beginning in the fall of 1979,
with the installation of the
seal, bimonthly sampling was
instituted at selected monitoring
points. Results of the sampling
showed that the impact of the
landfill on ground water has been
mitigated between November 1979
and November 1980. Visual obser-
vation shows that coloration in
the ponds adjacent to the land-
fill had improved significantly.
Analysis for water samples collec-
ted bimonthly had shown improve-
ment in the ground water as well
as retreat of the leachate plume.
In November 1979, specific conduc-
tivity (see Figure 5) indicated a
major plume moving westward toward
the reservoir as well as the sec-
ondary plume moving northward
towards Ponds 1, 2 and 3; By
November 1980, the secondary plume
had withdrawn toward the landfill
and the major westward plume had
been narrowed. In the summer
of 1980, four additional wells were
drilled on the western side of the
landfill to better define this
plume. Comparison of most specific
conductivity values in November
1979 with those of November 1980
have shown a decrease in the spe-
cific conductivity indicating a
decrease in the amount of leachate
generated.
The volume of water collected in
the pan lysimeters was measured
during the bimonthly sampling
rounds. The background pan ly-
simeter responded normally to
precipitation events. Little or no
water was collected from the pan
lysimeters below the membrane seal
indicating that the top seal is
effectively controlling infil-
tration into the landfill from
precipitation events.
CONCLUSIONS
A 20 mil PVC top seal can success-
fully be placed over a 10 hectare
(25 acre) landfill based on the
results of liner integrity tests.
The 20 mil PVC membrane maintained
its flexibility during the place-
ment of cover material which con-
tained coarse fragments. A series
of test pits excavated by hand to
the membrane showed that only a few
punctures were present in areas
which experienced high truck traf-
fic during placement of the cover
material. The few punctures that
were present were all from the top
down and were less than 2.5 cm
296
-------
O 1-23, 34-37 -MONITORING WELLS
O 24-29-SUCTION LYSIMETERS
D 30-33- PAN LYSIMETERS
———NOVEMBER 1979
NOVEMBER 1980
Figure 5. Specific conductivity in November 1979 and November 1980.
(1 in) length. The test pits
constructed to the top seal indi-
cated that the spring recharge was
intercepted by the membrane seal
and moved water laterally to the
line diversion ditches surrounding
the landfill. Pan lysimeters
placed below the top seal con-
firmed that little or no precipi-
tation moves through the top
seal. Specific conductivity
tests made of the lysimeters and
wells at the site show that the
plume of contamination resulting
from the landfill is receding
toward the landfill and concen-
trations in the vicinity of the
landfill are decreasing. Based
on the results of these investi-
gations, one year of monitoring
data, it is concluded that the
top seal is effectively control-
ling the infiltration of precipi-
tation to the landfill and signifi-
cantly reducing the contamination
effects of the site.
REFERENCES
1. Emrich, Grover H., Beck,
William W., Jr. and Tolman,
Andrews L. Top Sealing to
Minimize Leachate Genera-
tion, Case Study of the
Windham, CT Landfill.
297
-------
INVESTIGATION OF THE LIPARI LANDFILL USING GEOPHYSICAL TECHNIQUES
Joseph R. Kolmer
Woodward-Clyde Consultants
Plymouth Meeting, Pennsylvania 19^62
ABSTRACT
The implementation of a remedial program at a landfill site requires a good understanding of site
conditions as well as knowing the patterns of contaminant movement from the identified source
area. In order to acquire this basic information needed for remedial program design, a two phase
investigative program is recommended. The initial qualitative phase identifies the nature of the
problem, and the quantification phase provides the design data. This paper discusses the qualitative
investigation of the LiPari Landfill site in Gloucester County, New Jersey. This work was conducted
as part of the U.S. Environmental Protection Agency remedial action research program. Geophysi-
cal techniques were used extensively in this survey and provided information on locations of buried
metallic wastes within the landfill as well as the areal extent of contamination in the groundwater
system downgradient of the landfill site. The use of geophysics at the LiPari site has provided a
significant amount of data from which a quantification program for the site can be very effectively
formulated.
INTRODUCTION
In June of 1980, the U.S. Environmental
Protection Agency (US-EPA), Municipal Envi-
ronmental Research Laboratory, Cincinnati,
Ohio, awarded a research contract to
Woodward-Clyde Consultants (WCC) for the
preparation of a manual on how to select and
implement remedial action programs for haz-
ardous waste landfill sites. The methodology
and documentation for preparation of this man-
ual was to be developed by implementing a
remedial program at the LiPari Landfill site in
Gloucester County, New Jersey. The landfill
site investigation would be conducted in three
phases, the first of which being a qualitative
site assessment, and the second a quantitative
assessment from which a remedial program
would be selected and preliminarily designed.
The final phase would consist of implementa-
tion of the design remedial system. The total
term of the research project is four years.
This paper contains a discussion of the
first program phase, the qualitative investiga-
tion. Site history, existing site conditions and
documented pollution conditions resulting from
landfill discharge will be presented. Further, a
discussion of various geophysical investigation
techniques will be presented and their utility in
subsurface investigations at landfill sites will
be demonstrated. The information base for this
paper was obtained from the files of the US-
EPA in Cincinnati, Ohio, and Edison, New
Jersey. The files of the Gloucester County
Planning Department were also reviewed for
applicable information. Further, the described
geophysical surveys were conducted on site, and
the results of the survey were reduced and
interpreted by the WCC project staff.
BACKGROUND
The LiPari Landfill is located in Mantua
Township, Gloucester County, New Jersey (see
Figure 1). This landfill has been the subject of
strong controversy, court proceedings, and lim-
ited investigative studies since the early 1970s.
This activity stems from the fact that the
landfill has been sited as a major source of
surface water and groundwater pollution in the
Gloucester County area.
The total LiPari property is approximate-
ly 18 hectares in size and is detailed in Figure
2. The topography can be categorized as gently
rolling, except in the vicinity of the surface
drainage patterns where the slopes steepen.
Surface soils consist of sandy materials and,
based on a preliminary review of the site
298
-------
SCALE IN KILOMETERS
10
")
Figure 1. Regional Location Plan, LiPari Landfill Site, Pitman, New Jersey.
299
-------
conditions, it is believed that a large portion of
the precipitation infiltrates to the subsurface.
The major land use in the area is agricul-
tural. Relatively large apple and peach
orchards are planted throughout the surrounding
area, including the area of the LiPari property.
There is also a peach orchard planted directly
over the landfill. It was reported that this
orchard was planted by Mr. LiPari to test the
possible effects of methane gas and leachate on
the trees.
The landfill itself occupies 2A hectares of
the total 18 hectares. These 2A hectares are
divided into two flat lying areas of differing
elevation, separated by a steep slope. There is
no evidence of extensive erosion or develop-
ment of runoff flow channels over the landfill
area.
Two surface drainage systems flow near
the landfill, as shown in Figure 2. The main
drainage pattern, Chestnut Branch, has its
headwaters above the landfill area and, after
flowing past the site, discharges to Alcyon
Lake, approximately 370 meters downstream.
Rabbit Run, a small tributary to Chestnut
Branch, derives its headwater flow from a small
spring located adjacent to the landfill. As
shown in Figure 2, this stream flows along the
full length of the northwestern edge of the
landfill before it discharges to Chestnut
Branch.
DISPOSAL OPERATIONS
Mr. LiPari obtained the site property in
1958, and soon thereafter started a sand and
gravel quarry operation. It was reported that,
shortly after initiation of the quarry operation,
Mr. LiPari also started to accept waste materi-
als as backfill for the excavated portions of the
site. Apparently, the disposal operations fol-
lowed the quarrying pattern. The sand and
gravel were removed by cutting trenches, ap-
proximately 9 to 15 meters wide and 15 to 25
meters in length, and these trenches were back-
filled with waste materials. The depth of the
trenches is disputed, but reported values are in
the range of two to five meters. The nonsal-
able portions of the excavated sands and
gravels were used to cover the disposed wastes.
The wastes ranged from municipal type
refuse to industrial solids and liquids. No
detailed records of the quantity or types of
wastes were kept, but it has been estimated
that approximately 9,200 cubic meters of solids
were buried and approximately 11,000 cubic
meters of liquids were disposed. The liquids are
thought to be largely uncontained. It was
reported that drums of liquids brought to the
site were drained into the landfill and the
drums were reclaimed. Only the nonsalvage-
able drums were disposed. The volume of
liquids transferred to the site in bulk, if any, is
not known. Some of the reported wastes in-
clude cleaning solvents, paint thinners, paints,
dirty waste solvents, phenols or amines wastes
and residues, and resins or esther press cakes.
The only reported significant events dur-
ing the period of landfill operation were two
fires in 1969, and these fires were associated
with the drummed wastes. The first fire oc-
curred when a drum was being emptied into the
fill, and the second occurred when a bulldozer
struck a drum supposedly containing explosive
wastes. There were reportedly three drums of
explosive materials, and the disposition of the
remaining two is unknown. These remaining
drums are presumed to be in the landfill. Fol-
lowing these incidents, Mr. LiPari would not
accept liquids for disposal. The landfill opera-
tions were ceased in 1971.
GEOLOGIC AND GROUNDWATER
CONDITIONS
The LiPari Landfill is located within the
central Upland subprovince of the coastal plain
physiographic province. This region of Glouces-
ter County is underlain by u neon sol idated to
poorly consolidated late Cretaceous and
Tertiary age fluvial and marine deposits. The
surficial deposits in the site area consist of
upper Miocene (late Tertiary) age sands and
gravels belonging to the Cohansey sand forma-
tion. Lying unconformably beneath the
Cohansey are silty and clayey sands of the
middle Miocene age, Kirkwood Formation. The
thicknesses of the Cohansey Sand and Kirkwood
Formations, as reported in the literature, are
from 30 to 75 meters and W meters, respec-
tively.
Some information on the subsurface con-
ditions at the site was obtained from drilling
work conducted by others during 1979. Four
test borings were drilled, located as shown in
Figure 2. These borings were reportedly cased
and screened to a depth of six meters. The
completion methods, screen depths, etc., are
not known. Based upon the logs of these test
borings, a subsurface profile was constructed
and is shown in Figure 3. Strata 1 and 2 are
probably the Cohansey Sand and are approxi-
mately 8 to 11 meters thick. Stratum 3 is in
the Kirkwood and consists of two to five meters
300
-------
ALCYON LAKE
SUBURBAN DEVELOPMENT
APPROX. DIRECTION
OF GROUNDWATER FLOW
CHESTNUT BRANCH
APPROX. PROPERTY
BOUNDARY
SCALE IN METERS
Figure 2. LiPari Landfill Site Area, Pitman, New Jersey.
301
-------
o
to
50
40
OL
UJ
2
Z
<
>
35
30
NORTHWEST
B-l
/ V
APPROX. DISPOSAL
AREA
TANDY COVER
—
B"4
i CHEMICAL/MUNICIPAL/
I FILL /
SOUTHEAST
B-3
-APPROX. GROUND
SURFACE
0
ASSUMED DATUM
20
Figure 3. Subsurface Profile, LiPari Landfill Site, Pitman, New Jersey.
LEGEND^
(T) COHANSEY SAND FORMATION-ORANGE
TO GRAY, FINE TO COARSE SAND
WITH GRAVEL,StLT AND CLAYEY
LAYERS LOCALLY.
(?) COHANSEY SAND/KIRKWOOD FORM-
ATION - GRAY AND BROWN SILTY
FINE SAND.
(T) KIRKWOOD FORMATION - GRAY SILT,
TRACE FINE SAND AND CLAY.
©
4)KIRKWOOD FORMATION-GRAY SILTY
FINE SAND
(5) KIRKWOOD FORMATION - GRAY SILT,
TRACE FINE SAND
SCALE IN METERS
100
-------
of medium dense gray clayey silt and fine sand.
The lowermost deposits (Strata 4 and 5), en-
countered from depths of 15 to 20 meters,
consist of medium dense, gray, fine sandy silt/
silty fine sand. All four test borings terminated
within the Kirkwood Formation.
The preliminary geologic assessment of
the site indicated that near surface ground-
water would occur in the Cohansey Sand above
the Kirkwood Formation. The clayey silts in
the Kirkwood would tend to limit vertical
movement of the groundwater. The relation-
ship between horizontal and vertical flow is
dependent upon the relative hydrogeologic
characteristics of the Cohansey and Kirkwood
units and their interaction with the nature of
the chemical contaminants at the site. These
relationships would be detailed during the
follow-on quantitative assessment of the ex-
isting site conditions.
Groundwater level measurements taken
from the four existing wells at the site indi-
cated that the general direction of groundwater
flow is from the area of the landfill toward
Chestnut Branch, as indicated in Figure 2.
Since no aquifer characteristics were available
for the immediate site area, no estimate of
volumetric flow rates or seepage velocities
were made. Site inspection, however, did indi-
cate that the general direction of groundwater
flow was probably correct because wide areas
of contaminated groundwater seepage were
noted along the bank of Chestnut Branch in an
area that would be downgradient from the land-
fill site.
Based on the available information, it
appears that the groundwater flow is primarily
horizontal in the Cohansey Sands. These sands
appear to have a notably higher horizontal
hydraulic conductivity than the vertical con-
ductivity of the clayey silts underlying them.
Also, if the majority of the contaminant move-
ment was vertical, it is doubtful that the
contaminated groundwater seeps along Chest-
nut Branch would have developed in such an
extensive manner.
SITE INVESTIGATIONS
The quantification of groundwater, sur-
face water and contaminant conditions would
be done by installation of a series of monitoring
wells. These wells should be located so that a
maximum amount of information can be gath-
ered at each monitoring point. Further, the
wells drilled into the landfill area could possibly
encounter drums of hazardous and/or explosive
materials. It would be advantageous to have an
assessment of the lateral extent of groundwater
contamination and the location of buried waste
drums prior to installation of monitoring wells.
In order to obtain this desired information, a
series of geophysical surveys was conducted at
the site. The objectives of the surveys were to:
1. locate drums or pockets of waste filled
drums within the landfill area, and
2. preliminarily identify the areal extent of
contaminated groundwater leaving the
landfill site area.
Several geophysical methods were consid-
ered for locating the buried drums within the
landfill site. Most reports suggested that the
barrels were metal and, therefore, very likely
to produce magnetic anomalies as well as being
good reflectors of energy impulses. Consider-
ing these conditions, a metal finder (M-Scope),
a portable proton magnetometer, and a subsur-
face interface radar unit were selected as the
types of instruments best suited for the landfill
survey. In order to assess the areal extent of
contaminated groundwater, it was decided to
conduct a subsurface conductivity survey of the
site area. A brief description of each instru-
ment's theory of operation will be presented,
followed by a discussion of survey results.
Fisher M-Scope
The Fisher M-Scope pipe and cable finder
is scientifically designed to detect and accu-
rately locate buried metal pipes, conduits, and
miscellaneous metal objects. The equipment
detects the presence of metallic objects
through the use of an induced electromagnetic
field. The object is located and traced by
means of this induced field.
The instrument consists of two principal
component parts: a directional radio type
transmitter and a directional radio type re-
ceiver. The function of the transmitter is to
generate or induce in the buried metal object
the electromagnetic field referred to above.
The receiver locates the metal object by de-
tecting and tracing this electromagnetic field.
Proton Magnetometer
The proton precession magnetometer is so
named because it utilizes the precession of
spinning protons or nuclei of the hydrogen atom
in a sample of hydrocarbon fluid to measure the
total magnetic intensity. The spinning protons
behave as small spinning magnetic dipoles.
303
-------
These magnets are temporarily aligned or
polarized by application of a uniform magnetic
field generated by a current within the instru-
ment. When this current is removed, the spin
of the protons causes them to precess about the
direction of the ambient magnetic field, much
as a spinning top precesses about the gravity
field. The precessing protons generate a small
signal whose frequency is proportional to the
total magnetic field intensity.
The intensity of the magnetic field in a
specific location is dependent upon several fac-
tors, including the presence of metallic objects.
The depth of burial of the metals cannot be
assessed, but their presence will be reflected by
higher magnetic field intensities.
Subsurface Interface Radar
The subsurface interface radar (SIR) is an
advanced impulse radar system. The system's
transducer is moved along a line while the
recorder produces a continuous high resolution
profile of the subsurface. A unique feature of
the SIR system is its ability to automatically
generate a continuous map of subsurface fea-
tures, not a point-by-point sample from which
extrapolations must be made.
During operation, an electromagnetic im-
pulse is propagated into the subsurface soil
profile. When the impulse strikes an interface
between two materials of differing electrical
properties (i.e., a soil bedrock interface, a soil
water interface, a metal drum buried in the
subsurface, etc.), some of the transmitted im-
pulse energy is reflected and the remainder
continues on through the material to other
interfaces. The reflected signals are received
by the transducer and sent back to the system
control unit to be monitored on its oscillograph
and printed on a graphic recorder. The graphic
record of the signal wave is a close approxima-
tion of the interfaces one would see in a
vertical wall of a trench dug along the path of
the transducer. Metallic drums and pipes pro-
duce a unique signature pattern in the recorded
data. These signature patterns were used to
correlate the results of the subsurface inter-
face radar survey with the proton magnetome-
ter survey.
Conductivity Meter
The EM34-3 conductivity meter manufac-
tured by Geonics, Inc., is designed to measure
subsurface conductivity using an inductive elec-
tromagnetic technique. The instrument gives
direct readings in millimhos per meter, and
surveys are carried out by traversing the ground
in a grid pattern with two wire coil units. The
principle of operation is that the depth of
penetration is independent of terrain conductiv-
ity and is determined solely by the instrument
geometry, i.e., the intercoil spacing and coil
orientation. The instrument can be used at
fixed spacings of 10, 20 or ^0 meters. The
depth of penetration can be stepped from a
minimum of 7.5 meters to a maximum of 60.0
meters.
GEOPHYSICAL SURVEY INVESTIGATION
RESULTS
The geophysical surveys were conducted
in two areas: over the landfill itself, and
downgradient from the landfill toward the
groundwater discharge point along the bank of
Chestnut Branch. Figure 4 shows the areas
which were surveyed with each geophysical
technique and also indicates which figures show
the results of a particular survey. The follow-
ing paragraphs briefly describe the results
obtained with each geophysical technique.
Proton Magnetometer
The landfill site was staked to form a grid
having 7.6 meter centers. A magnetometer
survey was conducted on this grid spacing, and
a contour map portraying the various patterns
of magnetic intensity is shown as Figure 5.
The results of the magnetometer survey
indicate isolated areas within the landfill where
there are anomalously high magnetic values.
Buried metallic objects would locally induce
these types of magnetometer readings, of which
chemical waste drums may be included. The
lateral and vertical limits of the magnetic
anomalies should not be considered as exact
boundaries. The lateral extent is effected by
the data point grid spacing. Closer spacing of
the grid pattern would more accurately deline-
ate the boundaries of the magnetic highs. The
size and/or number of metallic objects and
their arrangement would also influence magne-
tometer values as would the depth of burial
and/or the distribution of metallic objects with
depth. Therefore, the limits of the suspected
areas containing buried metallic objects are
approximate, but these limits do represent
zones within the landfill that should not be
considered for test boring locations.
Subsurface Interface Radar
An SIR survey was conducted on the site
during July of 1980. Test runs were made with
304
-------
ALCYON LAKE
AREA OF CONDUCTIVITY SURVEYS
(SEE FIGURES 7,8 * 9)
SUBURBAN DEVELOPMENT
AREA OF
MAGNETOMETER
AND SUBSURFACE
INTERFACE RADAR
SURVEYS
(SEE FIGURES 5,6
APPROX PROPERTY
SCALE IN METERS
Figure 4. Geophysical Survey Plan, LiPari Landfill Site Area, Pitman, New Jersey.
-------
40
CONTOUR INTERVAL = 100 GAMMAS
MAGNETIC HIGH
6—- CONTOURS » 100
Figure 5. Magnetometer Survey, LiPari Landfill Site, Pitman, New Jersey.
-------
various transducers to determine the most
practical for the given field conditions. A high
resolution transducer was selected. A major
portion of the area, along the same 7.6 meter
grid lines as used in the magnetometer work,
was surveyed. The SIR survey lines are shown
in detail in Figure 6 along with a partial listing
of the areas where buried metallic objects were
detected.
As with the magnetometer survey, the 7.6
meter grid spacing is a factor in determining
the number of metallic objects which can be
located. The SIR system is more selective than
the magnetometer in that only objects directly
below the transducer are detected and record-
ed. The magnetometer measures the magnetic
field in the area of the observation point, and a
metallic object would not have to be directly
below the instrument to have an influence.
This difference in measurement techniques
must be considered when evaluating and corre-
lating data from each of these instruments.
Because of the narrow detection width on the
SIR unit, the signatures from metallic drums
will probably not appear at the location of each
magnetic anomaly. Where drums are seen on
the SIR, however, magnetic highs should and in
fact did occur. In general, a comparison of the
locations of metallic objects determined
through the radar survey with the magnetom-
eter survey showed a good association of
metallic objects with magnetic highs. Because
of the broader area coverage of the magnetom-
eter survey, the magnetic data were considered
to be more useful in developing the overall
boring location plan within the landfill.
As a check on both of the above methods
of investigation, an M-Scope survey was also
conducted. This survey was limited to specific
areas where both high and low magnetic read-
ings had been obtained. The M-Scope survey
results correlated very well with the
magnetometer survey results.
Conductivity Meter
The conductivity survey was conducted
over the area east of the landfill, as shown in
Figure k. As discussed above, preliminary field
work indicated that groundwater movement was
from the landfill toward Chestnut Branch, and
thus this site area was selected for the survey.
It was considered that contaminants migrating
from the landfill would increase the conductivi-
ty of the groundwater. This increased
conductivity could be measured and plotted,
thus providing a preliminary definition of the
area! extent of the contaminant plume. The
primary survey of the area was run on a grid
spacing of 7.6 meters, with an effective inves-
tigation depth of 7.5 meters. In order to verify
the initial results, a secondary survey was run
on a grid spacing of 15.2 meters, with a pene-
tration depth of 15 meters. The results of both
surveys were comparable and are shown in
contour map form in Figures 7 and 8. It is
considered that the conductivity pattern shown
in these figures is an approximation of the
plume of contaminated groundwater down-
gradient from the landfill site.
QUALITATIVE ASSESSMENT SUMMARY
The results of the above discussed geo-
physical surveys and the file information
reviewed have provided a basic understanding
of the site conditions, at least in a qualitative
sense. The approximate locations of metallic
wastes within the landfill have been assessed,
and the location of the contaminant plume in
the groundwater system downgradient of the
site has been estimated. A composite map
showing the results of both the magnetometer
survey and the conductivity survey is provided
as Figure 9. These data will be used as the
basis for developing the work to be conducted
during the quantification phase of the site
assessment program.
This work essentially completes the quali-
tative assessment phase of the site
investigation. As can be seen, a significant
amount of information has been gathered via
geophysical techniques, and these data will
facilitate the development of the quantitative
phase of work. It should be realized that
geophysical techniques have a tremendous
potential in the assessment of subsurface
conditions in and around landfill sites. It is
important, however, that these geophysical
applications be made with an understanding of
the capabilities and limitations of the specific
technique and equipment. As an example, the
techniques used in this research study relied
heavily on electromagnetic wave propagation.
There is very little interference to these types
of techniques at the LiPari site. These same
techniques may not have been applicable if
there were large areas of buried utilities or
radio transmission towers in the area. The site
conditions must be considered when selecting
the types of geophysical surveys or other
remote sensing surveys to be conducted.
307
-------
* METALLIC OBJECTS SHOWN
ON SI R
— TRAVERSE LINE
SCALE IN METERS
20
40
Figure 6. Subsurface Interface Radar Survey, LiPari Landfill Site, Pitman, New Jersey.
-------
LANDFILL
\
CONTOUR INTERVAL: 5M ILLI MHOS/METER
SCALE IN METERS
I . ... I
0 50
Figure 7. Conductivity Survey, 7.5 Meter Depth, LiPari Landfill Site, Pitman, New Jersey.
-------
LANDFILL
SCALE IN METERS
5O
CONTOUR INTERVAL = 5MILLIMHOS
/METER
Figure 8. Conductivity Survey, 15.0 Meter Depth, LiPari Landfill Site, Pitman, New Jersey.
-------
0 20 40
CONTOUR INTERVAL: 100 GAMMAS
CONTOUR INTERVAL = 5 Ml LLI MHOS / METER
-6 CONTOURS x 100
Figure 9. Composite Magnetometer and Conductivity Survey Results, LiPari Landfill Site, Pitman, New Jersey
-------
REMEDIAL ACTIONS AT UNCONTROLLED HAZARDOUS WASTE SITES
Nancy S. Neely, James J. Walsh
Dennis P. Gillespie, and Frederick J. Schauf
SCS Engineers
Covington, Kentucky
ABSTRACT
During the Summer of 1980, a nationwide survey was conducted to determine the status
of remedial measures applied at uncontrolled hazardous waste disposal sites. Remedial
actions were found to have been implemented at drum storage areas, incinerators, and
injection wells, but most frequently at landfills, dumps, and surface impoundments. At
the sites receiving such remedial actions, ground water was found to be the most commonly
affected media, followed closely by surface water.
Remedial activities encountered usually consisted of containment and/or removal
wastes. The survey determined that a lack of sufficient funds and/or improper selection
of corrective technologies were responsible for remedial actions having been applied
effectively only a portion of the time.
Nine sites were studied in detail to document typical pollution problems and remedial
actions at uncontrolled waste sites. Of these, remedial actions were completely effective
at only two. Technologies employed at these nine sites represented (1) containment, (2)
removal of waste for incineration or secure burial, (3) surface water controls, and/or
(4) ground water controls.
INTRODUCTION
Government agencies and private citi-
zens have become increasingly aware of
enviornmental problems associated with
unsound disposal and transport of hazard-
ous materials. An ABC News-Harris poll,
conducted after the headlines of Love
Canal, N.Y., and Elizabeth, N.J., found
that 76 percent of those surveyed consid-
ered the dumping of toxic chemicals to
be a serious problem. Ninety-four percent
of those interviewed wanted federal stan-
dards prohibiting such dumping made strict-
er than they are now. [1]
The magnitude of hazardous waste
problems can be reviewed by examining the
quantity of generated waste and the safety
of disposing the waste. The U.S. Environ-
mental Protection Agency (EPA) Office of
Solid Waste (OSW) estimates that 336 Mil-
lion metric tons (wet weight) of industrial
wastes are produced annually in the United
States and that the yearly growth rate
is about 3 percent. About 10 percent of
of this industrial waste is estimated to
be potentially hazardous. [2]
Eight EPA regions estimated that 4
percent of all disposal sites (32,000)
receiving hazardous waste might pose a
significant problem. [3] In contrast,
another study [4] estimates that 90 per-
cent of all hazardous waste has been dis-
posed in an unsound manner. A third study
[3] conducted for the OSW used both these
percentages along with the total number
of sites accepting hazardous waste as esti-
mated by the regions and by the report's
author to predict the number of sites that
might pose significant problems. The total
number of sites that contain hazardous
waste was thereby calculated to range from
32,000 to 51,000 and the number posing
significant problems to range from 1,200
to 34,000.
When one considers the number of sites
which could pose significant problems and
the number of sites which have already
endangered the environment, it is evident
312
-------
that comprehensive programs for enforcement
and remedial action response are needed.
The Solid and Hazardous Waste Research Div-
ision (SHWRD) of EPA's Municipal Environ-
mental Research Laboratory conducts research
and development to meet the needs of reme-
dial responses to uncontrolled hazardous
waste sites. This research and development
program is investigating technologies for
the prevention, control, and concentration
of hazardous substances released from un-
controlled waste sites. The types of re-
medial action initiated at uncontrolled
sites include more permanent, long-term
remedial actions, as well as short-term,
emergency responses. EPA SHWRD in Cin-
cinnati, OH is responsible for addressing
permanent, long-term responses and the
Oil and Hazardous Materials Spills Branch
(OHMSB) in Washington, D.C. for emergency,
short-term responses. [4]
Several federal laws (i.e., Resource
Conservation and Recovery Act, Clean Water
Act, Safe Drinking Water Act, and Toxic Sub-
stances Control Act) as well as state and
common laws provide enforcement provisions
which enable the government to attempt to
return a polluted or endangered environment
to its original state and to force the re-
sponsible party to reimburse the government
for funds expended. However, the nature of
improper disposal of hazardous waste makes
proof of liability difficult, since the
cause and effect relationship between the
act of disposal and eventual harm may not
appear for 20 years or more. [5]
To date, the two major laws which
have provided federal assistance for
remedial action at uncontrolled hazardous
waste sites are the Clean Water Act (CWA)
and the Resource Conservation and Recovery
Act (RCRA). Under Section 311 of the
CWA, money is available for emergency
remedial action at sites where release
of oil or hazardous materials threatens
navigable waters. However, land spills
that do not directly threaten surface
water are not covered under Section 311.
The authorities under CWA and RCRA
enable EPA to (1) supply limited assis-
tance for enforcement related investi-
gations (e.g., chemical analysis, site
investigation, technical assistance);
(2) take emergency remedial action where
navigable waters are threatened; and (3)
take legal action in cases where sites
pose an imminent hazard. Remedial actions
at sites where navigable waters are not
threatened can only be initiated and funded
by states, local governments, and respon-
sible parties. Usually in cases where 311
funds are not used, extensive time is in-
volved in identifying the problem, the
responsible party, and the remedial measure
which is likely to be successful. Even
more time is often required in getting
the responsible party to clean up such
sites, either voluntarily or through court
action. As a result of these conditions,
"Superfund" has been proposed to fill
the legislative gap described above.
PROJECT DESCRIPTION
In an effort to determine the type
and effectiveness of past remedial actions
at uncontrolled sites, a nationwide sur-
vey of on-going and completed remedial
actions projects was conducted from May
to October 1980. The purpose of the
survey was to provide information and
examples of applied remedial action
technologies. Examples provided in the
form of case histories identify typical
problems, effectiveness, and costs re-
lated to implementing remedial actions
at such facilities.
In identifying remedial action sites,
emphasis was placed upon landfill, dump,
surface impoundment, drum storage, incin-
erator, and deep well injection facili-
ties. For the purposes of this survey,
a waste burial site was designated as
a "landfill" if it was permitted and
would include sites resulting from mid-
night dumping and roadside dumping. Sur-
face impoundments include pits, ponds,
and lagoons used for the treatment, stor-
age and/or disposal of wastewater or
sludge. Injection wells include subsur-
face disposal wells and for the purposes
of the survey include such sites as
abandoned mine shafts. Incinerators
included facilities which dispose of
wastes by burning.
Case history sites were chosen after
consideration of the following factors:
(1) legal actions which would hamper
an in-depth investigation, (2) the ex-
tent and nature of the environmental
problem associated with the site; (3)
the nature and effectiveness of the
applied remedial action; and (4) access
313
-------
to the site and appurtenant file infor-
mation.
SURVEY FINDINGS AND LITERATURE COMPARISON
Initially 199 sites were identified
as having some form of remedial action.
Thirty of these sites were later deleted
from the list due to either (1) lack
of sufficient information, (2) insuffi-
cient progress on planned remedial action,
or (3) the use of "low-technology" reme-
dial actions. "Low-technology" actions
were defined to include measures such
as (1) merely filling a lagoon with native
soil without instituting surface or ground
water controls, (2) discontinuing waste
receipts at a landfill without attempting
to properly close the facility, or (3)
clearing a drum storage facility without
regard to existing soil or water contam-
ination.
Various approaches and remedial
measures for the cleanup of uncontrolled
sites have been published. [4,6] Gener-
ally the procedure used in identifying
and carrying out the remedial measure
to be employed includes the following
steps:
• The problem is identified
and the site is investigated.
• The problem is assessed. The
urgency of applying immediate
short term remedial measures
is determined. The need for
long term actions needed to
solve the problem is determined
and the source of funding is
identified.
• Preliminary design and cost
estimates for candidate remdial
actions are prepared.
t Remedial actions are implemented
with the funds being supplied
by the responsible party or other
sources.
Table 1 is presented to briefly summarize
the standard remedial methods used at
an uncontrolled site. The reader is
referred to references 4 and 9 for a
more complete listing and a description
of remedial measures.
TABLE 1. REMEDIAL ACTION TECHNIQUES [4]
Surface Water Controls
Grading Operations
Revegetation
Surface Water Diversion
Surface Water Collection
Groundwater Controls
Impermeable Barriers
Permeable Treatment Beds
Groundwater Pumping
Bioreclamation
Contaminated Water Treatment
Plume Containment
Leachate Control
Subsurface Drains
Drainage Ditches
Liners
Leachate Recycle
Gas Migration Control
Pipe Vents
Trench Vents
Gas Barriers
Gas Collection Systems
Gas Treatment Systesm
Gas Recovery
Direct Waste Treatment
Excavation
Hydraulic Dredging
Land Disposal
Incineration
Solidificatlon
Encapsulation
In-Situ Treatment
Wet Air Oxidation
Neutralization/Detoxification
Microbial Degratlon
Treatment of Contaminated Sewer
and Water Lines
In-Situ Cleaning
Leak Detection and Repairs
Removal and Replacement
Chemical Treatment
Contaminated Sediments
Mechanical Dredging
Low Turbidity Hydraulic Dredging
Dredge Spoil Management
Revegetation
Physical/Chemical Treatment
The survey of 169 remedial action
sites revealed that a variety of remedial
measures included containment on-site,
chemical/physical treatment (neutrali-
zation of acids and bases, precipitation,
etc.), biological treatment (land spread-
ing, oxidation ponds, and underground
enhancement of native microbes using
fertilizer),"incineration, and removal
and burial in a secure landfill. The
remedial measure most often employed
consisted of containment and/or removal
of the hazardous waste. Cost was found
to be the prime determinant of the type
of technology applied. As a result,
the primary remedial goal has been pre-
vention of further contamination of the
environment rather than complete cleanup.
Complete cleanup can require millions
of dollars, sophisticated technologies,
and long periods of time.
When hazardous material was contained
in its original location, surface water
controls were generally constructed (e.g.,
grading, diversion ditches, revegetation,
surface sealing, etc.) In most instances
where the ground water was contaminated,
a major portion of the waste was removed
and sent to a secure landfill or incin-
erated and surface water controls were
constructed to secure the remaining con-
taminants. Implementation of controls
314
-------
for ground water cleanup is typically
more expensive and time-consuming than
implementation of surface water controls.
Accordingly, ground water remedial mea-
sures were implemented at only a few
sites. Ground water pumping was the
most often applied such control while
remedial measures such as bentonite
slurry trenches or sheet pile cutoff
walls were found at some spill sites.
Tables 2 through 5 were compiled
based upon information gathered during
the survey. Over 130 individuals were
contacted to compile the data. Some
of the factors which should be considered
in reviewing the data present in the
following survey tables include: (1)
the data was based solely upon the
immediate survey findings, as reported
by individuals contacted; (2) a "po-
tential" threat was not considered suffi-
cient for the site to be included in
the inventory; while (3) a probable (but
undocumented) contamination was taken
as a positive finding.
Table 2 indicates the types of dis-
posal facilities experiencing remedial
actions. It should be noted that the
total number of facilities (204) shown
in Table 2 does not coincide with the
number of identified sites (169). The
higher number is the result of different
types of facilities being located on
the same property. More surface im-
poundments and landfills were identified
as experiencing remedial action than
other types of disposal facilities. This
would be anticipated since surface im-
poundments and landfills are the most
prevalent types of disposal method.
TABLE 2. FACILITY TYPE AT REMEDIAL ACTION
SITES
Number of Sites
Status
Facility Type
Landfill
Dump
Drum Storage
Surface Impoundment
Injection Well
Incinerator
Spill
Total
Active
16
0
11
18
1
1
0
Inactive
37
27
25
37
3
5
23
Total
Number
53
27
36
55
4
6
23
204
Table 3 was compiled to determine
the geographical location of sites which
had undergone remedial action. During
conversations with federal officials,
it became apparent that many factors affect
the geographical distribution. For
example, industrial waste disposal sites
would typically predominate in those
states which have more industry. The
possibility of a large number of uncon-
trolled sites needing remedial measures
would increase with an increase in the
number of industrial disposal sites. How-
ever, the presence of more uncontrolled
sites needing remedial measures did not
necessarily mean remedial measures were
being applied. Institution of remedial
measures is dependent upon time and force
exerted by public officials, as well as
the environmental concern of the site's
owner/operator. In order to provide a
perspective of remedial action sites,
U.S. total waste sites, the third column
in Table 3 demonstrates the number of
high-risk hazardous waste sites which
still require clean-up activities.
TABLE 3. LOCATION OF REMEDIAL ACTION VS. UNCONTROLLED WASTE SITES
Remedial Action
State Sites*
Alabama
Alaska
Arizona
Arkansas
California
Colorado
Connecticut
Delaware
Florida
Georgia
Hawa 1 i
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
2
o
3
2
3
3
4
2
7
4
o
o
8
3
1
2
5
3
2
1
5
11
3
0
4
5
o
o
1
10
o
14
7
5
4
o
o
16
4
3
o
10
3
1
o
2
o
1
3
1
High-Risk Uncontrolled
Waste Sites*
11
o
0
1
2
12
2
o
3
0
0
0
o
1
1
o
3
3
0
4
3
5
7
1
0
2
o
o
3
g
1
5
5
o
2
o
3
21
2
3
o
2
1
o
o
1
2
o
o
0
Total 50 States
169
120
*As determined by this survey
+From reference I.
315
-------
Public awareness and environmental
consciousness were strong factors in im-
plementation of remedial measures. Pres-
sure exerted by state officials sometimes
forced companies and property owners to
implement corrective actions. Since reme-
dial action is generally a time-consuming
endeavor, the number of remedial action
sites was also dependent upon how long
ago the environmental concerns were em-
phasized. Legal action to identify "re-
sponsible" persons for remedial actions
generally took four to nine years. After
this time, the identified responsible
party either instituted remedial actions
at the site or declared bankruptcy (in
the process refusing to remedy the sit-
uation) .
A study conducted on 421 cases asso-
ciated with land disposal of industrial
and pesticide waste damage revealed that
the ground water was affected 65 percent
of the time and the surface water was
affected 40 percent of the time. [6]
According to Table 4 ground water and
surface water were the most often affected
media at the 160 remedial action sites.
Ground water was affected 65 percent of
the time; surface water, 56 percent; soil,
41 percent; air, 29 percent; and food
chain, 12 percent of the time. Frequently
a site affected more than one media.
TABLE 4. AFFECTED MEDIA AT 169 REMEDIAL
ACTION SITES
stated, these funds are available only
for endangerment of navigable waters.
Since any one site might require millions
of dollars, total funding from state,
county, or municipal sources was unlikely.
As a result of these high costs, more
than one party often funded remedial
activities.
Table 5 was compiled to determine
the general status of improvement that
occurred at sites which had undergone
remedial actions. A total of 180 separate
remedial action efforts were initiated
at the 169 sites. The pollution status
was considered to be "unimproved" when
the implemented remedial measure did not
correct the contamination problem. Usual-
ly lack of improvement was the result
of inadequate funds or the type of action
instituted. Improved refers to a remedial
measure which may have partly corrected
the problem, but some problems are still
experienced at the site. A remedied site
was one at which the problem had been
corrected; e.g., contaminated surface
water was returned to its natural state.
Based on these definitions, the last
column in Table 5 indicates that 46 per-
cent of corrective actions were not effec-
tive, 38 percent improved the pollution
problem, and 16 percent were completely
effective.
TABLE 5. POLLUTION AND REMEDIAL
ACTION STATUS AT 169 SITES
Affected Media
Ground Water
Surface Water
Air
Soil
Food Chain
To ta 1
Number
Of
Occu rences
110
95
49
69
20
343
Pollution Status
Un1*^ roved
Improved
Remedied
Total
Number
Planned
Actions
16
12
0
28
of Remedial
On- Go ing
Actions
49
36
3
B8
Actions
Compl eted
Actions
17
21
26
64
Total
82
69
29
180-
• A toul of 180 remedial activities were Identified tt the 169 sites
Generally the state, county, and/or
municipality attempted to persuade the
owner/operator of an uncontrolled facility
to voluntarily remedy the environmental
hazards. If this effort failed, legal
proceedings were instituted against the
responsible party. Depending on the
degree of hazard posed by the site, var-
ious government agencies funded the
remedial activities while legal respon-
sibility was determined by the courts.
Federal financial assistance for remedial
measures was usually found to be funded
under Section 311 of CWA. As previously
CASE STUDY FINDINGS
Case study sites were selected based
on a desire to represent a range of fac-
ilities, pollution, and remedial actions.
Tables 6, 7, and 8 present an overview
of the nine case histories. The nine
sites include two remedied and seven im-
proved sites. Remedial action applied
at the seven improved sites showed varying
degrees of effectiveness. The combination
of all nine sites covered contamination
of all media including ground water, sur-
face water, soil, air, and the food chain.
316
-------
Waste types involved included mercury,
arsenci, solvents, oil, tire wastes, inor-
ganic and organic waste, and septic waste.
The types of facilities examined included
surface impoundments, landfills, drum
storages, and incinerators. The techno-
logy employed consisted mainly of contain-
ment, removal of waste for incineration
or secure burial, and institution of
surface water and/or ground water con-
trols. Costs for implemented remedial
actions ranged from $250,000 to more
than $7,000,000.
TABLE 6. CASE STUDY SITE IDENTIFICATION
Site
No.
A
B
C
D
Name
01 in Corporation
Firestone Tire and
Rubber
Anonymous
Destructo/Carol awn
Location
Saltvllle, PA
Pottstown, PA
East Central, NY
Kernersville, NC
Waste Type
Mercury
Tires, SO- scrubber
waste, organic waste,
pigments, PVC sludge
Solvents, oils, paint
waste with PCB
Volatile/flammable waste
Remedial Action Technology
Graded and constructed erosion conrol structures. Removed
contaminants. Planning extensive remedial action ({23 million).
Recovery wells intercepted polluted ground water and recycled
it through their plant. Expected to be 100 percent effective.
Lagoons filled and capped. Diversion ditches and test wells
installed.
Two Phases: 1. Waste removed, incinerated or landfilled.
E Whitmoyer
Laboratories
F Western Sand and
Gravel
G Ferguson Property
H 3M Company
I UMtehouse/Allied
Petroleum
Myerstown, PA
Burrillvllle, RI
Rock Hill, SC
Woodbury, MN
Jacksonville, FL
Arsenic compounds
Septic plus hazardous
wastes
Solvents, heavy metals
Spent solvents, add
sludge
011, PCB
Contaminated soil removed and landfilled.
2. Waste removed* incinerated, landfilled. and deep
well injected.
Removed arsenic waste from lagoon, treated and discharged. Waste
piles of arsenic placed in concrete vault. Ground water treated
using purging wells. Some contaminated soil remains.
Four lagoons pumped, dried, and contents stored off and on-site.
Monitoring wells installed. Future remedial action planned.
Two Phases: 1. Contained with polyethylene and clay cap.
Installed surface water diversion ditches and
vent pipes in contained area.
2. Since phase one ineffective, removed liquid.
Still some sludge and drums left.
Pits emptied and contents burned. Barrier wells installed to
stop spread of contaminated ground water.
Mobile activated carbon unit dewatered pit, oil absorbed using
solid waste and earth. Future remedial action planned.
TABLE 7. CASE STUDY SITE BACKGROUND
Facility Type
Status
Remedial Action
= •= 5 s = fc
S 3
Site
NO.
4
B
0
E
F
G
H
I
•o
c
_l
X
X
X
X
f 1
— o
X
X
a "? c "5.
,/) ^ — t/i
x
X
X
X X
X
X
X
X X
i- O
3 S
X
X
X
x
X
X
X
X
X
c
1
X
X
X
X
X
X
3 i- O O
X X
X X
XX X
X X
XXX
X
X XX
1
X
X
!! II!
X
X X
X
X XX
X X
X XX
ill 111 at
XXX
X X
X X
X X
X X
XX X
X X
X XXX
D I
317
-------
TABLE 8 REMEDIAL ACTION COSTS AT CASE STUDY SITES
Site
No. Waste Type
A Mercury
B Tires, SO,, Scrubber
Udste, Organic
Waste Pigments,
PUC Sludge
C Solvents, Oils,
Paint Waste with
PCB
D Volatile/Flammable
Waste
E Arsenic Compounds
F Septic plus Hazard-
ous Wastes
G Solvents, Heavy
Metals
H Spent Solvents,
Acid Sludge
I Oil, PCB
Remedial Action
I
Surface
Water Controls
X
X
X
Ground
Water Controls
X
X
X
Direct
Waste Treatment
X
X
X
X
X
X
Contaminated
Water Treatment
X
X
Estimated
Cost
$ 400,000
250,000
Unknown
> 364,000
Unknown
> 1,000, 000*
143,178
> 7,000,000
> 253,000*
CONCLUSION
Remedial measures encountered during
this survey were usually confined to con-
tainment and/or removal of the hazardous
wastes. The primary goal was to prevent
further contamination of the environment
rather than to perform a complete clean-
up. Complete environmental cleanup of
ground water or surface water generally
requires sophisticated technology, addi-
tional money, and additional time. There-
fore, a responsible party with sufficient
funds and expertise must be located before
complete cleanup can occur. In most cases
sufficient funds have not been available
for effective remedial action. The U.S.
EPA is able to provide only limited funds
under Section 311 of the CWA. States
and local governments typically cannot
provide sufficient money for total clean-
up, since correction of one site may
require millions of dollars.
Based on the case studies and sur-
vey, the state-of-the-practice in remedial
action does not look favorable. Fully
46 percent of the time the applied
remedial action was ineffective-and only
a portion of all uncontrolled sites
have received some form of remedial action.
Remedial action applied at a site exper-
iencing problems was found to be totally
effective only 16 percent of the time.
It should be emphasized that the
numbers presented in this section are
based on assumptions by the persons per-
forming the survey and the opinions of
those interviewed. However, the per-
centage numbers should be a fairly
accurate representation of the state-
of-the-practice in remedial actions.
REFERENCES
1. Magnuson, Ed, et al. 1980. The
Poisoning of America. TIME,
Sept.22, 1980, pp.58-69.
2. Metry, Amer A., 1980. Comprehensive
Hazardous Waste Management. In
the Handbook of Waste Management.
Technomic Publishing Co., Inc.,
Westport, Connecticut, pp. 45-89.
3. Fred C. Hart Associates, Inc., 1979.
Preliminary Assessment of Clean
Up Cost for National Hazardous
Waste Problems. Prepared for the
Office of Solid Waste, Contract
#68-01-5063.
4. Hill, Ronald D., et al. 1980. USEPA
Research Program: Uncontrolled
Hazardous Waste Sites. In Pro-
ceedings of National Conferanee on
Management of Uncontrolled Hazard-
ous Waste Sites, U.S. Environmental
318
-------
Protection Agency, Washington, D.C.,
pp. 173-179.
5. Brandwein, David I., 1980. The Dim-
ensions of Corporate Liability.
In Proceedings of National Con-
ference on Management of Uncon-
trolled Hazardous Waste Sites, U.S.
Environmental Protection Agency,
Washington, D.C., pp. 262-268.
6. Ghassemi, Masood, et al. 1980. Com-
parative Evaluation of Processes
for the Treatment of Concentrated
Wastewaters at Uncontrolled Hazard-
ous Waste Sites. In Proceedings
of National Conference on Manage-
ment of Uncontrolled Hazardous
Waste Sites, U.S. Environmental
Protection Agency, Washington, D.C.,
pp. 160-164.
7. Paige, Sidney, et al. 1980. Prelim-
inary Design and Cost Estimates for
Remedial Actions at Hazardous
Waste Disposal Sites. In Proceed-
ings of National Conference on
Management of Uncontrolled Hazard-
ous Waste Sites, U.S. Environmental
Protection Agency, Washington, D.C.,
pp. 202-207.
8. Pease, Robert A. Jr., et al. 1980.
Management of Abandoned Site
Cleanups: Wade Property, Chester,
Pennsylvania. In Proceedings of
National Conference on Management
of Uncontrolled Hazardous Waste
Sites, U.S. Environmental Pro-
tection Agency, Washington, D.C.,
pp. 147-151.
9. Sills, Michael A., et al. 1980. Eval-
uation of Remedial Treatment,
Detoxification and Stabilization
Alternatives. In Proceedings of
National Conference on Management
of Uncontrolled Hazardous Waste
Sites, U.S. Environmental Pro-
tection Agency, Washington, D.C.,
pp. 192-201.
319
-------
BARREL AND DRUM RECONDITIONING
INDUSTRY STATUS PROFILE
C. J. Touhill
Touhill, Shuckrow and Associates, Inc.
Pittsburgh, Pennsylvania 15237
Stephen C. James
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
ABSTRACT
This report is an industry profile of drum reconditioning process
characteristics and the current status of pollutant generation and disposal.
An overview of the reconditioning industry describes number, location, and
types of facilities, and estimates the volumes of drums processed. Drum
characteristics, new drum production, and users are defined. Washing and
burning processes for reconditioning are described along with operating and
design criteria for individual unit operations. The descriptions define
typical industry practice. Processing procedures which influence product
quality and environmental pollutant generation are discussed. Processing of
pesticide containers is discussed especially with regard to the ability of
reconditioning procedures to detoxify pesticide residuals. Pollutant load-
ings are defined in terms of sources and pathways, major pollutant parame-
ters, and generation and disposition modes. Discharge data is given for
wastewater discharges and residues. Current status of pollution control
practice is defined in terms of processes and equipment, operating proce-
dures, disposal practices, removal efficiencies, and costs.
Introduction
For many people, the spent 208
liter (55~gallon) steel drum has be-
come the symbol of toxic and hazard-
ous wastes. Even though drum recon-
ditioning is intended to support en-
vironmentally desirable goals of
recycle, reuse, and safe disposal of
used drums, popular equating of
drums and hazards focuses adverse
attention upon the industry. The
Environmental Protection Agency (EPA)
recognizes the useful service per-
formed by reconditioners, but also
is alert to potential problems the
industry may encounter in meeting
high standards of environmental qual-
ity. Because of this awareness, EPA
contracted with the firm of Touhill,
Shuckrow and Associates, Inc. (TSA)
for the following purposes:
to define representative
practice for barrel and
drum reconditioning by
washing and burning
to determine the environ-
mental impact of recondi-
tioning processes
to recommend procedures for
designing, optimizing, and
retrofitting reconditioning
facilities to meet rigorous
environmental standards
to determine the capabili-
ties of the reconditioning
industry to process pesti-
cide and toxic chemical con-
tainers safely.
320
-------
The TSA project is being con-
ducted in three parts. In the first,
current status of the industry is
evaluated in terns of state-of-the-
art practices for both recondition-
ing and pollution control. Results
of this evaluation are the subject
of this paper. In the second part
of the program, which will be re-
ported upon separately, three recon-
ditioning sites were selected for
sampling and analysis to define more
closely pollutant sources and path-
ways. The final phase will be de-
velopment of recommendations for
design modifications, process opti-
mization, and retrofitting recondi-
tioning processes.
Five sources of information
were used to prepare the first phase
report: 1) publicly available lit-
erature, 2) transcripts of National
Barrel and Drum Association (NABADA)
forums on reconditioning problems
and techniques, 3) proceedings of
four International Conferences on
Steel Drums, 4) responses to a ques-
tionnaire sent by NABADA to its mem-
bership, and 5) visits to eleven
drum reconditioning plants.
Industry Overview
During 1979, about 250 recondi-
tioners processed more than 41 mil-
lion steel drums. More than 95 per-
cent were 208 liter (55-gallon)
drums; most of the rest were 114 li-
ters (30-gallons). About two-thirds
of the drums are reconditioned at
washing plants which process tight
head drums. The remainder, open
head drums, are burned in drum rec-
lamation furnaces. Facilities that
only wash drums account for 39 per-
cent of reconditioning plants.
Those that only burn account for 18
percent, whereas 43 percent of the
facilities perform both functions.
NABADA members represent only
48 percent of the reconditioning
plants in the country, but they pro-
cess more than 90 percent of all
washed drums and more than 97 per-
cent of those burned. For non-
NABADA reconditioning facilities, 70
percent recondition for service and
resale, and 30 percent are users who
process only their own drums. More
than half (52 percent) of the drums
washed are on a service or laundry
basis; 45 percent are for resale.
By comparison, only about one-third
of drums burned are on a service
basis; 62 percent are resold.
In 1972, the U.S. Department of
Commerce (1) determined that the
reconditioning industry had sales
of nearly $110 million (Table 1).
This would be more than $210 mil-
lion in current dollars. At that
time, more than 56 percent of the
sales were in the Middle Atlantic
and Great Lakes states. Based upon
present reconditioning volumes and
productivity rates (assumed to be
20 drums/day/employee), industry
employment is about 7,900. The in-
dustry presently is operating at
approximately 70 percent of design
capacity. The average plant has an
IQh day inventory, most of which is
stored on-site.
For many years, the annual ra-
tio of new drums produced to recon-
ditioned has been nearly 1 to 1.
Additionally, the ratio of new tight
to open head drums (4 to 1) has not
changed over the past 10 to 12
years. However, trends show a de-
crease in drum thickness. For ex-
ample, in 1969, 18-gage or heavier
drums comprised 64.3 percent of
those manufactured that year. In
March of 1980, that percentage had
fallen to 41.4. This reduces the
pool of potentially reconditionable
drums, because thinner drums are
less able to withstand the rigors
of transportation and recondition-
ing processes.
Table 2 shows new and recondi-
tioned drum usage according to three
sources. New drum end use is based
upon data in the Bureau of Census,
Current Industrial Reports on steel
shipping barrels, drums, and pails
(also known as the M34K reports) (2).
Reconditioned usage is based upon
data from the 1973 Checchi Report
for NABADA (3), and the recent
NABADA survey made in conjunction
with this project. A high percent-
age of oil and petroleum drums are
used and recycled. On the other
321
-------
TABLE 1. RECONDITIONER SALES BY REGION (1) (1972 DOLLARS)
REGION
New Encland
Boston SMSA
Middle Atlantic
New York SMSA
Newark SMSA
Philadelphia SMSA
East North Central
Cleveland SMSA
Detroit SMSA
Chicago SMSA
West North Central
South Atlantic
East South Central
West South Central
Houston SMSA
Dallas-Ft. Worth SMSA
Mountain
Pacific
Los Angeles SMSA
San Francisco SMSA
NUMBER OF
RECONDITIONERS
15_
6
M
18
13
16
ii
12
5
18
n
22
21
4
6
7_
3_2
11
6
SALES
(in thousands)
5,085
1,384
32,863
6,969
9,023
7,248
28,535
3,035
1,708
8,521
2,821
14,940
280
7,453
1,401
2,019
1,503
16,457
4,729
3,895
TOTAL
263
109,537
TABLE 2. NEW AND RECONDITIONED DRUM USAGE
USE
Food
Oil & Petroleum
Paint
Ink
Adhesive
Resins
Industrial Chemicals
Pesticides
Cleaning Solvents
Janitorial Supplies
Other
Unspecified
NEW DRUM
END USE (1979)
rn
5.1
15.2
(6.6
40.2
3.1
9.8
20.1
SHIPPERS TO
RECONDITIONERS
Checchi
37.
19.
_
9.
-
14.
2.
11.
8
0
6
1
9
3
NABADA
6
36
10
4
6
8
15
0
8
.8
.2
.0
.8
.8
.8
.6
.5
.8
5.3
1.
322
-------
hand, a low percentage of chemical
drums are reused. Hence, it is not
surprising that drums containing
spent industrial chemicals or chem-
ical residuals comprise a signifi-
cant number of drums found at aban-
doned hazardous waste disposal sites,
because such drums exit the user
system quicker.
Reconditioning Processes
Steel drums are processed by
either washing or burning. Because
tight head drums almost always are
washed, reconditioners frequently
refer to washing facilities as tight
head plants. Conversely, open head
drums are processed almost exclu-
sively by burning; hence burning op-
erations often are called open head
plants.
Washing Process
Most drum washing is done with
strong hot caustic solution. De-
spite common usage of this technique,
no two tight head reconditioning
plants are the same. Certainly
there are many similarities, but for
maintenance or enhancement of envi-
ronmental quality standards each
plant must be evaluated separately.
Nevertheless, a flow diagram which
generally represents the caustic
washing process and its many varia-
tions is shown in Figure 1.
In a washing plant, the follow-
ing operations generally are em-
ployed. After screening and drain-
ing upon receipt, drums are pre-
flushed using a strong hot caustic
solution. Subsequently, they pro-
ceed to a submerged caustic washing
tank. When drum contents are diffi-
cult to remove using caustic alone,
chains are inserted into the drum
along with caustic and the drum is
tumbled to dislodge adhering materi-
als. If drum contents cannot be re-
moved by chaining or are cleaned on-
ly with great difficulty, the drum
heads are cut off, thus converting
them to open heads, and they are
sent to a burning plant. About one-
third of washing plants remove rust
using hydrochloric acid washes.
Tight head drums then are rinsed,
dedented, shot blasted, leak tested,
and painted.
RECEIPT*
SCREENING
1
RETURN OF DAMAGED DRUMS
OR THOSE CONTAINING UN-
ACCEPTABLE MATERIALS TO
FIGURE 1
WASHING FLOW DIAGRAM
Some operating data derived
from the recent NABADA survey are
given below. Caustic concentra-
tions averaged 12.2 percent for pre-
flushing and 13.4 percent for sub-
merger or stripper solutions^ Tem-
peratures ^averaged 82°C (180°F) and
84°C (184°F) for preflush and strip-
per solutions, respectively. Near-
ly 88 percent of all washing plants
chain some drums. Of these, 22 per-
cent chain all drums received. A-
bout 38 percent of washing plants
use 20° Baume hydrochloric acid
323
-------
washes to remove rust.
Tight head drum dedenting is
done at an average pressure of 3.11
kPa (65 psi). Most reconditioners
(60 percent) used a pressure of 0.34
kPa (7 psi) for leak testing; how-
ever, others choose slightly higher
pressures so that the mean was 0.38
kPa (8.0 psi). In the NABADA survey,
drum rates through shot blasters
ranged from 50 to 450 drums/hr.
Responders to the NABADA ques-
tionnaire indicated that water use
rate per durm is 66 liters/drum
(17.5 gallons/drum). This is for
plants that wash, burn, or both.
For plants that washed only, the use
rate was 50 liters/drum (13.3 gal-
lons/drum) . Plants that only burned
had a rate of 41 liters/drum (10.9
gallons/drum). Those that perform
both functions had a considerably
higher rate.
Burning Process
A flow diagram for a typical
continuous tunnel drum reclamation
furnace operation is shown in Figure
2. In addition, Figure 3 is a
sketch of such an operation. At
most burning plants, drums are in-
spected upon receipt, and those con-
taining residues beyond plant crite-
rion for emptiness and those con-
taining unacceptable materials are
returned to the shipper along with
damaged drums. Some reconditioners
drain the drums before burning in
order to reduce temperature excur-
sions due to materials in the drum.
Others believe that the best way to
get rid of the residuals in the drum
is to burn them directly; thus,
draining before burning is avoided.
Some furnaces have water sprays
or steam injection at the inlet op-
ening to prevent flashbacks and pos-
sible operator injury. Others have
built-in distance barriers to reduce
operator 3xposure to flashbacks.
Conveyor belts move drums
through the furnace at an average
rate of from 6 to 8 per minute. Av-
erage residence time is 6.6 minutes.
Mean furnace temperature is 675°C
RETURN OF DAMAGED
DRUMS OR THOSE
CONTAINING UN-
ACCEPTABLE
MATERIALS TO
SHIPPER
RECONDITIONED DRUMs|
FIGURE 2
BURNING FLOW DIAGRAM
(1250 F), whereas the range encoun-
tered in the NABADA survey was 315°C
(600°F) to 980°C (1800°F). All drum
reclamation furnaces use afterburn-
ers to control air emissions. Af-
terburners are on all the time at
more than 80 percent of burning
plants. Afterburners operate 95 per-
cent of the time considering an av-
erage of all plants. Average tem-
perature and residence time for af-
terburners is 810°C (1490°F) and 0.5
second. The State of California re-
quires reconditioners who burn pesti-
cide drums to operate afterburners
at 900°C (1650°F) using a 0.5 second
residence time.
When drums exit the furnace,
they are either air-cooled or are
324
-------
-siicn
SECOHDIR1
I If. PORT
-B»FFU
FIGURE 3
TYPICAL DRUM RECLAMATION FURNACE WITH AFTERBURNER
(Reference No. 4)
water quenched. About 40
percent of burning plants have the
capability to quench, but not all
use it all the time. Some only op-
erate the water quencher when smoky
drums are being burned, or when
there is a visible emission from the
drum outlet opening. After cooling,
open head drums are shot blasted,
dedented, leak tested, lined, and
painted.
Natural gas is the preferred
furnace fuel. Where it is unavail-
able or uneconomical. No. 2 oil is
recommended.
Operating Procedures
Most reconditioners have estab-
lished procedures for drum receiving
and storage, and nearly two-thirds
have oil recovery systems. At a
typical plant, about 22,700 liters
(6,000 gallons) of oil are recovered
monthly.
About 5.4 percent of drums
bound for reconditioning eventually
are discarded. The bulk of these
are sold for scrap.
Almost all reconditioners re-
fuse to accept drums containing cer-
tain materials (Table 3). Pesti-
cides are refused by 83 percent of
reconditioners. Other drums fre-
quently refused are those formerly
containing ink and adhesives. The
industry processes over 200,000 used
pesticide drums per year, although
there is some uncertainty in this
figure. Most are burned, but a few
plants do wash significant numbers
of pesticide drums. Even though the
industry claims that few facilities
accept pesticide drums, most plants
have such drums scattered within
their inventories.
Plants that process pesticide
drums on a regular basis use special
handling and processing procedures.
325
-------
TABLE 3
TYPES OF DRUMS REUSED
FOR PROCESSING
(expressed as percentage of
companies having restrictions)
Pesticides 83.0
Ink 29.8
Adhesives 27.7
Paint 14.9
Toxic and Hazardous Materials 12.8
Mercaptans, TDI and other
Unneutralized Gas Odorants 10.6
Resins 8.5
Class B Poisons 6.4
Odorous Materials 6.4
Tar/Asphalt 4.3
PCB 4.3
Cadmium Compounds 4.3
The following are 2% or less:
Delaware Drums
Heavy Drums
Plastic Inserts
Excessive Residual
Unknown Contents
Smokers
Automobile Sound Deadener
Silicone
Acrylics
Carbon Disulfide
Chlorides
Cyanides
Mercury Compounds
Sodium
Sulfur Monochloride
Zinc Compounds
Formaldehyde
Organic Phosphates
Benzidine and its Salts
Dye
Industrial Chemicals
Carcinogens
Radioactive Materials
Results reported by others indicate
that burning (using afterburners at
900°C or 1650 F) effectively detoxi-
fies most pesticide drums (5). Ad-
ditionally, in washing plants, phos-
phorus and nitrogen-containing pes-
ticides lend themselves to
detoxification by alkaline hydroly-
sis. Many unanswered questions are
expected to be dealt with as results
of the sampling and analysis portion
of this EPA project become available.
Pollutant Loadings
The EPA project placed most em-
phasis on liquid and solid wastes in
terms of pollutant sources and path-
ways. Previous investigations indi-
cated that air emissions presented
little problem in drum recondition-
ing.
Most washing facilities either
recycle and reuse caustic and rinse
waters or discharge effluents into
public sewerage systems. Table 4
shows various methods for dealing
with liquid waste streams. About
half of all plants (including those
that burn), claim to discharge some
water into public sewers. Nearly 20
percent of washing plants claim to
have completely closed cycle or "ze-
ro discharge" systems. Only 10 per-
cent are direct dischargers after
treatment. Mean flow for a typical
facility is 56,800 liters (15,000
gallons) per day. Because such a
high percentage of plants discharge
to sewers, water quality limitations
for such discharges become very im-
portant. Typical limitations as de-
termined during the recent NABADA
survey are shown in Table 5.
Reconditioning wastewater is
characterized by high pH, COD, BOD,
and solids values. These and other
wast«;water data are given in Table 6.
There is wide variability in waste-
waters depending upon the types of
drums processed. Some of the mean
values shown in Table 6 are skewed
by one or more outlying data points.
Hence;, the data set for each of the
parameters in Table 6 was examined
to determine which values were out-
side of the 99 percent confidence
level. Eleven of the 58 showed that
some values were outside of this
confidence limit. Such values were
discarded and a new analysis made.
The process was repeated until all
data sets conformed to the 99 per-
cent confidence limit. The statis-
tically adjusted discharge values
are given in Table 7. This is not
to imply that these values are bet-
ter or worse than those in Table 6.
326
-------
TABLE 4
METHODS FOR DEALING WITH
LIQUID WASTE STREAMS
METHOD PERCENTAGE OF RECONDITIONERS
USING THE METHOD
Caustic Solution Rinse Water Runoff
Discharge to public sewer 7.8 42.5 42.9
Neutralize and discharge
to public sewer 13.7
Discharge to stream 0 2.5 21.4
Recycle and reuse 52.9 35.0
Treat at own treatment plant
and discharge to stream 2.0 10.0 35.7
Discharge some and recycle
some 21.6
Haul to landfill 2.0 7.5
Evaporate - 2.5
TOTAL 100.0 100.0 100.0
327
-------
TABLE 5
WATER QUALITY LIMITATIONS
FOR DISCHARGES TO PUBLICLY
OWNED SEWERAGE SYSTEMS
No
Parameter
BOD (biochemical
oxygen demand)
COD (chemical
oxygen demand)
Suspended Solids
Total Dissolved Solids
PH
Oil and Grease
Fats and Oils
Fuel Oil
Benzene
Cyanides
Formaldehyde
Gasoline
Naphtha
Pesticides
Phenol
Volatile Organics
Chlorides
Fluorides
Phosphates
Total Kjeldahl Nitrogen
Temperature
Antimony
Arsenic
. of Reported
Limitations
12
5
12
4
13
11
2
1
1
8
1
1
1
4
5
1
1
1
2
1
6
5
7
Ranae*
Mean Value*
229
345
266
625
5.3 to 9.8
162
100
0
0
2.1
5.0
0
0
0
0.5
5.0
250
0
21
no limit
140°F
1.0
0.2
Low
20
25
25
500
4.5
0
100
-
-
0.1
-
-
-
-
0.3
-
-
-
12
110°F
0
0.05
High
no limit
450
no limit
1000
10.5
1000
100
-
-
10
-
-
-
-
0.5
-
-
-
30
150°F
5.0
0.5
328
-------
TABLE 5 (con't)
No. of Reported Range'
Parameter Limitations
Barium
Beryllium
Bismuth
Boron
Cadmium
Chromium (total)
Chromium (hexavalent)
Cobalt
Copper
Iron
Lead
Manganese
Mercury
Molybdenum
Nickel
Rhenium
Selenium
Silver
Strontium
Tellurium
Tin
Uranium
Zinc
7
4
4
4
10
9
6
4
9
6
11
7
5
4
9
4
7
7
4
4
4
1
11
Mean Value* Low
2.9
0.3
0.3
0.6
1.1
6.5
0.7
0.05
1.0
4.0
0.9
2.0
0.007
0.03
2.4
0.1
0.6
1.1
0.1
0.1
0.8
0
5.4
1.0
0
0
0.5
0.002
0.05
0.05
0
0 .2
1.0
0.1
0.05
0
0
0.5
0
0.02
0.05
0
0
0
-
1.0
— 1 • 1 —
Hiqh
5.0
1.0
1.0
1.0
3.0
25
1.0
0.2
3
5.0
5
5.0
0.02
0.1
10
0.5
2.0
5.0
0.5
0.5
1.0
-
15
otherwise.
329
-------
TABLE 6
WASTEWATER DISCHARGE ANALYSES*
u>
8
PARAMETER
Aluminum
Antimony
Arsenic
Barium
Beryllium
Bismuth
BOD 5
Boron
Bromide
Cadmium
Calcium
TOC
COD
Chloride
Chromium (total)
Chromium (hexavalent)
Cobalt
Color (Pt/Co units)
Copper
Cyanide (total)
Cyanide ("A")
Fluoride
Iron
Lead
Magnesium
Manganese
Mercury
Molybdenum
Nickel
Nitrogen (Total Kjeldahl)
No. of
Observations
2
6
3
3
3
2
31
4
1
12
4
5
24
6
26
4
2
3
16
7
5
1
12
29
4
4
11
3
12
4
Mean
181.5
0.3
0.13
9.3
0
0
4,599
2.5
0
0.10
74
5,397
24,549
2,169
12.5
0.11
0.2
420
2.4
1.4
0.5
1.2
114
45.8
23
1.5
1.0
0
0.3
18.6
Standard
Deviation
195.9
0.4
0.23
9.0
0
0
9,882
4.3
-
0.26
60
8,455
64,889
1,540
43.3
0.12
0.2
312
5.7
2.0
0.4
-
294
154.5
21
2.4
1.9
0
0.4
15.3
Range
Low
43
0
<0.03
<1.0
0
0
10
<0.05
-
0.0002
28
71
91.5
6.7
0.023
0
0
60
0
0.2
0.2
-
1.6
0
5.5
0
0
0
0
6.2
High
320
1
0.40
10
0
0
44,133
8.9
-
0.90
162
20,412
310,909
4,144
224
0.22
0.3
600
23.0
5.9
1.2
-
1,041
682
54
5
5.9
0
1.0
40
"continued"
-------
TABLE 6 (con't)
PARAMETER
Nitrogen (Ammonia)
Nitrogen (Nitrate)
Oil and Grease
PH
Phenols
Phosphorus (total)
Potassium
Selenium
Silicon
Silver
Sodium
Strontium
Sulfate
Sulfide
Sulfite
Surfactants (MBAS)
TDS
Total Solids
Total Fixed Solids
Total Suspended Solids
Total Volatile Solids
Mineral Suspended Solids
Volatile Suspended Solids
Thallium
Tin
Titanium
Turbidity (NTU)
Zinc
No. of
Observations
4
5
37
39
21
8
2
4
3
4
4
2
4
3
1
2
4
7
1
46
2
1
1
1
6
2
3
33
Mean
13.3
4.6
10,228
10.9
43.8
34.3
33
0.12
75
0.01
10,157
0
7,747
14.6
<1
2.0
8,514
8,701
5,414
2,435
265
32
108
<0.1
1.1
115
33
24.0
Standard
Deviation
10.8
5.1
42,656
3.3
45 .5
25.1
15
0.23
33
0.02
9,643
0
8,384
22.1
1.6
9,264
12,013
_
4,113
92
_
_
2.4
163
10
46.5
* All values in milligrams/liter unless specified otherwise.
Range
Low
1.96
0.66
19.2
1.8
0.044
5.04
22.0
0
37
0
1,678
0
16
0.7
0.8
944
970
77
200
_
—
0
o
21
0.1
High
23
13.2
248,340
13.6
148
68.48
43
0 .47
94
0 .04
19 ,000
15,000
40 .0
3.1
22,000
35,000
24,000
330
_
_
6
230
39
228
-------
TABLE
STATISTICALLY ADJUSTED DISCHARGE VALUES*
Parameter
BOD5
Cadmium
COD
Chromium (total)
Copper
Iron
Lead
Oil and Grease
PH
Total Suspended
Solids
Zinc
No. of
Observations
27
11
19
25
14
11
24
33
37
44
30
Mean
1,262
0.03
4,011
4.0
0.7
29.3
2.1
901
11.4
1,669
10.5
Standard
Deviation
899
0.06
2,835
4.0
0.8
34.0
2.2
1,202
2.7
1,662
8.6
Range
Low High
10
0
91.5
0.023
0
1.6
0
19.2
4.8
77
0.1
3,617
0.15
8,638
11.0
2.5
54
6.7
3,970
13.6
5,195
34
* All values in milligrams/liter unless specified otherwise
It merely means that outlying values
have been discarded in an attempt to
present values which potentially
could be more typical.
There is some evidence of haz-
ardous organics in small concentra-
tions in some reconditioning waste-
water.
The biggest problem facing re-
conditioners is the need to provide
safe, economical, and environmental-
ly acceptable means for residuals
management. Evolving solid and haz-
ardous waste management regulations
are expected to have major impact on
the industry. Disposition methods
for drum residue are shown in Tables
8 and 9. Clearly, landfilling has
been and continues to be the predom-
inant method for residue disposal.
Use of incineration is increasing
and will continue to at an acceler-
ated rate as the impact of RCRA
regulations are felt.
On the average, drums received
by reconditioners contain 2.5 liters
(0.65 gallons) or 2.5 kilograms (5.4
pounds) of residues. As Table 10
reveals, caustic sludges have high
pH, COD, BOD, and oil and grease
concentrations, much as caustic
washing solutions, but in far great-
er amounts.
Presently, surface runoff at
reconditioning facilities is not be-
ing handled to any extent.
332
-------
TABLE 8
RESIDUE DISPOSAL METHODS
BY REGION IN 1973
METHOD
PERCENTAGE OF RECONDITIONERS
USING THE METHOD
Eastern
Region
Central
Region
Western
Region
Total
U.S.
Landfill
Burn and landfill
- on-site
- off-site
- no residue
Discharge to sewer
Store
Contract burning
Contractor hauling
58.
8.
16.
8.
4.
4.
-
3
3
7
3
2
2
72.5
-
13.8
6.9
-
-
3.4
3.4
80.0 67.
3.
20.0 15.
6 .
1.
1.
1.
1.
3
4
6
9
7
7
7
7
TOTAL
100.0
100.0
100.0
100.0
TABLE 9
METHODS FOR DEALING WITH
RESIDUES (NABADA SURVEY)
METHOD
Discharge to sewer
Contractor hauling
On-site land disposal
On-site burning
Landfill off-site
Return to emptier
Burn and landfill
Dispose with other sludges
Store
Own treatment plant
Drain to caustic solution
Off-site burning
PERCENTAGE OF RECONDITIONERS
USING THE METHOD
Concentrated
Liquid Drainage
6.0
40.0
4. 0
16. 0
24 .0
2.0
_
Sludges and
Solid Drainage
1. 9
38. 5
7 . 7
-
36.5
1.9
5. 8
Burned
Ash
-
32. 3
-
-
38. 7
3. 2
-
6.0
2.0
1.9
5. I
25 . !
TOTAL
100.0
100.0
100. 0
333
-------
TABLE 10
CAUSTIC SLUDGE COMPOSITION*
No. of
Parameter Observations
PH
Oil and Grease
Floating Oil
TDS
Total Solids
Suspended Solids
BOD
COD
TOC
Arsenic
Barium
Cadmium
Calcium
Chlorides
Chromium (total)
Chromium
(hexavalent)
Copper
Cyanides
Fluorides
Iron
Lead
Manganese
Mercury
Nickel
Phenol
6
5
1
3
5
4
1
1
3
2
1
4
1
2
6
2
4
4
2
4
4
2
4
5
4
Average
12.3
117,654
323.7
192,533
368,047
150,822
12,000
580,000
174,333
1.64
651
9.6
1,687
1,390
199
0.64
2,393
10
17.5
24,922
4,554
290
0.48
29.2
199
Range
Low
11.4
3,826 240
-
110,000 314
11,837 547
2,150 394
-
-
35,000 420
0
-
1.9
-
880 1
0.01
0.009
22 9
3.0
9.0
1,087 76
375 15
30
0.00147
0.83
19.05
High
13.2
,000
-
,600
,000
,000
-
-
,000
3.28
-
13.5
-
,900
469
1.28
,000
19
26
,800
,100
550
1.51
82
500
334
-------
TABLE 10 (con't)
Parameter
No. of
Observations
Range
Average
Low
Phosphates
Potassium
Selenium
Silica
Silver
Sodium
Sulf ates
Zinc
Specific Weight
Flash Point
Ash Content
2
1
2
2
4
1
1
6
1
1
1
7,
2,
5,
8,
12,
6,
>
500
010
30.5
325
2.3
455
300
791
10.08
210°F
14.36%
2,100 12,900
-
8.7 52
3,300 7,350
0.5 5
-
-
25.95 39,000
-
-
.3
.6
*A11 values in milligrams/liter unless specified otherwise.
Pollution Control Practice and Costs
Major potential alternatives
for air pollution control at drum
reclamation furnaces are: 1) direct
combustion in the 650°C (1200°F) to
900°C (1650°F) temperature range
with an afterburner residence time
of 0.3 to 0.7 seconds; 2) catalytic
oxidation in the 315°C (600°F) to
510°C (950°F) range; 3) sorption us-
ing activated carbon, silica gel or
other materials; and 4) scrubber sys-
tems. Worldwide, the direct combus-
tion and afterburner method is the
one most frequently used. Other
methods usually are adjuncts to this
primary method.
All U.S. reconditioning drum
furnaces have afterburners to con-
sume particulates and organics not
burned in the furnace main chamber.
About 37 percent of drum burning fa-
cilities use air pollution control
equipment and techniques in addition
to afterburners. Typically, equip-
ment includes scrubbers, packed tow-
ers, baghouses, and dust collectors.
Some methods used to optimize exist-
ing equipment are automatic ducts
to regulate air flow, and entry and
exit air curtains for better and
more efficient combustion.
For washing plants, about 55
percent have in-place or, under con-
struction their own wastewater treat-
ment plants. Table 11 lists equip-
ment used by reconditioners, some of
which is related to water pollution
control. Other equipment mentioned
in the NABADA survey was coalescing
plate separators for oil removal,
oil skimmers (often the endless belt
or rope type), and packaged waste-
water treatment systems. Only two
manufacturers have sold a significant
number of pollution control devices
to the industry: one is a diatoma-
ceous earth filter, and the other is
a dissolved air flotation rinse wa-
ter clarification system.
Many reconditioners fabricate
their own pollution control systems
as opposed to using commercially a-
vailable products. Operating pro-
cedures such as preflushing, stream
segregation, and cascading water use
are important adjuncts to pollution
control equipment.
335
-------
TABLE 11
WASHING PLANT RECONDITIONING EQUIPMENT
Equipment Type
Percentage of Reconditioners
Using Such Equipment
Flushomatics(TM)
Caustic Filters
Vacuum Drying Systems
Automated Flushers
Automated Upenders
Oil/Water Separators
Screens for Gross Particle
Removal
Dissolved Air Flotation
Flocculation/Sedimentation
Coagulation Feeders
26.8
31.7
73.2
48.8
80.5
61.0
56.1
17.1
34.1
7.3
Table 12 lists new drum prices
for the year 1977 and 1978, as well
as prices for March of 1979 and 1980.
Figures were taken directly from the
M34K reports of the Bureau of the
Census. Prices of new and recondi-
tioned drums also were part of the
NABADA survey. These data are re-
ported in Table 13. Note that the
NABADA survey prices do not reflect
drum gage or condition. Figures,
however, compare well with those in
Table 12.
A typical reconditioner pays a-
bout $400 per month in sewer sur-
charges. The bases for surcharges
at several locations are shown in
Table 14.
Almost 90 percent of water used
by reconditioners is purchased from
local public or private water dis-
tribution systems at an average cost
of $0.23/1000 liters ($0.86/gallons).
Other utility costs, principally gas
and electricity, average approxi-
mately $66,000 per year for a typi-
cal plant.
Average residue disposal costs
are $0.15 to $0.17 per reconditioned
drum.
Based upon the recent NABADA
survey, it is estimated that U.S.
reconditioners have about $12,700,000
of installed pollution control equip-
ment. Of this amount about
$9,100,000 is undepreciated (based
upon 10-year life and straight-line
depreciation). Hence, it is clear
that most of the equipment is fairly
new. Table 15 identifies types of
pollution control equipment along
with estimated capital costs and
depreciation.
The NABADA survey determined
that a typical reconditioning facil-
ity spends about $45,000 annually
for operations and maintenance re-
lated to pollution control. Table 16
shows the distribution of these ex-
penditures. Most of the costs for
afterburners are for supplemental
fuel.
A question on the NABADA survey
asked for approximate unit costs per
drum for pollution control. Re-
sponders claimed costs of $0.38/drum
for washing and $0.35/drum for burn-
ing. An attempt was made to verify
this figure in the following way.
Component costs are: capital amor-
tization, operations and maintenance
costs, and residual disposal costs.
336
-------
TABLE 12
PRICES OF NEW DRUMS
(Bureau of the Census Survey)
Tight Head
18 gage and heavier
19 and 20 gage**
PRICE (in dollars)*
March
1980
18.07
18.39
March
1979
17.26
16.86
Year
1978
16.51
16.33
Year
1977
15.50
14.43
Open Head
18 gage and heavier
19 and 20 gage**
22.37
16.82
19.33
15.60
18.42
14.71
17.77
13.97
* Drum price is at the point of production. It includes the
net sales price, f.o.b. plant, after discounts and allowances,
exclusive of freight charges and excise taxes.
** Includes 20/18 gage drums.
TABLE 13
PRICES OF NEW AND RECONDITIONED DRUMS
(NABADA Survey)
Drum Type
New Tight Head
New Open Head
Reconditioned Tight Head
Reconditioned Open Head
Laundry/Service Fee
Mean
Price ($)
17.47
19.42
11.74
11.89
5.78
Standard
Deviation
2.66
4.31
1.33
1.79
1.18
Range
Min . Max .
13.50 27.00
15.00 34.00
9.00 15.19
9.75 15.50
4.00 9.80
337
-------
TABLE
14
BASKS 1'OK SEWERAGE SURCHARGES
Volume Cost
Factor
(C/1000 gal)
26.7
-
11
92
69. 3
-
BOD Cost Factor
( C/pound )
less
exc lusion
(mg/1)
-
4.3 305
3.0 0
5.6 300
7.9 250
4.6 240
TSS
2
3
7
5
8
3
. 0
. 3
. 9
.6
. 6
. 1
Cost Factor
( C/pound )
less
exclusion
(mg/1)
0
400
0
300
250
300
COD Cost Other
TKN Cost Factor Factor Costs
(C/pound) (C/pound)
less
exclusion
(mg/1
2.0
-
- Sewer Connec-
tion Charge =
$4. 53/mo.
- - Minimum Bill=
$25.00/mo.
-
5.5 25 - If a pollutant
25.4
is less than
80% of the ex-
clusion value,
then there may
be a credit a-
gainst the to-
tal surcharge.
Flat infiltra-
tion - inflow
charge=25. 4
-------
TABLE 15
POLLUTION CONTROL EQUIPMENT
(in thousands of dollars)
Equipment Type
Incinerators/Afterburners
Dust Collectors/Scrubbers
Paint and Spray Controls
Water Pretreatment &
Treatment
Tanks & Clarifiers
Dissolved Air Flotation
Auxiliary Water Pollution
Equipment
Water Filtration
Evaporator
TOTAL
Capital Cost
5,100
400
970
4,740
830
430
61
170
34
12,735
Undepreciated Cost
3,600
220
580
3,620
580
340
44
138
17
9,139
TABLE 16
OPERATIONS AND MAINTENANCE
COSTS FOR POLLUTION CONTROL
(in thousands of dollars)
0 & M Category
Incinerators/Afterburners
Dust Collectors/Scrubbers
Paint and Spray Controls
Water Pretreatment & Treatment
Tanks & Clarifiers
Dissolved Air Flotation
Auxiliary Water Pollution Equipment
Other
TOTAL
Annual 0 & M Cost
1,880
28
148
2,840
306
142
3
107
5,454
339
-------
Installed capital of $12,735,000
(from Table 15) at 8 percent inter-
est (from 1977) for 10-year useful
life was used to allocate amortiza-
tion allowance. For 41,204,000
drums, the per drum cost is $0.067.
Operations and maintenance costs
from Table 16 yield a per drum cost
of $0.132. Residue disposal costs
for landfilling cited earlier were
about $0.16. The total of these
three elements is $0.36/drum.
Values cited by reconditioners
at technical forums have been up-
dated to current costs. These are
$0.37/drum for U.S. reconditioners
and $0.44/drum for Japanese. Hence,
it appears that NABADA survey fig-
ures are accurate. Additionally, it
implies that reconditioners have
good insight into various cost
components.
References
1. U.S. Census of Wholesale Trade:
1972. Commodity Line Sales,
Report WC 72-L, U.S. Department
of Commerce. November 1975.
Current Industrial Reports.
Steel Shipping Drums and Pails,
Bureau of the Census,
M34K(79)-12 Report.
U.S. Department of Commerce.
February 1980.
Checchi Report for the National
Barrel and Drum Association,Inc.
Checchi and Associates,
Washington, D.C. 1973.
Danielson, John A., Editor.
Air Pollution Engineering Manual
(Second Edition) , Air Pollution
Control District, County of
Los Angeles, Report AP-40, U.S.
Environmental Protection Agency,
Research Triangle Park,
North Carolina, May 1973.
Wilkinson, R.R., G.L. Kelso, and
F.C. Hopkins. State-of-the-Art
Report: Pesticide Disposal
Research. EPA-600/2-78-183,
U.S. Environmental Protection
Agency, Cincinnati, Ohio,
September 1978.
340
-------
BENCH SCALE ASSESSMENT OF CONCENTRATION
TECHNOLOGIES FOR HAZARDOUS AQUEOUS WASTE TREATMENT
Alan J. Shuckrow
Andrew P. Pajak
Touhill, Shuckrow-and Associates, Inc.
Pittsburgh, Pennsylvania 15237
ABSTRACT
This paper describes portions of the experimental phase of an ongoing
program to evaluate the applicability of several concentration technologies
to treatment of hazardous aqueous wastes. Studies are being carried out at
the Ott/Story site in Muskegon, Michigan using groundwater which has been
severely contaminated by numerous organic compounds. Specifically, bench
scale laboratory treatability studies including activated carbon adsorption,
resin adsorption, aerobic and anaerobic biological treatment, and stripping
are under investigation. Most treatment technologies studied to date have
been moderately effective in reducing the levels of organic contamination.
However, a process train consisting of granular activated carbon adsorption
followed by activated sludge treatment can achieve high levels of treatment
for short periods of time. Virtually complete removal of organic priority
pollutants can be achieved under some conditions.
Introduction
The objective of the project
discussed in this paper is to iden-
tify and evaluate technologies for
concentrating hazardous constituents
of aqueous waste streams. As report-
ed previously (2,3), this objective
is being met through a multi-phased
program involving literature review,
desktop evaulations, and laboratory
bench scale treatability studies.
Early in this project it was
concluded that the major hazardous
waste problem facing the public sec-
tor is contamination from waste dis-
posal sites, specifically leachates
and contaminated ground and surface
waters. Wastewater compositions,
however, are diverse and often vary
at a given site. Moreover, few ac-
tual applications of treatment tech-
nology to this type of hazardous
aqueous waste problem exist.
On the basis of an extensive
literature review and desktop analy-
sis of concentration technologies,
the following unit processes were
identified as having potential broad
application to the aqueous contami-
nation problems associated with
waste disposal sites:
Biological treatment
Chemical coagulation
Carbon adsorption
Membrane processes
Resin adsorption
Stripping
These, however, must be supplemented
with ancillary processes such as
sedimentation and filtration.
Since hazardous aqueous waste
streams vary widely in composition
and often contain a diversity of
constituents, in general, no single
unit process is capable of providing
optimum treatment. Rather, arrange-
ment of individual processes into
process trains is necessary to
341
-------
achieve high levels of treatment in
the most cost-effective manner.
Thus, having identified the most
promising unit concentration tech-
nologies, the next step was to for-
mulate process trains which com-
bined technologies in a fashion
which would provide broad spectrum
treatment capability. The objective
was to identify process trains which
would produce high quality effluents
when applied to the wide range of
waste stream compositions likely to
be encountered. Five such process
trains incorporating the selected
concentration technologies were for-
mulated. Details describing the se-
lection of these unit processes and
the five process trains previously
have been reported (2) .
Subsequent to the desktop tech-
nology evaluations, experimental
treatability studies were initiated.
These experimental investigations
are described below.
Bench Scale Studies
Several alternative approaches
for conducting laboratory evalua-
tions of the unit process identified
earlier were considered. Actual,
rather than synthetic, wastewater
was the desired feedstock because
the complexity of wastewater compo-
sition at the various hazardous
waste problem sites could not be
duplicated adequately by a synthetic
wastewater.
On the basis of several factors,
including availability of quantita-
tive data describing problem nature
and magnitude, absence of pending
litigation which would limit infor-
mation transfer, cooperative rela-
tionships between current site own-
ers and the regulatory agencies, and
an ongoing feasibility study to
identify and evaluate clean-up tech-
niques being conducted by the state
agency, the Ott/Story Chemical Com-
pany site (now owned by Cordova
Chemical Company) in North Muskegon,
Michigan was selected. Groundwater
in the area had been contaminated by
the disposal and poorly controlled
storage of chemical production waste-
waters by previous facility owners.
Because of possible changes in waste-
water composition during shipment
and prolonged storage, and logistics
problems associated with shipment of
the substantial quantities of waste-
Vcter required, a treatability lab-
oratory was established at the site.
The characteristics of ground-
water at the Ott/Story site are
shown in Table 1. In addition to
these major pollutants, more than 70
other organic pollutants were detect-
ed at concentrations > 10 yg/1. Or-
ganic compounds are the primary pol-
lutants; heavy metal concentrations
are believed not to be a problem.
TABLE 1. GROUNDWATER QUALITY
AT OTT/STORY SITE
8-12
5400 mg/1
400 - 1500 my/I
64 raq/1
*ND - 32,500 ug/l
5-6 ,570 , q/1
60 - 19,850 .g/1
5-14,280 , g/1
350 - 111,000 ,g.
ND - 7,370 g/1
• 5 - 1,590 ng/1
5 - 5,850 .g/1
/I
In designing the bench scale
studies, it was decided to first ex-
amine unit processes and subsequent-
ly, as a data base was developed, to
evaluate process trains. The bench
scale evaluation approach is out-
lined below:
1. Pretreatment investigations
a. neutralization
fc. chemical coagulation and
precipitation
c. solids/liquid separation
d. air sparging
2. Batch studies
a. carbon adsorption
b. resin adsorption
3. Batch sequential studies
a. air sparging followed by
carbon adsorption
b. carbon adsorption followed by
342
-------
air sparging
c. carbon adsorption followed by
resin adsorption
4. Continuous flow studies
a. steam stripping, packed bed
column
b. granular activated carbon
columns
c. biological treatment
d. biophysical treatment
Results from pretreatment in-
vestigations, batch studies, and
batch sequential studies have been
reported previously (1). To brief-
ly summarize:
• Results on major priority pollut-
ant removals measured in these
studies are summarized in Table
2.
• Carbon adsorption reduced almost
all organic priority pollutants
to less than GC/MS detection lim-
its.
• Resin sorption proved to be only
slightly less effective than car-
bon sorption. Most organic pri-
ority pollutants were reduced to
below detection limits; all were
reduced by at least 98%.
• All volatile organic priority
pollutants were reduced to less
than detection levels by air
stripping. Removals for other
organic priority pollutants
ranged from 4 to 96%.
. Carbon treatment of the air
sparged groundwater generally
resulted in reduction to less
than detection limits for the
remaining organic priority pol-
lutants. All were reduced by
more than 98%.
• Despite good removals of priority
pollutants, a significant resid-
ual TOC (301-455 mg/1) was mea-
sured in all treated samples.
This residual represents uniden-
tified non-priority organic pol-
lutants.
The remainder of this paper de-
scribes continuous flow studies
which have been conducted or are in
progress. Because study efforts
will continue until mid-1981, the
results are incomplete and should be
considered as preliminary subject to
further investigations.
Steam Stripping
Continuous flow, steam strip-
ping experiments were conducted us--
ing a 76.2 cm by 4.8 cm (ID) packed
column. The apparatus was operated
at feed stream flow rates of 40 to
80 ml/min, overhead (condensate)
flow rates of 3.5 to 9.2 ml/min
(overhead to feed flow ratios of
0.064 to 0.14), no reflux flow, feed
stream TOC concentrations of 480 to
610 mg/1, and durations of 1 to 4
hours after establishing steady
state operation within the available
operational controls.
As indicated in Figure 1, which
illustrates the results of these
studies, TOC concentration in the
bottoms ranged from 300 to 400 mg/1
and appeared to be independent of
the overhead;feed ratio. Average
TOC reduction between feed and bot-
toms was 34%. These observations
indicate a major constraint associ-
ated with steam stripping - it is
necessary to further treat a bot-
toms waste stream having a flow on-
ly slightly less than the feed flow.
It should be noted that maintaining
steady state operation of the appa-
ratus proved to be very difficult.
The minimum overhead to feed flow
ratio which could be achieved was
0.064, significantly higher than
the desirable range of 0.02 to 0.05.
It was concluded that, because bot-
toms TOC concentration appeared to
be independent of system flow rates
and feed TOC levels over the ranges
examined and because additional
treatment of the bottoms is neces-
sary, steam stripping would likely
be a costly, yet only moderately
effective, concentration process.
Adsorption - Granular Activated
Carbon and Resin
Continuous flow granular acti-
vated carbon (GAC) studies were con-
ducted at the unadjusted groundwater
pH (9.3 to 10.0) using Filtrasorb
343
-------
TABLE 2. TOC AND MAJOR ORGANIC POLLUTANT REMOVALS
DURING BATCH SEQUENTIAL STUDIES
Study 1 (mg/1)
Compound
TOC
Benzene*
Benzole Acid
Camphor
Chloroform*
1 , 1-Dichloroethane*
1 , 2-Dichloroethane*
1 , 1-Dichloro-
ethylene*
Dimethylaniline
Ethylaniline
Methylene Chloride*
Toluene*
Raw
Waste
638
7.8
0.17
4.0
1.4
1.2
111
0.06
17.0
3.3
0.06
2.6
Resin
Sorption
Effluent
455
0.17
ND
0.04
ND
ND
0.23
ND
0.25
ND
ND
ND
Carbon
Sorption
Effluent
332
0.01
0.18
ND
ND
ND
0.01
ND
ND
ND
ND
ND
Study 2 (mg/1)
Raw
Waste
720
5.3
0.30
3.9
2.0
1.6
14
1.0
15.0
3. 8
0.07
3.6
Air
Sparge
Effluent
641
ND
0.02
0.47
ND
ND
ND
ND
0.61
0.60
ND
ND
Sparge plus
Carbon
Sorption
Effluent
301
ND
ND
0.01
ND
ND
0.01
ND
0.08
ND
NA
ND
(Note: 27 organic pollutants detected at 0.01 to 0.31 mg/1;
none found in effluents.)
ND - Not Detected at detection limit of 0.010 mg/1
NA - Not Analyzed
* - Priority Pollutant
300 GAC. The studies used three or
four glass columns arranged in se-
ries with sampling ports located at
the influent and effluent ends of
each column. Each column is 122 cm
by 2.54 cm (ID) with 91.4 cm of GAC
in each column. The system was op-
erated in a downflow mode at a load-
ing rate of approximately 1.35 l/m2s
(2 gpm/ft2). This provided an emp-
ty bed contact time (EBCT) of ap-
proximately 15 minutes per column.
Influent TOC concentration varied
substantially, ranging from 316 to
950 mg/1.
Results of a typical study are
illustrated in Figure 2. Results
are reported as removal percentage
rather than effluent concentration
because of influent TOC fluctuations.
Generally, after only 3 to 10 bed
volumes (BVs), TOC removal decreased
to 50%. TOC leakage reached 90% af-
ter about 200 to 240 BVs were pro-
cessed and continued at this level
until up to 500 BV had been pro-
cessed. As would be expected, efflu-
ents from columns 2 and 3 are inter-
mediate between columns 1 and 4.
Available GAC column performance
data corresponds to batch sorption
study data.
Continuous flow resin
344
-------
investigate a combined sorption/bi-
ological system. The rationale em-
ployed was to attempt to utilize
the sorbent to protect the biologi-
cal system from toxic materials.
Thus, the sorbent could be allowed
to leak relatively high concentra-
tions of organics which would be
degraded in the subsequent biologi-
cal process. Two such process
trains were investigated:
1. FS 300 granular activated carbon
followed by activated sludge
(GAC/AS), and
2. XE-347 resin followed by acti-
vated sludge (RES/AS).
Operating conditions for the
GAC/AS process trains are summa-
rized below:
. GAC - three 91.4 cm by 2.54
cm columns in series,
0.0815 m3/rn2/d (2 gpm/ft2)
hydraulic loading, and
34 min EBCT
• AS - one liter reactors
with hydraulic retention
times (HRT) of 6 to 16 hr;
GAC effluent neutralized
and nutrients added; mixed
liquor suspended solids
ranged from about 2000 to
8000 mg/1; and 5000 mg/1
PAC in aeration chamber
during PAC studies.
Performance of this process
train under various conditions is
illustrated in Figure 3. It should
be noted that the GAC process was
operated approximately 6 hr/day,
five days per week whereas the AS
process was operated continuously.
This necessitated collection and
storage of GAC effluent for 1 to 3
days to maintain continuous feed to
the AS reactors. Some decrease in
TOC levels were noted during- the
storage periods. Losses most like-
ly were due to volatilization and/
or biological degradation. Thus,
it is assumed that similar reduc-
tions would have occurred in the AS
reactors. The following observa-
tions have been made on the basis
of data presented in Figure 3:
• GAC pretreatment of raw
groundwater permits develop-
ment of a culture of aerobic
organisms capable of further
treating GAC effluent. In
excess of 95% TOC removal
can be achieved by this pro-
cess during the period which
GAC removal of TOC exceeds
30%. After this initial pe-
riod, process train perfor-
mance declines as GAC per-
formance declines. Effluent
TOC could be maintained at
< 100 mg/1 only for short
time periods and only when
GAC performance was at its
peak.
• These data indicate that
some fraction of TOC which
initially is sorbed by GAC
begins to leak through the
system after a short period
of operation. This fraction
of TOC which leaks through
the GAC system is not toxic
to AS but does not appear to
be removed or reduced either
biologically or by air
stripping associated with AS
aeration.
• Operation of the AS process
at HRTs ranging from 6 to 16
hr, with or without powdered
activated carbon added to
the aeration chamber seems
to have little impact on
process performance (based
upon TOC removal).
• Overall system performance
was maintained at 75-85%
TOC removal (effluent TOC of
100 to 185 mg/1) for about
21 days (46 retention times
for AS and > 110 BVs for
GAC) .
• Visual observations and typ-
ical mixed liquor analyses
(MLSS and MLVSS) suggest
that the biological systems
could survive in and utilize
GAC pretreated groundwater
even after GAC performance
had declined to about 10%
TOC removal. However, AS
effluent contained about
34S
-------
4000
3000 -
en
§
O
o
2000-
1000
Overheads
O
Bottoms
AA
006 008 0.10 012
Overhead : Feed Ratio
014
Figure 1. Results of steam
stripping studies.
adsorption studies were conducted
using XE-347 carbonaceous resin.
Three columns similar to those used
for GAC studies were charged with
792 to 835 cm3 of resin and were
operated at loading rates of 2.95
to 3.79 BV/hr. EBCT ranged from
16 to 20 min. TOG breakthrough
data for one representative study
are shown in Figure 2 . Breakthrough
characteristics were similar to
those of GAC except that XE-347 TOG
removal declined more rapidly. TOG
removal diminished to < 50% after
about 5 BV were processed and ap-
peared to stabilize at about 10% for
at least 120 BV.
Biological Treatment
Several attempts were made to
acclimate an activated sludge cul-
ture to raw groundwater using 350 ml
reactors. All attempts, however,
100 150 200 250
BED VOLUMES
Figure 2. TOO removal by adsorption.
were minimally successful. Neither
a conventional activated sludge nor
a commercial microbial culture could
be acclimated. Slight loading fluc-
tuations encouraged growth of a
light colored, filamentous biomass
which settled poorly. About 60% TOC
reduction was achieved; however,
stripping due to aeration appeared
to account for about two-thirds of
this removal. Addition of trace
elements and nutrients, and pH ad-
justment to pH 7.0 to 7.5 did not
aid acclimation to raw groundwater.
Addition of powdered activated car-
bon (PAC) at aeration chamber con-
centrations of about 10,000 mg/1
also did not aid acclimation to raw
groundwater or improve TOC removal
or mixed liquor appearance.
Adsorption/Aerobic Biological
Treatment
Because of the apparent toxici-
ty of the groundwater to a biologi-
cal treatment system and the rapid
breakthrough of TOC in adsorption
systems, it was decided to
346
-------
1000n
800-
600-
01
E
o
o
400-
200 -
virgin carbon added
virgin resin added
act- Slud9e unit 1 eff '••*•,
adsorp. unit effl.
10
20
30
DURATION (days)
Figure 3. Performance of adsorption/activated sludge process
200 mg/1 TOC at this time.
As shown in Table 3, limited
analyses made during operation of
the GAC/AS train suggest that high
levels of organic priority pollut-
ant removals can be attained even
with effluent TOC concentrations of
100 to 200 mg/1. Almost all of the
organic priority pollutants detect-
ed in raw groundwater were removed
consistently to less than the level
of detection (0.01 mg/1) by the pro-
cess train. One consistent feature
of these data and previous GC/MS
analyses from batch carbon adsorp-
tion studies is the early leakage
of 1,2-dichloroethane. A few other
compounds (benzene, methylene chlo-
ride, and toluene) also were detect-
ed to have broken through the car-
bon in some batch and continuous
flow studies. Acid and base-neu-
tral ex tractable compounds generally
did not break through the GAC pro-
cess. Data in Table 3 also indi-
cate that the activated sludge pro-
cess completely removed the few or-
ganic priority pollutants leaking
through the GAC system even though
TOC removal declined. The contin-
ued removal of organic priority pol-
lutants may be due to stripping.
An effort was made to determine
if specific organic priority pollut-
ants were concentrated in the bio-
logical sludge. In the one sludge
sample collected to date, no organic
priority pollutants were detected at
a 0.01 mg/1 detection level.
An off-gas sample from the
aerated reactor was collected using
a cold trap (acetone and dry ice) to
condense off-gas vapors.
Air flow to the reactor was approx-
imately 2 1/m and the collection
period was 4 hours. The following
organic priority pollutants were de-
tected in this sample:
Methylene Chloride 1.02 yg/1 air
1,2-Dichloroethane 1.04 ug/1 air
Benzene 0.25 yg/1 air
Perchloroethylene 0.125 yg/1 air
Toluene 0.088 wg/1 air
347
-------
TABLE 3. TOC AND PRIORITY POLLUTANT DATA FOR
GRANULAR ACTIVATED CARBON/ACTIVATED
SLUDGE PROCESS TRAIN (mg/1)
Collected on
Day 2*
Compound
TOC
Total Cyanide
CNA
Total Phenol
Methylene chloride
1 , 1-Dichloroethene
1 , 1-Dichloroethane
Trans-1 , 2-dichloro-
ethane
Chloroform
1 , 2-Dichloroethane
1,1, 1-Trichloroethane
Trichloroethylene
Benzene
1 , 1 , 2-Trichloroethane
Perchloroethylene
Toluene
Chlorobenzene
Phenol
2-Chlorophenol
2 , 4-Dichlorophenol
1 , 2-Dichlorobenzene
Dibutyl phthalate
Raw
Ground-
water
2
1
2
0
9
72
7
0
1
0
0
2
0
0
0
0
0
637
NA
NA
NA
.1
.6
.4
.06
.8
.6
.06
.2
.11
.49
.3
.23
.025
.040
.010
.085
ND
GAC
Effl.
380
NA
NA
NA
0.029
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
Collected
Days 9 and
Raw
Ground-
water
0
<0
16
14
0
0
0
0
25
0
0
1
0
1
0
0
0
0
0
0
929
.11
.05
.06
.17
.04
.70
.39
.03
.5
.07
.9
.97
.29
.028
.036
.010
.077
ND
GAC
Effl.
0
<0
<0
0
0
0
0
1
0
0
0
604
.21
.05
.16
.01
.01
.02
ND
.06
.4
.04
ND
.02
ND
ND
.05
ND
ND
ND
ND
ND
ND
on
10
As
Effl.
90
0.23
<0.05
<0.10
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
0.05
Collected on
Day 17
GAC
Effl.
770
0.23
<0.05
<0.10
0.16
ND
ND
ND
ND
0.05
ND
ND
ND
ND
ND
0.01
ND
ND
ND
ND
ND
ND
AS
Effl.
183
0.20
<0.05
<0.10
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
NA - Not Analyzed
ND - Not Detected
No other priority pollutants detected
* - Refers to Figure 3.
As indicated in Figure 3, on
day 64, the pretreatment process was
changed from GAC adsorption to resin
adsorption (carbonaceous resin
XE--347) . The operating conditions
for two resin adsorption pretreat-
at 0.01 mg/1 detection limit
ment studies were as follows:
• 3 columns in series
• columns were 2.54 cm diameter
. total BV = 792 to 835 cm3
• downflow operation at 41 to
50 ml/min (2.95 tO 3.79 BV/hr)
• EBCT ranged from 20 to 16 min
348
-------
Results, as shown on Figure 2, in-
dicate that TOC removal diminished
to < 50% after ^ 5 BV were loaded
and appeared to stabilize at 10-20%
for at least 120 BV. The charac-
teristics of the TOC breakthrough
curve are similar to those of GAC
except that XE-347 TOC removal de-
clined more rapidly.
As shown in Figure 3, the AS
units loaded with resin treated
groundwater were not able to reduce
TOC levels to < 100 mg/1 even for a
short period after virgin resin was
placed on line. Performance at 4
hr and 8 hr HRTs did not differ.
Adsorption/Anaerobic Biological
Treatment
With the thought of minimizing
potential air pollution due to
stripping of volatile organics dur-
ing biological treatment, an at-
tached-growth upflow anaerobic fil-
ter currently is being investigated.
The process is preceeded by GAC
pretreatment. Operating conditions
for the process train are as flows:
• GAC column - 91.4 cm by
2.54 cm ID glass
• GAC influent flow -
1.75 ml/min
. GAC EBCT - 3.8 hr
• UAF column - 70.5 cm by
4.76 cm ID glass
• UAF organic loading rate -
0.423 to 0.847 kg TOC/m3/d
(26.4 to 52.9 Ib TOC/1000
ft3/d)
. UAF EBCT - 13.1 to 22.8 hr
• UAF temperature 35°C
The UAF reactor was filled with ce-
ramic berl saddles and then with
sludge from a well operated munici-
pal wastewater sludge anaerobic di-
gester. The UAF reactor initially
was fed raw sewage for eight days
prior to converting to GAC pre-
treated groundwater.
Performance data on the GAC/
UAF process train are shown in Fig-
ure 4. TOC removals by the various
steps in the process train are sum-
marized below:
removal
GAC/UAF train
GAC process
UAF process
average range
66%38-81%
31% 10-46%
50% 12-67%
Gas production during the initial
month of operation has averaged
505 ml/g TOC fed. Figure 4 indi-
cates that UAF effluent TOC in-
creased as TOC leakage from the GAC
pretreatment process increased.
However, TOC removal and gas produc-
tion data cannot be related to or-
ganic loading rate or other opera-
tional data such as sludge pH,
sludge total alkalinity, or vola-
tile acids production. In an at-
tempt to bring the pH into a range
reported to be most optimal (pH 7.2
to 7.6), the GAC influent pH was
adjusted to ^ pH 7.0 to 7.5. This
has had no apparent effect on per-
formance. Initial results indicate
that about 40% (117 mg/1) of the
TOC in UAF effluent can be removed
by air stripping.
Recently, this process train
was modified by adding activated
sludge treatment following the UAF
process. A comparison is being
made of TOC removal by stripping and
biological activity.
Summary
Desktop evaluations of candi-
date concentration technologies for
treatment of hazardous leachates and
contaminated groundwater were per-
formed in order to select and prior-
ity order technologies for labora-
tory study. These evaluations were
based upon the characteristics of
each technology and upon its known
effectiveness in concentrating ma-
terials in specified chemical
classes.
Technologies identified as
most promising for experimental
study were:
1. adsorption - carbon and resin,
2. biological treatment,
3. biophysical treatment,
4. chemical coagulation,
5. membrane processes - reverse
osmosis and ultrafiltration,
6. stripping - air and steam.
349
-------
0>
E
o
o
800 -i
600-
400-
200-
converted from sewage to groundwater
GAC infl.
-anaerobic filter effl.
m= — wm
1
i
5
1
g
i
13
I
17
1
21
I
25
\
29
1
33
DURATION (days)
Figure 4. Performance of GAC/UAF process
Subsequently, laboratory experiments
on contaminated groundwater at the
Ott/Story Chemical Company site in
Muskegon, Michigan were initiated.
Laboratory results indicate good
removals of volatile priority pol-
lutants by batch air stripping. In
addition batch carbon and resin
sorption reduce all organic priori-
ty pollutants by greater than 98%.
However, high effluent TOC levels
have been measured subsequent to all
applied treatments. This TOC repre-
sents organic contamination by non-
priority pollutants.
Steam stripping in a packed
column resulted in concentration of
TOC in the overhead condensate
stream. However, within the range
of feed flow and overhead flow rates
investigated, TOC of the bottoms
could not be reduced below about 400
mg/1. Overhead TOC approached 4000
mg/1 at an overhead flow rate of
approximately 6% of the feed flow.
FS 300 granular activated car-
bon (GAC) employed in continuous
flow small size columns was not cap-
able of sustaining high levels of
TOC removal. TOC removal declined
to < 50% after processing < 5 BV at
hydraulic loading rates of ^ 1.85
gpm/ft2. Within 100-160 BV loaded,
TOC removal declined to 10 to 15% and
remained at this level for up to 200
BV. GAC adsorption was capable of
achieving high levels of organic pri-
ority pollutnat removals even when
TOC removal had declined to 35% and
effluent TOC levels were ^ 600 mg/1.
In both batch and continuous flow ad-
sorption studies, some volatile pri-
ority pollutants were detected in the
effluent. None of the acid or base-
neutral extractable organic priority
pollutants detected in the raw
groundwater were found in GAC efflu-
ent after processing up to 71 BV.
Continuous flow small size resin
adsorption (XE-347 carbonaceous resin)
studies demonstrated TOC breakthrough
characteristics similar to those for
GAC adsorption. However, TOC break-
through occurred more rapidly.
350
-------
GAC pretreatment of raw ground-
water permits development of a cul-
ture of aerobic organisms capable of
further treating GAC effluent. In
excess of 95% TOC removal can be
achieved by this process during the
period which GAC removal of TOC ex-
ceeds 30%. After this initial peri-
od, process train performance de-
clines as GAC performance declines.
Several organic priority pollutants
were detected in off-gas from AS
reactors; these included methylene
chloride, 1,2-dichloroethane, ben-
zene, tetrachloroethylene, and tol-
uene. No organic priority pollut-
ants were detected (at a detection
limit of 10 i-g/l) in an AS biomass
sample.
Anaerobic treatment (upflow
anaerobic filter, UAF) of GAC pre-
treated groundwater is possible.
However, UAF performance appears to
decline as GAC performance declines
(although changes in UAF organic
loading rate do not appear to ef-
fect UAF performance). Initial re-
sults indicate that about 40% of the
TOC in UAF effluent can be removed
by air stripping. Overall the
GAC/UAF process train performs more
poorly than the GAC/AS process train
with an upper TOC removal limit of
81%. Removal has averaged 66% and
ranged from 38 to 81%.
Acknowledgemen t
The work upon which this paper
is based was performed pursuant to
Contract No. 68-03-2766 with the
Environmental Protection Agency.
The cooperation and assistance
of Cordova Chemical Company, the
present owner of the Ott/Story site,
is greatly appreciated.
References
1. Pajak, A.P., A.J. Shuckrow, J.W.
Osheka, and S.C. James. 1980.
Concentration of hazardous con-
stituents of contaminated
groundwater. In: Proceedings
of the Twelfth Mid-Atlantic
Industrial Waste Conference.
Bucknell University, Lewisburg,
Pennsylvania.- pp. 82-87.
Shuckrow, A.J., A.P. Pajak, and
J.W. Osheka. 1980. Concentration
Technologies for Hazardous
Aqueous Waste Treatment. Inter-
im Report for Contract No.
68-03-2766, U.S. Environmental
Protection Agency, Cincinnati,
Ohio.
Shuckrow, A.J., A.P. Pajak, and
C.J. Touhill. 1980. Hazardous
waste concentration technologies.
In: Treatment of Hazardous Waste,
Proceedings of the Sixth Annual
Research Symposium.
EPA-600/9-80-011, U.S. Environ-
mental Protection Agency,
Cincinnati, Ohio. pp. 50-61.
351
-------
APPLICATION OF REMOTE SENSING TECHNIQUES TO EVALUATE
SUBSURFACE CONTAMINATION AND BURIED DRUMS
SHWRD Annual Research Symposium
Harold J. Yaffe
Nancy L. Cichowicz
Robert W. Pease, Jr.
The MITRE Corporation
Bedford, Massachusetts 01730
ABSTRACT
Several remote sensing techniques (ground-penetrating radar, electrical resistivity, metal
detection and seismic refraction) were employed to investigate subsurface chemical contam-
ination and buried drums at an uncontrolled hazardous waste site in Rhode Island. The
techniques were applied in conjunction with direct sample collection to support the selec-
tion of a long-term abatement alternative for the site. The results of the field investi-
gation are given and a comparison of the remote sensing techniques is presented. Recommen-
tations for accomplishing systematic investigations at other abap^oned hazardous waste
sites are also given.
INTRODUCTION
This paper describes the application of
several remote sensing techniques in con-
junction with direct sample collection at an
uncontrolled hazardous waste site to a) de-
termine the extent and nature of the buried
drum and subsurface cherxical contamination
problem and b) support the selection and de-
sign of a long-term abatement approach. The
use of the following remote sensing tech-
niques were also demonstrated:
• ground-penetrating radar
• electrical resistivity
• seismic refraction
• metal detection.
The focus here is on the techniques and
their results as opposed to the selection of
the preferred abatement alternative. The
latter is discussed in a separate technical
report [Cichowicz, et al. (1)J.
The uncontrolled hazardous waste dump
site which was investigated is located in
Coventry, Rhode Island, approximately 20
miles southwest of Providence. Ths site
encompasses approximately 7.5 acres of
cleared ground surrounded by woods and wet-
land in a relatively rural area of the state
(see Figure 1). An undetermined quantity of
chemicals had been placed into the ground
both by the burial of 55-gallon drums in
five separate locations and by direct dis-
charge into trenches (see Figure 2). A
swamp, located northwest of the site, is the
surface discharge area of chemicals leaching
from the dump. This swamp discharges to a
small pond which is a source of irrigation
water for a cranberry bog located approxi-
mately one mile from the swamp's outlet. To
date, no evidence of chemical contamination
in the pond has been found, based on sam-
pling conducted by the Rhode Island Depart-
ment of Environmental Management (DEM) and
the U.S. Environmental Protection Agency
(EPA), Region I.
State of Rhode Island officials were
alerted to the dumping activities by a fire
352
-------
Source: USGS Coventry Center, R.I. 7.5 minute quadrangle
Figure 1. Topographic niap of the hazardous
waste site in Coventry, Rhode Island.
and explosion in September 1977. A court
order issued in November 1977 prohibited the
property owner from continuing dumping ac-
tivities or otherwise altering the site.
From the end of 1977 to mid-1979, the DEM
conducted field investigations to quantify
the seriousness of the situation.
In October 1979, the DEM contracted
with The MITRE Corporation to conduct a sys-
tematic site assessment, and in April 1980
the investigation continued under funding by
the EPA/SHWRD. Although the investigation
of the Coventry site was conducted in two
discrete phases with separate project re-
ports, overall project continuity was main-
tained. Phase I funding was shared by the
DEM and EPA/SHWRD (the latter funded all
chemical analysis and the preliminary eval-
uation of abatement methods) and Phase II
was completely undertaken by EPA/SHWRD.
SITE INVESTIGATION
The ultimate purpose of the Phase I and
Phase II investigations was to support the
selection of one of the following long-term
abatement methods:
• site encapsulation
• leachate collection and treatment
• drum removal and chemical disposal
or the "no action" alternative.
The techniques employed for data col-
lection during the Phase I effort were:
electrical resistivity; metal detection;
installation of monitoring wells; and chem-
ical analysis of soil, ground water, and
surface water. The field methods employed,
data collected, conclusions drawn, and rec-
ommendations made to the DEM are documented
in the Phase I project report [Pease, et al.
(3)]. Although the extent of the problem
was defined and abatement options were pre-
liminarily evaluated, certain key pieces of
information (concerning the presence of
fracturing or contamination of the bedrock
and the condition and number of the buried
drums) needed to be ascertained before a
L£GEND
L-AND SURFACE CONTOURS (FEET ABOVE MSLI
00000 STONE FENCE
R\\\\\^ AREAS OF HIGH METAL CONTENT DETECTED NEA« GROUND SURFACE
Q VISIBLE METAL DRUMS
LAND SURFACE CONTOURS BASED ON USGS CONTOURS AN
WHERE DUMPING ACTIVITIES HAVE DISTURBED GROUND
I I ~T~
Figure 2. Outline of trench locations at
the Coventry site as determined by metal
detection.
353
-------
TABLE 1. TECHNIQUES USED IN PHASE II TO PROVIDE INFORMATION
NEEDED TO SELECT AN ABATEMENT ALTERNATIVE
Alternative
Additional Information
Required to
Select Alternative
Technique to
Obtain Information*
1. No Action
2 . Drum Removal and
Disposal (excavation,
testing, and proper
disposal of drums and
contents, and contami-
nated soils)
3. Site Encapsulation
(construction of imper-
meable barriers around
source of pollutants)
4. Leachate Collection and
Treatment
a. Limited Option
(interceptor trenches
constructed adjacent
to site walls)
b. More Complete Option
(interceptor trenches
constructed 600 feet
downgradient of site
walls)
• condition of source
(drums)
• state of nearby pond
• contaminant underflow at
swamp
• ultimate disposition of
all pollutants
• condition of source
(drums)
• condition of soil
• condition of source
(drums)
• condition of bedrock
• condition of source
(d rums)
• condition of bedrock
• same as above
• radar, exploratory
excavation
• additional wells,
chemical analysis of
soils and water samples
• radar, exploratory
excavation
• exploratory excavation,
chemical analysis of
soil samples
• radar, exploratory
excavation
• seismic refraction, core
drilling, deep wells
• radar, exploratory
excavation
• seismic refraction, core
drilling, deep wells
• same as above
Metal detection had previously been used to locate trenches; electrical resistivity to
delineate leachate plume. Radar could have been employed in lieu of or in conjunction
with metal detection, as recommended for other sites; potential radar effectiveness was
not known at the time of the initial survey.
permanent solution could be selected. The
Phase I report presented the abatement op-
tions, identified the necessary additional
information, and made recommendations for
immediate and near-term actions to protect
the public health and to collect additional
data.
The relationships among the abatement
methods, the additional information needs at
the conclusion of the Phase I study, and the
techniques employed in Phase II to obtain
that information are shown in Table 1.
Phase II was undertaken by MITRE with EPA/
SHWRD funding (ground-penetrating radar,
seismic refraction, bedrock sampling, and
chemical analysis) and by the DEM under in-
ternal state funding (exploratory excavation
of one trench). A project report [Cichowicz,
et al. (2)] which describes the use and
evaluation of the remote sensing techniques,
and a supplementary project report [Cicho-
wicz, et al. (1)] which describes the eval-
uation of abatement alternatives are cur-
rently being published by EPA/SHWRD.
REMOTE SENSING TECHNIQUES
Ground-Penetrating Radar
The technique of ground-penetrating
radar involves the repetitive propagation
354
-------
of short-time duration pulses of electromag-
netic energy in the radio frequency range
downward into the ground from a broad band-
width antenna on (within a few inches of)
the surface. Reflections from subsurface
interfaces are received by the antenna dur-
ing the off period of the pulsed transmis-
sion, processed electronically, and recorded
to yield a continuous profile of subsurface
conditions as the antenna/transmitter-re-
ceiver unit is moved across the ground sur-
face. The depth d to an interface, or the
surface of a "target" such as a metal drum,
is calculated from d = (vt)/2, where v is
the wave velocity (equal to c/^/e^", where c
is the velocity of light and er is the rel-
ative dielectric constant of the material in
which the wave is propagating) and t is the
pulse travel time.
Ground-penetrating radar has been used
in such applications as archeological sur-
veys, locating sewer lines and buried cables
prior to construction activities, and pro-
filing lake and river bottoms. The applica-
tion to locating buried drums of chemical
wastes is relatively new, and further re-
finements both in the technology and in data
interpretation are anticipated.
The field survey was conducted by the
equipment manufacturer, Geophysical Survey
Systems, Inc. (GSSI). The equipment used
was GSSl's Surface Interface Radar System 7.
The survey of the trench areas, representing
approximately two acres, took two days.
Following experimentation with two alterna-
tive antennas and center frequencies, GSSI
Model 3105AP operating at a center frequency
of 300 MHz and GSSI Model 3102 operating at
600 MHz, the latter was chosen for most of
the survey due to its improved spatial reso-
lution at shallower depths. The operating
depth varies approximately as the inverse
square of the frequency, all else being
equal.
One large trench (labeled the West
Trench in Figure 2) located by the metal de-
tection survey was surveyed with the 300 MHz
antenna set at a nominal depth of 25 feet,
later calibrated at 24.4 feet, based on
average soil conditions. The other trenches
(labeled Northwest, Northeast, and South)
were subsequently surveyed using the 600 MHz
antenna set at a nominal depth of 12.5 feet.
The survey was conducted according to a
rectangular grid. All trenches were sur-
veyed longitudinally by using parallel radar
transects at spacings of ten feet. Trans-
verse transects, or cross-cuts, were made at
intervals of 20 feet for the Northeast Trench
Trench and 40 feet for the West and North-
west Trenches. The antenna unit was pulled
along each transect manually, and the data
recorded by wire connection with equipment
located in a stationary van on the site,
which also served as the power source. The
major equipment components were a control
unit with cathode ray tube display, a tape
recorder, a graphic (chart) recorder, and a
solid state inverter.
The radar beam has a spread of +45° in
the fore and aft directions, and +20° later-
ally. Any target detected within this beam
will be recorded as being directly below the
point of the surface where the signal is
transmitted and received, and signals are
reflected only from surfaces perpendicular
to the direction of the signal. The use of
a 10-foot grid spacing thus resulted in a
sampling approach, as opposed to full cover-
age of the subsurface volume. However, even
with a very fine grid, a fraction of the
buried drums would be missed by the radar
due to a) their orientation or b) their be-
ing "shielded" by metal drums closer to the
surface, since metal is a near-perfect re-
flector of radar energy.
Illustrative data from the survey are
shown in Figures 3, 4, and 5. The ground
surface is at the top of each figure where-
as the bottom of the figure corresponds to
a depth of approximately 12 feet. The ver-
tical dashed lines are markers produced
electronically in the field, which corre-
spond to 10-foot intervals. Figure 3 illus-
trates a subsurface profile where there are
no buried drums. This profile was taken
over an area of undisturbed soil.
In Figure 4, there are a number of in-
dividual targets identifiable by the char-
acteristic hyperbolic "signature".. This
signature results from the increased travel
time between the target and the antenna when
the beam approaches or moves away from the
target versus when it is directly over the
target. Each reflecting target will produce
three characteristic hyperbolas. It is pos-
sible for a skilled interpreter to distin-
guish between the signature caused by a drum
or boulder either in the field or from the
recorded data. In the field, a metal object
can be determined instrumentally by compar-
ing the polarity of the target signal to the
background signal. A metal object which is
essentially a perfect reflector, will pro-
355
-------
duce a signal that is "in phase" with back-
ground. An object such as a boulder will
produce a signal reversed from background.
Close examination of the recorded data will
also show another reflecting signal produced
by a'metal object in addition to the three
characteristic hyperbolas. The fourth image
found above the characteristic three is also
caused by the reflection from metal and
would not be present on the data if the tar-
get were a boulder.
Figure 5 shows a blurred effect which
is interpreted as being caused by a concen-
tration of contaminants, as from a leaking
drum. Chemical analysis by the DEM has
shown that some of the chemicals being re-
leased from the drums are ionic, which is a
characteristic that would increase the at-
tenuation of the radar signal strength.
Therefore a noticeable blurred contrast
relative to the average strength is observed
on data taken over a trench containing a
high concentration of contamination.
Metal Detection
The entire 7.5 acre site in Coventry
was surveyed by personnel from Fred C. Hart
and Associates, Inc. with a Fisher M-Scope
(Model TW-5) metal detector. This equipment
is designed for locating buried metal ob-
jects by inducing an electromagnetic field
around the object in response to radiation
from a transmitter. The average depth of
detection for metal objects is dependent on
the amount of background "noise". Thus in
areas free of buried metal, the probable
depth of detection was approximately six to
eight feet. In areas of buried drums, the
sensitivity setting of the instrument had
to be cut back, resulting in a potential
Figure 3. Radar profile taken outside
trench boundary.
Figure 4. Radar profile taken within trench
boundary showing "signature" of buried
drums.
356
-------
Figure 5. Radar profile taken within trench
boundary showing buried drums and chemical
contamination (blurriness near center of
photo).
depth of metal detection of approximately
four to five feet. In areas where buried
drums were suspected, based on disturbed
ground or the initial gross scan of the
overall site, the survey was conducted by
traversing closely-spaced grid lines.
Electrical Resistivity
The electrical resistivity of a geolog-
ical formation depends upon the conduction
of electric current through the particular
subsurface materials. Since most of the
geologic formations that contain water have
high resistivities, the electrical resistiv-
ity of a saturated rock or soil is primarily
a function of the density and porosity of
the material and the concentration of the
conducting ions within the saturating fluid.
In a resistivity survey, an electric cur-
rent is passed into the ground through a
pair of current electrodes and the potential
drop is measured across an inner pair of po-
tential electrodes. The "apparent resistiv-
ity" is determined by the equation, Ra =
27rA(V/I), where A is the electrode spacing,
V is the potential difference, and I is the
applied current. The depth of penetration
is controlled by the distance between the
electrodes (called the A-spacing) and is
approximately equal to half of this dis-
tance. Varying the A-spacing allows resis-
tivity measurements to be taken in the form
of either lateral or depth profiling.
Both types of profiling methods were
conducted at the hazardous waste site in
Coventry using a Bison Instruments Model
2350B Earth Resistivity meter powered by a
90-volt battery. A fixed electrode spacing
of 20 feet was used for the lateral profiles
in the areas of the trenches and the swamp
where the depth of ground water contamina-
tion was suspected as being shallow. Two
lateral profiles using a fixed A-spacing of
50 feet were also conducted approximately
2000 feet west and north of the immediate
site walls, where it was suspected that the
contamination might be detected at greater
depths.
Additionally, seven depth profiles were
conducted in the vicinity of the West Trench
located previously by the metal detection
survey. Electrodes were set at intervals
of 1, 2, 4, 8, 16, and 32 feet at each pro-
file location, permitting a maximum depth of
investigation of approximately 15 feet. Re-
sistivity readings were taken at each spac-
ing interval for the left, right, and cen-
tral spacings. This particular spacing in-
terval was chosen based upon the intents of
the investigation, which were to identify
changes in subsurface contamination and to
locate the boundary defining the bottom of
the trench. The lateral surveys were con-
ducted by Fred C. Hart and Associates, Inc.
and the vertical surveys by Stephen A. Alsup
and Associates, Inc.
Seismic Refraction
The seismic refraction method is based
on the principal that elastic waves (mech-
anical rather than electromagnetic) travel
through different subsurface strata at dif-
ferent velocities. Elastic waves are intro-
duced to the ground surface by an energy
source, usually a small explosion or a ham-
mer blow on a steel plate for shallow in-
vestigations. The refracted waves are
357
-------
detected by small seismometers (geophones)
located on the surface at various distances
from the energy source. A seismograph re-
cords the travel time between the vibration
and the arrival of the elastic wave at the
geophones. Plotting arrival time versus
distance from the energy source to geophone
from a series of seismograph records enables
the determination of strata depths and their
seismic velocities through the use of simple
refraction theory.
Seismic refraction profiling of approx-
imately 2,850 linear feet was performed at
the Coventry site in two days of field work
by Stephen A. Alsup and Associates, Inc. A
Geometries/Nimbus Model ES1210F Multichannel
Seismograph was used to record and collect
the voltage outputs from 12 geophones spaced
at 20-foot intervals for each refraction
spread. The energy source used to initiate
each record and shock wave was a 30-pound
weight drop or 10-pound sledge hammer blow
on a steel plate with an attached impact
start switch.
The seismograph, which includes a digi-
tal memory of waveform from each data chan-
nel, allows repetitions of the elastic waves
from a series of hammer blows thereby en-
hancing the ability to detect signals and
pick arrival times. The use of the hammer
drop as the energy source (which is prefer-
able to using small explosive charges in
this type of investigation) would be more
difficult without the digital memory of the
seismograph. Impact points for this survey
were at the end of, and quarterly along, the
refraction spread, providing a locus of
depth calculations at 80-foot intervals
along each spread. Data continuity and rep-
etition were achieved by repeating end shots
where refraction lines were longer than one
spread length.
RESULTS OF FIELD STUDIES
Plume Delineation
Having determined the information need-
ed to evaluate the long-term abatement al-
ternatives, the Phase I investigation was
planned. A principal component of the site
investigation was the installation of shal-
low monitoring wells in order to collect
soil and water samples and to determine
ground water elevations.
Because natural conditions at the site
were such that measurement of electrical
resistivity was expected to be successful, a
lateral profiling survey was performed to
facilitate the placement of monitoring wells.
In addition, a depth profiling survey was
conducted to determine vertical contamina-
tion patterns. Figure 6 shows the apparent
resistivity values obtained during the lat-
eral profiling survey, plotted on a contour
map. The measurements taken using both the
20- and 50-foot A-spacing are given. Rather
than showing that the plume was moving di-
rectly toward the swamp in a northwesterly
direction as suspected from measurement of
water levels in existing monitoring wells,
the apparent resistivity values gave an in-
dication of two distinct plumes (a western
and a northern plume) at different depths
with each having a separate source. With
increasing distance from the site, however,
the two plumes joined finally to discharge
in the swamp.
The western plume, which was defined
primarily using the 20-foot A-spacing, ap-
pears to be generally within 10 feet of the
surface. Most of the plume moving toward
the north was defined using the 50-foot A-
spacing and appears generally deeper than
20 feet below the surface. However, some
shallow contamination is also apparent along
the northern border of the site near the
trenches. Shallow bedrock off the northwest
corner of the site was considered the most
likely explanation for the high apparent re-
sistivity values between the two plumes, al-
though this explanation was later proven to
be false (see next subsection). Hence, the
results of the resistivity survey suggest
that additional monitoring wells be located
to determine the existence of the shallow
bedrock and to substantiate the presence of
two separate plumes.
Additionally, a discovery of a contami-
nant source along the partly grass-covered
western edge of the site was important in-
formation, since it served to indicate how
far south and west the monitoring well pro-
gram ideally should extend. Locating this
additional source of contamination may also
have been possible using the radar technique
based upon comparison of signal strength.
As previously mentioned, several depth
profiling resistivity surveys were conducted
in the vicinity of the West Trench to deter-
mine vertical changes in contamination.
Figure 7 is a plot of the apparent resistiv-
ity versus approximate depth below the sur-
face for several profiles taken over and
358
-------
outside of the trench boundary. Presenta-
tion of the data in this manner allows com-
parison of the "normal" or background resis-
tivity patterns observed outside the West
Trench where no contamination is expected,
to the patterns observed over and downgradi-
ent from the trench where the occurrence of
contamination is anticipated.
Profiles 1 and 6, taken outside the
major plume boundary, as indicated by the
lateral resistivity survey, are indicative
of background patterns. Profile 2 was taken
in the open, unfilled trench where no drums
were buried, but bulk chemicals were sus-
pected to have been discharged. This pro-
file shows the effect of free-standing water
in the trench and perhaps some slight con-
tamination within the first few feet of the
surface, and then an increase in resistivity
to near background conditions. This indi-
cates that either the open trench was not
used for bulk dumping of significant amounts
of chemicals with high electrical conductiv-
ity (since conductive effects decrease
rapidly below several feet of depth), or
that significant portions of the chemicals
have been removed by leaching. Profiles 3,
4, 5, and 7, taken over or downgradient from
the West Trench, show the varying degrees of
contamination within the plume. A very high
contentration of drums may explain the ex-
tremely low resistivity measured at the lo-
cation of profile A. The volume of highly
conductive chemicals at this location is in-
ferred as being very large in order to have
such a strong effect on the resistivity mea-
surements. A very localized concentration
of conductive chemicals would not be expect-
ed to have such a strong effect with this
particularly wide electrode spacing (32
feet).
Figure 8 shows the same seven profiles,
but on a plot of "cumulative" resistivity
versus approximate depth below the surface.
Slope changes or breaks in the cumulative
curves generally indicate the depth of the
underlying unit, and the direction of slope
change indicates the relative resistivity
Figure 6. Contour map of apparent resistivity values.
359
-------
APPARENT RESISTIVITY (OHM - FEET)
Figure 7. Apparent resistivity depth profiles for several locations near the West Trench.
CUMULATIVE RESISTIVITY (OHM - FEET)
Figure 8. Cumulative resistivity depth profiles for several locations near the West Trench.
values of the subsurface materials. Thus,
an increasing slope signifies that the un-
derlying unit has a higher relative resis-
tivity, whereby a decreasing slope denotes a
lower relative resistivity in the underlying
unit. In accordance with the interpretation
of Figure 7, Figure 8 shows higher relative
resistivities with depth (no contamination)
for profiles 1 and 6. Interpretation of
profile 2 is more straightforward using the
cumulative resistivity plot. Slight de-
creases in slope, denoting lower resistivity
or presence of soil moisture or contamina-
tion, are evident between two and five feet
and below ten feet. Profiles 3, 4, 5, and
7 show a continued decrease in slope with
depth, which is expected because of their
locations within the buried drum area or
downgradient from the trench.
Determination of the depth of the bot-
tom of the West Trench was not possible
using the vertical resistivity plots. It
is not clear, however, whether the particu-
360-
-------
lar A-spacing used for these profiles (32
feet) was wide enough to allow adequate pen-
etration necessary to detect the bottom of
the trench. A second possibility is that
there is no detectable change in contamina-
tion at the bottom of the trench thereby
causing a continuation of the low resistiv-
ity readings beyond the region of buried
drums.
Fifteen monitoring wells were installed
following the lateral resistivity survey.
Refusal depths, tentatively assumed to re-
flect the approximate top of bedrock, did
indicate a mound in the bedrock surface off
the northwest corner of the site. Four
wells in this vicinity were dry, which also
gave credence to the results obtained from
the resistivity survey, namely the existence
and location of two plumes. In addition,
soil samples taken from these same locations
were much less contaminated than soil sam-
ples taken from borings located within the
plume boundaries. Consideration of these
factors seemed to indicate that ground water
flow was being diverted around a bedrock
mound and this had resulted in the detection
of high apparent resistivity values in this
area. It was found by subsequent (Phase II)
bedrock drilling, seismic refraction survey,
and chemical analysis of soil and ground
water that the bedrock mound did not exist
and that contaminated ground water was in-
deed traveling in this location. In gener-
al, the ground water is at a greater depth
below the surface in this region than the
other surveyed areas, thus resulting in
higher relative resistivity values, and sub-
sequent incorrect interpretation.
Determination of Bedrock Topography
Complete verification of the shallow
bedrock off the northwest corner of the site
was not possible until the bedrock coring
and the seismic survey were performed. The
drilling showed that the refusal depths of
the previous borings had actually been due
to boulders and/or very dense till. At each
boring location, the bedrock (a granite
gneiss) was discovered to be 10 to 30 feet
deeper than anticipated. The seismic survey
indicated that the bedrock surface was gen-
tly rolling, varying from approximately 10
feet below ground surface near the swamp to
approximately 70 feet below ground surface
on top of the site.
The boring drilled in the area between
the two plumes showed that the bedrock was
highly weathered and fractured. A piezome-
ter installed in the fractured bedrock indi-
cates that the granite gneiss is hydrauli-
cally connected to the unconsolidated
glacial deposits. Therefore, ground water
is not being diverted around a shallow bed-
rock mound as had been inferred from the
resistivity survey and Phase I drilling, but
is actually moving over this area toward the
swamp at depths greater than 20 feet. A
ground water sample taken from this well was
found to contain a diverse assortment of
volatile organic pollutants similar in con-
centration to samples taken from wells with-
in the two plumes.
A seismic refraction profile was per-
formed over the West Trench in an experimen-
tal attempt to determine the depth of the
base of the buried drums. Neither ground-
penetrating radar nor metal detection was
able to show the lower boundary of drums and
resistivity depth profiles revealed no read-
ily-interpretable trends. Knowing the depth
of trenches is critical to the estimation of
the number of drums in each trench. A re-
mote sensing method that can effectively
determine depth of drums would greatly aid
other similar investigations for the deter-
mination of drum number and cost estimates
of abatement techniques.
The results of the seismic profile of
the West Trench are shown by Figure 9.
Three distinct velocity units are shown on
the profile and it is believed that the
upper one, varying from 7 to 14 feet, repre-
sents the disturbed soil surrounding the
buried drums. However, this interpretation
needs to be confirmed through test drilling
and/or excavation.
Buried Drum Location and Number
The location and dimensions of the
trenches used to estimate the number of
drums in each trench were based on a combin-
ation of data from metal detection, ground-
penetrating radar, and the exploratory ex-
cavation. The results from the seismic
profiling of the West Trench were used to
estimate the lower limit for the bottom of
the trenches, even though these data have
not been confirmed. For the purposes of
estimating the number of drums contained in
the trenches, the angle of the vertical side
walls was assumed to be 60°, the angle of
the declining surface of drums 45°, and the
angle of descent of the trench ends 45°.
The angle of repose for disturbed site soil
361
-------
100
I
200
I
490
LEGEND.
5200
SHOT POINTS
INFERRED SURSURFACE
VELOCITY BOUNDARIES
SEISMIC WAVE VELOCITIES
IN FT/SEC
Source: S.A. Alsup and Assoc., Inc.
Figure 9. Subsurface profile of the West
Trench as determined by seismic refraction.
is approximately 45°, but excavated side
walls were shown to maintain a steeper
slope.
Since the radar probed to a depth of 12
feet in contrast with the four to six feet
in the vicinity of the trenches for metal
detection, the radar would be expected to
present a somewhat more accurate indication
of trench boundaries. The radar found two
trenches in the "Northeast Trench", versus
the single trench identified previously with
metal detection; the explanation for this is
not known. On the other hand, the radar
data for the West Trench had to be supple-
mented by data from the metal detection.
The radar provided, in addition, some
useful qualitative information on the way
drums were placed and on the trench con-
struction. For example, although there were
isolated instances where drums appeared to
be neatly stacked, this was the exception
rather than the rule; the drums for the most
part appeared to be randomly stacked based
on the radar data, and at least in the top
eight or so feet below the surface (where
individual drums most clearly could be iden-
tified) the drums appeared to be present in
clusters as opposed to being uniformly dense
throughout a trench. Also, the top surface
of the drums displayed an "angle of repose"
from the sides to the center of the trench
cross-section.
The radar was not able to detect the
bottom of the trenches, partly because the
upper drums masked what was beneath. Even
in the West Trench, where a 25-foot nominal
depth was probed, the trench bottom could
not be located from the data. The radar
data can often be used, however, to deter-
mine the interface between the sides of the
trenches and the undisturbed soil. Radar
signals from within the trench are general-
ly stronger than signals from outside the
trench. This contrast is attributed to the
fact that the disturbed soil within the
trench has a higher dielectric constant be-
cause it is more porous and has a greater
moisture content than undisturbed soil. For
future work at other sites, it is suggested
that deep radar probing at and just outside
a trench boundary may be successful in de-
termining the maximum depth of drums, de-
pending on the steepness of the side of the
trench relative to the radar beam, the clar-
ity of the radar signal at this depth, and
the subsurface material at the given site.
In order to produce estimates for the
number of drums remaining buried, a theoret-
ical trench geometry described earlier was
employed. It is also assumed for the pur-
pose of the drum estimates that a two-foot
layer of soil covered the top of the burial
area and two nominal trench depths of 14 and
22 feet were used in order to bracket the
range determined from remote sensing and di-
rect excavation. The bottom of the trenches
are assumed to be level with no irregular-
ities. Straight sides for the horizontal
widths and lengths have also been assumed.
Two densities of drums (percent of vol-
ume of drums within trench volume below the
cover layer of soil) were used for the drum
number estimates: 90 percent and 50 per-
cent. A drum density of 90 percent repre-
sents the closest packing arrangement pos-
sible for cylinders without regard to inter-
ferences imposed by the actual geometry of
the trench boundaries. An actual drum den-
sity of 54 percent was calculated for the
Northeast Trenches using the results obtain-
ed from the DEM site representative combined
362
-------
with the theoretical trench geometry. The
calculated 54 percent density was rounded
off to 50 percent for the lower limit calcu-
lations of the drum number estimates.
The estimated range of the number of
drums remaining buried at the Coventry site
is found in Table 2. The drum estimate was
performed by calculating the volume of each
trench and multiplying the volume by the as-
sumed drum density to yield the total volume
of drums. The estimate for the number of
whole drums is provided by dividing the to-
tal volume by the volume of a single drum
(7.35 ft3). As Table 2 shows, the overall
range varies by a factor of two and a half,
from 16,700 to 44,700, while the more likely
range based upon the observed depth of the
Northeast Trenches is less than a factor of
of two, from 25,000 to 44,700.
The above estimates are for whole, un-
crushed 55-gallon drums. The numbers will
necessarily increase if some are crushed,
enabling closer packing. The drum number
estimates can be corrected for the presence
of crushed drums by multiplying by g/(f + g
- gf), in which f represents the fraction of
crushed drums and g is equal to the ratio of
the volume of a whole drum to the volume of
a crushed drum. If g = 2 and f = 0.3, for
example, as indicated by the exploratory ex-
cavation of the Northeast Trenches, there
would be 18 percent more drums (whole plus
crushed); however, there would be 17 percent
fewer whole drums.
Prior to the radar survey, an estimated
range of drums was made which was substan-
tially lower than the estimates presented
here. The earlier analysis plausibly as-
sumed that the trenches with buried drums
were of similar construction to that of an
unfilled trench on the site. As a lesson
for other similar sites, it is wise to keep
in mind that without the benefit of more ac-
curate information, the "worst case" corre-
sponds to a steep sided trench (with angle
of repose depending on local soils as well
as the method of trench construction) with
depth approximately equal to the water
table, to bedrock, or to the maximum fea-
sible excavation depth.
CONCLUSIONS AND RECOMMENDATIONS
Uncontrolled or abandoned hazardous
waste sites present varying degrees of dif-
ficulty to investigators. For example,
abandoned sites which are extensive in
area, rural (with hindering vegetation), or
in areas of complex geology and hydrology,
represent troublesome environments for in-
vestigation. Therefore it is important to
develop approaches for thorough, but rapid
and cost-effective assessments of these
difficult situations. In most cases, a
well designed and executed investigative
program will include remote sensing tech-
niques in addition to direct measurement.
Premature action to drill wells, collect
and analyze various air, water, and soil
samples, or perform excavation without
careful planning and proper integration of
available techniques may result in unneces-
sary adverse exposure to hazardous condi-
tions and in an inaccurate or incomplete
understanding of the total problem.
TABLE 2. ESTIMATED NUMBER OF BURIED DRUMS
BASED ON EXTRAPOLATION OF BEST AVAILABLE DATA
Trench
Location
Northwest
West
South
Total
Maximum
Drum Density
d = 14 ft
14,800
13,500
1,700
30,000
d -
22
20
2
44
22 ft
,400
,200
,100
,700
Drums Randomly
Stacked
d = 14
8,200
7,500
1,000
16,700
ft d -
12,
11,
1 ,
25,
22 ft
400
200
200
000
Remote sensing techniques may be used
to provide reasonably accurate assessments
of subsurface contamination, the location
and extent of buried drums, and other data
needs for determining appropriate -methods of
abatement. It must be stressed, however,
that not all critical information can be ob-
tained remotely, since each of the tech-
niques has limitations, both theoretical and
site-specific, and consequently that direct
sampling should be undertaken at every un-
controlled hazardous waste site.
Notes: d = nominal trench depth
Random stacking indicated by results of excavation of
Northeast Trenches,, approximated by 50 percent drums,
earth by volume in trench below 2 ft cover and assumed
trench geometry, as described in the text.
Drums are assumed to be uncrushed, 55-gallon drums.
Table 3 summarizes the purpose, advan-
tages, and limitations of each of the four
remote sensing methods used at the Coventry
site. It is important that this type of in-
formation be consulted prior to development
of an investigatory program. Even though
363
-------
TABLE 3. COMPARISON OF REMOTE SENSING TECHNIQUES
Technique
Purpose
Advantages
Limitations
EIgcj^rleal Resistivity
Lateral Profiling
Depth Profiling
Seismic Refraction (Non-
explosive Method)
Metal Detection
Ground-Penetrating Radar
• determine lateral extent of
contaminated ground water
facilitate placement of mon-
itoring wells and optimize
their number
monitor changes in plume
position and direction
indicate change in contamina-
tion with depth
establish vertical control in
areas of complex stratigraphy
determine depth and topogra-
phy of bedrock
determine depth of trench
containing buried drums
locate areas of high metal
content (e.g., buried drums)
• locate buried objects (e.g.,
buried drums)
• provide qualitative infor-
mation regarding drum density
* detect interfaces between
disturbed and undisturbed
soil (e.g., bottom of tren-
ches)
• detect plumes of high chemi-
cal concentration
• procedure less expensive than
drilling
• procedure more rapid than
drilling
• equipment light-weight, able
to be hand carried
• survey may be conducted in
vegetated areas
same as above
• procedure less expensive than
coring or excavation
• procedure more rapid than
coring or excavation
• survey may be conducted in
vegetated areas
• procedure less expensive than
excavation or radar
• procedure more rapid than ex-
cavation or radar
• equipment light-weight, able
to be hand-carried
• survey may be conducted in
vegetated areas
• procedure less expensive than
excavation
• procedure more rapid than ex-
cavation
• procedure deeper-penetrating
than metal detection
• procedure yields more infor-
mation than metal detection
• procedure may be used over
paved areas
• limited ability to detect
non-conductive pollutants
• technique unsuitable if no
sharp contrast between con-
taminated and natural ground
water
• interpretation difficult if
water table is deep
• interpretation difficult if
lateral variations in stra-
tigraphy exist
• interpretation difficult if
radical changes in topogra-
phv are not accounted for in
choice of A-spacing
• technique unsuitable in paved
areas or areas of buried con-
ductive objects
same as above
technique unsuitable if no
sharp velocity contrast be-
tween units of interest
(e.g., trench containing
buried drums and surrounding
soil)
oad
for vehicle
depth of penetration varies
with strength of energy
source
low velocity unit obscured
by overlying high velocity
units
• interpretation di f ficult in
regions of complex stratigra-
phy
i technique unsuitable for the
detection of non-metallic
objects
i technique unsuitable for ob-
jects below five feet
i technique unsuitable for de-
termination of number or ar-
ranpement of buried objects
technique unsuitable for
vegetated areas
data requires sophisticated
interpretation
underlying objects obscured
by those above
survey requires access road
for vehicle
there are disadvantages inherent to each
technique, proper sequencing and phased
studies can potentially result in an overall
optimized approach. It must be emphasized
that as the study progresses, preliminary
conclusions will necessarily be modified and
the nature of direct sampling activities
will need to be evaluated continuously. It
is recommended that final conclusions not be
drawn solely from the results of remote
sensing methods.
To accomplish site investigations in
the most efficient manner, a systematic ap-
proach is necessary to take advantage of the
information that can be extracted from re-
mote sensing methods. In addition, a sys-
tematic approach allows a reduction in the
364
-------
time and cost, and an increase in the effec-
tiveness of direct sampling.
In general, the following two objec-
tives must be addressed by all investiga-
tions at uncontrolled hazardous waste sites:
• determination of the nature and
extent of the problem and the
resulting effects on public
health and the environment (both
actual and potential)
• determination of environmentally
sound and cost-effective methods
to effectively abate the problem
(if abatement is deemed neces-
sary) .
The first activity of an investigation
should be identification of specific data
needed to meet each objective. After this
has been accomplished, the various tech-
niques available for data acquisition, both
remote and direct, can be evaluated with re-
gard to the type of information that can be
obtained from each in relation to the spec-
ific conditions at the site. Although not
always the case, it may be reasonably as-
sumed that remote sensing techniques should
be used in advance of the more direct data
acquisition methods of borings or excava-
tions. This is not intended to imply, how-
ever, that all direct sampling should be
held in abeyance. There have been numerous
instances in which emergency action is de-
pendent upon immediate results from air,
water, and soil sampling, and for such
cases remote sensing techniques should be
used secondarily.
Both the selection and sequence of re-
mote sensing and direct data collection
techniques should be determined based upon
the specific needs and circumstances of the
given site. Additionally, the limitations
of the remote sensing techniques presented
in Table 3 should be kept in mind. Even the
best combination of results obtained remote-
ly provides only an approximate representa-
tion of subsurface condition. Finally,
since the cost of such site surveys tends to
be only a very small fraction of the total
cost of ultimate solutions, it is generally
cost-effective to apply several overlapping
techniques at a site to complement one an-
other and refine the results to support im-
plementation of the preferred long-term
solution.
ACKNOWLEDGMENTS
The work on which this paper is based
was supported under MITRE contracts with
U.S. Environmental Protection Agency (EPA),
Solid and Hazardous Waste Research Division,
and the State of Rhode Island, Department of
Environmental Management. The encouragement
and interest of D. Sanning and S. James of
the EPA are especially noted. Subcontracted
field services not mentioned in the text
were provided as follows: test borings and
well installation by R. F. Geisser and Asso-
ciates, Inc. and Guild Drilling Co., Inc.;
laboratory services by Energy Resources
Company (ERCo); topographic surveying by
Caputo and Wick, Ltd.; and geotechnical
support by Geotechnical Engineers, Inc.
The contributions to the site investi-
gation of Paul J. Stoller of MITRE are also
acknowledged. The results and conclusions
presented are those of the authors and do
not necessarily reflect the views of the
sponsoring agencies or other subcontractors.
Portions of this paper were presented
at the EPA National Conference on the Man-
agement of Uncontrolled Hazardous Waste
Sites, October 15-17, 1980, Washington, D.C.
and appear in the proceedings under the
title: "Remote Sensing for Investigating
Buried Drums and Subsurface Contamination at
Coventry, Rhode Island" by H. J. Yaffe,
N. L. Cichowicz, and P. J. Stoller.
REFERENCES
1. Cichowicz, N.L., R.W. Pease, P.J.
Stoller, and H.J. Yaffe. 1980. Eval-
uation of Abatement Alternatives:
Picillo Property, Coventry, Rhode
Island. MITRE Technical Report-80W
00253, The MITRE Corporation, Bed-
ford, Massachusetts. 83pp.
2. Cichowicz, N.L., R.W. Pease, P.J.
Stoller, and H.J. Yaffe. 1980. Use
of Remote Sensing Techniques in a
Systematic Investigation of an Un-
controlled Hazardous Waste Site.
MITRE Technical Report-80W00244, The
MITRE Corporation, Bedford, Massa-
chusetts. 69pp.
3. Pease, R.W. et al. 1980. Hazardous
Waste Investigation: Picillo Prop-
erty, Coventry, Rhode Island. MITRE
Technical Report-80W00032, The MITRE
Corporation, Bedford, Massachusetts.
142pp.
365
-------
CONCEPTUAL COST ANALYSIS OF REMEDIAL ACTIONS
AT UNCONTROLLED SITES
HOWARD L. RISHEL
SHEILA M. KENNEDY
SCS Engineers
Long Beach, California 90802
OSCAR W. ALBRECHT
U.S. Environmental Protection Agency
Municipal Environmental Research Laboratory
Cincinnati, Ohio 45268
ABSTRACT
A document has been prepared, in response to the 1980 "Superfund" legisla-
tion, which gives enforcement officials a tool with which they can quickly
compare the relative costs of several remedial action scenarios at inactive
landfill and surface impoundment disposal sites. This report outlines the
process of costing remedial action unit operations (at a range of site
sizes), and combining the unit operations to yield relative cost data for
an entire remedial action scenario. Examples are given of costing one unit
operation at five scales of portrayal, and of costing one remedial action
scenario which is comprised of several unit operations. The assumptions
and limitations of the methodology are given.
Introduction
During 1980, the U.S. Congress
enacted "Superfund" legislation,
which was proposed to provide funds
for the U.S. Environmental Protec-
tion Agency (EPA) to assist in the
mitigation of pollution problems at
uncontrolled waste disposal sites
through remedial actions. The re-
sponsible offices within EPA (the
Oil and Special Materials Division,
and the Office of Enforcement)
requested the Office of Research
and Development to provide tech-
nical information to support this
process. As a part of this effort,
SCS Engineers undertook to review,
compile, update, and integrate
existing data on the costs of such
remedial actions, in terms of dis-
crete unit operations which could
then be combined to construct con-
ceptual remedial action scenarios.
This type of review-and-update
approach was considered more appro-
priate than additional conceptual
designs efforts because much con-
ceptual design work has already
been done. The design work which
exists, however, is scattered, in-
complete, and inconsistent in meth-
odology; much of it is out of date,
and vague either about the methods
used to arrive at a cost figure or
about what components the cost fig-
ure i ncluded.
Through the review-and-update
approach, we hope to impose a con-
sistent methodology on the existing
data in terms of scope, location,
time frame, and cost compu-
tations. In addition, we hope to
flesh out the missing details,
where necessary, and present the
366
-------
TABLE 1. SCALES OF OPERATION FOR LANDFILL DISPOSAL SITES
Scale of Operation
1. 9.07 tonnes/day
2. 45.36 tonnes/day
3. 90.72 tonnes/day
4. 272.16 tonnes/day
5. 453.60 tonnes/day
Refuse-to-Soil Ratio
1:1
1.5:1
2:1
3:1
4:1
results in a
uniform format, with a minimum of
overlap between the individual unit
operations. The resulting document
presents this existing data in a
framework of a broad and consistent
methodology, with enhanced detail.
It should be emphasized here
that no new conceptual design work
was done for this project. Where
data was incomplete, some detail
was filled out, but the thrust of
this work was to enhance previously
existing conceptual design data and
make them more available and useful
to enforcement personnel respon-
sible for overseeina the retrofit
operations.
Because the document was in-
tended for use in the retrofitting
of uncontrolled sites, the unit
operations examined are limited to
those appropriate for the clean-up
of closed or abandoned sites.
Methodoloqy
In developing and characteriz-
ino unit operations for remedial
action at waste disposal sites, we
developed hypothetical site pro-
files and their associated unit
operation profiles. Costs for each
unit operation at five scales of
site operation were then computed
at three cost levels: high and low
U.S. averages, and the price esti-
mate for a single city, Newark, NJ
(mid-1980 dollars).
Site Profiles
Site profiles, or hypothetical
disposal sites, were developed for
landfill disposal and for surface
impoundments (disposal ponds). Each
of these was portrayed at five
scales of daily operation. The
resulting site profiles were con-
figured to conform to uniform sets
of design criteria and environmen-
tal conditions. For both landfills
and surface impoundments, the se-
lected scales of operation were
developed in terms of daily input.
This emphasis on daily input is
consistent with the usual view of
landfill practices and the assump-
tions that surface impoundments are
intended as temporary storage, with
relatively short retention times.
Landfil1s--Table 1 shows the
scales of operation for landfill
disposal sites. The range of scale
sizes was develooed from data pre-
sented in References (2) and (7).
Although the scale of operation for
each landfill is different, the
followina design criteria are held
constant:
• The surfae-e area for each
landfill is square.
• All landfills are cut and
cover operations, with cut
slopes at a ?:1 ratio and
fill slopes at a 3:1 ratio.
367
-------
TABLF 2. SCALES OF OPERATION FOR SURFACE IMPOUNDMENTS
Scale of Operation
( Influent)
1. 10 m3/day
2. 50 m3/dav
3. 500 m3/day
4. 5,000 n3/da.y
5. 50,000 m3/day
He tent ion Time
3 days
3 days
10 days
10 days
10 days
• Operation at each landfill
was 260 days/year for ten
years before the site was
closed.
• The compaction rate was
0.596 tonnes/ comoacted m .
Figures 1 and 2 show the lay-
out of the hypothetical landfill
site, without reference to the
scale of operation.
To allow comparison of reme-
dial actions between sites operat-
ino at different scales, the fol-
lowing environmental conditions are
also held constant:
• Ground surface and ground
water oradient are at a 1%
slope.
• Ground water is 4.0 m (13.1
ft) below the ground sur-
face .
• Low permeability strata
(<10~6 cm/ sec) is 15 m (50
ft) from the ground sur-
face.
• Unconsol i da ted earth, mate-
rials have a permeability
of 10~ cm/ sec or greater.
Surface Impoundments--Table 2
showsthescalesof operation for
surface impoundments. The ranqe of
scale sizes was developed from data
in References (3) and (8). In this
case, the scale of operation is
given in terms
influent.
of cubic meters of
Fiaure 3 shows the plan of the
hypothetical surface impoundment
without reference to the scale of
operation. The following design
criteria were common to all of the
surface impoundment site profiles:
The ponds
u n 1 i n e d.
are square and
• Berms were constructed from
soils excavated during pond
construction, and have 3:1
side slopes.
• The site operated 365 days/
year for ten years before
closure or abandonment.
• Sediment was removed from
oond bottom every two
years.
• Wastewater contained 100
mg/1 settleable solids.
• Density of solids was 2
g/ml .
• Sludge is 70% moisture by
weight when removed.
• Wastewater was recirculated
after allowing 3 to 10 days
for soli ds sett!i ng.
e Because of short detention
time and sludge on bottom,
precipitation, evaporation,
368
-------
GROUND SURFACE(
SLOPE
GROUND WATER
-*• FLOW OF GROUND WATER
ism
77&
,'yv- v'v/v^xvT^'0
Low Permeability Strata
Where:
Total volume of refuse Inr)
Total volume of soil (m )
Height of landfill above ground surface (m)
Depth of landfill below ground surface (m)
Top side of landfill (m)
Bottom side of landfill (m)
I3'.
Side
Area
of
of
1
1
andfl
andfl
11
11
at
at
ground
ground
surface
surface
(ml
(m2)
Figure 1. Typical side view of landfill
-------
FLOW OF
GROUND WATER
Figure 2. Typical top view of landfill.
-------
BERM
H
TOTAL
LENGTH
GROUND SURFACE
-— t
POND WATER SURF C__
FREEBOARD
GROUND WATER TABLE
4m
ism
Low Permeability Strata (Aquiclude or Aquitard)
Where:
• (Total length) = impoundment surface area.
• Berm side slopes constructed at 3 horizontal to 1
vertical.
• Berm top width nominal 2m, maximum 3 meters.
• 0.5m freeboard designed into all ponds.
Figure 3 . Side view of surface impoundment
-------
and percolation losses are
considered negligible when
compared to the volumes in-
volved.
• Averaoe percolation _5_8.14
1/day/n2 (0.2 qal/daj/ft2).
To help compare remedial
actions for each of the scales of
operation, the followina environ-
mental conditions surrounding each
pond were developed:
t Ground surface and ground
water gradient are at a \%
si ope.
• Ground water is 4.0 m (13.1
ft) below the ground sur-
face .
• Low permeability strata
(aauielude or aquitard;
K<10"6 cm/sec) located at
15 m (50 ft) from ground
surface.
• Unconsolidated earth mate-
rials haye a permeability
of _^10~s cm/sec.
Unit Operation Profiles
Thirty-five unit operations
were identified which might be used
as part of a remedial action effort
at an uncontrolled hazardous waste
disposal site. These operations
are shown in Table 3 for both im-
poundments and landfills. Examina-
tion of this table shows some
apparent overlap between the two
lists, but in fact each operation
was configured separately for the
distinctly different design cri-
teria and environmental conditions
assumed in each site profile.
All of the unit operations
shown were drawn from the litera-
ture and only those for which the
literature contained adequate con-
ceptual designs and cost data were
addressed. in some cases a part of
the necessary data was missinq and
was suppli ed by us.
Alternative unitoperations--
When "a pol 1 uti on problem exists, a
number of unit operations may be
used interchangeably. The follow-
ina list qives some unit operations
which may be used either conjunc-
tively or as alternatives for each
other.
1. For elimination of contam-
inated site runoff, and
prevention of precipita-
tion from entering a land-
fill or closed impound-
ment, the following are
conjunctive or alternative
unit operations:
- Contour gradina and sur-
face water diversion
- Surface seal 1nq
- Revegetation
- Berm construction/
reconstruction.
2. For prevention of leachate
formation, the following
are conjunctive/alterna-
tive unit operations:
- Contour gradina and sur-
face water diversion
- Surface seali nq
- Grout curtain
- Sheet pilinq
- Slurry trench
- Wei 1 extraction
- Well point system
- Chemical fixation.
3. For the control of
leachate/contaminated
ground water migration,
the followina are conjunc-
tive/alternative unit
operati ons:
- Grout curtain
- Grout bottom seal
- Sheet piling
- Slurry trench
- Well point system
- Well extraction
- Well injection
- Underdrains.
4. For gas (methane and other
volatile hydrocarbons),
the following are con-
junctive/a! terna-tive unit
operations:
- Perimeter gravel trench
- Gas migration control -
passive
372
-------
TABLE 3. UNIT OPERATIONS USED AS REMEDIAL ACTIONS
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
Landfills
Contour grading and
water diversion
Surface sealing
surface 22.
23.
Revegetatlon 24.
Bentonite slurry trench 25.
cutoff wall 26.
Grout curtain 27.
Sheet piling cutoff wall 28.
Bottom seal ing 29.
Drains 30.
Well point system 31.
Deep well system 32.
Injection 33.
Leachate handling by 34.
subgrade irrigation 35.
Chemical fixation
Chemical injection
Excavation and reburial
Ponding
Dike construction
Perimeter gravel trench
Treatment of contaminated
waste
Gas migration control -
passive
Gas migration control -
active
Impoundments
Pond closure and contour
grading of surface
Surface sealing of closed
impoundment
Revegetation
Slurry trench cutoff wall
Grout curtain
Sheet piling cutoff wall
Grout bottom seal
Toe and underdrains
Well point system
Well extraction system
Well injection system
Leachate treatment
Berm reconstruction
Excavation and disposal
at secure landfill
373
-------
- Gas mi oration control -
active.
Cost, Compilation
Once a remedial action unit
operation was defined in terms of
its intended use (with respect to
the landfill or surface impoundment
site profiles) and the extent to
which it could be used in conjunc-
tion with or as an alternative to
other unit operations, the cost of
the operation was defined for each
scale of operation associated with
the site profiles. To do this, the
unit operation was broken down into
component reauirements. Each com-
ponent was further defined in terms
of sub-components (labor, mate-
rials, and equipment), and costs
were assigned to each component and
subcomponent. The assigned costs
are in terms of mid-1980 dollars
for U.S. upper and lower cost aver-
ages (for the continental 48
states), as well as for the example
location of Newark, NJ.
After costs were assigned to
each component, conceptual desiqn
capital and operatina cost esti-
mates were accumulated and allow-
ances for overhead and contingen-
cies were applied. Total and aver-
age life cycle costs were then com-
puted for each scale of unit opera-
tion. These cost averages were
then plotted and graphically inter-
polated to estimate scale economies
for each unit operation.
Estimation of ComponentCosts-
For the most part, the 1980 Dodge
and Means Guides (4,6) were used to
obtain the costs needed. The costs
were then expressed in terms of
metric units.
Regional adjustment indexes
presented in the Dodge Guide (4)
were used to modify the metric ver-
sions of the cost estimates for
geoaraphical differences. These
indexes were applied to obtain
revised material and labor costs
for the U.S. low, U.S. high, and
Newark, NJ estimates. No index was
applied to equipment costs, since
it was assumed that equipment costs
are the same nationwide. Because
the Dodqe Guide (4) and Means (6)
present costs differently, assump-
tions were made so that the re-
gional adjustment indexes could be
used for both texts. For example,
in the Means, labor costs were not
identified as a separate entry, but
were included as part of installa-
tion. Thus, whenever the Means was
used to present costs, the Podae
Guide Regional Adjustment Index for
Labor was applied to installation
costs.
Freouently, both Dodqe and
Means did not itemize costs into
categories of labor, material, and
equipment, but simply presented a
"total" estimate. Depending upon
which reference was used, the fol-
lowing rules were applied: in the
Dodge Guide, if only a total cost
was presented, an averaoe labor/
material index was applied to the
unit cost; in the Means, the
"total" costs include an overhead
allowance of ?5%. This allowance
was removed before the labor/mate-
rial index was anplied. In all
cases, costs were adjusted so that
overhead allowances were not in-
cluded at the subcomponent level.
As the scale of operation
changed, the quantity of any one
component required for a unit oper-
ation also chanqed. The cost of
each component presented in the
unit operation conceptual design
cost tables typically includes the
sum of costs for any material,
labor, or eouipment subcomponents.
These total costs for each compon-
ent do not include overhead and
contingencies. Once all the com-
ponents within a unit operation
were costed, the costs were summed,
giving a subtotal capital cost for
the unit operation. This subtotal
capital cost was then used to ob-
tain an overhead allowance (always
25%), and a contingency allowance
(between 10 and 40%, dependinq upon
the unit operation). The subtotal
capital cost was added to the over-
head and contingency allowances to
obtain the estimate of total unit
operation capital cost. (See
example below.) This method was
374
-------
TABLE 4. DERIVATION OF INFLATION PATES FOR O&M INPUTS
Type of O&M Input
Re la ted Published
Cost Index:
March 1980 Index
Value:
Average Annual Index
Percent Increase Since
Base Year (i.e., 12.75
years since mid-1967
when index eaualed 100)
Assumed Future
Inflation Rate:
Electricity
Electric Power
305.7
All Other O&M Inputs
Consumer Price (CPI-W)
239.9
9.160%
9.2%
7.104%
7.1%
used for all
t i o n.
scales of unit opera-
Life Cycle
c a p i tlTT and operating
determined for lower and
averaaes and for Newark,
averaae life cycle costs
puted to en-ure that any
cost comparisons of unit
could be equitably accomplished.
C o s t i n_q_- -Once total
costs were
upper U.S.
total and
were corn-
subsequent
operations
Although operation and mainte-
nance (O&M) cost estimates are for
1980, as the first year of opera-
tion, O&M component quantity re-
quirements were estimated to accu-
rately reflect requirements for
each of the first ten years of
remedial operations. This 10-year
life of the conceptual desiqns
means that life cycle evaluation of
operating costs only addressed sub-
sequent inflation and appropriate
discounting of these O&M components
costs to their mid-1980 present
values. It was further assumed that
capital costs would not be amort-
ized and discounted, but would be
considered as fully incurred in the
first year of operation. As a
result of these assumptions, aver-
age annual compoundina inflation
rates for electricity and for all
other O&M components were derived
using estimates from the April 1980
Survey of Current Business (1).
These inflation rates were derived
as shown in Table 4.
In determininq the present
values of future expenditures, the
March 1980 Gross National Product
Implicit Price Deflator of 174.51
(5) was similarly evaluated in
terms of its 1972
estimate an annual
tion rate of 7.4%.
assumed 4% social
rate was added to
annual di scoun t
1i fe cycle cost
base year to
general infla-
To this, an
time preference
create a total
rate of 11.4%. The
methodology was
then followed, in which inflated
operating costs were discounted to
their mid-1980 present values, and
summed with total capital costs, to
determine total life cycle costs
over the 10-year life span of each
unit operation. Average life cycle
costs were then computed by divid-
ing this total by the site pro-
file's daily scale of operation.
The contract report of this
work includes a set of tables for
each unit operation, qiving the
cost of each component, plus over-
head and contingency allowances,
O&M costs, and life cycle costs, at
each scale of operation. One set
of these tables is given as part of
the costing example below (Tables
5-10).
375
-------
ScaleEconomies--The averages
found by d i v i d in q tFtal life cycle
costs by the daily acceptance rates
of the site profiles were plotted
and graphically interpolated to de-
termine unit operation scale econo-
mies. Graphical interpolation by
visually fitting curves to the
points was performed in preference
to statistical curvefit technioues.
It should be emphasized that
although the upper and lower curves
represent the "typical" range in
life cycle average costs, they were
derived usina Dodae Guide city
index numbers for labor and mate-
rial subcomponents. As a result,
they do not represent extremes, and
are not statistically derived con-
fidence bounds.
These figures (one for each
unit operation) are given in the
contract report of this work, and
an example is given below (Figure
4).
Example unit operation
An example of remedial action
unit operation can be taken in the
revegetation of a landfill (Unit
Operation 3, Table 3). All other
unit operation were derived simi-
larly and reported in the same for-
mat in the contract report; each
consisted of six tables, followed
by one figure.
Table 5 describes this
unit operation in general terms,
beginning with a brief statement
regarding the use of this unit
operation. It then provides a
description of how this operation
is configured, and which other unit
operations are commonly used in
conjunction with revegetation. The
table ends with the critical as-
sumptions for applying this unit
operation to the site profiles.
This format is followed in each of
the general description tables.
The next five tables describe
how the revegatation unit operation
would be achieved for each of the
five scales of landfill operation.
Table 6 shows the component re-
quirements and their associated
costs when the revegetation unit
operation is applied to the 9.07
tonnes/day landfill site. In this
table both capital and O&M compon-
ent reouirements are identified and
cos ted in terms of mid-1980 dol-
lars.
As mentioned previously, the
costs associated with each compon-
ent may represent the sum of vari-
ous labor, material, and equipment
costs incurred in accomplishing
that component. For example, costs
for the "mulching" component are
for applying mulched hay over 9.86
ha (2.43 acres). The labor costs,
according to the 1980 Dodge Guide,
are typically $14.00/ha ($34.50/
acre). This number was adjusted
for regional labor cost differen-
tials and multiplied by the number
of hectares involved; these labor
costs ranged from $50 to $110. The
value for Newark, NJ was $100.
Material (hay) and equipment costs
were added to these labor costs to
obtain the mulching cost estimates
seen in Table 6. A similar process
was followed in costing each capi-
tal and O&M component.
Once all of the capital com-
ponents had been identified and
costed, allowances for overhead and
contingencies were added to com-
plete the capital cost portions of
these estimates. In all unit oper-
ations a 25% overhead allowance was
assumed. This assumption is partly
based on the fact that the Means
construction cost guide (6) also
assumes a 25% allowance for con-
tractor's overhead. A contingency
allowance to cover unforseen cost
additions was also applied to each
captial cost subtotal. In general,
the contingency allowance ranged
from 10% to 40% depending on the
extent to which the unit operation
was expected to encounter unforseen
difficulties.
As explained previously, each
of the O&M component costs were
escalated for future inflation and
then discounted to their present
values (in mid-1980 dollars).
These present values were summed
376
-------
TABLE 5. UNIT OPERATION 3 - REVEGETATION
Use: Revegetation helps to physically stabilize the earth material
and reduce infiltration; it also serves to minimize erosion
of the cover material by wind and water.
Confi guration: Revegetation involves first grading the landfill,
covering it with a suitable, fertile soil, adding soil
supplements, and then seeding.
Conjunctive Uses: Contour grading and gas migration control sys-
tems are used as cost components.
Assume: a) Entire surface of landfill is revegetated.
b) 0.6m of clay and silt loam will be used for landfill
cover.
c) Clay and silt loam are easily accessible; transporta-
tion costs are not included.
d) Mulch will be used to stabilize soil until vegetation
takes hold.
e) Native grasses will be used for seed.
377
-------
TABLE 6. REVEGETATION UNIT OPERATION AT LANDFILL OPERATION OF
9.07 TONNES/DAY
Total Unit $
Capital Cost of Components
(mid-1980)
Area Preparation
1) Excavation and Grading
Refuse (5,130 m3)
2) Hydroseeding )
3) Soil Supplements )
4) Mulching
Capi tal Cost (subtotal )
Overhead Allowance (25%)
Contingency Allowance (10%)
Capital Cost Total
0 & M Costs
Lower U.S.
670
8,130
880
270
9,950
2,490
1,000
13,440
Upper U.S.
1,160
9,420
1,240
390
12,210
3,050
1,220
16,480
Newark, NJ
1,040
9,090
1,110
350
11,590
2,900
1,160
15,650
a) Grass Mowing
(2.43 acres)
(6 mowings/year)
Refertili zation
(2.43 acres)
(1 x per year)
0 & M Cost Total
Total Life*
Cycle Cost
(over 10 years)
Average Life*
Cycle Cost
(By tonnes/day)
420
240
2,090
660
390
2,790
600
340
2,600
*See Text for Methodology and Assumptions.
378
-------
TABLE 7. REVEGETATION UNIT OPERATION AT LANDFILL OPERATION OF
45.36 TONNES/DAY
Total Unit $
Capital Cost of Components
(mid-1980) Lower
Area Preparation
1) Excavation and Grading
Refuse (17,075 m3)
2) Hydroseeding }
3) Soil Supplements )
4) Mulching
Capital Cost (subtotal )
Overhead Allowance (25)
Contingency Allowance (10%)
Capital Cost Total
0 & M Costs
a) Grass Mowing
(8.23 acres)
(6 mowings/year)
Ref erti 1 i zation
(8.23 acres)
(1 x /year)
0 & M Cost Total
Total Life*
Cycle Cost
2
27
2
33
8
3
44
1
2
62
U.S.
,290
,030
,990
910
,220
,300
,320
,840
,320
820
,140
,880
Upper
3
31
4
1
40
10
4
55
2
1
3
84
U.S.
,960
,330
,220
,300
,810
,200
,080
,090
,220
,320
,540
,930
Newark, NJ
3
30
3
1
38
9
3
52
1
1
3
78
,540
,250
,780
,170
,740
,690
,870
,300
,980
,150
,130
,690
(over 10 years'
Average Li fe*
Cycle Cost 1,390 1,870 1,730
(By tonnes/day)
*See Text for Methodology and Assumptions.
379
-------
TABLE 8. REVEGETATION UNIT OPERATION AT LANDFILL OPERATION OF
90.72 TONNES/DAY
Capital Cost of Components
(mid-1980)
Area Preparation
1) Excavation and Grading
Refuse (27,685 m3)
2) Hydroseeding )
3) Soil Supplements )
4) Mulching
Capital Cost (subtotal )
Overhead Allowance (25)
Contingency Allowance (10%)
Capital Cost Total
0 & M Costs
a) Grass Mowing
(13.38 acres)
(6 mowings/year)
Referti1i zation
(13.38 acres)
(1x1 year)
0 & M Cost Total
Total Life*
Cycle Cost
(over 10 years)
Average Li fe*
Cycle Cost
(By tonnes/day)
Total Unit $
Lower U.S. Upper U.S. Newark, NJ
3,710 6,440 5,760
43,820 50,790 49,050
72,730
2,160
1,340
3,500
102,240
1,130
4,860
1,480
53,870
13,470
5,390
6,860
2,120
66,210
16,550
6,620
6,130
1,910
62,850
15,710
6,290
89,380
3,600
2,140
5,740
137,770
1,520
84,850
3,240
1,870
5,110
127,930
1,410
*See Text for Methodology and Assumptions.
380
-------
TABLE 9. REVEGETATION UNIT OPERATION AT LANDFILL OPERATION OF
272.12 TONNES/DAY
Total Unit $
Capital Cost of Components
(mid-1980)
Area Preparation
1) Excavation and Grading
Refuse (65,100 m3)
2) Hydroseeding }
3) Soil Supplements )
4) Mulching
Capital Cost (subtotal)
Overhead Allowance (25)
Contingency Allowance (10/0
Capital Cost Total
0 & M Costs
Lower U.S.
8,770
103,060
11,490
3,490
126,810
31,700
12,680
171,190
Upper U.S.
15,230
119,440
16,220
5,010
155,900
38,980
15,590
210,470
Newark, NJ
13,610
115,350
14,510
4,500
147,970
37,000
14,800
199,770
a) Grass Mowing
(31.65 acres)
(6 mowings/year)
Referti1i zation
(31.65 acres)
(1 x/year)
0 & M Cost Total
Total Life*
Cycle Cost
(over 10 years)
Average Life*
Cycle Cost
(By tonnes/day)
5,100
3,170
8,270
240,§10
890
8,520
5,060
13,580
324,950
1,190
7,620
4,430
12,050
301,350
1,110
*See Text for Methodology and Assumptions.
381
-------
TABLE 10. REVEGETATION UNIT OPERATION AT LANDFILL OPERATION OF
453.60 TONNES/DAY
Total Unit $
Capital Cost of Components
(mid-1980)
Area Preparation
1) Excavation and Grading
Refuse (93,935 m3)
2) Hydroseeding }
3) Soil Supplements )
4) Mulching
Capital Cost (subtotal)
Overhead Allowance (25%)
Contingency Allowance (10%)
Capital Cost Total
0 & M Costs
a) Grass Mowing
(45.74 acres)
(6 mowings/year)
Ref erti 1 i zation
(45.74 acres)
( 1 x/year)
0 & M Cost Total
Total Life*
Cycle Cost
Lower U.S.
12,680
148,710
16,600
5,060
183,050
45,760
18,300
247,110
7,380
4,570
11,950
347,850
Upper U.S.
22,020
172,340
23,460
7,250
225,070
56,270
22,510
303,850
12,360
7,320
19,680
469,760
Newark, NJ
19,680
166,430
20,970
6,510
213,590
53,400
21,360
288,350
10,980
6,400
17,380
434,870
(over 10 years)
Average Life*
Cycle Cost
(By tonnes/day)
770
1,040
960
rSee Text for Methodology and Assumptions.
382
-------
over a 10-year life cycle of future
site maintenance and then added to
the capital cost total to determine
total life cycle cost. Averaqe
life cycle cost was computed by
dividing this total by the land-
fill's former daily scale of opera-
tion.
Similar cost analyses for ap-
plying revegetation to the other
scales of landfill operation appear
in Tables 7-10.
The life cycle average costs
from Tables 6-10 were plotted in
Figure 4. As mentioned above, the
cost points on Figure 4 are average
life cycle costs per tonne of waste
deposited during the daily opera-
tion of the landfill. The curves
presented on this fiqure were
created by visually interpolating
between the points to graphically
denict scale economies. Because of
the small number of points per
curve, statistical curvefit methods
were not attempted.
Application and limitations
The contract report document
produced as a result of this
review-and-update work consists
primarily of unit operation costs
at the component and sub-component
levels, and average life cycle
costs for each scale of operations.
By applying this cost information
to the report's discussion of
alternative or conjunctive unit
operations, enforcement personnel
will be able to configure and cost
complete remedial action scenarios.
The following is an example of
how a remedial action scenario in-
volving several unit operations may
be costed by using the scale eco-
nomy figures for each constituent
unit operation. The scale of opera-
tion was purposely selected to be
different from the scales used in
developing the unit operation con-
ceptual design cost estimates.
The Problem
The hypothetical hazardous
waste landfill site is producing
leachate from surface water intru-
sion; the leachate is contaminating
nearby ground water. It has also
been found that the bottom 1 m of
the landfill is in ground water.
The scale of operation is 300
tonnes/day.
Remedial Action Scenario Solution
The solution presented here is
only one of many ways that the
problem outlined above may be re-
duced, and is aiven to show how a
long-term response action can be
developed for a problem site. The
solution uses the following unit
operations as part of the remedial
action scenario:
• Contour grading and surface
water di versi on .
0 Revegetation.
• Sheet piling cutoff wall.
Contour grading and surface
water diversion promote and channel
surface water runoff by grading the
landfill and creating earth berms
to drain water away from the land-
fill. Revegetation helps to sta-
bilize the earth materials and
reduce infiltration. The sheet
piling cutoff wall is used as a
physical barrier to lower the water
table around the landfill, so that
the waste present is no long satu-
rated .
Table 11 is a breakdown of
average life cycle costs for unit
operations used to clean up a land-
fill which received 300 tonnes/day
of hazardous waste during fts oper-
ational days. Each of the figures
shown was obtained by reading the
average life cycle cost curve for
that unit operation at the stated
scale of operation. The unit oper-
ation costs are then summed to give
the total average life cycle cost
for the proposed remedial action
scenario.
It should be pointed out that
although this process of reading
costs from figures is not neces-
sarily inappropriate for certain
383
-------
453.60
90.72
LANDFILL SCALE OF OPERATIONS IN TONNES/DAY
Figure 4. Scale economies of revegetation
384
-------
TABLE 11. LONG-TERM REMEDIAL ACTION AVERAGE LIFE CYCLE
COSTS FOR A 300-TONNE/DAY LANDFILL
Unit Operations Lower U.S. Upper U.S. Newark, MJ
Contour Grading and 730 870 840
Surface Water Diver-
sion
Reveoetation
Sheet Pil ing Cut-off
Wai 1
Total Average Life
Cycle Costs
850
4,500
6,080
1,180
6,500
8,550
1,080
5,600
7,520
applications, it does introduce
measurement errors. It may also
ignore component duplications or
scale economies resulting from the
combination of several related unit
operati ons.
While the intention of this
unit operation costing is to allow
comparison of the relative costs of
several remedial action scenarios,
it must be emphasized that the
resulting cost figures are averages
and not absolute costs. The U.S.
high and low figures do not repre-
sent actual limiting values, but
qive an average of ranges across
cities in the continental United
States. Because these figures are
averages and not absolutes, they
should not be used to determine the
actual or expected cost of a reme-
dial action scenario, but to com-
pare the relative costs of several
scenarios.
Factors which should be con-
sidered by anyone using the docu-
ment are: 1) the costs for eouip-
ment are not permitted to vary
across the nation, and 2) all costs
are in terms of mid-1980 dollars,
except O&M costs.
Summary
A total of thirty-five con-
ceptual design unit operations, ad-
dressing either landfill or surface
impoundment disposal sites, were
identified in the literature and
their cost estimates were updated.
The results were presented in terms
of a consistent format and method-
ology.
As part of this updating pro-
cess, revised site profiles for
landfill and impoundment conditions
were developed at each of five
scales of operation. Remedial
action unit operation conceptual
designs appearing in the literature
were then refined to best address
each operating level of these land-
fill or impoundment site profiles.
Costs were estimated for each
of the unit operations according to
a consistent computational methorl-
oloqy, in which capital and operat-
ing cost estimates were combined to
determine total and average life
cycle costs for the example loca-
tion of Newark, NJ, as well as for
the average lower and upper U.S.
costs, within the continental
states. All costs were estimated
in terms of mid-1980 dollars.
The average life cycle costs
for each unit operation were plot-
ted against the five scales of site
operation. Cost curves were graph-
ically interpolated between point
estimates to determine unit opera-
tion scale economies.
Because complete remedial ac-
tion scenarios for uncontrolled
385
-------
sites typically consist of several
unit operations, much remains to be
done, even from a conceptual design
cost perspective, in identifying
and costina the most promisina sce-
narios. Such an effort might in-
clude a systematic evaluation to
determine the most prevalent pollu-
tion problems occurrina at uncon-
trolled landfill and impoundment
sites. Once this set of "typical"
pollution cases has been deter-
mined, likely remedial action sce-
narios could be configured usinq
unit operations developed in this
study. The resulting composite
cost estimates for these scenarios
could then be compared to determine
the relative cost advantaoes of
each alternative.
References
Current Business Statistics
Survey of Current Business,
Vol . 60, No. 4 U.
ment of Commerce,
Economic Analysis.
1980.
S. Depart-
Bureau of
April
Fred C. Hart Associates, Inc.
Analysis of the Technology,
Prevalence and Economics of
Landfill Disposal of Solid
Waste in the United States -
Volume II. EPA Contract No.
68-01-4895, U.S. Environmental
Protection Agency, Office of
Solid Waste, Washington, D.C.
1979. 97 DP.
Geraghty and Miller, Inc. Sur-
face Impoundments and Their
Effects on Ground-Water Quality
in the United States. EPA-
570/9-78-004, U.S. Environmen-
tal Protection Agency, Office
of Drinking Water, Washington,
D.C. 1978'. 275 pp.
McMahon, L. A. 1980 Dodge Guide
to Public Works and Heavy Con-
struction Costs. McGraw-Hill,
New York, New York. 1979.
National Income and Product
Tables. Survey of Current
Business, Vol. 60, No. 4. U.S.
Department of Commerce, Bureau
of Economic Analysis. April
1980.
Robert S. Means Company, Inc.
Buildina Construction Cost Data
1980. 1979. 371 pp.
SCS Engineers. Study of Ongo-
ina and Completed Remedial Ac-
tion Projects: Survey Results
and Recommended Case Studies.
Environmental Protection Agency
Solid and Hazardous Waste Re-
search Division, Cincinnati,
Ohio. 1980.
SCS Engineers. Surface Im-
poundment Assessment in Cali-
fornia. EPA Contract No. 68-
01-5137, U.S. Environmental
Protection Agency, Office of
Drinking Water, Washington,
D.C. 1980.
386
-------
REGULATING ILLEGAL DUMPING OF HAZARDOUS WASTES
Edward Yang, Bob Anderson, and Roger Dower
Environmental Law Institute
Washington, D.C. 20036
ABSTRACT
Efficient regulation of hazardous waste management practices is an important conerstune
of implementation of Resource Conservation and Recovery Act (RCRA). However, efficient
regulations can not be achieved without considering compliance as an important variable
in determining appropriate level of standards and requirements of the regulations. In
order to do this, our paper develops a compliance model based on the relative costs of
compliance and noncompliance. The model assumes the financial consideration of the
hazardous waste managers to be the primary motivation to comply or not. The model is
constructed on an industry level to address the policy-makers' concern on the percentage
of industry compliance. The model also allows the policy-makers to weight the analysis
by the various policy variables, such as number of firms, employment size, value of
shipment and enforcement strategy. Using the data gathered from several case studies
of illegal dumping of hazardous waste, the model was partially applied to show the
cost differential between proper and improper management practice. Although the data
limitation prevented a full operation of the model, the possible effects of compliance
cost on the level of industry compliance are demonstrated.
Introduction
In the past, environmental management
models have often stressed deriving a soc-
ially optimal level of regulation where
social demand for environmental quality
equals the social supply of it.l The ap-
plication of the models has encountered
several difficulties. First, the determin-
ation of an optimal level of environmental
quality,is often done in the political arena
which is difficult to be explicitly incor-
porated into the model, since it is based
on a social preference function. Second,
even for a given set of social preferences,
the difficulty of measuring social welfare
loss due to misallocation of resources of-
ten deters the operation of the model,
Third, much of the regulatory inefficiency
may come from not using the environmental
standard in a most efficient way, rather
than from not being able to determine the
socially optimal level of environmental
quality. The first two difficulties have
been recognized and dealt with by a large
body of social and economic research.2
However, it is only recently that the in-
efficiency involved in standard or criteria
setting has been rigorously examined. 3
This paper examines the problem of illegal
dumping of hazardous waste in the light of
such inefficiency through the use of a com-
pliance model.4,5 The use of a compliance
model is especially important for regula-
tions that are difficult to enforce, such
as illegal dumping of hazardous waste,
where the release of waste is often un-
detected and its harmful effects are dif-
ficult to demonstrate.
Until recently the implicit assump-
tions underlying environmental regulatory
analyses have often been that the question
of compliance is a separate issue from
the one of determining optimal level of
environmental protection. When such as-
sumption is unrealistic, the net effect
is that the social cost of regulation be-
comes underestimated. In these analyses
where the optimal level of pollution abate-
ment is determined by the demand and sup-
ply of the abatement, such underestimation
387
-------
can provide misleading results. In reality,
full compliance can not be taken for grant-
ed. Furthermore, the setting of regulatory
requirement inherently influences the level
of .compliance through compliance cost. In
other words, the cost of compliance becomes
a crucial variable in determining the suc-
cess of the regulatory mechanism.
In studying hazardous waste policy
we are interested in the relationship be-
tween the cost of complying with the RCRA
regulations and cost incurred by illegal
dumping. Prior to RCRA, much of the impro-
per release or dumping of the waste was
done to save disposal cost. Although, some
careless dumping may have arisen from igno-
rance concerning the hazardous nature to
the waste, to perceive all of the past im-
proper waste disposal activities as being
caused by ignorance is naive. Costs do
affect a Firm's decisions; even with strict
RCRA regulations defining hazardous wastes
and their legal disposal, improper disposal
will continue as long as there is a signi-
ficant reduction in disposal costs result-
ing from noncompliance.
On the applied level, the purpose of
this paper is to examine the effects of
RCRA on illegal dumping. More specifically,
we can ask the question, "Will RCRA regula-
tions promote the optimal amount of compli-
ance?" The very posing of this question
implies that we are not taking full indus-
try compliance for granted. Furthermore,
the question implies that the level of com-
pliance may depend on the stringency and
structure of the regulations. We will di-
rect our attention to the financial factors
underlying the decision to comply. The
groundwork for our compliance model appear-
ed in the literature on the economics of
crime.6 However, it was not until Viscusi
and Zeckhauser's recent work on optimal
standards that the level of compliance was
linked to the cost of compliance. Their
work showed that there is a point in set-
ting standard beyond which further tighten-
ing of that standard will reduce the over-
all effectiveness of the regulation. The
regulator should strive to approach that
point as closely as possible without going
beyond it. Although, the work is path-
breaking in its theoretical content, the
application of this model remains to be at-
tempt; the information requirement of this
model is prohibitive.
In pursuing the purpose of this pa-
per, we have put some effort into recon-
structing the Viscuci and Zeckhauser (VZ)
model so that it can be more susceptible
to application. We will first illustrate
the concept of optimal standard through a
number of graphs and then proceed to con-
struct a model that can utilize information
on cost differential between compliance
and noncompliance. The model will also
allow the regulator to perceive compliance
from an industry level, given information
on the cost differential distribution
across the industry. In addition, the
policy maker will be able to weigh the
criteria for choosing optimal standard
with policy variables of his or her con-
cern. The advantage of this model is that
it can be utilized on several levels de-
pending upon the availability of relevant
information. We will show that in the
case of full information the optimal stan-
dard can be easily derived.
Ideally, the paper would then pro-
ceed to apply the above model to determine
whether the RCRA regulatory requirements
are optimal in soliciting the maximum level
of industry compliance. But even on the
simplest level of the above model such a
task is beyond the resources of this study;
considerable amount of additional effort
in collecting data on costs of compliance
will be required. As we will explain, in
a later section, there are limitations
in qualifying nonmonetary costs, among
which is the value of health and environ-
mental damages.
We will be able to partially apply the
model to some of the data on cost differen-
tials from the case studies that we have
conducted. Such a step is performed only
for an illustrative purpose, since without
further data it would be impossible to con-
clude whether the RCRA's regulatory require-
ments are optimal. Nevertheless, this ex-
cercise should show the operational feasi-
bility of the model.
The Concept of Optimal Regulatory Require-
ments
Given the legislative objective of
RCRA to prevent future improper disposal of
hazardous wastes, the regulatory agency
must choose a set of requirements that
would be most effective in achieving this
388
-------
objective. We start by examining a firm's
decision to comply with a certain standard
required by regulation. In Figure I we
have Firm A's marginal cost curve, Ca, for
complying with a continuing level of stan-
dard, and a marginal cost curve, NC, for
not complying with the required standard.
With the horizontal axis representing an
increasing level of standard, cost curve Ca
slopes upward because we assume that it is
more expensive to comply with the more
stringent standards. The cost curve, NC,
is constant. This assumes that regardless
of what level of standard the firm incurs,
fines, penalties and third party liability
for noncompliance remain the same. However,
it is possible that noncompliance penalties
and fines or the probability of getting
caught may change with new standards (e.g.
the RCRA manifest system) so that NC be-
comes nonconstant. Since it is not our
purpose to deal in-depth with the structure
of noncompliance cost, we will merely point
out that it consists of the expected value
of loss from being caught and whatever cost
is incurred during the illegal activity.
The loss can be nonmonetary, such as cor-
porate image and fear of imprisonment in
the case of a criminal penalty.
Cost
C
E
A
^
X
NC2
Nr
1
S1
S* S3
Standard
Figure I:
Firm's Decision to Comply or
Not
Whether Firm A will comply depends on
whether the standard is to the left or
right of point A, where the marginal non-
compliance cost equals the marginal compli-
ance cost. If the standard is set at S-"-,
firm A will certainly comply since the cost
of noncompliance is greater than the cost
of compliance by the amount CE. However,
if the standard is set at $3, the firm
would rather not comply. Of course one can
raise the cost of noncompliance to NC2 so
that S3 becomes a compliance point. This
occurs when penalties and fines of enforce-
ment efforts are raised.
When industry has numerous producers
with varying cost curves, the regulator
faces the situation depicted in Figure II.
Cost
Figure II:
S2 S3
Compliance Decisions on An
Indus t ry _Leve1
There are five firms with Firm A
and Firm E having the highest and lowest
cost structures respectively for disposal
of certain hazardous wastes. Let us say
that before the promulgation of any regu-
lation, all firms were handling waste in
a manner equivalent to standard S^. As-
sume a regulator imposes the standard Si.
When this occurs all of the firms, with
the exception of Firm A, would comply and
improve their disposal practice to S^. The
improvement in environmental quality due to
Firms B, C, D, and E complying with stan-
dard S-"- is shown in Figure III.
Environmental
Benefit
Standard
Figure III: Environmental Improvement Due
to Additional Firms in compliance
If the regulator were to decide to push the
standard to S2, only firms C, D, and E
would upgrade their practice and comply.
Firm B, not only would not comply, but may
go back to SO since it would not be in
compliance anyway. This additional in-
crease in the standard cause two effects:
One is the additional improvement in the
environment, as registered in Figure III,
the other is a reduction in environmental
quality due to Firm B dropping out of the
compliance group and going back to its
original level of waste handling practice
indicated as S . This negative effect is
rsgistered by curve ED in Figure IV.
389
-------
Environmental Loss
Standard
Figure IV: Environmental Degradation Due
Firms Dropping out of Compliance
ihe shapes of El and ED curves warrant some
examination. The environmental improvement
due to firms complying with the standards
is shown by El increasing at a decreasing
rate. The underlying assumption is that
as the standard becomes increasingly strin-
gent, fewer firms will be willing to fur-
ther improve their practices to be in com-
pliance. This can be seen in Figure II as
we move along the horizontal axis repre-
senting higher levels of standards. At the
same time, the environmental degradation
due to the firms dropping out of the com-
pliance group increases because they are
further away from the original level S .
In other words, the opportunity cost of
losing a firm to noncompliance is higher
at a high level of standard. Thus the ED
curve rises at an increasing rate in Figure
IV.
The optimal standard (the one that re-
sults in the largest improvement in envi-'
ronmental quality) can be determined by
combining the two curves, El and ED, as
shown in Figure V. At S* the net environ-
Units of Environmental
Quality ED/
A, "
Standard
Figure V:Maximum Benefit from Optimal Std.
mental improvement is maximized. In an
economic framework, when the marginal cost
of raising the level of standard equals
the marginal benefit of the move, the stan-
dard is at its optimal level. Maximum com-
pliance is achieved through the cost mini-
mization behavior of the industry.
Compliance Model
It is not difficult to see that the
cornerstone to the application of the above
conceptual model is the ability to identify
the marginal benefit and loss at each level
of the standard. The benefit is derived
from the firms that will remain in the com-
pliance group and upgrade their practices.
The loss comes from the firms that will
drop out of the compliance group and re-
turn to their original improper practices.
The link between the levels of standard
and the entry and drop rates of the two
groups are the cost differentials between
compliance and noncompliance. The model
to be presented primarily builds around
the relationships between the marginal be-
nefit and loss of the standards and the
distribution of cost differentials across
the firms that will be affected by the
regulation.
Our compliance model can best be il-
lustrated graphically. Figure VI presents
a histogram showing the number of firms
for each cost differential category for a
given level of standard. The positive
quadrant's horizontal axis represents how
much more expensive it is to comply with
the standards than not to. The negative
quadrant's horizontal axis represents how
Number of Firms
E
D
C
R
A
Cost Differential
(Compliance cost -
Noncompliance cost)
Figure VI: Cost Diffarential Distribution
for Given Standard
much less expensive it is to comply with
the standard than not to. Assuming that
firms respond to cost advantage, the right
quadrant contains the firms that would not
comply and the left quadrant contains the
firms that would. Figure VI is simply
another way of presenting the information
contained in Figure II, except it is for
one standard at a time instead of one firm
for each cost differential. We now have
the number of firms for each cost differen-
tial categorgy. The latter change reduces
the information requirement of the model.
390
-------
The static situation in Figure VI
can be made dynamic by introducing differ-
ent levels of standard. Changing the stan-
dard would move the histogram along the
horizontal axis. Standards increasing the
cost of compliance will shift it to the
right; ones decreasing the cost will shift
it to the left. For example, the regulator
can increase the stringency of the standard
to the point at which category C enters the
noncompliance group. This generalization
is necessary since optimization of the
standard must allow for variation in the
level of standard. However, this is done
through the employment of some assumptions.
First, the ranking of the categories must
be transitive: if A is greater than B and
B is greater than C, A must be greater than
C. Second, the ranking must be invariant
as to the choice of the level of standard.
These assumptions are not unreasonable, al-
though the latter one may encounter trouble
when the nature of the standard changes,
allowing firms to trade places. For the
model to operate we also need to be able
to identify the relationship between the
standards and the cost differentials. Such
a relationship can be either derived from
engineering data or estimated from several
sample points. Given these assumptions we
can assess the portion of industry under
compliance.
Based on our conceptual model, as we
incrementally tighten up the standard,
there will be a marginal benefit and loss.
Looking at Figure VI the question is,
"what is the marginal benefit and loss when
we shift the histogram so that one category
of the firms will move into the noncompli-
ance region?" We know some of the factors
determining the marginal benefit and loss:
the number of firms leaving the compliance
region, the ones remaining in the region,
the volume of waste generated by each
category, and the level of standard where
the firms drop out of compliance. We can
even replace the assumption of constant
marginal noncompliance cost with one of
variable schedule based on firm size, since
large firms tend to have a higher detection
rate when they do not comply. Such factors
can all be included into the model by as-
signing weights to the categories, as long
as there exists a relationship between the
factor and the cost differential. Looking
at Figure VI it is obvious that the weigh-
ting factor is the number of firms. In
other words, the assumption is that the
number of firms in compliance or noncompli-
ance is the policy variable with which we
are concerned. Another choice is to weight
the categories by the volume of waste han-
dled. Or if it is considered important that
firms with a large cost differential, due
to their smallness in size, are fnore im-
portant the firms toward the rieht in
Figure VI should be weighted heavier.
Once the relevant factors are weighted, we
can construct a net benefit index represen-
ted by the vertical axis in Figure VII.
Cost Differential
Figure VII
Distribution of Net Benefit by Cost
Differential
The histogram is now turned into a smooth
curve for the purpose of illustration.
TIow we can relate the cost differential
back to the optimality condition. The
areas under the curve are the net benefits
of having the firms, based on their cost
differentials, in the two regions. A
standard should be chosen so that the
vertical axis representing the breakeven
point hits the top of the curve. At this
point the standard is set so that its
marginal loss, the slope of curve just
to the right of the vertical axis, equals
its marginal benefit, the slope just to
the left of the axis.
Some comment on the shape of the curve
in Figure VII is required, since the ex-
istence of optimality depends upon it. It
was demonstrated earlier that without any
weighting scheme, an optimal point exists
because of the shapes of El and ED in
Figure V. Note that the vertical distance
between the two curves corresponds to the
curve in Figure VII. The question is
whether weights based on other factors will
change the curves so much that an optimal
point can not exist. For example, what
happens if the distribution of the numbers
of the firms in each category produces two
peaks, one at each end? It is then clear
that two standards instead of one are
necessary to maximize the net benefit.
What if small firms are considered all
391
-------
important to be kept in compliance? Such
a weighting will create a peak at the very
end on the right-hand side of the curve so
that as soon as a standard is increased
from the minimum, the peak will intersect
the'vertical axis; further strengthening
the standard will be undesirable. What if
each category is valued equally? It is
clear that no optimum exists since the
curve will be a simple plateau. The re-
sult is again in accordance with the
weighting scheme; without any ranking of
the category one can set the standard any-
where. In summary, the resulting optimal
standard always reflects the underlying
assumptions considered to be important in
setting of standards.
Limitations
It is clear that two types of diffi-
culties will be encountered in utilizing
the above compliance model. The first one
is still the information requirements.
These are: 1) the cost differential be-
tween legal and illegal disposal of the
waste, and 2) the distribution of the cost
differentials across the firms. In regards
to the first type of information, one can
obtain market values of wastes that are
legally disposed of off-site, based on
what the disposer charges. For the wastes
that are disposed of on-site, the cost has
to be constructed from the firm's engineer-
ing data.
The cost of illegal disposal is more
difficult to obtain. It consists of sev-
eral parts. One is the actual cost of il-
legally dumping the waste. This usually
involves straightforward calculation of
the cost of hauling the waste from the
generator to the dumping site. In some
cases, there may be little or no cost in-
volved, e.g. disposing the waste in the
existing landfill that does not comply
with the RCRA standards. However, the
major portion of the cost of illegal dump-
ing comes from paying for the damages
caused by the dumping when detected. This
cost can be difficult to obtain, since as-
signing values to damages to health and
the environment still faces various tech-
nical difficulties. ^hysical well-being
and environmental goods are not traded on
markets. Estimating the severity of the
damage is also often impossible since
scientific knowledge can not always estab-
lish the links between the exposure and
final effects. However, since the deci-
sion to dispose of the waste legally con-
cerns liability for damages rather than
the true value of the damages, two factors
can serve as bench marks. One of them is
the financial ability of the illegal dis-
poser to pay for the damage. Any assess-
ment beyond the ability to pay will simply
force the disposer to declare bankruptcy.
When damages are likely to be high the
disposer uses the net worth of his opera-
tion as the cost of noncompliance. The
other factor is what the court assesses as
damage. This means financial data of the
firms involved with hazardous wastes mana-
gement and past court decisions can be used
to derive the cost differentials. Similar-
ly, although the cleanup cost is difficult
to estimate, its accuracy will not affect
the cost differential once it is beyond the
financial ability of the disposer. Finally,
we have to consider fines-and penalties.
These costs, relatively small compared to
the others, should be calculated at the
maximum allowable by law. However, one
complication may come from valuing corpor-t
ate image and imprisonment. Firms realize
that a negative image, such as one of
social irresponsibility in dumping hazar-
dous waste may lead to loss of business.
It is also likely, since RCRA allows for
criminal charges, that a corporate person
would avoid possible jail sentences at a
high cost. These intangible effects must
be incorporated into the cost of noncompli-
ance. A difficulty also arises when one
has to translate these costs into expected
value. This requires knowledge of perceiv-
ed possibilities of detection and convic-
tion. In other words, the disposer adjusts
the cost of noncompliance by the possibili-
ty that he will be penalized and held res-
ponsible for the damages. For regulations
that have been in effect for a while, some
type of detection rate can be calculated
from the enforcement record. The RCRA re-
gulations, being relatively new, probably
would not have such data. The researcher
may have to assume several hypothetical
detection rates and conduct a sensitivity
analysis.
There are also difficulties in obtain-
ing information on the distribution of the
cost differential between compliance and
noncompliance across the firms that are
affected by the regulation. But this fac-
tor is crucial since determination of opti-
mal regulatory standards can only be made
at the industry level. For industries with
few firms such data may be constructed
392
-------
from various existing industry profile data,
or simply a survey of ongoing disposal tech-
niques. For larger industries that generate
and manage their own hazardous waste, an
accurate account of the cost differentials
may be impossible. An approximation has
to be made through two steps. First,
classes of representative firms have to be
selected and their cost differentials cal-
culated. Second, the distribution of
these representative firms have to be as-
sumed based on available data.
The second type of difficulty is to
decide on the weighting scheme to be used
for arriving at the net benefit. We admit
that on technical grounds that more re-
search is needed. However, on a concep-
tual level such practice can be defended.
This is because it simply makes explicit
the assumptions that regulators utilize.
The weighting scheme also allows for sen-
sitivity tests which enable the ultimate
decision maker to form a range of alterna-
tives, rather than accept or reject a
single standard.
Construction of Costs Differentials?
As aforementioned, since we do not
have the relevant data to apply to the
compliance model completely, this section
will only present the mechanics of calcu-
lating the cost differential between com-
pliance based on past cases of illegal
dumping. By analyzing the various types
of cost in these cases, we can make some
preliminary judgement as to whether firms
of similar operation and size will comply
to the existing RCRA regulations. The as-
sumption underlying the judgement is that
the cost of proper disposal that the il-
legal dumpers refused to incur represents
the cost of complying with the existing
RCRA regulations. The assumption is rea-
sonable since the ultimate effect of the
RCRA regulations is to induce proper dis-
posal. The exercise in this section will
end after cost differentials are calcula-
ted for the firms in the case studies.
But theoretically the calculation can be
repeated for firms of other sizes until
enough categories can be created for sub-
sequent analysis.
Each of the four case studies includ-
ed for the cost differential analysis will
be briefly summarized first. Different
types of costs for the cases are calcula-
ted and put infio. For illegal
dumping or noncompliance we have the direct
cost and expected cost. The former is the
actual cost involved in illegally dumping
the wastes. The latter is the expected
cost resulting from the release being de-
tected and the disposer found responsible.
It consists of clean-up cost, fines and
penalties and health and environmental
damages. Although, the last item is listed
in Table I, it is not quantified. Qualita-
tive descriptions are made to the extent
possible. The case studies also
provided the direct cost of compliance or
legal disposal of wastes. These costs are
estimated by quoting what legitimate waste
disposers would charge for disposing the
wastes that was illegally dumped. These
cases are:
A. Bullit County, Kentucky (Valley of
the Drums)
The Valley of the Drums is located in
a rural area in Bullitt County, Kentucky,
fifteen miles south of Louisville. From
1967 through 1978 over 17,051 drums of
chemical wastes were illegally stored on
the ten-acre site. A number of drums were
buried or emptied directly into pits.
Soil and water samples taken by EPA have
detected 197 organic chemicals and 28
metals, including 21 suspected carcinogens,
mutagens, and teraogens and 26 priority
water pollutants.
1. Direct and Expected Costs of
Illegal Disposal
An estimate of the direct cost of il-
legal disposal or organic chemical wastes
on the site is derived from the fee charg-
ed to industry to transport and dispose of
the wastes and the total volume of wastes
illegally dumped.
Taylor, the owner of the site, charged
the industries involved 2.50 dollars per
drum to transport and dispose of chemical
wastes on his property. At the time of
the sites closure in 1979, 17,051 drums of
organic chemical wastes were present on
the site. Combining these figures, the
total direct cost of illegal chemical waste
disposal on the site is 42,627 dollars.
The expected costs of illegally dis-
posing of wastes on the site include the
cost of cleaning up the property, the esti-
mated cost of health and environmental da-
393
-------
TABLE I
Costs of Legal and Illegal Dumping of
Hazardous Wastes from Selected Cases
Case
Bullitt
County
Kentucky
Chester
Penn.
North
Carolina
Dover
N.J.
Chemical
Organic
Compo-
nents
Hazar-
dous
materi-
als
PCB
Organ-
ic
Chemical
Illegal Dumping
Volume
17,051
Drums
50,000
30,000
gal-
lons
5,500
bar-
rels
Direct
Cost
42,627
375,000
30,000
20,000
Expected Cost
Clean Up
Cost
3,148,000
1,250,000-
3,000,000
1,580,000
44,350
Health and
Environment
Damage
*
*
*
*
Fines and
penalties
(max)
10,000
10,000
10,000
10,000
Legal Dumping
Direct Cost
255,000-426,000
1,750,000-2,500,000
.90,000-120,000
21,175-30,250
*Nonquantifiable
Source: ELI Case Studies
-------
mages, and fines or penalties imposed on
the owner of the property.
The total cost of cleaning up the site
is expected to reach $3,148,000. U.S. EPA,
Kentucky's Department of Natural Resources
and private industry have provided the
funding for clean up efforts thus far. EPA
has contributed the largest share, with
$328,000 spent for general clean-up opera-
tions and installing a recharge filtration
system, and $200,000 spent for analysis of
soil and water samples. In addition, the
clean-up operations have cost EPA an un-
determined amount in general administrative
expenses.
The state of Kentucky has spent
$24,391 in connection with the Valley of
the Drums. The majority of this has been
spent on arranging and supervising the
clean-up efforts. In addition, the State
expects to spend $10,000 to clean and re-
charge a filtration system on the site.
Five industrial firms who generated
much of the waste dumped at the site have
agreed to pick up and properly dispose of
the wastes that they can identify as having
originated with them. The five major gene-
rators of waste at the site expected to
spend approximately $85,000 to remove
1,300 of the 17,051 drums present on the
surface of the site.
Kentucky's state officals estimate
that an additional $2.5 million will be
required to completely analyze, remove,
and properly dispose of the wastes dumped
at the Valley of the Drums, bringing the
total expected clean-up costs to $3,148,
000.
The expected health and environmental
costs of the -illegal disposal of chemical
wastes at the site have not been quanti-
fied. Presumably, there has been a loss
in the recreational value of Wilson Creek
which abuts the site. The State of Ken-
tucky has advised against fishing in the
creek due to PCB contamination of fish.
Property damages have also not been
quantified for the illegal disposal of
chemical wastes on the site. However, the
real property value of the land was asses-
sed at $10,000 for County tax rolls in
1978. No one has been willing to purchase
the property although the owner has been
trying to sell the dump site. Therefore,
the value of the property has, presumably,
been decreased.
The third category of expected costs
from the illegal disposal of chemical
wastes on the site are fines or penalties
envoked against the owner of the site. At
the present time no such action has taken
place. However, pursuant to the Compre-
hensive Environmental Response, Compensa-
tion, and Liability Act of 1980, (Superfun±) ,
an illegal disposer of hazardous wastes,
as defined under Section 102 of the Act,
shall upon conviction be fined not more
than $10,000. The expected costs of fines
or penalties against the owner of the site
are therefore, quantified in this summary
as the maximum allowable, or $10,000.
2. Direct Cost of Legal Dumping
Proper disposal of the wastes found
at the Valley of the Drums would have cost
between $15 and $25 per drum in 1976. The
direct cost of properly disposing of the
17,051 drums of wastes present on the sur-
fact of the site would have cost $255,000
in 1976 at the cost of $15 per drum and
$426,000 at the cost of $25 per drum.
B. Chester Pennsylvania
Between 1974 and 1977 approximately
50,000 drums of chemical wastes were il-
legally stored and dumped on an industrial
plant site along the Delaware River in
Chester, Pennsylvania. The Pennsylvania
Department of Natural Resources discovered
the illegal site in 1977 and ordered it
shut down. Two parties were involved in
the illegal disposal: Wade, the owner
of the site, and ABM disposal services, a
waste transport company. In addition,
Wade also profitted from an unidentifiable
amount of drums to a near by waste recy-
cles.
1. Direct and Expected Lost of
Illegal Disposal
The total direct cost of illegal dis-
posal of hazardous wastes on the Chester
Site is $375,000. This figure is derived
from combining the amount ABM waste trans-
port services paid wade, the owner of the
site to accept the drums of waste and the
$6.00 profit wade gained from emptying
the drums of waste and selling the emptied
drums to a recycler.
395
-------
The expected losts from the illegal
disposal include Clean-up costs, health
and environmental damages, and the expec-
ted cost to the illegal disposer from
fine.8 or penalties.
Total Clean-up costs for the Chester
Site are expected to range from $1,250,000
to $3,000,000. A breakdown of these costs
includes $650,000 to remove drums from the
site, and the remainder for clean-up of
the Delaware River, removing the surface
layer of soil from the site, and whatever
measures may be needed to prevent the fur-
ther spread of chemical contaminants.
Health and environmental damages have
not been quantified for the Chester site.
A substantial amount of heavy metals have
contaminated the Delaware River and bottom
sediment. U.S. EPA and the Pennsylvania
Department of Natural Resoruces are current-
lyinvestigating the amount of damage to the
river and nearby groundwater services.
Property damages which have resulted
primarily affect the value of the land
where the wastes were illegally dumped.
The owner has stated that the property was
worth approximately $4,000,000, however,
the most prospective buyers have offered
for it is $750,000.
There have not been any penalties or
fines assessed against the owner of the
site. However, for purposes of this analy-
sis we have used the $10,000 maximum allow-
able penalty for violation of Section 102
of (Superfund).
2. Direct Cost of Legal Disposal
Proper disposal of the hazardous wastes
present in the Chester site would have cost
approximately $35 to $50 per drum. This
estimate is based on the average waste dis-
posal charges of legally permitted disposal
facilities located in the vicinity of the
site. Combining these figures with the
volume of waste illegally disposed at the
site, the total cost of legal disposal of
the wastes would have ranged from $1,700,
000 to $2,500,000.
C. PCB in North Carolina
During a two mongh period in 1978
Robert Burns illegally disposed of approxi-
mately 30,000 gallons of transformer oil
along 211 miles of rural foods in North
Carolina. Federal regulations issued in
the srping of 1978 permitted the disposal
of PCB oils only in incinerators; since no
incenerators were operating in 1978 the oil
could only be stored. However, Burnes con-
tracted with Ward Transformer Company to
illegally dispose of the waste, for this
Ward received a kick-back from Burnes.
1. Direct and Expected Losts of
Illegal Disposal
The total direct cost from the illegal
disposal of PCBs is $30,000. This figure
was derived from the amount paid to Burnes
by Ward Transformer Company, the supplier
of the PCB oil. Ward Transformer Company
contracted with Burnes to dispose of the
PCB oil at $1.70 per gallon. However,
Burnes agreed to pay Ward 70c per gallon
as a cash kick-back, making the actual cost
of disposal $1.00 per gallon^
The expected costs from the illegal
disposal of PCB's include clean-up costs,
damages to health and the environment, and
penalties or fines imposed against the il-
legal disposer.
Clean-up costs for the contaminated
roadsides are expected to total at least
$1,580,000. These costs include removing
the contaminated soil from the road
shoulders, reshaping the shoulders, hauling
the contaminated material to the disposal
site and constructing a proper disposal
site.
The health and environmental damages
from the PCB dumping have not been quanti-
fied. Agricultural produce was destroyed
or not harvested as a result of the PCB
spill but the extent of the loss was not
monetized.
Charges have been brought against
Ward and Barnes, however, no fines have
been invoked. Burnes was sentenced to 3-5
years in prison, a trial on Federal charges
against Burnes and Ward is slated for
July 1981. The maximum allowable fine
under Section 10s of the Act is $10,000.
This figure has been used as the maximum
expected fine in this summary.
2. Direct Cost of Legal Disposal
The total cost of the legal disposal
of the PCB contaminated oil would have
ranged from $3.00 - $4.00 in 1978. These
396
-------
figures are based on the charge per gallon
for inceneration of hazardous chemicals at
an approved facility owned by Rollins
Environmental Services in Texas. The total
cost of legal disposal may be derived from
combining the charge per gallon by the
volume of PCB contaminated oil that was
discharged. At $3.00 per gallon the total
charge for disposal would have been $90,
000 and at $4.00 per gallon the cost of
proper disposal would have been $120,000.
D. Dover, N.Y.
During an eight month period in 1971
a scavenger trucker under contract to
Union Carbide Corproation disposed of be-
tween 5,000 and 6,000 barrels of chemical
wastes on a former chicken farm in Dover
Township, New Jersey. The wastes included
aromatic hydrocarbons, benzene, tobrene,
styrene, xylene, ketones, alcohols, and
phenolic resins.
1. Direct and Expected Costs of
Illegal Disposal
The direct losts of illegal disposal
of hazardous wastes in Dover, New Jersey
total $20,000. Union Carbide paid the
scavenger trucker $3.50 per drum to haul
away and dispose of approximately 5,500
drums of waste.
The expected costs of the illegal
disposal of waste include clean-up costs,
damages to health and the environment,
and fines and penalties levied against the
illegal disposer.
Clean-up operations on the site are
expected to total $49,350. Dover Twonship
has already spent an estimated $10,000 to
inspect and supervise the removal of waste
dumped at the Reich farm. Union Carbide
has spent approximately $15,750 to remove
4,5000 drums from the site.
The health and environmental damages
have not been quantified for the illegal
disposal of wastes on the site. However,
Union Carbide has given the owners of the
property $10,000 and reimbursement for
the cost of drilling a new well as compen-
sation for damage done to their property
as a result of dumping the waste. In ad-
dition 140 local residents have received
$140,000 in total from Union Carbide as
compensation for the loss of their wells.
Both instances of compensation were initi-
ated as a result of groundwater contamin-
ation from the illegal dump site. The
State of New Jersey is also expected to
receive up to $60,000 from Union Carbide
in compensation for costs connected with
any groundwater contamination that may
occur after April of 1977.
The maximum fine or penalty allowed
under Seciton 102 of Superfund is $10,000.
This figure has been quantified as the
maximum expected cost from fines or penal-
ties in the illegal disposal of wastes on
the site.
2. Direct Cost of Legal Disposal
The direct total cost of the legal
disposal of the wastes would have ranged
from $21,175 to $30,250. These figures
are derived from the cost per drum of
incineration of chemical wastes at a legal
facility in New Jersey. Incineration
would have cost $3.85 to $5.50 per drum,
totalling $21,175 and $30,250 respectively
for the 5,500 drums.
Cost Differentials with Varying Probability
The costs in Table I are transformed
into cost differentials with varying pro-
bability of the disposer being held res-
ponsible for the release in Table II.
The numbers with negative signs indicate
it is still cheaper to comply than not to.
Where the positive numbers indicate the
cost of compliance has exceeded the cost
of noncompliance, we can expect the incen
tive to dump illegally to be strong. In
the case of the Valley of Drums in Bullit
County, Kentucky, cost of noncompliance
does not exceed that of compliance until
the probability becomes.1. The same dis-
poser would probably illegally dump again
if the probability of detection is equal
to .1 or less. The case of Chester, N.C.
is clearly an example where costly regu-
lation can prompt illegal dumping. Even
at the certainty of 1.0 it is probably
cheaper not to comply. On the other hand,
both the case of North Carolina and the
one of Dover, N.J. show tendency to comply
even for small probability of being detec-
ted. The reason for the illegal dumping
must have been either not knowing the con-
sequence of the dumping or simply believing
that one would not be caught.
The above results are subjected to a
number of qualifications. First, as we
397
-------
have noted, health and environmental damages
are not incorporated into the numbers pre-
sented in Table II. Of course, if these
damages can be recovered, there will be an
increase in incentive to comply. On the
other hand, when the damages are in the
area of millions of dollars, it is likely
that the disposers do not have the financial
profile of the disposers.
Summary
The qeustion of how to minimize illegal
dumping or maximize compliance in the hazar-
dous waste regulations is not only complex
on the theoretical level, but also a dif-
ficult one to address on the empirical
level. Although, economic theory can ex-
plain the mechanics of decision-making in
compliance or noncompliance on a firm level,
the extension of the theory to the compli-
ance question on an industry level has been
only occurred recently. Although the ap-
plication of an industry compliance model
still faces serious difficulties due to
lack of relevant data, the recognition of
the concept that overly burdensome regula-
tions can be counterproductive is a note-
worthy step foreward. (In time, relevant
information may accumulate to produce
adequate data base for the model.) Never-
theless, even if the information existed,
it should be clear that the marginal bene-
fit of regulation is often defined in the
political arena; any result from a compli-
ance model is based on a certain set of
assumptions. Regulatory analysis should
make them explicit as does the weighting
scheme discussed in this paper. Such
efforts should promote the regulatory
efficiency much needed as our private in-
dustry is asked to bear more regulatory
constraints.
Footnotes and References
1. Freeman, A. Myrick, Haveman, H. Robert
and Kneese, V. Allen, The Economics of
Environmental Policy, John Wiley & Sons,
Inc. New York, 1973, p. 33.
2. For the difficulty in identifying social
preference see Arrow, J. Kenneth "A Dif-
fculty in the Concept of Social Welfare,"
Readings in Welfare Economis, Arrow,J.
Kenneth and Scitousky, Tibor (Ed).
Richard Iwrin, Inc., Illinois, 1969.
Research in theory of economic sur-
plus has dealt with the difficulty of
calculating loss of social welfare due
the misallocation of resources. See
Currie, M. John; Murphy, A. John and
Schmitz, Andrew "The Concept of Eco^
nomic Analysis," The Economic Journal
No. 324, Vol. LXXXI, Dec. 1971.
3. W. Viscusi and R. Zechauser, "Opti-
mal Standards with Incomplete Enfor-
cement," Public Policy, 27:437 (1979)
4. Illegal dumpins in this section includes
any hazardous waste disposal, treatment
and storage practice that causes impro-
per release damaging to health and
environment.
5. The distinction between environmental
compliance model and environmental
management model is that the former
model operates for any given level of
environmental quality. Unlike the
latter, it does not determine the
socially optimal level of envrionmental
quality. In other words, once the
optimal level of environmental quality
is determined through some process, be
it political, economic or legislative,
the compliance model decides the most
efficient level of standard to achieve
this specified level.
6. For example, see Becker, Gary, "Crime
and Punishment: An Economic Approach,"
in G. Becker and W. Landes (EDs.) Es-
say se in the Economics of Crime and
Punishement, Columbia University Press,
New York, 1974, pp. 1-54.
7. We have suppressed the footnotes iden-
tifying the source of the cost estima-
tion due to their number.
398
-------
TABLE II
Differentials in Cost Between Legal and
Illegal Disposal of Hazardous Wastes-"-
Bullit County
P. 2
1.0
.5
.3
.1
CF3
-2,860,000
-1,232,000
- 603,000
26,220
Chester
P
1.0
.5
.3
.1
CF
177,500
963,750
1,273,250
1,592,750
North Carolina
P
1.0
.5
.3
.1
CF
-1,515,000
- 720,000
- 402,000
84,000
Dover
P
1.0
.5
.3
.1
CF
1
-53,600
-24,000
-12,000
200
•k^ost differentials are calculated from Table I.
Probability of detection and conviction
3Cost Differential
When there is a range, the mean is used.
Source: ELI Case Studies
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing]
REPORT NO.
EPA-600/9-81-002b
I. RECIPIENT'S ACCESSIOf*NO.
TITLE AND SUBTITLE
5. REPORT DATE
Land Disposal: Hazardous Waste
Proceedings of the Seventh Annual Research Symposium
March 1981
6. PERFORMING ORGANIZATION CODE
AUTHOR(S)
Edited by David Shultz
Coordinated by David Black
8. PERFORMING ORGANIZATION REPORT NO.
PERFORMING ORGANIZATION NAME AND ADDRESS
Southwest Research Institute
P. 0. Drawer 28510
San Antonio, TX 78284
10. PROGRAM ELEMENT NO.
BRD1A DU109
11. CONTRACT/GRANT NO.
68-03-2962
12. SPONSORING AGENCY NAME AND ADDRESS
13. TYPE OF REPORT AND PERIOD COVERED
Municipal Environmental Research Laboratory--Cin.,OH
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
Final - 9/20-3/81
14. SPONSORING AGENCY CODE
EPA/600/14
15. SUPPLEMENTARY NOTES
Project Officer Robert E. Landreth, 684-7871
16. ABSTRACT
The Seventh Annual SHWRD Research Symposium on land disposal of municipal
solid waste and industrial solid waste and resource recovery of municipal solid
waste was held in Philadelphia, Pennsylvania, on March 16, 17, and 18, 1981. The
Durposs of the symposium were (1) to provide a forum for a state-of-the-art review
and discussion of on-going and recently completed research projects dealing with
the management of solid and industrial wastes; (2) to bring together people concerned
with municipal solid waste management who can benefit from an exchange of ideas
and information; and (3) to provide an arena for the peer review of SHWRD's overall
research program. These proceedings are a compilation of papers presented by the
symposium speakers. The technical areas covered in the Land Disposal: Municipal
Solid Waste are gas and leachate production, treatment and control technologies
and economics. The areas covered in Land Disposal: Hazardous Wastes are hazardous
waste characterization, transport and fate of pollutants, hazardous waste containment,
land treatment of hazardous wastes, hazardous waste treatment, uncontrolled sites/remed
action, and economics. Municipal Solid Waste: Resource Recovery include the
areas of equipment and processing, recovery and use of materials, environmental
aspects and economics/impediments and special studies.
al
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS
COS AT I Field/Group
Leaching Collection, Hazardous Materials,
Disposal, Treatment, Soils, Groundwater
Pollution, Waste Treatment, Methane, Gases,
Linings
Solid Waste Management
Sanitary Landfills
Hazardous Waste
Leachate
13B
13. DISTRIBUTION STATEMENT
Release to public
19. SECURITY CLASS (ThisReport)
Unclassified
21. NO. OF PAGES
418
20. SECURITY CLASS (Thispage)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
400
i US GOVERNMENT PRINTING OFFICE 1961-757-064/0244
------- |