c/EPA
United States
Environmental Protection
Agency
Office of Research and
Development
Washington, DC 20460
EPA/600/R-93/033
March 1993
Environmental Monitoring
Issues
Results of Workshops
Held in July 1992 as
Part of EPA's Eighth
Annual Waste Testing and
Quality Assurance Symposium
-------
EPA/600/R-93/033
March 1993
ENVIRONMENTAL MONITORING ISSUES:
RESULTS OF WORKSHOPS HELD IN JULY, 1992 AS PART
OF EPA'S EIGHTH ANNUAL WASTE TESTING
AND QUALITY ASSURANCE SYMPOSIUM
Office of Modeling, Monitoring Systems and Quality Assurance
Office of Research and Development
U.S. Environmental Protection Agency
Washington, D.C. 20460
Printed on Recycled Paper
-------
DISCLAIMER
The workshop was funded in part by the United States Environmental Protection Agency
(EPA) under a Cooperative Agreement with the American Chemical Society (ACS) and the
report was prepared with contract support from SAIC under contract number 68-W2-0027 and
The Cadmus Group under contract number 68-W2-0004.
The document has been subjected to administrative review and has been approved for
publication as an EPA document. It presents the viewpoints and ideas expressed by the
participants at a public meeting and therefore should not be construed as necessarily representing
the viewpoint of the Agency. Mention of trade names or commercial products should not be
considered to be an endorsement or recommendation for use.
11
-------
CONTENTS
Page No.
Foreward v
Acknowledgements vi
Abbreviations vii
1. INTRODUCTION 1
2. PREDICTING THE ENVIRONMENTAL IMPACT OF OILY MATERIALS 2
2.1. BACKGROUND 2
2.2. SUMMARY OF POSITION STATEMENTS 3
2.2.1. Regulatory Program Perspective 3
2.2.2. Regulated Community Perspective 5
2.2.3. State-of-Science Perspective 7
2.3. CHARGE TO THE WORKSHOP PARTICIPANTS 8
2.4. RESULTS OF WORKGROUP DELIBERATIONS 8
2.4.1 Definition of Oily Waste 8
2.4.2 Analytes of Concern 9
2.4.3. What Approach Should Be Used in Developing a New Testing
Procedure? 9
2.4.3.1. New Laboratory Leach Testing Methods 11
2.4.3.2. Mathematical Models to Estimate Mobility 12
2.4.3.3. If the Necessary Tools Are Not Presently Available, How
Should Such a Test Method Or Model Be Developed And
Evaluated? 12
2.5. AGENCY FOLLOWUP TO WORKSHOP . 13
3. ISSUES ASSOCIATED WITH ADOPTION OF PERFORMANCE-BASED
METHODS 14
3.1. BACKGROUND 14
3.2. SUMMARY OF POSITION STATEMENTS 15
3.2.1. Regulatory Program Perspective 15
3.2.2. Perspective of the In-House Laboratory from the Regulated
Community 17
3.2.3. Perspective of the Commercial Laboratory Industry 19
3.2.4. State-of-Science-Perspective 21
111
-------
CONTENTS
Page No.
3.3. CHARGE TO THE WORKSHOP PARTICIPANTS 22
3.4. RESULTS OF WORKGROUP DELIBERATIONS 22
3.4.1. How Should Acceptable Performance Be Defined? 22
3.4.2. What Documentation Should EPA Require to Verify Compliance
with Performance Standards? 25
3.4.3. How Will the Change to Performance Standards Affect the
Procurement Process by Both EPA and Commercial
Laboratories? 26
3.5. AGENCY FOLLOWUP TO WORKSHOP 26
4. CHARACTERIZING MIXED WASTES 27
4.1. BACKGROUND 27
4.2. SUMMARY OF POSITION STATEMENTS 28
4.2.1. Regulatory Program Perspective 28
4.2.2. Regulated Community Perspective 29
4.2.3. State-of-Science Perspective 31
4.3. CHARGE TO THE WORKSHOP PARTICIPANTS 33
4.4. RESULTS OF WORKGROUP DELIBERATIONS 33
4.4.1. Entombment 34
4.4.2. Incineration 34
4.4.3. Solidification 35
4.4.4. Other Issues 35
4.4.5. Other Waste Management Options 36
4.5. CONCLUSIONS 37
4.6. AGENCY FOLLOWUP TO WORKSHOP 37
5. CHARACTERIZING HETEROGENEOUS MATERIALS 38
5.1. BACKGROUND 38
5.2. SUMMARY OF POSITION STATEMENTS 39
5.2.1. Regulatory Program Perspective 39
5.2.2. Regulated Community Perspective 41
5.2.3. State-of-Science Perspective 42
5.3. CHARGE TO THE WORKSHOP PARTICIPANTS 42
5.4. RESULTS OF THE WORK GROUP DELIBERATIONS 43
5.4.1. Should the Agency Change to Attribute Testing for Properties that
are Not Averageable? 43
5.4.2. What is the Highest Practicable Degree of Confidence that can be
Required? 44
5.5. AGENCY FOLLOW TO WORKSHOP 44
IV
-------
CONTENTS
Page No.
APPENDIX A: ISSUES ASSOCIATED WITH ADOPTION OF
PERFORMANCE-BASED METHODS A-l
David Friedman INTRODUCTION A-l
James M. Conlon and
William Telliard PERFORMANCE-BASED METHODS-
one view from the regulators A-4
Reid Tait INDUSTRY PERSPECTIVE A-15
Carla Dempsey INDUSTRY PERSPECTIVE A-25
Billy J. Fairless, Ph.D. SCIENTIFIC PERSPECTIVE A-39
APPENDK B: PREDICTING THE ENVIRONMENTAL IMPACT OF
OILY MATERIALS B-l
David Friedman INTRODUCTION AND REGULATORY
PERSPECTIVE B-l
Clifford T. Narquis INDUSTRY PERSPECTIVE B-7
Larry P. Jackson SCIENTIFIC PERSPECTIVE B-24
APPENDK C: CHARACTERIZING HETEROGENEOUS MATERIALS C-l
David Friedman INTRODUCTION C-l
Charles A. Ramsey REGULATORY PERSPECTIVE C-4
David Reese INDUSTRY PERSPECTIVE C-12
John P. Maney, Ph.D. SCIENTIFIC PERSPECTIVE C-16
APPENDK D: CHARACTERIZING MIXED WASTES D-l
Dr. James A. Poppiti INTRODUCTION D-l
Susan Jones REGULATORY PERSPECTIVE D-5
C.S. Leasure INDUSTRY PERSPECTIVE D-ll
W.H. Griest and
J.R. Stokely, Jr. SCIENTIFIC PERSPECTIVE D-20
APPENDIX E: LEACHABnjTY PHENOMENA: Recommendations and Rationale
for Analysis of Contaminant Release by the Environmental
Engineering Committee E-l
-------
FOREWORD
The U.S. Environmental Protection Agency (EPA) has several goals in sponsoring the
Waste Testing and Quality Assurance Symposia series. These include serving as a forum for
working with the public and other regulatory officials on measurement, monitoring, and quality
assurance issues involved with implementing the hazardous waste management program, and
exchanging information on new developments in waste and environmental media sampling and
testing methods.
As a means of soliciting input from all sectors of the monitoring community, the Agency
in 1992 added a series of workshops on monitoring issues important to EPA's regulatory
programs. The workshops dealt with four issues:
• Adoption of performance-based methods,
• Predicting the environmental impact of oily materials,
• Characterizing heterogeneous waste, and
• Characterizing mixed wastes.
The Symposium series is cosponsored by EPA's Office of Solid Waste and Emergency
Response (OSWER) and Office of Modeling, Monitoring Systems and Quality Assurance
(OMMSQA) within the Office of Research and Development (ORD). Due to the importance of
issues discussed in the workshops, the Agency is continuing its efforts to address these issues.
This document informs persons who were not able to participate in the workshops of the
resulting ideas and information, solicits additional information on possible approaches to address
the issues, and review activities the Agency has undertaken to followup on the workshops.
Comments on the issues should be sent to the author at the following address:
David Friedman
Office of Research and Development (RD-680)
U.S. Environmental Protection Agency
Washington, DC 20460
VI
-------
ACKNOWLEDGEMENTS
This report presents the results of a series of workshops held as part of EPA's Eight
Annual Testing and Quality Assurance Symposium. The success of the workshops was due to
the efforts of a number of individuals and organizations.
The Agency wants to thank the speakers who catalyzed the discussions by presenting
ideas of options for the participants to consider:
Michael Conlon, U.S. Environmental Protection Agency
Carla Dempsey, Lockheed Engineering and Sciences Co.
Billy Fairless, U.S. Environmental Protection Agency
Wayne Griest, Oak Ridge National Laboratory
Larry Jackson, Consultant
Susan Jones, U.S. Environmental Protection Agency
Gene Klesta, Chemical Waste Management
Craig Leasure, Los Alamos National Laboratory
John Maney, Consultant
Clifford Narquist, BP Research
James Poppiti, Department of Energy
Charles Ramsey, U.S. Environmental Protection Agency
David Reese, Safety-Kleen Corp.
James Stokely, Oak Ridge National Laboratory
Reid Tail, Dow Chemical Co.
Special thanks are due to the staff of SAIC who served as workshop facilitators and
scribes; to Larry Jackson and James Poppiti who assisted in organizing the workshops and who
prepared the workshop summaries from which this report was prepared; and to the staffs of The
Cadmus Group and Versar, Inc. who supplied editorial and document composition assistance.
The author also wishes to thank Gail Hansen and Alec McBride of the Office of Solid
Waste (OSW) for their help and guidance in developing the workshops and in reviewing the
report.
vu
-------
ABBREVIATIONS
ASTM
API
ALARA
AEA
CERCLA
CWA
CAA
DQO
DOE
DOT
EC
EMMC
EMSL-LV
EP
FIFRA
ICP-MS
ISO
INEL
LDR
NRC
NEIC
OSWER
OSW
OMMSQA
ORD
OWEP
PCBs
QA
QC
QAPjP
RMW
RCRA
RSD
SDWA
SAB
TDS
TCLP
TSCA
U.S. EPA
WAP
American Society for Testing and Materials
American Petroleum Institute
As Low As Reasonably Achievable
Atomic Energy Act
The Comprehensive Environmental Response, Compensation,
and Liability Act
Clean Water Act
Clean Air Act
Data Quality Objective
Department of Energy
Department of Transportation
Environment Canada
Environmental Monitoring Management Council
Environmental Monitoring and Support Laboratory , Las Vegas
Extraction Procedure
Federal Insecticide, Fungicide, and Rodenticide Act
Inductively Coupled Plasma-Mass Spectrometer
International Organization for Standardization
Idaho National Energy Energy Laboratories
Land Disposal Restriction
Nuclear Regulatory Commission
National Enforcement Investigation Center
Office of Solid Waste and Emergency Response
Office of Solid Waste
Office of Modeling, Monitoring Systems and Quality Assurance
Office of Research and Development
Oily Waste Extraction Procedure
Polychlorinated Biphenyls
Quality Assurance
Quality Control
Quality Assurance Project Plan
Radioactive mixed waste
Resource Conservation and Recovery Act
Relative Standard Deviation
Safe Drinking Water Act
Science Advisory Board
Total Dissolved Solids
Toxicity Characteristic Leaching Procedure
Toxic Substances Control Act
U.S. Environmental Protection Agency
Waste Analysis Plan
vui
-------
1. INTRODUCTION
Each issue workshop consisted of two parts: a plenary session and a working session.
During the plenary session, participants heard viewpoints on each issue from:
• A representative of EPA (the regulatory perspective),
• One or more representatives of the regulated industry or the laboratory services
industry (the regulated community's perspective), and
• A scientist active in the field to review the state-of-science.
Appendix A contains the text of all presentations.
Following the plenary session, the moderator presented the participants with their charge,
and the approximately 200 participants broke into three working groups to address each problem
presented. Participants were assigned to different breakout sessions to ensure that all sectors of
the monitoring community were represented in each workgroup.
This report highlights the principal points presented by the speakers during the plenary
sessions and the results of the working groups. The inputs from each working group on a single
issue have been combined to provide a consolidated picture of the information, ideas, and
suggestions raised by the participants.
Each workshop provided a forum for members of the environmental monitoring
community from industrial, State, Federal, and private sector groups to share ideas and
information.
-------
2. PREDICTING THE ENVIRONMENTAL IMPACT OF OILY MATERIALS
2.1. BACKGROUND
Prevention of ground water contamination historically has been one of EPA's highest
priorities in implementing the Resource Conservation and Recovery Act (RCRA) hazardous
waste management program. To that end, the Agency has developed and promulgated fate and
transport models, test methods, and regulatory standards to control the management of wastes
whose properties might pose a hazard to ground water. The Agency has further demonstrated
its concern for ground water by including the Toxicity Characteristic (40 CFR 261.24) among
the hazardous waste identification characteristics.
One type of waste that can pose a hazard to ground water is oily waste, which can, in
some cases, migrate like a liquid even though it appears to be a solid. Oily wastes take many
forms, including liquids of widely varying viscosity, contaminated soils, sludges, and tarry
"plastic" masses. Because oily wastes result from a wide variety of commercial processes and
applications, these wastes are broadly distributed and have a large volume.
Over the years, a number of laboratory extraction methods have been applied to the
problem of predicting what might migrate from oily wastes managed under landfill conditions.
Among the test methods that have been developed and employed to identify those wastes that
might pose an unacceptable hazard are: EPA methods 1310 [Extraction Procedure, (EP)], 1311
[Toxicity Characteristic Leaching Procedure, (TCLP)], and 1330 [Oily Waste Extraction
Procedure, (OWEP)].
All current approaches have serious deficiencies in predicting the mobility of toxic
chemicals from oily wastes. Method 1311 (TCLP) underestimates the mobility of many oily
wastes due to filter clogging problems, its degree of precision, and many operational problems.
Conversely, Method 1330 (OWEP) probably overestimates mobility, since this method emulates
a worst possible case scenario. No available laboratory mobility procedure is truly satisfactory
-------
as each method suffers from one or more problems. The goal of the workshop was to advance
the state-of-the-art with respect to techniques for characterizing the leaching potential of oily
materials.
Because mobility estimation is such a large and complicated topic, the workshop focused
on only one aspect: predicting which substances might leach from oily material that have been
improperly placed in a landfill. Participants discussed how to predict either the first or the
highest concentration of material that would be released overtime from such waste to the soil
immediately below it (the initial source term). This information is used in the fate and transport
models that predict the final toxicant concentration at a specified distance from the disposal site.
2.2. SUMMARY OF POSITION STATEMENTS
2.2.1. Regulatory Program Perspective
Mr. David Friedman, ORE), presented the Agency's position that, "Given the importance
of this issue, it is imperative that accurate, precise, and usable approaches to characterizing the
mobility of oily materials be developed." Predicting a material's leachability requires
consideration of the disposal conditions and the fate and transport models for which the
characterization is to serve as the release term. No universal test procedure is likely to be
developed that will accurately replicate all disposal conditions. There are many disposal
scenarios that are of concern. However, for the purposes of this effort, the priority waste
management scenario that EPA's Office of Solid Waste (OSW) selected to be modeled in this
workshop is placement of the waste into or on the ground (e.g., in an unlined pit, lagoon). This
scenario requires considering a number of parameters, including:
• Temperature of the waste,
• Rainfall regime,
• Amount of waste,
-------
• Leachate-waste contact time, and
• Soil types.
The challenge lies in predicting which constituents, at which concentrations, will leave
the disposal point and move toward an aquifer. Pollutant migration will result from a
combination of movement of the "liquid" phase of the oily waste and from extraction of
toxicants by the action of rainwater or landfill leachate. The information obtained from the
characterization serves as the input to the fate and transport models being developed as part of
other Agency initiatives.
As input parameters, the model requires information on the amount and composition of
both the aqueous and non-aqueous phase liquid portions of the waste and the composition of the
leachate that might be generated by action of surface waters on any "solid" material present in
the waste material. At this time, the Agency does not have a precise way to define either
"aqueous-phase liquid" or "nonaqueous-phase liquid."
As a practical matter, it is important that the laboratory test methods selected to predict
the environmental impact of oily wastes:
• Be simple to use (not require sophisticated equipment or constant attention by a
technician),
• Be inexpensive to perform,
• Take as little time as possible to perform (ideally no more than 24 hours),
• Give consistent results (i.e., RSD < 50%),
• Be rugged, and
• Not generate wastes (e.g., generate little, if any, solvent waste and waste
contaminated media).
-------
From a regulatory perspective, it is important that the tests:
• Be accurate (relative to predicting behavior of waste in the environment) and
• Have a high degree of freedom from false negatives (any errors that
underestimate the threat to the environment).
The Agency has identified three approaches that may provide the necessary estimate of
possible environmental impact of oily wastes under the target disposal scenario:
• One approach defines as mobile all organic materials (e.g., 50+ percent of
solvent extractable material is carbon) that have a defined degree of flowability
under conditions of moderate temperature (e.g., 20°C) and pressure (e.g., 10
psi).
• A second approach develops a single generic laboratory procedure to:
a. separate the material into that portion of the material that can be expected to
flow away from the disposal point (i.e., would behave as a "solid"), and
b. estimate the concentration in ground water leaving the disposal point that
might result from ground water solubilizing the hazardous constituents from both
the mobile and nonmobile fraction of the material.
• A third approach relies on a series of laboratory test procedures to evaluate the
waste material. These procedures would be keyed to the waste properties and
could be representative of site specific conditions of the disposal environment.
2.2.2. Regulated Community Perspective
Mr. Clifford Narquist, of BP Research who represented the American Petroleum Institute
(API), presented a regulated industry viewpoint.
Many groups, including Environment Canada (EC), American Society for Testing and
Materials (ASTM), and EPA, have participated in developing improved leachability and
contaminant fate/transport tests and models. Despite this work, predicting the potential
-------
environmental hazard associated with oily wastes remains problematic. TCLP makes the
assumption that aqueous and nonaqueous liquids will behave identically, both within the landfill
and upon their hypothetical release. Little technical support exists for this position.
The risk from light nonaqueous phase liquids (i.e., oils) appears to result mainly from
contaminants partitioning from the oil into the ground water. Free oil, itself, poses little risk
because of a lack of exposure. As oil moves through the soil, the soil immobilizes it. Toxic
constituents partition into the water phase from this point of immobilization. Because people
are highly unlikely to drink free-phase hydrocarbons, leaching the oil fraction along with the
solids appears a sensible approach to treating this fraction.
The disposal scenario as depicted by EPA does not reflect current waste disposal
management practices. Specifically, an EPA-OSW survey documented that municipal landfills
no longer accept liquid-type wastes. New test methods, therefore, should be based on actual
waste management scenarios.
With respect to the fate and transport model, the infinite source assumptions currently
used are unrealistic. One improvement would be to design transient, declining source terms into
the model. Further, the leaching test and model need to consider the processes that transpire
in the vadose zone. Among these processes are biodegradation, hydrolysis, volatilization, and
soil retardation. These processes are not considered in the current regulatory approach.
Until work on the behavior of nonaqueous materials and the prediction of their movement
is more developed, the nonaqueous liquids should be treated like the solid phase of the waste and
should be subjected to the same extraction fluid. To the extent that the hazardous constituents
are released into the extractant, they should be combined with the aqueous extract generated
from the waste solids.
Before developing new test, a consistent and easily applied physical, hydrologic, and
geochemical representation should be developed for the waste management scenario of concern.
-------
The development of leach tests and fate and transport models should be coordinated. These new
procedures should follow the recommendations of EPA's Science Advisory Board (SAB)
Environmental Engineering Committee, Leaching Subcommittee. This approach would place
the new tests and models on a firmer scientific basis and lead to better decisions.
2.2.3. State-of-Science Perspective
The state-of-the-science in predicting the environmental impact of oily materials was
presented by Mr. Larry Jackson, a private consultant. As with the API presentation,
considerable importance was placed on using the rigorous scientific approach of the SAB
Leaching Subcommittee.
The presentation proposed a series of evolutionary changes to the current approach based
on existing EPA methods, experience gained from using the TCLP, and a reconsideration of
some existing models and data on the fate and transport of fuel-derived materials. The
presentation also offered four suggestions for improving the current approach:
Adopt a flowable materials test to separate free liquid phase hydrocarbons from
the contaminated solid matrix. Such a test would permit scientists to
independently evaluate the potential impact of the free-phase material and the
contaminated soil independently. A test similar to EPA's proposed liquid release
test was described.
Modify the TCLP using approaches already allowed in other RCRA methods.
These modifications could overcome some mechanical problems associated with
the current procedure.
Adopt a new method of contacting the leach medium with the waste.
Descriptions of several column leach testing procedures that are consistent with
the SAB recommendations were provided. This approach offers options that
allow the evaluation of both lighter-than-water and heavier-than-water materials.
Use a new model to determine the release potential of oily wastes. The proposed
model uses simple well-founded tests and basic physical/chemical parameters of
the waste and potential pollutants. With this model, the effect of soils on the fate
and transport behavior of individual compounds can be assessed.
-------
2.3. CHARGE TO THE WORKSHOP PARTICIPANTS
Before breaking into three workgroups, the participants were asked to address the
following questions:
1. What is the "best" approach to use to predict the nature and concentration of the
components that will leach from oily wastes placed in an unlined landfill
environment?
2. If the necessary tools are not available, how should such a test method or model
be developed and evaluated?
3. What form should a cooperative development program take?
• How should it be organized?
• Who might the cooperators be?
• How long would it require?
2.4. RESULTS OF WORKGROUP DELIBERATIONS
During their deliberations, the workgroup participants dealt with a number of aspects of
the problem including the definition of oily waste and analytes of concern. A summary of the
discussions follows.
2.4.1 Definition of Oily Waste
Two workgroups felt strongly that the Agency should revisit how "oily waste" is defined
and which constituents of an oily waste should be of concern.
Participants discussed two approaches to defining the universe of wastes that should be
considered oily wastes. The first classifies wastes based on physical properties such as viscosity
8
-------
and density. The second approach defines oily wastes by listing wastes or residues resulting
from a given industry or waste generation process. The workshop participants believed that
developing appropriate testing schemes requires more information about oily wastes than is
currently available. Specifically, participants expressed a need for EPA to conduct a survey that
will better define the sources of oily wastes, the volumes generated, and the current
disposal/treatment options.
2.4.2 Analytes of Concern
Participants voiced concern that whatever future approach the Agency uses to characterize
oily wastes that approach should consider more of the compounds listed in Appendix VIII than
are currently included in the Toxicity Characteristic. This concern was based partly on the need
to protect human health and the environment. In addition, participants stated that the list of
regulated compounds will probably grow. This growth in turn will impact the design and
conduct of any test. Some participants felt that, unless all analytes likely to be of concern were
considered at the beginning, any test developed may require later modification. People objected
to spending time developing tests that would rapidly become outmoded.
2.4.3. What Approach Should Be Used in Developing a New Testing Procedure?
This was the central question addressed by the groups, and consideration of this issue
occupied most of the discussion time. The participants identified two basic criteria to apply to
any new test method development effort:
1. The test must address the procedural problems encountered when applying the
TCLP to oily wastes (i.e., time consuming nature of the TCLP, its high cost, its
high variability, and its inability to handle materials that are difficult to filter).
2. The new test(s) should be scientifically valid. Where possible, tests should be
capable of representing the actual disposal conditions (soil types, rainfall, etc.)
-------
and must simulate the behavior of oily materials in the environment. New tests
should consider such properties of the oily waste as amount of flowable material,
density of the liquid phase, viscosity, and the material's flow characteristics in a
soil environment. Special efforts should be made to validate these models with
data from real-world samples disposed in actual landfills.
Members from discussion groups suggested that developing a tiered testing approach
would reduce the number of samples requiring leach testing. Specifically, an initial aggressive
test, such as compositional analysis of the total waste, would be used to screen the waste.
Wastes containing high levels of regulated constituents would be considered as hazardous without
mobility testing. Wastes with low levels of contaminants would be considered as nonhazardous.
The time consuming, mobility testing procedures would then be applied only to those wastes
containing moderate levels of constituents of concern. EPA would establish criteria for each
parameter that would initiate the complete leach testing procedure.
Another variation of this approach would employ a quick, relatively inexpensive
screening test (e.g., some type of leaching test) to identify wastes that have a high enough
potential of posing a leaching hazard to warrant definitive testing for leachability. As before,
EPA would have to establish new criteria for each parameter that would initiate further leach
testing.
The general consensus favored abandoning the TCLP as a means of characterizing oily
wastes. Only one participant wanted to retain TCLP as the principal predictive tool. Attempting
to modify the TCLP to improve its performance received support from only two participants.
The rest preferred a new approach. Participants discussed two general approaches for
predicting leachability: use of laboratory tests and use of mathematical models.
10
-------
2.4.3.1. New Laboratory Leach Testing Methods
Participants considered two approaches to new laboratory methods for evaluating
leachability: a multitest approach and column leach testing.
• Multi-Test Leaching Approach
A number of participants supported having a suite of leaching tests to evaluate the
environmental impact of oily materials. Given the diverse universe of oily
wastes, a single "best" test may not be feasible or scientifically valid. The use
of multiple tests will allow analysts to match the characteristics of the waste
and/or the management unit to a test designed for the specific situation.
In a related suggestion, some participants proposed a generic testing protocol with a range
of options for key testing variables. This approach would allow the selection of the most
appropriate test conditions for a particular waste. Variables could include the nature and amount
of leaching fluid, particle size of the waste, time, impact of site-specific soils, etc. The concept
would apply equally well to either batch- or column-type tests.
• Column Leach Testing
Several people supported the development of a column leaching technique as the
best method to simulate the movement of oily material in the environment. The
addition of inert materials to the column would immobilize liquid, provide
permeability, and allow efficient extraction of the waste material. The column
technique also permits determination of the effect of site-specific soils on the
waste and its leachate.
In developing new mobility procedures, the workgroup supported validating these
procedures against actual field conditions. With column tests, it will also be important to study
such factors as particle size and moisture content of the column materials, flow direction and
rate, viscosity of materials, etc.
11
-------
2.4.3.2. Mathematical Models to Estimate Mobility
Several participants suggested mathematical models to predict toxicant mobility, because
such models can describe the behavior of possible pollutants both in the waste matrix and the
disposal environment. The literature contains empirical models for predicting equilibrium
partitioning of compounds between oily materials and between aqueous liquids and soils. The
key variables in the models are the water solubility of the constituents of concern, the percent
carbon in the waste matrix and surrounding soil, and the various partition coefficients.
A modeling approach holds the promise as a means of reducing both the cost and time
needed for characterization. In addition, mathematical models can provide site-specific
evaluations of environmental impact.
2.4.3.3. If the Necessary Tools Are Not Presently Available, How Should Such a Test
Method Or Model Be Developed And Evaluated?
A thorough review of existing data from proposed modeling approaches would be useful
for directing future efforts. Collecting data from actual field sites would also aid in developing
and validating methods.
Participants expressed the opinion that the Agency should work with members of the
private sector and universities in developing any new methods. EPA should also involve
members of other Agency programs such as the Toxic Substances Control Act (TSCA), the
Federal Insecticide, Fungicide, and Rodenticide Act (FIFRA), and the Comprehensive
Environmental Response, Compensation and Liability Act (CERCLA) programs in the
development effort.
Participants recommended additional technology transfer activities similar to this
workshop. Suggested activities included symposia and seminars. Possible co-sponsors include
ASTM and API. International participation would also be desirable.
12
-------
The Agency should pursue an evolutionary approach that would bring about incremental
modifications in available methodology. Such an approach would lead to improved methods
within only 1 to 2 years. However, participants recognized that fully understanding how to
predict the mobility of oily material may require up to 10 years of research and development.
2.5. AGENCY FOLLOWUP TO WORKSHOP
Since the workshop, EPA has taken a number of steps to address this problem of
predicting the environmental impact of oily materials.
Taking to heart the suggestions that the Agency work with the private sector, EPA has
entered into a cooperative agreement with the ASTM Institute for Standards Research. This
agreement calls for working with members of the public and private sectors to develop improved
protocols for evaluating the mobility of oily materials.
In October SAB met to review the results of the workshop and the Agency's plans for
moving forward with the development of improved protocols for oily waste. Appendix A to this
report describes the approach under consideration.
As a first step, the Agency plans to use soil column studies to identify those oily wastes
for which the TCLP is inappropriate. At the same time, the Agency is determining which
management scenarios and subsurface fate and transport models are appropriate for use in
assessing the impact of oily waste on ground water. Based on the results of these studies, the
Agency will identify the input needs of the models for which the mobility procedures are
designed to serve.
13
-------
3. ISSUES ASSOCIATED WITH ADOPTION OF PERFORMANCE-BASED
METHODS
3.1. BACKGROUND
A number of regulatory programs are implemented by EPA that require collection of
environmental monitoring data. These include the Safe Drinking Water Act (SDWA); the Clean
Water Act (CWA); RCRA; CERCLA; the Clean Air Act (CAA); and FEFRA.
Selecting the appropriate measurement method is a key part of any monitoring program.
Over the years, in crafting its monitoring requirements, the Agency has primarily adopted a
policy of publishing test methods appropriate for complying with the various regulations. The
methods have been written with varying levels of detail. Rather than issuing flexible methods
and specifying required performance, the Agency generally has opted to give the analyst
mandatory instructions on how to conduct the analysis.
This approach has resulted in several problems:
1. Long delays have occurred in using new technology to cost-effectively
analyze samples.
2. In some cases, the required methods either give biased results for some
types of samples or do not yield the requisite sensitivity. Since in many
cases the methods make no provision for modification by the analyst, the
analysis can lead to results that do not meet the data quality needs.
3. Rigid specification of methods rather than data quality, frequently forces
laboratories to use expensive technology in place of suitable approaches
that are simpler and less costly.
14
-------
The Agency is aware of these problems and is attempting to solve them. One option
entails moving from a system that specifies methods to one that specifies performance and allows
flexibility in the methods used to conduct the analysis. Before implementing such an approach,
the Agency needs to resolve several questions. These include:
• How will acceptable performance be defined?
• What documentation is needed to verify compliance with performance standards?
• How will the change to performance based methods affect the procurement
process, both by EPA and by commercial clients?
The Agency sponsored this workshop to examine this issue and to bring together
members of the environmental monitoring community to share ideas and information on ways
to increase the utilization of innovative monitoring technologies by switching to performance-
based monitoring. This change has the potential to improve both the quality and cost-
effectiveness of environmental monitoring. Although many people have ideas on how to
implement performance-based monitoring, as yet, there is no agreement on which approach is
best.
3.2. SUMMARY OF POSITION STATEMENTS
3.2.1. Regulatory Program Perspective
Mr. Michael Conlon, EPA Office of Ground water and Drinking Water, presented his
views on the merits of adopting performance-based analytical methods.
Regulatory data needs to provide accurate information for decision making rather than
determining the absolute truth. This is especially true when pursuing absolute truth results in
expenditures that bring about little or no additional protection of public health. Limited spending
on regulatory data does not mean sacrificing scientific integrity, but rather means obtaining the
15
-------
needed data at the lowest cost. This need evaluation process has been termed the "data quality
objective process."
The control-based approach to obtain regulatory data specifies the exact procedural steps
that the analyst must follow. The major advantage of this approach is that one knows in advance
what must be done. The disadvantage is that the test results may not meet the acceptance
criteria and degrees of certainty normally associated with reliable analytical methods.
As an alternative to the control-based approach to analytical method specification, the
Agency is studying a performance-based approach. In a pure performance-based approach, the
user can employ any procedure so long as the specified quality of results is achieved. The major
advantage of a performance specification is that it assures the desired outcome. The major
disadvantage is that performance parameters must be specified to ensure that the results meet
minimum standards. In addition, performance must be verified immediately to institute
corrective action. Immediate verification may not necessarily be a disadvantage, because total
quality management requires monitoring quality at the earliest point in the process rather than
waiting until the end to determine if a problem has occurred.
The requirement to produce legally defensible data drives the method specification and
documentation process. For a performance-based approach to work, the documentation must
assure that the performance criteria were met. The challenge facing the technical community
consists of finding a means to have both flexibility of methods and confidence in the results.
For the foreseeable future, EPA is likely to rely on both reference and performance-based
methods, and the criteria for a performance-based method will likely be driven initially by the
reference methods.
16
-------
3.2.2. Perspective of the In-House Laboratory from the Regulated Community
Mr. Reid Tait, Dow Chemical Company, presented his views of the advantages and
disadvantages of using performance-based methods in industrial laboratories. Overall, the
concept offers a mechanism to:
• Implement the latest scientific and technological developments in analytical
chemistry,
• Minimize costs and paperwork,
• Maximize flexibility of method selection, and
• Improve the quality of results.
Before these benefits can be realized, some important questions must be answered. The
first question is primarily institutional in nature: How can industry satisfy the large number of
different regulatory program requirements for documentation without being overwhelmed by
paperwork? Historically, each program has required considerable documentation to support any
method variance. Can reasonable and consistent documentation standards be developed that
respect the regulators need to be protective of the environment and yet do not prove unduly
burdensome to industry?
Major technical questions are as follows:
• What criteria will be used to judge acceptability of the method, and what are the
criteria for rejecting a method? Making this determination entails documentation
of performance requirements. Will applying the data quality objective (DQO)
process adequately describe the requirements?
• Must the method be capable of yielding comparable results when performed by
other laboratories?
17
-------
Should the results of the method be confirmed by at least one independent
laboratory or should they be compared to existing methods believed to work for
the analytes of interest in the sample matrix?
Satisfactory answers to the major technical questions will result in the following benefits
being accrued to industry and regulator alike.
• Industry can apply its vast resource of analytical technology and trained personnel
to its problems.
• Regulators and other potential users will have immediate access to this technology
through the documentation supporting the methods.
• Incentives for industry investment in new, cost-effective technology will be
created.
• Both industry and the regulatory community will have a higher level of
confidence in the data because the quality assurance/quality control (QA/QC) data
accompanying the results will be tailored to prove compliance and control of the
analytical process.
The use of performance-based methods will place a premium on effective communication
between the industrial laboratory (or its laboratory service providers) and the regulators. A
means of determining "acceptable performance standards" expeditiously will have to be
developed. Sufficient documentation must to be developed to protect industry from conflict-of-
interest claims and the regulator from accusations of inadequate oversight and control. Industry
will face the possibility of being required to develop additional methods to meet the new
performance requirements.
Mr. Tail's final recommendation was that all methods should be performance-based.
18
-------
3.2.3. Perspective of the Commercial Laboratory Industry
Ms. Carla Dempsey, Lockheed Engineering and Sciences Company presented her views
of the major issues that performance-based standards raise for her industry. This presentation
discussed both advantages and disadvantages associated with the change to performance-based
methods.
Among the principle advantages is the overall elevation of the level of science in the
environmental analytical process. Performance-based methods will make maximal use of the
professional chemist's knowledge and experience. As a result, the chemist will improve the
quality of data, optimize the time and cost of method performance, and create new methods.
Additionally, sound scientific procedures can replace outdated or incorrect ones, improving the
overall quality of the environmental decision-making process.
This new approach will attract highly qualified and motivated professionals to
environmental chemistry laboratories. Currently, such professionals tend not to pursue careers
in environmental chemistry because of both the absence of good science and the boredom
associated with strict adherence to prescriptive methods. The competition between laboratories
to provide higher quality, more rapid, and cost-effective analysis will benefit the consumer of
laboratory services. The ultimate result will be to improve the overall quality of the analytical
chemistry industry.
The commercial laboratory industry has many of the same concerns as the in-house
industry laboratories. These include:
• How will acceptable performance be defined and measured, and what level of
documentation will be required?
• Will it be possible to obtain agreement among the multitude of regulatory groups
who are the ultimate end users of the data?
19
-------
The presentation also raised the following new issues:
How will analyses be procured? The current prescriptive methods are easy to
cost, but how will laboratories establish a fair market price for a performance-
based analysis prior to completion of the analysis?
• What will be used as a "benchmark" to establish method performance? Are there
existing "scientific performance standards" that can be used as benchmarks or are
all current standards method/instrument specific?
• Can current regulations that require use of specific methods be changed in a
timely way to allow implementation of a performance-based program for all
media?
• Who will select the methods used on projects? How will the laboratory have
input into the decision-making process that establishes project objectives?
Ms. Dempsey made some concrete recommendations about specific actions that would
aid in the implementation of performance-based methods. They include:
Defining performance parameters for specific types of environmental decisions
(site surveys, clean closure decisions, monitoring, etc.) as part of the DQO
process;
Providing a guidance document so that data users know what types of
performance data can be collected and how they are used;
Defining documentation requirements;
Promoting the formulation, distribution, and use of appropriate performance
evaluation materials (reference materials); and
Restructuring the procurement process.
The presentation concluded with an endorsement of the shift to performance-based
methods and a description of how the commercial laboratory sector could contribute to a joint
industry-Agency effort.
20
-------
3.2.4. State-of-Science-Perspective
Dr. Billy Fairless, EPA Environmental Services Division presented his views on the
state-of-the-science position in performance-based methodology at this point in time. He stated
that much of the technology necessary to effect the conversion exists, but that it must be
organized to promote the use of performance-based methods. The conversion can be
implemented quickly and will promote creativity and quality science, produce higher quality data
more cost-effectively, and reduce the incentive to commit fraud.
Dr. Fairless supported this latter point with the contention that, when faced with
requirements to use prescriptive methods known or believed to be of questionable value, some
people may adopt method improvements to try to obtain better data. Such changes may produce
data of questionable quality and may even be illegal when viewed in a regulatory sense.
Consequently, a person who pursues good science and sound environmental decision may
become liable for prosecution. Such a situation is in no one's best interests.
The presentation offered a three-step approach for effecting the conversion to
performance-based methods:
1. Fully describe all procedures used to generate data in a quality assurance project
plan (QAPjP).
2. Provide documentation that all procedures were in control during the data
generation process. Standard QC steps are adequate for the purpose if they are
carefully selected and executed.
3. Perform enough replicate sample and spike sample analyses to establish the
precision and accuracy of the determinations at the concentration of interest.
21
-------
At the conclusion of the presentation, the presenters fielded questions from the audience
before they broke into discussion groups.
3.3. CHARGE TO THE WORKSHOP PARTICIPANTS
After the four speakers gave their positions, the moderator asked participants to address
the following questions:
1. How should performance be defined?
2. What documentation should EPA require to verify compliance with performance
standards were met?
3. How will the change to performance standards affect the procurement process by
both EPA and commercial laboratories?
3.4. RESULTS OF WORKGROUP DELIBERATIONS
3.4.1. How Should Acceptable Performance Be Defined?
This question received the largest response from each group. The groups identified four
principle ways to define acceptable performance. Interestingly, each group independently
arrived at the same four approaches. The recommended approaches are:
1. Validation by comparison to reference method.
With this approach, EPA or some other appropriate standards-setting organization would
publish a reference method for the analysis. The performance of the reference method
would be defined in terms of precision, bias, sensitivity, or other appropriate criteria.
The method that a laboratory opted to use would be compared against the reference
22
-------
method's performance. As long as this method met the performance criteria, it would
be acceptable. Participants agreed that the necessary reference method and performance
data do not currently exist for many methods and analytes of interest. If a reference
method and performance data are unavailable for a particular method, participants
suggested that a laboratory perform a side-by-side comparison with an EPA, ASTM, or
other recognized method.
2. Demonstration of compliance with defined, generic OA/QC criteria.
Nearly all participants believed that a sound laboratory QA/QC program is essential for
demonstrating acceptable performance for performance-based analyses. Each standard
QC criteria (blanks, matrix spikes, duplicates, etc.) applies to this problem. Generic QC
criteria for regulatory programs might be set forth in a number of ways including
Chapter 1 of SW-846, QAPjPs, and project DQOs. Participants agreed that meeting or
exceeding criteria would constitute acceptable performance for a proposed method and
that such methods should receive approval. The Agency would specify documentation
guidelines that laboratories would have to meet to prove acceptability.
3. Demonstration of compliance with project-specific data quality requirements.
Adherence to project-specific DQOs should be sufficient to define acceptable
performance. If properly conducted, the DQO process determines project-specific data
quality requirements in terms of acceptance criteria that are independent of the method
of analysis. Method substitution cannot occur where data requirements are based on
results from an operationally defined procedure, such as TCLP or Total Dissolved Solids
(TDS). If the data requirements can be expressed in absolute terms, such as mg/kg or
pH units, with applicable acceptance windows, then any method should be acceptable
provided that the data meet or exceed the acceptance criteria.
23
-------
4. Demonstration of performance using performance evaluation materials representative of
the material being analyzed.
If the proposed method can produce analytical values within the acceptance range of the
certified value using appropriate reference materials, then the method should be
considered acceptable. This standard applies to methods that measure absolute
properties, such as concentration or pH. This method of defining acceptable performance
depends upon the availability of representative performance evaluation samples. The
Agency's support for using Performance Evaluation samples is necessary to promote their
commercial production.
Two other means of demonstrating acceptable performance drew less discussion. These
were:
5. Successful participation in a laboratory accreditation program.
Some participants believed that the Agency should accept data obtained by unapproved
methods, provided that the laboratory generating the data was certified under an
acceptable laboratory certification program.
6. Successful participation by the analyst in a training and certification program.
Some participants believed that the Agency should accept data obtained by unapproved
methods provided that the individual analyst could be certified under an analyst
certification program.
Neither proposal found widespread support.
24
-------
3.4.2. What Documentation Should EPA Require to Verify Compliance with Performance
Standards?
Many workshop participants noted that this question was closely related to the first
question. The type of documentation required depends on the approach selected to define
acceptable performance (See previous discussion.) Several individuals favored global
documentation procedures that would apply across all EPA regulatory programs and would
greatly simplify documentation problems and also reduce costs.
It was pointed out that establishing the legal defensibility of the data entails Agency-
sponsored documentation requirements similar to those in the TSCA and FIFRA programs.
Participants expressed considerable concern that laboratory liability is directly linked with
adherence to published Agency documentation requirements. Use of International Organization
for Standardization (ISO) standards for documentation was suggested as a possible alternative
to the Agency developing another set of requirements.
Nearly all participants believe that the documentation requirements should be based on
the QA/QC procedures that are commonly used in the laboratory. The documentation
requirements should be specified in the QAPjP and be based on the DQO process. In general,
participants agreed that developing an achievable and cost-effective DQO requires that the
laboratory become involved in the earliest phases of the process. A common complaint raised
during the meeting was that the laboratory is currently excluded from the DQO process,
sometimes resulting in unrealistic and/or impossible criteria.
Several people supported the recommendation made in the Dr. Fairless' presentation that
documentation should include:
25
-------
• Full descriptions of all procedures used to generate data in the QAPjP.
• Data indicating that all procedures were in control during the data generation
process. Standard QC measures are adequate for the purpose, provided that they
are carefully selected and properly executed.
• Enough replicate samples and spike samples to establish precision and accuracy
of the determination at the concentration of interest.
In addition, several people thought that the documentation should present the reason for
not using the "standard method."
3.4.3. How Will the Change to Performance Standards Affect the Procurement Process by
Both EPA and Commercial Laboratories?
Only one group discussed this question. The group agreed with the Dempsey
presentation that the procurement process would have to change and probably become more
complicated. The group believed that the increased complication of the procurement process will
impact large and small laboratories differently.
3.5. AGENCY FOLLOWUP TO WORKSHOP
As a followup to the workshop, the Agency's Environmental Monitoring Management
Council (EMMC) has undertaken a review of this issue and is currently developing a program
to expand the use of performance standards in EPA's regulatory programs.
A specific interoffice committee is working under the auspices of the EMMC, and the
committee's recommendation on what performance characteristics need to be monitored is
expected in March of 1993. Participation by individual program offices will follow, on a
program by program basis over time. The specific requirements for documentation will be
contained in individual program rules and regulations.
26
-------
4. CHARACTERIZING MIXED WASTES
4.1. BACKGROUND
A major challenge facing the Department of Energy (DOE) in its RCRA and CERCLA
related cleanup activities is the characterization of mixed waste. Mixed waste is a mixture of
radioactive and hazardous waste generally in a liquid or solid form. The presence of
radionuclides in a hazardous waste sample increases the procedural problems, the safety
precautions, and the cost of analyzing the hazardous component. The presence of radionuclides
also reduces or limits the options available for the management, treatment or disposal of the
waste. For these reasons, one should know precisely how the data will be used, what and how
much data are really needed, and how good that data must be.
In characterizing mixed wastes, samples may be so radioactive that any amount of
handling will pose a safety risk. In such cases, special procedures may need to be developed
to characterize the waste to determine the proper means of treatment, storage, or disposal.
Moreover, careful choice of the properties (in addition to radiological hazards) that need to be
determined has to be made to properly characterize the material with respect to the appropriate
options for treatment, storage, or disposal.
The goal of the session was to provide a forum for exploring ways to develop waste
analysis plans based on the characteristics of the waste materials and the available treatment,
storage, and disposal options for mixed waste management; how to balance environmental risk
and characterization costs; and potential hazards posed by mixed wastes.
27
-------
4.2. SUMMARY OF POSITION STATEMENTS
4.2.1. Regulatory Program Perspective
RCRA, Subtitle C, is a hazardous waste management program that emphasizes proper
waste characterization by facilities that generate, treat, store or dispose of hazardous waste.
Proper characterization is the first step in ensuring the regulator that the waste will be managed
under RCRA in a manner protective of human health and the environment. Waste
characterization usually requires waste analysis via some testing method. However, waste
testing is not always required if background information or process knowledge exists. This is
especially true for radioactive mixed waste which poses testing and sampling problems because
it contains both radiological and chemical hazards. Testing mixed waste is regulated under two
different acts the [Atomic Energy Act (AEA)] for the radioactive component and RCRA for the
hazardous component) with two distinct sets of management requirements.
Ms. Susan Jones, OSW, reviewed the RCRA requirements for mixed waste
characterization. EPA requires that the hazardous portion be properly identified by process
knowledge or testing. In addition to waste identification, waste analysis ensures that
incompatible wastes are not stored in the same management unit. Waste analysis also plays an
important role in determining land disposal restriction (LDR) requirements. Operators of
hazardous waste treatment, storage, and disposal facilities are required to formally document
analysis procedures used in a waste analysis plan (WAP) and to accurately identify the hazardous
waste by testing using existing published data, or process knowledge.
For mixed waste, EPA will consider reducing testing frequency, modifying testing
techniques, and incorporating ALARA (as low as reasonably achievable) practices on a case-by-
case basis. The regulated community has communicated to EPA its difficulty in meeting RCRA
testing requirements on radioactive mixed waste samples. Some problems include choosing
proper sample sizes while maintaining good ALARA principles, determining the number of
samples needed to establish the homogeneity of the waste, and obtaining representative samples
28
-------
of heterogeneous waste. EPA and the Nuclear Regulatory Commission (NRC) recently issued
a joint draft guidance document entitled "Clarification of RCRA Hazardous Waste Testing
Requirements for Mixed Waste" that provides an overview of the RCRA testing requirements.
It addresses concerns such as the potential radiation exposure of employees, elimination of
redundant testing and adjusting sample size to reduce radiation exposure while maintaining
plausible detection limits.
Many problems exist in characterizing mixed waste, but accurate characterization of
mixed waste may be successfully accomplished within the confines of RCRA by using published
data and process knowledge. Testing of radioactive mixed waste as a means of waste
characterization should only be done when absolutely necessary.
4.2.2. Regulated Community Perspective
Mr. Craig Leasure, of DOE's Los Alamos National Laboratory, reviewed the problem
of mixed waste characterization from the perspective of a regulated entity. Characterization of
radioactive mixed waste (RMW) is necessary to determine the appropriate treatment options and
to satisfy environmental health and safety regulations. Radioactive mixed waste can be classified
as contact handled (low-level) or remote-handled (high-level) waste. His paper focused on the
characterization of contact-handled RMW.
There are five major issues affecting the sampling and analysis of contact-handled RMW.
They are:
• The impact of screening requirements on holding times;
• The management of treatment options;
• The generation of secondary RMW;
29
-------
• Detection limits; and
• The determination of when material is radioactive.
Radioactive screening of samples can be a time-consuming process that can impact on
holding times. The screening process can require such a significant amount of time that analysis
can not be successfully completed within the required timeframe. Holding time requirements
must be modified to ensure that laboratories have a reasonable amount of time to complete
analysis. However, degradation of target compounds must remain minimal during storage. All
Government and industrial facilities that handle radioactive material must strictly follow NRC
rules or DOE Order 5480.11. The first level of screening is performed with hand-held devices
that determine dose rate and the level of personal protection required by field personnel handling
samples. The next required screening level ensures that Department of Transportation (DOT)
regulations are met and that contract laboratory radioactive material license limits are not
exceeded. Screening analysis may include, but is not limited to, gross alpha, gross beta, gamma
isotopic, and liquid scintillation counting.
RCRA characterization should not be required on a radioactive mixed waste prior to
treatment, if radioactive content is the driver for the type and extent of initial treatment. For
example, the only treatment option for a waste that contains significant quantities of fission
products may be vitrification. RCRA hazardous waste determination may indicate that the waste
contains trace levels of hazardous solvents. However, this information is not likely to change
the treatment of the material. The more significant risk would be worker exposure to radiation
during sampling and analysis.
Performing RCRA analyses on samples that are known to be radioactive and that are
suspected of being hazardous can generate secondary mixed waste. The solvents required to
perform extractions are hazardous, and the matrix is spiked with hazardous material. The
amount of secondary generated mixed waste can be reduced or eliminated by identifying and
30
-------
using extraction solvents that are not hazardous and by using process knowledge to minimize the
collection of samples and performance of analyses.
EPA requires that relatively large amounts of sample material be taken for analysis to
obtain a low detection limit. NRC recommends smaller amounts of sample material to reduce
worker exposure. Guidance is needed to optimize sample size and detection limits. If the
sample size is reduced by a factor of 50 to reduce worker exposure, then the detection limit will
rise by a factor of 50.
One fundamental mixed waste issue that must be addressed is the establishment of a
clear definition of what constitutes radioactive material. Presently, the definition of radioactive
material ranges from one atom of radioactivity to a material that is greater than 2 nCi/g of
radioactivity.
4.2.3. State-of-Science Perspective
Mr. Wayne Griest and Mr. James Stokely, both of DOE's Oak Ridge National
Laboratory, reviewed the current state of technology with respect to radioactive waste
characterization and discussed promising new developments in the field.
The following mixed waste issues were covered:
• What should be determined and how;
• The applications and limitations of current analytical methodologies;
• Promising new technologies; and
• Areas where further methodology research is needed.
The TCLP developed for disposal of hazardous waste in a landfill, may not be relevant
for RWM disposal. Mixed waste will not be disposed of in a municipal landfill. Regulatory
31
-------
agencies should permit more realistic leaching procedures for RMW, which reflect final waste
forms and modes for disposal or storage. Los Alamos National Laboratory is evaluating a brine
leaching solution for assessing the suitability of radioactive waste for disposal in the Waste
Isolation Pilot Plant.
RCRA allows process knowledge to be used for characterization if a well-defined process
generated the waste.
Collection of a representative sample of a waste is critical to generating valid
characterization data. For homogeneous low-level wastes, sampling is similar to nonradioactive
waste sampling, except additional precautions and constraints are taken to maintain the ALARA
principles. Additional research is needed in the sampling and analysis of heterogeneous waste.
Heterogeneous waste includes drums containing discrete items such as gloves, bottles, metal
items, and towel, and tanks with multiple phases.
More research should be conducted in sample preparation using microwave digestion and
EiChroM columns. Inductively Coupled Plasma-Mass Spectrometer (ICP-MS) also seems very
promising for radionuclides with long half-lives such as 99Tc, 129I, and 227Np. The development
of field analytical methods for the rapid determination of various analytes is under way at Oak
Ridge, Los Alamos, and the Nevada Test Site. The development and validation of field
analytical methods will be important in minimizing the costs of sample shipment and reducing
the number of laboratory analyses.
Idaho National Energy Laboratories (INEL) has adapted TCLP extraction equipment for
hot cell use and can perform the procedure using the required amount of sample on high-level
samples without significant exposure to workers.
32
-------
4.3. CHARGE TO THE WORKSHOP PARTICIPANTS
The participants were asked to address the following questions in their workgroup
sessions:
1. What kind of information is needed (and what quantity and quality) to safely manage a
mixed waste for a series of waste management options?
2. What scientific properties or constituents of the mixed waste should be determined?
The workgroups were asked to focus on the following waste management options:
• Entombment in a below-ground cavern;
• Incineration; and
• Solidification.
4.4. RESULTS OF WORKGROUP DELIBERATIONS
The workgroups discussed each waste management options on which they were asked to
focus. Some workgroups extended their discussions to include other waste management options
such as:
• Vitrification;
• Reinjection;
• Transmutation; and
• Interim warehouse storage.
33
-------
Discussions of these options, which may be considered as subsets of the original three
options, are also summarized in this report. Each workgroup generated a listing of the essential
information that must be collected for the safe handling of mixed waste for the various waste
management options. Summaries of the workgroup discussions on these parameters are given
below for each waste management option.
4.4.1. Entombment
• Physical matrix or waste type (i.e. sludge, filter cake, clothing, gloves,debris,
etc.) is required to provide fundamental information on the waste. This may be
obtained through process knowledge or direct testing;
• Offgas monitoring for fixed gases because of the potential danger of explosion;
• Compatibility information for different waste types that may come into contact
with each other resulting in adverse interactions;
• Free liquid presence both in the mixed waste and in the underground caverns.
Waste with free liquids should be treated to remove the liquid before placement;
• Radiological properties of the mixed waste including daughter products generated;
and
• Heat generation from the decaying mixed waste which would have important
ramifications concerning the increased mobility of the waste, compatibility of
various mixed wastes, and production of gas.
4.4.2. Incineration
Physical properties of the waste including its BTU content, ash content, moisture,
viscosity, and flashpoint;
Physical form of the mixed waste;
Chemical content of the waste including the levels of mercury, lead, PCBs
(especially tri- and tetra-chlorobiphenyl since they are precursors to dioxins), total
halogen content (due to presence of halogenated solvents), total sulfur, NOX, and
volatile radionuclides (e.g. H-3, C-14);
34
-------
• Specific activity of the waste from gross alpha, beta, and gamma measurements;
• Presence of specific radionuclides and special nuclear materials;
• Presence of materials resistant to incineration, such as teflon or steel pipes, which
will require separate disposal for the residues; and
• Any site-specific requirements.
4.4.3. Solidification
• Setting time for the solidification of the waste;
• Compaction strength, leachability, and stability of the stabilization product;
• Total organic content of mixed waste since the presence of organic materials may
interfere with the solidification process;
• VOC levels in the offgas generated during the curing phase; and
• Total anion content of the mixed waste since a high level of anions may prevent
stabilization or adversely affect the strength of the stabilization product.
In summary, the workgroups felt that general testing guidance outlined in 40 CFR 190
needs to be followed.
4.4.4. Other Issues
Workgroup discussion also included issues related to the information that is required in
the safe performance of the waste management options. These are summarized below.
In the case of entombment of waste, the need for performing leaching tests, especially
the TCLP test, is questionable. It was suggested that since no salt migration had occurred at the
site for tens of thousands of years it is not necessary to perform any leaching tests. If such
leaching tests are mandated, however, they should be carried out using a brine solution.
35
-------
For the incineration process, it was noted that the public would demand the monitoring
of all possible air contaminants covered under CAA. It was suggested that a list of target
analytes must be developed for mixed wastes or else all possible contaminants would be
required. It was also suggested that air monitoring could be made part of the permit process.
One workgroup discussed additional issues associated with incineration that should be
considered in deciding to minimize the characterization parameters while protecting the
environment and safety of personnel. Protection of workers taking samples may be either a
high- or low-cost issue depending on the complexity of the sample. Potentially expensive
analyses, such as those for volatile radionuclides and Polychlorinated Biphenyls (PCBs), could
be eliminated with process knowledge. Alternative methodologies which are better, faster, and
cheaper may become available. Risk of population exposure around the incineration operation
and DOT shipping regulations may also impact on the characterization requirements of the mixed
waste and the incineration residues.
A possible simpler approach to ensure that solidification is performed in a safe and cost
effective manner was suggested. It was proposed to first stabilize the mixed waste and evaluate
the stabilization product for compaction strength, leachability, and stability. If a problem is
encountered in the stabilization process (i.e the mixture fails to set or has a low-compaction
strength), then a detailed chemical evaluation should be performed on the mixed waste. If the
stabilized product passed the criteria for strength, stability and leachability, then costly chemical
analysis need not be performed.
4.4.5. Other Waste Management Options
Other waste management options were briefly discussed. It was noted that vitrification
is probably the best technology in terms of cost and safety, and requires the least
characterization; however, it requires the most development before it becomes fully
implemented. Deep-well reinjection is a waste management option requiring characterization
information similar to that for entombment. Transmutation of actinides and fission products to
36
-------
stable or more short-lived species may be a practical option, and the cost may not be much
greater than the cost for characterization. This will produce a waste that is hazardous only and
that may be treated and disposed of by conventional methods. It was suggested that interim
warehouse drum storage with an approved monitoring program may be the best management
option until we better understand the complexity of mixed wastes is better understood.
4.5. CONCLUSIONS
Generally, the workgroups recommended that the only information needed for
characterization is that suggested in 40 CFR 190. Attempts were made to answer the questions
posed to the workgroups. Better, more consistent definitions of types of radioactive wastes and
de minimis levels are needed. Finally, it was concluded that a well-defined decision tree,
including characterization, needs to be prepared initially (i.e. data quality objectives) and be
presented to the regulators in developing and implementing a waste management option for
mixed waste.
4.6. AGENCY FOLLOWUP TO WORKSHOP
DOE currently has a large amount of containerized mixed wastes in storage at its
facilities that need to be cost-effectively characterized prior to treatment. In order to determine
the waste properties needed to properly categorize the wastes for treatment, the EPA
Environmental Monitoring and Support Laboratory at Las vegas (EMSL-LV) in conjunction with
DOE is organizing a workshop to be held in 1993. The workshop will address the issue and
attempt to develop specific solutions to the characterization problem.
37
-------
5. CHARACTERIZING HETEROGENEOUS MATERIALS
5.1. BACKGROUND
In both the RCRA and CERCLA programs, a material often must be characterized to
determine if it possesses some property of concern or interest. Types of questions that may
require answering include:
• Is the material hazardous?
• Can it be safely or successfully managed using a specified treatment technique?
• How much supplemental fuel must be added to incinerate the waste?
When the material being characterized is homogeneous, conventional sampling and
subdividing approaches can be employed. For example, in a truckload of used foundry sand
with an average phenol concentration of interest, the material is in the form of relatively fine
particles with the phenol uniformly distributed on the surface. Conventional compositing and
subsampling can provide representative samples with an average composition very close to that
of the entire load. The actual task of obtaining each subsample from the appropriate point in
the truck may be very difficult, but the approach used is relatively straightforward.
However, insurmountable problems can develop when faced with a heterogeneous
material. Heterogeneity is a relative term and, among other factors, is a function of the
objectives of the characterization and the analytical sample size. This workshop was concerned
with wastes of such various particle size, waste consistency, or extraordinary concentration
gradients that the sampling and analytical objectives could not be met using traditional
approaches and standard techniques. This is an important and difficult problem that has been
of concern to many organizations involved in waste management.
38
-------
The workshop goal was to advance the state-of-the-art techniques for characterizing
difficult to sample wastes. This covers wastes with a highly variable nature, (e.g. wide ranges
of particle size, large concentration gradients, and mixtures of different waste materials (e.g.
dredge materials containing sludge, tires, construction debris, bottles, cans, etc.)). These types
of materials present challenges that traditional sampling and analysis approaches cannot meet
with the level of certainty required for modern waste management decisions.
RCRA has historically taken the position that the sample actually being analyzed does not
have to be representative of the material being characterized, but rather that the sum total of the
data must represent the property of interest. The problem presented was how to develop
practical solutions to characterize these difficult situations. In this session, the participants
discussed how to cost-effectively characterize materials that are inherently very heterogeneous
and present characterization difficulties.
5.2. SUMMARY OF POSITION STATEMENTS
5.2.1. Regulatory Program Perspective
Mr. Charles Ramsey, EPA's National Enforcement Investigation Center (NEIC),
reviewed some issues involved in implementing the RCRA regulatory requirement that a
"representative" sample of the waste be collected and analyzed. 40 CFR, Part 260.10 stipulates
that a representative sample, '... means a sample of a universe or whole (e.g., waste pile,
lagoon, ground water) which can be expected to exhibit the average properties of the universe
or whole." This definition has two aspects.
1. What is an average property?
2. How is the universe or whole defined?
The inability to answer these questions in clearly defined scientific terms leads to most of the
problems surrounding the characterization of heterogeneous wastes.
39
-------
Mr. Ramsey Proposed that hazardous waste characteristics are not "averageable"
properties. For example, corrosivity is expressed as pH that is a logrithemic function, while
ignitibility can be expressed as passing or failing the 60°C ignition test. With these
characteristics, the numerical mean is not meaningful; therefore, an average value cannot be
obtained.
Using the average value as a measure of hazard with respect to the toxicity and reactivity
characteristics may also present potential problems. When the material to be evaluated consists
of both a contaminated phase and an inert phase, use of the mean may result in a waste being
classified as nonhazardous even though a substantial portion of the waste might be above the
regulatory threshold and pose an environmental hazard.
Using "the most likely result" approach (also known as attribute testing) avoids many of
the above problems. It requires that a certain percentage (e.g., 50%, 75%, 95%) of the
individual sample results fall below the regulatory decision value. This avoids problems
associated with determining average values. Use of this approach, however, will require a
change in the regulations.
Even when the property of concern is a continuous variable and, therefore, averageable,
collecting representative samples of heterogeneous wastes is often extremely difficult. Many
heterogeneous wastes are discontinuous in terms of physical/chemical properties and, given the
small size of analytical samples relative to the discontinuity, it is not possible to collect (and
demonstrate) representative samples of the wastes. Therefore, to ensure that an adequate
characterization is conducted, a clear definition is needed of what constitutes the "universe or
whole" to be characterized and what constitutes an acceptable level of characterization (i.e.,
What degree of confidence, that the property is below the regulatory threshold, must be obtained
to characterize the waste as nonhazardous.). The universe of the whole must also be defined
to determine the period of time and space in which the samples must be collected.
40
-------
5.2.2. Regulated Community Perspective
An industrial perspective was prepared by Mr. David Reese, Safety-Kleen Corp., and
presented by Mr. Gene Klesta, Chemical Waste Management Corp. The presentation focused
on the types of samples received at recycle and waste disposal centers and the different types
of problems these materials represent relative to heterogeneous wastes found in the field. The
industry receives large quantities of containerized wastes from different sources which in the
aggregate form a very heterogeneous set.
Industry strongly supports considering the method of management when establishing
analytical requirements. The current requirement for full-scale analysis is, in many cases, not
necessary and represents a significant financial burden for generators and the waste management
industry. The industry recommends that wastes destined for recycling or treatment be exempted
from rigorous analysis if the waste management process and end products are thoroughly
characterized during operations and prior to disposal.
It is further recommended that the Agency permit the use of methods other than those
in SW-846 when conducting acceptance testing of wastes for treatment and recycling. Using
proper QA/QC procedures is adequate to ensure that reliable analysis are conducted. Again,
process monitoring and end product analysis provide sufficient protection of the environment.
The industry will work with the Agency to bring industry-generated methods into the public
sector.
The industry position is similar in concept to the Agency's interest in moving towards
performance-based methods using sound project QA plans, data quality objectives,and accepted
QA/QC procedures as the quality assurance tools to monitor performance.
41
-------
5.2.3. State-of-Science Perspective
Dr. John Maney, a private consultant, presented a detailed strategy for random and
nonrandom sampling of heterogeneous wastes. The presentation, described a unified way of
looking at all types of waste based on the source and degree of heterogeneity of the material.
Using flow charts and an extensive set of definitions and examples, the paper presents
concrete examples of options available to the field sampling team and the analytical chemist to
generate reliable analytical data from complex, heterogeneous materials. However, unless the
sample undergoes extensive comminution, Dr. Maney indicated that obtaining an analytical size
representative sample of a heterogeneous waste is almost impossible. (See the full paper in
Appendix A for details.).
5.3. CHARGE TO THE WORKSHOP PARTICIPANTS
Before breaking into three workgroups, the participants were asked to address three
questions:
1. Should EPA change from current practice (average testing) to attribute testing for
properties that are not averageable?
2. If so, what is the highest practicable degree of confidence (%) that could be required?
3. Should there be an override that says that if some samples are greater than "X", the
waste is hazardous even if the number of such samples are below the percentage
established in question 2?
A large part of each workgroup's time was devoted to discussion of the practical
definition of attribute testing and how it differs from average property testing. This decreased
42
-------
the time available to discuss the three issues, so no effort was made to address the third issue
within any workgroup.
5.4. RESULTS OF THE WORK GROUP DELIBERATIONS
5.4.1. Should the Agency Change to Attribute Testing for Properties that are Not
Averageable?
Approximately 90 percent of the participants believed that attribute testing should be
adopted for properties that are not averageable. This approach is more costly than average
property testing in those situations where the average property was determined by compositing
a series of field samples prior to analysis. There was some discussion that compositing of
samples prior to analysis is one way to determine an average property that avoids many of the
technical problems associated with averaging the results from a series of discrete sample
analyses.
Neither the attribute testing nor the average property testing approach eliminates all the
problems associated with the sampling of wastes prior to analysis. The nature of attribute testing
can make the sampling in many situations easier. The group agreed that approximately 80
percent of the error in characterizing waste arises from field sampling error that effects both
approaches equally.
The Agency should focus its efforts on providing sampling guidance for heterogeneous
wastes. Special efforts should be made to provide guidance for specific types of wastes such as
tires, telephone poles, construction debris, and military ordinance. Also, knowledge of the
process that generated the waste is an important part of the decision to use attribute testing and
would be a primary driver in selecting the sampling strategy.
43
-------
5.4.2. What is the Highest Practicable Degree of Confidence that can be Required?
Several participants felt the answer to this question should take into account the degree
of health and ecological risk associated with the material and the amount (mass or volume).
Therefore, each site should be considered unique and receive individual consideration. A site-
specific degree of confidence should be carefully developed as part of the data quality objective
process and be codified in the site-specific QAPjP.
It was generally agreed that a maximum of 10 percent of samples collected should be
allowed to fail the regulatory criteria. Sixty-six percent of the participants felt that a 4 to 6
percent failure rate was the most appropriate. All participants felt the Agency should publish
clear guidelines on determining if attribute testing should be applied to a specific situation.
5.5. AGENCY FOLLOW TO WORKSHOP
Subsequent to the workshop, Agency staff met to discuss what steps to take to address
the issue and the ideas presented during the session. It was decided that quantitative data are
needed to determine:
• What the practical effect would be if the attribute testing approach was used
instead of the current arithmetic mean.
• How the number of samples needed to achieve a defined degree of confidence
would vary across the different types of materials characterized in the RCRA and
CERCLA programs.
During the next year, the Agency plans to examine its data in an attempt to answer the
above questions. However, the Agency will greatly appreciate receiving any additional
information that will help determine if a change in approach will either improve the quality or
lower the cost of decision making. If anyone has data that might assist in this effort, EPA will
appreciate receiving such information. Such information should be sent to the author at the
address indicated in the FORWARD to this report.
44
-------
APPENDIX A:
ISSUES ASSOCIATED WITH ADOPTION OF PERFORMANCE-BASED METHODS
-------
ISSUES ASSOCIATED WITH ADOPTION
OF PERFORMANCE BASED METHODS:
INTRODUCTION
David Friedman
USEPA Office of Research and Development
401 M Street SW
Washington, DC 20460
BACKGROUND
EPA implements a number of regulatory programs which require the collection of
environmental data. These include SDWA, CWA, RCRA, CERCLA, CAA and FIFRA.
In any monitoring program, selection of the appropriate measurement method is a key
part of the process. Over the years, in crafting its monitoring requirements, the Agency has
primarily adopted a policy of publishing appropriate test methods for complying with the various
regulations. The methods have been written with varying levels of detail. However, rather than
issuing flexible methods and specifying required performance, in general, the Agency has opted
to give the analyst detailed instructions on how to conduct the analysis.
PROBLEM
This has resulted in several problems.
1. Long delays in bringing new technology to bear for cost-effectively analyzing
samples.
Praatlai onJufyl3,1992ai EPA WoHahcp I
en 'Performance-Based Hotels' A-l
-------
2. In some cases the required methods either give biased results for some types of
samples or do not yield the requisite sensitivity. Since in many cases the methods
do not provide for modification by the analyst, the analysis can lead to results that
do not meet the data quality needs.
3. By rigid specification of methods rather than of data quality, laboratories are
often forced into using expensive technology when a simpler, less costly approach
might be suitable.
PURPOSE OF TODAY'S MEETING
The Agency is aware of these problems and is trying to do something about it. One
option being looked at it to move from a system of rigid methods to one where performance is
specified and flexibility is given on how the analysis is conducted.
However, implementing this new approach is not easy. A number of problems and issues
need to resolved before the system can be changed. These include:
• How will acceptable performance be defined?
• What documentation, etc. is needed to verify compliance with performance
standards?
• How will the change to performance based methods affect the procurement
process, both by EPA and by commercial clients?
We in EPA are aware of the problems. We agree that something needs to be done.
However, while many people have ideas on how such a program can be implemented, not
everyone agrees on what is the correct approach. That is where you come in.
The two questions we are asking for your help in answering are:
Presented on Jufy 13. 1992 at EPA Workshop I
on -Performance-Based Methods' A~2
-------
How should performance be defined?
What documentation should EPA require to verify that the required performance
standard were met?
We also want your input on how you think a change to performance based methods
would affect the procurement process, both by EPA and by commercial clients?
If you think of any ideas, information, or suggestions that you feel the Agency should
consider when addressing this issue, please send them to us. Send your comments to:
David Friedman (RD-680)
US Environmental Protection Agency
Washington, DC 20460
We will need to receive your comments by August 21, 1992 in order to incorporate them into
the conference final report.
Prtsenud on Jufy 13, 1992 at EPA Wortshcp I
on 'Pafomancf-BasedUeAods' A'3
-------
PERFORMANCE-BASED METHODS-one view from the regulators
James M. Conlon
Director Drinking Water Standards Division (WH-550D)
U.S. Environmental Protection Agency
William Telliard
Chief, Analytical Methods Staff
Engineering and Analysis Division
U.S. Environmental Protection Agency
ABSTRACT
This paper will offer some definitions of performance based chemistry methods, discuss
some of the differences between the performance based and the controlled based methods and
describe some of the processes that the Environmental Protection Agency is using to sort through
the issues associated with the movement toward the broader use of performance based methods
in its regulatory programs. Some of the risks and benefits seen by the regulators will also be
described.
INTRODUCTION
As part of its efforts to examine the quality of the information and data underlying its
basic regulatory decisions, some parts of the Environmental Protection Agency are taking a
closer look at how and why they generate chemistry data. The fashion and form by which it
requires others to produce similar information is also being examined. More specifically several
programs are examining the question of whether the up-front specification of step by step
analytical chemistry methods is the only way, or in fact even the preferable way to cause data
to be generated by the regulated community. The step by step procedures articulated in the rules
Presenud on Jufy 13. 1992 as EPA Workshop I
on 'Performance-BasedMethods' A~4
-------
and regulations of the last twenty years are being closely examined with an eye toward their
being supplemented with the so called performance-based methods; methods whose only
specifications might relate to precision, accuracy, sensitivity, precision, and specificity rather
than step by step, cook book instructions.
This effort actually goes back a ways and has its roots in the review of the chemistry in
the water programs that was directed by amendments to the Clean Water Act in 1987. The
results of this review concluded that:
EPA should establish additional performance criteria for EPA chemical methods.
The Office of Water (OW) and the Office of Research and Development (ORD) should
agree on ways to improve the process for developing and standardizing methods.
EPA should develop quality control requirements for all test methods promulgated at 40
CFR Part 136.
EPA should establish an Environmental Methods Management Council to coordinate:
common analytical methods,
methods manuals, and
the development of uniform QA/QC requirements within methods.
At the same time, there was a growing concern that a number o the promulgated methods
were not necessarily giving us the best answers at the best price. In some cases we were getting
inferior information at higher prices and in all too many cases some of us came to realize that
we had very carefully crafted a system that fostered duplication and unnecessary redundancy of
methods without a clear cut benefit. The only purpose for collecting any of this data is to
provide sound, accurate information that can be used in making public policy decisions; that
have as their primary purpose the prevention or elimination of disease or damage to people or
environmental systems.
Presented on July 13. 1992 at EPA Workshop I
on 'Performance-BasedMethods' A~5
-------
Collection of information in search of absolute truth is appropriate for others but not
necessarily for those in the environmental protection business. This is especially true if it causes
expenditure of monies in pursuit of the ultimate detection level with little or no return in
additional protection of the public health. That does not mean that you sacrifice scientific
integrity but rather that you focus on the need for which the data are being generated to get the
most for your money. Some people have called these data needs data quality objectives.
Concerns over these types of issues as well as a specific examination of the number of
the methods that had been promulgated by the various program offices within the Agency caused
a number of people to begin to actively search for a mechanism that would help remove some
of the confusion. Hopefully this mechanism would also move toward a day when the analyst
has greater opportunity to actively improve on basic methods, and in turn improve the efficiency
of the overall decision processes if by no other way than making good data some more cheaply.
In the more conventional arenas associated with analytical chemistry on a day to day
basis there is a marked increase of interest in the question of how to add efficiency to our efforts
to improve the overall quality of the data. Interest in performance-based analytical methods also
lies in the problems associated with the development of uniform QA/QC requirements across
various methods. The existing suites of analytical methods developed by the various EPA
Program Offices do not currently use a consistent set of performance measures, and, where the
performance measures are the same in similar methods from different Programs, the
performance specifications often vary widely.
Performance versus Control-Based Methods
To understand a performance-based method, one must understand the alternative. The
alternative is a control-based method. In a pure performance-based method, the method user
would be able to choose any procedure or instrument so long as the specified quality of results
is achieved. In a control-based method, the procedural steps are specified and if these steps are
Presented on July 13, 1992 at EPA Workshop I
on "Performance-BasedMethods' A~6
-------
followed exactly, the desired quality of results will be achieved theoretically and the performance
does not need to be specified or measured. Control-based methods are required of course when
the technique itself defines the parameter. Examples are BOD, TOC, and COD, etc.
The military is a large user of control-based procedures and specifications. Military
instruction manuals are written in a terse series of steps that military personnel are expected to
follow exactly. In many instances, the person following these steps does not know the
performance that is expected and the performance may never be measured.
In practice, most analytical methods are both control-based and performance-based
because of the fashion in which methods are developed and edited for publication. A number
of agency methods provide procedural steps, and performance specifications that must be met.
EPA 600/1600 wastewater methods (40 CFR Part 136, Appendix A) are mostly
performance-based, in that many performance specifications are given in these methods and the
analyst must meet these specifications. The analyst is "permitted to modify the method to
improve separations and lower the cost of measurements" so long as the performance
specifications are met. However, the methods are somewhat control-based, in that the user is
not permitted unlimited discretion in modifying the method. For example, an infra-red
spectrophotometer cannot be substituted for the gas chromatograph specified in these methods.
The EPA Superfund Contract Laboratory Program (CLP) methods are control-based, in
that the procedural steps are specified in great detail and must be followed regardless of the
analyst better judgement. The advantage of a control-based approach in the CLP is that it leads
to fewer concerns about comparability when data for samples from a single site are generated
by a number of laboratories. Unfortunately, on occasion it also has the potential for generating
data of legendary quality.
Presented on Jufy 13, 1992 at EPA Workshop I
on •Performance-Based Hehods- A-7
-------
The major advantage of a performance specification is that it assures the desired result.
The major disadvantage is that the flexibility of the procedural steps allowed may result in a
compromise in performance, unless the performance is completely specified. In addition,
performance must be proven nearly immediately in order to institute corrective action. Total
quality management and other quality assurance techniques teach that it is better to monitor
quality at the earliest point in a process than to wait until the end to find out that problem has
occurred.
The major advantage of a control specification is that each step is controlled within a pre-
defined limit. In NASA and military hardware manufacturing programs, each step may be
witnessed by a quality control officer, thus assuring that the step was followed. The major
disadvantage is that when a performance test is finally made, the control may have been
insufficient to assure the desired performance.
Performance Specifications Identified by the Workgroup
Because of issues such as these and also in explicit response to recommendations in the
518 Report, EPA formed the Environmental Monitoring Methods Council (EMMC) in late 1989.
In turn, the EMMC formed a Panel on Methods Integration that consists of several workgroups,
among which is a workgroup on performance-based methods (the Workgroup). This Workgroup
is chaired by Orteria Villa, Director of the Central Regional Laboratory in EPA Region HI at
Annapolis, Maryland.
Initially the workgroup arrived at the following definition for a performance based
method:
"A monitoring approach that permits the use of appropriate analytical methods that meet
established, demonstrated method performance standards based on Agency data quality
needs. Minimum required elements of performance, such as precision, accuracy,
Presented on July 13, 1992 at EPA Workshop I
on 'Performance-BasedMethods' A-8
-------
sensitivity, specificity, and detection limit, must be specified and an available method
must be shown to meet the method performance standards."
However, in later discussions a somewhat different view of performance based methods
evolved.
A somewhat simpler definition is:
"A procedure containing the necessary methods performance specifications to assure the
desired quality of results."
Initially the EMMC Workgroup identified the following list of performance specifications
as important to analytical methods. At least one member of the workgroup believed that each
was "essential."
Precision
Bias
Sensitivity
Specificity/selectivity
Detection limit
Written method
Performance range
Performance objectives linked to regulatory objectives
QC and performance documentation/initial validation/routine performance documentation
Data review process/internal and external data audits
Performance evaluation samples
Analyst qualifications/proficiency
Several of the items on the list above are not performance specifications, but address the
desire to satisfy a particular program need. After further discussions the work group was able
to agree that most of these items are related to either precision or bias. Perhaps it should also
Praaatd on July 13, 1992 a! EPA Wortshcp 1
on 'Performance-Based MeAods' A~9
-------
be noted that a number of these characteristics are already cited in the 600 and 1600 series
methods.
In its most recent deliberations the Workgroup seems to have come to the conclusion that
the most appropriate way to proceed toward the adoption of a performance based ethic is more
along the lines of establishing not only performance criteria but also dealing explicitly with
questions of how much documentation would be required for the use of methods under routine
laboratory conditions. In addition, the workgroup also insists on the programs maintaining at
least a single reference method that is assured of being able to meet the stated data need.
This group also seems to believe that indicator of method precision include method
sensitivity, range of performance, and the ruggedness of a method. Precision criteria should
establish numeric, statistically-based quality control requirements related to measurement
replication and method calibration and should account for matrix effects.
Limitations of Pure Performance-based Methods
The number of true performance-based methods is limited in part because not every detail
of performance can be specified and in part because in a decentralized regulatory program it is
much easier to perform comparisons of techniques against a national standard that it to examine
the science behind procedures unique to each laboratory. Consider a method for the
determination of the 126 Priority Pollutants given in Appendix A to 40 CFR Part 423. The list
consists of organics, metals, cyanide, and asbestos. The organics consist of pesticides/PCBs,
non-pesticides, and 2,3,7,8-tetrachlorodibenzo-p-dioxin (dioxin).
Consider just the requirements for determining of non-pesticide/non-dioxin organics. A
specification can be written for each compound with a required precision, accuracy (recovery),
and detection limit. The laboratory making the measurement would then be free to choose the
technique used for each compound.
Praaiiai at Juty 13, 1992 al EPA Vforbhop I
on •Pcrfonnanct-Bascd Methods' A-10
-------
To an extent, this approach to the determination of organic compounds was tried by
EPA's Effluent Guidelines Division (EGD), the predecessor division to BAD, in the development
of the regulation of discharges from the Organics, Plastics, and Synthetic Fibers (OCPSF)
industry. An "Organic Chemicals Verification (OCV) Program" was initiated in the late 1970s
to characterize waste streams from OCPSF manufacturing plants. The laboratories were given
specifications for precision, recovery, and detection limit.
One of the outcomes of the program was that although the data produced were of a
defined quality, the laboratories exercised great latitude in their choice of extraction and cleanup
techniques, GC columns, and detectors. EPA's Science Advisory Board (SAB) reviewed the
OCPSF program and concluded that the laboratories had been given too much latitude in
selection of these techniques and that the program needed to be better controlled.
The Definition of a Method is Changing
The definition of an analytical method is changing from a series of steps performed by
an analyst to a series of steps that will result in a data set that will stand up to legal scrutiny.
Defendants involved in litigation over environmental measurements will attempt to attack the
data produced in order to win their case, and industries subject to regulation will scrutinize data
carefully to make sure that the regulation is in their best interest. If it can be shown that the
laboratory did not follow the analytical method, or if all of the method specifications were not
met, the defendant may prevail. Therefore the data package that results from the method must
demonstrate that all of the control- and performance-based specifications were met. This
requirement to produce a legally defensible data set changes the way in which performance
specifications are developed, in that every specification must be achievable.
Method writers frequently include specifications casually, without studying the
consequence of failing a given specification. This places the analyst or laboratory in an
unacceptable position. If a specification is failed, and the cause is attributable to a problem
Presented on July 13, 1992 at EPA Workhcp 1
at 'Performance-Based Methods' A-11
-------
outside of the control of the analyst or laboratory (e.g., a matrix effect), there is no recourse but
to continue running the sample (or QC test) over and over.
Similarly, specifications in analytical methods can be so stringent that they are difficult
or impossible to meet. Therefore, EPA must ensure that the desire to achieve perfect data is
tempered by what is realistically achievable. However, it is essential that the way to the realism
is fully documented.
For performance based methods to work with the Environmental Protection Agency there
will have to be some agreed upon documentation. The first purpose will be to assure that the
performance criteria have been met. To do so, it is likely that the agency will require at a
minimum a written description of the method that meets some specified format; results of initial
validation studies as well as the maintenance of routine performance information.
The actual values of the performance criteria themselves will of course derive in large
measure from the data needs but also from the reference methods. In many cases the reference
method will be at the cutting edge and the performance characteristics will be most difficult to
achieve using alternatives. However, it is likely that in at least as many cases, the analysts
outside the agency will be able to improve on the specified method, or in fact substitute a new
method that clearly delivers an equivalent value with a reduction in cost. The movement toward
the wider use of performance based methods by EPA will not be without its fits and starts but
it will happen. There will be concerns expressed and a number of those are being identified and
focused, and it is likely that more will be added to the list in coming months.
For the moment at least the following minimum list of benefits and risk of performance
based methods has emerged. This list seems to capture a number of the views and concerns of
the regulators who are familiar with the issue, but if you know regulators you know that it will
never be complete.
Presented on July 13, 1992 at EPA Workshop I
on 'Performance-BasedMethods' A-12
-------
Nonetheless as we see them today, the advantages of performance based methods seem
to be:
* FLEXIBILITY
* ALLOWS US TO DEFINE THE QUALITY OF THE DATA
* HAS POTENTIAL FOR IMPROVING DATA QUALITY
* HELPS ASSURE COMPARABLE (EQUIVALENT) DATA
* ALLOWS LABS TO ADOPT NEW TECHNOLOGY RAPIDLY
* ALLOWS LABS TO EMPLOY INNOVATIONS/CREATIVITY
* MAY MAKE COMPLIANCE MONITORING MORE
INTERESTING/ CHALLENGING
* SHOULD SPEED METHODS DEVELOPMENT
* SHOULD HELP EPA MEET CONGRESSIONAL DEADLINES
* SHOULD IMPROVE THE SCIENCE SUPPORTING THE EPA'S
MONITORING PROGRAMS
—and the disadvantages perceived are:
* MAY PLACE THE SMALL LAB AT A COMPETITIVE
DISADVANTAGE
* MAY REQUIRE BETTER TRAINED ANALYSTS
Presetted an July 13, 1992 at EPA Workshop I
<* 'Performanct-BasedMethods' A-13
-------
* MAY REQUIRE MORE EXTENSIVE DOCUMENTATION
* MAY REQUIRE MORE FREQUENT AND INDEPTH
INSPECTIONS
* MAY REQUIRE MORE HIGHLY TRAINED AND SKILLED
INSPECTORS
* MAY REQUIRE MORE PE CHECK SAMPLES
* MAY REQUIRE A DATA BASE TO TRACK METHODS BEING
EMPLOYED
On the other hand, the use of only control-based specifications could result in
uncontrolled variability of results because performance would not be tested, thus violating the
DQOs of precision and accuracy.
Therefore, EPA will continue to need to use both control- and performance-based
specifications in its analytical methods. With prudent use of these specifications, and further
method evolution based on improvements in technology, EPA believes that its data quality
objectives will met more easily and at a lesser cost.
Presented on Juty 13, 1992 at EPA Workshop I
on •Performance-BasedMethods' A-14
-------
ISSUES ASSOCIATED WITH ADOPTION OF
PERFORMANCE-BASED METHODS:
INDUSTRY PERSPECTIVE
Reid Tail
Dow Chemical Company
1261 Building
Midland, MI 48667
INTRODUCTION
Industry has complained for years about Environmental Protection Agency (EPA)
methods that do not work and give data that is at best questionable and costly to obtain. All of
us, industry and regulators, want data of known quality based on defensible methodology. The
main issue or question regarding any method is: Does it work and how well? The concept of
performance based methods can be supported by both industry and the agencies as a way of:
1. Removing road blocks to application of the latest scientific and technological
developments;
2. Minimizing paperwork and analytical costs;
3. Maximizing flexibility in method selection;
4. Improving the quality of the results.
Prcsmud on July 13, 1992 at EPA Workshop I
on •Paformanct-BasaiMethods' A-15
-------
STATEMENT OF THE ISSUE
As we consider the issue of performance based methods, we can briefly review the
experience of our Canadian neighbors. They used EPA guidelines as a starting point, but instead
of mandating or suggesting methods, they chose to set up performance based guidelines. The
Canadian government, contract laboratories, and industrial laboratories soon determined that
performance comparison could only be done in a clean matrix. Therefore, specifications were
based on clean matrices. The laboratories found that they could meet almost any detection limit
if they changed to higher analyte concentrations where better precision allowed them to
calculate lower detection limits. Without a definition of a detection limit at a given
concentration, almost any answer was possible.
If a laboratory could meet the requirements in a clean matrix, then the system was
working and the laboratory could do the analysis. Canadian industry found that the matrix effect
issue had to be explained, proven, and negotiated with the government. Does this circle sound
familiar? As David Friedman said in the opening remarks, "How will acceptable performance
be defined and how will we document it?"
Proving acceptable performance has been difficult in the past. It is one of the major
reasons for the high cost of doing business and getting the environmental job done. Much of
the work we do to prove acceptable performance is caused by the lack of trust between industry,
the public, and government agencies. If a company says it is using EPA method 8270, the
company, the contract laboratory, and the Agency know what can be expected from the method.
The permit writer feels a certain comfort knowing an EPA method is being used. If a permit
writer is asked to accept a method not found in Federal Regulation, EPA manuals, or tested by
the EPA, he/she must confer with technical staff and talk with the director. Meanwhile, the
much needed permit languishes on the desk. Everyone is afraid that the one being regulated is
trying to "put one over on them", or if they accept a company method, that the Agency has
"given in".
Presented on Jufy 13, 1992 at EPA Workshop 1
an 'Performance-BasedMethods' A-16
-------
A broader look at the present situation is that EPA methods (with few exceptions, i.e.
land ban requirements) are recommendations, and are to be used as guidance. This means they
can be modified to meet the matrix requirements or other data quality objectives. In practice,
the recommendations concept has not been applied. If it were, we would not be here having a
discussion on performance based methods. In general, all methods are performance based.
Industry and agencies alike should support the concept that if a method does not perform, change
it so it will, i.e., apply the best available technology (BAT).
Several problem areas which prevent the use of performance based methods need to be
considered from the regulated community point of view:
1. How can we minimize the amount of paperwork required to prove a method is
adequate in meeting data quality objectives and to present the necessary data in
a way that is correct and convincing to the non-technical, non-trusting, and
skeptical regulator or permit writer?
2. How can we demonstrate that a company or commercial laboratory method can
be performed by any laboratory and get the same results?
3. What criteria will be used to judge the acceptability of the method? What is the
criteria for rejection?
4. Should a proposed method be checked by at least one independent laboratory or
against existing methods which can be applied to the sample matrix and to the
analytes of interest?
5. How cost effective will performance based methods be? Will they save us money
or will industry have to pay commercial laboratories to develop a method for each
matrix or data quality objectives (DQO)?
Presented on July 13, 1992 at EPA Wortshcp 1
°n 'Performance-BasedHeShods' A-17
-------
6. How can we build trust between industry and the regulator? Will the attitude of
the regulators be "this is the level we want you to get to and I do not care how
you do it-just do it"; "I want the method to have no bias."; "Don't give me that
matrix problem excuse-it is your stuff and it is your problem"?
7. Acceptable performance guidelines need to be defined and documented. Is the
DQO concept based on performance methods the answer?
It is not my intent to answer all of the above questions in the limited time allowed for
this presentation, but to provide points to explore during the discussion period.
PERFORMANCE-BASED METHODS AND LABORATORIES
The advantages and disadvantages of performance based methods will, to some degree,
depend on the company and whether it has or wants to use its own laboratories and methods for
analysis. There are companies that have their own laboratory with the latest technology and
companies which choose to use one or more commercial laboratory to do their work.
What are some of the advantages and disadvantages of performance based methods from
the viewpoint of the regulated community?
Using Industry Laboratories
To the large company which chooses to use its own laboratory, the advantages and
disadvantages are:
Presented en Jut? 13, 1992 a: EPA Workshop I
on 'Performance-BasedMethods' A-18
-------
Advantages:
1. The company has a high experience level with the application of the methods to
their compounds of interest. The limitations of the method are usually
understood.
2. The methods are already developed.
3. The people are already trained.
4. There is less cost to the taxpayer, since the Agency will not need to develop
additional methods.
5. The rate of development of new cost effective methods to meet the performance
standards will increase.
6. Manpower intensive and time consuming methods will be eliminated since there
are cost incentives for improvement.
7. Less complex methods could be used, since the methods would be designed for
the company's compounds of interest (not the universe).
8. The company can submit the results with an appropriate level of QA/QC data and
supply a certification of accuracy and responsibility.
9. The Agency can accept the data with a minimum of QA/QC review since it is the
responsibility of the data supplier to defend its data.
Presented on July 13,1992 at EPA Workshop I
on 'Pafomancc-BasedMethods' A-19
-------
Disadvantages:
1. The Agency is not familiar with the method, and has historically required a
considerable amount of documentation to prove beyond any question, that the
method will answer the question and meet the performance criteria or DQOs.
2. The new technology may not be understood by the regulators. A lack of
understanding of a technology or a method makes Agency personnel suspicious
of the intent of the regulated community and very uncomfortable with any method
changes suggested to them.
3. The methodology has come from the very person who is being regulated.
Conflict of interest is always a question.
4. Negotiations will require time. Additional time will be needed in order to make
sure the data will be accepted by the Agency.
5. Considerable time would be required to compare company developed methods
against methods the Agency is familiar with.
6. Intralaboratory studies may not have been done; this makes it difficult to assess
the methods ruggedness.
7. Developing matrix specific methods could be more costly to companies.
Using Commercial Laboratories
Companies using commercial laboratories may find several advantages and a few
disadvantages. They are:
Presented on July 13, 1992 at EPA Worbhcp I
an 'Performance-BasedMahals' A~20
-------
Advantages:
1. Superior methods or technology will be developed by laboratories to give them
a competitive edge.
2. More cost effective methods are available from which to choose.
3. There is more experience with a variety of matrices.
4. It can be less expensive in those cases where the commercial lab has methods
already developed.
5. The Agency may be familiar with the commercial laboratory's methods and have
confidence in them; thus, less time is required to convince the Agency that the
correct methods have been applied.
6. The conflict of interest question is minimized.
Disadvantages:
1. Communications become more difficult between the client and the laboratory.
This is and has been a serious problem since the client carries the legal
responsibility for the data generated by the laboratory.
2. The regulated company cannot agree to any negotiated performance standards
until the contract laboratory has spent some time analyzing the company's
samples.
ld on July 13. 1992 at EPA Workshop 1
on 'Performance-BasedMotets' A~21
-------
3. Development of a method to meet the performance standards could be very
costly.
RECOMMENDATIONS
All methods should be performance based and meet minimal QA/QC requirements instead
of the present overkill of redundant standardizations and checks. These increase cost and
decrease productivity in an effort to eliminate ALL questions. Data is seldom or ever without
qualification; it is seldom or ever absolute.
Let's keep the system simple, get back to the basics, and report the data with the
following information:
1. The instrument type
2. Any special sample preparation steps
3. The method for entering the sample into the testing device
4. The instrument operating conditions
5. The tuning requirements have been done and their
results (mass spectrometric analyses)
6. Standardization procedures, including linear range
7. Any blank and duplicate data
8. Spike recovery yields for the analytes and surrogates
9. Sample chromatograms
10. Detection and quantitation limits in the matrix of interest
11. Calculation results, including:
a. Units of measurement
b. All complete equations used
The amount of detail required will decrease to the following:
Presented on July 13, 1992 ai EPA Workshop I
on •Performance-Based Ueshods' A.-22
-------
1. Report the method used
2. Note any deviation from the method and why
3. Report blank results
4. Report spike, duplicate spike, and duplicate sample results
5. Report calibration checks
6. Note that tuning results were within specifications
7. Report the data and any qualifiers
8. Report detection or quantitation limits of the
reported data
The above may need to be modified for the different types of methods used. Most of
data that needs to be reported is already being done and should require little or no changes.
CONCLUSIONS
Any method should be acceptable as long as it produces data of known and acceptable
quality. The data reported to an Agency should state the method used and the results of spike
recoveries, duplicates, blanks, quantification limits, and (where applicable) surrogates.
Performance based methods will allow greater flexibility, and if properly instituted,
would be cost effective. The methods will meet the DQOs and answer the question or questions
being asked about the site, monitoring wells, etc, without using exotic methods designed for the
universe. Performance based methods, used in conjunction with agreed upon data quality
objectives, can increase the development of new and better methods, be cost effective, and
address environmental needs.
However, the best available technology or performance based technology cannot be
effectively applied until mistrust, lack of knowledge, and poor communications between the
Praenled on Jufy 13, 1992 a! EPA Workshop 1
on 'Performance-BasedMethods' A-23
-------
regulated community and the regulatory agencies are eliminated, or at least minimized. This
will require improved communications, honesty, trust, and the education of each other.
Presented on July 13, 1992 as EPA Wortshcp 1
on "Performance-Based Meshods' A-24
-------
ISSUES ASSOCIATED WITH ADOPTION OF
PERFORMANCE-BASED METHODS:
INDUSTRY PERSPECTIVE
Carla Dempsey
Program Development Manager - EPA Programs
Lockheed Engineering and Sciences Company
1725 Jefferson Davis Highway
Suite 300
Arlington, VA 22022
INTRODUCTION
Commercial laboratories face different problems in considering how to implement per-
formance-based methods than those which are faced by regulators, EPA and other scientists, and
the regulated community. However, the problems that these groups must consider will drive the
primary issues and problems faced by the commercial laboratory industry in implementing the
program.
The number, type and severity of problems that commercial laboratories will encounter
when implementing a program for performance based methods depends almost entirely upon the
scientific and policy decisions that are made initially. These decisions include the definition of
performance, determining the type, volume, and submittal requirements for information required
to demonstrate and document performance, constraints imposed on selection of methods, the role
laboratories will have in assisting data users in their planning of the analytical work, and
structure of the bidding/costing requirements for competition and completion of work. The ease
of implementation will also depend upon how clearly data users define and articulate the types
of samples and methods required for data needs of their projects, or alternatively what role the
Praaued on July 13, 1992 at EPA Workshop I
on •Ptrfomancc-BasedMethods' A-25
-------
laboratory will have in defining the method requirements. Additionally, the availability of
quality matrix-specific performance evaluation materials utilized initially to evaluate and optimize
methods and then used during ongoing operations to assess performance of methods will impact
the ease of implementation and regulatory assessment of laboratory performance.
Science and Policy Questions
Some of the science and policy related questions and problems surrounding this issue that
will affect ease of implementation include:
• How will acceptable performance be defined?
• How will acceptable performance be measured?
• How will performance be documented in terms of the report and data that must
accompany the sample results?
• How will analyses be procured for sampling and analysis events that have data
needs that are not "static?" (e.g. a performance based method may be easily
procured, for a process effluent sample can be adequately described for a
procurement. In this case, the samples do not change significantly and the
decisions based on the use of the data do not change. However, how can
procurement be accomplished for many site samples if the samples are from many
Superfund sites? In this case the composition of samples differs substantially
from site to site. Also the types of decisions that will be supported by sample data
vary significantly from site to site and from one phase of the site remediation to
another.)
• Is the "performance" of different methods on the same matrix/site/sample
significantly different, and if so, what defines the method performance that
provides the benchmark to which all other method data will be compared? (This
issue is important for consideration in determining how to use data generated
from several methods on the same study and in cross-comparisons among
studies.) How will method bias be documented and what limits for acceptability
for method bias will be established. Will the limits for acceptability vary
depending on how the data will be used?
Presented on Jity 13, 1992 at EPA Workshop I
on •Performance-BasedMethods' A~26
-------
• Do any environmental methods exist that have been evaluated by strictly scientific
performance rather than instrument-specific performance measurements that could
serve as a prototype performance method for implementation purposes? (An
example of scientific performance criteria is the acceptable retention time delta
as defined by comparison with a standard and the separation/overlap distance of
peaks on a GC trace as defined by comparing standards to samples, rather than
requiring use of one particular column and knowing what performance is required
for that one column.)
• Can current regulations that require use of specific methods be changed in a
timely way to allow implementation of a performance based program for all
media?
• Is there a potential for all EPA Program Offices (and other government agencies)
that require laboratory data to come to consensus and accept the performance
based approach?
• Who will select methods for use on projects? Will the laboratories have any
scientific latitude in applying methods to optimize the performance or by
substitution of a better performing procedure or will this be precluded by need to
define operating parameters so that data from numerous laboratories are
comparable and can be utilized on the same study?
Business Aspects of Performance-Based Methods
The questions and problems surrounding this issue that pertain to the business aspects
related to change to performance based methods can be relatively simple or extremely complex
depending upon the policy and scientific decisions made initially in structuring the program:
• How will the EPA procure laboratory services in a manner that assures that the
price and quality of data are fairly evaluated in a competitive process?
• How will the other government agencies and their contractors or other users write
contracts to procure performance-based laboratory services?
• What role will the laboratory have in assisting/selecting the methods that can best
be utilized for evaluation of particular samples
Presented cnjufyl3,1992
-------
Who will specify the type of technical information and documentation
requirements that will be considered acceptable format? Will the format be
tailored for each method or will it be universal in that requirements for QC
information and results?
Acceptable performance depends on the project objectives. How will
performance objectives be stated to the laboratory and how will they be factored
into the bid price of the data?
Will all government agencies, EPA Program Offices and other entities that model
their requirements after EPA accept and implement this approach? If this does
not occur, the cost of analysis may increase due to implementation of different
procedures that add to the complexity of laboratory operations, rather than a
decrease in cost based on real cost of the chemistry involved.
Viewpoint from the Commercial Laboratory Perspective
Implementing performance based methods is important for the commercial laboratory
industry for several reasons. Many of the reasons are similar to those shared by scientists,
regulators and the regulated communities. Some of these reasons for supporting change to
performance based methods stem from benefits that may be achieved, including:
• Data that is more useful for its intended purpose will be produced for sound
decision making, particularly if laboratories are able to select one of the myriad
of methods currently required by regulations and optimize the method both for
establishing continual performance and for determining limit of detection and
quantitation on a realistic schedule with samples.
• Science will benefit, as the output of the best professional judgments and practices
and efforts of professional chemists.
• Better, less costly, and less time-intensive procedures that produce quality data
for decisions can be developed and implemented if the field is "deregulated".
• Data will be assessed in a scientifically valid manner instead of basing value on
requirements that are not all scientifically valid and/or not useful to data users.
Presented en July 13, 1992 ai EPA Wortohcp I
on 'Performance-Based Methods' A-28
-------
• Sound scientific procedures may promote reason in establishing "holding times"
based on scientific data so that sample analyses can be prioritized based on user's
needs, not on artificial limits imposed by regulatory or contractual requirements.
• Scientific judgment may be used to optimize method performance to cost-
effectively produce useful data, rather than follow rigid protocols exactly as
prescribed. (Optimal results on specific samples are not often obtained if methods
are not optimized for the samples.)
• Environmental chemistry laboratories, both public and private sector may be
better able to attract and retain professional chemists that stay challenged in their
positions and advance the collective body of environmental chemical analysis
knowledge.
• Per sample cost of analyses may be minimized, allowing for better statistical
sampling to occur and more samples to be analyzed at lower cost per sample to
allow for more informed decision making.
• Competition would be increased among laboratories based on ability to complete
analyses on samples using cost-effective methods and modifications to the
methods that are cost-effective and meet customer's needs. Laboratories could
then stay competitive based on excellence, based on quality of data as measured
by the usefulness of data for decisions, its cost, and overall customer service.
• The current practice of bidding contract work at the price required to win may
be eliminated. Therefore, competing with laboratories that complete the analyses
to the "letter-of-the-contract but with marginal quality could be replaced with fair
competition among laboratories that produce quality data. This change may serve
to improve the overall baseline of quality for the industry.
Unfortunately, there are also many apparent advantages in maintaining the status quo,
primarily because the potential impact of implementation cannot be understood until the exact
nature of the program is defined. Therefore, unless the program is well thought out with indus-
try providing input into it, the change may be viewed as very difficult to implement. Consider
some of the advantages to the status quo that occurs because of the competitive bid procedures
used to compete for business and conduct business after contracts are awarded:
• Contracting and bidding structure is established, well understood and the primary
means of competition for business.
Presented en July 13, 1992 at EPA Wortshap I
on 'Performance-RasedMethods' A~29
-------
• Customers "order" certain analyses from the laboratory that have clearly defined
contract and procedural requirements.
• The evaluation criteria by which the usual analytical data is measured is
established.
• The data output requirements and software used to generate this data is in place
and well understood by most data users
• The price per sample is based on adherence to specific requirements in methods
and delivery of specific data deliverables. Therefore, the competition is based on
well understood and conduct of the same methods and delivery of the same data.
Other considerations that have also emerged because of the current way laboratory
analyses are conducted but are not generally perceived as positive are:
• Laboratory staff at commercial laboratories has been trained to perform
laboratory analyses by rote so the skill level has been established by criteria that
currently do not require use of expert professional judgment, as will be required
by more analysts in using performance based methods.
• Laboratory equipment purchase has been driven by regulatory method
requirements, the ease at which the data systems produce required data formats,
and for a myriad of reasons not related to which instruments produce the best
data for specific purposes. The laboratory industry may therefore not have
purchased the equipment that is most versatile in the laboratory and/or that can
perform the best analyses for particular samples, but will have to compromise on
optimal performance based on lack of the best equipment until capital equipment
expenditures are made.
• In the past, laboratories performed the analyses as specified by the customer with
minimal consultation concerning the optimal selection of methods since
regulations specified most of the selections. In the future, how will customers
decide which methods are optimal? Will the laboratories provide assistance?
How will the work be accomplished "optimally" if the bid price dictates how
much time the analysis can take? How can fair competition among laboratories
occur if the work is not well defined in order to establish the basis of compe-
tition?
Presented en July 13, 1992 at EPA Workshop I
an •Performaxt-BasedMethods' A-30
-------
STATEMENT OF POSITION FROM THE COMMERCIAL LABORATORY
PERSPECTIVE
Change must occur so that the chance for data required for sound decision making to be
delivered to the data user is maximized. Continuing with the status quo limits the chance that
data needed to make sound environmental decisions will be delivered.
Why Labs Want Performance Based Methods
The first reason is that this approach will minimize duplication of methods. Currently,
five methods for broad spectrum volatiles analysis are accomplished at the Lockheed Analytical
Laboratory (LAL): 624, CLP, 8240, 8260, and 524.2. At least 6 methods for organochlorine
pesticides are set up and operational. The differences between the methods are minor and most
have no bearing on data quality. However, it is a very large effort to keep standards,
procedures, QC requirements, analytical specifics, and reporting requirements separate and
accomplish all requirements correctly. The duplication of effort and complications arising from
this is a major cost for most environmental laboratories. Further, this expended cost and effort
has no discernible benefit on either quality or throughput. In fact, it diverts the expertise in the
laboratory from true quality and throughput issues. It prevents the laboratory from expending
time on optimizing method performance for an analytical procedure so that bias, limit of
detection, quantitation limits and overall performance of the methods are effectively established.
The second reason is that methods remain stagnant in a non-performance based system.
As an example of this, consider one pesticides method. Until approximately 1987-1988, the
CLP laboratories were required to use 1978 technology (packed columns). When the CLP
pesticides method was changed in 1990, the technology was updated to 1985 technology. The
procedures in the protocols are written as requirements, so innovation and improvements that
may improve performance are prohibited.
Presold on Jufy 13, 1992 at EPA Workshop I
on 'Performance-Based Methods' A~31
-------
Possible Solutions to the Problem, Including Advantages and Disadvantages
Before possible solutions to the problem are discussed, performance based methods must
be defined. Some methods that are commonly thought of as performance based are really
process based methods even though they include some measures of performance together with
a myriad of process requirements. For instance, the CLP currently includes both process and
performance requirements. An example of one of the restrictive process based requirements is
the requirement for extraction of samples. If a 400 ml beaker is required, it is a violation of
the contract to perform the extractions in a different type container. Many parts of some EPA
data validation procedures are also largely process based. Much of the focus is not on results
of analyses, but on the process used to generate results (e.g. was the DDT retention time greater
than 12 minutes?)
Some actions can be taken to solve the problem:
Defining performance parameters for specific types of decisions is a logical
extension of the data quality objective (DQO) process. For example, some data
is collected to determine if gross violations or gross level of contamination has
occurred. Should performance parameters and requirements on those parameters
for data needed for this purpose be the same as that required for data needed to
determine if clean-up levels have been reached? Clearly, the decisions are
different and therefore we must consider the need for performance parameters and
their associated requirements, as well as the needed data documentation
requirements that are based on what decisions the data will be used to support.
In some instances, such as screening for gross contamination, kit tests may be
cost effective and performance parameters may include only a blank check,
standard known analyte check and a reading of the samples. Clearly data
collected to evaluate effectiveness of clean-up must be supported by more
rigorous performance parameters and requirements. Such requirements will
include information that allows bias of method to be evaluated, as well as
precision and capability of method to perform on the samples. We must be
flexible in allowing for differences in performance parameters, requirements, and
documentation requirements and assure that all requirements are driven by the
needs to evaluate the data with respect to the decisions it is collected to support.
Presetted en July 13, 1992 at EPA Workshop I
en •Performance-Based Methods' A~32
-------
• Provide structure in a guidance document so that data users know what types of
performance data can be collected and how it can be used in both very general
and very specific examples. Included in this document should be the use of
performance evaluation samples and how they can be used both to establish bias
of methods and also to evaluate continuing performance. The two uses of
evaluation materials must be well understood by data users so that the materials
are not misused. An example of their misuse is assuming that acceptable
performance on one sample provided to a laboratory assures the data user that the
method used by the laboratory has an "acceptable level of bias" or that the data
is acceptable for specific decisions. In order to effectively use performance on
one evaluation sample, the laboratory must first establish the bias and precision
of the method by many replicates of performance evaluation materials of similar
matrix to the samples being evaluated. Only after method precision, accuracy and
stability are demonstrated can performance evaluation materials be effectively
used to predict performance on samples.
• Define what data needs to be collected to document performance for the samples
being analyzed by that method. Many performance parameters can be collected
to document method performance. However, information that must be
documented for any sample includes: (1) methodology used for both sampling
and analysis, and supporting information including detection and quantitation
limits established for the analytical method, (2) data establishing the response of
a blank sample associated with the samples, (3) calibration information showing
the method is in control and the standard curve brackets the samples, (4)
precision information provided by replicates, and (5) accuracy information
provided by control samples and spiked samples.
• Promote the formulation, distribution and use of quality matrix-specific
performance evaluation materials for both establishing method bias and day-to-day
performance and also to assessing performance of laboratories on an ongoing
basis by regulators.
• Restructure procurement process. One alternative to the restructuring can be
envisioned to be a two-step process if the need for data of a particular type is
large enough to warrant optimizing a method specifically for the samples. Phase
one could be called the Development Phase and would consist of developing the
optimal method, performance criteria and associated requirements to be used on
specific samples. Performance evaluation materials that have qualitative/semi-
quantitative to quantitative values associated can be generated for use in
evaluating laboratory performance during this phase. The Analysis Phase can
then be competitively bid based on specifications for use of a method that can
meet very specific performance criteria and requirements that can be met, and
specific performance criteria documentation that must delivered. The successful
Presented on Jufy IS, 1992 at EPA Workshop I
ai 'Performance-BasedMethods' A~33
-------
laboratory then can be "creative", use a method that will produce the quality of
data needed, and be assured that the evaluation of data acceptability will be
grounded on performance criteria and requirements that are meaningful and
achievable for the samples to be analyzed.
• Alternatively, commercial clean-up companies, industries that require RCRA
permit-mandated analyses and other industrial clients may elect to award a
services contract in order to complete the Development Phase and Analysis Phase
in a smooth process. This approach encourages cooperative problem solving and
loyalty to a laboratory based on service, price, and quality of resultant data. The
approach is attractive, but is complicated in many cases because the requirement
for work is to be subcontracted to the lowest bidder.
• Another alternative is for the government to allow award of service contracts
instead of firm-fixed price contracts to the laboratories to "perform" quality work
on samples rather than use the same methods and same quality for all samples,
no matter what the intended use of the data. This approach has been tried and
has proven to be very expensive when compared to fixed price laboratory
services. However, these comparisons were conducted by only comparing price
per sample, and did not measure the usefulness of resultant data to the user.
Another argument against this type of approach is that it can foster ineffective use
of time and resources.
• A final alternative is for the commercial laboratory community and the regulators
to define categories of samples that are cataloged as difficult to easy based on
very specific descriptions, such as oily waste to well-water samples. The
laboratories could then bid competitively on each type of sample category. If the
samples as received in the laboratory failed to meet the definition of the category,
the laboratory would then work with the customer delivering the samples to
redefine the sample category prior to beginning analysis. Additionally, the
laboratories could define charges for additional analytical procedures that may be
required on a case-by-case basis for specific types of samples and define with the
customers and regulators the justifications based on data quality that would invoke
their use and additional charges to be incurred. For this approach to be feasible,
it is mandatory that performance requirements for methods used by laboratories
be well established and meet well-defined performance requirements on specific
sample matrices. Since it is unrealistic to expect all very difficult samples to be
able to be effectively analyzed by even the best method available, the implications
of this situation must be carefully determined prior to implementing this type of
program.
Presented on July 13. 1992 at EPA Workshop 1
on -Performance-BasedMethods' A"34
-------
Agency Actions
Agency actions that can be accomplished to implement the suggested solutions include:
• Fully define the program so that it can be implemented based on full
understanding of its scope and potential impact.
• Gamer the support of all EPA Program Offices and other agencies that now have
similar requirements that require regulatory specified methods.
• Define what performance criteria are critical, important and mandatory for any
analysis. Prioritize existing criteria based on value in defining the quality of data
for different generic an specific purposes. This action was also included in the
previous discussion included in the Possible Solutions to the Problem Section.
Implementation Impediments
Impediments to implementing the solution have been implicitly included in the descrip-
tions of the options. However, one impediment that is usually not considered is the dilemma
faced in attempting to determine how to utilize past data collected under different requirements.
One must use historical data and the lessons learned from generating the data as the baseline for
establishing the new system that will be used. Past data must be utilized to the full extent
possible based on the decisions to be made and the information content of the data. In
evaluating the data for potential use in specific decisions, regulators and other data users may
gain insight into what data must be gathered to establish performance and what performance
requirements are mandatory for specific decisions. Since history cannot be changed, we must
use the past data and past experience to establish more effective means to collect data for future
decisions. We must use lessons of the past to craft the best program for the future and not allow
past experience and the volume of data generated by past programs to be used as a reason to
impede changes and improvements.
PraaOfd enjufyl3,l992at EPA Workshop I
on 'Performance-Based Uahods' A~35
-------
Commercial Laboratory Assistance
The commercial laboratory industry can assist EPA in solving the problems in several
ways:
• Provide EPA with data demonstrating the within laboratory biases that occur
using different methods on the same samples/standards, based on past and future
results.
• Assist EPA in defining the set of performance criteria that provides the most
information to enable users to assess data fitness for use in particular applications.
• Provide EPA with data for specific methods demonstrating what performance
objectives on critical criteria can be met with real waste samples, which has the
benefit of producing data that can be used to determine if methods are capable of
producing required quality of data on real samples and if laboratories can achieve
required performance.
• Assist EPA in structuring the type of competitive bid process that could be used
to procure performance-based analyses. Past performance-based schemes were
effective in many ways but had several flaws that made it difficult for the system
to work, which were discovered and resolved after implementation. These types
of lessons learned from the commercial laboratories perspective may help in
avoiding releaming the same lessons.
• Provide information on cost impact of selecting different performance indicators
and the costs/benefits for frequency of performance data collection.
• Provide general information on the cost/benefit of changing to a performance-
based approach based on potential changes in procedures that could allow use of
less costly methods than those currently required by regulations
CONCLUSIONS
Change is needed in order to deliver data needed to support sound decisions. Change
is also needed to stimulate the industry to use the best new techniques and methods to solve
problems that cannot be tackled using the restrictive protocols required today. Therefore,
Presented on Jufy 13, 1992 at EPA Worktop I
on "Performance-BasedMethods' A~36
-------
change must occur and will best proceed if sound scientific and business decisions are the basis
for implementation. Industry and government must plan the program cooperatively so that it
will be successful.
From our perspective, we prefer that the methods include mandatory surrogate spiking
requirements and ongoing verifications of performance, such as laboratory control samples,
blanks, and other such information. The resultant program should enable a laboratory to select
cost effective methods for analysis if spiking, blank, detection limit and other "science based "
performance requirements are met.
OTHER PERFORMANCE BASED METHODS DISCUSSIONS
Laboratories have heard previous discussions of performance based programs which
sound like "dial a precision and accuracy number". They seem to be based on the implicit
assumption that laboratories have a variety of methods sitting on the shelf where precision and
accuracy are directly proportional to price. The client then (1) determines the appropriate
precision and accuracy, (2) tells the laboratory to obtain that precision and accuracy, and (3)
obtains data that meet the requirements at the lowest possible cost. This sounds wonderful, but
it is a simple approach that has no basis in reality. Some reasons follow:
"Someone once said that the secret to making money in the laboratory business
is doing a whole lot of what you are good at." Laboratories like to have single
methods set up to accomplish particular analyses. Setting up multiple methods
of a single analysis is costly, leads to confusion in the laboratory, and increases
the frequency of errors. A laboratory would have to perform multiple methods
that produce data with different precision and accuracy, and the precision and
accuracy would have to be established for each one of these methods in order to
be accomplish the "dial a precision and accuracy" approach.
The greatest source of variability in environmental measurements is field sampling
and environmental heterogeneity. The place to customize precision and accuracy
numbers for a project is during the design of the sampling plan. Field duplicates
Praeaed en Jufy 13, 1992 at EPA Worbhcp 1
on •Performance-Based Methods' A~37
-------
may add more information than that obtained from more stringent controls on the
laboratory in establishing the accuracy of measurements.
Data quality and laboratory efficiency usually go hand in hand. An ongoing focus
of most laboratories is to improve both the quality and efficiency of the analytical
procedures. As quality is improved, the number of reruns is lowered. As wasted
analytical effort is eliminated, both quality and throughput are improved. With
the exception of screening methods, the most cost effective analysis is frequently
also the highest quality. If precision and accuracy need to be tightened beyond
what is considered reasonable for the most cost-effective analytical procedure,
the most cost effective and prudent way may be to conduct duplicate analyses or
take duplicate samples.
Presented on July 13, 1992 at EPA Workshop I
on 'Performance-Based Methods' A~38
-------
ISSUES ASSOCIATED WITH ADOPTION OF
PERFORMANCE-BASED METHODS:
SCIENTIFIC PERSPECTIVE
Billy J. Fairless, Ph.D.
USEPA Environmental Services Division
25 Funston Road
Kansas City, KS 66115
INTRODUCTION
I am recommending that the environmental monitoring community and the regulatory
agencies should replace existing requirements specifying in detail how data are to be generated
with guidelines describing how the quality of data should be documented. This recommendation
probably goes beyond what David Friedman had in mind when he asked me to discuss my
experiences with performance-based methods. However, I have concluded that it is the
performance of the system of methods, rather than a single method, which is only a part of the
system, that should be the focus of our attention. This change would give data users the ability
to estimate the risk they take when making a decision and it would be more cost effective than
current practices. It could be implemented quickly and it would promote creativity and quality
science, it would produce higher quality data, it would reduce the incentive to commit fraud and
the potential for controversial legal issues.
Because guidelines for documenting the quality of environmental data do not exist, the
environmental community has made large time and dollar investments performing actions they
assume will improve the quality of the data they generate. Included in this list of actions is the
development and use of sample collection, handling, preservation, and analytical methods.
These are often specified by a regulatory authority or by an applicable contract. In addition,
various options to certify laboratories, field scientists, and analysts also are included or are under
Presented on Jufy 13, 1992 at EPA Workshop I
en •Performance-Based Methods' A~39
-------
consideration. One thing that all of these practices have in common is the fact that they will not
provide a description of the quality of the data managers use to make their decisions; which is,
in my opinion, the most important quality assurance product the scientific community should be
providing to environmental managers. Instead, existing requirements are based on unproven
assumptions that data quality will both be adequate and will be improved if the actions described
above are taken. Although these assumptions may be true part of the time, they are not always
true and may not be true for most data collection activities. Why should we continue to base
important decisions involving millions of dollars on unproven assumptions, when the quality of
the data supporting a decision can be obtained and documented with less time and for fewer
dollars than are now being spent?
WHAT IS REQUIRED TO DOCUMENT THE QUALITY OF SCIENTIFIC DATA?
Three kinds of information are needed to document the quality of scientific data. The
first of these is done primarily before the samples are taken and are usually contained in what
I call an approved Quality Assurance Project Plan (QAPP).
The first kind of information required is a description of all procedures used to generate
the data. These procedures should be in sufficient detail that another scientist could reproduce
the work described. The documentation (QAPP) should include a description of the decisions
to be made with the data to be collected and how the data will be used (perhaps with other
existing information) to make each decision. For example, a lawn of a residence might be
considered to be contaminated if the average soil lead concentration is above 500 ppm at a 95 %
confidence level. The documentation should contain a description of the design of the monitoring
network, each station in the network, how and when samples are to be collected, handled and
analyzed. The QAPP should also include a description of how the precision and accuracy of the
data are to be measured.
Presented on July 13, 1992 al EPA Workshop I
on 'Performance-BasedMethods' A-40
-------
The second kind of information needed is evidence that all procedures were actually in
control during the time that data were being generated. This is usually done by showing the
absence of a measurable amount of contamination and by using various performance evaluation
standards and sample spikes. The QAPP must describe how the data generator will show that
each procedure was in control so it is known that all data were generated under "in controlled"
conditions. These quality control data are collected when the environmental data are collected.
If control of a procedure is lost, data generation is simply stopped until the problem is corrected.
The third kind of information needed is a measure of the accuracy and precision of the
environmental data at the concentration of interest. This information would normally include
a knowledge of the distribution (normal,log normal, etc.) of the data and would usually be based
on replicate samples or spikes.
REASONS WHY IT IS BETTER TO DOCUMENT DATA QUALITY THAN TO INVEST
IN PRACTICES THAT MAY OR MAY NOT PROVIDE USEABLE DATA
It is Cost Effective
My recommendation requires the preparation and acceptance of one guideline. Existing
regulations require development of many methods and then tracking compliance with these
methods. In addition, such things as good laboratory practices and laboratory certification are
required by some programs and states even though there is no evidence that these practices
improve data quality by a measurable amount.
It Produces Higher Quality Data
My recommendation includes a measure of data quality. It allows the science to be
optimized to the decision and to physical conditions such as sample matrix and other pollutants.
If the measure of success is data quality rather than adherence to a specified procedure, those
Presaged on Jufy 13, 1992 at EPA Workshcp I
on. 'Performance-BasedMahals' A-41
-------
procedures and practices that most efficiently provide the quality needed will quickly replace
procedures and practices that are neither efficient or cost effective.
It can be Implemented Quickly
Implementation does not require that a government or a private Agency establish an
elaborate bureaucracy to manage the program. It only needs agreement that the recommendation
has merit, preparation of the required guidelines describing how each of us should document the
quality of the data we generate and use followed by regulatory approval to use the guidelines
rather than existing methods when appropriate to do so.
It Promotes Creativity
Between 5 and 10 years are required to get regulatory approval for the general
community to use a "new" method. Therefore, most environmental scientists work 5 to 10 years
behind the state-of-the-art in their chosen discipline. As a result, there is less incentive for
environmental scientists to develop methods specific to their waste, which reduce waste, that are
more efficient, sensitive, accurate or precise because such methods are not easily incorporated
into the regulatory system. Should we not provide an incentive for scientists to develop and use
the highest quality methods they can and is there any doubt they will do so if that incentive is
provided? I expect the fact that environmental regulations effectively prevent environmental
scientists from working at the state-of-the-art of their profession has damaged the quality of
environmental science more than any other single factor.
It is Consistent with Total Quality Management
The recommendation prevents most data quality problems by requiring better planning
with an emphasis on consensus building during preparation and approval of the QAPP. It
Presented on Jufy 13, 1992 al EPA Workshop I
on 'Performance-BasedUeShods' A-42
-------
includes a measure of the desired product, data quality, that can be used to measure the
effectiveness of the different processes that produced the data.
It Reduces the Incentive to Commit Fraud
When scientists are required by regulation or contract to increase their costs by taking
actions that do not improve data quality, there is a natural tendency to take shortcuts.
Unfortunately, many of these shortcuts are illegal and totally unacceptable even though they
would be acceptable to the scientific community. My recommendation provides the flexibility,
in most cases, to practice quality science within the recommended guidelines.
'"This paper describes the views of the author and does not reflect EPA policy.
Praaaed on Jufy 13, 1992 at EPA Workshop I
on 'Pafomante-BasedMethods' A-43
-------
APPENDIX B:
PREDICTING THE ENVIRONMENTAL IMPACT OF OILY MATERIALS
-------
PREDICTING THE ENVIRONMENTAL IMPACT OF OILY MATERIALS:
INTRODUCTION AND REGULATORY PERSPECTIVE
David Friedman
USEPA Office of Research and Development
401 M Street SW
Washington, DC 20460
BACKGROUND
Prevention of groundwater contamination has historically been one of the EPA's highest
priorities in implementing the RCRA program. To that end, the Agency has developed and
promulgated test methods, fate and transport models, and regulatory standards to control the
management of wastes whose properties might pose a hazard to groundwater. Scientists are
concerned with oily waste due to its volume, toxicity, and potential for causing severe ecological
damage. Such wastes take many forms including: liquids of widely varying viscosity,
contaminated soils, sludges, and tarry "plastic" masses.
Oily wastes have some unique properties. They can migrate like a liquid but appear to
be a solid. Because they result from many commercial processes and applications, they are
broadly distributed, of very large volume, and of tremendous commercial importance.
In developing the hazardous waste identification characteristics, EPA highlighted its
concerns with protecting ground water resources by developing the Extraction Procedure
Toxicity Characteristic (40 CFR 261.24). The characteristic relies on laboratory procedures to
predict toxicant mobility.
Presented on Jufy 14, 1992 at EPA Worlahcp D on Predicting
tkt Environmental Impact cf Oify Materials' B~l
-------
Over the years a number of laboratory extraction methods have been applied to the
problem of predicting what might migrate from oily wastes managed under landfill conditions.
Among the test methods that have been developed and employed to identify those wastes which
might pose an unacceptable hazard are: EPA methods 1310, 1311 and 1330 (Extraction
Procedure, Toxicity Characteristic Leaching Procedure and Oily Waste Extraction Procedure).
The current approaches all have deficiencies with respect to predicting the mobility of
toxic chemicals from oily wastes. Methods 1310 (EP) and Method 1311 (the TCLP)
underestimate the mobility of many oily wastes due to filter clogging problems, their precision
is less than desirable, and they have certain operational problems. Conversely, Method 1330
(OWEP) probably overestimates mobility since it emulates a worst possible case scenario. None
of the available laboratory mobility procedures is thus totally satisfactory.
PROBLEM
Given the importance of this issue, it is imperative that accurate, precise, and usable
approaches to characterizing the mobility of oily materials be developed. That is what we are
here for today.
The problem of mobility estimation is too large and complicated for us to try and solve
all its aspects in a half day. Therefore, we will focus on just one aspect of the problem -
predicting the initial source term. To put it another way, we want to predict the highest
concentration of material that might be released from the waste to the soil immediately below
the point of disposal for some reasonable amount of time. This information would then feed into
the fate and transport models used to predict the final toxicant concentration at some distance
away from the disposal area.
Presented on Jufy 14, 1992 at EPA Workshop n on Predicting
fa Environmental bnpaa of Oiiy Materials' B-2
-------
DISPOSAL SCENARIO OF CONCERN
The priority waste management facility scenario that EPA has selected to be modeled in
this workshop is placement of the waste into or on the ground (e.g., landfill or lagoon). Within
this scenario a number of parameters need to be considered. These include:
• Temperature (assume temperate conditions),
• Rainfall regime,
• Biodegradation,
• Hydrolysis,
• Soil types (assume soil underlying the waste management unit has a
porosity similar to that of sand), and
• Amount of waste (assume amount is large enough so that it can be
considered to be infinite).
FATE AND TRANSPORT MODEL CONSIDERATIONS
To properly manage oily wastes to protect ground water sources from contamination by
waste constituents, an adequate model to predict the fate and transport in the subsurface
environment is needed. The Agency is developing a model to simulate the migration of aqueous
and nonaqueous phase liquids and the transport of individual chemical constituents which may
move by convection and dispersion in each phase.
As input parameters, the model needs information on the amount and composition of both
the aqueous and nonaqueous phase liquid portions of the waste as well as the composition of the
leachate that might be generated by action of surface waters on any "solid" material that may
have initially been present in the waste material. At this time, the Agency does not have a
precise way of defining either an "aqueous phase liquid" or a "nonaqueous phase liquid".
Praeaai at Jufy 14, 1992 as EPA Workshop U an Predicting
fa Emvcnmoual Impact of Oify Materials' B-3
-------
REGULATORY PERSPECTIVE
In an ideal situation, an effective approach to evaluating oily wastes would:
• Be simple to use (not require sophisticated equipment, nor constant
attention by a highly trained technician),
• Be inexpensive to run,
• Take as little time as possible to perform (ideally no more than 24 hours),
• Be accurate (relative to predicting behavior of waste in the environment),
• Be precise (i.e., be reproducible),
• Be rugged (capable of characterizing a broad range of waste types and
constituents of concern), and
• Not generate wastes (e.g., generate little if any solvent waste and waste
contaminated media).
The characteristics that the approach should have (maximum desirable values for each
parameter) are:
• A high degree of freedom from false negatives (any errors tend toward
overestimation of threat to environment),
• Precision (RSD <50%),
• Relatively low cost,
• Taking as little time as possible to perform (<24 hours), and
• Ruggedness.
Presented on Juiy 14, 1992 at EPA Worbhop U on Predicting
theEnvirnonemaUmpactcfOify Materials' B-4
-------
OPTIONS FOR CONSIDERATION
I. Develop a two-component mobility test that determines the fraction of the waste
which is flowable (capable of physical movement under the influence of gravity
and overburden pressure) under the conditions of the test. Suggested conditions:
room temperature and 50 psi. Define as mobile all material that is flowable
under terms of the test plus the aqueous extract of the non-flowable fraction.
Under this option, the procedures used are independent of waste properties and
disposal environment.
II. Develop a single generic laboratory procedure to estimate what disposal point
concentration would result from aqueous leaching of the hazardous constituents
from both the mobile and non-mobile fraction of the material. Under this option,
the procedures used are independent of waste properties and disposal
environment.
in. Employ a series of laboratory test procedures to evaluate the waste material.
These procedures would be keyed to the fate and transport model to be employed
to evaluate the data. The procedure also would be independent of the properties
of the material under evaluation.
The questions we would like you to address are:
• What would be the "best" approach to use in order to predict the nature
and concentration of the components that would leach from oily wastes if
the waste were to be placed in an unlined landfill environment?
• If the necessary tools are not presently available, how should such a test
method or model be developed and evaluated?
Praenitd on Jufy 14, 1992 at EPA Woriahop n on PrttUaini
fa Environmental Impact of Oify Materials' B~5
-------
What form should a cooperative development program take? How could
it be organized? Who might the cooperators be? How long would you
expect it to take to develop the necessary tools?
If you think of any ideas, information, or suggestions that you feel the Agency should
consider when addressing this issue, please send them to us. Send your comments to:
David Friedman
US Environmental Protection Agency
401 M St. SW (RD-680)
Washington, DC 20460
We will need to receive your comments by August 21, 1992 in order for them to be incorporated
into the conference final report.
Presented on Jufy 14, 1992 at EPA Wortohop D on Predicting
the Environmental Impact of Oify Materials' B~6
-------
PREDICTING THE ENVIRONMENTAL IMPACT OF OILY WASTE:
INDUSTRY PERSPECTIVE
STATEMENT OF ISSUE
Clifford T. Narquis
BP Research
4440 Warrensville Center Road
Cleveland, OH 44128-2837
Managing solid waste in an environmentally sound manner is a subject of high concern
to industry, EPA and the public. However, we must have regulatory tools which accurately
reflect the environmental hazard, and analysis tools which accurately assess the potential impact
of various management approaches. We need to bring the best science to bear on the evaluation
of the potential environmental threat from oily waste disposal, considering the likely management
scenarios. One way EPA has decided to regulate certain oily waste in the past is by listing the
waste as hazardous under RCRA. This approach identifies a material as an environmental threat
based on possible (but not necessarily realistic) mismanagement scenarios. A second way to
regulate the waste is to determine if it is hazardous using the toxicity characteristic (TC) and the
Toxicity Characteristics Leaching Procedure (TCLP) test. This test is used to determine whether
a waste is hazardous or not based upon a specific leachability test and municipal landfill disposal
scenario. The options explored in this paper serve to promote thinking about new approaches
for identifying the environmental threats and thereby to better focus regulations dealing with the
management of these materials. This will be done by introducing and critiquing current most
common predictive methods and presenting potential avenues for more accurate methods.
Although it is nearly impossible to precisely define the term "oily waste", the following
analysis can provide a basis for further discussion:
Presented at July 14, 1992 at EPA Workshop U on Predicting
&t Environmental Impact of Oily Uaurials' a-I
-------
a) An oil is generally an immiscible or relatively insoluble liquid, varying in
composition but consisting of organic constituents. Petroleum oils principally
consist of hydrocarbons; vegetable and animal oils are glycerides, and fatty
acids; and essential oils are terpenes, alkaloids, etc.
b) An oily waste is an industrial process waste or residual bearing oil in visual
and/or measurable proportions.
c) Oil in oily wastes can occur in any matrix, including: sorbed to dry solids; in
sludges or slurries; multi-phase liquids or sludges/slurries with multi-phase
liquids, if water is present. Proper treatment and disposal of all such matrices is
a concern of the petroleum industry.
d) Analysis of oils in oily wastes can be accomplished by techniques such as Total
Petroleum Hydrocarbons (TPH) (not constituent-specific) or TCLP (constituent-
specific). In a number of contexts, the procedures of methods such as these
serve to define what is meant by "oil" and "oily waste."
e) Oily wastes possess a wide variety of compositions and physical and lexicological
properties.
Some examples of oily waste include petroleum refinery sludges, such as oil- water
separator sludge and dissolved air floatation froth, storage tank bottom sludge, used oil and
others. Expanded beyond the petroleum community there are many types of oily wastes (POTW
sludges, polymer plants, timber processing, iron and steel, pulp and paper, meat packing,
slaughterhouse, leather tanning, coil coating,restaurants, and miscellaneous foods including meat,
dairy and vegetable based oils and fats, etc.).
Currently there are a variety of state and local programs designed to address potential
environmental impacts of the management of various types of oily materials, such as E&P
wastes, spill residues and UST wastes. A number of RCRA listed and toxicity characteristic
wastes are also regulated under federal programs. EPA is currently evaluating the possible
listing of additional petroleum refining wastes.
Presented on Jufy 14, 1992 at EPA Workshop Q on Prtdiaoif
At Environmaaai Impact of Oily Materials'
-------
Unfortunately, the current analytical methods for determination of the environmental
threat of petroleum constituents in wastes and oily materials via the TCLP test and model and
RCRA listing system remain controversial. USEPA, academia, and the regulated community
are continuing efforts to identify a sound, reproducible methodology to accurately assess these
threats. In fact, EPA has recently proposed a rule to address the over-regulation of listed wastes
created by EPA's "mixture" and "derived-from" rules. This initiative, called the Hazardous
Waste Identification Rule, could have major impacts on the classifications and management of
hazardous and nonhazardous industrial wastes, including oily waste.
An approach that the American Petroleum Institute (API) has suggested to the Agency
is concentration-based exclusion criterion coupled with contingent management. It is a two-
tiered process for determining whether wastes captured under the RCRA listing rule should or
should not continue to be regulated as hazardous. One tier would allow wastes with constituents
below health based levels (with an appropriate multiplier to account for dilution and attenuation)
to be deemed nonhazardous provided the waste does not exhibit a RCRA characteristic. A
second tier would allow wastes that contain constituents below somewhat higher health based
levels to be deemed as nonhazardous, provided these wastes are managed in certain
environmentally protective ways. This second approach would alter the current system by
basing a waste's classification on how it is actually managed and not how it could be
hypothetically mismanaged.
On the test method side, many, including Environment Canada, ASTM and the USEPA,
have been involved with the development of improved leachability and contaminant fate/transport
tests and models. Despite this work, predicting the potential environmental hazard associated
with oily wastes remains problematic.
Praaatd on July 14, 1992 at EPA Woriahcp U on Pre&ctni
fa Enviratmoaal Impact of Oify Materials' B~9
-------
TCLP APPROACH
The toxicity characteristic leaching procedure (TCLP) was developed as a way to evaluate
the threat of solid waste disposal under "...a mismanagement scenario for toxic wastes which
constitutes a prevalent form of improper management—namely, the co-disposal of toxic wastes
in an actively decomposing municipal landfill which overlies a groundwater aquifer..." (Fed.
Reg., May 8, 1990). The TCLP is a leaching and acidic aqueous extraction test. The test was
designed to model mismanagement of the disposal of process wastes. The Toxicity
Characteristic (TC) rule itself, and constituent-specific limits associated with the TC, define
wastes as hazardous on the basis of the concentrations of certain toxic constituents. The TC and
its constituent-specific limits were developed in large part to protect human health from
contamination of drinking water aquifers. The TCLP test, as currently interpreted, is applicable
to those wastes which produce a separate non-aqueous phase as well as those which do not.
Specifically the model system that forms the basis for the regulatory limits imposed by
the current toxicity characteristic is one which assumes that the waste is disposed of in a
municipal hazardous waste landfill where it is leached by acidic landfill liquids, emerges from
the landfill bottom into underlying groundwater whereupon it migrates to an hydraulically down-
gradient drinking water well (see Figure 1).
The current Toxicity Characteristic defined-method for determining environmental risks
uses a three-part system, consisting of a physical model (TCLP), coupled to a mathematical
model (EPACML), coupled to a lexicological model. The TCLP simulates constituent leaching
from a landfill, EPACML simulates constituent transport from a landfill to a drinking water
well, and the lexicological model relates drinking water concentration to health-effects. The
TCLP was not designed as, and fails as, a multi-phase model. This is because the oil phase is
simply treated as water. Additionally any multi-phase capability within EPACML was ignored
due to the TCLP output.
Presented on July 14, 1992 at EPA Workshop n an Predicting
the Environmental Impact ofOify Materials' B-10
-------
"Tl
o
o
s
in
ff
Groundujatar
-------
Significant technical aspects of the TCLP simulation can be summarized as follows (see
Figure 2).
• No vadose zone-bottom of the landfill is in direct contact with the ground water.
• The disposal of waste liquids (oil and water) are equally mobile.
• The liquids are not leached or diluted but elute directly from the landfill into the
ground water.
• Infinite source-liquids continue to be released forever regardless of amount of
liquids in the original waste.
• The solids are leached with a 20:1 volume of acidic "landfill leachate" which then
enters the ground water.
• Infinite source - the hazardous constituent concentration in the initial 20:1
leachate volume continues to be leached from the material forever, regardless of
mass of constituents in the original waste.
• The liquids and leachate travel through the ground water to a drinking water well.
Attenuation and dilution reduce concentrations by a factor of 100.
• Oil moves as water.
• Hazardous constituent concentrations achieve steady-state in the well at which
time the well-owner drinks two-liters/day for 70 years (oil and all).
VALIDITY OF THE TCLP APPROACH
Oily wastes provide a great challenge to those charged with evaluating their potential
impact on the environment. Unfortunately the design of the TCLP test in concept, methodology
and fate/transport modeling, inaccurately predicts the behavior of waste containing separate-
phase oil and organic constituents. One shortcoming is that it forces a generic disposal scenario
which may be reasonable for some cases but impossible for others. Specific problems include
operational problems with the test procedure, the assumption that oil behaves identically to water
Presented en Jufy 14, 1992 at EPA 'Workshop n on Predicting
the Eaviraunauol Impact qf Oily Materials' o~\.Z
-------
"ri
»»4«
(TO
§
KJ
1/1
CO
n>
ff H
s n
Q co
Is
31
o
Landfill
1. Do vadosa zone
2. Hquaous and non-aquBous
liquids treated identically
3. Liquids not loachod or diluted
prior to entoring groundinator
9. Individual
drinks 2 L/dag
for 70 years
Waste
liquids
oil G
ID a tor
6. Infmlta sourcs
of loachate
. Infinite source
of masto liquids
5. Leachate enters
groundroator directlg
7. Straight dilution 6 //*'
attenuation by factor
of lOQ
0. Oil moves as tnater
-------
in the environment, the validity of the disposal scenario, and invalid contaminant fate/transport
assumptions.
The test system was not designed for multi-liquid phase materials. This results in
operational problems with the TCLP methodology including non-reproducible free oil
breakthrough, filter clogging, and difficulties with volatiles equipment. The zero headspace
extractor (ZHE) test equipment is difficult to clean. Some volatile chlorinated compounds are
transformed within the TCLP extraction (Bricka, et al, 1991). EPA has, to date, not provided
approved test methods which are validated for the analysis of metals in non-aqueous liquids (55
Fed.Reg. 4444).
One of the initial steps in the TCLP test is pressure filtration of the waste. For some oily
wastes, non-aqueous liquid may be expressed. This liquid is segregated from the remaining
solids which are then acid-leached. This acid leachate is combined with the non-aqueous liquid
to produce the "TCLP leachate" which is compared to hazardous/nonhazardous criteria.
Implicit in this procedure is the assumption that both aqueous and non-aqueous liquids
will behave identically, both within the landfill and upon their hypothetical release. EPA has
been able to provide little, if any, support for this critical portion of the TCLP.
"The initial liquid/solid separation problems are due to the tendency for some material, such
as certain types of oily wastes, to clog the 0.45 um filter and prevent filtration This
problem is serious, since materials which do not pass the 0.45 um filter are treated as solids
even if they physically appear to be a liquid. These (liquid) wastes are then carried through EP
extraction as a solid."
"This is particularly serious for oily wastes, since oils have been known to frequently
migrate to ground waters. It is important for the liquid (sic)/solid separation to treat, as liquids,
those materials which can behave as liquids in the environment."
Presented on Jufy 14, 1992 at EPA Workshop U an PretHctoig
fa Environmental Impact of Oity Materials' B-14
-------
"As indicated below, EPA believes that the liquid/solid separation technique.... reduces
variability....and that it also provides a more adequate differentiation between those materials
that behave as liquids in the environment, and those materials which behave as solids. "(51 Fed.
Reg. 21658)
As we gain experience with risk evaluations, we see that the risk posed by light, non-
aqueous phase liquids (i.e. "oil") appears to be mostly due to dissolved contaminants in drinking
water. The calculated risk due to free oil is not great due to the lack of exposure. As it moves
through the soil, oil will be immobilized in the soil and from that point may partition into the
water phase according to constituent solubilities. Any mobile oil migrating to a water well does
not represent a 2L/day, 70 year hazard since it is not realistic to project that anyone will drink
free-phase hydrocarbons daily for their entire lives. Therefore, it makes some technical sense
to leach the oil fraction with the acidic medium along with the solids.
The disposal scenario as depicted by EPA is not an accurate description of current waste
disposal practices. An EPA-OSW survey, several years old already, documents that liquid-type
wastes are no longer being accepted by municipal landfills (51 Fed. Reg. 21655). New test
methods based upon actual waste management situations would give more accurate results than
those based on generic hypothetical scenarios.
Industry has commented upon the shortcomings of the current TCLP/CML model. For
example, the infinite source assumptions require contaminant mass to continue to be available
for introduction into ground water until steady state is achieved. This is unrealistic. One
improvement would be to design transient, declining source terms into the model. Further, there
is no consideration of a vadose zone although we know it exists and future landfill regulations
will require the presence of a vadose zone. The TCLP is not designed to handle the separate
organic phase flow. The current TCLP system does not take into account aerobic
biodegradation, volatilization, or retardation. Hydrolysis is apparently being considered at this
juncture, but is not currently part of this system.
Praauai cnjufy 14, 1992 at EPA Workshop U on Pratiaoig
fa Environmental Impact of Oiiy Ualehais' B-15
-------
The unilateral application of TCLP to multi-phase wastes, especially those containing oily
materials, is unsupported and inappropriate. There is no evidence that non-aqueous liquids
behave as aqueous liquids in a landfill. Indeed, such liquids have an affinity for the solid
materials in the landfill which could cause contaminants to be less mobile than predicted by the
TCLP.
Until work on the behavior of non-aqueous materials and the prediction of their
movement is more mature, the non-aqueous liquids should be treated like the waste itself and
be subjected to the same extraction with acidic fluid. To the extent that hazardous constituents
are released into the extractant, they should be combined with the aqueous extract generated
from the waste solids.
THE EFFECTS OF THE CURRENT TC SCENARIO MODEL - SOME IMPACTS
A number of wastes from the petroleum industry, such as waters from tank drawdowns,
ground-water extraction, and hydrotesting, are or may be subject to the TC rule even though
there is no conceivable way that these materials would ever find their way into a landfill.
The RCRA Corrective Action Program could potentially generate large quantities of
petroleum-contaminated soils. On-site and in-situ management techniques are not accurately
represented by the TCLP. In addition, a number of states currently have effective response
programs for clean up of spills and other releases of petroleum into the environment. States are
concerned that application of the TCLP (particularly if TCLP is a poor estimator of
environmental threat) will seriously impact operation and effectiveness of these programs by
adding unnecessary and unwarranted hazardous waste handling requirements to wastes which
don't pose a threat. We can no longer afford to waste large sums of money handling solid waste
in a manner which over estimates the actual environmental threat.
Presented on Jufy 14, 1992 at EPA Workshop a an Predicting
the Environmental Impact of Oify Materials' B-16
-------
A WAY FORWARD - SAB LEACHABILITY SUBCOMMITTEE
Given the problems with the applicability of the TCLP to multi-phase waste, are there
any alternatives? What are the potential ways forward?
Last year, a report was issued by the Science Advisory Board (Environmental
Engineering Committee, Leachability Subcommittee), entitled "Recommendations and Rationale
for Analysis for Contaminant Release." It contained nine recommendations:
• A variety of contaminant release tests and test conditions which in corporate
adequate understanding of the important parameters that affect leaching should be
developed and used to assess the potential lease of contaminants from sources of
concern.
• Prior to developing or applying any leaching tests or models, the controlling
mechanisms must be defined and understood.
• A consistent, repeatable and easily applied, physical, hydrologic and geochemical
representation should be developed for the waste management scenario of
concern.
• Leach tests and conditions (stresses) appropriate to the situations being evaluated
should be used for assessing long-term contaminant release potential.
• Laboratory leach tests should be field-validated, and release test accuracy and
precision established before tests are broadly applied.
• More and improved leaching models should be developed and used to complement
laboratory tests.
• To facilitate the evaluation of risk implications of environmental releases, the
Agency should coordinate the development of leach tests and the development of
models in which release terms are used.
• The Agency should establish an inter-office, inter-disciplinary task group,
including ORD to help implement these recommendations and devise an Agency-
wide protocol for evaluating release scenarios, tests, procedures, and their
applications.
Presented on Jufy 14, 1992 at EPA Woriahcp 77 on Predicting
^ Environmaual Impact cf Oily Materials' B~17
-------
The task group should also be charged with recommending what the appropriate
focal point(s), responsibilities, and organizational, budgetary and communication
links should be within the Agency for the most effective, continued and ongoing
support and pursuit of research, development, and utilization of methods and
procedures.
To fully accomplish all of the recommendations will be costly and time- consuming.
However there can be no alternate to core research on contaminant release and transport
methods. SAB identified approximately 30 leach tests which are used internationally to attempt
to evaluate environmental threat of wastes. Unfortunately SAB concludes that each method
suffers from shortcomings.
APPLICATION TO OILY WASTE
If we wish to improve upon the system, there are two basic options:
1. Stick with the physical/mathematical model basis of the TCLP and improve
accuracy by modeling multi-phase transport and remove assumptions predicated
on long-term human consumption of immiscible product.
2. Replace with a single alternative model, either physical or mathematical.
We understand that EPA is exploring various enhanced modeling systems for multi-phase
disposal scenarios. These would include multi-phase flow within the unsaturated zone, and
partitioning between aqueous, oil, and air phases within the soil. Also included would be
saturated zone groundwater pollutant transport models which are more accurate. Industry favors
these developments as tools to better understand oily waste disposal impact.
On the other hand, we must currently deal with an inappropriate TC rule. Currently,
industry must comply with the TC rule, which means it must run TCLP tests on oily wastes.
Presented on July 14, 1992 at EPA Workshop II on Predicting
the Environmental Ijnpact ofOify Materials' B-lo
-------
This has resulted in a disastrous situation. The TCLP was not designed to accurately assess the
environmental threat of oily materials. Therefore it does not. However, decisions on the
"proper" management of these wastes are being made on the basis of a flawed test.
Industry has had to deal with the TC for many years. We have modified our waste
management approaches and strategies, we have complied with TC and land disposal
requirements and we are preparing to fully comply with Corrective Action. Unfortunately,
changes to the TC at this point may be just as disruptive and costly as compliance with the TC
has been to date. Modifications must be done carefully and deliberately, always using the best
possible science to ensure accuracy, not just consistency.
API supports, as mentioned in the Statement of Issue section, a concentration- based
exclusion coupled with contingent management for exempting listed hazardous waste from
subtitle C requirements. This would address the "inappropriate scenario" dilemma by
incorporating elements of actual management approaches instead of one hypothetical approach.
The regulated community has volunteered to work with EPA both as individuals,
individual companies, and through trade organizations. We will continue to offer such
assistance. For myself, I see continued interaction between EPA- OSW and the American
Petroleum Institute. Typical industrial support to EPA includes offering technical comment,
procuring wastes, providing waste generation and characterization data, and participating in
round-robin testing of new methods.
Praaaed on July 14, 1992 at EPA Workshop U on PraHaing
fa Environmoaal Impact of Oily Materials' B~19
-------
TABLE 1 - EXTRACTION TESTS
Test Method
Leaching Fluid
Liquid: Solid Ratio
Maximum
Particle Size
Number of
Extractions
Time of
Extractions
I. STATIC TESTS (LEACHING FLUID NOT RENEWED)
A. Agitated Extraction Tests
TCLP(lSll)
EPTOX(1310)
ASTM DM 3987-85
California Wet
Leachate Extraction Procedure
(MOE, Ontario)
Quebec R.S.Q. (MOE,
Quebec)
French Leach Test (AFMOR,
France)
Equilibrium Extraction
(Environment Canada)
Acetic Acid
0. 1 N Acetic Acid Solution,
pH 2.9, For Alkaline Wastes
0. 1 M Sodium Acetate Buffer
Solution, pH 5.0, For Non-
Alkaline Wastes
0.5 N Acetic Acid
(pH = 5.0)
ASTM Type IV Reagent Water
0.2 M Sodium Citrate
(pH = 5.0)
Acetic Acid
2 MEQ/G
Inorganic 0.02 MEQ/G
Organic Distilled Water
DI Water
Distilled Water
20:1
16: 1 during extraction
20:1 final dilution
20:1
10:1
20:1
10:1
10:1
4:1
9.5mm
9.5 mm
As in environment
2.0 mm
As in environment
Ground
9.5 mm
Ground
1
1
1
1
1
1
1
1
18 hours
24 hours
18 hours
48 hours
24 hours
24 hours
16 hours
7 days
-------
TABLE 1 (continued)
Test Method
Multiple Batch Leaching
Procedure (Environment
Canada)
Material Characterization
Centre-4 (Material
Characterization Centre)
Oily Waste (1330)
Synthetic Precipitation
Leaching Procedure (1312)
Equilibrium Leach Test
Leaching Fluid
Acetic Acid Buffer, pH 4.5
Choice
Soxlet with THF and Toluene
EP on Remaining Solids
Variable
Distilled Water
Liquid: Solid Ratio
4:1 or 2:1
10:1
100 g:300 ml
20:1
20:1
4:1
Maximum
Particle Size
9.5mm
2 fractions
74-149 mm
9.5 mm
9.5 mm
150 urn
Number of
Extractions
Variable
1
150-425 mm
3
1
1
Time of
Extractions
24 hours
20 days to
10 years
24 hours (EP)
18 hours
7 days
B. Non-Agitated Extraction Tests
Static Leach Test Method
(Material Characteristic
Centre- 1)
High Temperature Static
Leach Test Method (Material
Characterization Centre-2)
Can be Site-Specific
Same as above but at 100° C
VoI/Surface 10 /im
Vol/Surface 10 /xm
r\
40 mm surface area
40 mm surface area
1
1
>7 days
>7 days
C. Sequential Chemical Extraction Tests
Sequential Extraction Tests
0.04 m Acetic Acid
50:1
9.5 mm
15
24 hours per
extraction
D. Concentration Build-Up Test
Sequential Chemical
Extraction
Five Leaching Solutions
Increasing Acidity
Various from 16:1 to
40:1
150 /im
5
Varies from 2
to 24 hours
-------
TABLE 1 (continued)
Test Method
Standard Leach Test,
Procedure C (University of
Wisconsin)
Leaching Fluid
DI Water
Syn Landfill Leacliate
Liquid: Solid Ratio
10:1,5:1
7.5:1
Maximum
Particle Size
As in environment
Number of
Extractions
3
Time of
Extractions
3 or 14 days
II. DYNAMIC TESTS (LEACHING FLUID RENEWED)
A. Serial Batch (Particle)
Multiple Extraction Procedure
(1320)
MWEP (Monofill Waste
Extraction Procedure)
Graded Serial Batch (U.S.
Army)
Sequential Batch
ASTM D4793-88
Waste Research Unit Leach
Test (Harwell Laboratory,
UK)
Standard Leaching Test:
Cascade Test
SOSUV, Netherlands
Same as EP TOX, then with
Synthetic Acid Rain (Sulftiric
Acid; Nitric Acid in 60:40%
Mixture)
Distilled/Deionized Water or
other for
Distilled Water
Type IV Reagent Water
Acetic Acid Bufferd pH = 5
Distilled Water
HN03 pH 4.0
20:1
10: 1 per extraction
specific site
Increases from 2: 1 to
96:1
20:1
1 bed vol 5 el ut ions
10 bed vol >6 elutions
20:1
9.5 mm
9.66 or monolith
N/A
As in environment
Crushing
Crushing
9 (or more)
4
>7
10
>11
5
B. Flow Around Tests
IAEA Dynamic Leach Test
(International Atomic Energy
Agency)
DI Water/Site Water
N/A
One face prepared
>19
24 hours per
extraction
18 hours per
extraction
Until steady
state
18 hours
2 to 80 hours
23 hours
>6 months
-------
TABLE 1 (continued)
Test Method
ISO Leach Test (International
Standards Organization)
ANSI/ANS 16:1 (American
National Standard Institute/
American Nuclear Society)
DLT
Leaching Fluid
DI Water/Site Water
DI Water
DI Water
Liquid: Solid Ratio
N/A
N/A
N/A
Maximum
Particle Size
Surface polishing
Surface washing
Surface washing
Number of
Extractions
>10
11
18
Time of
Extractions
> 100 days
90 days
196 days
C. Flow Through Tests
Standard Leaching Test:
Column Test (SOSUV, The
Netherlands)
Column ASTM D4874-89
DI Water
HN03 PH = 4
Type IV Reagent Water
10:1
One void volume
As in environment
As in environment
7
1
20 days
24 hours
III. OTHER TESTS
MCC-55 Soxhlet Test
(Material Characteristic
Center)
Acid Neutralization Capacity
Dl/Site Water
HNO, Solutions of Increasing
Strength
100:1
3:1
Cut and washed
150 jim
1
1
0.2 ml/min
48 hours per
extraction
References:
1. Compendium of Waste Leading Tests. Waste Water Technology Centre, Environment Canada, Final Draft, May 27, 1989.
2. Private discussions with Gail Hansen, Office of Solid Waste, U.S. EPA.
-------
PREDICTING THE ENVIRONMENTAL IMPACT OF OILY MATERIALS:
SCIENTIFIC PERSPECTIVE
Larry P. Jackson
Ted Varouxis
Associated Design and Manufacturing Company
Alexandria, VA
INTRODUCTION
This paper is intended to stimulate discussion into new or better ways to evaluate the
potential release of regulated substances from oily wastes. The paper discusses some options
to the currently approved procedures to determine the concentrations of regulated organic
chemicals released into the groundwater regime from improperly managed oily waste. The paper
also describes a proposed method to evaluate the fraction of an oily waste which is flowable
under the influence of gravity or overburden pressure if the material is improperly disposed in
a landfill. The options presented cover, in part, some of the major technical concerns of the
Environmental Engineering Committee of the Environmental Protection Agency's (EPA) Science
Advisory Board (SAB) in their October, 1991 recommendations to the EPA Administrator.
This paper is prepared from the perspective that accurate, reliable, and cost-effective
analytical procedures can be developed to properly characterize and manage potentially
hazardous oily wastes. The paper accepts the premise that the regulatory community must
proceed carefully and the "worst case scenario" will be considered in any proposed solutions.
The paper seeks to incorporate some of the suggestions of the EPA Science Advisory Board that
methods should take into consideration real world factors such as:
Presented on Jufy 14, 1992 ai EPA Workshop U on Predicting
the Environmental Impact of Oity HOerials' B-24
-------
source matrix properties,
contaminant properties,
leachant properties,
fluid dynamics,
chemical and physical properties of the waste,
temporal/spatial dependence,
measurement methods, and
physical models.
TECHNICAL ISSUES
Neither the regulatory nor the regulated community has successfully proposed methods
to properly characterize the potential for environmental impact for oily wastes. Existing leaching
tests are known to be technically and mechanically deficient, and no method exists to measure
the amount of flowable, oily material which may be released from a waste. Solutions to these
problems have not been discussed to any extent in the published literature nor in the proceedings
of symposia and workshops. These issues are recognized as the major unaddressed problems in
evaluating the pollution potential of oily wastes. Any scenario which proposes to assess the
pollution potential of this class of wastes must address these issues. This section describes the
current state of the technology in these areas.
The Oily Waste Extraction Procedure (OWEP, EPA Method 1330A)2 is designed to
evaluate the potential for an oily waste to release metals under aqueous leaching conditions.
OWEP separates the solid material from the oil by solvent extraction. The solid phase is then
leached by Method 1310A, Extraction Procedure Toxicity Test2 and the extracted oil analyzed
directly for the metals of interest. The results of the analyses of the two fractions are combined
mathematically. It is generally conceded that this overestimates the leaching potential of the
waste. If the method is applied to the analysis of regulated organic constituents, all of the analyte
will be deemed leachable which is incorrect. It should be noted that the OWEP has never been
suggested as appropriate for organic constituents.
Presented on July 14, 1992 at EPA Workshop U on PretHcang
fe Environmental Impact of Oily Materials' B-25
-------
The current approach for analyzing the leaching potential of solid waste, EPA Method
1311, Toxicity Characteristic Leaching Procedure (TCLP)2 differs from the OWEP in that TCLP
attempts to determine the aqueous leachability of the waste for both inorganic and organic
constituents in a single leach test. It is very difficult to conduct in a reproducible manner.
Mechanical problems with the test make it time consuming to perform and frequent reanalysis
is required. Precision between replicate tests is very poor. Equipment cleanup is a major
obstacle to laboratory productivity. Costs can run to several thousand dollars per sample for
difficult-to-handle samples. The major problems found in conducting the TCLP are:
• Handling of the sample is messy, effecting weighing of proper amounts into the
extraction vessels. Loss of volatiles occurs.
• Proper sub-sampling of multi-phasic materials is difficult. Samples frequently
contain oil, water, and solids. Isolation of solids for extraction is arduous.
• The tumbling action of the two liter extraction vessels forms emulsions making
isolation of the aqueous leachate difficult.
• Separation of the leachate from the solid residue after extraction is frequently
impossible because the oily material clogs the filter. This is especially serious
when using the zero headspace extractor (ZHE) since the test must be repeated
if this happens.
• Oily wastes frequently yield aqueous leachates and free organic material which
must be separated and analyzed separately, doubling or tripling the analytical
costs.
• Equipment cleanup is very time consuming, minor amounts of residual organic
material can carry over and contaminate succeeding samples.
These problems are sufficiently severe that both regulators and regulated community
have lost confidence in the utility of the method to estimate the potential environmental hazard
of oily wastes.
Presented on Jufy 14, 1992 at EPA Workshop fl on Predicting
fa Environmental Impact of Oify Materials' B-26
-------
PROPOSED APPROACHES
This section discusses four proposed approaches to improving the technical and/or
procedural methods for determining the potential environmental impact of oily materials. They
are:
• Adopt a flowable materials test
• Modify the TCLP.
• Adopt a new method of contacting the leach medium with the waste.
• Devise a new model for determining the amount of a regulated substance released
from an oily waste by aqueous leaching mechanisms.
Approach 1 - Flowable Materials Test
The EPA has laid the groundwork for a Flowable Materials Test (FMT) in the 1991
proposed rule making for the Liquid Release Test (LRT), EPA Method 90963. The Agency has
published two reports describing the test for its original application, namely to detect the release
of any free liquid from material destined for land disposal4'5. The test places a 76mm
diameter by 10mm high sample in a confined chamber under a 50 psi load for ten minutes to
force the release of free liquids.
If the device is modified to provide an tight fitting piston/barrel arrangement (identical
to the design of the zero headspace extractor, the ZHE) and the indicator paper holder is
replaced with a reinforced screen and fluid collection vessel, it will be capable of applying the
necessary degree of pressure to the sample necessary to simulate overburden pressure. The
screen will allow for the effective escape and collection of the flowable material from the solid
mass. Both the retained solids and the collected flowable material can be analyzed separately
using one or more of the suggested experimental changes described in the following sections.
Presented on July 14, 1992 at EPA Workshop D at Prettirtinx
die Environmental Impact of Oify Materials' B-27
-------
Approach 2 - Modify the TCLP
Modification 1 - Addition of Inert Substrate.
One basic problem with the ability to conduct TCLP is the physical nature of oily
material and the impact that has on the conduct of the test as discussed above. The EPA has
addressed this type of problem in Methods 3540 and 35502, Soxhlet Extraction and Sonication
Extraction respectively, where inert adsorbents are added to the waste to provide a free flowing
material with sufficient permeability to allow for efficient extraction. The same approach can
be taken with the TCLP.
The addition of a high surface area, inert matrix lite silica beads (or sand) will
effectively immobilize the free phase organic material and provide a free flowing medium for
sample preparation (sub-sampling) and extraction. The increased surface area will promote
solubilization of the organic components into the extraction medium. This approach will be
effective for both free liquids and oily solids. If an aqueous phase is also present, adsorption of
the oily material should facilitate separation of the aqueous material prior to extraction. The
presence of the substrate surface as a host site for oily material will minimize the formation of
emulsions during tumbling of the waste/leachant mixture, provided the viscosity of the organic
material is sufficiently high that the shear forces of the tumbling action do not separate the liquid
material from the solid. After the tumbling sequence, the solid substrate and absorbed oily
material will settle to the bottom of the leaching vessel and eliminate or minimize the amount
of free organic liquid floating at the surface of the solution, making the filtration step much
easier and clogging less likely.
This type of sorbent bed closely resembles the real world case of oily material spilled
onto or migrating through soil columns until it no longer moves under the force of gravity. This
model of oil coated soil represents the most common real world source of potential pollutant
release from oily wastes.
Presented en July 14,1992 at EPA Wortihcp U on Preehcang
the Environmental Impact of Oily Materials' B-2o
-------
Modification 2 - Use of Fritted Stainless Steel Filter.
Regardless of whether or not the method is modified by the addition of an inert substrate
to immobilize the oily material, the filtration step of the method can be improved. Agency
funded research of improved filtration media led to the development of a sintered stainless steel
filter that overcame many of the clogging problems". The modification has not been added to
the method at this time, but it has been used without formal regulatory adoption, with some
intractable wastes. Other approaches to improve filtering should be examined, such as the use
of thick pads (several mms) of non-woven glass or plastic fibers and powdered filter aids as pre-
filters. These are physical changes to the filter apparatus; they should be permitted as long as
the modifications can be shown not to alter the composition of the filtrate by absorption of
analytes or allow for the passage of particles with a nominal size greater than 0.7 micron.
Approach 3 - Adopt a New Leaching Technique
The mechanical forces that act on oily waste during the TCLP tend to separate the oil
from the substrate that was part of the original waste material or the inert material added in
Approach 2. This leads to the formation of emulsions and/or free liquid phases which coat and
clog the filters during the filtration step. Column leaching configurations that are less physically
aggressive than tumbling can be used as the leaching model for oily wastes. The permeability
of the waste material is the key property of the waste which must be controlled if a column
technique is to work (permeability also impacts the efficiency of any extraction process). Use
of inert sorbents, as in Approach 2, can provide the necessary permeability to allow for uniform
flow of the aqueous medium through the waste bed and promote effective leaching. Uniform
flow would be provided by pumping the leachant through the system.
Flow rates can be adjusted to minimize the shear forces which might dislodge oily
material from the waste. Flow direction can be changed based on the density of the organic
fluids. Downward flow for materials lighter than water and upward flow for materials heavier
Presented on July 14, 1992 at EPA Wortshcp U on Predicting
fa Environmental Impact of Oily Uaierials' B~29
-------
than water will minimize the likelihood that oily material will separate from the substrate during
testing. If the fluids do separate, they will not find their way into the leachate reservoir without
passing through the substrate bed where they will re-deposit on the surface. Minimum flow
volumes per unit mass of waste will become the operational control of the test rather than the
tumbling time that is now used.
The dynamic flow conditions of the column test also allow for efficiencies in subsequent
sample analysis. Modern solid phase sorbents for both organic and inorganic analytes can be
placed between the pump outlet and the head of the column to collect and pre-concentrate the
analytes for future analysis. The use of these sorbents also allows the introduction of "fresh
leachant" to the top of the column of waste in a manner similar to the way fresh ground or
surface water would contact the waste in the real world scenario. This would promote maximum
release of the target analytes.
The column leach model proposed here resembles the real world case where ground or
surface water percolates through oily material that adheres to the soil, more closely than does
the tumbling action of the TCLP. The column leach approach can be extended to evaluate the
attenuation of solubilized materials by a representative soil, by placing a soil layer in the same
extraction column as the waste or by passing the leachate through a second column placed in
series with the waste containing column. This allows for the development of a modular test
sequence in which the same test used to characterize a waste leachate is used as the source term
for attenuation studies, which may be conducted as part of a site-specific risk assessment. This
strategy is in keeping with the SAB's recommendation to the EPA.
Approach 4 - Use a Totally Different Leaching Model
The current TCLP and the two options previously discussed are alternative physical
models of the leaching process. Most problems resulting from these approaches center around
sample handling, leaching, and filtering the leachate. To avoid some of these problems, the
Prcsaaat on July 14, 1992 at EPA Worktop U on PrexUfOng
die Environmental Impact of Oify Materials' B-30
-------
Agency should consider using well-developed, existing theoretical and experimental models of
the leaching of materials from oily matrices as well and the migration and interaction of the
soluble components with soils. Both the EPA and the American Petroleum Institute (API) have
published significant papers on the approach7'8'9'10'11. There is ample experimental
evidence that these models are a good first order approximation of the amounts material actually
found from aqueous leaching of oily materials. They apply to both oily solids and flowable oily
materials. The models depend on the amount of the target analyte present in the waste, the
analytes physical/chemical properties, and the chemical properties of the soil. Some of the more
important features of these models are discussed below.
The American Petroleum Institute (API) published a review of historical data relating fuel
composition to the aqueous solubility of its various components . The review investigated the
relationship between the solubility of the pure hydrocarbon components in water and the amount
found in aqueous solutions that had been allowed to equilibrate with fuels (1:10 fuel/water ratio).
The study defined the partition coefficient for this process K^ by equation 1:
C/
(1)
where: Cf = concentration of the component in the fuel, g/L
Cw = concentration of the component in the water, g/L
This property is related to the solubility of the pure component in water, S, for a group
of six aromatic compounds by equation 2 which has a correlation coefficient of r = 0.99:
Log Kfo = - 0.884 log S + 0.975 (2)
Prtsailtd on Jufy 14. 1992 at EPA Workshop U an Predicting
&t Enviraonaaal Impart of Oily Materials' B-31
-------
As should be expected, the relationship between S and K^ is a function of the class of organic
compounds being considered (aromatic, aliphatic, olefinic, etc.). When six additional
compounds, one aromatic, two olefinic, and three aliphatic, were considered, the best fit
equation describing the relationship between S and K^ became
Log
= - 1.018 log S + 0.706
(3)
and the correlation coefficient, r1, dropped to 0.87. Table 1 compares the experimental data for
eleven of the compounds for which data was experimentally determined with the data derived
from equation 3. Most of the data compare favorably with the normal range of allowable
differences between replicate analytical determinations.
TABLE 1
Comparison of Observed and Estimated Hydrocarbon Concentrations
from a Standard Gasoline in Equilibrium with Water (1:10)
HYDROCARBON
Benzene
Toluene
2-Butene
2-Pentene
Ethylbenzene
o-Xylene
m-Xylene
Butane
1,2,4-Trimethylbenzene
2-Methylbutane
Pentane
CONCENTRATION
OBSERVED
58.7
33.4
2.4
2.4
4.3
6.9
11.0
2.7
1.1
3.7
1.0
ESTIMATED
58.8
37.8
3.2
1.7
3.2
4.7
9.2
5.2
1.8
6.2
2.2
Presented en July 14, 1992 al EPA Workhcp 11 on Predicting
the Environmental Impact (fOify Materials'
B-32
-------
This type of model works well for those cases where the oily waste matrix is the primary
determinate in the partition coefficient in the matrix/water distribution. The 1984 API report
discusses the situation where the amount of oil is very small compared to the total mass of
organic carbon in the soil/sediment/waste matrix and in effect represents oil absorbed on soil .
Equation 4 applies to these situations.
where: 1C = soil/water partition coefficient
K|JC = the organic carbon partition coefficient
foc = weight percent organic carbon in the substrate.
The organic carbon partition coefficient, KgC, is related to the octanol water partition
coefficient, Kow, by equation 5.
LogKoc = log KQW ~ 0.317 (5)
These models are based on the physical and chemical properties of the target analytes and
the substrates with which they are associated, be it a flowable liquid, solid waste, or soil.
Accurate and precise methods exist for experimentally determining the input variables to the
models. These variables include but are not limited to:
• analyte concentration in total waste,
• percent organic carbon in soil or waste matrix,
• partition coefficients (fuel/water, soil/water, octanol/water), and
• water solubility of target analytes.
Presented on July 14, 1992 at EPA Workshop H on Predicting
At Environmental Impact of Oily Materials' B-33
-------
As the data base is expanded, relationships among classes of organic compounds, water,
soils, and wastes will emerge. These relationships will lead to better empirical and theoretical
understanding of the physical and chemical factors controlling the release of materials to the
environment. New materials can be evaluated without detailed experimental studies by analogy
with similar compounds, wastes, and soils.
Use of this type of approach helps fulfill the SAB's recommendation that more rigorous
scientific procedures be used to determine the potential for release as well as environmental
impact. This approach also meets the recommendation that rugged tests that are less susceptible
to waste matrix effects be used. The approach also uses many of the same parameters used in
determining fate and transport and important measures of environmental risk; therefore, a more
unified model of environmental impact can be developed.
ROLE OF THE EPA AND PUBLIC SECTOR GROUPS
The EPA can serve as a catalyst for the necessary research studies for needed to improve
and develop reliable analytical methods. EPA also can lead in measuring the important physical
and chemical properties, of the analytes, wastes, and soils, that define analyte behavior in the
environment. Members of the public sector can contribute laboratory support and technical
expertise to developing the necessary methodology and demonstrating the applicability and
ruggedness of the methods.
These workshops are an ideal expression of how this should work.
Presented on July 14, 1992 at EPA Wortshop tt on Predicting
the Environmental Impact of Osfy Uaurials' B~34
-------
APPENDIX C:
CHARACTERIZING HETEROGENEOUS MATERIALS
-------
CHARACTERIZING HETEROGENEOUS MATERIALS:
INTRODUCTION
David Friedman
USEPA Office of Research and Development
401 M Street SW
Washington, DC 20460
BACKGROUND
Throughout the RCRA and CERCLA programs, one is faced with the task of
characterizing a material to determine if it possesses some property of concern or interest.
Types of questions one may be answering include:
• Is the material hazardous?
• Can it be safely or successfully managed using a specified treatment technique?
• How much supplemental fuel might I have to add to incinerate the waste?
PROBLEM
When the material being characterized is homogeneous, conventional sampling and
subdividing approaches can be employed. For example, take the case of a truckload of used
foundry sand whose average phenol concentration is of interest. The material is in the form of
relatively fine particles. Conventional compositing and subsampling will provide relatively small
representative samples whose average composition is very close to that of the whole load. The
actual task of obtaining each subsample from the appropriate point in the truck may be very
difficult, but the approach used is relatively straightforward.
Presented at EPA Workshop ffl July 16,1992
'Ouraaeraaig Heterogeneous MaUrials' C-1
-------
However, what happens when one is faced with a heterogeneous material? When we
speak of heterogeneity, we need to keep in mind that this is a relative term and is a function of
the objectives of the characterization and the analytical sample size. The concern of today's
workshop is generally with wastes of such various particle size, waste consistency or
extraordinary concentration gradients that the sampling and analytical objectives cannot be met
using traditional approaches and standard techniques. When it is not possible to obtain a
representative sample, what shall we do?
This is an important and difficult problem. It is one that has been of concern to many
organizations involved in waste management. For example, in March 1991, the Environmental
Protection Agency, the Department of Energy, and ASTM's Committee D-34 conducted a
workshop to examine how to characterize heterogeneous wastes that were also contaminated with
radioactive materials. A report has recently been issued that describes the results of the
workshop1. However, many problems and issues remain to be addressed and we have
convened today's workshop to address some of the issues.
We would like your help with addressing the following specific issues.
Identify characterization situations that present difficult problems (generically identify
both the type of material and situation).
• Develop specific guidance laying out the process to be used to adequately and
cost-effectively address the problem. Describe what criteria should be used to
guide development of solutions/approaches to specific problem situations. Please
be as quantitative as possible on when trade-offs may need to be made and how
to select among options.
• If research work needs to be done to develop a solution, please identify what
work needs to be done.
1 Characterizing Heterogeneous Wastes: Methods and Recommendations, EPA/600/R-92/0,
February 1992.
Present*! as EPA Wortshop mjuty 16, 1992
'Characterizing Heterogeneous Materials' \^ L
-------
Since EPA would like to solve monitoring problems through cooperative ventures,
what form should a cooperative development program take? How can it be
organized? Who might the cooperators be?
If you think of any ideas, information, or suggestions that you feel the Agency should
consider when addressing this issue, please send them to us. Send your comments to:
David Friedman (RD-680)
US Environmental Protection Agency
Washington, DC 20460
We will need to receive your comments by August 21,1992 in order for them to be incorporated
into the final conference report.
Presented at EPA Workshop m July 16, 1992
'Ouroaeridng Heterogeneous Materials' C~3
-------
CHARACTERIZING HETEROGENEOUS MATERIALS:
REGULATORY PERSPECTIVE
Charles A. Ramsey
USEPA National Enforcement Investigations Center
Box 25227 Building 53 Denver Federal Center
Denver, Colorado 80225
INTRODUCTION
Characterization is the process used to determine whether solid waste is also a hazardous
waste. Characterization is arrived at by following "testing structure." A testing structure is the
particular sampling, analysis, and data interpretation steps that are employed to measure a waste.
The purpose of this paper is to examine the testing structure used to measure characteristics
(ignitability, corrosiviry, reactivity, and toxicity) of heterogeneous solid waste that would make
it hazardous waste as defined in 40 CFR part 261.21-24.
To date, there has been little progress in developing a testing structure to characterize
heterogeneous waste. EPA regulations state that characterizing waste includes determining the
average property of the "universe or whole", 40 CFR part 260.10. The problems with the
regulations are that "average" lacks scientific validity and "universe or whole" is not defined.
These two items are an integral part of the testing structure.
To address these problems, two areas must be explored: 1) policy, regulations, and
guidance issues and 2) scientific concerns. This paper focuses on heterogeneous solid waste
because it is a worst case scenario. A testing structure that works for heterogeneous solid waste
will work for any solid waste.
Presented at EPA Workshop mJufy 16. 1992
'Characterizing Heterogeneous Materials' C~4
-------
POLICY, REGULATIONS, AND GUIDANCE ISSUES
Many problems with characterization of heterogeneous materials under RCRA can be
traced to the definition of what type of sample must be collected and analyzed. The required
sample is a "representative" sample which is defined in 40 CFR part 260.10. "representative
sample means a sample of a universe or whole (e.g., waste pile, lagoon, ground water) which
can be expected to exhibit the average properties of the universe or whole." At least two
problems arise with this definition: what is the average property and what is the universe or
whole?
Average
Does average signify the arithmetic average or the "most likely result" of the target
population? To illustrate, if the data points are 5.7, 5.7, 5.8, 5.9, and 0.2, the arithmetic
average is 4.7. If the regulatory limit is 5.0, the best point estimate of the average is that it is
below the regulatory limit. However, if confidence intervals were required, there would be little
confidence in this conclusion.
If average, however, signifies the "most likely result" (the point at which one half of the
values are above or below~the median) one would conclude that 80% of the data points are
above the regulatory limit. This conclusion is also an expression of a higher degree of
confidence than the conclusion with arithmetic average. More of the values agree with this
conclusion (4 of 5 are above the limit) than agreed with the conclusion based on arithmetic
average (1 of 5 were below the limit).
With the arithmetic average approach, it is possible to have well over half the population
above the regulatory limit and still achieve compliance. With the "most likely result" approach,
not more than one half of the population can be above the regulatory threshold and be in
compliance. The "most likely result" is sometimes referred to as "attribute" testing. An
Praemed at EPA Workshop W July 16, 1992
'Characterizing Hesavfeutna Mourials' \_ ~ D
-------
attribute testing structure enables one to conclude if a waste is or is not hazardous-not the
degree of hazard.
For clarification of what is meant by average, the Agency provided Chapter Nine of
"Test Methods for the Evaluation of Solid Waste, Physical/Chemical Methods" (SW-846). This
guidance suggests that average signifies the arithmetic average.
A problem with this definition can be illustrated by the following scenarios. One of the
RCRA characteristics is ignitability which includes analyzing samples for flashpoint. The
characteristic is present when the waste flashes at a temperature less that 60° C. In many cases
a number is not recorded, only the presence of a flash. The regulations do not state that a waste
which flashes at 10° C is any more hazardous than one that flashes at 59° C. This logic fits
well with "most likely result" or percentage (which does not require numeric results), but is
inconsistent with the arithmetic average. With ignitability testing sometimes the analysis
produces numbers and other times it does not. Without numbers, arithmetic averages can not
be calculated, but the "most likely result" can be determined. Another characteristic is
corrosivity which can be determined by pH. Several pH values cannot be averaged to determine
the average property of a waste. First of all they are log values and adding them to derive an
arithmetic average is actually a multiplication, thus yielding the wrong information.
Furthermore, from chemical principles, the final pH depends on the buffering capacity of the
individual components. If a waste in question has pH measurements of 2 and 4 (the regulatory
limit under RCRA is 2.5), it is incorrect to conclude that the average pH is 3 and therefore the
waste is not hazardous. Arithmetic average is also not scientifically valid for either toxicity or
reactivity.
Alternatively, the "most likely result" approach will produce valid results for each RCRA
characteristic test. If five samples were randomly selected from a waste and four flashed, then
80% of the samples exceeded the regulatory limit. If "most likely result" is synonymous with
greater than 50%, then the conclusion that the waste is hazardous can be made with confidence.
Presented at EPA Worbhcp m Jafy 16,1992
'Characterizing Heterogeneous Materials' C~O
-------
Universe or Whole
The waste population to be characterized is not defined in the regulations. If the
"universe or whole" is undefined with respect to both temporal and spatial requirements, it is
not possible to design a valid testing structure. This is illustrated through some of the following
scenarios.
If there are several piles of waste, must a decision be made for each pile or for the whole
of all the piles. What if one pile is above the regulatory limit and the other pile is below? What
if several piles are below the regulatory limit, but a few are above? What if the average is
below; are any of the piles hazardous?
What if the average property of a pile is below the regulatory limit? Some portions of
the pile could be above the limit, and others below. What if the portion of the pile above the
limit was loaded into a truck to be hauled to a non-hazardous waste landfill and this was
sampled? Is this waste which was not hazardous in the pile now hazardous when in the truck?
What if the waste produced today is above the regulatory limit, but the waste produced
yesterday and the waste produced tomorrow is below the regulatory limit, is the waste below
the regulatory limit on average or not?
"Universe or whole" must be defined or it is impossible to know where to sample and
how to interpret the results.
SCIENTIFIC CONCERNS
While it is always possible to sample, analyze and produce data for any waste, the
problem with heterogeneous materials is the confidence that a correct compliance decision was
made based upon a valid testing structure. A compliance decision is affected by the purpose of
Praenud at EPA Workshop U] Jafy 16, 1992
"Oioractmanj Haerogauons Hauriols' C~7
-------
the measurement, use of proper sampling and analytical methods, and correct interpretation of
the results. The variability within heterogeneous materials exacerbates the bias and imprecision
of the results. As bias and imprecision increase and the average approaches the regulatory limit,
compliance decisions become more difficult. If, however, the proper testing structure was
employed, it would dramatically improve the decision making process and increase the
confidence in the results.
A proper testing structure should have the following characteristics:
It must be scientifically valid. Basic scientific principles must not be violated. There
are instances where the ideal scientific solution can be adjusted for simplicity, politics,
and economics, but the solution must remain scientifically valid. Agency regulations,
policy, and guidance must never be inconsistent with basic scientific principles
("Safeguarding the Future: Credible Science, Credible Decisions", EPA/600/9-91/050).
It must be consistent. All elements of the testing structure are dependent on each other
for their validity. Sample collection must anticipate the analyses and data interpretation
to be performed to ensure that the proper samples are collected. Likewise, without
detailed knowledge regarding sample collection, data interpretation is meaningless.
Proper Sampling
Proper sampling primarily depends on the regulatory question (e.g., presence, trends,
absence, etc.), the universe or whole (population), and the confidence that the correct
compliance decision can be made. These three considerations-question, population,
confidence—must be definitively addressed before a sampling plan is developed and implemented.
Presoiial al EPA Workshop W July 16. 1992
'Cfaracterizinf Hetfrofenfaus Uaterials'
-------
Proper Analysis
Proper analysis for RCRA characteristics is not an issue, because three of the four
characteristics have required test methods referenced in the regulations (reactivity methods are
only guidance at this time). The correct analytical result is the one obtained by following the
prescribed analytical method.
Proper Data Interpretations
Proper data interpretation is important for making correct decisions. Data interpretation
ties the analytical results back to the population with some specified degree of confidence to
determine if the sampling and analysis answered the question about the waste. Proper data
interpretation includes:
Making the proper type of confidence statement. The statement does not need to be
statistical; it can be an expression of professional judgment. If statistical statements are
made one must recognize that there are many types (e.g., confidence interval of the
mean, tolerance limits, prediction intervals, etc.). Not all statistical statements are valid
for any particular compliance problem and care must be chosen to select the correct one.
Establishing degree of confidence in the statistical statement. It is possible that one can
make a certain statistical statement (or use professional judgment), but if one is not very
confident in the statement it serves no purpose proving compliance or non-compliance.
With many types of statistical statements, certain assumptions are made regarding the
distribution of the data. If incorrect assumptions are made, the resulting statistical statement and
degree of confidence might not be correct. The guidance given in SW-846 makes certain
assumptions regarding the distribution of data (normally distributed) which are usually incorrect.
By definition, heterogeneous waste (as well as most environmental data) defies these assumptions
Praaaed at EPA Workshop D7 July 16, 1992
'Qtamclercjng Heserofauous Materials'
-------
to a degree that can make the decision and confidence statements based on arithmetic average
incorrect. The usual indicators of non-normal data are outliers and skewed distributions. With
the "most likely result" no distribution is assumed (non-parametric) and therefore the problems
with heterogeneous waste do not affect the accuracy and confidence of the decision.
CONCLUSIONS
Characterization of heterogeneous waste depends on the development on a valid testing
structure. Currently, only portions of the testing structure are in place. Two items that need
to be addressed are:
The "most likely results" (percentage) is the valid statistic for making compliance
decisions to characterize waste. The current use of arithmetic average is scientifically
invalid for the RCRA characteristics and can lead to wrong compliance decisions. There
are at least two percentage options that currently exist in the RCRA regulations. One is
the empty drum regulation (40 CFR part 261.7) which states that drums with a capacity
of less than or equal to 110 gallons are considered empty if no more than 3% is left (and
the contents have been removed using common practices). Drums with a capacity of
greater than 110 gallons are allowed 0.3%. The other percentage option is found in the
Land Ban regulations (49 CFR part 268) which requires that no more than 1 % of the
population can be over the treatment standard.
With this new system, the Agency must decide what percentage of the waste it
will allow to be over the regulatory limit and still classify the waste as non-hazardous.
To be the most protective of the environment, 0% should be adopted although it is
probably an unrealistic standard.
"Universe of whole" must be clearly defined. This is not a scientific issue, but rather
a policy issue, The term requires clarity and specificity as to both time and space. Some
Praaaed at EPA Workshop m July 16, 1992
'Qaracteraaif Heterogeneous Materials' V_-~ 1U
-------
examples might be: one day of production, one week of production, one pile, all piles
currently on the property, 100,000 Kg, etc. Defining "universe or whole" as one waste
stream would ignore the temporal problems. Without knowledge of exactly what the
waste population is in terms of both time and space, it impossible to develop a testing
structure to characterize it.
Presented at EPA Workshop m July 16, 1992
'Characterizing Heterogeneous MaSeriols' C-11
-------
HETEROGENEOUS WASTE CHARACTERIZATION:
INDUSTRY PERSPECTIVE
David Reese
Safety-Kleen Inc.
PO Box 92050
Elk Grove, IL 60009-2050
INTRODUCTION
Heterogeneous as defined by Webster in the context of a waste is "consisting of dissimilar
or diverse ingredients or constituents". The complexity of this issue is illustrated with a drum
of heterogeneous waste in figure 1.
Characterization of heterogeneous wastes presents a number of unique challenges for the
industrial community. The principal challenges are: sampling, methodology of analysis,
frequency of testing, and controlling/ reducing costs.
SAMPLING TECHNIQUES
Routine sampling techniques consisting of sample segregation, compositing,and particle
size reduction are often inappropriate when trying to homogenize a heterogeneous waste.
Collecting the sample can become a formidable task when the waste is comprised of multiple
large particle size components. SW-846 sampling containers often are not suitable for sample
containment and larger vessels are required. Loss of volatiles due to headspace in the vessels
affects the validity of the results. Present particle size reduction techniques such as grinding
only exacerbates the problem by driving off volatiles.
Presented a: EPA Workshop m July 16, 1992
'Oiaructerizatf Heterogeneous Materials"
-------
TESTING REQUIREMENTS AND PROCEDURES
State and Federal regulations often require extensive testing protocols in Waste Analysis
Plans (WAPs) and permits. Regulators routinely require the use of SW-846 analytical methods.
These methods often are not appropriate for a variety of waste streams (organic solvents, oily
wastes, by products of recycling such as distillation bottoms). The use of methods other than
SW-846 should be accepted for alternative test procedure approval and not rejected on the sole
basis they are not SW-846. Most of the quality assurance/ quality control principals and
guidelines can be incorporated into alternative test methods resulting in improved data with only
procedural and or hardware modifications. Unfortunately these methods are not considered valid
by regulators.
The present sampling and testing requirements fail to consider the volume and frequency
of the waste being handled. It is extremely costly to test every fifty-five gallon drum according
to SW-846 guidelines, while testing every incoming rail car might be feasible. Current sampling
frequencies often overlook one of the key issues facing a heterogeneous hazardous waste
handler, that being, the end use of the waste. If small volumes of wastes are going to be mixed
with other wastes for processing in terms of recycling, and the facility is permitted to handle the
waste, the result will have no impact on the process, thus, sampling and testing need to reflect
the use of the waste while providing a safe working environment. The product of the process
should become the focus of testing.
Continuing the status quo will drive many companies out of business and create large
companies which will monopolize hazardous waste handling while passing the costs on to the
generator and ultimately the consumer. This is due to the excessive costs associated with testing
every handled waste. Multiple testing of the same waste is redundant and provides no greater
assurance of safety, process control, product quality.
Presotud ta EPA Worbhep mjufy 16. 1992
'OarnctercBig Heterogeneous Materials' C»" 13
-------
CONCLUSIONS: DEVELOPING ALTERNATIVES
Recognize Alternative Test Methods
State and federal regulators must become accepting of industry generated methods of
analysis for complex waste streams handled on a daily basis.
Develop New Methods
EPA in conjunction with industry should work together to facilitate the rapid development
and approval of new methods. This could be accelerated by using industry developed methods.
Encourage Recycling
Recycling needs to be encouraged by minimizing input testing and focussing on
by-products of the process. When a facility is fully permitted to handle wastes, analysis does
not affect the processing or end use. This adds unnecessary costs to the recycling process
through testing of incoming, in process, and final products.
Account for End Use
Prescribing testing must account for the end use of the waste. Some end uses such as
alternative cement kiln fuel present minimal hazards, thus testing should be relevant to the waste
destination.
Balance Cost with Severity of Hazard
The health and safety models need to be incorporated into the testing protocol by
focussing on hazards and not over testing of materials that pose little or no threat.
Presetted at EPA Workshop HI July 16, 1992
'Oaraaerizing Heterofeneaa tfaterials' C-14
-------
Obstacle to Implementation
The largest obstacle for implementation of the above alternatives is the magnitude of the number
of individual state regulators with differing view points. The result of negotiating fifty different
WAPs in fifty states poses limitations on industries ability to comply fully at all times. It is
impossible for a business to operate fifty different ways on the same type of waste stream.
REFERENCE
Characterizing of Heterogeneous Wastes: Methods and Recommendations; EPA 60Q/R-92/033,
Feb. 1992
Praenled at EPA Workshop CJ July 16, 1992
'Characterizing Haavgeneaa Materials' C-15
-------
CHARACTERIZING HETEROGENEOUS MATERIALS:
SCIENTIFIC PERSPECTIVE
John P. Maney, Ph.D
Environmental Measurements Assessments
5 Whipple Road
Hamilton, MA 01982
INTRODUCTION
A discussion of heterogeneous waste characterization is best started with a brief review
of definitions.
Sample: a part or piece taken or shown as representative of a whole group. (Webster's
Unabridged Dictionary, Second Edition)
Sampling: the act, process or technique of selecting a suitable sample.
(Webster's 7th Collegiate Dictionary)
Representative Sample: A sample of a universe or whole (e.g. waste pile, lagoon,
ground water) which can be expected to exhibit the average properties of the universe
or whole. (CFR260.10)
Homogeneous: uniform structure or composition throughout.
(Webster's 7th Collegiate Dictionary)
After studying the above definitions and applying them to waste characterization, it
becomes apparent that sampling of an ideal homogeneous waste will always result in a sample
that represents the properties of the waste (assuming that the sampling process itself does not
introduce contamination or allows for selective loss of waste components).
Presented at EPA Workshop W July 16, 1992
'Ouraaejuing Heterogeneous Materials' C~ 16
-------
Heterogeneous: consisting of dissimilar ingredients or constituents. (Webster's 7th
Collegiate Dictionary)
Unfortunately, many of the real-world wastes are not homogeneous, but are
heterogeneous, which makes collection of a representative sample a challenging task. Prior to
discussing the representativeness of heterogeneous wastes, the following points can be made
regarding homogeneity and heterogeneity.
• Homogeneity and heterogeneity are diametric terms.
• Homogeneity and heterogeneity are relative to objectives.
• Homogeneity and heterogeneity are related to spatial and temporal distribution.
• Homogeneity and heterogeneity are sample size dependent.
A determination regarding the homogeneity and heterogeneity of a waste is made by
comparing the visual, physical or chemical composition or properties of different samples. The
more heterogeneous a waste is, the less homogeneous it is and vise-versa. Thus, when either the
homogeneity or heterogeneity of a waste is defined the other is also defined. This relationship
should be kept in mind, since for the sake of readability, the remainder of this discussion will
often refer to only one term.
Heterogeneity is relative to objectives and perspective. A non-random mixture of fine
silver nitrate powder and large crystals of silver nitrate will be considered to be heterogeneous
by the analyst performing particle size determination while the analyst performing the TCLP for
lead would consider the material homogeneous. Likewise, while the chemical properties of
99.99999% pure uranium are homogeneous, a nuclear chemist may consider it highly
heterogeneous on an isotopic level.
Presented al EPA Workshop mJufy 16. 1992
'OiaraOerKing Heterogeneous Materials' C-" 1 /
-------
The spatial or temporal distribution of dissimilar waste constituents or concentrations
gradients can also affect the measured heterogeneity of a waste. If a representative distribution
is spread over a greater area or length of time than that represented in a sample, then sample
to sample heterogeneity will be greater than when a sample encompasses the distribution of the
dissimilar waste components or concentration levels. For example, the measured heterogeneity
of an effluent, whose concentration of phenol linearly cycles from lOOppm to 5ppm and back
to lOOppm every two minutes, will be a function of what portions, or if the entire cycle is
included in sequential samples.
Assuming that the concentration of the target parameter varies with particle size, sample
to sample heterogeneity will depend on whether a sample is not only large enough to
accommodate all the different sized constituents of a material but will also depend on whether
the sample is large enough to accommodate representative amounts of the various sized
constituents of a waste. To better understand this concept consider a 2 liter waste container that
has one gram nuggets of cadmium randomly distributed through-out an otherwise homogeneous
and cadmium free matrix. If one gram samples are collected and analyzed for cadmium, the
sample to sample heterogeneity of the waste will vary dramatically and will depend upon
whether the sample did or did not contain a cadmium nugget. However, the heterogeneity will
appear to be substantially less if 30 gram samples are collected and analyzed in total for
cadmium. Thus the measured heterogeneity of the same waste varies according to the size of the
sample. That is why for the following discussion, unless otherwise indicated, the sample size
used to determine heterogeneity of a waste will be the analytical sample size. Heterogeneity of
a waste will be measured according to the ability of an unaltered analytical sample to
reproducibly represent the average properties and/or chemical constituents of the waste unit of
interest. (The analytical sample is the mass/volume of sample submitted to analysis not the field
sample, e.g. the 1 to 2 gram sample required by the acid digestion specified in Method 3050 .
Unaltered means that the sample was not subjected to homogenization or pulverization steps. A
waste unit is the population or unit of waste that is being subjected to evaluation, e.g. drum, a
wastepile or landfill).
Presorted at EPA Workshop mJuty 16, 1992
'Characterizing Heterogeneous Materials' C~ 1 u
-------
Representative Database: a database, generated by the collection and analysis of more
than one sample, which together represent the average properties of the universe or
whole.
It is often impossible to collect a single sample which is representative of a heterogeneous
waste. The average properties of a heterogeneous waste are better represented by collecting a
number of waste samples such, that all portions of the waste under study have an equal chance
of being sampled. Analyses of these samples can be compiled into a database which should be
representative of the waste. It is more correct to think of a representative sample in terms of a
representative database, which is closer to the statistical use of the term, sample.
In light of all the attention directed towards heterogeneous waste, it is important to note
that from a sampling perspective, the significance of heterogeneity is not in this quality, itself,
but in the fact, that heterogeneity hinders or prevents the generation of a representative database.
The need for representativeness is the driver behind the interest and studies into heterogeneous
waste characterization.
Random: being a member of , consisting of, or relating to a set of elements that have
a definite probability of occurring with a specific frequency; being or related to a set
whose members have an equal probability of occurring. (Webster's 7th Collegiate
Dictionary)
Randomness is a critical factor in the characterization of hazardous waste. Accurate
characterization requires that the sampling process reflect the randomness or lack of randomness
of the waste. If the waste parameter of interest is distributed in a random fashion a proper
randomly designed sampling program should generate a representative set of samples. If the
waste parameter is not randomly distributed (refer to discussion of strata in the following
section) then a simple random sampling plan may not accurately characterize the waste.
Presented at EPA Workshop W July 16, 1992
'Characterizing Hcterogaieaa Materials' L~ 1;»
-------
TYPES OF HETEROGENEOUS WASTE
If the average properties of the waste unit of interest is not reproducibly represented in
analytical samples, then the material is considered heterogeneous. Such heterogeneous wastes
can exist in many forms;
RANDOM HETEROGENEOUS
• Randomly heterogeneous in composition/property.
• Randomly heterogeneous in particle size.
• Randomly heterogeneous in particle size and composition/property.
NON-RANDOM HETEROGENEOUS (STRATIFIED)
• Non-randomly heterogeneous in composition/property.
• Non-randomly heterogeneous in particle size.
• Non-randomly heterogeneous in particle size and composition/property.
EXCESSrVELYNON-RANDOMHETEROGENEOUSCEXCESSIVELYSTRATIFIED)
• Excessively non-randomly heterogeneous in composition/property.
• Excessively non-randomly heterogeneous in particle size.
• Excessively non-randomly heterogeneous in particle size and
composition/property.
Assuming that all particle sizes in the waste can be accommodated by the analytical
sample size and that the analytical method is applicable to all waste constituents, then randomly
Presented at EPA Workshop mJufy 16, 1992
'Characterizing Heterogeneous Materials'
-------
and non-randomly heterogeneous waste can be characterized by a database generated from the
collection and analysis of random or systematic samples. Regarding non-randomly heterogeneous
waste a more precise estimate of mean (representative) concentrations can be obtained with a
stratified sampling approach.
Excessively non-random heterogeneous wastes are defined as those wastes that have
such dissimilar components, that it is not practical to use traditional sampling approaches to
generate a representative database. Nor would the mean concentrations reflected in such a
database be a useful predictor of a given subset of the waste that may be subjected to
evaluation, handling, storage, treatment or disposal, (i.e. the level of uncertainty is too
great).
Stratum: a portion or subgroup having a consistent distribution of the target
parameter, and a different distribution than the rest of the waste.
Waste strata can be thought of as different portions of a waste population, which may
be separated in time or space or by waste component, and each portion has internally similar
concentration distributions of the target parameter, through-out. A landfill may display
spatially separated strata, since old cells may contain different wastes than new cells. A
wastepipe may discharge temporally separated strata, if night shift production varies from the
day shift.
There are wastes which do not display any identifiable temporal or spatial
stratification, yet the target compound distribution is excessively erratic. For these wastes it
may be helpful to consider a third type of stratification - stratification of the waste by waste
component.
Component: A waste component is any identifiable article, discrete unit or
constituent of the waste, which is randomly or non-randomly distributed through-out
the waste.
Presented at EPA Workshop m Jufy 16. 1992
'Otcraaersing Heterofeneous Materials' C~21
-------
Stratification by waste component is easily applied to wastes that contain easily
identifiable particles such as large crystals or agglomerates, rods, blocks, gloves, pieces of
wood, concrete, etcetera. Strata separated by waste component is a different but key concept.
Separating a waste into strata according to waste components is useful when a specific kind
of waste component is not randomly distributed through-out the waste and when a
contaminant or property of interest is correlated with the waste component. This type of
strata is different since the components are not necessarily separated in time or space but are
usually intermixed and the properties or composition of the individual components are the
basis of stratification. For example, automobile batteries which are mixed in an unrelated
waste would be a waste component which could constitute an individual strata if lead was a
target parameter. If one were to sequester the batteries they would have a consistent
distribution which was different from the rest of the waste. However, if the concentration of
the target parameter is similar in different components and the particle size is such that all
the components are represented in the chosen sample size, then there is no purpose in
stratifying by waste component. Strata by component is an important mechanism for
understanding the properties of excessively non-random heterogeneous waste and for
designing appropriate sampling and analytical efforts for these types of waste.
Table 1 summarizes mechanisms for segregating strata as well as the discriminating
qualities of strata. Since the mechanism and qualities are independent of each other there are
a total of 9 possible types of stratified wastes.
Presented at EPA Workshop m July 16, 1992
'Characterajnf Heterogeneous Maltriah' C-22
-------
TABLE 1. Strata Considerations
Mechanism for Segregating Strata
Spacial
Temporal
Component
Discriminating Qualities
Composition/Property
Particle Size
Composition/Property and Particle Size
The contamination in different strata are often generated by different processes. The
different geneses of the contamination will usually result in a different concentration distribution
and mean concentrations. Thus, each strata will have their own distribution and mean
concentration levels. If a waste having strata of distinctly different concentration distributions
is sampled using a simple random sampling approach, the concentration distributions of the
different strata may result in a bi-modal or multi-modal distribution. Homogeneous and randomly
heterogeneous waste, which are not stratified will display a normal unimodal distribution. See
Figure 1.
Non-randomly heterogeneous waste will usually have a limited number of strata which
can be identified and sampled individually or sampled as one waste. When different strata are
sampled as one population, the properties of the different strata are averaged. If the different
strata are similar in composition, then the average concentrations will be a good predictor of
composition for subsets of the waste and will often allow the program objectives to be achieved.
As the difference in composition between different strata increase, the average values become
less useful in predicting composition/properties of individual portions of the waste, which may
be handled separately. In this later case, it is advantageous to sample the individual strata
separately and if an overall average of waste composition is needed, it is best calculated
mathematically using statistical information from each strata.
Excessively non-random heterogeneous waste has numerous strata, each of which contain
different distributions of contaminants and/or particle sizes, such that an average value for the
Presented at EPA Workshop jn July 16, 1992
'Characterizing HOfrogemeus Materials'
C-23
-------
Normal Bell Distribution
(Gaussian)
Concentration
cr
OJ
Bimodal
Distribution
Concentration
Concentration
Figure 1: Types of Concentration Distributions
PresetedasEPA Workshop mjufy 16, 1992
'Characterizing Heserogateaus Materials'
C-24
-------
waste would not be useful in predicting the composition or properties of individual portions of
the waste (i.e. statistically speaking the variance and standard error of the mean will be large).
The line of demarcation between non-random heterogeneous and excessively non-random
heterogeneous waste is not hard and fast. A waste can be considered excessively non-random
heterogeneous when mean values will not be representative of most portions of the waste and
when it is so heterogeneous that the waste cannot be cost-effectively sampled using traditional
sampling approaches and meet the project objectives.
A theoretical example of an excessively stratified waste could consists of commingled
waste from:
Source A - which generated a waste of varying particle size with a mean antimony
concentration of 20ppb and a standard deviation of 7ppb.
Source B - which generated a waste of varying particle size with a mean antimony
concentration of 7000ppm and a standard deviation of SOOOppm.
Source C - which generated a waste of varying particle size with a mean antimony
concentration of 3% and a standard deviation of 2%.
Source D - which generated a waste of varying particle size with a mean concentration
of 81% and a standard deviation of 17%.
To further complicate the above waste, imagine the waste being commingled with other
materials of various particle size that may or may not contain various contaminants in addition
to antimony.
To improve readability, the remainder of the document will refer to non-random
heterogeneous waste as stratified waste and excessively non-random heterogeneous waste as
excessively stratified waste.
PresaiUd at EPA Worfahop W July 16. 1992
'Otarsutersant Heterogeneous MOerials' C-25
-------
Figure 2 presents a grEpfiical depiction cf the types of special dsstrfubtions and heterogeneity. Letters
are waste constituents which are kJsnOcs! to ctfterconslfiaents ol the same fefier interns ofthe
composition, property, or partide size of Msrest, DSferert waste const'rtusias am desolated by
different letters.
Homogeneous
AAAB A A
/^BAAABABAAB>
BAABAAAABBAA'
'ABABABABABABAB'
,'BABABABABABABAB]
[ABABABABABABABAj
ABABAB ABABABAI
vBBBBBBBBBBBBB,
IBBBBBBBBBB
^BBBBBBBBB,
~
Stratified
Excessively Stratified
or
Random Heterogeneous
'YYY
( ^| AY Y Y YYY\
fYYYYYY^c'TYYYYYYl
'YYYYYlpl'YYYYYYI
'YYYYYj ^ Y Y Y Y Y Y/
iz,
Stratified
As can be seen in the stratified
heterogeneous waste above,
heterogeneity can vary over
time.
Figure 2: Types of Positional Distributions for
a Waste Unit Such as a Landfill
Presaged at EPA Workshop m Juty 16, 1992
'Characterizing Heterogeneous Materials '
C-26
-------
This figure depicts the process
by which a waste is classified
as a particular waste type.
This classification will be
based upon knowledge of the
waste, observations, or ideally
- preliminary sampling of the
waste.
If no significant variation is
detected between random
samples, the waste can be
considered homogeneous.
The waste is a heterogeneous
waste if there is variation
between individual samples.
If the waste constituents are
randomly distracted
through-out the waste, the
waste would be a randmoly
heterogenous waste. It
information describes a
correlation between waste
variation and time or spatial
variations or with certain waste
components or particle size,
!hen the waste would be
classified as a stratified or
excessively stratified waste.
Wastes that can be
cost-effectively sampled to
meet project objectives would
be classified as stratified
wastes. Those wastes which
consist of such strata that they
cannot be cost effectively
sampled are classified as
excessively stratified waste.
Knowledge
Observations
Preliminary Sampling
Homogeneous
Waste
Variation between
Samples
Randomly
Heterogenous
Variation Correlates with
Time, Space, Particle
Size, or Components
1
Can Waste Be Cost-effectively
Sampled to Meet Objectives
Excessively Stratified
Waste
Figure 3: Process for Classifying Wastes
Preserved at EPA Workshop HlJufy 16, 1991
'Oaracterizing Heterogeneous Materials'
C-27
-------
Table 2 details some of the issues and concentration distributions that are pertinent to the
different types of heterogeneous wastes. For example, sample size, the number of samples and
sample locations are not an issue when sampling a homogeneous waste, while they are critical
issues when attempting to collect a representative sample of a heterogeneous waste.
PARTICLE SIZE AND SAMPLING THEORY
As a result of a continuing need to improve environmental data, it has been recognized
that sampling is presently the greatest contributor to imprecision and inaccuracy. To minimize
this error, many have correctly turned to the mining industry, which has a relative wealth of
sampling theory and expertise. The theory and applicable expertise of the mining industry has
recently been documented by a recognized expert on sampling, Francis F. Pitard, (Pitard,
Francis F., "Pierre M. Gy's Sampling Theory and Sampling Practice, CRC Publishers). This
theory, which not only applies to field sampling but also to subsampling performed in the
laboratory - determines sample mass/volume as a function of the maximum particle size of the
material being sampled..
Based on the maximum particle size of a material, sampling theory suggests minimum
sample sizes (refer to Table 3.) and particle size reduction of the sample to minimize sampling
error. For example, a waste having a maximum particle size of 0.5 inches would require
collection and analysis of a minimum sample size of 2 Kilograms to keep the sampling error less
than 17%. Since a 2 kilogram sample is too large to be subjected to analysis, it would have to
be subjected to particle size reduction prior to analysis.
It would be very costly to routinely use existing sampling theory to minimize sampling
error. These increased costs would be driven by the large sample sizes and the requirement to
reduce particle size.
The large sample sizes, specified to minimize sampling error, would increase costs since;
Presented a: EPA Workshop EMuly 16. 1992
'Characterizing Heterogeneous Materials ' L- ~2.0
-------
• Sample containers and coolers would be larger and more numerous
o Larger samples would be shipped
• Larger sample storage and sample handling areas would be needed
• Large volume of unused samples would increase disposal costs.
• Health and Safety costs associated with exposure to larger samples.
Particle Size Reduction, (PSR), also has a substantial impact on cost because of the
manpower and capital investment for the required crushing and pulverizing equipment. In
addition the following are reasons for concern if particle size reduction was to be implemented
on a large scale;
• Laboratory contamination from fines.
• Damage to instrumentation from fines.
• Difficulty in cleaning PSR equipment to prevent cross-contamination.
• Loss of volatile and labile compounds and elements.
• Generation of a sample which has different properties and thus yields different
analytical results than the original sample.
• Increased Health and Safety Costs
Fortunately, the costs of employing large sample sizes and particle size reduction can
often be minimized if the waste history and Data Quality Objectives are communicated to and
discussed with the lab staff and field personnel.
Presented at EPA V/orkhap ffl Juty 16, 1992
'Characterizing Heterogeneous Materials'
-------
TABLE 3. SUGGESTED SAMPLE SIZE BASED ON MAXIMUM PARTICLE SIZE
ACCORDING TO PIERRE GY'S SAMPLING THEORY
SUBSAMPLE SIZE MAXIMUM PARTICLE SIZE
(grams) (centimeters)
1 .1
2 .13
3 .14
4 .16
5 .17
10 .21
20 .27
30 .31
40 .34
50 .37
75 .42
100 .46*
*The Toxicity Extraction Procedure and the Toxicity Characteristic Leaching Procedure allow
samples to contain particles as large as .95 centimeters.
Waters and soils can be contaminated in numerous ways, the most common mechanisms
being; direct discharge of the contaminant onto the soil or water and atmospheric fall-out from
fires, fugitive emissions, vents and stacks. Soils can also become contaminated from adsorption
of groundwater or surface water contaminants while waters can become contaminated by mixing
with contaminated waters or by leaching contaminants from contaminated soils or wastes.
Wastes can be contaminated by many mechanisms during generation or by mixing with
other wastes or further contaminated by the mechanisms described in the previous paragraph.
The different types of contaminants are aqueous liquids, non-aqueous liquids, gases, small to
large particles and multi-phased mixtures.
Pramled at EPA Workshop m July 16, 1992
'Otaraclerizing Haerogaieaa Materials' C-30
-------
How a contaminant will be dispersed in the matrix will be determined by the mechanism
of contamination, the type of contamination, the surface area and adsorptive properties of the
sample matrix, gravity, physical manipulation and the physical size of the contaminant at the
time of contamination. (Eventually, fate and transport will substantially affect dispersion of
contaminants.) The physical size of the contaminant can be referred to as the contaminant unit
and can be measured in centimeters. Liquids, dissolved contaminants and gases have a
contaminant unit on a molecular or atomic level, (e.g. atoms can be on the order of 1 X 10-8
centimeters in diameter). Solids and suspended solids contaminate on a microscopic to
macroscopic level, (i.e. their Contaminant Unit is equal to their particle size).
Particle size can have an impact on sampling error even when the contaminant unit is on
the atomic scale, if matrix particles are of different sizes and have different adsorptive
properties. If adsorptive properties are similar for different size particles, then the dramatically
larger surface area of smaller particles/volume would result in more atomic scale contaminants
being adsorbed to smaller particles. (Thus one could eschew particle size reduction, (PSR), and
still employ a smaller subsample size by excluding large particles, if the ease of using a smaller
sample size out-weighed the chance of a false-high concentration.)
When the contaminant unit is on a larger, particle scale then adsorptive properties of the
matrix will not affect contaminant distribution in the sample. For microscopic and macroscopic
particles, an accurate representation in the subsample will be affected by the number of particles
and the relative size of the particles to the subsample size. Microscopic particles will have a
contaminant unit substantially smaller than the sub-sample size and these small particles should
not be discriminated against during subsampling.
When the contaminant unit is macroscopic and approaches the size of the sample, the
error in sampling will increase, if large sample sizes or PSR is not employed. If the largest
particle size in the sample correlates with the contaminant unit, then the large sample size
specified by sampling theory pertains to the waste. As the contaminant unit becomes smaller as
Praaaed al EPA Workshop ffl Jufy 16, 1992
'Quraaer&ing Heterogeneous Materials' C"31
-------
compared to the size of the sample, the greater the opportunity to avoid PSR and to selectively
exclude larger particles from the sample without introducing substantial error., (If desired the
weight of the excluded particles can be used in the calculation of a final concentration. However,
this could result in a false low result, since the contamination, if any, associated with the large
particles is not accounted for.)
The mechanism of contamination, the type of contaminant and the contaminant unit will
impact the complexity and potential error of sampling. These impacts may be better understood
by reading the following examples.
Example 1
TYPE OF CONTAMINANT: Particles contaminated with lead
MECHANISM: Direct discharge of particle to soil
CONTAMINANT UNIT: Particles as large as 0.3 centimeters
AVERAGE LEAD CONCENTRATION: lOppm
HISTORY: A manufacturer of lead-alloy babbitt lining for bearings, stored its machining
waste in piles behind it facility. Periodically the machining wastes would be recycled into
new babbitt linings. After 50 years of employing this practice, the soil became
contaminated with particles of babbitt having a maximum particle size of 0.3 centimeters.
The contaminated soil consisted mainly of fine silts with less than 10% consisting of
gravel with stones having diameters ranging from 2 to 3 inches. The sampling team
uncovered a problem when they referred to the specified "minimum sample size" for 3
inch particles (i.e. 450kg). The problem of large sample size was avoided, since the
sampling team was aware of the "Type of Contaminant" and the "Mechanism" of
contamination and the "Contaminant Unit". Knowing that lead contaminated particles
were discharged directly to the soil, the sampling team was able to discard the large
stones with the realization that by discarding the stones, the resulting lead concentrations
would be a worst case. (The small lead contaminated particles would preferentially exist
in the fine silts as opposed to being adsorbed to the large stones and if lead had
dissolved, the dissolved lead would tend to adsorb to the silt which has the much larger
surface area.) The sampling team collected approximately 250 mis of soil and delivered
them to the laboratory for analysis.
Praatitd at EPA Workshop Dl July 16, 1992
'Otamclerizing Heterogeneous Materials'
-------
The analyst referred to sampling theory (refer to Table 3.) and determined that the
minimum sample size for a sample with a maximum particle size of 0.3 centimeters was 30g.
(The small percentage of larger gravel was not used to determine the maximum particle size.)
The analyst split the sample into 30g aliquots and randomly chose one for analysis. Since
the analyst did not have particle size reduction equipment and the largest sample volume he
could analyze was approximately 10 grams, he split the chosen aliquot in thirds and analyzed
all three for lead. After analysis, the concentrations were averaged.
Example 2
TYPE OF CONTAMINANT: Ionic lead dissolved in an aqueous solution (Atomic scale)
MECHANISM: Direct discharge to soil
CONTAMINANT UNIT: Atomic scale, (approximately 1 X 10-8 centimeters for atoms
and somewhat larger for ion diameters)
AVERAGE CONCENTRATION: 10 ppm
HISTORY: The above described babbitt manufacturer also employed an acid-treatment
process. The aqueous waste was discharged into a tank which allowed the lead
contaminated water to leach into the surrounding soils. The surrounding soils consisted
of fine silt with 15% of its volume consisting of stones ranging in size from 0.25 to .5
inches. The sampling team collected a 2 Kilogram sample to minimize the sampling
error. The samples were sent to the laboratory for analysis. Since the analyst knew that
the soil was contaminated by lead on an atomic scale, he discarded the small stones, split
the sample and analyzed a 2 gram aliquot. (The analyst knew that the much higher
surface area of the silt would have adsorbed orders of magnitude more lead than the
stones, so that this approach yielded a worst-case analysis. To check this hypothesis, the
analyst performed an acid leach on a few stones and no detectable quantities of lead were
measured.)
The above examples show how, with knowledge of the type and mechanism of
contamination and the contaminant unit the sampling team and the analyst can decrease the costs
of handling large samples and particle size reduction without substantially increasing sampling
Presented at EPA Workshop mjuly 16, 1992
'Qvracterizing Heterogeneous Uaterials' C"33
-------
error. Without the site history, large sample sizes or particle size reduction would have been
required or one would risk the introduction of a substantial error into the measurement process.
Thus communication with the analyst regarding the type and mechanism of contamination is an
important means of controlling costs and minimizing errors.
The above sampling strategies are not ideal since they allow for assumptions and
interpretation and these assumptions and interpretations could lead to substantial intentional and
unintentional errors. However, the above sampling strategies are more practical, more likely to
be implemented and may be a significant improvement over the alternatives. Lastly, when these
sampling strategies are implemented by an experienced personnel who are aware of the DQOs,
site history, the types and mechanisms of contamination and the contaminant unit, the errors
associated with interpretation and assumptions should decrease substantially.
It is important to note, that application of these sampling strategies to wastes is much
more difficult than for soils and subsampling of wastes may frequently require a strict
application of sampling theory. Wastes are a more difficult media, since the complexities of
waste generation often preclude knowledge of how a contaminant of interest is dispersed within
a waste and whether distribution will be a function of particle size.
EXCESSIVELY STRATIFIED WASTES
The remainder of this discussion will concentrate on excessively stratified wastes, the
most difficult wastes to characterize. Compositional or particle size heterogeneity or a
combination of both compositional and particle size heterogeneity can be the cause of excessive
stratification. This discussion will address each of these causes and will concentrate on sampling
issues, since sampling is now recognized as the greatest contributor to imprecision and bias.
However, before addressing the different causes of heterogeneity, it will be useful to
consider an issue which should be common to all waste characterization efforts.
Presented at EPA Workshop JH July 16, 1992
'Characterizing Heterogeneous Materials'
-------
The first step in characterizing any heterogeneous waste is to gather all available information on
the;
• Need for waste characterization,
• Objectives of waste characterization,
• Pertinent regulations, consent orders, and liabilities,
• Sampling, shipping and laboratory health and safety issues,
• Generation, handling, treatment and storage of the waste,
• Existing analytical data and exacting details on how it was generated,
• Treatment and disposal alternatives.
These types of information will be used in the planning of the sampling and analytical
effort. The planning process should be detailed and address the issues defined in EPA's data
quality objective (DQO) process. The different disciplines ( e.g. sampling, chemistry,
engineering, statistics needed to properly understand and exploit the above types of information
must be present during the planning process. If enough information is available, the planning
process will uncover the existence of excessive stratification which will prevent achievement of
objectives. If information is lacking, a preliminary sampling effort would be advisable, and if
done properly should detect the existence of excessively stratified wastes.
Excessively stratified waste can not be cost-effectively characterized by traditional
methods and this fact usually becomes apparent during the planning process. The following
discussion will consider approaches which in effect will lessen the level of stratification and
allow for more cost-effective characterization. Some of these approaches will require changes
in objectives, waste handling or disposal methods, and some will require compromises, but all
approaches will require the above types of information.
Presented at EPA Workshop m July 16, 1992
'Oaractercinf Heterogeneous Materials' C"35
-------
Of the following types of excessively stratified waste, the difficulty in characterizing a
waste increases from those that have strata based solely on particle size, to those which consist
of compositional strata to those which have strata of varying composition and particle size. In
fact, the difficulty in characterizing waste, which consists of strata of different particle size, is
simply a matter of determining that composition is constant across different particle sizes. If the
composition is found to be constant across the different particle sizes, the waste should be easily
characterized.
The more problematic types of waste, which have enumerable strata of different
composition or a combination of different composition and particle size are much more difficult
to characterize. The approach to characterizing these wastes usually has to be determined on an
individual and unique basis.
Excessive Strata of Different Sized Particles
Wastes having excessive stratification due only to different sized particles will by
definition have the same composition or property (i.e. homogeneous or randomly heterogeneous)
through-out its different strata. The strata can be separated in space or in time. Unless one is
attempting to measure particle size, this waste is the simplest of the excessively stratified waste
types to characterize. All particles in these types of wastes are usually generated by the same
process, (e.g. smelter slag and the previous example of silver nitrate powder and crystals),
which is the reason for similar composition across all particle sizes.
The complexity of dealing with these types of wastes is in proving that the waste has
similar composition, (i.e. mean levels and concentration distribution of the parameter of interest)
across the varying particle sizes. This determination can be made by using knowledge of the
waste or by sampling the different sized particles to determine if there are significant
compositional differences. If the determination is made using knowledge of the waste, it is
advisable to at least perform limited sampling to confirm the determination.
Presented al EPA Workshop m July 16, 1992
'Characterizing Heterogeneous Materials' C-36
-------
The characterization process is greatly simplified, once a determination has been made
that the waste has similar composition or properties across the various particle sizes. The
sampling and subsequent analysis can be performed on particles which are readily amenable to
the sampling and analytical process and the resulting data can be used to characterize the waste,
in its entirety.
It is important to periodically verify the assumption that the different particle sizes are
composed of materials having the same concentration levels and distributions of the contaminant
of interest. This verification is especially important when there are any changes to the waste
generation, storage, treatment or disposal processes. Similarity of composition between particles
has to be verified for each parameter of interest. The effect of different particle size must also
be considered when measuring properties such as the Toxicity Characteristic Leaching Procedure
(TCLP) .
Excessive Strata of Different Composition or Composition and Particle Size
Wastes having excessive stratification due only to composition or property will have
similar particle size through-out its different strata. The strata may be separable in space, time
or by component or source. Stratifying the waste should simplify the characterization process.
Wastes having excessive stratification due to both composition/property and particle size
are usually the most difficult wastes to characterize. The strata can be separated in space, in
time, or by component or source.
Figure 4. summarizes an approach to characterizing these types of excessively stratified
wastes. If a waste is excessively stratified, traditional methods of sampling will not allow
objectives to be cost-effectively achieved. To cost-effectively sample an excessively stratified
waste, one must use a non-traditional approach. The non-traditional approach may involve
modification to the sampling, sample preparation or analytical phase of the process. If after
PresatUd at EPA Workshop Ul July 16. 1992
'Charocterizaig Heterogeneous Materials' C-37
-------
Figure 4. Approach for the Characterization of Heterogeneous Waste
Is Waste
Excessively Stratified?
Y
Can Sampling Be Modified?
"I
Can Sample Prep
Be Modified?
N
"1
Change Handling,
Treatment,
Disposal of Waste Or
Target Parameter
Change Sampling
and
Analysis
Objectives
N
N
N
Y
Use Traditional Random,
Stratified or Systematic
Sampling
Will Modified
Approach
Allow Waste
to Be
Cost-Effectively
Sampled
And
Objectives
To Be Met?
Figure 4: Approach for the Characterization of Heterogeneous Waste
Presented as EPA Workshop m Jufy 16, 1992
'Characterizing Heterogeneous Materials'
C-38
-------
modifying the approach to sampling and analysis the objectives still can not be achieved in a
cost-effective manner, then the original plan of waste handling, treatment or disposal has to be
examined and changed so the waste can be characterized according to new and achievable
objectives.
The following subsections discuss approaches that can be employed to make excessively
stratified waste more amenable to a cost-effective sampling approach.
Design of the Sampling Approach
The first efforts to resolve the difficulty in characterizing an excessively stratified waste
are usually focused on the sampling aspects of the project. This is a logical place to start and,
if a successful sampling approach is designed, the project objectives can be cost effectively
achieved.
The difficulty in sampling excessively stratified waste can result from:
1) Various particle sizes and waste consistency which makes sampling difficult and
traditional sampling approaches cost prohibitive.
2) Extraordinary concentration gradients between different components or
enumerable strata that lead to such excessive variance in the data, that project
objectives can not be achieved.
3) Wastes which exhibit both of the above properties.
A strategy for designing a sampling plan for such excessively stratified waste includes
the following five steps;
Presented at EPA Workshop ffl Jufy 16, 1992
'Characterizing Heterogeneous Materials' V_-jy
-------
1) Select the target parameters.
2) Determine whether these parameters are correlated with; particle size space time
components, or sources.
3) Determine if any waste components or strata can be eliminated from sampling
because they do not contribute significantly to the concentration of the target
parameter.
4) Determine if small particles in a stratum represent the stratum as well as large
more difficult to sample particles. If yes, sample the smaller particles and only
track the volume contribution of the larger particles. (See Particle Size and
Sampling Theory).
5) Determine if contamination is innate or surface adsorbed. Is the contamination
surface adsorbed which would allow the material to be representatively sampled
by wipe sampling? Can large particles be wiped and smaller particles extracted,
leached or digested. Can waste be stratified according to impervious and non-
impervious waste and sampled and analyzed accordingly?
To understand how this strategy would work, consider a hypothetical scenario - a storage
area containing 4000 drums of waste generated over a 15 year period. The drum contents are
excessively stratified and contain a myriad of wastes from process waste; destruction and
construction debris such as wood, concrete; lab wastes including broken glassware, paper, empty
bottles; etcetera. The appearance of the combined drum contents could best be described as a
municipal landfill in drums which appears impossible to characterize.
Prcsaued a EPA Workshop m Juty 16, 1992
'Characterizing Heterogeneous Materials' C--40
-------
1) WHAT ARE THE TARGET PARAMETERS?
None of the waste contents are known to be listed waste so the target parameters are the
TCLP and other hazardous waste characteristics. In addition, groundwater modeling has
indicated that the storage area is the source of a plume contaminated with solvents and
beryllium.
2) ARE THE TARGET PARAMETERS CORRELATED WITH AN IDENTIFIABLE
STRATA OR SOURCE?
The source of beryllium is traceable to one process, whose waste should be easily
identifiable if drum markings are not legible enough to determine the source. The solvents are
likewise traceable to a machine shop which would have disposed of its waste in easily identified
drums.
Testing will have to be performed to determine if there is any correlation with particle
size, space, time or components in the waste.
3) CAN ANY WASTE COMPONENTS OR STRATA BE ELIMINATED?
Historical information indicated that 400 drums of construction debris were generated
during construction of a new warehouse. The information indicates that the virgin nature of the
materials may make these drums candidates for not sampling or less intensive sampling.
Likewise, the source of beryllium contamination is a beryllium sludge which exists in
drums by itself or in drums commingled with shredded packing material and laboratory wastes
that were generated during physical testing of the beryllium product. If the materials commingled
with the beryllium waste are known not to be a source of contamination, the commingled
Presented at EPA Workshop fl7 Jufy 16, 1992
'OansaersBig Heterogeneous Materials' C~41
-------
material can be discriminated against during sampling and only the beryllium sludge sampled
and the volume contribution of the commingled material noted.
4) ARE CONTAMINATION LEVELS CORRELATED WITH PARTICLE SIZE?
Some of the older beryllium sludge has dried and formed a cementaceous aggregate of
different particle sizes. Since the sludge is known to be homogeneous within a batch by process
knowledge and preliminary sampling data, sampling can be restricted to the more easily sampled
- smaller particles sizes.
5) IS CONTAMINATION INNATE OR SURFACE ADSORBED
The waste from the machine shop consists of varied material from fine metallic filings
to large chunks of metal and out-of-specification metal product. Since the only contamination
in the machine shop is solvents and cutting oils and the waste matrix is impervious, the
contamination is surface adsorbed in nature. Thus sampling of these wastes will consists of the
sampling of fines which will be subjected to extraction, wipe sampling of the large metallic
objects and notation of the volume contributions of the different particle sizes.
It is essential that all assumptions , (i.e. any correlations), be verified by at least
knowledge of the waste and preferably confirmed by exploratory sampling and analyses.
In the above hypothetical case, the proposed strategy for characterizing the 4000 drums
resulted in:
The identification of two large strata that constitute the majority of the waste (i.e. the
beryllium sludge and the solvent and cutting oil contaminated machine shop waste).
The elimination of the need to sample 10% of the drums, (i.e. the construction debris),
if preliminary testing verifies waste disposal information.
Presented at EPA Workshop IU July 16,1992
'Characterizing Heterogeneous Materials"
-------
Simplified sampling of the beryllium commingled waste by restricting sampling to the
beryllium sludge and not the other commingled materials.
Simplified sampling of the cementaceous beryllium sludge by limiting sampling to the
more easily sampled small particles.
Simplifying the sampling of the machine wastes since the source of contamination is
surface adsorbed and not innate to the waste materials.
A less expensive and doable sampling design which will result in a more precise estimate
of the mean and will generate strata specific information which will be valuable if the
strata are eventually treated separately.
The following subsections describe additional strategies that can be employed if the above
sampling strategies are not applicable to a waste or if they are applicable but by themselves will
not allow the project objectives to be cost-effectively met.
Modification of the Sample Preparation Method
As discussed in the Introduction, heterogeneity is analytical sample size dependent. The
greater the particle size and the greater the variance of the concentration of the target parameter,
the greater the heterogeneity for a given analytical sample size. To minimize the measured
heterogeneity and to accommodate large particle sizes, traditional sample preparatory methods
can be altered.
In the laboratory, the term "sample preparation" is commonly meant to include two
separate steps; 1) the subsampling of a field sample to generate an analytical sample, and 2) the
preparation of the analytical sample for subsequent analysis.
Regarding subsampling, the previously discussed logic for field sampling (refer to Section
3.1.1) is also applicable for the generation of analytical samples. That is, knowledge of
concentration distributions within the waste can be used to simplify subsampling by:
Praaaed at EPA Workshop m Jufy 16, 1992
'Characterizing Heterogeneous Materials'
-------
1) Eliminating any waste components or strata that do not contribute significantly to
the concentration of the target compound;
2) Discriminating against large particles and only select small particles if small
particles represent the waste as well as large particles; and.
3) Surface wiping larger particles and extracting or digesting fines if surface
contamination is the source of the target parameter.
If the above approaches are not applicable to a field sample, the field sample will have
to be subjected to particle size reduction (PSR) prior to subsampling or the sample preparation
method will have to be modified to accommodate the entire field sample.
PSR is useful for handling field samples, which have particles too large for proper
representation in an analytical subsample. The intent of PSR is to decrease the maximum particle
size of the field sample so that the field sample can then be split and or subsampled to generate
a representative subsample. The difficulties in applying PSR to waste samples are:
1) Not all materials are easily amenable to PSR (e.g. stainless steel artifacts);
2) Adequate PSR capabilities and capacities do not normally exist in environmental
laboratories;
3) PSR can change the properties of material ( e.g. leachability)
4) PSR can be a source of cross-contamination; and
5) PSR is often not applicable to volatile and labile compounds.
Presented al EPA Workshop m July 16, 1992
"Charaaeranf Heterogeneous Materials ' C-44
-------
Modification of sample preparative methods can include the extraction, digestion or
leaching of much larger sample masses than specified. The advantage of this approach is that
the resulting extract, digestate or leachate are relatively homogeneous which simplifies
subsampling. This approach is particularly important for volatile organic compounds which may
suffer from substantial losses if subjected to PSR. For volatile organic compound analysis, larger
portions of the wastes can be subjected to methanol extraction or possibly the entire field sample
could be subjected to heated headspace analysis as one sample or as a series of large aliquots.
Prior to modifying a sample preparatory method, especially a method associated with a
property such as the Toxicity Characteristic Leaching Procedure (TCLP), it is advisable to
consult the end-user of the data and the pertinent regulator if appropriate.
Modification of Analytical Method
The analytical phase of a sampling and analytical program allows another opportunity to
simplify the characterization of an excessively heterogeneous waste. Examples of different
classes of analytical methods are ;
Screening methods,
Portable methods,
Field Laboratories methods,
Non-intrusive methods,
Innovative methods, and
Fixed laboratory methods.
Screening, portable and field laboratory methods have the distinct advantage that they
allow for the cost-effective analysis of more samples. These methods not only generate more
precise data but the greater number of samples make it easier to detect correlations between
concentration levels and waste strata or components. Also some screening methods may analyze
a larger sample volume than what is traditionally submitted to a fixed laboratory.
Presented at EPA Workshop m July 16, 1992
'Characterizing Heterogeneous Materials' C-45
-------
Non-intrusive methods can be useful when there are health and safety issues regarding
exposure to the waste. These methods may also be used to qualitatively or semi-quantitatively
evaluate large volume wastes.
An EPA Document entitled "Characterizing Heterogeneous Wastes EPA 600/R-92/033"
has a solid discussion of non-intrusive methods and new methodologies that are being developed
for the characterization of heterogeneous wastes.
Modification of the Waste Handling, Treatment Disposal Plan
If the modifications discussed in the previous subsections are not applicable to a given
waste or when they are applicable but still do not allow the objectives to be cost-effectively met,
then the reasoning behind the original program must be examined. The original need behind the
waste characterization objectives has to be examined and an approach for simplifying the
characterization process must be devised.
For example, assume that the need behind waste characterization objectives for a certain
program was the common requirement to determine if a waste is hazardous prior to waste
disposal. An initial attempt to characterize the waste; 1) failed to meet the objective, 2) indicated
that the waste was excessively stratified, and 3) proved that portions of the waste are hazardous.
After reviewing this preliminary information and the costs to attempt a defensible
characterization of the waste, it could be decided that all the waste will be assumed to be
hazardous and treated as hazardous waste. Under this scenario, the needs change and now
compliance with the land-ban requirements may become the issue. Assume that the waste was
incinerated, then the less heterogeneous and more easily sampled incinerator ash would be
sampled in lieu of the original excessively stratified waste.
An other example would be a large laboratory operation that generates drums of gas-
chromatography (GC) vials containing dissolved standards, solvents and numerous contaminants
Presented at EPA Workshop mJuty 16, 1992
'OaracUrizing Heterogeneous Materials' C~46
-------
extracted from samples of different origins. Although, the vial contents were individually
analyzed by GC for certain parameters, the contents were not analyzed for enough parameters
to characterize the waste. A quick review of this waste would preclude further and cost
prohibitive analyses of every vial. Thus, an alternative approach would have to be devised. An
alternative approach could require that the vials be crushed and the vial contents and the solvent
used to rinse the broken vials would be collected in a receiving drum. This approach would
convert drums which contained hundreds of different vials, each of which could have their own
independent concentrations of contaminants, into two relatively homogeneous waste strata, i.e.
solvent rinsed glass and contaminated solvent.
CONCLUSIONS
The previous discussion reviewed heterogeneity issues and proposed a single and
systematic approach to the characterization of hazardous waste. This approach is based upon the
evaluation of wastes in terms of their degree and type of stratification - a factor which drives
the degree of difficulty in sampling a particular waste.
This discussion also proposed stratification in an additional dimension than the traditional
spacial and temporal dimensions. The stratification of waste according to waste components and
sources can simplify the characterization process and provide needed strata information.
This systematic approach and the additional mechanisms for stratifying waste is intended
to aid in the characterization of wastes especially those that are excessively stratified.
Presented a! EPA Workshop ffl Juty 16, 1992
'Characterizing Heterogeneous Materials' C-47
-------
APPENDIX D:
CHARACTERIZING MIXED WASTES
-------
CHARACTERIZING MIXED WASTES:
INTRODUCTION
Dr. James A. Poppiti
US Department of Energy
Germantown, MD
INTRODUCTION
Certainly, within the Department of Energy's RCRA and CERCLA related clean up
activities, mixed waste characterization presents some real challenges. By mixed waste, we
mean a mixture of radioactive and hazardous wastes generally in liquid or solid form and
combinations thereof. The presence of radionuclides in a hazardous waste sample presents
additional handling and procedural problems beyond those presented by non-radioactive
hazardous wastes. In addition, their presence limits the options available for managing the
waste. This, in turn, dictates the kinds of data needed to select and implement the best option.
The cost of analyzing hazardous constituents in mixed waste samples is considerably
greater than that for analysis of the same constituents in non-radioactive waste samples. This
increased cost of analysis makes it all the more important to give careful consideration to what
samples need to be collected. Determining what samples need to be collected depends on how
the data are going to be used. Only when you know precisely how the data are going to be used
can you then determine what kind and how much data need to be collected and the necessary
data precision, accuracy, and reliability. In most cases, taking extra samples and performing
unnecessary analyses is just wasteful, in terms of time and money. In addition, when dealing
with mixed waste, one is faced with significant safety problems. Accordingly, we should
routinely plan our data collection programs so that, before the first sample is ever taken, we
know precisely how we are going to use the data, what and how much data we really need, and
how good the data must be.
Presented an Jufy 16. 1992 a! EPA Workshop IV
on 'Characterizing Mixed Wastes' D~l
-------
CONCERNS
In this session, we will be looking primarily at two concerns.
The first is: if the sample is so "hot" (i.e., radioactive) that any amount of handling
would pose a safety risk, then how does one characterize the waste for purposes of determining
the proper means of treatment, storage, or disposal?
The second is: given the limitation of possible treatment, storage, and disposal options
for mixed waste, there may be only selected types of analyses that need to be performed. If the
samples have significant amounts of radioactivity, then what properties (in addition to
radiological hazards) need to be determined in order to properly characterize the material with
respect to the appropriate options for treatment, storage, or disposal? A sub-part of this concern
is: how do we go about minimizing characterization costs while maintaining a high level or
environmental protection?
PURPOSE OF TODAY'S SESSION
We will use today's session as a forum for exploring:
• How to develop waste analysis plans tailored to the characteristics of the waste
materials and the available treatment, storage, or disposal options,
• How to go about balancing environmental risk and characterization costs,
• Options for mixed waste management (i.e., treatment, storage, and disposal),
• Potential hazards posed by mixed wastes.
The options for mixed waste management that should be considered in the three sessions
include:
Presented on July 16. 1992 at EPA Workshop IV
on 'Characterizing Mixed Wastes'
-------
• Warehouse storage in drums,
* Entombment in a below-ground cavern,
• Incineration,
• High temperature glassification, and
• Solidification using cement, asphaltic, or lime processes.
WORKSHOP ORGANIZATION
This issue session will consist of overview presentations by our three experts. First,
Susan Jones will present the regulatory point of view. She will be followed by Craig Leasure,
who will give the industrial point of view. Then, Wayne Griest will give us an overview of the
current state-of-the-science in the area of mixed waste characterization.
Following these presentations, you will have an opportunity to ask clarifying questions.
Then, we will divide into three workgroups to address the issue. The purpose of these
workgroup sessions is to solicit your ideas and recommendations that EPA can use in addressing
the mixed waste characterization issue.
We would like for each of the three workgroup sessions to focus on the following
questions:
1. What kind of information is needed (and what quantity and quality) in order to safely
manage a mixed waste for at least each of the following management options:
• Entombment in a below-ground cavern,
• Incineration,
• Solidification using cement, asphaltic, or lime processes?
Presented on July 16,1992 at EPA Workshop TV
on 'Qxsraaerizsnf Mixed Wastes' D-3
-------
2. What scientific properties or constituents of the mixed waste should be determined?
Examples include:
• Physical form (e.g., liquid content),
• Radiological properties,
• Physical properties (e.g., strength, heat content, flammability),
• Leachability,
• Chemical properties (e.g., corrosivity),
• Metals,
• Organics, and
• Ionic species (e.g., CN, SO4, Cl).
3. How should one go about balancing risk and characterization costs? Of course, in
situations where there are tradeoffs between characterization costs and total project costs,
then our concern would be with total project costs. (For example, there have been cases
where, by spending an additional $200,000 to do a more precise characterization, the
total project costs were reduced by several million dollars). It should be clear to all of
you that we are not attempting to balance costs versus benefits - rather we want to
explore how to go about minimizing costs while maintaining a high level of
environmental protection. The EPA has had considerable success in this area through
application of the data quality objective (DQO) process.
Each workgroup has a facilitator and a scribe. After you have completed your
deliberations, we will turn over the results to EPA to consolidate the information into a report
summarizing the thoughts, ideas, and conclusions presented. We anticipate sending a copy of
the report by the end of the year. In addition, EPA will use this information when formulating
its approach to the problem. Similarly, DOE will make full use of this information.
Presented at Jufy 16, 1992 ai EPA Workshop TV
on 'Characterizing Hutd Wastes' D-4
-------
CHARACTERIZING MIXED WASTES:
REGULATORY PERSPECTIVE
Susan Jones
United States Environmental Protection Agency
State and Regional Programs Branch
401 M Street SW
Washington, DC 20460
INTRODUCTION
Subtitle C of the Resource Conservation and Recovery Act (RCRA) is a "cradle to grave"
hazardous waste management program (from waste generation to ultimate disposal) that places
importance on proper waste characterization by facilities that generate, treat, store or dispose
of hazardous waste. Accurate characterization of waste by facilities determines if and/or how
wastes will be managed under RCRA. Proper characterization also is the first step in ensuring
the regulator that the waste will be managed under RCRA in a manner protective of human
health and the environment. A major component of waste characterization is waste analysis.
Waste analysis via testing is often the most reliable, accurate and defensible way to characterize
hazardous wastes. However, waste testing is not always required, or desirable for certain waste
streams. This is especially true for radioactive mixed waste which poses testing and sampling
problems because it contains both radiological and chemical hazards. Testing mixed waste is
further complicated by virtue of being regulated under two different statues (the Atomic Energy
Act (AEA) and RCRA) with two distinct sets of management requirements.
Presented on Jufy 16, 1992 at EPA Worbhap IV
on 'QuraOerizbig Mixed Waiter' D-5
-------
STATEMENT OF ISSUES
This discussion involves a review of what Environmental Protection Agency (EPA), as
a regulator, regards as the essential waste characterization requirements under RCRA, why EPA
requires such information, and what concerns EPA has in sampling and testing mixed waste
streams.
Hazardous Waste Analysis Requirements for Generators
Waste characterization begins at the point of generation of a solid waste. Initially, a
generator is tasked with determining whether his or her solid waste is a hazardous waste that is
regulated under Subtitle C of RCRA (see 40 CFR 262.11). This may simply involve applying
knowledge of how the waste is generated, knowledge of the hazardous characteristic of the
waste, or matching the waste to a hazardous waste listing that describes the waste and, for
certain listings, the process that generates it. If the generator lacks data or process knowledge
of the generated waste stream, he or she may be obligated to test or send a representative sample
off-site for analysis. For purposes of characterizing mixed waste, EPA requires that the
hazardous portion is properly identified by process knowledge or testing. If more than one
hazardous waste is in the radioactive waste stream, then all hazardous wastes must be identified,
and if shipped off-site, reported on the manifest. Any records of such waste identifications,
through waste analyses, testing, or determinations made on process knowledge, must be retained
for at least three years (40 CFR 262.40) as proof of compliance with RCRA.
Besides waste identification, waste analysis serves to ensure that incompatible wastes are
not stored in the same management unit and the composition of the unit is not incompatible with
the waste stream to prevent hazards. Waste analysis by generators also plays an important role
in meeting the land disposal restriction (LDR) requirements discussed below.
Presented en Jufy 16, 1992 a EPA Vforiahap IV
on 'Qvsmcteriziitg Mixed Waita' D~6
-------
Hazardous Waste Analysis by Treatment, Storage, or Disposal Facility (TSDF)
Owner/Operators
Owners/operators of hazardous waste treatment, storage, and disposal facilities also must
accurately identify hazardous waste that they manage. However, unlike generators, TSDFs are
specifically required to obtain a chemical and physical analysis of a representative sample of the
wastes managed on site or received from off-site, and formally document analysis procedures
used in a waste analysis plan (WAP). These analyses may include, in part, data developed by
the generator, existing published or documented data on the hazardous waste, or data on waste
generated from similar processes. The requirements and frequency of such analyses are
specified in the waste analysis plan and may be modified to account for hazards presented by
the waste stream. For mixed wastes, the waste analysis requirements afford the flexibility to
reduce testing frequency, modify testing techniques, and incorporate the principles of ay low as
reasonably achievable (ALARA).
It is important to the regulator that TSDFs gather accurate and current analysis data to
ensure that wastes that are analyzed are indeed those the facility can treat, store and dispose as
specified by their permit, and are compatible with the processes and units used at the facility.
Also important, is the analysis of waste shipments from off-site to verify the identity of the
waste on the manifest and verify that the land disposal restriction (LDR) notification of 40 CFR
268.7 is accurate.
Special Waste Analysis Requirements for the Land Disposal Restrictions (LDR)
In addition to the initial waste characterization, RCRA requires some degree of analysis
for determinations by generators that their waste is restricted or not restricted under the LDRs,
determinations by generators, treaters, or disposers that a land disposal restricted waste meets
a required treatment standard, and verifications by off-site facilities through corroborative testing
that the hazardous wastes received are in compliance with LDR treatment standards and
Presented on July 16. 1992 at EPA Workshop IV
en 'Charaaercijig Mixed Wastes' D~7
-------
prohibitions. Process knowledge may be used for the LDR determinations required of
generators (i.e., determinations whether a waste is land disposal restricted and determinations
that a waste meets treatment standards). However, some amount of testing is required for
treaters and disposers to verify that the treatment standards have been met prior to disposal,
unless the treatment standard is a specified technology.
The addition of the land disposal requirements has increased the waste characterization
burden on facilities, which in the case of mixed waste, has raised special concerns. In
particular, there is concern of increased exposure risk to workers from the additional sampling
and analysis of mixed wastes. To address this, EPA has emphasized in the LDR Third Third
rule (55 FR 22669, June 1, 1990) and the draft guidance entitled "Clarification of RCRA
Hazardous Waste Testing Requirements for Mixed Waste" that the RCRA permit writer has the
flexibility hi the WAP, on a case-by-case basis, to minimize incremental exposure to radiation
by reducing the frequency of testing mixed waste and allowing more reliance by the TSDF on
analysis information provided by the generators or treaters of mixed waste.
Purpose and Use of Waste Analysis Information
As mentioned earlier, waste analysis is important in determining whether a solid waste
falls within the scope of RCRA and how a waste will be managed once it identified as RCRA
hazardous. EPA's primary goal in requiring some level of waste analysis is to ensure that the
mixed waste is managed in a safe operating environment and in a manner that is protective of
human health and the environment. EPA uses tools, such as the records kept by generators on
waste analysis and a facility's WAP, to verify that the data collected and analyzed by facilities
is of high quality, to track the movement and management of hazardous waste by ensuring that
all parties managing a waste stream have identical information, to determine compliance with
RCRA permit conditions and to collect information for enforcement initiatives and legal
proceedings. These tools also may benefit generators and facilities to plan waste management
activities, provide consistent procedures for analyzing wastes, ensure that information is not lost
Presented on July 16, 1992 at EPA Workshop IV
on 'OtaraOfrKing Mixed Wastes'
-------
during personnel changes or absences, and minimize potential mismanagement of hazardous
wastes.
Special Concerns in Sampling and Testing Mixed Waste
The regulated community has communicated to EPA (and the Nuclear Regulatory
Commission) their difficulty in meeting RCRA testing requirements. Problems include
difficulties in using EPA test methods, choosing sample sizes, obtaining a representative sample
for heterogenous waste, determining the number of samples needed to establish the homogeneity
of waste, and complying with ALARA in conjunction with complying with RCRA analysis
requirements. To address some of these mixed waste testing concerns, EPA and NRC recently
issued a joint draft guidance document entitled "Clarification of RCRA Hazardous Waste Testing
Requirements for Mixed Waste," which provides an overview of the RCRA testing requirements
and addresses concerns related to potential radiation exposure, redundant verification testing, and
sample size needed for an accurate test. The guidance emphasizes reliance on process
knowledge by generators when feasible, use of surrogate materials for high activity wastes that
represent the hazardous properties of the waste, and the use of permit conditions which
incorporate ALARA considerations to minimize or eliminate redundant testing by off-site
facilities. In addition, the guidance points out that flexibility exists in mixed waste sampling and
analysis. For example, choosing a sample size smaller than the 100 gram minimum sample size
recommended in a Toxicity Characteristic Leaching Procedure test protocol is allowable if the
results show that the test is sufficiently sensitive and measures the constituents of interest at
regulatory levels prescribed by the test. EPA and NRC are currently reviewing comments on
the guidance which is scheduled for release in the fall.
ADDRESSING THE ISSUES
Much work and coordination with interested parties has to be done to iron out all the
issues associated with testing mixed waste. In the near term, EPA is interested in further
Presented on Jufy 16, 1992 al EPA Wortshcp IV
on •Oaracteraing Mixed Wastes' D-9
-------
investigating the role that reliance on process knowledge might play in limiting exposure from
testing, the appropriate types of information that should serve as a basis for the process
knowledge determination, and the reasonable limits on using process knowledge so it is not
misused. We are also interested in determining the appropriate framework for applying ALARA
considerations to adjust or optimize the sampling frequency and waste analysis parameters in a
RCRA facility's WAP. Furthermore, we want to gain knowledge of the other ALARA concerns
or practical consideration associated with mixed waste testing which are not mentioned here or
in the draft guidance. Finally, we hope to explore additional scenarios where surrogate materials
may be characterized in lieu of running RCRA tests on specific waste streams that may pose a
significant radiation exposure potential.
CONCLUSION
Analysis of hazardous wastes is an integral part of the RCRA program that EPA. uses to
meet its mission in providing protection to human health and the environment through safe
management practices. Although, many problems exist in characterizing mixed waste, EPA
feels that accurate characterization of mixed wastes may be successfully accomplished within the
confines of RCRA, using minimal testing. Through the waste analysis plan and other regulatory
avenues, such as case-specific permit conditions, we should be able to have the flexibility to
address mixed waste testing issues as they arise. However, this may only happen through
cooperation of the regulators with the regulated community in determining reasonable approaches
that offer minimal risk, while meeting the objectives of waste characterization.
Presented at Jufy 16, 1992 a EPA Workshop IV
en 'Characterizing Mixed Wastes' D-10
-------
CHARACTERIZING MIXED WASTES:
INDUSTRY PERSPECTIVE
C. S. Leasure
Environmental Chemistry Group (EM-9)
Los Alamos National Laboratory
Los Alamos, NM 87454
INTRODUCTION
Physical and chemical characterization of radioactive mixed wastes (RMW) is necessary
for determination of appropriate treatment options and to satisfy environmental regulations.
Radioactive mixed waste can be classified as two main categories: contact-handled (low level)
RMW and remote-handled RMW. This discussion will focus mainly on characterization of
contact-handled RMW.
The characterization of wastes usually follows one of two pathways: 1) characterization
to determine necessary parameters for treatment or 2) characterization to determine if the
material is a hazardous waste. However, wastes sometimes can be declared as hazardous waste
without testing and then treated as hazardous waste.
STATEMENT OF ISSUES
Characterization of radioactive mixed wastes poses some unique issues that will require
special solutions. Below, five issues affecting sampling and analysis of RMW will be discussed.
Impact of Screening Requirements on Holding Times
Radioactive screening of samples can be a time-consuming process that can impact
holding times. Holding time requirements must be modified to ensure that laboratories handling
Proofed on July 16. 1992 at EPA WoHahop IV
on 'Charucurizing Mixed Wastes'
-------
RMW samples can meet holding times without compromising either the health and safety of the
workers, Nuclear Regulatory Commission (NRC) regulations, or the ability of the analysis to
accurately reflect the bulk concentration of the target analytes. Currently, Environmental
Protection Agency (EPA) specified sample holding times may not accurately reflect the amount
of time that samples can be stored without compromising the analysis results. However, many
regulators will not accept data from an analysis on a sample that misses holding time even
though the concentration of target compound that would cause action is one to two orders of
magnitude greater than the detection limit of the method. For example, the holding time for
water samples for semivolatile organic analysis (Method 8270) is 7 days (1). For a compound
such as naphthalene, which has an action level related to land disposal near 60 parts-per-billion
(2) and estimated quantitation limit of 10 parts-per-billion, sample degradation must be extensive
before reaching a level where a decision is affected. The amount of time for degradation of
target compounds to reach one order of magnitude may be months or years when the sample is
preserved and stored at low temperature (3).
Radioactive screening can be extensive and time consuming, as discussed below. For
samples that are suspected of containing radioactive materials, screening must be conducted to
determine the gross amount and type of radioactivity for safe handling and shipping purposes.
All government and industrial facilities that handle radioactive materials follow strict rules for
their handling, either NRC requirements (4) or Department of Energy (DOE) Order 5480.11 (5).
First, screening must be conducted at the sampling site to immediately determine what level of
personal protection for field personnel is adequate to handle the samples. This first level of
screening usually is performed with hand-held devices with detection levels appropriate only to
measure levels of radioactivity that could impact workers. The second level of screening, which
cannot be met with hand-held screening devices, is required to ensure that Department of
Transportation packaging requirements (6) and off site contract laboratories' radioactive materials
license limits as specified by the NRC are met. This level of screening can entail up to four
different analyses (gross alpha/beta, gross gamma, gamma spectroscopy, and liquid scintillation
counting) depending on the expected radioactive content of the sample (Knowledge of Process).
Presented at July 16, 1992 at EPA Workshop IV
on 'Cfarazterizuig Mixed Wastes' D-12
-------
Only gamma spectroscopy and liquid scintillation counting can provide definitive isotopic
information. Consequently, while the radioactive screening is being conducted, sample cuts for
chemical analysis are stored in a fashion to minimize worker exposure (usually a combination
of shielding and distance from personnel).
These two levels of screening require time, even when systems are in place to
expeditiously perform this screening. Once the data are generated, decisions can be made as to
how the samples must be packaged for transport and where the samples can be analyzed (which
laboratory, and the impact on radioactive materials license limits). These decisions are
dependent on the knowledge of expected isotopes, because even different isotopes of the same
element can require different handling or have different license limits. For shipping, the worst
case isotope expected can be used in the Department of Transportation packaging determination.
A worst case isotope can also be used for the materials license impact determination; however,
for the same sample the specific isotope may be different than the worst-case DOT isotope. If
the specific isotope or isotopes cannot be estimated, then isotope-specific analyses must be
conducted. Isotope-specific analyses generally require significant time (ca. weeks). For
example, the analysis of strontium-90 is indirect, meaning that the daughter, yttrium-90, is
separated from the sample and yttrium-90 is allowed to grow in for approximately 4 weeks. At
that time, the yttrium-90 is again separated from the strontium-90 and the yttrium-90 is analyzed,
providing data to calculate the amount of strontium-90 based upon the half-life of strontium-90.
One caution, however, must be made in regard to gross screening of radioactivity.
Sometimes the radioactive content in samples can be scattered in a non-uniform manner, which
can cause significant problems with the screening data. For example, cores from drilling
operations are usually subsampled with a cut from the core selected for radioactive screening.
This selection can be done using hand-held devices for guidance to pick the worst-case spot.
However, for alpha contamination, a hot spot could be present and not detected. (Alpha
particles can be attenuated from the bulk material and that portion of the core could contain
several orders of magnitude more radioactive material than was determined from the gross
Presented on Jufy 16, 1992 at EPA Workshop IV
on 'Oiaracterizing Mixed Wastes' D~13
-------
screening. Therefore, the gross screening data may not always be indicative of the bulk or
localized radioactivity. RMW samples that comprise more than one phase are an additional
complication.
Even without the added complexity of heterogeneity of the radioactive content in RMW
samples, the time required for screening prior to shipment of samples can be as little as several
hours to as much as several weeks. For samples requiring analyses where holding times are a
requirement (many organic analyses and some inorganic analyses) .the impact to the laboratories
from reduced holding time windows can be significant. Therefore, for RMW samples, holding
times that make technical sense must be determined and assigned.
Management of Treatment Options
For radioactive mixed waste, Resource Conservation Recovery Act (RCRA)
characterization should not be required prior to treatment, if the radioactive content is the driver
as far as the type and extent of initial treatment. For example, for a waste that contains
significant quantities of fission products, the only treatment option may be vitrification.
However, if this waste also contains trace levels of RCRA hazardous solvents, a traditional
RCRA hazardous waste determination, which will cause exposure of the workers to radiation,
will likely not change the treatment of the material. If one were to conduct a risk assessment
on this scenario, the significant risk in the entire process would be the exposure of the workers
performing the RCRA sampling and analysis. Therefore, a RCRA analysis or analyses prior to
treatment would not provide any value-added data. What is likely to be more important from
the treatment perspective, in this case, would be the BTU content of the waste. The
determination of the BTU content of the waste is a simple process, which can be safely
performed on RMW with minimal worker exposure. For RMW where the treatment process
is designed for the radioactive material (since the radioactive material is usually the risk driver),
RCRA or any other environmental compliance analysis only should be performed on RMW after
treatment. One caveat on this statement should be that this is appropriate only if the treatment
Presented on Juty 16, 1992 at EPA Wortshcp TV
on 'Oiaraaercung Mixed Wastes' D-14
-------
process (or facility) is designed to meet requirements for containment of any hazardous or
radioactive substances generated, such as offgas.
Generation of Secondary RMW
When performing RCRA or Contract Laboratory Program analyses on samples that are
known to be radioactive and that are suspected of being hazardous (suspect RMW), some of the
analyses will generate RMW because the required solvents are themselves hazardous or because
the matrix is spiked with hazardous materials. One example of this problem is the extraction
of semivolatile organics from waste using methylene chloride. If a suspected RMW is extracted
with methylene chloride, that suspect RMW is now definitely radioactive mixed waste from the
methylene chloride residue. Even if the waste, prior to analysis, was actually radioactive only,
and no semivolatile organics were present in the waste before analysis, the material becomes
RMW as a result of the analysis. This is a case where more RMW is generated from the
analysis than was initially present.
This secondary RMW must then be managed. However, the secondary RMW can take
additional forms, such as filters, disposable wipes, gloves, and other consumables generated
from not only the analysis, but also from sampling. Currently, the only treatment, storage, and
disposal option that may be available to commercial laboratories performing analyses that
generate RMW is to capture these wastes and package and ship them back to the source of the
samples for ultimate treatment, storage, or disposal. At this time, alternative treatment options
for Land Disposal Restricted mixed waste is extremely limited. The added costs to resolve this
problem could exceed the actual cost of analysis. Two things are needed to minimize the
problem of secondary RMW generation. First, new environmental compliance analyses must
be developed that do not themselves use materials that are considered hazardous waste when
disposed. Second, Knowledge of Process must be used as much as possible to minimize the
collection of samples and performance of analyses, which themselves can create RMW.
Presented on July 16, 1992 at EPA Workshop IV
on 'Otaracurcang Uaed Wastes' D~15
-------
Another potential problem, not necessarily a current issue for regulators, but one that
could become a significant issue later, may be the question of who will be responsible for the
final decontamination and decommissioning of commercial laboratories that perform analyses on
RMW.
Detection Limits
Most EPA recommended or required analyses use relatively large amounts of sample
material, then digest or extract and preconcentrate to obtain low detection limits. However, the
DOE and Nuclear Regulatory Commission ALARA (As Low As Reasonably Achievable)
principle requires laboratories to handle as small amount of sample as possible. Consequently,
for samples with moderate to high levels of radioactivity, small amounts of sample are handled
that satisfy health physics protection requirements but cause detection limits to be higher than
for typical hazardous waste samples. For example, Oak Ridge National Laboratory published
a method for the determination of semivolatile organics in radioactive tank waste (7). Because
of ALARA considerations, they used only 20 mL of sample for extraction rather than the
recommended 1000 mL. Hence, their detection limits for most of the target compounds were
approximately 50 times higher than for a typical hazardous waste sample. Some latitude in
meeting detection limits should be given in cases where ALARA (personnel safety)
considerations dominate.
What is needed is guidance on the relative risks posed by hazardous and by radioactive
materials. Ultimately, one could foresee a table or graph that one could use to evaluate the risk
from the level of radioactive material vis-a-vis the risk from the level of hazardous material so
that the true driver of risk could be assessed. Once the true driver is established, one could
determine sample designs and analysis schemes that would determine needed actions. However,
this is a much more complicated issue than one would first imagine. If one tries to develop the
plot, multiple exposure scenarios must be explored. For a single exposure scenario and for a
Presented en July 16, 1992 al EPA Worluhcp IV
on 'OiaraaeTizing Mixed Wastes' D-16
-------
simple waste that contains one radioactive substance and one hazardous substance, a plot could
be developed.
At each end of this plot, either the radioactive substance or the hazardous substance drive
the risk. Where this gets very complicated, is in the vicinity where the two lines intersect,
where the total risk is a combination of both the radioactive and hazardous substances. As can
be seen, this is very complicated when only two components are present; with more components,
it becomes a significant problem. Rather than try to provide specific guidance, a better
suggestion would be broad guidance such that if the expected risk from the radioactive content
of the waste (based upon knowledge of process or screening) is expected to be greater than 10
times the risk posed by the hazardous portion of the waste, the need for strict characterization
of the hazardous waste is unnecessary or can be reduced. However, if the risk posed by the
hazardous portion of the waste is expected to be less than 10 times or greater than the risk posed
by the radioactive portion of the waste, then strict characterization of the waste may be
appropriate.
In the case where strict characterization of the hazardous portion is not indicated,
allowances should be made for limited characterization that does not adversely impact the health
and safety of the workers.
Measurement of Properties Appropriate for Management
When characterizing RMW, only those treatment properties that are appropriate for the
management of that RMW should be measured. For example, the Toxicity Characteristic
Leaching Procedure was designed with the municipal landfill model in mind. For the
Department of Energy (DOE) however, RMW will never be placed into a municipal landfill.
RMW will either be incinerated, vitrified, or placed in long-term storage. Therefore,
measurements appropriate for the selected treatment scheme or those properties necessary to
make a treatment decision should be measured. For vitrification or incineration, the BTU
Presaaed on July 16, 1992 at EPA Workshop IV
on 'Characterizing Mixed Wastes' D-17
-------
content, or only those substances that may affect the efficiency of the process or final product
should be measured prior to treatment. For long-term storage of wastes that have been vitrified,
incinerated, or not treated, only those properties that affect the management of the long-term
storage facility (e.g. any mobile substances) should be measured. Since the likely long-term
storage option for those materials may be in a salt mine, then the teachability of the materials
by brine would be a more appropriate measurement.
WHEN IS A MATERIAL RADIOACTIVE?
All parties involved in the treatment, storage, or disposal of radioactive or RMW
materials must agree on a determination as to when a material is considered to be radioactive.
A proposal for a definition that could withstand scrutiny both from a risk basis as well as one
that is in place already would be 2 nCi/g. One of the fundamental issues that must be addressed
when considering radioactive mixed waste is the fact that there are clear definitions and rules
regarding when a waste is considered to be hazardous. However, there are currently no clear
rules or definitions as to when a waste (or any material) is considered to be radioactive. The
definition can range from one atom of added radioactivity (current DOE waste management
philosophy) up to 2 nCi/g , which is a Department of Transportation definition of radioactivity
as it applies to shipment (8). With the fact that treatment, storage, and disposal options for
RMW are extremely limited, it would seem prudent to determine a de minims level for
radioactivity for determination that a waste is radioactive.
CONCLUSIONS
The five issues that affect the sampling and analysis of radioactive mixed wastes
discussed above are likely only a small part of the overall issues revolving around RMW.
However, if some of these issues can be resolved or if we can move toward a successful
resolution of these issues, the ultimate goal of protecting our workers, the public, and the
environment can come closer to realization.
Presented on Jufy 16, 1992 at EPA Workshop IV
on 'CtamaeriringMixed Wastes' D-lo
-------
REFERENCES
1. "Test Methods for Evaluating Solid Waste Physical/Chemical Measurements," U.S.
Environmental Protection Agency, Third Edition, Volume IB, 1986.
2. Code of Federal Regulations, Tide 40, Part 268.43, Table CCW, 1992.
3. Maskarinec, ES&T (I'm tracking down this reference)
4. Code of Federal Regulations, Title 10, Part 2-C, 1992.
5. Radiation Protection for Occupational Workers, Order 5480.11, Department of Energy,
December, 1988.
6. Code of Federal Regulations, Title 49, Part 173.417, 1992.
7. Tompkins, B.A., Caton Jr., Fleming, G.S., Garcia, M.E., Harmon, S.H., Schenley,
R.L., Treese, C.A., Griest, W.H. Anal. Chem. 1990, 62, 253-257.
8. Code of Federal Regulations, Title 49, Part 173.403, 1992.
Presented on July 16, 1992 at EPA Workshop IV
on 'Oaraaeriang Mixed Wastes' D~19
-------
CHARACTERIZING MIXED WASTES:
SCIENTIFIC PERSPECTIVE*
W. H. Griest and J. R. Stokely, Jr.
Analytical Chemistry Division
Oak Ridge National Laboratory
Oak Ridge, Tennessee 37831-6120
INTRODUCTION
This paper is concerned with the following issues concerning the physical and chemical
characterization of radioactive mixed wastes (RMW): (1) what should be determined and how;
(2) the applications and limitations of current analytical methodologies, (3) promising new
technologies, and (4) areas where further methodology research is needed. Constituents to be
determined, sample collection, preparation, and analysis are considered. The scope concerns
mainly low level and very low level RMW that allow contact handling and analysis by Nuclear
Regulatory Commission or Agreement State-licensed commercial laboratories. The paper also
discusses high level RMW that will be characterized in laboratories with special shielded or
contained facilities.
WHAT SHOULD BE DETERMINED?
The characterization of RMW is driven by regulatory and treatability requirements.
Currently, the regulatory requirements are defined jointly by the Environmental Protection
Agency (EPA) and the Nuclear Regulatory Commission (NRC) under the Resource Conservation
Research sponsored by the Office of Technology Development, Laboratory Management
Division under US Department of Energy Contract DE-AC05-04OR21400 with Martin Marietta
Energy Systems, Inc.
Preserved on Juty 16, 1992 at EPA Workshop TV
on 'Characterizing Mixed Wastes' D-20
-------
and Recovery Act (RCRA) and the Atomic Energy Act, respectively, with enforcement of both
delegated to states having EPA Authorizations and NRC Agreements. Specific constituents and
concentrations or activities are specified. From a scientific viewpoint, the Toxicity
Characteristic Constituents and the toxicity characteristic leaching procedure (TCLP) scenario
of waste co-disposal with municipal refuse in a landfill are not necessarily relevant to RMW
disposal. The source terms (the wastes or their treated forms) and the localities of release to the
environment (leaching and migration to drinking water supplies) are very different for RMW
than for the solid wastes that RCRA was intended to regulate. Making radioactivity a Toxicity
Characteristic and setting de minimis levels of activity would remove the uncertainties leading
to current "no added radioactivity" policies which classify all hazardous wastes as RMW until
they are proven otherwise.
Regulatory agencies should permit more realistic leaching procedures for RMW which
reflect their ultimate forms and modes of disposal or storage. Simultaneously, it should be a
high priority in the Department of Energy (DOE) to develop the scientific understanding of the
leaching behavior and ultimate fate of the leached compounds in the different waste disposal and
treatment options. For high level wastes, this should include evaluations of the changes in waste
composition and properties over time which could affect the long-term waste performance (e.g.,
decay of radionuclides and radiolysis of organics). This knowledge could considerably simplify
and "target" the waste characterization needed for evaluation of waste form aging and
environmental behavior. A good example is the concentrated brine leaching which will be
evaluated at the Los Alamos National Laboratory (LANL) for assessing the suitability of
radioactive wastes for disposal in the Waste Isolation Pilot Plant. Leachability and fate are
included in the studies.
The constituents and properties that should be measured for waste treatability
requirements are complex and not fully understood. The constituents and waste properties that
affect the waste form for the major treatment technologies under consideration obviously must
be identified, and concentration and property range "windows" acceptable to the treatment
Presented on July 16, 1992 at EPA Vfarbhop TV
on 'Oiaraaercnig Mixed Wastes' D-21
-------
processes must be determined. These will define the RMW characterization requirements from
the treatability standpoint. An integrated approach was used by Rockwell International in
selecting constituents and establishing detection limits for treatability and regulatory analyses of
waste candidates for grouting (1). "Predictor" (or "group indicator") analyses and process
knowledge also were included in their approach.
THE ROLE OF PROCESS KNOWLEDGE
The RCRA allows process knowledge to be used for characterization where a well-
defined process generated the waste. For RMW, process knowledge is of limited use, because
low level wastes generally are not well-defined in composition and data for contaminated
environmental samples or for wastes generated from the cleanup or other handling of high level
wastes is usually nil. Historical records on old wastes (both low and high level) are usually
sketchy at best. The main use of process knowledge in such cases appears to be in assisting
characterization efforts by eliminating some constituents from analysis. For example, although
many of the inactive nuclear waste storage tanks at the Oak Ridge National Laboratory (ORNL)
lack useful source records, herbicides have been excluded from their RMW characterization
because the total amount of herbicides used at the lab (estimated from laboratory purchase
records) would not exceed the TCLP regulatory levels if disposed entirely in the tanks. This
exclusion does not hold for all DOE sites.
HOW CAN THE DETERMINATIONS BE MADE?
The following section discusses sampling and analytical issues, and offers suggestions for
resolving some of these problems.
Prcsaaed on Jufy 16, 1992 at EPA V/orbhop IV
on 'OaroaercsKf Mixed Wastes' D~22
-------
Sample Collection
Collection of a representative sample of a waste is critical to generating valid
characterization data. For low level wastes, sampling is conducted similarly to that for non-
radioactive wastes but with the additional precautions and constraints of the ALARA principle
(keeping radiation exposure and contamination of personnel and facilities as low as reasonably
achievable). The main area requiring advances in strategy and technology for low level wastes
is in sampling heterogeneous wastes. Drums or bags containing discrete items such as towels,
gloves, bottles, and metal parts, or tanks with multiple phases (e.g., oily, aqueous, and sludge
layers) are obviously heterogeneous. Other wastes may lack obvious heterogeneity, such as
single phases with composition gradients (as observed in an ORNL waste tank)(2). The current
options for sampling heterogeneous wastes include segregation and subsampling, multiple
sampling, or whole waste homogenizing or compositing and sampling. Heterogeneity also is a
problem in sampling high level wastes (e.g., stratified wastes in the Hanford tanks). For
situations where there are only a few heterogeneous samples to be analyzed, these approaches
are feasible, but laborious. Where there are many samples to be collected and analyzed, current
technology is not practical.
Physically collecting the sample, let alone sufficient amounts or representative portions
of samples from locations with restricted access is a major technology area requiring
development. The presence of high radiation fields (a major problem with high level waste
tanks at some DOE sites such as the Hanford reservation) further complicates the problem.
Volatile organics are a class of compounds particularly difficult to sample properly under such
conditions. Radiation-hardened samplers with articulated arms are under development but are
critically needed now. Development of improved sampling equipment and their acceptance by
regulatory agencies should be expedited.
Praaaai on Jufy 16, 1992 at EPA Workshop IV
on 'OamaerizBig Head Wastes' D-23
-------
General Analytical Consideration
Although low level wastes are not simple to prepare and analyze, there are no unique
"show stoppers" in their characterization. Regulatory sample preparation and determination
methods such as those listed in the SW-846 manual (3) can be performed essentially without
modification in radiochemical hoods, and the measurements can be conducted per the specified
protocols with the instruments located in contamination zoned laboratories. The main
operational difference from work with nonradioactive samples is the use of appropriate
precautions in sample and extract handling and laboratory waste management to observe the
ALARA principle and to minimize the spread of radioactive contamination. The major problem
in the characterization of low level wastes will be the disposal of contaminated laboratory
wastes. Often, secondary laboratory wastes generated in the preparation of RMW for analysis
and in the clean-up of hoods after sample preparation can exceed the mass or volume of sample
analyzed. This shows the need for the development and regulatory acceptance of "less waste
intensive" sample preparation methods to reduce analytical costs.
Sample Handling
Low level samples can be contact-handled in radiochemical hoods. Higher activity wastes
require handling and analysis with containment or shielding, which does not limit sample
operations and productivity. The permissible activities for radiochemical hood operations
currently vary in magnitude as well as units of specification among the DOE laboratories.
Generally, they range from about 0.1 to 45 jiCi (microCuries) of activity and up to 1 gram of
total Pu in solution for alpha emitters, and dose rates of 10 to 500 mrem/hr at contact for
beta/gamma emitters. Radiotoxicity and the complexity of the operation (and therefore the
potential for a spill or other release of contamination) are also considered in health physics
guidelines. For samples containing higher activities than permitted in radiochemical hoods,
samples with alpha emitters are handled under containment in glove boxes to prevent the spread
of highly radiotoxic nuclides whose radiation has little penetrating power. In contrast, a hot
Presented an Jufy 16, 1992 at EPA Workshop TV
on 'Cfaracterxing Mined Wastes' D~24
-------
cells (or, at least, sample shielding with lead bricks) is required for the much more penetrating
gamma (and beta) emitters.
Productivity
Almost any operation that can be performed on a bench or in a hood can be conducted
in a glove box or hot cell. However, the methods and equipment must be adapted for the latter,
highly skilled operators are required to safely perform the work, and productivity is reduced in
the much more restricted working environment. With glove boxes, samples and reagents are
bagged in, and prepared extracts and wastes are bagged out through sealed bag ports to prevent
direct contact of the glove box atmosphere with the laboratory air. Sample manipulations are
performed by reaching through heavy walled rubber gloves, which severely restrict manual
dexterity and freedom of motion. Sample introduction into hot cells is equally difficult, and in
addition, the work is performed using master-slave remote manipulators which require a high
degree of eye-hand coordination. Depending upon the complexity of the operation, productivity
can be decreased (or conversely, costs are increased) by factors of 2 to 5-hold for glove box
operations and up to 10-fold for hot cell operations.
Holding Times
Another consideration is that regulatory holding times designed for environmental samples
are very hard, if not impossible, to meet in some instances because of the delays caused by the
additional sample characterization needed to define safe handling practices, and the reduced
productivity from working under containment or shielding. The relevance of these holding times
is highly questionable for RMW materials which in some cases have been in storage or decades.
Reconsideration of this issue is needed.
Presented on Jufy 16, 1992
-------
The Toricity Characteristic Leaching Procedure (TCLP)
Operationally, the TCLP (4,5) consists of two experimental procedures, leaching the
wastes and analyzing the leachates for regulated constituents. For low level wastes that can be
leached in a radiochemical hood, no major difficulties are anticipated in conducting the leaching.
Cleaning the leaching equipment (especially the zero headspace extractor [ZHE]) will cause a
waste disposal problem because of the generally messy nature of equipment cleanup operations
and the radioactive contaminated lab wastes generated from the cleaning operations.
Because of the relatively large sample sizes (25 - lOOg) and bulky, complex apparatus
and equipment (e.g., the ZHE and tumbler) specified in the regulations, few laboratories have
adapted the TCLP leaching procedures to glove box or hot cell operations for high level RMW.
The nonvolatile leaching has been adapted to glove box and hot cell operations in only five DOE
laboratories, with the main deviation from the EPA protocol being a reduction in scale. The
Idaho National Engineering Laboratory (INEL) is the only laboratory conducting the full-scale
non-volatile leaching in a hot cell. The ZHE is much more complicated and difficult to adapt
to glove box or hot cell use, but work in this area is underway. At least two companies are
designing full-scale, lower cost ("disposable") ZHEs for hot cell use, and ORNL and INEL are
designing ZHEs from alternate apparatus such as gas syringes. Reductions in TCLP sale to
accommodate ALARA, limited sample availability, or more practical hot cell equipment for high
level waste leaching must be developed and accepted by regulatory agencies. The recent draft
of the joint EPA/NRC guidance on mixed waste testing (6) suggested a regulatory agency
realization of this need. Until such equipment and procedures are developed and approved, the
direct analysis option makes the most sense for RMW from both a scientific and a practical
standpoint.
A final problem with the TCLP that is not specific to RMW is the interference of acetate
with the analytical procedures. In particular, acetate from the leaching fluids is carried into the
semivolatile organics analysis (SVGA) extracts, where it seriously degrades the performance of
Presented on Jufy 16, 1992 as EPA Workshop TV
on 'Characterizing Mixed "Wastes' D-26
-------
the capillary GC-MS determinative method. Acetate also comes derivatizing agents in the
preparation for herbicides determination by GC-ECD. The use of SW-846 GC methods 8040
(for phenols), a combination of 8121/8091/8081 (for hexachlorobenzene, dinitrotoluene, and the
pesticides), a modification of HLPC method 8123 (for herbicides), and 8260 (volatiles plus the
remaining extractable organics) allow determination of all TCLP organics without interference
from the acetate (7). The HPLC method is being further adapted to include all of the acidic
extractable TCLP organics without any required sample preparation. Regulatory agency
acceptance of HPLC methods for these compounds would permit faster and cheaper analyses
with less laboratory waste and personnel exposure.
No difficulties are anticipated with the determination of the characteristics of corrosivity,
flammability, or reactivity in low level RMW. Flammability determinations can be a problem
when high level RMW samples require shielding or containment. INEL is developing a flash
point apparatus with a pressure-sensing membrane to allow this determination to be conducted
in a hot cell or glove box.
Organic Chemical Analysis
Problems dealing with organic chemicals in RMW are discussed in this section.
Sample Preparation
For low level RMW, current ultrasonic or soxhlet (solids) or separatory runnel or
continuous extractor (aqueous liquids) SW-846 or Contract Laboratory Program methods can be
used with little modification and careful handling and containment to minimize personnel
exposure and prevent spread of contamination. However, considerable laboratory waste that is
contaminated with radioactivity (therefore a low level waste and possibly RMW from toxic
solvents or spikes) can be generated and will have to be disposed. With higher activity RMW,
Presented on July 16, 1992 at EPA Wortshop IV
on "Characterizing Mixed Wastes' D-27
-------
these problems are compounded with the difficulties of working in a glove box or hot cell, often
with smaller sample amounts than specified in SW-846 protocols (which raises detection limits).
Several extraction technologies may be useful in low and high level RMW analysis, and
expedited acceptance of them by regulatory agencies will help to alleviate these difficulties.
Work at several laboratories (including EPA-sponsored studies) has demonstrated that a variety
of analytes can be efficiently recovered form solids by supercritical fluid extraction (SFE).
LANL recently has shown that carbon dioxide with a water modifier can efficiently extract many
semivolatile organics (including phenols) from simulated RMW. The ability to extract relatively
small masses of sample (in a glove box or hot cell, if need be) and analyze the extracts on-line
via transfer lines to a GC-MS outside the box or cell greatly reduces lab waste generation,
radiation exposure of staff, labor, and sample preparation/analysis time. For liquids samples,
solid phase extraction using packed columns (as in some EPA 500 and 600 series methods),
extraction membranes, and open plastic tubing (as at LANL) promise rapid and reproducible
methods with far less solvent usage and operator exposure than current solvent extraction
methods. Both of these technologies can be conducted relatively easily under shielding or
containment, and they could be readily automated to further improve productively. Further
development and optimization of these powerful tools, and their acceptance by regulatory
agencies clearly should be a priority.
For higher level wastes, the presence of high concentrations of nitrite has caused
problems in aqueous liquid sample preparation for the SVGA. Foaming during acidification and
release of oxides and nitrogen, which creates artifacts by nitrating organic compounds
(Westinghouse Hanford has observed nitration of at least two SVOA surrogate standard phenols),
are some of the problems caused by nitrite (2). A means of selectively removing or lowering
the nitrite, or of avoiding the need to acidify before extraction would be helpful in improving
sample handling and reducing potential artifacts. Other problems encountered with high level
wastes include high alkalinity of sludges and also unidentified interferences in Hanford tank
wastes which quickly degrade volatile organic analysis (VOA) and SVGAS GC-MS performance.
Presented on Jufy 16. 1992 al EPA Wortihap IV
on 'Characterizing Mixed Wastes' D-Zo
-------
The latter is not removed by CLP purification methods. Clearly, there are major technology
needs for clean-up of high level samples.
The automation of sample preparation and analysis will increase laboratory productivity
and reduce operator exposure to radioactive samples. In the near term, only segments of
procedures are automated (e.g., VGA purge and trap or gel permeation chromatography and
fraction concentration), but in the long term, efforts such as the DOE contaminant Analysis
Automation Program should provide the methods and equipment. This effort should be
continued.
Determination of Regulated Organic Compounds
Low level RMW do not present major problems in the organics determination steps, and
regulatory agency protocols can be followed as prescribed. The instruments preferably should
be located in contamination zoned laboratories to contain any radionuclides which may be carried
over into sample extracts. The concentrated extracts are usually decontaminated by several
orders of magnitude and thus could be analyzed in non-zoned laboratories (if the extracts are
analyzed for gross alpha and gross beta activities to confirm decontamination), but it cannot
always be assumed that complexing agents will not be present in the samples. Complexing
agents would facilitate the transfer of radionuclides into the SVGA or PCB/pesticide extracts.
The volatile organics analysis (VGA) must be conducted in a zoned laboratory because the
sample preparation (purge and trap) is best carried out on-line with the GC-MS.
For high level RMW, sample amounts are often limited by compliance with the ALARA
principle or by sampling constraints. This increases analysis reporting limits. Means of
increasing the fraction of the SVGA extract for example, injected into the GC-MS would offset
the smaller sample sizes. Application of the SVGA extract to an adsorbent, evaporation of the
solvent and thermal desorption into the GC-MS are being evaluated at ORNL. On-line SFC-GC-
MS as being developed at LANL also may overcome this problem. Use of more sensitive MS
Presented on July 16, 1992 at EPA Workshop IV
on 'Characterizing Mixed Wasus' D-29
-------
instrumentation (i.e., the ion trap mass spectrometer) also should be examined. Validation of
these methodologies and regulatory agency acceptance of their use is needed. Alternatively, it
would be more reasonable for regulatory agencies to relax organic and inorganic compound
regulatory limits for samples whose primary hazard is a high level of radioactivity.
DOE-Unique Organics
A major area requiring methods development is the determination of compounds that are
unique to DOE RMW in their complexity and concentrations, i.e., chelators, extractants,
decontamination agents, their radiolytic/chemical/thermal decomposition products, and chelated
radionuclides. The organic matter suggested to be present in some DOE RMW by total organic
carbon measurements is rarely accounted for by regulatory organics determinations (2,8), and
in some cases is accounted for by these compounds. Although these compounds are not
currently regulated, they are important for their influences upon treatment (e.g., grouting
behavior) and mobility of radionuclides (e.g., environmental transport of radionculides from
leaks of radioactive wastes). These compounds are not determined using current regulatory GC-
MS methods because of their high polarity, water solubility, and thermal instability. At present,
analysis of these compounds is performed by an incompletely evaluated method including waste
drying and derivatization followed by GC or GC-MS. High performance liquid chromatography
and capillary zone electrophoresis interfaced with mass spectrometry appear to be good
candidates for identifying and measuring DOE-unique compounds. The very small sample
volume requirements (microliter volumes for nanoliter injections), high efficiency (up to one
million plates per meter of column), and natural applicability to polar, water-soluble, ionic
species make the latter particularly attractive. Development and validation of methods will be
needed for these specialized constituents.
Presenud on July 16. 1992 at EPA Workshop TV
on 'Charoaerizing Hoed Wastes' D-30
-------
Inorganic and Radiochemical Analyses
This section presents the problems that are unique to the analysis of inorganic and
radiochemical analytes of RMW.
Sample Preparation Methods
Low level RMW sample preparation (where radiochemical hood containment is sufficient)
for inorganic and radiochemical analyses presents few additional problems over nonradioactive
samples. Analyses of low level wastes are readily feasible with current technologies, although
there are areas where methodology improvement is desirable to facilitate their turn-around.
One problem unique to RMW is the spectral interferences caused by the presence of
uranium, thorium, plutonium, and other actinides in sample digests analyzed for metals by
inductively coupled plasma - atomic emission spectrometry (ICP-AES). The concentrations of
actinides in the sample digests can be reduced using tri-n-octylphosphine oxide extraction or
extraction chromatography on the unique EiChroM TRU-Spec chelator columns (9) developed
at Argonne National Laboratory (ANL). Evaluation of these preparation methods is nearly
complete at ORNL. Regulatory acceptance of these clean-up methods would improve ICP-AES
performance with samples heavily contaminated with actinides.
A second problem unique to RMW, but limited to high level wastes, is the requirement
to conduct the metals analyses of sample digests under containment or shielding. Many DOE
laboratories have placed the ICP-AES, ICP-MS, or flame-AA burner, and the GF-AA or
electrothermal vaporization ICP-MS furnace in glove boxes, and INEL has the only x-ray
fluorescence source in a hot cell. It would greatly facilitate metals analyses of high level
samples if methods were available to reduce the radionuclide content of the digests.
Transuranics can be removed from the sample digest with essentially no effect upon RCRA and
EPA Target Analyte List metals using TOPO extraction or the EiChroM columns, but a major
Presented at July 16, 1992 at EPA Workihop TV
an "Characterizing Mixed Wastes' D-31
-------
technology development need is for methods to isolate or remove beta- and gamma-emitting
radionculides from regulated metals.
Sample dissolution/digestion preparation methods improvements would benefit both
metals and radiochemical analyses. These include high pressure sealed tube dissolution methods,
improved microwave dissolution methods, and methods for obtaining clean digests or organic
matrices. Removal of salts from the digests is also highly desirable. The high solids content
they contribute to digests interferes with metals analyses and especially gross alpha activity
determinations. Salts also interfere with alpha spectrometry unless lengthy separations are
performed. One promising method, the isolation of alpha emitters by extraction chromatography
with the EiChroM TRU-Spec columns, should be validated for RMW and accepted by regulatory
agencies.
With high level RMW samples, high concentrations of alkali metal nitrates in the sample
digests cause corrosion of the graphite furnace in the GF-AA analysis of metals, and probably
also of the electrothermal vaporization furnace of ICP-MS. A means of separating the heavy
metals from nitrates would reduce graphite furnace corrosion and maintenance. Interferences
in As and Se determinations by the hydride method have been observed at Westinghouse
Hanford in samples with high Bi concentrations, possibly indicating the need to remove Bi
(which also can consume hydride) from the digest.
The colorimetric determinations of cyanide and sulfide are subject to interferences from
co-distilled compounds such as oxides of nitrogen and elemental chlorine and iodine. Methods
to reduce the concentrations of these species would increase the reliability of colorimetric anion
analyses. Preferably, however, regulatory agencies should accept ion chromatography for
cyanide and sulfide analysis determinations in RMW because it is not subject to these
interferences. Several laboratories are using this method successfully.
Presented en Jufy 16, 1992 al EPA Workshop IV
on 'Characterizing Mixed Wos!a' D-32
-------
Inorganic and RadioncuUde Determinations
No major problems exist for the inorganic and radiochemical determinations of low level
wastes. However, laboratory operations would be considerably facilitates by more rapid, high
sample load methods to measure gross alpha and gross beta activities. This would speed up
sample turn-around because these are generally the first analyses conducted upon sample receipt
in the laboratory, and they are also usually the highest sample load. Automation of these
determinations if called for. Sample preparation and transfer to counters (which already have
automated sample changers) is a good candidate for robotics, and the prognosis for a near-term
payoff is very good.
Both radiochemical and inorganic analyses of wastes would benefit considerably from the
evaluation and regulatory agency acceptance of TCP-mass spectrometry (ICP-MS), particularly
for multielement scanning and for the determination of long-lived, low specific activity
radionuclides . Traditional sample preparation and long counting times lengthen turn-around
times for determinations of radionuclides such as ^Tc, Np, and I. Among the attractive
features of ICP-MS are multielement capability, high sensitivity, small sample requirements (at
least with electrothermal vaporization), direct analysis of solids and slurries with laser ablation
or electrothermal vaporization, and its ease of automation. Also, RCRA metals probably could
be determined with nigh accuracy using isotope dilution techniques. ICP-MS analytical
technology should be a high priority for validation and acceptance by regulatory agencies.
For higher level RMW, determination methods for total fissile material by ICP-MS or
by delayed neutron counting after neutron activation need to be evaluated and accepted.
Field Analytical Methods
Certainly, RMW characterization would benefit from the development, validation, and
acceptance of methods for rapid sample screening and analysis in the field. These methods
Presented on July 16, 1992 at EPA Workshop IV
on 'Characterizing Mixed Wastes' D-33
-------
would eliminate sending samples with little or no organics, metals, or radionuclide content back
to the laboratory for expensive analyses. Work is underway at ORNL and LANL with ion trap
mass spectrometry, immunoassay, and other fieldable techniques for organics determinations.
Field methods for rapidly screening samples in the field for selected metals and gross alpha and
beta activity would be a useful adjunct to the organics screening methods. The latter also could
reduce the time for laboratory radiochemical analyses needed for health physics protection.
Examples of important work in this area include the method adaptation work which will be
underway shortly at LANL for field gross alpha and gross beta measurements using the Protean
system, and gross gamma and ^H measurements by other methods. A field test or a portable
x-ray fluorescence screen for selected metals (10) was supported by the Army. The Nevada Test
Site is also supporting work adapting the Frisch grid detector for alpha spectroscopy in the field.
CONCLUSIONS
The main conclusions from the scientific perspective are that, although there are
difficulties working with low level and very low level RMW, there are no major problems or
"show stoppers" preventing their characterization in appropriately equipped and licensed
commercial laboratories. The main problem probably will be the disposal of secondary
laboratory wastes generated in the analysis of low level wastes. However, validation and
regulatory agency acceptance of new technologies or adaptations of current technologies would
greatly facilitate the characterization of the large sample load expected for DOE. Development
and validation of field analytical methods also will be important in minimizing the costs of
sample shipment and laboratory analyses and may help in meeting currently very restrictive
holding time requirements. Higher level RMW requiring more stringent containment or
shielding pose much greater characterization problems in their own right.
Two roles will be important in successfully meeting this challenge. DOE and other
agencies must identify and adapt new laboratory and field analytical technologies, evaluate their
performance, and transfer these technologies among the DOE sites and to the commercial sector.
Presented on Jitfy 16, 1992 at EPA Workshop IV
on 'Choranenzing Mixed Wastes' D~34
-------
The EPA must facilitate the means of gaining regulatory agency acceptance of new and adapted
methods. It also must officially recognize the limited applicability of the TCLP to RMW, and
the high desirability of minimizing the characterization required of highly radioactive wastes.
Presented on July 16. 1992 at EPA Wortshop IV
en 'CharaettrizBig Mixed Wastes' D-35
-------
REFERENCES
1. W.G. Richmond, "Waste Characterization Program Development for Hanford Grout
Disposal," RHO-RE-SA-177 P, Rockwell International, Richland, WA (April, 1987).
2. J.W. Autrey, D.A. Costanzo, W.H. Griest, L.L. Kaiser, J.M. Keller, C.E. Nix, and
B. A. Tomkins, "Sampling and Analysis of the Inactive Waste Storage Tank Contents
at ORNL," ORNL/RAP-53, Oak Ridge National Laboratory, Oak Ridge, TN (August,
1989).
3. Test Methods for Evaluating Solid Waste. Volume IB, Laboratory Manual Physical/
Chemical Methods; SW-846, Third Edition, United States Environmental Protection
Agency, Office of Solid Waste and Emergency Response, Washington, D.C. (November,
1986).
4. Federal Register. 55(611. 11798-11863 (Thursday, March 29, 1990).
5. Ibid.. 55(126), 26986-26994 (Friday, June 29, 1990).
6. Federal Register. 57(591 10508 (Thursday, March 26, 1992).
7. M.P. Maskarinec, "Analysis of TCLP Extracts: Problems and Potential Solutions," 32nd
ORNL/DOE conference on Analytical Chemistry in Energy Technology, Gatlinburg, TN,
October 1-3, 1991.
8. A.P. Toste, TJ. Lechner-Fish, DJ. Hendren, R.D. Steele, and W.G. Richmond,
"Analysis of Organics in Highly Radioactive Nuclear Wastes," J. Radioanalytical and
Nuclear Chemistry. 123(1). 149-166 (1988).
9. E.P. Horowitz, et. al., "Concentration and Separation of Actinides from Urine Using a
Supported Bifunctional Organophosphorus Extractant," Anal. Chim. Acta. 238. 263-271
(1990).
10. R.A. Jenkins, F.F. Dyer, R.L. Moody, C.K. Bayne, and C.V. Thompson,
"Experimental Evaluation of Selected Field Portable Instrumentation for the Quantitative
Determination of Contaminant Levels in Soil and Water at Rocky Mountain Arsenal,"
ORNL/TM-11385, Oak Ridge National Laboratory, Oak Ridge, TN (October, 1989).
PresenUd on July 16, 1992 at EPA Workshop IV
on 'Characterizing Mixed Wastes' D~36
-------
APPENDIX E:
LEACHABILTTY PHENOMENA: Recommendations and Rationale for Analysis
of Contaminant Release by the Environmental Engineering Committee
-------
!> « \
^^ c<^
UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON, D.C. 20460
EPA-SAB-EEC-92-003
OfFICEOF
^. .„«-.-- THE ADMINISTRATOR
October 29, 1991
Honorable William X. Reilly
Administrator
U.S. Environmental Protection Agency
401 M Street, S.W.
Washington, D.C. 20460
Subject: Leachability: Recommendations and Rationale
for Analysis of Contaminant Release
Dear Mr. Reilly:
The Leachability Subcommittee (LS) of the Science Advisory
Board's Environmental Engineering Committee (EEC) has prepared
the attached recommendations and rationale on leachability, an
important release term related to solid wastes and contaminated
soils, for your consideration.
Over the past decade, the EEC has reviewed a number of EPA
issues involving leachability phenomena and noted several
problems relating to this release term that were common to a
variety of EPA offices. The Committee believed that these common
problems would be best called to the Agency's attention through a
general review of leachability phenomena.
Drafts of this report on leachability have been reviewed at
a series of Subcommittee/ Committee, and Executive Committee
meetings over the past 18 months. This included both a session
on February 26, 1990, devoted to assessing the Agency's varied
needs on leachability-related information, and a Technical
Workshop on May 9, 1990. The workshop assisted in determining
how leachability phenomena should be used to determine how a
waste will leach when present under various scenarios in the
environment.
The following recommendations have been developed. First,
in regard to leachability test development we recommend:
a) incorporation of research on processes affecting
leachability into EPA's core research program to better define
and understand principal controlling mechanisms,
b) development of a variety of contaminant release tests,
rather than focusing on mimicking a single scenario,
-------
c) development of improved release and transport-
transformation models of the waste matrix to complement the
leaching tests, and
d) field validation of the tests and models/ and
establishment of release-test accuracy and precision before tests
are broadly applied.
Next, in regard to the application of such tests and models,
we recommend:
e) use of a variety of contaminant release tests and test
conditions which incorporate adequate understanding of the
important parameters that affect leaching in order to assess the
potential release of contaminants from sources of concern. A
medical analogy is that no physician would diagnose on the basis
of one test showing only one aspect of the problem,
f) development of a consistent, easily applied, physical,
hydrologic, and geochemical representation for the phenomenon or
waste management scenario of concern,
g) identification and application of appropriate
environmental conditions for tests in order to evaluate long-term
contaminant release potential as required under varying statutes,
and
h) coordination between the Agency's programs which develop
leachability tests with those that develop the environmental
models in which the release terms are used.
Finally, we recommend:
i) establishment by the Agency of an inter-office, inter-
disciplinary task group, including ORD to help implement these
recommendations, and
j) development of an Agency-wide protocol for evaluating
release scenarios, tests, procedures, and their applications.
These recommendations are made with the anticipation that an
improved understanding of the fundamental scientific principles
that control contaminant release and transport within a waste
matrix will allow better regulatory and technical decisions to be
-------
made in cases where the potential exists for leaching of
contaminants into the environment.
We are pleased to be of service to the Agency, and hope that
you will find this effort useful. We look forward to your
response to the recommendations cited above.
Dr. Raymond c. Loehr, chairman Mr. Richard A. Conway, chairman
Executive Committee Environ. Engineering Committee
Dr. C. H. Ward, Chairman
Leachability Subcommittee
-------
United States Science Advisory Board EPA-SAB-EEC-92-003
Environmental Protection (A-101F) October 1991
Agency
&EPA Leachability
Phenomena
Recommendations and
Rationale for Analysis of
Contaminant Release by the
Environmental Engineering
Committee
Printed on Recycled Paper
-------
NOTICE
This report has been written as a part of the activities of
the Science Advisory Board/ a public advisory group providing
extramural scientific information and advice to the Administrator
and other officials of the Environmental Protection Agency. The
Board is structured to provide a balanced/ expert assessment of
scientific natters related to problems facing the Agency. This
report has not been reviewed for approval by the Agency; hence/
the comments of this report do not necessarily represent the
views and policies of the Environmental Protection Agency or of
other Federal agencies. Any mention of trade names or commercial
products does not constitute endorsement or recommendation for
use.
-------
ABSTRACT
The reachability Subcommittee (LS) of the Environmental
Engineering Committee (EEC) of the EPA Science Advisory Board
(SAB) conducted a self-initiated study and prepared a report on
the topic of leachability phenomena. The intent of this report
is to provide recommendations and rationale for analysis of
contaminant release to the staff in the various offices of the
Environmental Protection Agency (EPA). The nine recommendations
from the report are highlighted as follows:
1) A variety of contaminant release tests and test condi-
tions which incorporate adequate understanding of the important
parameters that affect leaching should be developed and used to
assess the potential release of contaminants from sources of
concern.
2) Prior to developing or applying any leaching tests or
models, the controlling mechanisms must be defined and
understood.
3) A consistent, replicable and easily applied, physical,
hydrologic, and geochemical representation should be developed
for the waste management scenario of concern.
4) Leach test conditions (stresses) appropriate to the
situations being evaluated should be used for assessing long-term
contaminant release potential.
5) Laboratory leach tests should be field-validated, and
release test accuracy and precision established before tests are
broadly applied.
6) More and improved leach models should be developed and
used to complement laboratory tests.
7) To facilitate the evaluation of risk implications of
environmental releases, the Agency should coordinate the
development of leach tests and the development of models in which
the release terms are used.
8) The Agency should establish an inter-office, inter-
disciplinary task group, including ORD to help implement these
recommendations and devise an Agency-wide protocol for evaluating
release scenarios, tests, procedures, and their applications.
9) Core research on contaminant release and transport within
the waste matrix is needed.
Key Wordsi leachability, leachability phenomena, leach tests and
methods, leaching chemistry, leaching models
ii
-------
LEACHABILITY SUBCOMMITTEE
ENVIRONMENTAL ENGINEERING COMMITTEE
of the
SCIENCE ADVISORY BOARD
Chairman;
Dr. C. Herb Ward, Professor and Chairman, Department of
Environmental Science and Engineering, Rice University, Houston,
Texas
Vice-Chairman;
Dr. Ishvar P. Murarka, Senior Program Manager, Land and Water
Quality Studies, Environment Division, Electric Pover Research
Institute, Palo Alto, California
Members and Consultants;
Dr. Larry I. Bone, Environmental Quality Department, The Dow
Chemical Company, Midland, Michigan
Dr. Yoram Cohen, Professor, Department of Chemical Engineering,
University of California at Los Angeles, Los Angeles, California
Mr. Richard A. Conway, Senior Corporate Fellow, Union Carbide
Corporation, South Charleston, West Virginia
Dr. Linda E. Greer, Senior Scientist, Natural Resources Defense
Council, Inc., Washington, D.C.
Dr. J. William Haun, Director of Research 6 Development,
General Mills, Inc. (retired), Maple Grove, Minnesota
Dr. Wayne M. Kachel, Consultant Manager, Pilko and Associates,
Inc., Houston, Texas
Dr. Raymond C. Loehr, Professor and H.M. Alharthy Centennial
Chair in Civil Engineering, University of Texas, Austin, Texas
Dr. Frederick G. Pohland, Professor and Edward R. Weidlein
Chair of Environmental Engineering, University of Pittsburgh,
Pittsburgh, Pennsylvania
Dr. Paul V. Roberts, Professor, Department of Civil
Engineering, Stanford University, Stanford, California
Dr. Walter M. Shaub, Technical Director, Coalition on Resource
Recovery and the Environment, U.S. Conference of Mayors,
Washington/ D.C.
Dr. Mitchell J. Small, Professor, Department of civil
Engineering, Carnegie-Mellon University, Pittsburgh, Pennsylvania
iii
-------
Science Advisory Board Staff;
Mail Address: U.S. Environmental Protection Agency
Science Advisory Board (A101F)
401 M Street, S.W.
Washington, D.C. 20460
Designated Federal Official
Or. K. Jack Kooyoomjian
Staff Secretary
Mrs. Marcy Jolly
Assistant Staff Director
Mr. A. Robert FlaaX
Director
Dr. Donald G. Barnes
Chairman, Executive Committee,
Science Advisory Board
Dr. Raymond C. Loehr
Chairman, Environmental Engineering; Committee
Mr. Richard A. Conway
iv
-------
TABLE OF CONTENTS
I. EXECUTIVE SUMMARY 1
II. INTRODUCTION 2
III. WHAT ARE THE NEEDS OF THE AGENCY AND REGULATED
COMMUNITY? 4
IV. ASSESSMENT OF CURRENT PRACTICE 5
V. RECOMMENDATIONS FOR IMPROVED LEACHABILITY
DETERMINATIONS TO FILL GAPS BETWEEN NEEDS AND
CURRENT PRACTICES 7
APPENDIX A - BACKGROUND ON LEACHABILITY AS A
SELF-INITIATED ACTIVITY OF THE
SCIENCE ADVISORY BOARD 33
APPENDIX B - LEACHABILITY WORKSHOP PROGRAM 34
APPENDIX C - LEACHABILITY NEEDS, USES, TESTS,
CONCERNS, AND ISSUES 35
C-l - Superfund Remedial and Removal Programs
C-2 - Office of Water, Sediment Criteria Program
C-3 - Office of Solid Waste
C-4 - Office of Toxic Substances
C-5 - R&D Perspective of the R.8. Kerr Lab,
Ada, Oklahoma
C-6 - R&D Perspective of the Office of Modeling,
Monitoring and Quality Assurance,
Washington, D.C.
C-7 - Criteria and Objectives of Leaching Tests
and Modeling Considerations for Partitioning
Tests from the Perspective of Regulators
Versus Industry
APPENDIX D - CREDITS AND ACKNOWLEDGEMENTS 46
APPENDIX E - GLOSSARY OF TERMS AND ACRONYMS 49
-------
LIST OF TABLES
TABLE 1 - EXTRACTION TESTS 20
TABLE 2 - TEST REQUIREMENTS, USES OF TESTS, AND PROGRAMATIC
NEEDS FOR LEACHABILITY TESTS BY THE AGENCY AS
PERCEIVED BY TEE SCIENCE ADVISORY BOARD'S
ENVIRONMENTAL ENGINEERING COMMITTEE 25
TABLE 3 - SCIENTIFIC CONSIDERATIONS IN DESIGN AND
INTERPRETATION OF LEACHABILITY TESTS 27
LIST OF FIGURES
FIGURE 1 - CONCEPTUAL VIEW OF LEACHING IN A WASTE UNIT 6
Vi
-------
I. EXECUTIVE SUMMARY
In waste management, including managing the effects of
spills or other releases which are sources of underground
contamination, a critical issue is the assessment of the
potential for constituents to leach to the environment. The
Environmental Engineering Committee (EEC) of the Science Advisory
Board (SAB) undertook a study of this issue because it noted
several common problems relating to this release term as it
reviewed, over the past decade, various leaching tests and risk
models for several EPA offices. Tests such as the Extraction
Procedure (EP) and the Toxicity characteristic Leaching Procedure
(TCLP) had, and continue to have, scientific limitations, yet
were being inappropriately and in some cases widely used. Often
tests were developed without rigorous review. A self-initiated
study seemed appropriate to define the leachability problem
better and to offer advice on its resolution.
The EEC established a Leachability Subcommittee (LS) that
addressed:
1) Needs of the Agency and regulated communities to
quantify leachability (releases) of contaminants to the
environment.
2) State-of-the-art and science related to fundamental
principles and practice in predicting leaching of constituents
from wastes, contaminated soils, and other sources.
3) Recommendations to improve the scientific understanding
and application of leaching tests.
Workshops were held, literature was analyzed, and
findings were discussed over an 18-month period leading to the
preparation of this report.
The various needs for tests and models to predict leaching
are defined. Tests developed and used in the U.S. and Canada are
summarized. The scientific considerations important in design
and interpretation of leachability tests are presented. This
information, expert advice and analysis by workshop participants,
and reviews by SAB members, resulted in guidance which should, if
progressively implemented, significantly strengthen the Agency's
ability to assess appropriately leaching of contaminants from
hazardous wastes, contaminated soils and other sources.
This guidance, in the form of nine recommendations, is
summarized as follows:
1) A variety of contaminant release tests and test
conditions which incorporate adequate understanding of the
important parameters that affect leaching should be developed
-------
and used to assess the potential release of contaminants from
sources of concern.
2) Prior to developing or applying any leaching tests or
models, the controlling mechanisms must be defined and
understood.
3) A consistent, replicable and easily applied, physical,
hydrologic, and geochemical representation should be developed
for the waste management scenario of concern.
4) Leach test conditions (stresses) appropriate to the
situations being evaluated should be used for assessing long-term
contaminant release potential.
5) Laboratory leach tests should be field-validated, and
release test accuracy and precision established before tests are
broadly applied.
6) More and improved leach models should be developed and
used to complement laboratory tests.
7) To facilitate the evaluation of risk implications of
environmental releases, the Agency should coordinate the
development of leach tests and the development of models in which
the release terms are used.
8) The Agency should establish an inter-office, inter-
disciplinary task group, including ORD, to help implement these
recommendations and devise an Agency-wide protocol for evaluating
release scenarios, tests, procedures, and their applications.
The task group should also be charged with recommending what the
appropriate focal point(s), responsibilities, and organizational,
budgetary and communication links should be within the Agency for
the most effective, continued and ongoing support and pursuit of
the research, development and utilization of methods and
procedures.
9) Core research on contaminant release and transport within
the waste matrix is needed.
II. INTRODUCTION
In both hazardous and non-hazardous waste management, one of
the most critical issues is the assessment of the potential for
constituents contained in the source material to leach or
otherwise be released to the environment. Approaches to estimate
potential release of organic and inorganic constituents and their
subsequent environmental migration and associated health risks
are important in many situations (e.g., pollution prevention,
risk reduction, restoration-remediation and hazard identi-
fication) .
-------
This review has been initiated by the Environmental
Engineering Committee of the Science Advisory Board because 1)
the Committee has been reviewing Agency actions which require
definition of the potential for releases from wastes and their
transport to human and environmental receptors where exposure can
occur/ and 2) the Committee has previously reviewed the
scientific and technical basis for two tests for leaching
potential intended for particular uses: the Extraction Procedure
(EP) and the Toxicity Characteristic Leaching Procedure (TCLP)
(See for instance, US EPA, Science Advisory Board, Report of the
Environmental Engineering Committee, Report on the Review of EP-
III. A Procedure for Determining the Leaching Potential of
Organic Constituents from Solid and Hazardous Wastes, July 19,
1984). In addressing and reviewing Agency proposals, the
Committee has repeatedly observed and commented on the scientific
limitations of the EP and TCLP tests. Many of the proposed uses
for the tests have been inappropriate because the waste
management scenarios of concern were not within the range of
conditions used in the development of the tests themselves.
In most cases of inappropriate use of the EP or TCLP tests,
the justification given was that it is necessary to cite
"standard" or "approved" methods. Even if it is acknowledged
that the tests cannot be applied without significant change in
the test protocol itself, the need to use a previously "approved"
test has been cited.
In a contradictory set of pressures, some offices have
devised new tests to suit particular needs, e.g., the "oily waste
extraction test," when it was considered necessary. Only rarely
have such new or modified leaching tests been subjected to a
rigorous review of their precision, accuracy or technical bases
comparable to that applied to the EP and TCLP tests.
There are many laboratory tests that have been devised to
obtain estimates of the potential for contaminant release. These
tests are generally characterized as either static or dynamic.
In all instances, aqueous solutions have been utilized as the
leaching fluid. Solid-to-liquid ratios of 2:1 to 20:1 have been
prescribed. Leaching times of 18 hours to several days, and in
some tests, years are required. Various tests specify single to
over 20 extractions, and particle sizes froa 2 ma to monolith
proportions. Table 1 (page 20), provides a summary of over 30
tests designed to help determine the potential for contaminant
release. Although a wide range of leaching tests exist, a
conceptual framework for their application is generally lacking.
In preparing this report, the Committee has sought the best
technical input available on "state-of-the-art" knowledge of the
leaching phenomenon. It has also sought and received extensive
information on the needs of regulatory and enforcement programs
for reliable leaching predictions and their interpretations.
-------
Building on its previous experience in technical reviews, on
outside input, and on its own expertise, the Committee offers
advice on vays to resolve the conflict between the need for
"standard" tests and the need for tests better adapted to the
circumstances to which the data are to be applied.
This self-initiated study focused on the following three
questions:
1) What are the needs of the Agency and regulated community
to quantify leachability or release of contaminants to the
environment?
2) What is the state-of-the-art and science dealing with
the fundamental principles that should be considered in
predicting leaching of chemicals from wastes, contaminated soils,
and other sources?
3) What should be or could be done to improve the
scientific understanding and application of leaching tests in
future risk analysis?
III. WHAT ARE THE NEEDS OF THE AGENCY AND REGULATED COMMUNITY?
The Subcommittee convened a one-day session on February 26,
1990, in Washington, D.C., devoted to assessing the Agency's and
other's (private sector and citizen groups) varied needs for
leaching tests and information. The findings are summarized in
Table 2 (page 25) and are detailed in Appendix C (pages 35-45).
Table 2, summarizes the Subcommittee's understanding of the
wants, uses, and needs for leachability tests/data within the
Agency. At least six program offices have expressed interest in
such information. The wants, or what the program offices would
like, are varied. Most focused on a Beans to predict field
conditions. All offices expressed a desire for a method(s) to
appropriately classify a waste. Given that such a test(s) did
exist, the offices would use it such as to set standards,
primarily through simulating risk. The Office of Toxic
Substances appears to have the broadest uses. Just as the wants
and uses are varied, so too are the needs. Consistently, all
offices see leachability tests as a means of demonstrating
compliance; a use for which most leaching tests were not
originally intended.
The EPA, through mandates of the RCRA, CERCLA, CWA and
associated regulatory programs, has required chemical testing and
other laboratory procedures to predict the possible hazards of
chemicals potentially released into the environment. The intent
of leaching/extraction tests is to reliably estimate the poten-
tial amount and/or rate of contaminant release under worst case
environmental conditions, thus enabling remedial, prevent-ative
-------
and anticipatory management actions to be taken to protect human
health and the environment.
IV. ASSESSMENT OP CURRENT PRACTICE
The Committee assessed the state-of-the-art in leachability
determinations through several Beans:
a) Participation in a Workshop on Contaminant Migration
at Rice University on December 15-16, 1989, organized
by the National Center for Ground Water Research
for the EPA Robert 8. Kerr Environmental Research
Laboratory in Ada, OK., in cooperation vith the
University of Texas at Austin and the Electric Power
Research Institute.
b) Holding a Leachability Workshop under the auspices
of the US Environmental Protection Agency, Science
Advisory Board on May 9, 1990 in Washington, D.C.
The Workshop Program and list of speakers are given
in Appendix B, page 34.
c) Review of key references assessing current leach-
ability tests, e.g., Compendium of Waste Leaching
Tests, Wastewater Technology Centre, Environment Canada,
May 27, 1989 (See Table 1, page 20 for summary
of extraction tests).
The findings reported also reflect the personal experiences and
expertise of the members of the Leachability Subcommittee.
The Leachability Workshop was conceived as a vehicle for
knowledgeable scientists, engineers and practitioners in the
field to focus on the scientific principles and issues relating
to leachability. The purpose of the Workshop was to conduct a
review of the scientific principles involved with leachability
phenomena. Various experts discussed relevant topics such as
test methods, their descriptions and capabilities for application
to the leaching of organics and inorganics, the leaching of
stabilized materials, physical-chemical mechanisms, leaching
chemistry of organics and inorganics, and alternative approaches
to laboratory tests.
The Workshop assisted the Leachability Subcommittee in
summarizing the fundamental scientific principles that control
leachability (Appendix B, page 34, and Figure 1 page 6_), and
determining how they can be applied to predict the extent to
which contaminants will leach when disposed under various
potential environmental scenarios. The Contaminant Migration
Workshop, the Leachability Workshop, and the review and
assessment of key references, provided the background for
formulation of recommendations on how basic principles can be
-------
Fluid composition '
(e.g.,pH, dissolved
gases) and quantity
Leachates
Biogeochemical (precipitation/
dissolution, adsorption-
desorption, redox, biotic
transformation and
partitioning) reactions
between solids and liquids
Washing
Dissolution of solubility-
controlling solid
/— Matrix dissolution
/—Cosolvent
/ present
Time or Pore Volumes
(b)
-—Depletion—-
/— Microbial
/ degradation
Aqueous
partitioning
Time or Pore Volumes
(c)
FIRGURE 1. Conceptual View of Leaching in a Waste Unit.
(a) Generation of leachates;
(b) Potential leaching stages for inorganic contaminants;
(c) Generation of organic leachates.
-------
used on a consistent basis for improving or developing decisions
related to leachability.
A conceptual view and sunnary of the major processes and
interactions that can occur in leaching of a waste matrix is
presented schematically in Figure 1, page 6> The generation of
leachates is depicted (Figure la) and is the sun of several
geochemical processes involving reactions within a physical
mixture of different forms of the sane element. For example, in
fly ash, an element such as cadmium may occur as simple oxide
salts accumulated on the surface of ash particles, as an element
within the glass matrix, and as a post-combustion product of
aqueous reaction such as CdCCU. Potential leaching stages
areillustrated for an inorganic contaminant in Figure Ib.
Reaction of the readily available and highly soluble fraction,
such as the surface oxide salts, provides high solute
concentration during initial washing or leachate generation.
This stage is followed by considerably decreased concentrations
of solutes as the readily available fractions have been leached
or transformed into less soluble forms (such as the CdCO3).
These transformed solids are referred to as solubility-
controlling solids. Dissolution and depletion of these
solubility controlling solids and the bulk matrix are important
determinants of the temporal change and characteristics in
leachate generation following initial washing.
The generation of organic leachates (Figure Ic) also
involves several biogeochemical processes. For example, the
leachate concentration of an organic compound may be controlled
by its water solubility (aqueous partitioning). However, in the
presence of a cosolvent, the leachate concentration is usually
controlled by the solubility of the constituents in the organic
phase rather than in water and nay be substantially increased.
Microbial degradation, abiotic transformations and physical
partitioning to gaseous and solid phases can further alter the
pattern of leachate generation.
V. RECOMMENDATIONS FOR IMPROVED LEACHABILIT.Y DETERMINATIONS TO
FILL GAPS BETWEEN NEEDS AND CURRENT PRACTICES
Based on a broad-ranging review and analysis of needs and
information available on leaching phenomena, the Leachability
Subcommittee has developed the following recommendations:
1) A variety of contaminant release tests and test
conditions which incorporate adequate understanding of the
important parameters that affect leaching should b« developed and
used to assess the potential release of contaminants from sources
of concern.
In scientific and technical terms, no "universal" test
procedure is likely to be developed that will always produce
-------
credible and relevant data for input to all decision Baking
exercises. There is nothing inherently bad about having a vide
variety of test conditions and Methods, to cover the range of
needs, if each is defensible in view of the scientific and
technical understanding of the basic processes involved.
Provisions for adequate margins of safety and varied scenarios
should be made. A wealth of knowledge already exists vhich can
form the essential basis for quantitative determinations of the
release of contaminants. Appropriate physical, chemical and
biological factors should be selected for a specific test or
model to reliably estimate contaminant releases. Important
information includes: waste characteristics/ Mobilizing fluid
characteristics, intrinsic behavior of chemicals, and most likely
reactions to occur under varying hydrologic, chemical and
atmospheric conditions (Table 3, page 27.).
Chemical and physical characteristics of a waste are
significant determinants of leachate composition. The nature and
conditions of occurrence of leachable constituents, rather than
the total amounts, often dictate the consequential release of
contaminants from the waste. Moreover, the speciation of
constituents of concern, for example, heavy metals, cannot or has
not in most cases been reliably quantified. This can confound
reliable anticipation of potential contaminant release, further
contributing to uncertainty in contaminant release assessment.
Waste or matrix heterogeneity is another complicating factor.
In addition to waste and matrix characterization, it is
similarly important to characterize the leaching medium (i.e.,the
fluid) which contacts the waste material. Terrestrial
porewaters, including groundwaters, have broadly defined fluid
parameters and show tremendous diversity in those characteristics
which influence solubility and chemical behavior of mobilized
waste chemical constituents. Also, fluid which contacts the
waste can be influenced by the presence of contaminants from
other wastes, such as at a Superfund site where waste oil has
been codisposed with PCBs.
2) Prior to developing or applying any leaching tests or
models, the controlling nechanisas Bust be defined and
understood.
Contaminant release (and eventual fate) in a field
environment is an extremely complex phenomenon involving multiple
phases and multiple constituents. In order to provide a proper
conceptual framework for a leachability scenario, a recommended
first step in any leaching test or model should be to identify
all significant mechanisms that can ultimately determine release
and environmental fate of the contaminants.
After identifying mechanisms, an understanding of how they
(directly or indirectly) influence release and environmental fate
8
-------
should be established. Past experience suggests that identifying
the principal controlling mechanisms is often straightforward.
However, anticipating the interrelationships among concurrent and
often competing mechanisms can be much more difficult, yet
critical in the characterisation of leachability.
In developing the conceptual framework for a leachability.
scenario, attention should be given to accounting for all
significant phenomena — be they physical, chemical, or
biological —and their potential interactions. At a minimum, the
following phenomena should be considered: fluid characteristics
and flow dynamics, source matrix morphology and chemistry,
chemical reactions (equilibrium or kinetic), biotic reactions,
and temporal and spatial dependence. These are discussed in
relation to the principles identified throughout this report (see
also Figure l and summary Tables 1, 2 and 3 for overview
information).
To understand and predict the leaching of organic
constituents of concern, it is important to consider the presence
of organic solubilizers such as solvents and oil in addition to
wastes. Under the conditions of codisposal with solubilizing
agents, the extent of leaching is usually controlled by the
solubility of the constituents in the organic phase rather than
in water. However, dissolved solvents can, to a lesser extent,
affect constituent solubility in the aqueous phase.
Experimental data obtained over the last decade indicate
that the solubilizing effect of some agents can dramatically
increase the concentrations of normally insoluble organic
constituents. In some instances, organics have been shown to
enhance the mobility of inorganic constituents, most likely
through complexation. In these cases, aqueous extractions to
estimate the extent of leachability of the waste may seriously
underestimate the magnitude of release for constituents of
concern.
Biotic reactions are known to be important in some
circumstances and should be considered to fully simulate
leachability. Both inorganic and organic constituents frequently
undergo biological transformations within a source matrix. These
transformations can directly change the chemical environment and
composition of the leachate, and contribute to other secondary
effects such as changing the nature of the leaching fluids and
the setting within which leaching occurs.
The effect of biotransformation has been considered in some
instances. The EP, the TCLP and other tests use an organic acid
in an attempt to simulate codisposal with degradable materials.
The oily waste extraction procedure requires extraction of oil
from the solids prior to leaching, because biodegradation in
-------
nature will remove the oily fila allowing Bora intiaate contact
with the leaching fluid.
It is often difficult to reproduce or simulate in batch
extraction procedures, biotransformations which occur in field
situations/ in part because the temporal and spatial
considerations for each are so different. Batch extractions are
designed to evaluate equilibrium processes, while biotrans-
fonnations are rate-limited. However, biotransformation studies
are not inherently incompatible with column or dynamic leaching
tests, and as such should be incorporated in those cases where
they could be important. In attempting to integrate the
assessment of biotransformation phenomena into a leaching test,
it is critical to consider rate limitations in choosing the
test's duration. The effects of biotransformation should be
considered in interpreting or applying leachability results in
cases where it can occur.
3) A consistent, replicable and easily applied , physical,
hydrologic and geochemical representation should be developed for
the waste Management scenario of concern.
Central to devising leaching tests and models is the
development of a sound conceptual framework of the physical
system which is to be simulated. This framework requires proper
and relevant identification of the precipitation-dissolution
reactions by systematically examining the fluid and waste
characteristics as well as the intrinsic chemical behavior of the
constituents. This includes the equilibrium or steady state
conditions as well as the reaction rate representing the dynamics
of the leaching process. Reaction rate is an important, yet
poorly understood, factor.
The complexities of fluid-solid waste interaction dynamics
can occur in "real world" field conditions in which fluid
parameters can change spatially and temporally with corresponding
shifts in constituent release behavior. A constituent may
initially occur in a highly soluble form, but Bay subsequently be
transformed to highly insoluble precipitates, thereby rendering
it immobile. Similarly, a constituent may be leached at a
greater rate due to the very acidic or very alkaline pH of the
leaching fluid. Dissolution of the protective matrix material may
increase the potential for transport of the leaching fluid and
the release of contaminants.
The Subcommittee believes that rate-limiting chemical and
microbial reactions often play a pivotal role in governing
leaching rate and contaminant fate. Consequently, situations can
occur in which equilibrium concepts say not apply to some {and
probably even to the majority) of the contaminant release (and
transport) scenarios. It is expected that, in numerous
instances, equilibrium-based projections of leachate levels can
10
-------
only provide an asymptotic (e.g./ upper or lower limit) estimate
of leachability tbat is not experienced at actual disposal sites
during the management period of interest. In addition,
hydrologic conditions and their spatial and temporal variations
can impact on rates of dissolution and precipitation/ mass
transfer/ and diseguilibria. These effects and their impacts are
not well understood, despite their important contributions to
leaching of constituents to the environment. Despite these
limitations, determinations of hazard for various contaminant
release (and transport) scenarios are a. necessity in order to
accommodate the realities of statutory requirements and the
attendant regulatory requirements.
If the intent is to match a test to an environmental
situation with respect to contact time, a distinction could be
made between situations where the waste is contained within a
lined landfill (long contact time), or where the waste is
underlain by a very porous medium or is in a flowing surface
water (short contact time). Likewise, the leaching liquid-to-
solid ratio at a waste management site is functionally
(dynamically) related to rainfall, infiltration rates, the
presence or absence of a cap over the waste, the quantity of
waste, and other site-specific factors. Exceptions do occur
based on site-specific factors. For example, if porous media
have been plugged, this would increase contact time. Because of
plugging phenomena, significant retention of waste leachate has
been observed in some municipal landfills (such as in the Long
Island, New York area), despite the fact that they are unlined
and underlain by a very porous medium (sandy soil).
Explicit selection of leachate tests to best match repeated
or continual leaching may require a determination of whether the
particular management scenario involves wastes in a lined and/or
capped landfill, or in more open systems in which greater contact
exists with the surrounding environment. In the former case, the
leachate may only be drained from the waste once. In the latter
case, multiple leaching and contact times can reasonably be
expected to occur. In most cases, multiple leaching can probably
be expected/ although contact times, liquid-to-solid ratios, pH
and other environmental factors may vary, not only for different
scenarios, but also for successive leaching events in a given
environmental setting. This variability can be difficult to
predict and simulate.
The nature and influence of physical dimensions (e.g.,
particle size and shape) of the waste matrix is difficult to
predict for test development purposes. As a general principle,
accuracy and reliability should be improved if the waste is
leached in the form that is present in the environment, if it can
be expected and demonstrated that the form of the waste will
remain relatively unchanged with time. In other words, waste
matrix dimensions should simulate what is expected in the
11
-------
environment over the time period of concern. It also follows
that the dimensions range in test samples should reflect that of
the vastes of concern.
A noteworthy issue is that some wastes in the environment
clearly have dimensions which are too large to be examined with a
particular laboratory test apparatus. Depending on
accommodations to particular requirements of the test apparatus,
this may or may not be a serious problem. If only a modest size
reduction is required, such as reduction of a football-sized
object to an average of one inch diameter/ this is likely not to
be a serious problem with respect to the development of
leachability information, because the surface to volume
perturbation is much less severe than had the requirement been to
reduce the football-sized mass to milled wastes of a millimeter
or less size range. The surface-to-volume ratio between large
particles and those of an inch or so in size is not greatly
different. However, diffusion limitations are intimately related
to grain size and matrix uniformity, and the time scale is
proportional to the square of the length scale. Size reduction
may also cause changes in surface chemistry, redox conditions and
availability of reaction sites.
One practical way to limit the size problem in batch tests
is to mill or crumble wastes which do not have strengths
appreciable enough to survive in the environment. Accurate
predictions of leaching potential from wastes with intermediate
strength may be better handled by subjecting them to sequential
leaching accompanied by sequential particle size reduction. At
the other extreme, exceptionally durable materials are best
handled by leaching in an "as is" condition. Column tests,
provided that channel effects are minimized or eliminated, are
more amenable to testing wastes "as is" with respect to waste
matrix size than are batch tests. Secondary tests which evaluate
strength in waste management environments include, for example,
the Toxicity Characteristic (TC) structural integrity test,
unconfined compressive strength, freeze-thaw (ASTM Method 4843-
88) and wet-dry (ASTM D4842-89) tests. Most batch leachate tests
require wastes to be tumbled. It has been shown that if wastes
are tumbled "as is", those which are not strong enough to survive
in the environment will, in fact, break up during tumbling, while
only strong materials will survive and remain intact (Bone et al,
"Modification of the TCLP Procedure to Accommodate Monolithic
Wastes," Fifth Annual Waste Testing and Quality Assurance
Symposium, July 24-28, 1989). Additional testing may be
necessary to determine whether the stabilized waste will remain a
monolith under varying environmental conditions. Thus it is
generally best to limit sample size reduction even in batch
tests.
12
-------
4) Leach test conditions (stresses) appropriate to the
situations being evaluated should be used for assessing long-term
contaminant release potential.
The best way to estimate the extent of contaminant release
from a waste matrix of interest is to have a test that reflects
realistic field conditions (Table 3, page 27). However, the
regulatory and statutory framework for decision makers often
seems to require that estimates be made for the maximum potential
for contaminant release in a specific scenario (Table 2, page
25). This involves testing under extreme conditions or stresses,
employing physical parameters such as fine particle size,
temperature and pH set at high contaminant solubility, high ratio
of leaching fluid to sample, high degree of mixing, and long
fluid/particle contact time. But, while it may seem necessary
from a regulatory and compliance perspective to use maximums in
leach tests, every effort should be made to design realistic
tests which simulate those actual worst case field leaching
conditions that can be reasonably postulated to occur at some
frequency in relevant waste management conditions.
In order to adequately characterize any particular field
scenario for leach tests, the relevant environmental conditions
postulated, and the degree to which they should be applied to
samples undergoing contaminant release testing, should be
carefully established and should take into consideration the
nature of the regulatory decisions that are required. Moreover,
any extrapolation of a set of conditions or stresses appropriate
for one purpose should not be applied for other applications
without reasonable verification of relevance. Extrapolation of
tests designed for one purpose to another purpose should be
scientifically defensible. A suggested approach to development
of an array of leach tests follows:
a) First, identify the set of regulatory decisions that
will be made with the test results.
The decision set could include (but is not limited to): (1)
a decision whether or not a waste should be classified as
hazardous; (2) the extent of leaching from a large volume waste
in order to determine suitability for waste utilization or
alternative management with or without restriction; (3) a
determination of release potential to support an estimate of risk
to human health or the environment; (4) a determination of
solidification/stabilization effectiveness to provide a basis for
a containment design; and, (5) selection of containment,
treatment or remediation technologies.
b) Convene a panel of individuals that represent a cross-
section of the regulatory community, the regulated community,
academia, and environmental/public interest groups for the
13
-------
purpose of defining the array of conditions for each type of
test.
inputs should be sought from the broad technical (expert)
community, for example, ASTM Committee D-34 on Waste Management
and other technically credible groups of scientists, engineers
and practitioners in the field. Consideration should be given to
obvious stress factors such as: the appropriateness of sample
size reduction, leaching fluid pH and buffering capacity,
leaching temperature, waste-to-leaching fluid ratio, number and
sequence of leaching steps, contact time, agitation mode,
biological action, and exposure to freeze/thaw as well as wet/dry
cycles.
c) Develop a hierarchial framework for reasonable stresses.
The set of possible stresses can be ranked from moderate to
severe, and identified with appropriate regulatory decisions,
depending on the conditions to which they are applied and the
importance of the decisions.
Using the above rationale, as the need for a new decision is
identified, its placement within and relevance to the decision-
making framework of prescribed stresses (and extant leach tests)
should be clear. While some procedures must be developed to make
decisions in direct compliance with regulations, and require
either site-by-site or type of application assessments, this
recommended exercise could help to avoid the inappropriate
application of a leach test or related test that has been
developed for another purpose.
5) Laboratory leach tests should be field-validated, and
release test accuracy and precision established before tests are
broadly applied.
Numerous tests have been developed by the EPA, ASTM and
others in order to estimate contaminant release and subsequent
transport through soil matrices (Table l, page 20 and Appendix C,
page 35). Some of the test methods have been subjected to
extensive precision studies involving multiple laboratory samples
and analysis, while other tests have clearly only been subjected
to minimal or very limited evaluation, as in single laboratory
precision studies. Furthermore, the accuracy of leaching test
results has not been, for the most part, subjected to field
verification. Anecdotal evidence suggests that the accuracy and
reliability of laboratory test results for field application are
questionable, primarily due to the simplifications or
approximations utilized. In principle, the accuracy and
precision of contaminant release predictions can be improved by
matching the controllable test variables more closely to the
environmental conditions that actually are encountered under
field conditions. But it is questionable whether many leaching
14
-------
tests are adequately predictive of some reasonable worst case
scenarios.
The number and types of analytes for which the tests have
been evaluated and applied are likely to evolve continuously.
Currently, all of the tests are designed for metals, semi-
volatile, and non-volatile organics, and only a few are valid for
volatile organics. Consequently, there are considerable data to
determine the precision of the tests for metals. This is
fortunate, since metals leachability is much more sensitive to
factors which are difficult to control, such as pH, ionic
strength and particle size. Thus, in view of current knowledge
about metals leachability, much can be inferred about the
fundamental processes involved in leaching of inorganic
constituents from wastes.
As was suggested in the SAB EEC Leachability Workshop and
technical briefing of May 9, 1990, review of the reliability and
precision of metals leachability test results leads the
Subcommittee to conclude that test precision is probably
satisfactory, particularly in comparison to the reliability of
test methods and variables associated with other factors which
are used in conjunction with leachability to arrive at
environmental risk assessments. Although the precision of any
test used for regulatory decisions or environmental risk
assessments should always be evaluated, test accuracy is more
important than test precision.
The state of scientific capability for leaching test
interpretation indicates that, in order to provide a realistic
estimate of the leachability of a specific waste in a given
environment, site-specific conditions must be fully considered.
Consideration must be given to all factors that have the
potential to impact leachability in either a positive or a
negative fashion. For example, cosolvent effects can greatly
facilitate the movement of contaminants out of the waste matrix,
while biological activity can increase or decrease the release of
contaminants as well as transform contaminants prior to their
release into the environment.
Therefore, the Subcommittee recommends that, through the
Office of Research and Development, EPA carry out a comprehensive
"field validation" of leaching tests and establish laboratory
accuracy and precision. The results should then be factored into
guidance for the improvement of leaching tests.
6. More and improved leach models should be developed and
used to complement laboratory tests.
Unlike the advances in models for transport and fate
predictions, development of mathematical models to predict
contaminant leaching is in its infancy. Only a small number of
15
-------
"leachate generation" models are proposed (e.g., HELP/ FOWL™,
UNIFAC), and to date these have had limited use.
While various laboratory tests can, in principle, be used to
physically "model" a contaminant release scenario, more and
improved "mathematical" models for leaching predictions should be
developed and employed to complement laboratory tests.
Simple equilibrium (or as warranted, more comprehensive
dynamic) models could be utilized to analyze data obtained from
various leaching tests. These investigations could then be used
to evaluate the applicability of such models as a leach test
adjunct, or in more direct application, as an approach to
estimating field leachability.
Contaminant release and transport models currently play an
important role in the Agency's regulatory decision-making
process. For example, the HELP model is used in the delisting
regulation to project hydraulic flux through landfills. Further,
the HELP model, in conjunction with the EPACML, is used to
predict the dilution attenuation of contaminants from the bottom
of a landfill to the nearest well. This and other models hold
the promise of significant utility if: (1) they are sufficiently
comprehensive and reliable for predicting the transport and fate
of contaminants of concern; and (2) the data base necessary
(including leachate composition) for model use is adequate and
reliable. These criteria limit the conditions under which a
model can be applied.
Potential pitfalls in the use of models should be examined
prior to their application, to ensure that their results are
reliable (Refer to the SAB Resolution on Use of Mathematical
Models by EPA for Regulatory Assessment and Decision-Making (EPA-
SAB-EEC-89-012), January 1989). Generally, such a review can
proceed consistent with principles outlined in prior SAB
deliberations on the generic use of models. Examples of concern
to be addressed include: Agency over-reliance on models to the
exclusion of the acquisition of needed data; the extent to which
models are based on a fundamental representation of the relevant
physical, chemical and biological processes that can affect
environmental systems; the extent to which models have been
validated with laboratory and field data; the analysis of
sensitivity and uncertainty impacts on models and model
predictions; and, the need to ensure adequate peer review of
model development and utilization. In circumstances where
laboratory and field data fail to confirm the adequacy of a
model, it is inappropriate to use the model for decision making
until improvements of an acceptable nature can be implemented.
Laboratory and field tests should be utilized to establish
the conditions under which model simulations can be used to
extrapolate laboratory and field data. A model should not be
16
-------
used to predict leachability or transport in scenarios that are
outside the scope of the model applicability.
The sensitivity of a model to specific input data parameters
should be established; this indicates the level of effort needed
in the determination of the parameters for model evaluation.
Pertinent questions include: How accurate should the input data
be? Can input data be estimated by analogy rather than obtained
from actual field measurements? Following these judgments, an
appropriate data base can be developed for a model input.
Analysis should attempt to identify whether a different
outcome might have been realized had more representative data
been available, i.e., the expected model outcomes associated with
varying degrees of data uncertainty should be established.
Consistent with input data requirements, a model can then be used
to predict leachability behavior in a specified field scenario.
7) To facilitate the evaluation of risk implications of
environmental releases, the Agency should coordinate the
development of leach tests and the development of Bodels in which
the release terms are used.
Leaching tests characterize the "source terms" for transport
and fate models. Yet almost all transport and fate models assume
that leachates are of constant concentration and of infinite
duration and quantity. In reality, source terms (leachates) are
a function of time, space, waste properties, and leaching fluid
characteristics.
Numerous models presently are available to describe the
transport of chemicals through porous material, including both
the saturated and unsaturated zones. The hydrologic or fluid
flow models (such as EPACML, HELP) could be improved to consider
the chemistry and microbiology of contaminant release within the
source waste matrix. The resulting leaching predictions would
then be dynamically included in the transport analysis. The
Subcommittee recommends that models used by the Agency be
modified to couple source leaching masses -with the transport and
fate predictions. Such linked models would more accurately and
precisely predict environmental concentrations to quantitatively
evaluate risk implications, albeit at the cost of greater
computational and data collection effort.
8) The Agency should establish an inter-office, inter-
disciplinary task group, including ORD to help implement these
recommendations and devise an Agency-wide protocol for evaluating
release scenarios, tests, procedures, and their applications.
The task group should also be charged with recommending what the
appropriate focal point(s), responsibilities, and organizational,
budgetary and communication links should be within the Agency for
the Most effective, continued and ongoing support and pursuit of
17
-------
the research, development and utilization of methods and
procedures.
The Subcommittee's discussions on Agency "needs" with the
various program offices pointed out that a variety of
applications and regulatory decisions depend on appropriate
"release" tests and waste matrix transport and fate analyses
(Table 2, page 25 and Appendix C, page 35). Therefore/ it is
recommended that an inter-office/ inter-disciplinary task group
be established to aggressively formulate implementation plans for
the development of scientifically defensible leaching tests and
models for the many contaminants and applications. This task
group should include experts in the field of hydrology, soil
science, analytical chemistry, environmental chemistry and
biology, mathematical modeling and environmental engineering.
9) Core research on contaminant release and transport
vithin the waste natrix is needed.
Consonant with the underlying concept of EPA's core research
initiative and its intent to provide and sustain knowledge and
expertise responsive to both current and future risks to human
health and the environment, it is apparent from the preceding
discussions that issues associated with leachability, and
specifically methods to adequately measure and predict leaching
from an assortment of waste matrices, are and should remain a
priority focus.
Unfortunately, consensus with respect to the use of
leachability testing protocols has yet to be attained. It
appears that methods currently advocated neither fully satisfy
short-term or long-term needs, nor do they withstand the rigors
of scientific scrutiny to an extent that scientifically
supportable management decisions can be made.
Therefore, the present state-of-knowledge concerning
leaching phenomena under a broad range of waste management
scenarios of regulatory and scientific interest should be fully
analyzed and reported. Based on the outcome of this task, an
integrated, active program of research, including exploration of
potential and actual risks associated with leaching of
constituents, should be developed.
The core research program regarding leachability should
embrace, to the maximum extent feasible, all underlying issues
pertinent to leachability. This should include, but not be
limited to the following: basic mechanisms, potential test
procedures, analytical methods, predictive model development,
performance standards, and regulatory initiatives. Accord-
ingly, a complementary understanding of the scientific and
operational issues of various waste management options, as well
as the intricacies of the associated environmental settings, is
18
-------
required. For instance, a possible approach may be to develop a
waste management matrix which defines both current and potential
future treatment, storage, use or disposal practices and
associated environmental circumstances, and then develop
companion testing protocols to simulate each situation. These
tests should consider the operational phase as well as the pre-
installation and post-closure periods.
Such an approach as recommended above should provide
better correspondence between the results of testing protocols
intended to simulate actual conditions under both short-term and
long-term conditions. Whether these objectives can be
accomplished with laboratory, pilot or field-scale simulations
would be part of the challenge of the core research initiative.
However, it could be anticipated that both short-term or
accelerated screening and long-term field-scale simulations may
need to be developed as an essential adjunct to each selected
waste management alternative. The ultimate goal would be to
provide operational as well as regulatory (and remedial) control,
thereby enhancing the potential for more meaningful assessments
of environmental and health risks.
19
-------
TABLE 1 - EXTRACTION TESTS
I. STATIC TESTS (LEACHING FLUID NOT RENEWED)
A. AGITATED EXTRACTION TESTS
TEST METHOD
TCLP (1311)
EP TOX (1310)
ASTM 03987-85
CALIFORNIA WET
LEACHATE EXTRACTION
PROCEDURE (HOE,
ONTARIO)
QUEBEC R.S.Q
(HOE, QUEBEC)
FRENCH LEACH TEST
(AFNOR, FRANCE)
EQUILIBRIUM
EXTRACTION
(ENVIRONMENT CANADA)
MULTIPLE BATCH
LEACHING
PROCEDURE
;f SIT ':?>t.<:.>;.>
NUMBER OF
LEACHING FLUID IIOUID:SOL!D RATIO MAXIMUM PARTICLE SIZE EXTRACTIONS
ACETIC ACID 20:1 9.5 mm 1
0.1 N ACETIC ACID
SOLUTION, pH 2.9, FOR
ALKALINE UASTES
0.1 N SODIUM ACETATE
BUFFER SOLUTION, pH 5.0,
FOR NON- ALKALI HE UASTES
0.5N ACETIC ACID 16:1 DURING EXTRACTION 9.5 mm 1
(pH«5.0) 20:1 FINAL DILUTION
ASTH TYPE IV REAGENT UATER 20:1 AS IN ENVIRONMENT 1
0.2 H SODIUM CITRATE 10:1 2.0 mm 1
(pH-5.0)
ACETIC ACID 20:1 AS IN ENVIRONMENT 1
2 HEQ/G
INORGANIC 0.02 MEQ/G 10:1 GROUND 1
ORGANIC DISTILLED UATER
DI UATER 10:1 9.5 nm 1
DISTILLED UATER 4:1 GROUND 1
ACETIC ACID 4:1 OR 9.5 irni VARIABLE
BUFFER, pH 4.5 2:1
TIME OF EXTRACTIONS
18 HOURS
24 HOURS
18 HOURS
48 HOURS
24 HOURS
24 HOURS
16 HOURS
7 DAYS
24 HOURS
-------
TEST HETHOO
UACHING FLUID
TABLE 1 - EXTRACTION TESTS (Continued)
LIQUlDjSOLIO RATIO MAX IHUH PARTICLE SIZE
NUMBER OF
EXTRACTIONS
TIME OF EXTRACTIONS
MATERIAL CHARACTER-
IZATION CENTRE-4
(MATERIAL CHARACTER-
IZATION CENTRE)
CHOICE
10:1
2 FRACTIONS
74 - 149 mm
ISO - 425 mm
1
20 DAYS TO
TO YEARS
OILY UASTE
(1330)
SOXLET WITH THF AND
TOLUENE EP ON
REMAINING SOLIDS
100G:300ML
20:1
9.5 inn
24 HOURS (EP)
SYNTHETIC PRECIPI-
TATION LEACHING
PROCEDURE (1312)
VARIABLE
20:1
9.5 mm
18 HOURS
EQUILIBRIUM
LEACH TEST
DISTILLED UATER
4:1
150
7 DAYS
B. NON-AGITATED EXTRACTION TESTS
TEST METHOD
LEACHING FLUID
NUMBER OF
LIQU|D;SOtlD RATIO MAXIMUM PARTICLE SIZE EXTRACTIONS TIME OF EXTRACTIONS
STATIC LEACH
TEST METHOD
(MATERIAL CHARACTER-
ISTIC CENTRE-1)
HIGH TEMPERATURE
STATIC LEACH TEST
METHOD (MATERIAL
CHARACTERIZATION
CENTRE-2)
CAN BE SITE
SPECIFIC
SAME AS ABOVE
BUT AT 100-C
VOL/SURFACE 10 7 DAYS
> 7 DAYS
C. SEQUENTIAL CHEMICAL EXTRACTION TESTS
TEST METHOD
LEACHING FLUID
LIQUIDrSOlID RATIO MAXIMUM PARTICLE SIZE
NUMBER OF
EXTRACTIONS
TIME OF EXTRACTIONS
SEQUENTIAL
EXTRACTION TESTS
0.04 M ACETIC ACID
50:1
9.5 mm
15
24 HOURS PER
EXTRACTION
-------
D. CONCENTRATION BUILD-UP TEST
TEST METHOD
LEACHING FLUID
TABLE 1 - EXTRACTION TESTS (Continued)
NUMBER OF
LIQUIDlSOlID RATIO MAXIMUM PARTICLE SIZE EXTRACTIONS
TIME OF EXTRACTIONS
SEQUENTIAL
CHEMICAL EXTRACTION
STANDARD LEACH
TEST, PROCEDURE C
(UNIVERSITY OF
FIVE LEACHING SOLUTIONS
INCREASING ACIDITY
Dl WATER
SYN LANDFILL
LEACHATE
VARIES FROM 150 |un 5
16.1 TO 40.1
10:1, 5:1 AS IN ENVIRONMENT 3
7.5:1
VARIES FROM 2
TO 24 HOURS
3 OR 14 DAYS
WISCONSIN)
II. DYNAMIC TESTS (LEACHING FLUID RENEWED)
A. SERIAL BATCH (PARTICLE)
TEST METHOD
MULTIPLE
EXTRACTION
PROCEDURE
(1320)
MUEP
(NONOFILL WASTE
EXTRACTION PROCEDURE)
GRADED SERIAL BATCH
(U.S. ARMY)
SEQUENTIAL BATCH
ASTM D4793-ft8
WASTE RESEARCH
UNIT LEACH TEST
(HARWELL LAB-
ORATORY, UK)
STANDARD LEACHING
TEST: CASCADE TEST
LEACHING FLUID
SAME AS EP TOX, THEN
WITH SYNTHETIC ACID
RAIN (SULFUR 1C ACID:
NITRIC ACID IN 60:40%
MIXTURE)
DISTILLED/DEIONIZED
WATER OR OTHER FOR
SPECIFIC SITE
DISTILLED WATER
TYPE IV REAGANT WATER
ACETIC ACID
BUFFERED pH=5
DISTILLED WATER
HMO] pM 4.0
NUMBER OF
L1QUID:SOLID RATIO MAXIMUM PARTICLE SIZE EXTRACTIONS TIME OF EXTRACTIONS
20:1 9.5 DID 9 (OR MORE) 24 HOURS PER
EXTRACTION
10:1 PER 9.5 mm OR 4 18 HOURS PER
EXTRACTION MONOLITH EXTRACTION
INCREASES FROM N/A >7 UNTIL STEADY STATE
2:1 TO 96:1
20:1 AS IN ENVIRONMENT 10 18 HOURS
1 BED VOL 5 ELUTIONS CRUSHING >11 2 TO 80 HOURS
10 BED VOL >6
ELUTIONS
20:1 CRUSHING 5 23 HOURS
SOSUV, NETHERLANDS
-------
B. FLOU AROUND TESTS
TABLE 1 - EXTRACTION TESTS (Continued)
TEST METHOD
IEACHIHC flUlO
IIQUID:SOLID RATIO MAXIMUM PARTICLE SIZE
NUMBER OF
EXTRACTIONS
TIME OF EXTRACTIONS
IAEA DYNAMIC LEACH
TEST (INTERNATIONAL
ATOMIC ENERGY AGENCY)
Dl UATER/SITE WATER
N/A
ONE FACE PREPARED
>6 MONTHS
ISO LEACH TEST
(INTERNATIONAL
STANDARDS ORGANI-
ZATION)
Dl UATER/SITE UATER
N/A
SURFACE POLISHING
>100 DAYS
ANSI/ANS 16.1
(AMERICAN NATIONAL
STANDARDS INSTITUTE/
AMERICAN NUCLEAR
SOCIETY)
Dl UATER
N/A
SURFACE UASHING
11
90 DAYS
K)
UJ
DLT
Dl UATER
C. FLOU THROUGH TESTS
N/A
SURFACE UASNING
18
196 DAYS
TEST METHOD
STANDARD LEACHING
TEST: COLUMN TEST
LEACHING FLUID
Dl UATER
HNOj pH«4
LIOUID:SOLID RATIO
10:1
NUMBER OF
MAXIMUM PARTICLE SIZE EXTRACTIONS
AS IN ENVIRONMENT 7
TIME OF EXTRACTIONS
20 DAYS
(SOSUV, THE
NETHERLANDS)
COLUMN ASTM 04074-09 TYPE IV REAGENT UATER
ONE VOID VOLUME
AS IN ENVIRONMENT
24 HOURS
-------
111. OTHER TESTS
TABLE 1 - EXTRACTION TESTS (Continued)
TEST METHOD
LEACHING FLUID
NUMBER OF
IIQU|D:SOLID RATIO MAXIMUM PARTICLE SIZE EXTRACTIONS
TIME OF EXTRACTIONS
HCC-5S SOXHLET TEST
(MATERIAL CHARACTER-
ISTIC CENTER)
DI/S1TE UATER
100:1
CUT AND WASHED
0.2 ML/HIM
ACID NEUTRALIZATION
CAPACITY
HNO, SOLUTIONS OF
INCREASING STRENGTH
3:1
150
-------
TABIiB 2 - TEST REQUIREMENTS, USES OF TESTS, AMD PROGRAMATIC
MEEDS FOR LEACHABILITY TESTS BT THE AGEHCT
AS PERCEIVED BY THE SCIEHCE ADVISORY BOARD'S
EHVIROMMEHTAL ENGINEERING COMMITTEE
PROGRAM OFFICE *
TEST REQUIREMENTS 8F SC O8W OTS RSKERL OMMQA
Simple method O z Z O O Z
Field Analocr
Model Source Term
Predict Leachincr
Method Validation
TIE Compatible
Surface Water Interface
Mismanaaement Predictor
Waste Classification
"No Reasonable Risk"
Determination
USES OF TESTS
Demonstrate Federal/
State Compliance
Simulate Risk
Set Standards
Compare Waste
Management Strategies
Compliance/
Clean-Up Goals
Examine "Worst Case"
Apply to Human Health
Identify Toxicants
New Product Information
Establish "Equivalency"
To Thermal Destruction
Evaluate Lined and
Unlined Units
Z Z 0
Z O Z O
Z Z O
Z Z 0
Z 0
Z Z O
z z z
z z z z
z
z z z z
z z z z
Z 0 Z Z
Z 0 Z Z
ZOO O
Z Z 0
0 Z 0 0
0 Z Z Z
z
z
Z Z 0
O
z
z
z
O
z
z
O
O
z
O
O
0
z
O
z
O
0
z
z
z
O
z
z
0
z
z
O
z
z
z
O
z
O
0
O
25
-------
TABLE 2 - (Continued)
PSOGSAM OFFICE *
PBOGRAMATIC HEEDS FOB TESTS 87 SC OSW
Flexibility X X O
Multiple Tests X X O
Standardized Protocols X X
Compliance XXX
Remedial Desicrn X O O
Bioloaical Response O X O
Matrix Data XXX
EP/TCLP X X
Acid Rain Leaching1 O O X
Multiole Extraction O O X
Oily Waste Extract O O X
Equivalent Tests
Non-Destructive Tests O O O
* SF = Super fund Program, US EPA
SC = Office of Water, Sediment Critria
OSW = Office of Solid Waste, US EPA
OT8 = Office of Toxic Substances, US EPA
OTS
O
X
O
O
O
O
X
X
X
BSXESL
O
X
0
X
X
0
X
X
X
X
X
0
OMMQA
O
X
O
X
0
O
X
O
O
X
X
O
0
Program, US EPA
RSKERL = R.s. Kerr Environmental Research Lab, US
OMMQA = ORD/Office of Modeling, Monitoring
Assurance, US EPA
X = Yes, a Blank Signifies No
0 = Sometimes or Occasionally
, and
EPA
Quality
26
-------
TABLE 3 - SCIENTIFIC COH8IDERATIOHS IH DKBIGH AND
IHTERPfiETATIOH OP LEACHABILITY. TESTS
PROPERTIES/CHARACTERISTICS
Source Matrix Properties
IMPORTAHCE AND TEST RAMiriCATIOHS
Matrix properties affect availability
and accessibility of contaminants.
Chemical composition (functional Relates to containment and leaching
groups/ carbon content, etc.) environment.
Morphological structure
(amorphous vs. crystalline)
Surface area (surface-to-
volume ratio)
Surface physics (e.g./ charge/
tension)
Matrix heterogeneity
Pore structure (volume/
distribution)
Pore liquid volume (degree
of saturation)
Pore liquid composition
Permeability
Ease of saturation (time,
pressure required)
Pooling (micro- and macro-
reservoirs/ field capacity)
Biodegradability
Toxicity
Affects access and containment.
Determines interface for sorption;
can affect pE.
Controls access and flow in the
pores matrix.
Creates morphological and chemical
differences.
Constricts flow/ retains gas/
serves as "small" reactor.
Contributes to effective leachate
volume.
Modifies leachate composition.
Affects flow regime and residence
time.
Leachant/vaste interface may be
limited by the rate and extent of
saturation.
Extracting fluid ("leachant") and
leachate retention; provides in
situ reaction opportunity.
Matrix can change properties due
to biodegradaton.
May limit biodegradation of the
contaminants or matrix property
changes.
27
-------
TABLE 3
PROPERTIES/CHARACTERISTICS
Buffer capacity
2. Contaminant Properties
Chemical composition
(continued)
33CPORTAHCE AHD
B*WTFICATIOMS
Concentration
Toxicity
Biodegradability
Heterogeneity
Diffusivity
Solubility
Volatility (boiling point,
Henry's constant)
3. Leachant Properties
Initial chemical composition
Aqueous/non-aqueous
Availability of buffer capacity
affects leachability and diffusion,
and regulates efficiency and
nature of biodegradation.
Leachability is a function of the
nature of the contaminants.
Different chemicals or the same
contaminant in a different
physical or chemical form
exhibit distinct differences
in leachability.
Concentration gradients affect
leaching rate and equilibria.
Reduces efficiency of concurrent
bio trans format ion.
Compound structure and environmental
conditions affect biodegradability.
Affects containment and availability
for reaction and/or leaching.
Determines transport via diffusion.
May affect mass transport and limit
removal and/or reaction.
controls liquid/vapor phase transport
and loss of contaminants during
leaching and analysis.
Leachant properties control
solubilization/dissolution and mass
transport processes.
Could contain contaminants, may be
aggressive, may change as the
leaching process proceeds.
Normally water (distilled, tap, or
site) as modified by test protocol.
28
-------
TABLE 3 (continued)
MTTBRISTICS IMPORTAHCB AHD TEST BAMIFICATIOHS
Hydrophobic-hydrophilic
nature **
Gas (oxygen, carbon dioxide
content, etc.)
Density
Buffer capacity
Viscosity
pH
4. Fluid Dynamics
Flow gradient
Flow regime (laminar vs.
turbulent)
Surfactants or hydrophobic solvents
could enhance leaching.
Affects pH and nature of biological,
physical and chemical interactions.
Contributes to hydraulic gradient
effects and hydraulic conductivity.
Controls pH change.
Affects flow regime, saturation, and
hydraulic conductivity.
May or may not control the leaching
process.
Fluid dynamics in a given system
dictate contact time and opportunity,
vhich affect reaction extent and mass
transfer. Fluid dynamics have
important ramifications to the
selection of leaching mode, e.g.,
continuous column, batch, sequential
batch; equilibrium (intended to
represent specific, worst case
scenarios), non-equilibrium; dynamic
(mixed), static.
Affects transport of contaminants by
dispersion, convection, and
advection. Also affects mechanism of
mass transport. Agitation may be
used to generate a maximum gradient.
Flow path could lead around or
through the waste. Cracks or inter-
connected pores would short-circuit
flow.
Affects contaminant transport and
gradient. Restricts the applica-
bility of Darcy's Law.
** NOTE: This is also a contaminant characteristic which determines
affinity to leach or to be bound in a matrix.
29
-------
TABLE 3 (continued)
PROPERTIES/CHARACTERISTICS
Flow pattern (intermittent vs
continuous)
System Properties
Precipitation/dissolution/
reprecipitation
Solubilization (capacity/
limits)
IMPORTAHCE AHD TEST
Could weather and/or disintegrate
waste matrix. Impacts on the
concentration gradient and
transport.
Operationally determined by the
leachant/vaste interaction.
Potential removal/release process,
Ability to remove contaminants by
dissolution.
Other chemical reaction and
reversibility
Complexation
Sorption/desorption
Partitioning
Cosolvency
Common ion effect
Redox environment
pH
Temperature
Mass transfer or equilibrium
limitations
May change contaminant behavior
and/or structure.
Affects transport, solubilization
and possible surficial binding
of metals and organometallic
compounds.
Contributes to the retention or
removal of solutes.
Affects equilibrium opportunity and
spatial and temporal distribution.
Enhanced removal by solvent
mixtures, affects distribution
of solutes.
Could delay the removal of contamin-
ants associated with More than one
anion.
Could affect opportunity for biolog-
ical or chemical transformations
and reactivity.
Major influence on biological,
physical and chemical transformation
processes.
Affects reaction rates, solubility,
pore pressure, etc.
Need to determine which dominates.
30
-------
TABLE 3 (continued)
PROPERTIES/CHARACTERISTICS PCPORTAHCE AKP TEST RAMIFICATIOH8
Dependence
Contaminant recharge
Temporal and spatial limitations may
accelerate or retard leaching.
Important element of tests involving
site-specific simulations.
Aging dynamics
Weathering effects (dis-
solution surface washing,
wet/dry, freeze/thaw)
Physical and chemical properties may
change in time.
Long-term humidity and temperature
changes affect matrix integrity.
Biodegradability
Barometric fluctuation
Leachant volumes (contact
time)
7. Monitoring methods
Precision/accuracy (overall)
Environmental sampling/
sample preservation/holding
time (environmental samples
and leachate test samples)
Leaching test
Leachate preservation/storage
Aerobic and anaerobic transformation
of and within the matrix.
Impact gradients, dispersion,
and groundvater movement, and
behavior of gases and volatiles.
Major consideration vhen selecting
the leachant/waste (or source
matrix) ratios. Leachant/source-
matrix interface over an extended
period of time could result in the
depletion of the contaminant or in
the erosion of the matrix.
Method of monitoring could influence
the test and their results.
To be defined by the data quality
objectives.
Affects results; plans and standards
•ay be available.
Selected in accordance with
objectives. Considers time,
environmental conditions, and
site specifity.
Sample components change with time.
31
-------
TABLE 3 (continued)
PROPERTIES/CHARACTERISTICS IMPORTAHCE AHD TEST
riCATIOHS
Analytical (sample preparation
and test method)
Testing schedule (time)
Reproducible/ specific/ and efficient
methods are available. Analytical
procedures need to be appended vith
appropriate protocols.
Could affect reproducibility,
interpretation/ and comparability
of data.
8. Physical Modeling
Comparability vith scenario
to be simulated
Congruence vith scenario
Similar in form and arrangement.
Governs applicability of results.
32
-------
APPENDIX A - BACKGROUND ON LEACHABILITY AS A
SELF-INITIATED ACTIVITY OF THE
SCIENCE ADVISORY BOARD
Over the past decade, the Environmental Engineering Committee
(EEC) of the Science Advisory Board (SAB) has reviewed a number of EPA
subjects and issues involving leachability phenomena either as a major
or minor factor in the review. In these various reviews/ the Committee
has noted a number of problems and issues, relating to leachability
phenomena that were common to a variety of programs, rules and Agency
procedures. The Committee believed that these common problems and
issues, would be best called to the Agency's attention through a
general set of recommendations on leachability phenomena, rather than
in the specific, individual reviews.
Believing that the scientific principles of contaminant
leachability need broader understanding and exposition, the EEC has
undertaken the initiative, with the concurrence of the Executive
Committee, to conduct this self-initiated review to:
1) Consider the fundamental scientific principles that can
reliably describe contaminant release/transport. In particular, to
consider the controlling characteristics of the source, the leaching
media and the importance of dynamic considerations; and
2) Suggest how the scientific principles can be applied to
determine how a waste will leach when present in the environment,
according to a prescribed scenario.
The Leachability Subcommittee (LS) was formed by the EEC. The
group convened a project scoping and planning session in Houston, Texas
on December 15-16, 1989, immediately following a Workshop related to
this topic. The LS then followed this with a one-day session in EPA's
Headquarters Office in Washington, D.C. on February 26, 1990, devoted
to assessing the Agency's varied needs for leachability-related
information. The day's activities and findings were then discussed
with the full EEC on February 27, 1990. This was followed by a
Workshop on Leachability on May 9, 1990 in Washington, D.C. The
Workshop was conceived as a vehicle for distinguished scientists,
engineers and practitioners in the field to focus on the scientific
principles and issues relating to leachability phenomena. The Workshop
was video taped, so that those unable to attend from EPA or any other
interested parties could have the benefit of this exchange of
information.
The Leachability Workshop assisted the LS of the SAB's EEC to
better define the fundamental scientific principles that control
leachability. Further, the workshop assisted the SAB and the attendees
in ascertaining how leachability phenomena and tests can be applied on
an appropriate and consistent basis to determine how a waste will leach
when present under various scenarios in the environment.
33
-------
APPENDIX B
LEACHABILITY WORKSHOP PROGRAM
May 9. 1990
Welcome and Administrative Remarks Or. C.H. Ward
Mr. Richard A. Conway
Statement of Issues and Needs Dr. Raymond C. Loehr
Test Methods: Descriptions, Dr. Pierre C6te'
Capabilities, Organics-Inorganics
Leaching of Stabilized Materials Dr. Paul Bishop
Physical-Chemical Mechanisms: Dr. Marvin Dudas
Concepts on Interactions of Solids-
Liquids, Liquid-Liquid, Sol ids-
Li quids'Gases
Technical Problems and Challenges Mr. Robert L. Huddleston
for Regulators and the Regulated
Leaching Chemistry of Inorganics Dr. John M. Zechara
Leaching Chemistry of Organics Dr. P. Suresh Rao
Alternative Approaches to Dr. Carl Enfield
Laboratory Tests (Modeling)
Concluding Remarks Dr. C.H. Ward
Dr. Ishwar P. Murarka
CONVENERS AND SPEAKERS
Dr. C.H. Ward, Chairman, Leachabitity Subcommittee, Rice University, Houston, Texas
Dr. Ishuar P. Murarka, Vice-Chairman, Leachability Subcommittee, Electric Power Research Institute, Palo Alto, California
Mr. Richard A. Conway, Chairman, Environmental Engineering Committee, Union Carbide Corporation, South Charleston,
West Virginia
Dr. Raymond C. Loehr, Chairman, Science Advisory Board, University of Texas, Austin, Texas
Dr. K. Jack Kooyocmjian, Designated Federal Official, US EPA, Science Advisory Board
Dr. Donald G. Barnes, Director, US EPA, Science Advisory Board
Mr. A. Robert Flaak, Assistant Staff Director, US EPA, Science Advisory Board
Dr. Pierre Cote1, Zenon Environmental, Inc., Burlington, Ontario, Canada
Dr. Paul Bishop, University of Cincinnati, Cincinnati, Ohio
Dr. Marvin Dudas, The University of Alberta, Edmonton, Alberta, Canada
Mr. Robert L. Huddleston, Conoco, Inc., Ponca City, Oklahoma
Dr. John M. Zachara, Battelle Pacific Northwest Laboratories, Rich Iand, Washington
Dr. P. Suresh Chandra Rao, University of Florida, Gainesville, Florida
Dr. Carl Enfield, R.S. ICerr Environmental Research Laboratory, Ada, Oklahoma
34
-------
APPENDIX C - LEACHASaiTT NEEDS. USES, TESTS. COKCEXXS. AMD ISSUES
C-1 - SUPERFUW REMEDIAL AMD KENOVAL PKOG8ANS
KED FOR USE OF LEACHIMfi TESTS USES OF LEACXAflRITT AMD LEACH TESTS
- Need to comply with Federal and State laws that are
applicable, or relevant and appropriate (e.g.,
RCRA)
- Need to approximate real world conditions
- Need for methods which provide input to groundwater
modeling
- Need for standardization of Leaching methods
(including K^ for specific applications and data uses
- Need for methods development for predicting long-term
Leaching potential
- Need for validation of Leaching methods
• Leaching, extraction and other chemical test are
typically needed to provide a variety of data on
either untreated or solidification/stabilization
-------
C-1 - SUPERFUND REMEDIAL AMD REMOVAL PROGRAMS (Continued)
TYPES Of DATA REQUIRED, SPECIAL CONCERNS, ISSUES AMD
CONSTRAINTS (e.g., variables that can affect test
results.)
- Superfund allows flexibility in choice of
teachability test based on site specific
conditions and needs
- National Contingency Plan (NCP) outlines program
goals and expectations which drive the remedy
selection process. Some of the expectations are:
- EPA expects to use treatment to address principle
threat wastes (e.g., highly toxic, mobile, etc.)
- EPA expects to use engineering controls (e.g.,
containment) to address wastes which pose a
relatively low long-term threat or where
treatment is impracticable
- EPA expects to return usable ground waters to
their beneficial uses whenever practicable
- Parameters that can affect test results include:
- Sample heterogeneity
- Curing time
- Liquid-to-solid ratio
- Extraction time, number and frequency
- Leaching medim
- Superfund is unique in that the program can use
flexibility on a site-by-site and case-by-case
technical basis for selection of remedy:
- Additional goals in the NCP aim to treat waste
that are the principal threat (e.g., highly
mobile, toxic, etc.)
- Probably wilt not excavate and treat non-mobile
wastes which are not the principal threat
- Generally, there is an expectation goal to
reduce toxicity, nobility and volume of the
waste by 90S to 99X (whether or not this
presumption is valid)
- Superfund may use multiple tests. Various leach
tests (e.g., 18 hours versus 90 days or more) are
employed. There is a large variety, depending on
the waste and the disposal scenario:
- The more agreement and consensus on appropriate-
ness of tests, the acre iirportant are the tests
to the program (e.g., decisions in New Jersey
and California should be consistent.)
- Superfund waste recsoval activities have wore
flexibility than remedial activities. However,
the science and technical decisions wjst withstand
scrutiny.
- For Superfund, RCRA protocols lay not be
appropriate because:
- The Superfund application is • different
purpose than for what the RCRA protocol
was originally devised.
- Lack of standardized protocols in Superfund
- Many aethods exist with • lot of variability
- Technical uncertainty on waking sense of the
varied forms of data to nake a decision.
TYPE OF TESTS
- Note that all treated wastes (not just solidified/
stabilized (S/S) treated wastes) need to be evaluated
to determine how protective they are in specific
Management scenarios
While Superfund program must comply with Federal
and State regulatory requirements (e.g., RCRA TCLP
methodology), other tests may be utilized to
approximate real world conditions. Single or multiple
Methods may apply to a site. Additional methods may
include the following listed below:
- Short-ten* extraction tests (hours to days)
- Leaching tests (weeks to years)
- Column Leach Test
- EP (Method 1310) (Was used in the past, but is
superseded by the TCLP)
- TCLP (Method 1311)
- TCLP with cage modification (This has never been
promulgated and appears to have reproducibility
problems)
- California Waste Extraction Test (Cal WET)
- Multiple Extraction Procedure (MEP)
- Synthetic Acid Precipitation Leach Test
- Monofilled Waste Extraction Procedure
(Hethod 1312) (HWEP)
- Materials Characterization Center Static
Leach Test
- American Nuclear Society Leach Test
- Dynamic Leach Test (DLT)
- Shake Extraction Test
- Others
36
-------
C-2 - OFFICE OF UATEX, SEDIMEKT OUTESIA PROGRAM
HEEDS FOR USE OF LEACHIMG OR OTHER TESTS
Simple methods that can be simply used by field
people for all types of sediment and water
environments to address the toxicity of sediment:
- Heed for an in-the-field practical method where
each method is not turned into a "research project"
- The AET (Apparent Effects Threshold) method uses
a preponderance of evidence approach
- The AET method has no relevance to sediments
that are exposed to leaching conditions
- Addresses the toxicity of in-piace sediments
Methods to provide sediment criteria decisions to
evaluate the following:
- Most likely scenario
- Methods to be applicable to human health, aquatic
life or wildlife protection
- Ability to generate numerical criteria for
specific chemicals
Toxicity Identification Evaluation (TIE) procedures
to identify and quantify chemical components
responsible for sediment toxicity, such as:
- Techniques for the identification of toxic
compounds in aqueous samples containing
Mixtures of chemicals
- Interstitial water toxicity method TIE procedures
are implemented in three phases to evaluate:
- Pore water toxicity,
• Identify the suggested toxicant, and
- Confirm toxicant identification.
USES OF LEACHABILITT AW) LEACH TESTS
- Sediment criteria decisions for:
- Applicability of Method to human health, aquatic
life or wildlife protection
- Predicting effects on different organisms
- Suitability for in-place pollutant control
- Suitability for source control
- Suitability for disposal actions
- Suitability for different sediment types
- Suitability for different chemicals or classes
- Ability to generate numerical criteria for
specific chemicals
- Toxicity Identification Evaluation (TIE) potential
uses:
- Use of pore water as a fraction to assess
sediment toxicity
- In conjunction with TIE procedures, can provide
data concerning specific compounds responsible
for toxicity in contaminated sediments
- Ability to identify specific toxicants
responsible for acute toxicity in contaminated
sediments
37
-------
C-2 - OfFICE OF WATER, SEDIHBCT CRITERIA PROGRAM (Continued)
TYPES Of DATA REQUIRED. SPECIAL COMCEJUJS. ISSUES AND
COXSTSAIirTS (e.g., variables that can affect test
results.)
- Types of data required:
- Biological response data (either acute or
chronic)
- Choice of Test Organism
- Practical concerns of method choices:
- Ease of use
- Relative cost
- Tendency to be conservative
- Level of acceptance
- Ability to be implemented by laboratories with
available/typical equipment end handling facilities
- Degree to which results lend themselves to the
following:
- Interpretation
- Environmental applicability
- Accuracy and precision
- Concern for leaching of hazardous substances or
hazardous materials to food, plants, groundwater
and sediment
- Concern for monofilling and impacts on groundwater:
- There are differences in sorptive capacity
(reductions) observed in monofill versus an
increase in sorptive capacity in well-aerated soil
(plow zone)
- Groundwater to surface water issues
- All current sediment criteria development efforts
address the toxicity of in-plaee sediments. As a
result, research activities have not focused on what
happens to sediment-bound chemicals when exposed to
leaching conditions:
- Some research suggests that for non-ionic organic
contzrainants, the presence of organic carbon in
leaching materials nay be responsible for some
binding of contaminants, as well as Mobility
- For Metals, it is possible that the leaching
conditions Might provide for the release of
significant levels of metals, because of the
expected reduction of the acid volatile sulfide
(AVS) content of Meny sediments
- AVS binds up significant levels of Metals and
it lost when exposed to oxidizing conditions
TYPE Of TESTS
Aeration tests
B i oassays
Equilibrium Partitioning Approach
Solid phase extraction tests
Graduated pH test
Filtration tests
Reversed phase. Solid Phase Extraction
(SPE) tests
Oxidant reduction test
EDTA addition test (The EDTA - Ethylene
diaraine tetraacetic acid test)
Hollow block with a semi-permeable
membrane, which is inserted into the
sediments
38
-------
C-3 - OFFICE OF SOtID WASTE
IEH>S FOR USE OF LEACHING TESTS:
Statutory requirement of RCRA to look at
reasonable worst-case mismanagement and
Maintenance scenarios to assess:
- Landfill scenarios (sanitary and monofill)
- Maintenance scenarios
- "Mismanagement11 scenarios
- Acid leaching scenarios
- Pump and treat systems
- Effect of covers
- Effect of Liners
To characterize the leachate source term, such as:
- Finite versus infinite source of waste
- Effect on leachate quantity
- Effect on leachate quality
To determine if a waste is hazardous or non-
hazardous. (Such determinations are built into
RCRA.):
To determine how a particular material needs to be
managed:
- Need a decision-making tool to evaluate
leachabi City
- Need to examine various scenarios and their
validity
USES OF LEACHABtLm AND LEACH TESTS:
To model and simulate risks
To assess risks
To direct regulatory decisions
To come to grips with the complexities of
teachability phenomena (i.e., the source term itself
is very difficult and complex)
To answer questions pertaining to what kind of
contaminant levels are appropriate to be left in the
soil or removal from the soil (e.g., clean closure)
TYPES OF DATA REQUIRED, SPECIAL CONCQUIS,
ISSUES AND CONSTRAINTS (That is, variables
that can affect test results.):
- Factors affecting teachability include, but
are not necessarily limited to the following:
- Size of materials
- Permeability of solid
- Time dependency of leaching
- Contact time (a crucial issue_
- Type of contact (tumbling versus stirring)
- Ratio of leaching fluid to waste (i.e.,
solubility versus mass-limited; also,
infinite source versus finite source)
- Changes in the waste itself (e.g., doe to
biodegradation, chemical changes, anaerobic
versus aerobic, hydrology, and climate
changes)
- Issue of scenarios where organics may be
bound better under acid, rather than neutral
or basic circumstances (Many industrial
landfills are highly on the basic side)
TYPE OF TESTS
- EP (Extraction Procedure) (Method 1310)
- TCLP (Toxicity Characteristic Leaching
Procedure) (Method 1311)
- TCLP with cage modification
- Acid rain leaching tests for large volume
wastes (Method 1312)
- Multiple Extraction Procedures (MEP) for
delisting
- Oily Waste Extraction Procedure (OWE?) for
delisting
39
-------
C-4 - OFFICE OF TOXIC SUBSTANCES
HEEDS FOR USE OF LEACHING TESTS
-TSCA is a "No Unreasonable Risk" Statute
- The only alternative technologies acceptable, for
instance for PCB's, must demonstrate that they are
equivalent to thermal destruction of PCB's (e.g.,
solidification of PCB's would cause a problem under
these TSCA criteria.)
- TSCA is also a cost-benefit statute:
- Possible use of uaivers by EPA Regional
Administrators
- With this (cost-benefit) constraint, there
could be a problem between what the Superfund
and the Office of Toxic Substances programs
says are "low" and/or acceptable concentrations
- Need some tests to determine the long-term
effectiveness of these problems
USES OF LEACHABILITY AND LEACH TESTS
- Particularly interested in teachability data,
especially for new product information in the PHN
(Pre-Hanufacturing Notification) program for new and
existing chemicals
- OTS is in need of information on teachability for
establishing the equivalent comparison in a chemical
waste landfill, which is the only non-destructive
method that is authorized (that is not to say that
treatment processes that are non-destructive
would not be examined)
TYPES OF DATA REQUIRED, SPECIAL CONCERNS,
ISSUES AMD CONSTRAINTS (This is, variables
that can affect test results.)
- Equivalency test data, for alternative
technologies to thermal destruction
- Cost-benefit data on alternative
technologies to incineration:
- The Office of Toxic Substances 1s in
need of inf onset ion that examines the
alternative technologies to incineration
- The alternative technologies to incineration
Bust demonstrate equivalency
- Need to know what kind of teachability
criteria would be needed to obtain
a revised equivalent to incineration
- Also, need to know what kind of laboratory
testing should be required to give results
equivalent to incineration
- RID is needed to answer above.
TTPE OF TESTS
- Equivalency tests (as compared to the
incineration alternative) need to be developed
- At present, disposal of waste in chemical
waste landfill is the only non-destructive
Method that is authorized
- Other treatment processes that are non-
destructive could be possible, but to date,
no research has occurred to develop such
non-destructive tests which demonstrate the
equivalence to incineration
- Need some tests to look at the long-term
effectiveness of these problems
40
-------
C-5 - UO PERSPECTIVE OF THE R.S. EERR LAB. IN MM, OKLAHOMA
HEEDS FOR USE OF LEACHIKG TESTS
The RCRA statute requires development and use of
teachability data and tests
Mission of the Lab. and the technical support center
is to further understanding of subsurface media to:
- Remediate contaminated vadose and saturated zones
- Provide a technical support bridge between the
Lab. and the EPA regional and state regulators
- R.S. Kerr Lab. is primarily a user and not a
developer of different Leaching test methods
and alternative procedures:
- EPA ORD, EPA OSU, ASTM, NRC and others develop
leaching tests
USES OF LEACHA8ILITY AM) LEACH TESTS
To answer questions pertaining to what kind of
contaminant levels are appropriate to be left in the
soil, so that there will not be a problem later
(i.e., so that leachate through the soil will be
protective of groundwater). Hence, this leads to
the following questions:
- What kind of leaching test can be used to
develop criteria for "safe levels?"
- What site specific processes must be considered
in the criteria development process?
- What kind of vadose zone models should be used?
TYPES OF DATA REQUIRED, SPECIAL CONCERNS.
ISSUES AMD CONSTRAINTS {That is variables
that can affect test results.)
- To develop a "fanrily of procedures" and to
determine when it is appropriate to use each
procedure to answer site-specific questions
- Ability to get source term in model is a
problem
- Need to act conservatively, particularly *
with organics, but need to use "common
sense*
- There is considerable competition between
different tests
• To date no evaluation has been mede as to what,
in fact, are the reasonable stresses, such as:
- When to grind or not to grind a waste
- When to apply acid and at what strength
and duration
- Issues with appropriate tests for
radionuclides and mixed wastes:
- What kind of test will be applicable to
low-level radioactive material?
- Is EP Tox. or TCLP appropriate? Under
what circumstances?
- There are no "standard" RiD leaching tests
or migration potential evaluation protocols
for Superfund on-site remedies
- For the coobustion residue area, the
following issues illustrate special concerns
and constraints:
- The ability to get the appropriate source
ten* is usually a basic problem
- The RSD program uses a lot of different
procedures to develop a credible data base
- In this context, ORD is primarily a user.
and not a developer of Methods for leaching
tests
- The RIO staff are using a lot of different
procedures to develop a credible data base
TYPE OF TESTS
- EP (Method 1310)
- TCLP (Method 1311)
41
-------
C-6 - RID PERSPECTIVE OF THE OFFICE OF MODELING,
MONITORING, AND QUALITY ASSURANCE, WASHINGTON. D.C.
USES OF LEACHING TESTS
- Determining regulatory status of waste
- Determining effectiveness of treatment
processes which are designed to reduce
teachability
- Determining teachability of waste inder
different management scenarios
TYPES OF TESTS NEEDED (SCENARIOS TO BE
MODELED)
- Sanitary Landfill co-disposal:
- Lined unit
- Unl ined unit
USES OF TRANSPORT TESTS
- Predicting transport through media as input to
fate and transport Bodels (i.e., serve as the
source ten* for the models)
TYPES OF TESTS NEEDED (ENVIRONMENTS TO BE
MODELED)
- Vadose Zone:
- Moist
- Dry
- Saturated Zone:
Honowaste disposal
Dedicated, nixed waste unit
Uncontrolled contaminated soil
- Sandy
- Clay
- Loan
- Calcarious
- Acidic
Effect of waste on LeachebiIity/transport
of nterial from other wastes
42
-------
C-7 - CRITERIA AMD OBJECTIVES Of LEACHING TESTS MO MODELING CONSIDERATIONS FOR PARTITIONING TESTS FRCH THE
PERSPECTIVE Of REGULATORS VEBSUS INDUSTRY
INDUSTRY VIEW OF THE PERSPECTIVE
OF THE REGULATORS
- Sinplicity (e.g., simplicity of test protocol)
• Reproducibility
- Conservatism, if real is* is not practical
- One/few tests to fit tost vastes
- Use to delineate "non-hazardous" versus "hazardous"
wastes (e.g., differentiating between wastes needing
RCRA "C" standards management and wastes that only
need "DH controls.)
[NOTE: The above is in contrast to a large part of
the Agency's response, particularly
Superfund and the Kerr Lab.]
PERSPECTIVE OF INDUSTRY
- Simplicity only where appropriate,
- Reproducibility
- Realisia in approximating site and waste specific
conditions
- Test variations to account for waste differences
- Suitable for differentiating between different
disposal conditions
- Use to define acceptable disposal conditions
TYPICAL CURRENT PRACTICES OF INDUSTRY IN LEACKATE
TESTI IK
- Determine regulatory status of waste, such as
hazardous waste e.g.. Subtitle C which is subject to
land ban), EP, TCLP and SPLP (e.g.. Method 1312)
- Determine reasonable worst-case releases from waste
residual oils, and contaminated media
- Determine effectiveness of waste treatment process
(e.g., solidification/stabilization)
- Provide basis for landfill design
- Account for biological activity (rarely) in soil
colim effects
- Deal primarily with large volune. Mono-filled
wastes
DISTINCTIONS TO BE HADE BETWEEN LEACHING
TESTS AND MODELING COHSI DERATIONS FOR
REGULATORS AMD INDUSTRY
- There is a distinction between leaching
tests and partitioning tests:
- Current leaching test* are
standardized regulatory tests
not requiring site-specific
analyses (e.g., Modeling)
- Partitioning tests are used for data
input into mathematical models for site-
specific subsurface Migration analysis
- Leaching tests are not for site-specific
analysis of contaminant transport
- Regulators and Industry do not want to
spend time, money and resources applying
leaching tests which are not:
- Applicable for given conditions
- Realistic, given site specific conditions
type and strength
- Cannot be used to identify prudent waste
management and disposal activities
- "False positives" and "false negatives" are
costly from both the regulator and industry
point-of-view
PERSPECTIVE OF PUBLIC
- Desire for conservatism
- Tests subject to public review
- Reflective of reasonable worst-case
environmental transformation
43
-------
C-7 - CRITERIA MD OBJECTIVES OF LEACHIHG TESTS A» NODELIBC COHSIDERATIWS FOR PARTITIOMIMC TESTS FROH THE
PERSPECTIVE OF REGULATORS VERSUS IW3USTRT (Continued)
BASIC HEEDS
o A Rational Construct of the Lcachinfl Phenomenon
- Avoid regulatory expediency, conservatism
- Aim for well founded approach
o Understand and Clearly Define the Role of Basic Mechanisms
- Weathering, Dissolution
- Desorption
- Diffusion, Permeation
- Mechanics/Chemistry of
Stabilization/Solidificat ion
o Develop a Leachate Test/Model Interface
- Tests representative of important phenomena
- Test results usable in diffusion, flow models
o Validate Model Results through Field Studies
- Realistic Test Cells
- Actual Disposal Situations
BASIC BELIEFS
o All Materials Leach
o Our Goal Should be an Acceptable Rate of Release To
The Environment
o Leachate Tests Provide a Useful Measure of
Environmental Availability
- Excellent predictor of water borne contamination
- Of Secondary value for volatile emissions
o Leachate Tests Should Fairly Represent the Actual
Mechanics in the Field
o Properly Constructed Leachate Experiments Coupled
With Technically Sound Analysis of Flow Phenomena Can:
- Provide environmentally acceptable disposal
- Define risks posed by contaminated Bedia
44
-------
C-7 - CRITERIA AID OBJECTIVES Of LEACH IMG TESTS AMD MODELINC CONSIDERATIONS FOR PARTITIONING TESTS FROM THE
PERSPECTIVE Of REGULATORS VERSUS INDUSTRY (Continued)
EFFECTS Of SOC LEACHATE TEST PARAICTERS
o Waste/Liquid Ratio
- Most allow equilibrium/saturation
- Worst case for total dissolved Materials
- Saturation wight suppress relative solubility of
some constituents
o Leaching Tine
- Host approach equilibrium
- No residence tine/advection relationship unless a
column test
o Ntirber of Extractions
- Single extractions tell little of leaching phenomena
- Multiple extractions can provide more information
o Surface Effects
o Diffusional Fluxes
o Compositional Changes
o Particle Size
- Size reduction provides rapid equilibrium, most
conservative, but unrealistic results
o Ignores permeation, diffusion rate
o Defeats solidification/stabilization nechanisns
o Leaching Medium (Eluant) Composition
- Organic acids represent specialized case
o To simulate municipal co-disposal
o To provide buffered system, stable Ph
o Acetates, citrates preferentially extract metals
o May confuse analysis
Inorganic acids to represent "acid rain"
o More realistic for mono-wastes
- Total acidity
o Often high to counteract high alkalinity wastes
o Overestimates rate and amount of neutralization
o Agitation
- Speeds Equilibrium
- Mot representative of actual conditions
45
-------
APPENDIX D - CREDITS AND ACKNOWLEDGEMENTS
The reachability Subcommittee (LS) of the Science Advisory
Board's (SAB) Environmental Engineering Committee (EEC) wishes to
acknowledge the many people and offices for their numerous
contributions to this self-initiated study to arrive at
recommendations and rationale for analysis of contaminant
release. While listing names provokes the opportunity to miss
contributors by omission, we believe that the following
contributors to the efforts of the LS and its parent EEC deserve
a special "thank-you" and acknowledgement for their time,
energies and efforts toward improving this product and the
perspective of the SAB members and consultants in this very
complex area.
With respect to an information gathering session conducted
on February 26, 1990 in which the LS attempted to assess the
varied needs of the Agency, the following persons are recognized
for their participation in this exercise, as well as
participation in the follow-up activities after this session:
Mr. Harry Allen, EPA, Office of Emergency and Remedial
Response (OERR), Emergency Response Division (ERD), Edison, New
Jersey
Ms. Robin Anderson, EPA, OERR, Hazardous Site Control
Division (HSCD), Washington, D.C.
Ms. Joan Blake, EPA, Office of Toxic Substances (OTS),
Exposure Evaluation Division (BED), Washington, D.C.
Mr. David Friedman, EPA, Office of Research and Development
(ORD), Modeling, Monitoring Systems and Quality Assurance Office,
Washington, D.C. (Formerly with the Office of Solid Waste (OSW),
Characterization and Assessment Division (CAD))
Ms. Gail Hansen, EPA, OSW, CAD, Washington, D.C.
Mr. Alexander C. McBride, Chief, Technical Assessment
Branch, EPA, OSW, CAD, Washnigton, D.C.
Ms. Lynnann Kitchens, EPA, ORD, Office of Environmental
Engineering and Technology Demonstration (OEETD), Washington,
D.C. (Now with ORD's Waste Minimization, Destruction and Disposal
Research Division (WMDDRD) within the Risk Reduction Engineering
Laboratory (RREL), Cincinnati, Ohio)
Mr. M. R. Scalf, EPA, ORD, R.S. Kerr Environmental Research
Laboratory, Ada, Oklahoma
Mr. Carlton Wiles, Chief, Stabilization Section of the
Municipal Solid Waste Residuals Branch,WMDDRD, RREL, ORD, EPA
Cincinnati, Ohio.
46
-------
Mr. Christopher Zarba, EPA, Criteria and Standards Division
(CSD), Office of Water Regulations and Standards (Now known as
the Health and Ecological Criteria Division, Office of Science
and Technology ), Washnigton, D.C.
Dr. Linda E. Greer, Natural Resources Defense Council,
Washington, D.C. (At the time she was with the Hazardous Waste
Treatment Council (HWTC), but represented herself at the meeting,
and not the HWTC)
Mr. Phillip A. Palmer, E. I. duPont DeNemours & Co.,
Engineering Department, Newark, Delaware
The above participants in the February 26, 1990 information
gathering session deserve special recognition for helping develop
Appendix B which summarizes the various uses and needs of
leaching tests. Dr. K. Jack Kooyoomjian of the SAB Staff and
Designated Federal Official to the EEC and its LS deserves
particular recognition for synthesizing the above information.
Additionally, Mr. Samuel Rondberg's assistance in setting up
Appendix B and assisting in the numerous and often tedious edits
is greatly appreciated.
With respect to the Leachability Workshop and Technical
Briefing which the LS held on May 9, 1990, the following persons
are recognized for their contributions as participants in
briefing the LS members and consultants:
Dr. Paul Bishop, Department of Civil and Environmental
Engineering, University of Cincinnati, Cincinnati, Ohio
Dr. Pierre Cote, ZENON Environmental, Inc., Ontario, Canada
Dr. Marvin Dudas, Department of Soil Science, The University
of Alberta, Edmonton, Alberta, Canada
Dr. Carl Enfield, R.S. Kerr Environmental Research
Laboratory, Ada, Oklahoma
Dr. Robert Huddleston, Conoco, Inc., Ponca City, Oklahoma
Dr. P. Suresh Chandra Rao, Institute of Food and
Agricultural Sciences, Soil Science Department, University of
Florida, Gainesville, Florida
Dr. John Zachara, Battelle Pacific Northwest Laboratories,
Geochemistry Section, Richland, Washington
With respect to Figure 1 in the text which depicts a
conceptual view of leaching in a waste unit, the following person
deserves recognition:
Dr. Ishwar Murarka, Program manager for Land and Water
Studies, EPRI, Palo Alto, California
47
-------
With respect to the development of information on Extraction
Tests (Table 1), the following persons deserve particular
recognition:
Dr. Larry I. Bone, Environmental Quality, Dow Chemical,
Midland, Michigan
Ms. Gail Hansen, EPA, OSW, Characterization and Assessment
Division, Washington, D.C.
With respect to the development of Table 3 which lists the
scientific considerations in design and interpretation of
leachability tests, the following persons deserve particular
recognition:
Mr. Richard Conway, Senior Corporate Fellow, Union Carbide
Corporation, South Charleston, West Virginia
Mr. Peter Hannak, Union Carbide Chemicals and Plastics
Company, Inc., South Charleston, West Virginia (Formerly of
Alberta Environmental Centre)
And last, but not least, we would like to offer a special
thanks to Mrs. Marcy Jolly, Secretary to the Leachability
Subcommittee, for typing and retyping the numerous drafts of this
report.
Of course, this report would never have materialized without
the perseverance of the LS and the EEC after the need for a self-
initiated report on leachability was recognized. Special
recognition is made to Dr. Ward, Chairman of the LS and to Dr.
Murarka, Vice-Chair of the LS for guiding the report to a
successful conclusion.
48
-------
APPENDIX E - GLOSSARY OF TERMS AND ACRONYMS
AET
ANS
ANSI —
ARARSs
ASTM —
AVS
BOAT —
C
CERCLA
CWA
DI
DLT
EDTA —
EEC
EP
EPA
EPACML
FOWL™
HELP —
IAEA —
ISO
MCC
MEP
MEQ/G -
MIN
ML
MM
MWEP —
OMMQA -
N
N/A
NCP
NRC
OMMQA -
ORD
OSW
OTS
OSW
OTS —
OWEP —
PCB's -
pH
PMN
R&D
RCRA —
RSKERL
APPARENT EFFECTS THRESHOLD
AMERICAN NUCLEAR SOCIETY
AMERICAN NATIONAL STANDARDS INSTITUTE
APPLICABLE OR RELEVANT AND APPROPRIATE
AMERICAN SOCIETY OF TESTING MATERIALS
ACID VOLATILE SULFIDE
BEST DEMONSTRATED ACHIEVABLE TECHNOLOGY
CENTIGRADE
COMPREHENSIVE ENVIRONMENTAL RESPONSE, COMPENSATION
AND LIABILITY ACT (ALSO KNOWN AS "SUPERFUND")
CLEAN WATER ACT
DEIONIZED
DYNAMIC LEACH TEST
ETHYLENE DIAMINE TETRAACETIC ACID
ENVIRONMENTAL ENGINEERING COMMITTEE (SAB/EPA)
EXTRACTION PROCEDURE TOXICITY
U.S. ENVIRONMENTAL PROTECTION AGENCY (US EPA, or
"THE AGENCY")
EPA COMPOSITE MODEL FOR LANDFILLS
FOSSIL FUEL COMBUSTION WASTE LEACHING
HYDROLOGIC EVALUATION OF LANDFILL PERFORMANCE
INTERNATIONAL ATOMIC ENERGY AGENCY
INTERNATIONAL STANDARDS ORGANIZATION
DISTRIBUTION COEFFICIENT
LEACHABILITY SUBCOMMITTEE (EEC/SAB/EPA)
MOLE (MOLARITY)
MATERIAL CHARACTERISTIC CENTER
MULTIPLE EXTRACTION PROCEDURE
MILLI EQUIVALENT PER GRAM
MINUTE
MILLILITER
MILLIMETER
MONOFILL WASTE EXTRACTION PROCEDURE
ORD/OFFICE OF MODELING, MONITORING, AND QUALITY
ASSURANCE, US EPA
NORMAL (NORMALITY)
NOT APPLICABLE
NATIONAL CONTINGENCY PLAN
NUCLEAR REGULATORY COMMISSION
OFFICE OF MODELING MONITORING AND QUALITY ASSURANCE,
ORD/EPA
OFFICE OF RESEARCH AND DEVELOPMENT, US EPA
OFFICE OF SOLID WASTE, US EPA
OFFICE OF TOXIC SUBSTANCES, US EPA
OFFICE OF SOLID WASTE, US EPA
OFFICE OF TOXIC SUBSTANCES, US EPA
OILY WASTE EXTRACTION PROCEDURE
POLYCHLORINATED BIPHENYLS
NEGATIVE LOG OF HYDROGEN ION CONCENTRATION
PRE-MANUFACTURING NOTIFICATION
RESEARCH AND DEVELOPMENT
RESOURCE CONSERVATION AND RECOVERY ACT
R.S. KERR ENVIRONMENTAL RESEARCH LABORATORY, US EPA
49
-------
APPENDIX E- GLOSSARY OF TERMS AND ACRONYMS - Continuation
SAB SCIENCE ADVISORY BOARD (EPA)
SC OFFICE OF WATER, SEDIMENT CRITERIA PROGRAM, US EPA
SF SUPERFUND PROGRAM, US EPA
SPE SOLID PHASE EXTRACTION
SPLP SYNTHETIC PRECIPITATION LEACHING PROCEDURE
S/S SOLIDIFICATION/STABILIZATION
SYN SYNTHETIC (IN REFERENCE TO SYNTHETIC LANDFILL LEACHATE)
TC TOXICITY CHARACTERISTIC
TCLP TOXICITY CHARACTERISTIC LEACHING PROCEDURE
THF TETRAHYDROFURAN
TIE TOXICITY IDENTIFICATION EVALUATION
TSCA TOXIC SUBSTANCES CONTROL ACT
MM MICRO MOLES
UNIFAC — UNIVERSAL FUNCTIONAL ACTIVITY COEFFICIENT MODEL
WET WASTE EXTRACTION T.EST
50
------- |