RCRA RISK/COST POLICY MODEL PROJECT
PHASE II REPORT
Submitted to:
Office of Solid Waste
U.S. Environmental Protection Agency
June 1982
ICF INCORPORATED International Square
1850 K Street, Northwest, Washington, D. C. 20006
-------
RCRA RISK/COST POLICY MODEL PROJECT
Phase 2 Report
Submitted to:
Office of Solid Waste
U.S. Environmental Protection Agency
June 15, 1982
prepared by:
ICF Incorporated
Clement Associates, Inc.
SCS Engineers, Inc.
ICF Incorporated
-------
CONTENTS
Page
Abstract
Preface
1. OVERVIEW 1-1
Develop a Data Base 1-3
Specify Technologies 1-3
Specify Costs 1-3
Specify Releases to the Environment 1-4
Score for Inherent Hazard 1-5
Adjust for Environmental Medium 1-7
Calibrate Inherent Hazard Scores 1-7
Adjust for Environmental Settings 1-8
Generate Alternative Strategies 1-9
Set Parameters 1-9
Choose Goals, Array Alternatives, and Perform Sensitivity
Analysis 1-9
Select Strategy 1-10
Assumptions Limit Use of the Model 1-10
Improvements Are Planned 1-11
2. TECHNOLOGIES 2-1
Technologies Are Matched to RCRA Waste Streams 2-3
Model Allows Only Technically Feasible Combinations 2-3
Costs are Direct Capital and Operating Expenses 2-3
Model Uses Infinite Time Horizon and 3 Percent Real
Discount Rate 2-8
Cost Estimates Are for Typical Facilities 2-9
Onsite, Local, or Long-distance Transportation Is
by Truck 2-10
Releases to the Environment Are Either Routine or Accidental 2-11
Release Rates Partly Depend upon Waste Characteristics 2-12
Waste Characteristics/Technology Assumptions Determine
Pollutant Migration into Different Media 2-12
Treatment Technologies Alter Waste Streams 2-12
Phase Separation Divides Solids and Liquids 2-16
Component Separation Isolates Constituents 2-16
Chemical Transformation Changes or Destroys Constituents 2-20
Chemical Fixation/Stabilization Solidifies Waste 2-21
Incineration Destroys Constituents 2-21
ICF Incorporated
-------
Page
Transportation Allows Wastes to Use Landfill and Incineration
Capacity in Other Areas 2-21
Disposal Technologies Provide for Long-Term Management of
Wastes 2-23
Landfills Bury Wastes 2-23
Surface Impoundments Hold Liquid Wastes 2-23
Land Treatment Degrades Wastes 2-26
Deep Well Injection Pumps Wastes Underground 2-26
Ocean Disposal Disperses Wastes 2-26
3. WASTE STREAMS 3-1
Risk Scoring System Assesses Human Health Risks Based on
Tenfold Differences in Inherent Hazard and Exposure 3-1
Scoring System Is Based on Theoretical Model 3-3
Inherent Hazard Is Expressed as a Single Score 3-6
Data Are Primarily from Secondary Sources 3-9
Complex Chemistry and Missing Data Are Problems 3-11
140 Compounds Are Scored 3-11
Exposure Adjustments Are Made for Three Media 3-12
Air 3-22
Surface Water 3-22
Ground Water 3-24
Data Are Primarily from Secondary Sources 3-24
Complex Chemistry and Missing Data Are Problems 3-25
Data Base Includes Estimates of Uncertainty 3-25
Case Studies Calibrate the Scoring System 3-26
Four Case Studies Are Used 3-35
Model Uses 83 Waste Streams 3-35
Waste Volumes Are Based on Published Data 3-36
Wastes Are Grouped by Characteristics 3-37
Data Are Poor 3-38
Data Base Includes Waste Streams Representing
32 Metric Tons 3-31
ICF Incorporated
-------
Page
4. ENVIRONMENTS 4-1
Environments Change Risk 4-1
Thirteen Environmental Categories Are Based on Three Media
Plus Oceans 4-1
Air Medium Is Distinguished in Terms of Three
Levels of the Potential Population Exposed 4-1
Model Uses Two Levels of Assimilative Capacity
for Surface Water (Hydrology) 4-2
Two Levels of Ground-water Contamination Potential
(Hydrogeology) Are Used 4-3
Risk Scores Adjusted for Environmental Characteristics 4-3
Modifications for Population Exposure 4-3
Modifications for Hydrography 4-7
Modifications for Thickness of Dispersion Channel 4-7
Adjustments Reflect Tenfold Differences 4-7
Site-Specific Problems Must Be Examined 4-7
Volume of Waste and Capacity of Facilities Must Be Considered 4-8
5. ALTERNATIVE STRATEGIES 5-1
Set Parameters 5-2
Choose Goals 5-3
Array Alternatives 5-4
Results of Example Strategy 5-4
Perform Sensitivity Analysis 5-6
Select Strategy 5-6
REFERENCES
APPENDICES
1. Treatment Technologies
2. Transportation Technologies
3. Disposal Technologies
4. Release Rates for Treatment Technologies
5. Distribution of Releases to Air, Surface Water, and Ground-water
6. Waste Stream Descriptions
7. Model for Environmental Exposure
8. Dose Conversion Calculations
9. Calibration of Risk Scoring System Using Case Studies
10. Inherent Hazard Scoring Data Base Summary (complete data bound separately)
11. Exposure Adjustment Data Base (bound separately)
ICF Incorporated
-------
ABSTRACT
The RCRA Risk/Cost Policy Model establishes a system that allows users to
investigate how trade-offs of costs and risks can be made among wastes,
environments, and technologies (W-E-Ts) in order to arrive at feasible regulatory
alternatives.
There are many components in the system. Eighty-three hazardous waste streams
are ranked on the basis of the inherent hazard of the constituents they typically
contain. The system assesses these waste streams in terms of the likelihood and
severity of human exposure to their hazardous constituents and models their
behavior in three media -- air, surface water, and ground water. The system also
incorporates the mechanisms by which the constituents are affected by the
environment, such as hydrolysis, biodegradation, and adsorption.
A second integral part of the system is the definition of environments in which
the hazardous components are released. Thirteen environments including a special
category for deep ocean waters are defined on the basis of population density,
hydrology, and hydrogeology. The system adjusts the exposure scores of the waste
streams' hazardous constituents to account for their varying effects in the three
media in each of the environments.
The third component of the system consists of the technologies commonly used to
transport, treat, and dispose of the hazardous waste streams. This includes 3
types of transportation, 21 treatment technologies, and 9 disposal technologies.
The system determines costs and release rates for each of these technologies based
on the model's existing data base. It also incorporates estimates of capacities of
the technologies, the amount of waste to be disposed of, and the proximity of the
wastes to the available hazardous waste management facilities.
EPA's purpose in developing the RCRA Risk/Cost Policy Model is to assist
policymakers in identifying cost-effective options that minimize risks to health
and the environment. The framework of the system is intended as a screen -- to
identify situations that are of special concern because of the risks they pose and
to determine where additional controls may not be warranted in light of the high
costs involved. The framework uses a data base that is too imprecise and general to
be the sole basis for regulations. The results of the model will be used in more
detailed Regulatory Impact Analysis to determine whether some type of regulatory
action is warranted.
ICF Incorporated
-------
PREFACE
Early in 1981, the staff of EPA's Office of Solid Waste (OSW), at the request of
OSW senior management, began work on the RCRA Risk/Cost Policy Model (referred to
then as the Regulatory Policy Project). The purpose of the project was to develop
an analytic tool that would perform three functions. First, it should help EPA
policymakers understand the complex interactions of hazardous waste and ways of
managing that waste. Second, it should assist them in determining how stringent a
program is needed to protect human health and the environment from hazardous waste.
Third, it should serve as a screen to identify situations where regulatory action
may be most promising. The goal was to be accomplished by gathering information on
the risks and costs of managing different wastes using different technologies in
different locations or environmental settings.1 Policymakers could then use this
information in developing regulatory priorities. The project was to provide a
continuous and efficient way to assist management in evaluating alternatives and
testing hypotheses.
One immediate use of the results will be to single out waste management
situations that seem either especially risky or comparatively safe. After detailed
further investigation of their risks and costs, policymakers may decide to ban
especially risky situations or require stricter controls than now exist. In the
case of comparatively safe situations, controls may be loosened if the information
provided by further analysis confirms that these situations are, indeed, relatively
harmless.
Another immediate use of the results will be to help guide the selection of
alternatives in the Regulatory Impact Analyses (RIAs) that are required under
Executive Order 12291. These RIAs now cover all forms of land disposal,
incineration, land treatment, storage and treatment, surface impoundments,
industrial boilers, the disposal of waste oil, and the location of facilities in
seismic areas and floodplains. This project will help the RIAs by analyzing the
interaction of alternatives under each RIA and will project their combined effect
by looking comprehensively at the entire waste management situation.
Three consulting firms -- ICF Incorporated; SCS Engineers, Incorporated; and
Clement Associates, Incorporated -- started work on the project in August 1981. In
October, EPA made available a Phase I Progress Report on general features of the
methodology. The present report details the contractors' efforts to develop such a
tool since then. First, and most important, it describes a model for assessing and
comparing the risks and costs of a wide variety of waste management situations.
Second, it contains the contractors' first attempt to construct a data base as
comprehensive as possible within their constraints of time and resources and to
make their assessments of risk and cost on the basis of these data, which are often
scanty. Third, it outlines what we intend to do next and discusses options for
further improving or extending the model's analytic capability.
1Another original goal of the project was to assess the feasibility and
administrative cost of using alternative regulatory mechanisms, such as
performance standards and design standards. This task has, however, been
incorporated into other projects.
ICF Incorporated
-------
Planned future work includes determining which areas of the model and data base
most need improvement, and testing, how sensitive the model is to changed
assumptions, data, and hypotheses. We will, of course, explore further those areas
about which there is great uncertainty and that appear to affect significantly the
results of the model. The model will then be subject to continual revisions as
needed.
We recognize that it is impossible to anticipate all potential concerns about a
model such as the one we have developed, but we will attempt to cover some. First,
the project is designed to assist policymakers at a very general level. It is not
sufficiently detailed or scientifically precise to support regulations. The tool
will, however, enable policymakers to clarify issues and to test the effect of
different value or policy judgments. At the very least, the model will attempt to
identify those areas where different assumptions or judgments are most critical.
Second, the tool is intended to extend policymakers' ability to consider options,
not to eliminate their judgments. Third, the project is not designed to exclude
those outside EPA from participating in a public debate over policies. To the
contrary, it is intended to improve the quality of the debate by making judgments
and possible outcomes more explicit. Finally, our purpose is not to devise a
strict "degree of hazard" system that would divide all hazardous wastes into a
single set of categories differentiated only on the basis of inherent hazard. Our
system recognizes, instead, that how and where different wastes are managed
determine how risky or safe a situation is (its degree of hazard) and attempts to
differentiate among these situations. The groupings or categories of wastes used
in the model are not assumed to be adequate for subdividing wastes that EPA lists as
hazardous. Special, tailored regulations based on distinctions among waste streams
and management situations must rely on more precise scientific and management
information than this project uses.
The development of the model is a shared achievement, possible only through the
interaction of many individuals. Those who made particularly significant
contributions include Lawrence G. Buc (EPA), Joseph Kirk (project officer from
ICF), Ian Nisbet (Clement Associates), Michael McLaughlin (SCS Engineers), and
Steven Shelton (associated with SCS Engineers). The interest, comments, and help
of many others including numerous members of OSW, the Office of Research and
Development, and other offices within EPA; environmental organizations; numerous
trade associations; the Office of Technology Assessment; and, of course, the
contractors' staffs have also enhanced the value of the work. The quality of this
document was improved immeasurably by Arline Sheehan (EPA) and by the editorial
review of Pat Fox.
Curtis Haymore
EPA Project Officer
ICF Incorporated
-------
CHAPTER 1
OVERVIEW OF PROJECT
This project establishes an analytic framework that can identify many of the
possible trade-offs among the risks and costs of various regulatory approaches more
comprehensively and more precisely than in the past. For any waste management
situation, the degree of risk and cost depends on the interaction among wastes,
environmental settings, and technologies for treating, transporting, and disposing
of wastes. This project presents a framework that, for very general waste-
environment-technology (W-E-T) combinations, summarizes the best available
scientific and engineering information about risks and costs. It also includes a
measure of the uncertainty associated with the estimates.
The project is a tool to assist in evaluating alternative policies. It is not a
scientific inquiry; rather it combines data from scientific and economic studies to
gain a general and as nearly comprehensive view as possible of hazardous waste
management.
The framework is helpful in developing potential answers to many difficult
policy-related questions. Examples of such questions are:
• Which regulatory strategies offer superior risk
reduction per unit of cost? How should priorities be
set among wastes and environments? Where are the
serious problems? What are the safest and cheapest
technologies?
• What is the appropriate balance to strike between
how much society should pay (in prevention and
abatement costs) to achieve environmental safety of
varying levels?
• Where should additional research effort be focused?
Where is there a lack of data on potentially serious
problems or potentially useful technologies?
• What are the interactions among technologies?
Which are substitutes for one another, and which are
not? How do cost differentials affect their use?
The W-E-T framework is merely one way of arraying and comparing information
about different possibilities. The manner in which we develop and compare the risk
and cost information for different situations is schematically presented in Exhibit
1-1, which is an overview of the entire approach. Detailed discussions of all the
components are included in the report. Chapter 2 discusses the technologies we
defined, Chapter 3 covers waste streams and risk scoring, Chapter 4 explains the
use of environmental settings, and Chapter 5 contains an example strategy which
applies the data. This overview chapter describes the main features and
limitations of the model and summarizes what improvements we hope to make. The
steps necessary to develop a data base of risk and cost values are shown in the
large box at the top of Exhibit 1-1. The large box at the bottom of the exhibit
shows the steps required to generate alternative strategies for discussion and
further analysis.
ICF Incorporated
-------
1-2
EXHIBIT 1-1
RCRA REGULATORY POLICY TOOL: SCHEMATIC OF CONCEPTUAL FLOW
Develop a Data-Base
TECHNOLOGY
Specify
Technologies
I
Specify
Costs
Specify
Releases to
Environment
ENVIRONMENT
Adjust for
Environmental Setting
WASTE
Score for
Inherent
Hazard
b
Adjust for
Environment
I
Calibrate
Generate Alternative Strategies
Set
Parameters
~
waste
volumes
~
capacities
o
treatment
steps
~
transporta-
tion
Choose Goals
(low risk,
low cost)
Array
Alternatives
I
Perform
Sensitivity
Analysis
Select
Strategy
ICF Incorporated
-------
1-3
DEVELOP A DATA BASE
The following sections describe each of the steps identified in Exhibit 1-1
that are necessary to develop a data base.
Specify Technologies
The model applies three classes of technologies to manage hazardous waste:
treatment, disposal, and transportation. Treatment technologies alter in some way
the characteristics (volumes, concentrations, and so forth) of the waste streams.
The present model uses 21 generalized treatment technologies, each representing
perhaps several closely related technologies. The way they alter the
characteristics of the waste stream is meant to be typical of these processes and
to represent a facility similar in size and effectiveness to many now in use. The
technologies are not necessarily "average" or median in any statistical sense. No
exotic technologies are now in the model, although they could be added fairly
easily. Disposal technologies, of which the model now has nine, are similarly meant
to represent typical facilities, but do not result in redefined waste streams. The
three transportation technologies--onsite, local, and long distance—do not affect
the composition of the waste streams, but do allow waste streams to be treated or
disposed of in different locales.
Both treatment and disposal technologies in the model have certain wastes they
cannot handle. Filter presses, for example, cannot accept very dilute wastes,
wastes that are already solid, or wastes in which the primary constituent is
dissolved rather than suspended. The incineration modules, similarly, will not
accept waste streams that are primarily metals rather than organics.
The data base includes every combination possible of feasible treatment and
disposal options. The only constraints are that every waste stream have one
disposal step, one transportation step, and no more than three treatment steps.
Specify Costs
In this analysis, costs include only real resource costs, that is, the value of
goods and services lost by society as a result of using resources to comply with a
regulation. For the purposes of this project, this translates into direct capital
costs and operating costs associated with various waste management technologies.
Capital and operating costs each have direct and indirect elements of cost
associated with them. Direct capital costs include expenses to acquire and install
assets (land, facilities, equipment). Among direct operating costs are materials,
labor, utilities, maintenance of equipment, and transportation. We do not include
indirect costs (such as engineering and design costs, licenses and permits, and
administrative overhead) because these items usually represent relatively fixed
percentages of direct costs. Adding a flat percentage to all direct costs would
not change the relative ranking of costs; thus, we chose not to estimate indirect
costs separately.
We specify costs in dollars per unit of weight of the nonwater portion of the
waste stream. Costs are consistent with risks in a temporal sense since they
compare total risk reduction with discounted (3 percent) life cycle costs of the
technologies.
ICF Incorporated
-------
1-4
Estimating costs of waste management technologies poses a number of difficult
problems, but there are more data, and these data on cost are of substantially
higher quality than those on risks. It is possible to design a treatment-
transportation-disposal chain and develop specific costs if we have some estimates
about the type and volumes of wastes being handled. Unlike the situation in
estimating risks, there is considerable practical experience in the engineering
community that provides a solid foundation on which to base estimates of cost.
It is also an important element of this model to predict how waste will shift
between technologies as prices of treatment and disposal vary. To do this requires
relatively precise estimates. The range over which costs vary is also fairly
limited. Thus, we selected a cost scale with divisions that distinguish between
costs that are at least two times as great as other costs. Cost scores that
represent doubling differences are still very broad, but we believe will enable the
model to capture the most important shifts. We established ranges around the
typical scores to recognize special situations and thus provide the basis for
sensitivity testing.
We expect that the typical costs in the model are overestimates of the actual
direct capital and operating costs. This is because the model does not account for
the redesign of the production processes and other waste-handling methods of waste-
generating plants to minimize total costs. It will often be less expensive to
redesign a process, segregate waste, or take other measures than to use many
hazardous waste treatment and disposal steps.
Transportation costs are based on costs for a 6,000-gallon tank truck
transporting waste locally (a one-way trip of 25 miles) or long-distance (a one-way
trip of 250 miles). Onsite transportation costs are included in the treatment and
disposal technology costs.
Specify Releases to the Environment
There are three media into which waste can be released to the environment --
air, surface water, and ground water. There are two classes of releases: routine
or accidental. The former include leachate generation and migration from landfills
and surface impoundments and air emissions from open tanks, land treatment
facilities, and incinerators. Accidental releases result from random spillage,
which we estimate on the basis of the probability of the spill and of the quantity
spilled. Both types of releases often depend on the characteristics of the waste
as well as the technology used to manage it. By combining these two classes of
releases, the model can compute the total release to each medium from each
technology. The total release to the environment is then the sum, by medium, of the
releases from each treatment, transportation, and disposal step.
There are few quantitative data on actual release rates from different types of
facilities. This information is, however, critical to any attempt to compare the
risks of different waste management situations. Thus, most of these estimates are
based on the best judgment of the engineering contractors, supported by their
experience with different types of facilities. Individual sites may, of course,
vary greatly in how much pollution they generate. The estimates are, again,
subject to change as additional data become available.
ICF Incorporated
-------
1-5
In many cases, a waste must be transported from the point of generation to
final disposal. When that occurs, the model includes an additional environmental
release based on the probability of an accident and a release based on reported
spill rates due to transfer of wastes. The data base is derived from U.S.
Department of Transportation accident rates and assumptions regarding the expected
release as a result of each accident.
Score for Inherent Hazard
In the third column of Exhibit 1-1, labeled "Waste," the first step is to score
chemical compounds for inherent hazard. This step requires defining what hazards
will be considered. To be complete, the model should include hazards posed to
human health and the environment and effects on the economy. At this stage of
development of the model, we have only included adverse effects to human health.
We hope to develop methodologies for including other types of effects in the model
eventually, but we believe that risks to human health are almost always the
predominant element in the social cost of hazardous waste pollution. Ecological
damage is usually closely correlated with risks to human health: only rarely does
chemical pollution pose a severe threat to the environment without also posing a
threat to human health. Some economic risks, such as damages to resources
(fisheries, for example), are generally related to human health risks. Other
economic effects, such as the expense of cleaning up or preventing contamination
(purifying or replacing water supplies, for example) serve as substitutes for other
ecological and human health risks. For these reasons, we believe that by
considering risks to human health alone, we are capturing the most significant
hazards from hazardous waste management.
Since chemical compounds affect human health in many ways, it is cumbersome and
difficult to catalog those effects separately throughout a general model. The
alternative is to find some basis for comparing effects to allow a single or
selected few measures to be used. We chose this latter approach. We recognize that
value judgments about the relative importance of various health effects are
necessary to formulating a regulatory approach and that it would be helpful for
policymakers to develop a consensus about relative importance. In the model, we
give all effects equal weight, but we intend to test the sensitivity of this
assumption to alternative weighting schemes. As a first approximation, the model
attempts to relate all health effects at a given probability of occurrence per unit
dose. We define the overall risk presented by a hazardous waste facility as the
expected (or most likely) number of people affected within a specified period of
time.1
The model uses two scales to estimate the two broad classes of effects on
health -- dichotomous and graded. Dichotomous (all-or-none) responses include
those caused by carcinogens, teratogens, and mutagens. Graded responses include
illnesses, such as those that cause liver and other organ damage, that range from
1There is concern that, although this approach is certainly appropriate for
policymaking, it may give insubstantial weight to situations where relatively few
people are at high risk. Another approach is to consider only the risk to
individuals and ignore how many persons are exposed. A third, and our preferred,
approach is to use both criteria since both are clearly important to a fair and
efficient program. The model allows us to test many variations to determine how
sensitive it is to varying assumptions about maximum individual risk and the extent
of the population exposed. This is one of the critical areas for further analysis.
ICF Incorporated
-------
1-6
slight to substantial impairment depending on the dose. The model computes a risk
score for inherent hazard of substances in each of the two classes. For
dichotomous responses, the data are usually expressed as or can be converted to a
probability of response per unit dose of a compound. For graded responses, the
model uses the minimum effective dose (MED) required to produce a measurable
adverse effect as the basis to assign scores. This scale is matched to that of unit
risks (for dichotomous responses) by assuming that a dose of MED/10 in human
studies corresponds roughly to a probability that 1 percent of the exposed
population will suffer substantial adverse effects. Both scales assume that
effects correspond to dose linearly over the range we consider them. The scales
also generally assume no threshold level of effects.
The model uses order-of-magnitude estimates for risks. This level of
generality is appropriate at present for a number of reasons: first, the quality of
the data is seldom more precise that this, and second, there is insufficient
information for most chemicals regarding inherent hazard and environmental
behavior (see next section) to permit precise estimates of risk. These two factors
demonstrate that scoring for risks to human health is subject to large
uncertainties and in many cases depends on scientific judgment. Because the
results of the model will be used at a high level of generality (to set priorities
rather than to design specific regulations), more specific estimating procedures,
which would be much more time and resource intensive, should be reserved for the
few high-priority situations that the model will indicate. The order-of-magnitude
scale we developed for scoring risk, therefore, distinguishes between risks that
are at least 10 times as great as other risks. Order-of-magnitude differences in
risk lead naturally to broadly defined waste categories, environments, and
technologies rather than to specifically designed technological solutions to
particular disposal problems.
The October Phase I report presented a proposed methodology for scoring
inherent hazard by categorizing RCRA waste streams into about 15 groups based
primarily on toxicity to humans and persistence in the environment. Because of
many comments concerning the appropriateness of that categorization scheme and our
own difficulty in attempting to apply it, we have abandoned it. The replacement is
not wholly satisfactory, but better incorporates available waste stream analyses.
The model uses, at least temporarily, every waste stream for which data on
constituents, proportion, and volume are available to the contractors (83 at this
writing).
Each waste stream in the model has five defining characteristics generally
believed to be typical of actual waste streams and five more parameters for each
constituent of concern.2 Each waste stream is associated with the total volume
produced in each environmental setting by the industry that generates the waste.
Each waste stream, thus, becomes a category and is intended to be representative of
the waste streams in that industry. Where industries produce more than one waste
stream for which there are data, we allocated volumes based on the best information
available to us.
2Throughout this document, the constituent of concern is referred to as "X."
The presence of the constituent of concern is generally the reason that the waste
stream is considered hazardous under RCRA. The model permits the presence of
several constituents of concern in a waste stream.
ICF Incorporated
-------
1-7
A better way is needed to extrapolate both approximate hazard levels and
volumes to those waste streams for which detailed data are not available. Our
proposal in October presented a generalized methodology, but we and others found
too much inconsistency within at least some of the groups of waste streams to
continue to use them without substantially more specification. What is more
important, the October approach did not provide a way to estimate and allocate the
volumes of wastes in each category. The present approach provides for the
inclusion of all available data and a method for allocating volumes on the basis of
industries. Of course, gaps in the data may result in some biases, but the effects
of bias can be tested by sensitivity analysis.
The basic problems with any attempt to establish broad categories are: (1) the
lack of data, (2) the variability in levels of hazard between similar waste
streams, and (3) the difficulty in finding common ways to group constituents. The
most appropriate method for eventually grouping waste streams is, of course, based
on their properties and characteristics. Because the data on these properties and
characteristics are now limited, using properties of waste streams seems unlikely
to produce a general categorization scheme that is useful, at least in the sense of
prospectively analyzing its effects. Employing such a scheme as a regulatory
strategy, however, may still be worthwhile. In any case, much work is still needed
in this area, and we will continue to explore ways to overcome this problem.
The data base includes estimates of typical values plus a range around those
values, thus giving a measure of the uncertainty in the estimates. In the
sensitivity analysis stage, we will use this feature to identify areas where
additional, more detailed analysis would be helpful.
Adjust for Environmental Medium
The overall goal of scoring for the inherent hazard of compounds is to
determine the danger of their release to the environment. Compounds behave
differently depending on the environment, and the environments have varying
abilities to dilute, degrade, and transport the compounds. Thus, the eventual
level of hazard the compounds pose in a particular environment varies. To account
for these differences, the model adjusts each inherent hazard score so that there
are eventually three scores -- one for each medium (air, surface water, and ground
water).
The adjustment factors are primarily based on the half-life in each medium of
each of the wastes' constituents of concern. The model also uses several
attenuation factors -- hydrolysis, biodegradation, and volatilization -- and
considers bioaccumulation, adsorption, and wash out. On the basis of assumed rates
of movement and dispersion patterns for each medium, the model incorporates all
these influences into a single adjustment factor for each medium and compound. As
an example, a compound having a half-life of 3 minutes in air, 30 minutes in surface
water, and 6 hours in ground water (and having no special attenuation) would
receive the same score in each medium, because the model predicts these degradation
rates to pose essentially equivalent levels of exposure to humans.
Calibrate Inherent Hazard Scores
Although we were able, in principle, to extend our exposure models to
calculating exposures and risks to populations, we decided the models were too
generalized to provide a sufficient level of accuracy. We chose to calibrate the
ICF Incorporated
-------
1-8
scoring system by using case studies that relate population exposures or risks to
measure release rates. These case studies have known release rates', known or
measurable population exposures, and well-characterized dispersions. We sought,
but have been unsuccessful at finding, at least two case studies in each medium.
Those selected are the following:
• cadmium from a point source (into air);
• chloroform (into surface water);
• PCBs (into surface water);
• sodium chloride (rock salt) leaching (into ground water).
None of the case studies corresponds exactly to the simple dispersion characterized
by the exposure models, but they independently provide estimates of risk scores
that are relatively close to those predicted by the model (within one or two orders
of magnitude). We adopted the average value of 10 to correspond to a risk of one
person expected to be affected per year.
Adjust for Environmental Settings
The eventual goal of the scoring process is to compare the hazard of different
environmental releases of hazardous wastes. The calibrated scores of inherent
hazard by medium are converted to releases to the environment per year. For
example, a release of about 30 units of barium to air, according to the data now in
the model, poses a risk similar to a release of 63 units of nickel to surface water
or 10 units of zinc to ground water. The model computes these quantities for each
risk level. For any given release, therefore, its score is that corresponding most
closely to the quantity actually released.
The model calculates these scores on the basis of a standard environmental
setting: an area with a high potential for ground-water contamination (high risk),
a low assimilative capacity of surface water (high risk), and a medium density of
population potentially exposed (medium risk). For other situations -- low risk to
ground water and surface water and exposure to very many or very few persons -- the
model allows adjustments to be made. Based on the assumptions of the environmental
transport of hazardous wastes, the adjustments may alter a given risk score by up
to two orders of magnitude (although costs are unaffected). The combination of
high and low risks for ground-water and surface-water and three levels of
potentially exposed populations leads to 12 environmental combinations. (Two
ground water categories times two surface water categories times three population
categories equals 12 total categories.) A thirteenth category, ocean disposal, is
also used. These very broad categories are inadequate to examine a specific site.
They are useful, instead, at comparing the projected combined effects of many sites
and roughly comparing those sites to the combined effects of many sites in other
situations. Most important, this model cannot be used to evaluate particular
permit applications.
Our limited number of environmental settings poses other problems for specific
analysis. We do not, for example, attempt to introduce climatic considerations
until a strategy has been formulated. At that point, we review waste management
steps being applied to determine the extent to which one or more of them might be
influenced by particular geographic conditions. Freezing temperatures and the
requirement for enclosed systems would, of course, necessitate increasing the cost
of the typical system the model proposed.
ICF Incorporated
-------
1-9
GENERATE ALTERNATIVE STRATEGIES
The large box on the top of Exhibit 1-1 contains all the steps discussed so far
and represents, when complete, the data base on which the model operates. The data
are, in essence, the costs and risks of managing specified units of waste, not their
total effects. The large box at the bottom of the exhibit outlines the steps to
determine total effects by constructing and arraying alternative strategies. The
following sections describe each of these steps.
Set Parameters
Certain parameter-setting decisions must be made before we can assign total
risk and cost scores. They include the volume of the waste streams, the number of
treatment steps, the capacities of the various technologies, and the manner in
which transportation is included. It is possible, however, to change those
decisions to assess their effects. The volumes assigned to waste streams and the
capacities of different technologies are derived from EPA computer files of
information submitted by facilities on Part A of their permit applications and a
1980 EPA report, Hazardous Waste Generation and Commercial Hazardous Waste
Management Capacity -- An Assessment. We assume that capacities are limited only
for double-lined and single-lined landfills and 99.99 percent and 99.9 percent DRE
incinerators because other technologies can quickly respond to increased market
demand. We currently limit the number of treatment steps to no more than three.
While this encompasses most existing practices, the assumption could easily (but
perhaps not inexpensively) be relaxed.
Volumes of waste are assigned to three-digit zip-code geographic regions as are
capacities for landfills and incinerators described above. When capacities of
these technologies are filled in one region, the model compares the costs and/or
risks of technologies and may shift some of the waste to another region with
available capacity. Transportation costs and risks are then added as necessary.
Choose Goals, Array Alternatives, and Perform Sensitivity Analysis
The model generally consists of a data base, a set of limiting constraints
(described above), and a linear programming (LP) model. The LP model allows one to
set a goal (for example, to minimize cost) and to find the solution that best
fulfills the goal with information in the data base and within the given
constraints. We plan to compare the differences among four specific goals.
The first goal1 is to find the least costly way to achieve alternative levels
of safety (or risk). The result is a strategy that defines how all waste streams
are managed to achieve the specified goal. A separate computer run that minimizes
costs at each level of risk is conducted to produce a series of alternative
strategies that reflect different assumptions about acceptable levels of risk.
3In linear programming argot, goals are termed "objective functions." The
effect of changing data base values on the alternatives will be examined through
extensive sensitivity analysis. This is essential to understand how the model is
operating, which data need most to be improved, and how firm the results are. As a
starting point, we will vary the inherent hazard scores (adjusted by medium) by the
limits of their uncertainty estimates. Variations in volumes and concentrations in
waste streams will be tested similarly. This process will give many insights into
how well the entire model performs.
ICF Incorporated
-------
1-10
A second goal is to find the combinations of W-E-Ts that minimize risk for
given levels of cost. Again, a series of alternative strategies is generated.
Finding the lowest cost ways to achieve various levels of risk-cost ratios
constitutes the third goal. The fourth goal is to find the lowest risk ways of
reaching these different levels of ratios. We plan to compare these four goals to
determine whether significantly different strategies result. Should they differ
greatly, EPA may need to make some judgments about which criterion to use.
Regardless of the criterion, a series of alternative strategies is developed that
reflects different value judgments about how much public safety society can afford
to buy.
Select Strategy
On the basis of the ground-work laid previously, the ideal approach would be to
let policymakers find the strategy that most closely reflects their views. The
trade-offs will not be easy. Even if the data base were perfect (which it most
surely is not), the process of picking a strategy requires much deliberation and
reflection. With the given limitations of the data base, it also requires
restraint and some degree of interpretation.
An immediate use to which the results of the model will be put is as a screen to
help identify those especially risky or comparatively safe situations where
regulations might potentially be strengthened or loosened.
ASSUMPTIONS LIMIT USE OF THE MODEL
It would not have been possible to develop a policy analysis tool without
making a number of critical assumptions and simplifications. As a result of these
simplifications, for example, the risk scores are accurate to no more than an order
of magnitude and the cost scores to within a factor of two. This degree of
imprecision means that the results cannot be used in a specific regulation-making
context. It does not, we hope, hamper the tool's utility in setting directions for
policy and in establishing priorities among the many complex steps of the RCRA
regulatory program.
We could, of course, use our general methodological approach to reach specific
conclusions, but to do so would involve substantial effort and time, which should
probably be spent only on a very small number of regulatory options of the highest
priority. Even if we used the tool in such a limited fashion, we would have to make
substantial changes in the present assumptions.
Making assumptions and simplifications was necessary in order to produce a
policy analysis tool in a reasonably short period of time and with some economy of
resources. We believe that the theoretical underpinnings of our methodology are
sound, but strongly caution against the uninformed application of this tool.
Because of the level of generality at which we operated, it would be improper to
apply the risk and cost values to a specific situation. The tool is useful, we
believe, in its relative positioning of risks and costs.
In many instances, we applied scientific judgment to derive the risk scores,
particularly in using animal data to assess possible risks to human health and to
evaluate exposure to compounds. Likewise, we applied engineering judgment to
arrive at typical values for costs and release rates of the technologies we use in
managing wastes. Making specific judgments in individual cases requires much more
effort than we were able to expend.
ICF Incorporated
-------
1-11
Another major area of applied judgment is that of defining waste streams as
representative of the total wastes in the U.S. We used a limited number of waste
streams from the major industrial segments to approximate many others. This
introduces inaccuracies in volumes of wastes generated and the ability of existing
disposal technologies to handle the wastes.
Although a number of assumptions hamper specific analysis, we believe that the
tool is highly useful at the general level of application for which it is intended.
By combining many different waste streams and waste management chains, it allows
the decisionmaker to select the combinations of the highest priority from the
thousands of possibilities. On the basis of those selections, effective regulatory
decisions can be made.
Very few portions of this model would be expensive or difficult to change if
different assumptions are found to be more appropriate or if new data become
available. We hope that EPA can extend the model, improve the supporting data
base, and alter the built-in assumptions as a better understanding of environmental
processes is gained. The model is not designed to substitute for public
decisionmaking by making decisions seem automatic or easy. The model is, rather,
to be used, manipulated, tested, and verified as part of an exercise to understand
better the likely consequences of alternative regulatory decisions.
IMPROVEMENTS ARE PLANNED
This report summarizes the current state of the model. We expect to accomplish
or at least explore the following tasks over the next several months:
• Develop alternative strategies and compare results.
• Perform sensitivity analyses on:
different weightings of health effects;
effects of using different levels of populations at
risk;
effect of establishing maximum limits for exposure
within a W-E-T;
high and low estimates for risk and cost Scores.
• Compare the rankings of inherent hazards of 140
chemicals with other EPA and non-EPA risk assessment
projects to identify differences and resolve
discrepancies.
• Improve the quality of data on the characterization of
waste streams and on generation rates by verifying data
collected from secondary sources, by obtaining additional
information from primary sources (for example, state
files), and by incorporating data from EPA Annual Surveys
and RCRA questionnaires.
• Expand the risk methodology to take into account risks
other than to human health, primarily ecological effects.
ICF Incorporated
-------
1-12
• Expand the risk methodology to take into account
sudden health threats, such as fires and explosions.
• Improve the method for modeling the dispersion rate
and effects of different types of hazardous wastes to the
ocean environment.
• Incorporate storage facilities as a technology.
• Score additional chemical compounds that are present
in a few of the waste streams and are likely to increase
those waste streams' risk scores.
Following is a partial list of other possible improvements. It is unclear,
however, when (or whether) we will be able to undertake these improvements because
of resource constraints.
• Improve models for leachate generation and air
volatilization by taking into account additional site-
specific factors (for example, soil type and climate).
• Model the effects of synergism of comanaging
different types of hazardous wastes, especially for the
land disposal technologies (such as monofills versus
multiwaste landfills).
• Incorporate a method for modeling nonpoint releases
of waste to surface waters.
• Include additional technologies (for instance, wet
air oxidation, biological treatment, plasma arc
treatment, and high gradient magnetic separation) and
combine or eliminate those that are similar (such as the
vacuum filter and filter press).
• Devise a method for estimating the probability and
magnitude of release of hazardous waste to an aquifer
from an underground injection well from other than
casing failure.
• Include additional categories of environments based
on such factors as climate and soil type and revise the
methodology for computing release rates accordingly.
• Refine existing categories by obtaining firmer data
or by separating them into additional subcategories.
• Revise costing methodology to address separately
indirect costs and nonresource costs, especially for
third-party liability insurance.
• Incorporate estimates of future changes in the
composition and volume of waste streams.
We would, of course, appreciate suggestions concerning additions to this list and
opinions on which of these items are the most important.
ICF Incorporated
-------
CHAPTER 2
TECHNOLOGIES
The model now includes 21 treatment, 3 transportation, and 9 disposal
technologies (Exhibit 2-1). The model permits a series of up to three treatment
steps, a limitation that reduces the model's cost and complexity but which still
reflects the most common practices in hazardous waste management. A waste stream
need not, however, be treated. The model applies only one transportation and one
disposal technology at a time to a waste stream. A RCRA waste stream resulting from
metal plating might, for example, first be precipitated and the precipitant then
put through a filter press before being transported to a landfill.
We characterize each technology along three dimensions: (1) which wastes the
technology handles, (2) the cost of applying the technology, and (3) the amount of
waste released into the air, surface water, and ground water. We adopted these
dimensions for the following reasons:
1. Not all technologies are technically capable of handling
all wastes. Without restricting wastes to certain
technologies, the model could produce infeasible results.
For example, excluding filter presses from treating very
dilute waste streams and deep wells from injecting waste
streams with high suspended solids reflects the practical
limitations of these technologies.
2. Estimating the cost of applying a technology is a
necessary step to eventually analyze cost-risk trade-offs.
3. The amount of waste that is released into the environment
drives our estimates of the risk proposed by alternative
waste management practices.
Another key component of the model is its specification of the manner in which
a treatment technology changes the characteristics of a waste stream. The purpose
of treatment is to render the waste less hazardous or less costly in further
treatment and disposal. In contrast, the model uses the three transportation modes
to account for the added costs and releases of wastes that result from moving waste
to facilities in other areas either because of available treatment or disposal
capacity or lower risks.
The remainder of this chapter deals with each of the five points mentioned
above: the wastes that can be treated or disposed of using each technology, the
cost of the treatment or disposal, the environmental releases of hazardous
components, the alteration of waste streams through treatment, and the way
transportation is considered. Also included is a discussion of the disposal
technologies used. Descriptions of each technology and the detailed assumptions
and algorithms the model uses appear in Appendices 1-4.
ICF Incorporated
-------
2-2
EXHIBIT 2-1
TREATMENT, TRANSPORTATION, AND DISPOSAL TECHNOLOGIES
1 1
Treatment
i i
i
Transportation
i
i
I
i
Disposal
Phase Separation
1. Chemical coagulation
1. 0nsite
1.
Double-lined landfill
2. Filter press
2. Local
2.
Single-lined landfill
3. Centrifuge
3. Long-distance
3.
Unlined landfill
4. Vacuum filter
4.
Double-lined surface impoundment
Component Separation
5.
Single-lined surface impoundment
5. Evaporat ion/drying
6.
Unlined surface impoundment
6. Air stripping
7.
Land treatment
7. Steam stripping
8.
Deep well injection
8. Solvent extraction
9.
Ocean
9. Leaching
10. Distillation
11. Reverse osmosis
12. Carbon adsorption (PAC)
13. Ion exchange
Chemical Transformation
14. Chemical precipitation
15. Chemical destruction
16. Electrolytic decomposition
Chemical Fixation/Stabilization
17. Chemical fixation/stabilization
Incineration
18. Incineration -- 99.99% DRE*
19. Incineration -- 99.9% DRE
20. Incineration -- 99.0% DRE
21. Incineration — 90.0% DRE
*destruction and removal efficiency
ICF Incorporated
-------
2-3
TECHNOLOGIES ARE MATCHED TO RCRA WASTE STREAMS
Over 99 percent of the volume of waste streams now in the data base are waste
streams defined as hazardous under Subtitle C of the Resource Conservation and
Recovery Act. The exceptions are the addition of PCBs and asbestos wastes. Our
definition of hazardous waste streams (or RCRA waste streams) minimizes overlaps
between our waste streams and other forms of pollution such as air emissions
regulated under the Clean Air Act or wastewater discharges regulated under the
Clean Water Act.
For the purposes of this report, the RCRA waste stream includes waste materials
only after they have been separated from a process wastewater or air waste stream
(which may ultimately be regulated under an authority other than RCRA). A
wastewater treatment plant might, for example, consist of a wastewater screen, an
aeration basin, and a final clarifier, from which a point source discharge is made.
Although each of these steps could result in a residual hazardous waste stream, we
do not consider any of them RCRA waste stream treatment technologies. When,
however, the hazardous residuals are segregated from the wastewater flow and, for
example, subjected to volume-reduction treatment, we regard this waste stream as a
RCRA waste stream and include the volume-reduction technology in our analysis.
Model Allows Only Technically Feasible Combinations
The model prevents technologies from managing wastes that are technically
infeasible for them to handle. Exhibit 2-2 lists those wastes the model will allow
each technology to process. This list makes no judgment on whether particular
wastes should be managed by particular technologies. The appropriateness of
technologies and waste streams depends on the environmental setting and the
determination of how much risk and cost is acceptable, which is one aim of the
entire project. The data base, therefore, includes all technically feasible
combinations of technologies and waste streams, so that the linear programming
portion of the model will have a complete, yet feasible, data base with which to
work.
COSTS ARE DIRECT CAPITAL AND OPERATING EXPENSES
Exhibit 2-3 shows the unit costs of operating the technologies now in the
model. The model only accounts for real resource costs, that is, the value of goods
and services lost by society as a result of using resources to comply with a
regulation.1
Resource costs, which are the primary measure encouraged by Executive Order
12291, reflect the total impact on society rather than its distributional aspects.
In other analyses, particularly the Regulatory Impact Analyses now being performed
in the Office of Solid Waste, differential impacts on industries will be assessed
in detail. The use of the results of this project, which is designed as a screen or
1The costs of health care, the loss of economic resources, and the replacement
or protection of resources are not yet included in the model, but will be
considered as economic risks when they are later incorporated.
ICF Incorporated
-------
2-4
EXHIBIT 2-2
TECHNICALLY FEASIBLE WASTE STREAMS BY TECHNOLOGY
Technology
Technically Feasible Waste Streams
T reatment
Phase Separation:
Chemical coagulation
Filter press
Centrifuge
Vacuum filter
Component Separation:
Evaporation/drying
Air stripping
Steam stripping
Solvent extraction
Leaching
Distillation
Reverse osmosis
Where component X is present primarily (generally
> 80%) as a suspended solid.
Where the suspended nonwater fraction is less than
5 percent.
Where component X is present primarily (generally
> 80%) as a suspended solid.
When the suspended nonwater fraction is greater
than 5 percent.
Where component X is present primarily (generally
> 80%) as a suspended solid.
When the suspended nonwater fraction is greater
than 5 percent.
Where component X is present primarily (generally
> 80%) as a suspended solid.
When the suspended nonwater fraction is greater
than 5 percent.
Where the nonwater fraction is greater than 10
percent, and less than 70 percent.
Where component X is substantially more or less
volatile than water.
Where component X is primarily dissolved.
Where component X is substantially more or less
volatile than water.
Where component X is primarily dissolved.
All
Where component X is present primarily (generally
> 80%) as a suspended solid.
Where component X is substantially more or less
volatile than water.
Where component X is primarily dissolved.
Where the suspended nonwater fraction is near zero.
ICF Incorporated
-------
2-5
EXHIBIT 2-2 (continued)
TECHNICALLY FEASIBLE WASTE STREAMS BY TECHNOLOGY
Technology
Technically Feasible Waste Streams
Carbon adsorption (PAC)
Ion exchange
Chemical Transformation:
Chemical precipitation
Chemical destruction
Electrolytic decomposi-
tion
Chemical Fixation/
Stabilization
Incineration (all)
Where the suspended nonwater fraction is less than
5 percent.
Where component X is primarily dissolved.
Where the suspended nonwater fraction is near zero.
Where the total organic fraction of the waste is
near zero.
Where component X is primarily dissolved or
where oxidation/reduction of heavy metals is
contemplated.
Where component X is primarily dissolved.
Where component X is primarily dissolved.
Where oxidation/reduction of heavy metals is
contemplated.
All
All
Transportation (all)
All
Disposal
Landfills (all)
Surface impoundments
(all)
Land treatment
Deep well injection
Ocean
All
Where the suspended nonwater fraction is less than
10 percent.
Waste streams numbers 46, 50, 51, 52, 70-73, and 75
(see Appendix 6).
Where the suspended nonwater fraction is less than
0.5 percent.
Where specific gravity of the waste is greater than
1.0.
ICF Incorporated
-------
2-6
EXHIBIT 2-3
UNIT COSTS OF TECHNOLOGIES
Technology
Unit Cost
T reatment
Phase Separation
Chemical coagulation
Filter press
Centrifuge
Vacuum filter
Component Separation
Evaporation/drying
Air stripping
Steam stripping
Solvent extraction
Leaching
Distillation
Reverse osmosis
Carbon adsorption (PAC)
Ion exchange
Chemical Transformation
Chemical precipitation
Chemical destruction
Electrolytic decomposition
Chemical fixation/stabilization
Incineration
T ransportation
$1.00/kgd
1.00/kg
2.00/kg
0.50/kg
0.25/kg
16.00/kg
16.00/kg
16.00/kg
1.00/kg
8.00/kg
8.00/kg
1.00/kg
8.00/kg
0.50/kg
2.00/kg
64.00/kg
1.00/kg
b
Onsite
c
Local
3.30
per
cubic
meter
Long-distance
18.40
per
cubic
meter
Disposal
Double-lined landfill
$15.70
per
cubic
meter
Single-lined landfill
11.40
per
cubic
meter
Unlined landfill
8.25
per
cubic
meter
Double-lined surface impoundment
27.50
per
cubic
meter
Single-lined surface impoundment
16.00
per
cubic
meter
Unlined surface impoundment
4.80
per
cubic
meter
Land treatment
23.00
per
cubic
meter
Deep well injection
5.10
per
cubic
meter
Ocean
10.50
per
metric ton
Cost per kg of nonwater fraction.
^Cost depends on BTU content of waste.
cCosts of moving waste onsite are included in the treatment and disposal
costs.
ICF Incorporated
-------
2-7
guide for further analyses, must obviously be tempered by concerns about possible
uneven distributional effects. What are now excluded from the model are transfer
payments, that is, payments between parties that redistribute money but do not
consume resources such as taxes, fees, and insurance.2
At this stage of the project, we use direct capital and operating costs
associated with various waste management technologies. Direct capital costs
include expenses for acquiring and installing assets (land, facilities,
equipment). Among direct operating costs are materials, labor, utilities,
maintenance of equipment, and transportation. We do not include indirect costs
(such as engineering and design costs, licenses and permits, and administrative
overhead) because these items usually represent relatively fixed percentages of
direct capital costs (although not operating costs). We will reconsider including
indirect costs in subsequent analysis.
Since we are primarily examining existing facilities, siting costs (which are
indirect) are now excluded. We do not assume that it is equally difficult to site
all types of hazardous waste treatment or disposal facilities since the cost of
siting can vary greatly, but we intend to test assumptions about siting costs at a
later stage using sensitivity analysis. To the extent that differences can be
identified, we intend to adjust the cost scores.
Estimating costs of waste management technologies poses a number of difficult
problems. It is possible, however, to develop specific costs on a treatment-
transportation-disposal chain if we have good estimates about the type and volumes
of wastes being handled. There is considerable practical experience in the
engineering community that provides a foundation on which to base such estimates of
costs, unlike the situation in estimating risks.
Another important output of this model is to predict how wastes shift among
technologies as costs of treatment and disposal vary. To make such predictions,
the model requires more precise estimates for costs than for risks. The range over
which costs vary is also more limited than that for risk. Thus, we selected a cost
scale with narrower divisions than the one used for risks. This also provides more
variation in the lower portion of the scale, which is helpful because high-cost
treatment and disposal technologies are infrequently used. The cost score is used
in the model only after all the costs for all treatment, transportation, and
disposal are summed: the component costs are detailed but we score the total costs
on the basis of the range within which they fall. The scores now used in the model
distinguish between costs that are at least two times as great as other costs in
contrast with risk scores that reflect tenfold differences (see Chapter 3). We
estimated ranges around the typical scores to recognize special situations and, by
doing so, provide the basis for sensitivity testing.
2The cost of liability insurance is substantial for some firms and varies
greatly for different types of operations. Once the model incorporates economic
risks such as cleanup costs, the resource costs of liability insurance will be
fully incorporated except for the overhead of the insurance companies. To include
it now would, in effect, double count some of the effects on health while including
only one element of risk not related to health. We prefer to include those data
after the model has proven its usefulness in assessing risks to human health.
ICF Incorporated
-------
2-8
We expect that the typical costs in the model are overestimates of the actual
direct capital and operating costs. This is because the model does not account for
waste-generating plants redesigning their production processes and other waste-
handling methods to redistribute or reduce their waste discharges to minimize total
costs. It will often be less expensive to redesign a process, segregate waste, or
take measures other than to apply many treatment and disposal processes to existing
waste streams.
We specify costs in dollars per unit of the nonwater portion of the waste.
Specifically, the dollar costs are divided by the weight of the nonwater portion of
the waste stream. We will, however, be analyzing differences caused by using
alternative formulations.
Model Uses Infinite Time Horizon and 3 Percent Real Discount Rate
To construct a data base on risks and costs, it is necessary to make
assumptions, which can be varied and tested fo'r sensitivity, about how far into the
future to consider costs and risks. For ocean disposal, treatment technologies,
deep well injection, and land treatment, the model assumes that all costs and risks
(or releases to the environment of hazardous constituents) occur within the useful
life of the technology or shortly thereafter and are not continuing. Landfills and
surface impoundments will have perpetual care requirements for monitoring and
maintenance; we, therefore, compute those costs as continuing expenses.3
The present value of a dollar spent in the future is less than a dollar today;
how much less is measured by the discount factor applied for the time value of
money. The model uses a discount factor of 3 percent, which is consistent with
other cost analyses in the Office of Solid Waste and is roughly consistent with
historical patterns. The time value of money may be thought of as the difference
between expected inflation and long-term interest rates. There has until now been a
rough 3 percent differential, but today's high interest rates coupled with low
inflation would yield a value closer to 10 percent.
Although we use a 3 percent discount rate, we considered other rates for this
project. Exhibit 2-4 shows the effect of using various discount factors in
computing the total cost of the model double-lined landfill and unlined surface
impoundment. This exhibit shows that the choice of a discount rate barely affects
the relative costs of technologies that have grossly different patterns of costs
through time. A double-lined landfill, for example, has long-term and continuing
costs as compared to a facility with mostly near-term costs, such as an unlined
surface impoundment. The use of a 3 percent discount factor, therefore, results in
a unit cost that matches other studies, compares well with historical patterns, and
is insensitive to different values.
3The model now assumes, however, no additional risks (or possible associated
cleanup costs) from these facilities after 100 years on the assumption that they
have reached a steady state.
ICF Incorporated
-------
2-9
EXHIBIT 2-4
RATIO OF PRESENT VALUES OF UNIT COSTS
FOR SELECTED TECHNOLOGIES AND DISCOUNT RATES
! ! , !
Double-Lined Unlined Surface
Discount Rate Landfill Impoundment Ratio
(dollars/cubic meter)
i i i i
1
$18.40
$5.65
.31
3
15.70
4.80
¦ .31
6
13.50
3.95
.29
10
11.50
3.15
.27
Cost Estimates Are for Typical Facilities
Hazardous waste treatment and disposal technologies vary in size and
configuration and thus in capital and operating cost. Complete accuracy concerning
cost would require consideration of each variation in order to determine its effect
on unit costs. We do not believe that such detail is necessary, because our
analysis seeks only to estimate unit costs to the nearest factor of two. If during
sensitivity analysis we find greater accuracy is necessary, we will more accurately
define the sensitive technologies.
For this analysis, the model uses "typical" configurations of technology and
capacities. For many technologies this allows the use of broad definitions of
treatment and disposal and introduces only modest errors that are unlikely to
affect the cost scores. Thus, we can say that in the typical case, a simple
evaporation basin is less expensive than a more elaborate distillation process per
unit of waste. For incinerators, landfills, and surface impoundments, however, we
specified several types or levels of facilities, which allows more precise
estimates for these technologies.
We believe it is more important to make precise cost estimates for disposal
technologies and incineration than for treatment technologies. Disposal costs
typically dominate the total cost; treatment costs are usually incurred to avoid or
reduce disposal costs. We determined the average size of existing disposal
facilities, by type, using information from EPA's Part A permit application data.
By estimating the costs of necessary features and equipment to support a facility
of this average size, we calculated unit costs for operating such facilities.
Wherever possible, we used standard construction cost-estimating guides. If
necessary, we consulted the literature or made professional judgments. We then
computed the costs on a per unit volume of the total waste stream for those
technologies for which capacity depends upon volume, not weight. This approach
differs somewhat from other reported unit costs, which often list such costs on a
per ton
ICF Incorporated
-------
2-10
per ton or per 1,000 kilograms basis. For treatment technologies (except
incinerators), we used capacity cost curves as reported in the literature. We
chose the typical capacity of those technologies on the basis of facilities in
operation as reported by EPA and others and used the cost corresponding to units in
that size range. The estimated total costs for each treatment, transportation, and
disposal step in a chain handling a given waste stream are then summed. The total
cost is then divided by the weight of the nonwater fraction of the initial wa-ste
stream for comparison purposes and expressed as a score (Exhibit 2-5).
EXHIBIT 2-5
COST SCORE INTERVALS
i i i
Unit Cost
Score ($/kg of nonwater fraction)
i i i
0 $ 0.18
1 0.18-0.35
2 0.35-0.71
3 0.71 - 1.41
4 1.41 - 2.83
5 2.83-5.66
6 5.66 - 11.31
7 11.31 - 22.63
8 22.63-45.25
9 45.25 - 90.51
10 90.51 - 181.02
11 181.02 - 362.04
12 362.02 - 724.08
13 724.08 - 1,448.15
14 1,448.15 - 2,896.31
15 2,896.31 - 5,792.62
16 5,792.62 - 11,585.24
Onsite, Local, Or Long-Distance Transportation Is by Truck
A hazardous waste may be disposed of either where it is generated (onsite) or
at a disposal facility (offsite). Both kinds of disposal require a handling or
transportation step, each with its own release rate and cost. For onsite disposal,
we assume that a 1/4 mile separates the point of generation from the point of
disposal. This assumption is consistent with some previous EPA studies, and we
believe that it is representative of onsite disposal operations. We use this
transportation distance to calculate the environmental release rate. Since we
assume that the equipment used for onsite disposal is also used in transporting
wastes onsite, we include the cost of onsite transportation in the cost of the
disposal technology.
For this analysis, we use two kinds of offsite transportation: local, a
highway truck trip of 50 miles round trip, and long-distance, a round trip of 500
miles. We selected these lengths as typical of transportation distances, with an
order-of-magnitude difference between them. We believe it is appropriate to assume
that most shipments of hazardous waste are by highway, although it probably results
in both a higher release rate and a higher cost than would transporation by rail or
water.
ICF Incorporated
-------
2-11
Many different kinds and sizes of vehicles are used in transporting hazardous
wastes. For this analysis, we assume that transportation is by tank truck with a
6,000-gallon (about 23 cubic meters) capacity. Tank trucks are more expensive than
most other highway vehicles (about 2.5 times more expensive than a flatbed trailer,
for example), but tank trucks do not require as much costly loading equipment as
other vehicles. The 6,000-gallon capacity used for this analysis represents a tank
truck similar to large gasoline tractor-trailer combinations.
RELEASES TO THE ENVIRONMENT ARE EITHER ROUTINE OR ACCIDENTAL
Releases from technologies fall into two categories -- routine and accidental.
(See Exhibit 2-6 for a summary of the release rate factors for each technology.)
Routine releases include the following:
• evaporation and volatilization from landfills,
surface impoundments, and land treatment and from open
tanks during treatment;
• aeration, or the dispersion of waste droplets into
the atmosphere because of high-speed agitation
processes;
• leachate generation from landfills, surface
impoundments, and land treatment;
• routine spillage from poor handling techniques
and/or poor process design;
• runoff from land treatment;
• air emissions from incineration.
We made a number of simplifying assumptions regarding routine releases.
Leachate generation and migration from landfills, for example, depend on
precipitation, characteristics of the cover material, characteristics of the liner
material, efficiency of leachate collection, and hydrogeologic features of the
site. By using a simple water-balance model and assuming that all leachate is
saturated with the constituent of concern, we simplified the model of the complex
process of leachate generation. Specific assumptions and rationales for all the
technologies appear in the detailed discussion of technologies in the appendices.
In every case the assumptions are intended to identify significant (order-of-
magnitude) differences in routine release rates.
Accidental releases are due to random spillage and are, therefore, based on an
estimate of the probability of a release and an estimate of the quantity released.
If, for example, there is a 2.5 in a million chance per mile that a truck will be in
an accident, if the trip is 250 miles, and if 20 percent of the truck's contents are
projected to be lost in the accident, then the truck's release rate would be
2.5xl0-6 x 250 x 0.20 = .000125 or about 0.01 percent per trip. Accidental spillage
also includes large volume spills owing to severe equipment failure. This is
primarily a function of the age and maintenance history of the equipment and the
care with which the system is operated.
ICF Incorporated
-------
2-12
For releases from deep well injection, we assume the well casing fails and the
failure is not immediately detected. This is essential because a properly
installed deep well should have no release unless a failure occurs in the system.
Our approach to determining typical release rates relies on manufacturers' and
other published literature and mass transfer theory, especially for spillage and
aeration losses from treatment processes. In using data on reported releases, we
temper published data by applying our perception of current hazardous waste
management practices. We, more specifically, take account of the likely dependence
of releases upon the maintenance and operation of the particular treatment
equipment. As a result, we estimate these emissions for two cases: (1) best case,
for a relatively well-maintained and operated system; and (2) worst case, for a
relatively poorly maintained and operated system. For the "typical" case, we
estimate the total emissions on the basis of a subjectively weighted average of the
emissions for the best and the worst cases.
Release Rates Partly Depend upon Waste Characteristics
In estimating releases to the environment from the treatment, transportation,
and disposal technologies, we consider the characteristics of the waste as a
variable. The amount of waste that escapes into the air from an open-tank process,
for example, partly depends on the volatility of the waste. Of course, several
types of releases (for example, spillage and aeration losses) do not depend on the
characteristics of the waste. Where the model allows releases to vary, we have
included in Exhibit 2-6 the waste characteristics that cause them to vary.
Detailed descriptions of their derivations appear in the appendices.
Waste Characteristics/Technology Assumptions
Determine Pollutant Migration into Different Media
When a release occurs, the location, weather conditions, and the nature of the
waste determine what portion of the constituents of concern will migrate via
surface water, ground water, and air. In the model, we use a simpler approach.
Migration to air (volatilization) is assumed to occur as the release occurs.
Migration to surface water includes some portion of the remaining waste that is in
liquid form; that migration depends upon the imperviousness of the area (normally
expressed as a runoff coefficient). Although runoff coefficients vary widely from
place to place, we assume a typical value for industrial areas of 90 percent. Thus,
90 percent of the liquid constituents are assumed to migrate to surface waters, and
10 percent to ground water.
With the exception of releases related to transportation (about 20 percent of
which we assume remain in place after recovery operations are completed), we assume
that spilled material is not cleaned up. We assume further that the small
quantities involved in most routine releases from waste treatment facilities are
overlooked in typical operations.
TREATMENT TECHNOLOGIES ALTER WASTE STREAMS
The purpose of treating hazardous wastes is either to reduce the danger of
disposing of a waste or to reduce the cost of its eventual disposal. We define
waste streams in terms of the characteristics shown in Exhibit 2-7. These
characteristics change in response to the treatment used, but the amount of change
ICF Incorporated
-------
EXHIBIT 2-6
RELEASE RATES BY TECHNOLOGY
Evaporation/
Volatizatlon
Routine
Aeration
Leachate
Spillage
Runoff
Stack
Emissions
Accidental
Spillage
Deep Well
Treatment
Phase Separation
Chemical Coagulation
Filter press
Centrifuge
Vacuum filter
l.OxlO-5
1.0x10
-7
1.0x10-4
1.2xl0-3
4.3xl0"6
2.6xl0-4
l.lxlO"4
2.2xl0-4
l.lxlO-4
Component Separation
Evaporat ion/drying
Air stripping
Steam stripping
Solvent extraction
Leaching
Dist illat ion
Reverse osmosis
Carbon adsorption (PAC)
Ion exchange
Chemical Transformation
Chemical precipitation
Chemical destruction
Electrolytic decomposition
Chemical Fixation/Stabilization
Chemical Fixation/Stabilization
l.OxlO"7
1.0x10
1.0x10
-5
-5
l.OxlO-5
l.OxlO"5
l.OxlO-6
1.0x10
-7
1.0x10
-5
1.0x10-6
l.OxlO"4
l.OxlO-4
l.OxlO"6
l.OxlO"4
5.2xl0"7
l.OxlO"4
l.OxlO"6
2.6xl0-5
l.OxlO"4
7.7xl0"4
1.3xl0~4
2. lxlO-4
1.6xl0~4
3.0xl0-4
2.lxlO-4
1. 9xl0"*4
4.3xl0-4
4.3xl0"9
2.6xl0~4
l.lxlO-5
2.2xl0"4
1.5x10"
ho
I
I—1
CO
o
-n
3
O
o
n
T7
O
-i
b>
rh
(I>
Q.
Incineration
Incineration—99.99% DRE
Incineration—99.9% DRE
Incineration—99.9% DRE
Incineration—90.0% DRE
a—depends on the volatility of waste.
b—depends on the solubility and proportion of free liquids in the waste.
l.OxlO-4
l.OxlO-3
l.OxlO"2
l.OxlO-1
-------
EXHIBIT 2-6 (cont.)
RELEASE RATES BY TECHNOLOGY
Accldental
Evaporiza-
tion/
Volatization
Transportat ion
On-site
Local
Long Distance
Aeration Leachate Spillage
2.0X10"4
2.0xl0"4
2.0xlCT4
Runoff
Stack
Emissions
Spillage
1.3xl0"7
1.3x10-5
1.3xl0-4
Deep Well
DisDOsal
Double-lined landfill
Single-lined landfill
Unlined landfill
Double-lined surface impoundment
Single-lined surface impoundment
Unlined surface impoundment
Land treatment
Deep well injection
Ocean
1.2xl0-4
2.9xl0-4
a—depends on the volatility of waste.
b—depends on the solubility and proportion of free liquids in the waste,
c—depends on the solubility of the waste.
-------
2-15
EXHIBIT 2-7
CHARACTERISTICS OF HAZARDOUS WASTE STREAMS
Specified by User
1. Total weight including water (kg/day)
2. Percent nonwater by weight
3. Percent nonwater suspended by weight
4. Average specific gravity of suspended solids at 25°C.
5. Percent of X" by weight
6. Percent of X dissolved by weight
7. Molecular weight of X
8. Relative volatility (vapor pressure of X)/(vapor pressure of water) at
25°C.
9. Solubility of X in water at 25°C. (mg/1)
10. Net BTU content (BTU/kg)
Computed by Model
11. Percent nonwater dissolved by weight (#2-#3)
12. Percent of X suspended by weight (#5-#6)
13. Total X by weight (kg/day) (#5 x #1)
14. Total X dissolved by weight (kg/day) (#6 x if 13)
15. Total X suspended by weight (kg/day) (#12 x if 13)
16. Total X that is in fixed solid phase (kg/day) (only computed after
chemical fixation/stabilization)
17. Total weight excluding water (kg/day) (#1 - #2)
18. Total water by weight (kg/day) (#1-#17)
^constituent of concern
ICF Incorporated
-------
2-16
depends, as did release rates, upon the nature of the waste. Exhibit 2-8 shows
typical changes in the characteristics shown in Exhibit 2-6. These values are
included as illustrative of the changes made. For the changes to a given waste
stream, the algorithms in the appendices describe the actual changes that the model
makes .
We have incorporated 21 treatment technologies into the risk-cost framework.
We selected technologies that broadly represent the feasible methods for treating
hazardous waste. We recognize, however, that the list of technologies is not
exhaustive and suggest that others be added later. They should include some of the
emerging treatment systems now in the developmental stages.
The 21 treatment technologies can be classified into 5 groups: phase
separation; component separation; chemical transformation; chemical
fixation/stabilization; and incineration. A synopsis of each treatment technology
follows. We provide a detailed description of their functional aspects and the
algorithms we used to model their treatment of hazardous materials in the
appendices.
Phase Separation Divides Solids and Liquids
Chemical coagulation is used to produce settling of very small suspended solid
particles. Such solids may include viruses, bacteria and other colloidal inorganic
and organic material. Clarification or filtration is then used to separate the
solid from the liquid portion of the sludge.
Filter press is used to dewater and to concentrate the suspended solids in a
sludge stream. The filter press retains suspended solids from the bulk solution by
passing the fluid through woven media under pressure. The solids are retained on
the membrane while the filtrate, low in suspended solids, passes through the
membrane.
Centrifuge is used to dewater and concentrate the suspended solids in a sludge
stream by mechanical means. Centrifugal separators use gravity settling to
separate solids from the liquor, but improve the efficiency of the process by using
a strong centripetal acceleration instead of the force of gravity. The results are
more rapid separation and a solids cake containing less residual liquor.
Vacuum filter is used to dewater and concentrate the suspended solids in a
sludge stream by mechanical means. A filter removes suspended solids from the bulk
solution by drawing the fluid through porous woven media using a vacuum. The solids
are retained on the media while the filtrate, low in suspended solids, passes
through and is discharged.
Component Separation Isolates Constituents
Evaporation/drying is used to dewater sludge streams that do not contain
volatile compounds, particularly those with a high concentration of dissolved
solids.
ICF Incorporated
-------
EXHIBIT 2-8
TYPICAL CHANGES TO WASTE STREAMS
Techno Ioqy
Phase Separation
Chemical coagulation
FiIter press
Centri fuge
Vacuum f iIter
Waste Characteristic*
1
3
5
1
2
3
5
1
2
3
5
1
2
3
5
Typ i caI
Initial Cond i t i on
<0.5%
<0.5%
2-30%
2-10%
1-20%
2-30%
2- 5%
1- 5%
2-30%
2- 5%
1- 5%
Typical Reduction
After Treatment
50-99%
2%
0- 2%
50-90%
20-60%
20-50%
10-25%
50-80%
10-50%
10-30%
10-30%
50-90%
15-50%
15-30%
15-30%
M
I
o
3
O
o
-t
•U
0
1
11)
r+
(0
Q.
Component Separation
Evaportat ion/dryi ng
Solvent extraction
Leach i ng
1
2
3
5
Air stripping & Steam stripping 1
2
5
6
11
1
2
5
6
11
1
2
3
5
6
0-70%
0-70%
0-70%
1-20%
1-20%
1-20%
1-20%
0-20%
1-20%
1-20%
1-20%
0-20%
0- 2%
0- 2%
0- 2%
30-99%
70%
70%
0-70%
95%
50%
10-50%
10-50%
50%
60-90%
90%
10-90%
10-90%
90%
0-90%
2-20%
2%
0- 2%
0-20%
*Exhibit 2-6 lists all the characteristics. Only characteristics affected by the technology are shown.
-------
EXHIBIT 2-8 (cont.)
TYPICAL CHANGES TO WASTE STREAMS
3
O
o
"O
o
n
o>
rh
ID
Q.
Techno Iogy Waste Characteristic*
Component Separation (continued)
D i s t i I I a t i on
Reverse osmosis
Carbon absorption (GAC)
Ion exchange
1
2
3
11
1
2
5
11
1
2
3
6
11
1
2
3
6
11
TypicaI
Initial Cond i t i on
0-50%
0- 2%
0-20%
0-20%
0-20%
0-20%
0-20%
0-0.52
0-20%
0-20%
0-20%
0
0-20%
0-20%
Typ i caI Reduct i on
After Treatment
0-95%
50%
0-70%
30-90%
80%
5-20%
5-20%
5-20%
80%
75%
90%
10%
10%
80%
5-30%
0- 2%
0-30%
5-30%
N>
CD
Chemical Transformation
Chemical precipitation
Chemical destruction
Electrolytic decomposition
O
1
2
3
6
11
o
11
2
6
11
0-40%
2%
0-U0%
0-10%
0-20%
0-20%
0-20%
0-10%
0- 3%
0-10%
90%
2-10%
2%
0%
0%
10%
0%
20%
10%
0%
10%
-------
EXHIBIT 2-8
(cont.)
TYPICAL CHANGES TO WASTE STREAMS
Techno I og.v Waste Characteristic*
Chem i caI F ixat i on/Stab iIi za t i on
ChemicaI fixation/stabiIization
Incineration
Incineration -- 99.99% DRE
Incineration -- 99.9% DRE
Incineration -- 99% DRE
Incineration -- 90% DRE
1
2
9
1
2
13
10
1
2
13
10
1
2
13
10
1
2
13
10
Typ i ca I
Initial Condition
0-40%
50-100%
10,000 BTU/kg
50-100%
10,000 BTU/kg
50-100%
10,000 BTU/kg
50-100%
10,000 BTU/kg
Typical Reduction
After Treatment
(25% increase)
100%
0.01%
34-fold increase
oran i cs: 0.1
metaIs:
o rgan i cs
metaIs:
50%
95%
0.01%
0
17-fold increase
organics: 0.2%
metaIs:
organ ics:
metaIs:
50%
95%
0. i;
0
8.5-fold increase
organics: 0.4%
metaIs:
organ ics:
metaIs:
50%
95%
1%
0
o rganics:
metaIs:
95%
90%
95%
10%
0
-------
2-20
Air stripping is used to reduce the concentration of dissolved gases or vapors
in an industrial sludge stream. The two most common applications remove volatile
organics for reuse or for decreasing the load on downstream waste treatment
processes and removal of dissolved gases to prevent air pollution problems.
Steam stripping is used to remove moderately volatile organic liquids from the
sludge for recovery or to decrease the load on downstream waste treatment
processes. Steam stripping is, in principle, similar to air stripping, but can
remove species with lower vapor pressures by virtue of the heat and altered
equilibrium provided by the superheated steam.
Solvent extraction is used to extract specific compounds from the sludge for
recycling or to minimize their effect on treatment units downstream. The process
achieves partial separation of the liquid constituents of an industrial sludge by
virtue of their different solubilities in two insoluble liquid phases.
Leaching is used in industrial sludge treatment for selectively dissolving and
separating a portion of the influent solids from the bulk sludge.
Distillation is used to separate the constituents of a liquid mixture, by virtue
of their differences in vapor pressure, for the purpose of removing or reclaiming a
portion of the more volatile species. A properly designed and operated
distillation unit can yield nearly complete recovery of the volatile components
from the mixture.
Reverse osmosis concentrates the dilute dissolved species in RCRA waste
streams by using a differential pressure across a selectively permeable membrane.
Water is transported through vacancies in the molecular structure of the membrane
material. Even organic and ionic species of low molecular weight may be
concentrated by using reverse osmosis. This process is sensitive to various
chemicals that may damage the structure of the membrane.
Carbon adsorption (PAC) is used to remove soluble trace and refractory
organic compounds from (pretreated) waste streams. Two major types of carbon
adsorbants are used: powdered activated carbon (PAC) and granular activated carbon
(GAC), but PAC is the predominant treatment process for hazardous wastes.
Ion exchange concentrates dissolved ions from dilute waste streams. The ions
are adsorbed onto the ion exchange resin, then desorbed by washing with a strong
acid, base, or salt solution during a resin regeneration cycle. The most common
application for industrial waste streams is the removal of metals and inorganic
salts. Pretreatment is often required to remove soluble organics, which can react
permanently with the ion exchange resin. All suspended solids must be removed by
filtration to prevent resin plugging and interference with the adsorption-
desorption process.
Chemical Transformation Changes or Destroys Constituents
Chemical precipitation removes dissolved species from sludge by chemical
reaction to form insoluble compounds. The resulting suspended solids are separated
from the liquor by clarification or if highly concentrated by filtration.
ICF Incorporated
-------
2-21
Chemical destruction is used to "destroy" the hazardous component in a sludge
through chemical reaction to form one or more less hazardous or nonhazardous
species. This treatment category includes oxidation, reduction, or other reactions
in which the products remain in the dissolved phase and are retained in the RCRA
waste stream (such as ozonation and wet oxidation).
Electrolytic decomposition is used to transform hazardous ionic species in
industrial sludges to nonhazardous forms using an electric cell. This technology
is applied primarily in the treatment of spent plating solutions where chlorides
are oxidized to cause the decomposition of cyanides while heavy metal ions are
reduced to less toxic forms simultaneously.
Chemical Fixation/Stabilization Solidifies Wastes
Chemical fixation/stabilization is used to solidify a sludge stream by adding
chemicals that bind sludge components including water into a solid matrix. This
treatment is preferred when leaching or leaking of hazardous wastes from storage
vessels or disposal facilities could present severe hazards. There are several
different processes available for chemical fixation/stabilization, such as
silicate and cement, lime, and thermoplastic. For the purposes of the risk-cost
analysis, we considered only one generalized system. This is a treatment module
that could be better defined by subsequent efforts.
Incineration Destroys Constituents
Incineration destroys hazardous organic components by oxidation to form carbon
dioxide, water, and other combustion gases. The model provides four levels of
treatment by incineration, corresponding to four levels of destruction and removal
efficiency (DRE) for the constituent of concern (X) between the influent waste
stream and the final release of combustion products to the environment:
99.99 percent DRE implies 0.01 percent of X released
99.9 percent DRE implies 0.10 percent of X released
99.0 percent DRE implies 1.00 percent of X released
90.0 percent DRE implies 10.00 percent of X released
TRANSPORTATION ALLOWS WASTES TO USE LANDFILL AND
INCINERATION CAPACITY IN OTHER AREAS
The various treatment and disposal modules discussed in this chapter are not
equally available throughout the U.S. Even where facilities do exist, there are
practical limitations on the amount of wastes some facilities can accept. The high
capital costs associated with constructing some of these facilities, notably
landfills and incinerators, preclude the rapid addition of new capacity. These
practical constraints related to capacity become important when developing a
feasible nationwide strategy for hazardous waste management. Using an offsite
transportation module allows the model to make these geographic shifts while
accounting for the risks and costs of moving the waste.
ICF Incorporated
-------
2-22
The ideal situation would be to match treatment and disposal capacity with the
rate of waste generation on a facility-by-facility basis. Severe limitations on
the data available for both capacity and volume of waste, however, prevent applying
this approach. For example, the Part A permit application data provide the
approximate available capacity for landfills, but there exists no good source of
information to help us determine how much of this capacity is typified by our
double-lined landfill. In addition, we believe that many owners or operators of
treatment and disposal facilities may have overstated their available capacity in
the Part A applications.
In spite of these problems, we believe it important to estimate limitations on
the capacity of four of our treatment and disposal technologies: the double-lined
and single-lined landfills and the 99.99 percent and 99.9 percent DRE incinerators.
We expect that the balance of the treatment and disposal technologies will not be
constrained for capacity, either because excess capacity is available or new
capacity can be readily and cheaply added. Should a particular strategy seem to
require improbable amounts of capacity of other treatment or disposal technologies,
however, this constraint can be included in the model.
Although we believe that many owners or operators overstated capacity in their
Part A application data, we use that information as the basis for determining
available capacity for landfills and incinerators. As better information becomes
available, we will adjust our assumptions regarding capacity. In deciding how to
apportion the capacity reported in the Part A applications between the double- and
single-lined landfills and between 99.99 and 99.9 percent DRE incinerators, we
believe it appropriate to distinguish between onsite and offsite commercial
facilities. In the absence of specific data on the characteristics of commercial
and onsite landfills, we applied our experience to derive the following general
estimates of the mix among categories.
We assume that 25 percent of the capacity of commercial offsite landfills is
represented by our double-lined scenario and that the remaining 75 percent are
single-lined. Among onsite facilities, we assume that the double-lined scenario
represents 10 percent of capacity, the single-lined scenario 60 percent, and the
unlined scenario the remaining 30 percent.
For commercial offsite incinerators, we assume that 25 percent of capacity is
99.99 percent DRE incinerators and 33 percent, 99.9 percent DRE. For onsite
incinerators, we assume that 10 percent of the capacity is represented by the 99.99
percent DRE incinerator scenario and 25 percent by the 99.9 percent DRE scenario.
We use the capacity limitations only in the strategy formulation stage of
applying the model. As we discuss in Chapter 4, such limitations ration scarce
capacity among competing waste streams in a given environment. Lack of capacity in
one area can be offset by capacities in other regions of the country by using a
transportation technology. The objectives of the strategy determine whether we use
the limited capacity on a given waste stream.
The methodology by which we estimate volumes of waste streams is discussed in
Chapter 3.
ICF Incorporated
-------
2-23
DISPOSAL TECHNOLOGIES PROVIDE FOR
LONG-TERM MANAGEMENT OF WASTES
Disposal technologies handle the waste remaining after applying treatment
technologies and transportation and waste that is not treated. We use five types
of technologies in the model:
• landfills;
• surface impoundments;
• land treatment;
• deep well injection; and
• ocean disposal.
We also define three levels of landfills and surface impoundments to reflect the
range of design, operating practices, risks, and costs of operating these
technologies.
Landfills Bury Wastes
Landfilling -- the most common method of disposing of hazardous wastes --
usually involves placing the waste in a specially prepared excavation or trench and
then covering it with fill material. Modern landfills have a wide range of
specific features that depend upon the characteristics of the individual site.
The design of landfills is highly site specific; thus, no single landfill
configuration can be regarded as "typical." To reflect the range of available
design and operating practices, we defined three scenarios for hazardous waste
landfills: double-lined, single-lined, and unlined. All have the same total waste
capacity (674,000 cubic meters) and the same operating area (25 acres). Exhibit 2-9
outlines the features of the three scenarios.
We assume that excavated landfills can release wastes only to ground water
through leachate migration and to air by volatilization. Placing waste below
surface grade precludes releases directly to surface water unless the leachate
collection system fails. Accidental spills can, of course, occur, but we account
for them in the transportation technology.
Surface Impoundments Hold Liquid Wastes
Surface impoundments are often considered long-term storage facilities, but,
in regions of the country where evaporation rates are high, they can serve as
disposal sites. Evaporation reduces the volume of waste disposed of, and the site
can be eventually closed in a fashion similar to closing a landfill. Our approach
uses evaporation and ground water migration as the mechanisms that determine
disposal capacities. To reflect the differences in size, features, and
configurations, we define three typical scenarios, which we outline in Exhibit 2-
10. All three have the same surface area (2.5 acres), volume (16,660 cubic
meters), and evaporation rate (5,000 cubic meters per year). Waste capacity
differs widely among the scenarios, however, depending on the quantity of the waste
liquids lost through the liner or soil.
ICF Incorporated
-------
2-24
EXHIBIT 2-9
FEATURES OF THREE LANDFILL SCENARIOS
1 1
Feature
i i
Double-Lined
i I
Single-Lined
i i
l
Unlined
i
Leachate collection
Extensive
Some
None
Leachate treatment
Custom plant
Package plant
None
Intermediate cover
Twice daily
Once daily
At closure
Monitoring
8 wells
6 wells
None
Control of surface
runoff
Some
Some
Little
Final cover
Clay, artificial
liner, soil, &
revegetation
Clay, soil, and
revegetation
Topsoil
& revegetation
Postclosure care
1
Extensive
Some
Little
1
Common
Features
Life span
20 years
Area3
25(50) acres
Dimension
500 x 200 x 10m
Shape
3 straight walls
1 3:1 slope wall
Total void
962,267 m3
i
Capacity
70% of void =
673,587 m3
i
g
25 acres of operating area and 50 acres of total disposal facility.
ICF Incorporated
-------
2-25
EXHIBIT 2-10
FEATURES OF THREE SURFACE IMPOUNDMENT SCENARIOS
1 1
Feature
i i
1
Double-Lined
i
1
Single-Lined
i
l
Unlined
i
Leachate collection
and treatment
Yes
None
None
Berms
Extensive
Some
Little
Control of surface
runoff
Some
Some
Little
Monitoring
6 wells
4 wells
None
Final Cover
Clay, artificial
liner, soil &
revegetation
Clay, soil, and
revegetation
Topsoil &
revegetation
20 year capacity
100,000 m3
113,400 m3
268,000 m3
Postclosure care
|
Some
Little
None
i
1
Common
Features
1
Life span
20 years
Aread
2.5(5) acres
Dimension
100 x 100 x 3m
Side slope
3:1
Total void
23,800 m3
i
Volume**
70% void = 16,
660 m3
i
£
2.5 acres of surface impoundment and 5 acres of total disposal facility.
^About 30 percent volume (5,000 m3) of the liquid waste per year is lost
through evaporation. Thus, 5,000 m3 of liquid waste can be added annually,
totaling 100,000 m3 for 20 years. An additional 8,400 m3/year is added to the
unlined impoundment and 670 m3/year to the single-lined impoundment, owing to
ground-water releases (see Appendix 3).
ICF Incorporated
-------
2-26
Land Treatment Degrades Wastes
Land treatment -- also known as land farming, land cultivation, and soil
incorporation -- is a process by which industrial wastes are mixed with surface
soils and allowed to decompose. The objective in land treatment is to assure that
chemical constituents in the waste are retained and/or decomposed in the surface
layer. Microbial degradation is the principal decomposition mechanism, but
volatilization and chemical and photochemical degradation are also important
processes in land treatment. The process is not suitable for wastes too toxic or
too persistent to be degraded. Our model accepts only low-persistence halogenated
organics and other low-persistence hydrocarbons for land treatment.
On the basis of the Part A permit application data for land treatment
facilities and our professional judgment, we selected a 10-acre operating area as
typical for a land treatment facility.
Deep Well Injection Pumps Wastes Underground
Deep well injection is a process whereby wastes are pumped under pressure into
an area below the earth's surface. Our model estimates risks by analyzing the
probability of system failure. There are, of course, other significant issues,
including possible effects of the practice on deep geologic formations (resulting
in seismic activity) and the possibility that future generations may wish to use
the deep disposal areas for another (incompatible) purpose that are not amenable to
formal analysis. We also intend to explore adding risks from releases into
aquifers that are used for drinking water. We set the condition that, to be
injectable, the suspended nonwater fraction of the waste must be less than or equal
to 0.005, that is, 5,000 parts per million.
Ocean Disposal Disperses Wastes
Any waste that does not float is a technically feasible candidate for ocean
disposal. In ocean disposal, wastes are either dumped in bulk from special barges
or placed in containers (usually, weighted or perforated 55-gallon drums) and
dumped. Hazardous wastes are normally disposed of well offshore, in areas without
significant aquatic life that would be immediately affected by the practice.
We consider deep ocean waters a special case of surface waters. Thus, releases
from ocean disposal are releases to surface waters. Losses during transportation
on land and loading of wastes are considered transportation losses and are
discussed in the transportation section above. We assume that all of the liquid-
phase constituent of concern is immediately released to surface waters and that
deterioration and release of solid phase components begin almost immediately. We
set 10 years as typical for complete release of solid-phase constituents, an
assumption that we believe to be conservative, and assume that one-tenth of the
solid-phase constituents are released on an annual basis, together with all liquid-
phase constituents.
ICF Incorporated
-------
CHAPTER 3
WASTE STREAMS
This chapter has two sections. The first describes how we developed risk
scores for hazardous waste streams and their constituents; the second describes the
characteristics of the waste streams now in the model's data base. The second
section, begins on page 3-35.
RISK SCORING SYSTEM ASSESSES HUMAN HEALTH RISKS BASED ON
TENFOLD DIFFERENCES IN INHERENT HAZARD AND EXPOSURE
In this section, we describe the procedures used to assign risk scores to
different waste streams. The primary purpose of the scoring system is to assign
each W-E-T a numerical score that represents the risk of the waste stream and
differentiates it as efficiently as possible from the risks presented by other
waste streams.
The risk posed by a waste stream is related to a number of factors, the most
important of which are:
• the inherent hazard and physical properties of the
waste stream, which depends primarily on the nature of
the waste stream and its chemical constituents;
• the quantity of the chemicals released and the
medium into which they are released, which depend on
the size and nature of the waste stream, and on the
technology;
• the persistence of the chemicals, the rate of their
dispersion, and other factors governing exposure,
which depend primarily on the nature of the waste
stream and on the environment;
• the size of the human population potentially
exposed to the chemicals, which depends primarily on
the environment; and
• the presence of nonhuman systems that can be
adversely affected, such as sensitive ecosystems or
valuable resources, which depend primarily on the
environment.
Thus, risk involves all three components of the W-E-T matrix and the risk scoring
system accounts for all three.
In this phase of developing the model, we make four major decisions that
greatly simplify the problem and allow us to test our approach, while also
preserving the factors that will usually dominate any comparison of risk.
ICF Incorporated
-------
3-2
1. The scoring system is an index of risks to human health.
There are three major types of risk: to human health, to economic resources
(such as water supplies, fisheries, crops, or land values), and to nature (for
instance, ecosystems, endangered species, or the beauty of an area). Our judgment
is that risks from the mismanagement of hazardous wastes can, in most cases, be
subsumed under the human health score. Damage to economic resources usually occurs
because of actual risks to health (which, if discovered, result in shutting off of
water supplies, condemning contaminated food, and so forth) or perceived risks to
health (which may result in decreasing land values). Also, with few exceptions,
chemicals and waste streams that pose risks to wildlife and ecosystems also pose
risks to human health. Where a substance or a type of facility poses a
disproportionate risk to nonhuman resources, the model will, at a later date, be
able to assign it a higher score. Such exceptions might include the release of
chemicals that damage the ozone layer, the release of herbicides into aquatic
systems, or the location of facilities in areas that are the habitat of endangered
species. With the degree of aggregation used in this analysis, however, the only
W-E-Ts we have so far identified where a higher score can undoubtedly be justified
for nonhuman risks are those involving ocean disposal. Thus, at the present stage
of the model's development, separate scores for economic and environmental risks
have not been assigned. Scores for ocean disposal must, therefore, be interpreted
with caution.
2. The scoring system combines all types of health risk into a single
numerical score; a second score characterizes the range of uncertainty.
The primary purpose of the W-E-T matrix is to permit exploration of trade-offs
between the risk and cost of different waste management situations. Although human
health risk has many different components, it would be impractical at this stage of
the analysis to analyze these trade-offs if the individual types of human health
effects were assigned independent scores. The data on health effects, although
impressive in size and sophistication, are not yet comprehensive enough to be
comparable for most chemicals or types of effects. Rather than use many categories
of effects with few entries in each, we believe a more general measure (ideally, a
single •measure) is more useful for the present model. We believe that the
probability of an "incident" is such a measure. For these reasons, we combine the
scores. While this procedure groups into a single scale risks that are likely of
different severity, the model can be adjusted in the future to test whether
different weights for effects alter the results.
3. The scoring system characterizes the risks of waste streams by their
individual chemical constituents.
Most waste streams are complex mixtures of chemicals, and many are poorly
characterized chemically. For these reasons, there is comparatively little
information on the inherent hazard of waste streams per se, especially at the level
of aggregation used in this analysis. Hence, it is necessary to infer the inherent
hazard and other properties of waste streams from what is known about the
properties of their components. The risk posed by different types of streams is
ICF Incorporated
-------
3-3
then represented by the risk of the most hazardous components.1 Since the risks
posed by most waste streams are dictated by a few key components (usually the most
hazardous, persistent, and predominant), primary attention is given to these key
components. We decided to score chemicals that are known to be important
components in some waste streams. The procedure we applied has obvious
limitations: it does not account for interactions of hazardous substances
(synergism and antagonism), and it does not consider minor or unknown components.
Because of the level of aggregation used in the analysis, however, we do not
believe that these limitations are very serious.
4. The scoring system assigns scores to the nearest order of magnitude.
The scoring system is relatively simple and, in certain respects, crude, in
order to permit efficient application to a very broad range of circumstances. The
guiding principle in developing the risk scoring system is to match the level of
precision in scoring to the level of precision required to use the scores in
analyzing trade-offs among costs and risks. We determined that a relatively coarse
scale of risk scores is both appropriate and sufficient to differentiate among
situations. The most important reason is that the reliability of the assigned
scores is limited primarily by the paucity of data, especially on the environmental
persistence and dispersibility of chemicals. Although it is possible to assign
scores for inherent hazard for many chemicals on a finer scale and with, perhaps,
greater precision than we used here, there would be little point in expending
resources to do so since scores for exposure can only be assigned on a coarse scale.
Thus, we did not expend time and resources on details that would not contribute to
the precision in the total score.
For all these reasons, practical limitations and a wide variability in expected
results, we use a coarse scale to score risk. Specifically, we score the inherent
hazard and potential for exposure of waste stream constituents to the nearest order
of magnitude. That is, each higher score reflects a tenfold increase in risk to
humans if released into the environment.
Scoring System Is Based on Theoretical Model
There are two general problems in scoring risks to health posed by hazardous
wastes. First, most waste streams are complex mixtures, and many are poorly
characterized.2 Second, even if adequate information were available on the
composition of the wastes, for most chemicals, there is insufficient information on
inherent hazard and environmental behavior to permit precise estimates of risks.
For these reasons, scoring for risks to health is subject to large uncertainties
and in many cases depends on scientific judgment. The scoring scheme we developed
does, however, base application of scientific judgment on a formal scientific
underpinning. By providing rigorous definitions of the factors to be scored and by
1The risk posed by a given amount of a chemical in a given environment reflects
both its inherent hazard and its persistence and dispersibility in that
environment. Since chemicals vary in their persistence and dispersibility, it is
necessary to calculate their risks individually. It would not be appropriate, as
an alternative, to calculate an average score for the inherent hazard of a waste
stream because, for example, the more hazardous components may degrade rapidly,
before there is opportunity for significant human exposure.
2See the section beginning on page 3-35 for the waste streams now in the data
base.
ICF Incorporated
-------
3-4
specifying numerical equivalents for the scoring categories, it is possible to
assign fairly precise scores to a number of well-studied chemicals. These scores
can then guide the assignment of scores to less well-studied chemicals and to
mixtures. The following conceptual model of risks from hazardous wastes guided our
efforts, but the procedures we actually followed are described in the rest of this
chapter.
Many factors contribute to the risk posed by a W-E-T -- the number of people
potentially affected, the average intake per person per day, the probability of
response per unit of intake, and the severity of response -- and all are considered
independently. These factors are defined so that they interact multiplicatively in
contributing to the overall risk. We assign scores to these factors on an
approximate logarithmic scale and then sum them3 to provide a measure of overall
risk:
log (risk) = constant + (sum of scores)
The constant is unspecified; thus, the system initially provides a measure only of
relative risks.^
We define the overall risk presented by disposing of a waste as the expected or
most likely number of people affected. Since some chemicals persist essentially
forever after release into the environment, the scoring system also assesses the
risk of disposing of the waste indefinitely.5 The system allows for the option of
discounting risks that do not occur until 30 or more years in the future, but does
not now discount risks.
We define the major factors as follows:
Overall risk (expected number of people affected) = P • E • H,
where: P = Population at risk (number)
E = Exposure (average intake per person per day)
H = Inherent hazard (probability
of response per unit intake)
3Summing logarithmic values is arithmetically equivalent to multiplying their
standard values.
ukt a later stage, we calibrate the relative risk scores by analyzing a few
well-studied cases in which the absolute risk has been estimated. The scores can
then be used as approximate measures of absolute risk. That is not the primary
purpose of the system, however, and our scores should not be used to replace more
formal risk assessments in applications outside the scope of the W-E-T matrix.
sWe ignore the possibility that a compound may have more than one opportunity
to inflict harm on an individual. Metals, for example, may remain toxic forever
and threaten human health after recombining with many other chemicals through the
ages.
ICF Incorporated
-------
3-5
As presented above, the overall risk is equated with the number of people
affected, irrespective of the type of effect. In practice, some types of effects
are generally regarded as more "serious" than others. To take account of this, the
scoring system could be generalized by adding a (multiplicative) index of severity,
which would approach a value of 1 for the most "serious" effects, but might come
close to a value of 0 for effects judged to be least serious. The assignment of
indices of severity involves value judgments, which we have tried not to apply in
this phase of the study.
The submodel for exposure is:
Exposure (average intake per person per day) = Q*F,Di,In#A
where: Q = Quantity of material handled
F = Expected fraction released
Di = Dispersion factor (concentration in the ambient
environment per unit of release)
In = Intake factor (breathing rate, consumption of
drinking water, and so on)
A = Absorption coefficient
The last three subfactors vary considerably from one environmental medium to
another. Hence, it is necessary to assign and sum exposure scores separately for
each medium. The highest of the scores for the various media is essentially the
measure of risk because of the logarithmic scale (that is, any score less than the
highest is at most one-tenth as large and is not considered when rounding).
In our model, the overall risk depends linearly on the quantity of material
that is released. Unlike the other factors, quantity is not an inherent property
of the W-E-T. It can be reduced, at increased cost, by improving the level of
technology or by dispersing the waste stream to other facilities. As noted
earlier, one of the objectives of the analysis is to explore the trade-offs between
cost and risk resulting from such changes. Hence, the quantity released is not a
consideration in' the risk scoring system. The final risk score is an estimate of
the logarithm of the overall risk per unit of material released.
Because uncertainties are large in many cases, the system provides for
assigning a measure of uncertainty to scores for both inherent hazard and exposure.
Uncertainties arise from four major sources:
• lack of information (for example, on inherent hazard
or environmental behavior);
• site-to-site variability within a W-E-T (such as
quantities of materials released or hydrologic
conditions);
• variability related solely to the environment (for
instance, differences in susceptibility within the
population); and
• chance occurrence (such as probability of leakage).
ICF Incorporated
-------
3-6
Although these types of uncertainties are formally distinct, we assume that
they can be combined meaningfully into a single measure of probability. We assume
further that the distribution of probabilities (in this broad sense) for each
factor is about lognormal.6 Then the logarithmic score on each factor can be
characterized by a single expected value (mean) and variance, and the variances can
be combined by standard methods to yield a single measure of uncertainty in the
overall risk. These overall measures of uncertainty are, in practice, similar for
many waste streams; thus, they contribute little to the process of ranking. We
found some cases, however, where we judged the uncertainty in the risk scores to be
substantially larger than average. The system provides for identifying these cases
and for carrying estimates of uncertainty through the analysis.
A number of exposure factors (number exposed, intake factor, absorption
coefficient) are measured on numerical scales; thus, assignment of logarithmic
scores is relatively straightforward. The most difficult factors to score are
inherent hazard and environmental transport, which are discussed below.
Inherent Hazard Is Expressed As a Single Score
As discussed earlier, the system yields a single score for all the potential
hazardous effects of each chemical. In this respect, it differs from many other
scoring systems, which either are limited to one type of hazardous effect or yield
different scores for different effects or different circumstances of exposure.
Derivation of a single score from the complex body of data on the inherent hazard of
a chemical necessarily involves a number of simplifying assumptions. The
scientific issues underlying these assumptions are discussed in the following
paragraphs.
Effects of hazardous substances on an animal species depend on a number of
ways: the age, sex, and genetic characteristics of the individuals exposed; the
route of exposure; and the magnitude, timing, and duration of exposure. Such
effects may, therefore, be classified in the following ways:
• By animal species. Data obtained from experimental animals
are routinely used to predict safe levels of exposure for
humans, but this raises difficult problems in estimating
relative susceptibility and in scaling for differences in size,
life span, and so forth.
• By route of exposure. Most of the toxicity studies from
which data have been used to assign scores involve exposure
either by inhalation or by ingestion. In some cases, it is
possible to use data obtained in inhalation studies to predict
hazards resulting from ingestion and vice versa. These cases
usually arise when the chemical is absorbed into the body and
gives rise to effects at distant sites. Even in such cases,
however, it is often difficult to predict dose-response
relationships unless the relative amounts of the chemical
absorbed via the two routes are known. Where the effect of a
chemical is primarily at the point of contact with the body
sLognormal describes a probability distribution in which one expects many more
observations having slight differences from the true value than having great
differences.
ICF Incorporated
-------
3-7
(for example, the lungs, skin, or gastrointestinal tract), it
is often very difficult to use data obtained by one route of
exposure to predict hazards resulting from another.
• By frequency and duration of exposure. Acute exposure
usually refers to a single administration of a chemical;
subchronic exposure to administration for 1-6 months; and
chronic exposure to administration for more than 6 months.
• By type of effect. Responses to hazardous chemicals are
classified as "dichotomous" if an individual animal either
does or does not respond (if it dies, contracts cancer, or
exhibits birth defects) or as "graded" if the response is
measured on a continuous scale (change in weight or change in
activity of an enzyme). Note that the type of effect being
studied determines the form of the dose-response
relationship. Carcinogenic effects are often characterized
by linear, nonthreshold dose-response relationships, whereas
it is customary to characterize other types of response by
dose-response relationships that incorporate thresholds.
In assessing risks, we are generally concerned with the potential effects of
chronic exposure to low levels of hazardous chemicals. For this reason, we have
placed primary weight on the results of studies of chronic effects, either in
humans or in animals. Where subchronic studies are the only ones available, we use
them with the expectation that chronic effects would occur at somewhat lower doses.
To predict likely responses at low levels of exposure, it is usually necessary to
extrapolate from responses observed at high levels. For carcinogenic effects, it
is customary to fit the data to linear, nonthreshold dose-response models. For
other effects, we have developed a procedure for extrapolating from the minimum
effective dose (MED: the lowest dose that produces a clearly observable effect)
down to lower doses. Both procedures may overestimate risks, particularly at very
low levels of exposure, but we have matched them in such a way that they should give
comparable results.
The key feature of the scoring system is the definition of the inherent hazard
factor as the probability of response per unit dose. For carcinogenic chemicals,
this definition coincides exactly with the definition of "unit risk". The widely
adopted assumption is that dose-response relationships for carcinogenic effects
are likely to be linear and nonthreshold at low doses; thus, the slope of the dose-
response relationship is independent of dose and is defined as the unit risk. We
have calculated unit risks for a number of carcinogens, primarily on the basis of
animal data, although linear dose-response relationships can also be fitted to
human data. For the purpose of scoring carcinogenic chemicals in this study, we
have used estimates of unit risks derived by EPA, where they are available. In
other cases, we have either calculated unit risks ourselves by employing EPA's
procedures or used a simple approximation to derive an estimate of the unit risk.
This approximation is a linear dose-response function that provides a close
approximation to the unit risk when the observed incidence of cancer is less than
about 50 percent, but progressively underestimates the risk as the observed
incidence of cancer increases.
ICF Incorporated
-------
3-8
For hazardous effects other than carcinogenesis, it is not normal procedure to
calculate unit risks. For graded responses (such as impairment of kidney function
or neurobehavioral effects), we base the score for these responses on an estimate
of the daily dose required to produce an adverse effect in about one percent of the
exposed population. Because of statistical limitations, it is difficult, if not
impossible, to detect a frequency of response lower than 1 percent, either in
animal studies or epidemiological studies. In typical cases the smallest effect
that can be detected with statistical reliability is about 10 percent, ranging up
to 30 percent or more in studies using small samples or with high background
variability. The score is, accordingly, based on the MED required to produce a
measurable adverse effect. For an observed effect in humans, we then assume that a
MED/10 (that is, the number of milligrams per kilograms per day divided by 10) will
approximate the dose that yields a 1 percent probability of producing adverse
effects. We further assume that a generally linear relationship holds (at least
within the accuracy of this scoring system). Hence, for example, if the MED
observed in humans is 10 milligrams per kilogram of body weight per day, we can
assume that a dose of 1 milligram per kilogram per day will lead to roughly a 1
percent probability of producing adverse effects and assign the same score as to a
chemical with a unit risk of 0.01.
For an effect observed in animals, we base the score on the MED divided by 100
if the MED is obtained in a chronic toxicity study and by 300 if it is obtained in a
subchronic study. The difference between the factors of 100 and 300 in animal
studies and the factor of 10 applied to MEDs in human studies reflects two broad
generalizations: effective doses are expected to scale between humans and animals
in rough proportion to body surface (about a factor of 10 for both types of
studies), and MEDs in subchronic studies are typically on the order of three times
higher than those in chronic studies, for a combined factor of 30. 7
Although these scoring rules are relatively crude, we expect that they will
yield measures that are reliable at least to the order-of-magnitude accuracy that
is the objective of the system. Using these scores in a multiplicative risk model
involves, however, the implicit assumption that the relationship between dose and
probability of response is linear, an assumption that may overstate risk. The use
of linear dose-response relationships for noncarcinogenic effects is
controversial, but has two theoretical bases. First, the human population is
highly variable so that there is expected to be a wide spectrum of susceptibility
to many types of hazardous agents. Second, the human population is exposed to a
wide variety of agents; whenever the effect of one agent is to add to the background
of others, the relationship between incremental increases in dose and incremental
increases in the probability of response is likely to be linear. Thus, the dose-
response relationship for a variable population may be close to linear even when
one for a homogeneous group is not. The assumption of linearity is not necessarily
inconsistent with the existence of thresholds, but requires that population
thresholds should be very low and that the population dose-response curve be
roughly linear between the threshold and 10 percent response.
'These factors are not the same as "safety factors" that are conventionally
applied to no-observed-effect levels in animals to estimate safe doses for humans.
Our purpose here is not to estimate safe levels, but to estimate the doses at which
low frequencies of response are likely to occur.
iCF Incorporated
-------
3-9
The general assumption is that doubling the amount of material released in the
environment will about double the number of people likely to respond. In greater
generality, we could assume that the probability of response will vary as the kth
power of exposure over the range of exposures under consideration. The procedures
outlined above assume that k = 1. If it seems appropriate to assume a higher value
for k, this could easily be achieved by using the scoring system to weight the
exposure score by k and the inherent hazard score by 1/k. This procedure has not
been applied in this phase of development of the system, but could be implemented
easily in subsequent phases.
Exhibit 3-1 presents the scoring rules for inherent hazard in a flow chart
form. Higher scores are assigned to the more hazardous chemicals -- to carcinogens
with the higher unit risks and to noncarcinogens with the lower MEDs. Inherent
hazards (unit risks) are presented in units of 1 milligram per kilogram per day,
that is, in units of risk per unit dose. A unit risk of 0.05 per 1 milligram per
kilogram per day, for example, means that a constant dose of d milligrams per
kilogram per day leads to a lifetime risk of 0.05 d. The choice of numerical limits
for the scoring categories (for example, 0.0005 to 0.005 rather than 0.0001 to
0.001 or 0.0002 to 0.002)is arbitrary, since the inherent hazard scale is, in fact,
continuous.
It should be noted that we converted all doses and exposures to equivalent
doses in milligrams per kilograms per day, using conversion factors and intake
factors that are listed in Appendix 8.
data are primarily from secondary sources
We evaluated the inherent hazard of the substances we scored primarily by
reviewing secondary sources in order to determine the lowest exposure level at
which each has been shown to exert an adverse effect on health. Both human and
animal data were considered in this review. Some of the principal secondary
sources were the monograph series of the International Agency for Research on
Cancer (IARC), the National Academy of Sciences (NAS) series Drinking Water and
Health and Medical and Biologic Effects of Environmental Pollutants, NIOSH Criteria
Documents, Casarett and Doull s Toxicology, and EPA's Ambient Water Quality
Criteria Documents. In addition to these secondary sources, we consulted
carcinogenesis bioassays of the National Cancer Institute (NCI). Clement
Associates had primary literature for almost one-third of the chemicals. We also
conducted computerized searches of the literature.
The secondary sources listed above were generally based on critical, expert
scientific review and are usually reliable sources of information on inherent
hazard, at least for purposes of this scoring exercise. We did, however, build in
some safeguards. After identifying the secondary sources that indicated hazardous
effects at the lowest dose levels, for example, we obtained and briefly reviewed
the primary reference in each case. Where questions arose concerning the validity
of the toxicological evidence for a particular compound as reported in the
secondary sources, we also reviewed several primary references. We established our
inherent hazard score on the basis of the best-supported evidence.
ICF Incorporated
-------
3-10
EXHIBIT 3-1
FLOW CHART FOR SCORING INHERENT HAZARD
ICF Incorporated
-------
3-11
complex chemistry and missing data are problems
There were, of course, problems concerning inherent hazard scores. For certain
classes of compounds, particularly heavy metals, a complex mode of action and/or
complex chemistry made scoring difficult. In these instances, we prepared a
detailed toxicological profile before assigning a score, and we scored the most
serious reactions (for example, the effects of inhaling hexavalent chromium
compounds determined the score for chromium compounds) and the least serious. We
then selected a score that corresponded reasonably to the middle range of inherent
hazard of chemicals in the group. For example, although a score of 6 would be
assigned to mercury compounds on the basis of data on methylmercury, a score of 4 or
less would be appropriate for other mercury compounds. Hence, an overall score of
5 was assigned to mercury compounds as a class, and the uncertainty score was
increased to 1.5 to reflect the wide range of inherent hazards within the class
(see the section on uncertainty scores, below).
Data for toxicological evaluation were insufficient or absent for a few
compounds or classes of compounds. In these cases, we assigned scores on the basis
of analogies with compounds of similar structure for which data were available.
Polynuclear aromatic hydrocarbons (PAHs), for example, for which inherent hazard
data were not available, were scored as a group on the basis of the carcinogenic
potential of many of the PAHs and the evidence that PAHs do not occur naturally in
the pure state but in complex mixtures with many other PAHs.
Another group of chemicals that proved difficult to score were criteria air
pollutants (sulfur and nitrogen oxides). We assigned their scores on the basis of
information in EPA's Ambient Air Quality Criteria Documents, recognizing that their
effects on populations depend on their interactions with other chemicals and that
sulfur oxides are much less hazardous in aqueous media. Ignoring the atmospheric
conversion of hydrocarbons into photochemical oxidants, we scored them according to
their direct inherent hazard.
140 compounds are scored
We chose to score 140 chemicals to ensure adequate description of industrial
waste streams in terms of individual species. There were, however, constraints on
our data base in regard to the availability of data and of resources. Our scoring
procedure places primary weight on the results of studies of chronic effects in
either humans or animals, and, of the about 50,000 compounds for which there are
toxicological data, fewer than 5,000 have been characterized in terms of chronic
effects. The constraint regarding the availability of resources dictated the
number of chemicals scored. More can be scored and added to the system, as the need
arises. Work on defining waste streams was conducted after much of the scoring was
complete. It is interesting that, of the 140 compounds we have scored, only 53
appear in the waste streams now used in the data base. Only 14 compounds, however,
that are now in the data base do not yet have inherent hazard scores. Those
compounds are 1,1,1,2-Tetrachloroethane, 1,1,2-Trichloroethane, Phthalic
anhydride, Naphthoquinone, Paraldehyde, 2-Picoline, Aniline, Phenylenediamine,
Diphenylamine, Trichloropropane, Dichloropropanol, m-Dinitrobenzene, Benzyl
Chloride, and Benzotrichloride.
ICF Incorporated
-------
3-12
The substances we chose to score appear in Exhibit 3-2 (the exposure
adjustment, discussed in the next section, also appears in this exhibit). The
initial basis for selection was EPA's list of 129 priority pollutants. From this
list we deleted some chemicals believed to be of limited significance (either
because they occur in small quantities or because their contribution to the hazards
of waste streams is adequately represented by the scores assigned to structurally
similar compounds that characteristically occur with them). We added other
chemicals to the list on the basis of our knowledge that they are key components of
major waste streams. Appendix 7 summarizes the basis for the scores appearing in
Exhibit 3-2, including the types of effects on which they are based. Scores for
inherent hazard range from 0 to 6, which represents at least a millionfold range of
hazard.
Exposure Adjustments are Made for Three Media
We define exposure as the average intake per person exposed per day. To take
account of spatial variations in both population density (our proxy for the
potential number of people exposed) and environmental considerations, we use a
population-weighted average. Because the time scales and length scales for
environmental transport vary greatly from one environmental medium to another, it
is necessary to assign scores separately for each medium. To make the scores
commensurate (that is, to make the same score in each of the three media correspond
to the same risk), it is then necessary to develop a rigorous model of exposure for
each medium. Appendix 7 contains this generalized exposure model.
We assigned exposure scores to three media: air, surface water, and ground
water. Once a chemical enters the environment, there is a certain probability that
it can move from the point where it was released. Indeed, the distribution of some
chemicals -- carbon tetrachloride and DDT, for example -- is global in scope.
Distribution of a chemical in the environment reflects both its use (over a large
geographical area for a wide variety of applications) and properties (ability to
move in the environment).
When a chemical enters air or water, its dispersion (technically its intraphase
mass transfer and diffusion) is generally rapid because of fluid movements.
Dispersion of a chemical is not, however, solely a function of mass transfer but
also of chemical stability in any given environment. There are competing
processes: a chemical that is extremely stable in terms of environmental
degradation -- for example, carbon tetrachloride -- is dispersed throughout the
environment; a chemical that degrades rapidly is not dispersed far from its source.
Although intermedia transport is important to this movement of chemicals throughout
the environment, we cannot describe this process in detail since knowledge of such
transport is incomplete. For this reason, we scored chemicals by individual
medium; transport of a chemical between media is considered degradation.
Exposure adjustment scores (see Exhibit 3-3) are roughly equivalent to the log
of the expected half-life of each contaminant in each medium. A constant is
subtracted from each score so that a score of zero corresponds to time periods
considered insignificant in each medium in terms of risk; that is, risk is
effectively zero.
ICF Incorporated
-------
3-13
EXHIBIT 3-2
SCORES FOR 140 SUBSTANCES
Exposure Adjustments
Adjusted Inherent
Hazard by Medium
Compound
Inherent Air Surface Ground-
Hazard Water water
A BCD
Air
E
(A+B)
Surface
Water
F
(A+C)
Ground
water
G
(A+D)
Acenaphthene
Acenaph thylene
~Acetaldehyde
* Acetonitrile
Acrolein
Acrylamide
~Acrylonitrile
4-Aminobiphenyl
Anthracene
~Antimony
~Arsenic
~Asbestos
ot-BHC
3-BHC
y-bhc
*Bar ium
~Benzene
Benzidine
3
3
2
2
5
3
5
4
3
5
5
3
2
2
2
3
2
2
3
3
2
6
2
2
3
2
4
4
4
4
5
5
5
4
4
3
2
2
0
0
0
4
0
0
3
4
0
0
6
5
7
7
7
6
7
0
5
7
7
0
5
5
5
4
6
0
6
6
4
8
7
5
8
6
7
9
9
6
5
5
5
2
2
5
7
5
4
7
12
12
7
5
5
5
7
2
2
9
8
9
9
12
9
12
4
8
12
12
3
7-
7
7
7
8
2
ICF Incorporated
-------
3-14
EXHIBIT 3-2 (cont.)
SCORES FOR 140 SUBSTANCES
Adjusted Inherent
Exposure Adjustments
Hazard by Medium
Inherent
Air
Surface
Ground-
Air
Surface
Ground-
Compound
Hazard
Water
water
Water
water
A
B
C
D
E
(A+B)
F
(A+C)
G
(A+D)
:Benzo(a)anthracene
2
4
6
5
6
8
7
Benzo(a)pyrene
6
4
6
5
10
12
11
Benzo(g,h,i)perylene
3
4
8
5
7
11
8
Benzo(k)fluoranthene
3
4
6
5
7
9
8
3,4-Benzofluoranthene
2
4
6
5
6
8
7
Beryllium
6
4
2
2
10
8
8
Bis (2-chloroethoxy)-
3
1
0
0
4
3
3
methane
Bis(chloromethyl)ether
5
1
0
0
6
5
5
Bis(2-ethylhexyl)-
1
3
6
6
4
7
7
phthalate
Butane
0
3
0
7
3
0
7
Butene/Butadiene
0
2
0
7
2
0
7
'Cadmium
5
4
6
5
9
11
11
Carbon disulfide
5
6
0
7
11
5
12
rCarbon tetrachloride
3
8
0
6
11
3
9
'Chlordane
3
4
0
5
7
3
8
'Chloroacetaldehyde
4
3
0
6
7
4
10
4-Chloroaniline
2
3
1
5
5
3
7
rChlorobenzene
3
4
0
6
7
3
9
ICF Incorporated
-------
3-15
EXHIBIT 3-2 (cont.)
SCORES FOR 140 SUBSTANCES
Adjusted Inherent
Exposure Adjustments
Hazard by Medium
Inherent
Air
Surface
Ground-
Air Surface
Ground-
Compound
Hazard
Water
water
Water
water
A
B
C
D
E . F
G
(A+B) (A+C)
* (A+D)
~Chloroform
3
5
0
6
8
3
9
4-Chlor o-2-methylphenol
2
3
0
6
5
2
8
2-Chlorophenol
2
3
0
7
5
2
9
~Chromium (VI)
5
4
7
7
9
12
12
*Chrysene
4
4
6
5
8
10
9
~Copper
3
4
1
1
7
4
4
Creosote
2
1
7
4
3
9
~Cyanide
4
7
0
7
11
4
11
Cyclohexane
2
0
6
5
2
8
Dibenzo(a,h)anthracene
4
4
5
5
7
9
9
Di-n-butyl phthalate
2
3
3
6
5
5
8
~1,2-Dichlorobenzene
1
4
1
6
5
2
7
1,3-Dichlorobenzene
1
4
1
6
5
2
7
~1,4-Dichlorobenzene
2
4
1
6
6
3
8
3,3-Dichlorobenzidine
2
3
0
5
5
2
7
~1,2-D ichloroethane
3
4
0
6
7
3
9
~1,1-D ichloroethene
3
4
0
6
7
3
9
*Dichlor omethane
2
4
0
7
6
2
9
2,4-Dichlorophenol
2
3
1
6
5
3
8
ICF Incorporated
-------
3-16
EXHIBIT 3-2 (cont.)
SCORES FOR 140 SUBSTANCES
Exposure Adjustments
Adjusted Inherent
Hazard by Medium
Compound
[nherent
Hazard
A
Air
B
Surface
Water
C
Ground-
water
D
Air
E
(A+B)
Surface
Water
F
(A+C)
Ground-
water
G
(A+D)
2,4-Dichlorophenoxy-
acetic acid
2
3
0
5
5
2
7
1,2-Dichloropropane
2
4
0
6
6
2
8
1,3-Dichloropropene
4
2
0
6
6
4
10
0,0-Diethyl S-
(2-(ethylthio)ethyl)
ester of phosphoro-
thioic acid
4
4
3
5
8
7
9
0,0-Diethyl phosphoric
acid O-p-nitrophenyl
ester
4
4
3
5
8
7
9
1,2-Diraethyl hydrazine
4
2
0
6
6
4
10
2,4-Dimethyl phenol
4
2
2
6
6
6
10
Dimethyl phthalate
0
4
1
6
4
1
6
Dimethyl sulfate
4
0
0
0
4
4
4
4,6-Dinitro-2-
methylphenol
3
3
1
6
6
4
9
2,4-Dinitrophenol
3
3
1
6
6
4
9
' 2,4-Dinitrotoluene
3
3
1
6
6
4
9
2,6-Dinitrotoluene
3
3
1
6
6
4
9
1,4-Dioxane
2
3
0
7
5
2
9
Dodecane
2
3
0
6
5
2
8
Endrin
4
4
4
5
8
8
9
ICF Incorporated
-------
3-17
EXHIBIT 3-2 (cont.)
SCORES FOR 140 SUBSTANCES
Exposure Adjustments
Adjusted Inherent
Hazard by Medium
Compound
Inherent
Hazard
A
Air
B
Surface
Water
C
Ground-
water
D
Air
E
(A+B)
Surface
Water
F
(A+C)
Ground
water
G
(A+D)
*Epichlorohydrin 3
Ethylbenzene 2
Fluorene 3
Fluorine *(fluorides) 3
~Formaldehyde 5
*Hexachlorobenzene 3
*Hexachlorobutadiene 2
Hexachlorocyclo- 5
pentadiene
0
0
2
7
0
6
0
0
7
6
6
7
7
5
5
5
8
5
7
7
8
8
5
8
3
2
5
10
5
9
2
5
10
8
9
10
12
8
7
10
*Hexachloroethane
1
7 0
5
8
1
6
Hexachlorophene
3
4 2
5
7
5
8
Hexane
1
3 0
7
4
1
8
Hydrazine
4
2 3
6
6
7
10
Hydrogen chloride
3
4 N/A
N/A
7
N/A
N/A
Hydrogen sulfide
4
4 N/A
N/A
8
N/A
N/A
Indeno(l,2,3-c,d)-
pyrene
3
4 6
5
7
9
8
Isophorone
2
4 0
6
6
2
8
*Lead
5
4 3
2
9
8
7
ICF Incorporated
-------
3-18
EXHIBIT 3-2 (cont.)
SCORES FOR 140 SUBSTANCES
Inherent
Compound Hazard
A
Exposure Adjustments
Adjusted Inherent
Hazard by Medium
Air
B
Surface
Water
C
Ground-
water
D
Air
E
(A+B)
Surface
Water
F
(A+C)
Ground-
water
G
(A+D)
'Mercury
5
4
3
0
9
8
5
Methomyl
3
2
2
5
5
5
8
Methyl bromide
3
5
0
6
8
3
9
Methyl chloride
1
5
0
7
6
1
8
'Methyl ethyl ketone
0
3 ,
0
7
3
0
7
Methyl hydrazine
6
2
0
6
8
6
12
Methyl parathion
5
4
3
5
9
3
10
Naphthalene
1
4
0
6
5
1
7
2-Naphthylamine
2
3
3
6
5
5
8
'Nickel
4
4
2
2
8
6
6
'Nitrobenzene
2
4
0
5
6
2
7
Nitrogen oxides
3
3
N/A
N/A
6
N/A
N/A
2-Nitrophenol
2
3
0
6
5
2
8
4-Nitrophenol
3
3
1
6
6
4
9
N-Nitrosodimethylamine
4
3
0
5
7
4
9
N-Nitrosodiphenylamine
1
3
0
5
4
1
6
Octachlorostyrene
3
5
6
5
8
9
8
PCB-1016
5
4
6
5
9
11
10
PCB-1221
5
4
3
5
9
8
10
ICF Incorporated
-------
3-19
EXHIBIT 3-2 (cont.)
SCORES FOR 140 SUBSTANCES
Exposure Adjustments
Adjusted Inherent
Hazard by Medium
Compound
Inherent Air Surface Ground-
Hazard Water water
A BCD
Air
E
(A+B)
Surface
Water
F
(A+C)
Ground-
water
G
-(A+D)
PCB-1232 6
PCB-1242 6
PCB-1248 5
*PCB-1254 6
PCB-1260 6
*Parathion 6
*Pentachlorophenol 4
Phenanthrene 3
~Phenol 4
Pyrene 3
~Pyridine 2
Selenium 4
Silver 2
Sodium 2
Sulfur dioxide 3
2,3,7,8-Tetrachloro- 6
dibenzo-p-dioxin
*1,1,2,2-Tetrachloro- 4
ethane
4
2
0
2
3
4
5
3
3
2
1
3
1
7
5
7
N/A
5
5
5
5
5
5
5
6
5
6
5
5
7
5
5
N/A
5
10
10
9
10
10
10
7
7
6
7
6
8
6
6
7
8
10
6
8
8
10
11
9
7
5
5
6
3
11
7
9
N/A
11
11
11
10
11
11
11
10
8
10
8
7
11
7
7
N/A
11
10
*Tetrachloroethene
ICF Incorporated
-------
3-20
EXHIBIT 3-2 (cont.)
SCORES FOR 140 SUBSTANCES
Adjusted Inherent
Exposure Adjustments
Hazard by Medium
Inherent
Air
Surface
Ground-
Air Surface
Ground-
Compound
Hazard
Water
water
Water
water
A
B
C
D
E F
G
(A+B) (A+C)
(ArfD) -
~Thallium 5
~Toluene 1
~Toluene diamine 3
~Toluene-2,4-diisocyanate 6
o-Toluidine 1
hydrochloride
~Toxaphene 3
1,2,4-Trichlorobenzene 2
~1,1,1-Trichloroethane 1
*1,1,2-Trichloroethane 1
*Trichloroethene 4
2,4,6-Trichlorophenol 1
*Vanadium 5
Vinyl chloride 2
*Zinc 2
4
3
3
3
2
4
4
6
5
3
3
4
3
4
0
0
5
3
0
1
0
0
0
0
1
7
0
4
0
7
6
6
0
5
6
7
6
5
7
7
7
4
9
4
6
9
3
7
6
7
6
7
4
9
5
6
5
1
8
9
1
4
2
1
1
4
2
12
2
6
5
8
9
12
1
8
8
8
n
/
9
8
12
9
6
~Actually used in one of the data base's 83 waste streams
N/A—not applicable
ICF Incorporated
-------
3-21
EXHIBIT 3-3
EXPOSURE ADJUSTMENT SCORES
1 1
Half-Life
i i
Air
1
ab
Surface Water
i
1 1
be
Ground Water
i i
3 minutes
0
-
-
30 minutes
1
0
-
6 hours
2
1
0
3 days
3
2
1
30 days
4
3
2
1 year
5
4
3
10 years
6
5
4
100 years
-
6
5
1,000 years
-
-
6
di k
If bioaccumulation factor in fish is 10 , add K-2 points for K>3; leave
score unchanged for K<2.
If chemical is removed by conventional water treatment, subtract factor
based on degree of removal (90 percent removal = -1; 99 percent removal = -2).
c
If adsorbed to solid surfaces, subtract points according to degree of
adsorption (99 percent adsorption = -2; 99.9 percent adsorption = -3).
Note: Air scores include removal by atmospheric washout as well as
degradation. Surface-water scores include sedimentation and volatilization as
well as degradation. Ground-water scores include irreversible adsorption.
Although one or more mechanisms may attenuate the concentration of a given
compound in each medium, a single mechanism typically dominates (that is, is
greater than others by a factor of 10 as measured by half-life) the removal of any
specific compound. Thus, exposure scores are based on the half-life associated
with the dominant mechanism. Where two mechanisms are equally important (where the
half-lives differ by less than a factor of 10), scores are reduced by the log of 2.
In no instance did we find more than two attenuation mechanisms necessary to define
the half-life of a given compound.
The half-life of a substance in each medium includes both degradation processes
and processes resulting in transport of the compound to another medium (such as
evaporation from water to air). In the following sections we discuss briefly the
processes considered in each medium and summarize the main factors responsible for
attenuation and adjustment of a chemical's half-life by media, using chloroform as
an example.
Exhibit 3-2 shows the scores for inherent hazard and exposure. The adjusted
inherent hazard score is the sum of the exposure and inherent hazard scores.
ICF Incorporated
-------
3-22
The basic scale used for exposure scoring runs from 0 to 6, representing a 10-
millionfold range in environmental half-lives of chemicals. This range, however,
is extended even further by other factors. The most important of these is
bioaccumulation, which potentially raises the exposure scores for some chemicals in
surface water to as high as 9. Thus, the adjusted inherent hazard scores can range
from 0 to 15, representing 15 orders of magnitude variation in risk. Other
environment-specific factors such as population density and assimilative capacity
extend the range even further. In view of the wide range of scores assigned, the
limited precision of the available data and the simplified procedures used to
assign scores do not represent a significant limitation of the system.
air
We determined half-lives in air from the literature or extrapolated them from
data on structurally analogous compounds. We assumed that adsorption on dust
particles and subsequent washout are insignificant for volatile contaminants. We
treated nonvolatile particulates (inorganic compounds) equivalently and assigned
half-lives of 7 days (score = 3.5), a value typical of atmospheric washout.
Reaction with the ambient hydroxyl radical population was the major mechanism of
removal from the air of components in the vapor phase unless contrary information
was available.8
For chloroform, abstraction of hydrogen by the hydroxyl radical is the most
important degradation mechanism.9 The exposure scoring adjustments (see Exhibit 3-
3) imply a score for chloroform of 4.6, on the basis of an atmospheric half-life of
chloroform of about 7 7 days (derived from the equation t, = 0.20 years and
(k=l.7x10-13)).
surface water
We assumed attenuation of a compound in surface water occurs by one of four
mechanisms: oxidation, hydrolysis, biodegradation, or volatilization. We based a
chemical's score on what we judge is the most important mechanism.
8Half-lives (t^) for hydrogen abstraction by the hydroxyl radical have been
derived from reported rate constants (k) as t^ = 0.69/k[0H], where t^ is in
seconds, k is in cm3/molecule/sec, and the ambient hydroxyl radical concentration
is 10s molecules/cra3 (Atkinson et al., 1979). Scores are derived as air exposure
score = log t^ - 2.3.
9The reaction rate of chloroform with atomic singlet oxygen appears to be
somewhat greater than that of abstraction of hydrogen by the hydroxyl radical.
Ambient concentrations of the hydroxyl radical, however, are significantly greater
than of atomic singlet oxygen.
C13CH + OH' = C13C" + H20; C^C" + 0 -~ C13C02
C13C02 C0C12 + 0C1; 2C13C02 "* 2C0C12 + C12 + °2
ICF Incorporated
-------
3-23
We used two factors to adjust the scores: adsorption and bioaccumulation,
which decrease and increase the potential for human exposure, respectively. We
obtained half-lives for hydrolysis and biodegradation either from the literature or
by structural analogy. Where no data were found and structural analogy did not
indicate otherwise, we assumed hydrolysis and biodegradation to be unimportant.
Volatilization half-lives were calculated in the literature for vapor pressure
(P°), solubility (S°), and molecular weight (m), using the two-resistance theory
for transport across an interface.10
We estimated the potential for bioaccumulation from octanol-water partition
coefficients (P[oct]) when they were unavailable from the literature. They were set
equal to the log of P[oct] . We then adjusted surface-water scores by adding the
bioaccumulation factor minus 2 (a scaling adjustment). In cases where the
bioaccumulation factor minus 2 was less than 0 we used 0.
Chloroform has a bioaccumulation factor of 102. Hence, its exposure score
would not be adjusted. As an example of an adjustment, the bioaccumulation factor
for hexachlorobenzene is 106. Hence, we add a factor of 4 (6 minus 2), equivalent
to a 10-thousandfold increase in exposure to account for bioaccumulation.
Adsorption is an important attenuation mechanism for contaminants in both
surface water and ground water. We used values from the literature for Langmuir
isotherm coefficients (Q,b) or Freundlich isotherm preexponential factors (k) to
estimate the importance of adsorption.11 The equations were derived from the
appropriate isotherm equations and organic soil content correlations. We obtained
experimental Q, b, and k values from the literature on activated carbon adsorption.
For chloroform, although there are many plausible degradation mechanisms in
natural aquatic systems, little specific information exists on any of them. The
reactivity of chloroform is, therefore, difficult to estimate. Scientific evidence
1 "Volatilization half-lives are calculated from the following equation:
t^ = (5.4) • d • m* • (3.5 + 8.9xl02S°/P°m)
where t^ is in seconds, d is a depth (assumed to be 1 cm for these calculations), m
is in grams, P° is in atmospheres, and S° is in g/cm3 .
lxAn adsorption coefficient (Y) was calculated from Langmuir coefficients or
Freundlich isotherms, using the following equations:
Y = 1/2 log (3/4 Qb (x + 1/3)) - 1
or
Y = 1/2 log (3/4 k (x + 1/3)) - 1
where x is the organic content of the sediment (x is assumed to be 0.1 for surface-
water sediment and 0.02 for ground-water gravel).
ICF Incorporated
-------
3-24
supports hydrolysis as the most significant chemical reaction in aquatic
environments. The hydrolytic half-life of chloroform has been estimated variously
between one and several thousand years. Under typical ambient fresh-water
conditions, however, the half-life of chloroform is sufficiently long to assume
ultimate degradation only after transport from the aquatic environment to the
atmosphere. Thus, volatility is the controlling mechanism and was used to score
chloroform. The half-life of chloroform is 22 minutes, which yields an exposure
score of 0 (see Exhibit 3-3).
ground water
We assumed that only two attenuation mechanisms are important in ground water:
adsorption and hydrolysis. We treated both factors as described in the section on
surface water. Adsorption of contaminants onto soils, clays, and other materials
can, however, significantly attenuate the migration of contaminants. Thus, we
modify ground-water exposure scores in a manner analogous to those for surface
water, except we compute the score as 7 - Y (see footnote 11 for computations).
Where hydrolysis is unimportant, the contaminant's half-life is effectively
infinite, which we score as 7. Where hydrolysis is important, we calculate the
ground-water exposure score from its half-life (see Exhibit 3-3). 12
There are few data regarding the half-life of chloroform in ground water.
Estimates of half-life range from 4 to 4,000 years or more at lower pH. We have,
therefore, assumed hydrolysis to be unimportant for chloroform. Review of the
literature also suggests that adsorption of chloroform is not an important
mechanism. Thus, we score chloroform as 7.
data are primarily from secondary sources
We determined the exposure score for each substance primarily by reviewing
secondary sources, including EPA's compilations of data on fate and transport such
as the "Aquatic Fate of Priority Pollutants" and "Treatability Manual"; the OHMTADS
data base; and compilations of physicochemical properties in numerous texts.
Clement Associates had primary sources for almost one-third of the chemicals. We
also conducted computerized searches of the literature where appropriate.
The secondary sources mentioned above were generally based on critical,
scientific review. They are usually reliable sources of information on
environmental properties of chemicals, at least for the purposes of our scoring
exercise. After identifying from among the secondary sources the studies that
indicated the half-lives of compounds, we briefly reviewed the primary source in
each case. If there were questions concerning the validity of evidence for a
particular compound as reported in the secondary sources, we reviewed several
primary sources. Thus, we determined the exposure score on the basis of the most
supportable evidence.
I2Ground-water exposure score = t^(hyd)
- 4.3
ICF Incorporated
-------
3-25
complex chemistry and missing data are problems
As might be expected, we encountered a number of problems in scoring for
exposure. For certain classes of compounds, particularly heavy metals, a complex
mode of action and/or complex chemistry made scoring difficult. In these
instances, we prepared a detailed exposure profile before assigning a score, and we
scored the most persistent form to identify the upper bound. We then modified
these scores to correspond more closely to the middle range of exposure of
chemicals in the group (see the section on uncertainty scores, below).
For a limited number of compounds or classes of compounds, data for evaluating
exposure were insufficient or absent. In these cases, scores were assigned on the
basis of analogies with compounds of similar structure for which data were
available. Polynuclear aromatic hydrocarbons (PAHs), for example, for which
exposure data were not available, were scored as a group on the basis of the
evidence that PAHs do not occur naturally in the pure state but in complex mixtures
with many other PAHs.
Data Base Includes Estimates of Uncertainty
As explained above, the scoring system yields two independent numerical scores:
an estimate of the likely value of the risk and the range of uncertainty in this
estimate. The uncertainty score is intended to reflect incomplete knowledge,
variability in the behavior of the chemical (such as differences in inherent hazard
or in environmental behavior of different chemical forms), variability in the
environment, and so forth. Since uncertainty includes some components that are
primarily subjective, we made no attempt to establish rules for assigning
uncertainty scores. Our procedure was, instead, to assume a common value for the
overall uncertainty introduced by the many assumptions and simplifications of the
scoring system, including the exposure models. We assign a different score only
when specific properties of the chemical or the data base indicate clearly that the
range of uncertainty varies from the common value.
We assigned an uncertainty score of 1 for all chemicals for inherent hazard
unless there were specific reasons to assign a different score. This is equivalent
to the assumption that estimates of inherent hazard may generally be in error by a
factor of 10 (up or down). We assigned different scores in the following cases:
• For most heavy metals, we assigned a high score (usually
1.5, but 2 in the case of beryllium, nickel, and vanadium)
to reflect the wide variability in the inherent hazard of
different chemical and physical forms.
• For lead, we assigned a score of 0.5 to reflect the fact
that the inherent hazard of lead in humans is unusually
precisely defined and apparently varies little with the
chemical form.
• We assigned a score of 1.5 to those PAHs for which
specific data on inherent hazard were scanty or lacking and
for which inherent hazard scores were assigned by analogy.
ICF Incorporated
-------
3-26
• We assigned a score of 1.5 to PCBs, which are mixtures
and for which the inherent hazard depends strongly on the
concentrations of hazardous contaminants.
• We assigned a score of 1.5 to a few other compounds
(cyanides, 2,4-dimethylphenol), for which data on chronic
inherent hazard are lacking.
In cases where the uncertainty score is 1.5 or 2, the inherent hazard score was
modified downward by 0.5 or 1 unit, respectively, to correspond to the middle of
the likely range of inherent hazard.
The same general procedure outlined above was used to assign uncertainty scores
for exposure. We assigned a score of 1 except in the cases below:
• We assigned scores of 1.5 or 2 to several heavy
metals for which the environmental behavior is known
to vary in a major way according to the chemical form.
• We assigned a score of 1.5 to a number of chemicals
for which the lifetime in air was subject to
substantial uncertainty, either because data were
lacking or were conflicting.
• We assigned a score of 1.5 to a few other chemicals
in cases where exposure scores were substantially
dependent on analogies.
To combine uncertainty scores for inherent hazard and exposure, we assumed that
the probability distribution of each was lognormal and that the variances would be
additive. Hence, if the uncertainty scores for inherent hazard and exposure are
ut, and us> respectively, the overall uncertainty score U is:
[(ut)2 + (us)2]4
The uncertainty scores for the 140 selected chemicals are presented in Exhibit
3-4. Because we were conservative in assigning scores other than 1 except in well-
defined cases, the overall uncertainty score in the majority of cases was 1.4. The
higher scores reflect, however, a significantly greater range of uncertainty,
particularly those for heavy metals, whose scores generally range from 2.1 to 2.5.
Case Studies Calibrate the Scoring System
As we discussed above, we designed the scoring system to provide relative
measures of the risks presented by W-E-Ts. In later stages in the analysis (for
example, in considering whether specific levels of technology would reduce risks to
acceptable levels), we need to estimate absolute levels of risk to populations.
Although the exposure model developed in Appendix 7 matches the risk scores for
facilities that release the same chemical into different media, it was not intended
to be the basis for precise risk assessments (but could be adapted for this
purpose). Hence, we chose to calibrate the scoring system by using case studies in
which population exposures or population risks had been related to measured release
rates. The case study approach provided estimates of the dispersion of chemicals
under actual conditions and therefore appeared substantially better for this
purpose than attempting to apply generalized, theoretical models of dispersion.
ICF Incorporated
-------
3-27
EXHIBIT 3-4
UNCERTAINTY SCORES FOR 140 SUBSTANCES
Exposure Adjustments
Adjusted Inherent
Hazard by Medium
Compound
Inherent Air Surface Ground- Air
Hazard Water water
A B C D E
Surface
Water
F
Ground-
water
G
sj A2+B2 /a2+C2
-/ A2+D2
Acenaphthene
1.5 1 1 1 1.8
1.8
1.8
Acenaphthylene
1.5 1 1 1 1.8
1.8
1.8
Acetaldehyde
1 1 1 1 1.4
1.4
1.4
Acetonitrile
1 1 1 1 1.4
1.4
1.4
Acrolein
1 1 1 1 1.4
1.4
1.4
Acrylamide
1 1 1 1 1.4
1.4
1.4
Acrylonitrile
1 1 1 1 1.4
1.4
1.4
4-Aminobiphenyl
1 1 1 1 1.4
1.4
1.4
Anthracene
1.5 1 1 1 1.8
1.8
1.8
Antimony
1.5 1 1 1 1.8
1.8
1.8
Arsenic
1.5 1 1 1 1.8
1.8
1.8
Asbestos
1 1 1 1 1.4
1.4
1.4
a-BHC
1 1 1 1 1.4
1.4
1.4
3-bhc
1 1 1 1 1.4
1.4
1.4
Y-BHC
1 1 1 1 1.4
1.4
1.4
Barium
1 1 1 1 1.4
1.4
1.4
Benzene
1 1 1 1 1.4
1.4
1.4
Benzidine
1 1 1 1 1.4
1.4
1.4
ICF Incorporated
-------
3-28
EXHIBIT 3-4 (cont.)
UNCERTAINTY SCORES FOR 140
SUBSTANCES
Exposure Adjustments
Adjusted Inherent
Hazard by Medium
Inherent Air Surface Ground-
Compound Hazard Water water
A BCD
Air
E
Surface
Water
F
Ground-
water
G
\A2+B'
/a2+c2
Jp?+D2
Benzo(a)anthracene 1 111
1.4
1.4
1.4
Benzo(k)fluoranthene 1.5 1 1 1
1.8
1.8
1.8
3,4-Benzofluoranthene 1 111
1.4
1.4
1.4
Benzo(g,h,i)perylene 1.5 111
1.8
1.8
1.8
Benzo(a)pyrene 1 111
1.4
1.4
1.4
Beryllium 2 111
2.2
2.2
2.2
Bis(2-chloroethoxy)- 1 111
methane
1.4
1.4
1.4
Bis(chloromethyl)ether 1 111
1.4
1.4
1.4
Bis(2-ethylhexyl)- 1 111
phthalate
1.4
1.4
1.4
Butane 1 111
1.4
1.4
1.4
Butene/Butadiene 1 111
1.4
1.4
1.4
Cadmium 1 1 1.5 1.5
1.4
1.8
1.8
Carbon disulfide 1 1.5 1 1
1.8
1.4
1.4
Carbon tetrachloride 1 111
1.4
1.4
1.4
Chlordane 1 1.5 1 1
1.8
1.4
1.4
Chloroacetaldehyde 1 111
1.4
1.4
1.4
4-Chloroaniline 1 111
1.4
1.4
1.4
Chlorobenzene 1 111
1.4
1.4
1.4
IC F Incorporated
-------
3-29
EXHIBIT 3-4 (cont.)
UNCERTAINTY SCORES FOR 140 SUBSTANCES
Adjusted Inherent
Exposure Adjustments
Hazard by Medium
Inherent
Air Surface Ground
Air Surface
Ground
Compound Hazard
Water Water
Water
Water
A
BCD
E F
G
/a2+b2 /a2+c2
^a2+d2
Chloroform 1
111
1.4 1.4
1.4
4-Chloro-2-methylphenol 1
111
1.4 1.4
1.4
2-Chlorophenol 1
111
1.4 1.4
1.4
Chromium (VI) 1
12 2
1.4 2.2
2.2
Chrysene 1
111
1.4 1.4
1.4
Copper 1.5
111
1.8 1.8
1.8
Creosote 1
111
1.4 1.4
1.4
Cyanide 1.5
111
1.8 1.8
1.8
Cyclohexane 1
111
1.4 1.4
1.4
Dibenzo(a,h)anthracene 1
111
1.4 1.4
1.4
Di-n-butyl phthalate 1
111
1.4 1.4
1.4
1,2-Dichlorobenzene 1
111
1.4 1.4
1.4
1,3-Dichlorobenzene 1
111
1.4 1.4
1.4
1,4-Dichlorobenzene 1
111
1.4 1.4
1.4
3,3'-Dichlorobenzidine 1
111
1.4 1.4
1.4
1,2-Dichloroethane 1
111
1.4 1.4
1.4
1,1-Dichloroethene 1
111
1.4 1.4
1.4
Dichloromethane 1
111
1.4 1.4
1.4
2,4-Dichlorophenol 1
111
1.4 1.4
1.4
ICF
Incorporated
-------
3-30
EXHIBIT 3-4 (cont.)
UNCERTAINTY SCORES FOR 140 SUBSTANCES
Adjusted Inherent
Exposure Adjustments Hazard by Medium
Inherent
Compound Hazard
A
Air
B
Surface
Water
C
Ground
Water
D
Air
E
Surface
Water
F
Ground
Water
G
/a2+b2
n/a2+C2
Ja2+d2
2,4-Dichlorophenoxy- 1
acetic acid
1
1
1
1.4
1.4
1.4
1,2-Dichloropropane 1
1
1
1
1.4
1.4
1.4
1,3-Dichloropropene 1
1
1
1
1.4
1.4
1.4
0,0-Diethyl S- 1
(2-(ethylthio)ethyl)
ester of phosphoro-
thioic acid
1.5
1
1
1.8
1.4
1.4
O/O-Diethyl phosphoric 1
acid O-p-nitrophenyl
ester
1.5
1
1
1.8
1.4
1.4
1,2-Dimethyl hydrazine 1
1
1
1
1.4
1.4
1.4
2,4-Dimethyl phenol 1
1
1
1
1.4
1.4
1.4
Dimethyl phthalate 1
1
1
1
1.4
1.4
1.4
Dimethyl sulfate 1
1
1
1
1.4
1.4
1.4
4,6-Dinitro-2- 1.5
methylphenol
1
1
1
1.8
1.4
1.4
2,4-Dinitrophenol 1
1
1
1
1.4
1.4
1.4
2,4-Dinitrotoluene 1
1
1.5
1.5
1.4
1.8
1.8
2,6-Dinitrotoluene 1
1
1.5
1.5
1.4
1.8
1.8
1,4-Dioxane 1
1
1
1
1.4
1.4
1.4
Dodecane 1
1
1
1
1.4
1.4
1.4
Endrin 1
1
1
1
1.4
1.4
1.4
ICF Incorporated
-------
3-31
EXHIBIT 3-4 (cont.)
UNCERTAINTY SCORES FOR 140 SUBSTANCES
Exposure Adjustments
Adjusted Inherent
Hazard by Medium
Compound
Inherent
Hazard
A
Air
B
Surface
Water
C
Ground
Water
D
Air
-------
3-32
EXHIBIT 3-4 (cont.)
UNCERTAINTY SCORES FOR 140 SUBSTANCES
Exposure Adjustments
Adjusted Inherent
Hazard by Medium
Inherent
Compound Hazard
A
Air Surface Ground
Water Water
BCD
Air
E
Surface
Water
F
Ground
Water
G
J a2+b2 /a2+c2
/a2+d2
Mercury
1.5
112
1.8
1.8
2.5
Methomy1
1
111
1.4
1.4
1.4
Methyl bromide
1
111
1.4
1.4
1.4
Methyl chloride
1
111
1.4
1.4
1.4
Methyl ethyl ketone
1
111
1.4
1.4
1.4
Methyl hydrazine
1
111
1.4
1.4
1.4
Methyl parathion
1
111
1.4
1.4
1.4
Naphthalene
1
111
1.4
1.4
1.4
2-Naphthylamine
1
111
1.4
1.4
1.4
Nickel
1.5 1.5 1.5
2.5
2.5
2.5
Nitrobenzene
1
111
1.4
1.4
1.4
Nitrogen oxides
1
1.5 N/A N/A
1.8
N/A
N/A
2-Nitrophenol
1
1.5 1 1
1.8
1.4
1.4
4-Nitrophenol
1
1.5 1 1
1.8
1.4
1.4
N-Nitrosodimethylamine
1
1.5 1 1.5
1.8
1.4
1.8
N-Nitrosodiphenylamine
1
1.5 1 1
1.8
1.4
1.4
Octachlorostyrene
1.5
111
1.8
1.8
1.8
PCB-1016
1.5
111
1.8
1.8
1.8
PCB-1221
1.5
111
1.8
1.8
1.8
ICF Incorporated
-------
3-33
EXHIBIT 3-4 (cont.)
UNCERTAINTY SCORES FOR 140 SUBSTANCES
Exposure Adjustments
Adjusted Inherent
Hazard by Medium
Compound
Inherent
Hazard
A
Air Surface Ground
Water Water
BCD
Air
E
Surface
Water
F
Ground
Water
G
/a2+b2 /a2+c2
/a2+d2
PCB-1232
1.5
111
1.8
1.8
1.8
PCB-1242
1.5
111
1.8
1.8
1.8
PCB-1248
1.5
111
1.8
1.8
1.8
PCB-1254
1.5
111
1.8
1.8
1.8
PCB-1260
1.5
111
1.8
1.8
1.8
Parathion
1
1.5 1 1
1.8
1.4
1.4
Pentachlorophenol
1
111
1.4
1.4
1.4
Phenanthrene
1.5
111
1.8
1.8
1.8
Phenol
1
111
1.4
1.4
1.4
Pyrene
1.5
111
1.8
1.8
1.8
Pyridine
1
111
1.4
1.4
1.4
Selenium
1.5
1.5 1.5 1.5
2.1
2.1
2.1
Silver
1
111
1.4
1.4
1.4
Sodium
1
111
1.4
1.4
1.4
Sulfur dioxide
1
1 N/A N/A
1.4
N/A
N/A
2,3,7,8-Tetrachloro-
dibenzo-p-dioxin
1
111
1.4
1.4
1.4
1,1,2,2-Tetrachloro-
ethane
1
111
1.4
1.4
1.4
Tetrachloroethene
1
111
1.4
1.4
1.4
ICF Incorporated
-------
3-34
EXHIBIT 3-4 (cont.)
UNCERTAINTY SCORES FOR 140 SUBSTANCES
Exposure Adjustments
Adjusted Inherent
Hazard by Medium
Inherent
Compound Hazard
A
Air Surface Ground
Water Water
BCD
Air
E
Surface
Water
F
Ground
Water
G
7a2+b2 /a2*:2
/a2+d2
Thallium 1.5
111
1.8
1.8
1.8
Toluene 1
111
1.4
1.4
1.4
Toluene diamine 1
111
1.4
1.4
1.4
Toluene diisocyanate 1
111
1.4
1.4
1.4
o-Toluidine hydro- 1
chloride
111
1.4
1.4
1.4
Toxaphene 1
1.5 1 1
1.8
1.4
1.4
1,2,4-Trichlorobenzene 1
111
1.4
1.4
1.4
1,1,1-Trichloroethane 1
111
1.4
1.4
1.4
1,1,2-Trichloroethane 1
111
1.4
1.4
1.4
Trichloroethene 1
111
1.4
1.4
1.4
2,4,6-Trichlorophenol 1
111
1.4
1.4
1.4
Vanadium 2
111
2.2
2.2
2.2
Vinyl chloride 1
111
1.4
1.4
1.4
Zinc 1
111
1.4
1.4
1.4
ICF Incorporated
-------
3-35
four case studies are used
We looked for cases in which release rates were known, population exposures had
been measured or could be calculated, and dispersion of the chemical was well
characterized. We discovered that few cases have been studied in sufficient detail
to serve as the basis for a numerical calculation. Although many cases have been
described in which chemicals released from underground storage have contaminated
ground water, we found none in which both the quantity released and the steady
state concentration in the water had been measured. The best cases involving each
medium are:
• release of cadmium from a point source into air;
• release of chloroform into the Mississippi River;
• release of PCBs into the Hudson River;
• leaching of sodium chloride (road salt) into ground water.
We recognize that none of these cases corresponds exactly to the simple dispersion
characterized by our exposure models. Each is useful, however, in matching the
exposure model to an actual situation and hence in calibrating the exposure scoring
system.
For each of the four case studies, we were able to obtain numerical estimates
of the release .rate and either the magnitude of population exposure or of
population risk. By combining these estimates with our inherent hazard scores, we
were able to calculate both the risk score in our system and an independent
estimate of the magnitude of the population risk. For comparison, we scaled the
results from each case study to yield an estimate of the risk score that would
correspond to an expected risk of one case per year (that is, one person affected by
the chemical).
The four independent estimates of an expected risk of one case per year varied
between 9.1 and 11.3. The relatively close agreement between these estimates
confirmed that our exposure models had succeeded in matching the scoring systems
for all of the media. To develop the risk, technology, and cost analysis further,
we adopted the mean value of 10 and adjusted risk scores to this value as our best
estimate of the score corresponding to a risk of one person expected to be affected
per year.
MODEL USES 83 WASTE STREAMS
In order to investigate the appropriate level of control of various hazardous
wastes, it is necessary to characterize a manageable number of waste streams, a
process requiring a considerable amount of approximation and simplification. We
believe, however, that this process will achieve two major objectives:
• to define relative risks and costs associated with
alternate treatment and disposal options for
representative hazardous waste streams;
• to establish how the volume of hazardous waste can
be matched to existing treatment and disposal
capacity.
ICF Incorporated
-------
3-36
Waste Volumes Are Based on Published Data
To achieve these objectives, we had to characterize the wastes in terms of
their constituents of concern and to determine the quantity generated. Our major
source in assessing- quantities generated was a recent study carried out for EPA
(Hazardous Waste Generation and Commercial Hazardous Waste Management Capacity: An
Assessment, December 1980). Putnam, Hayes & Bartlett, Inc. (PHB) undertook the
waste characterization component of this study. As the most recent available
source of data, it was used as the basis for correlation with our waste
characterization process. It should be noted that, although the PHB data are
considered to be a "nearly comprehensive estimate of industrial waste generation,"
a number of other known sources are excluded from their estimates:
Hazardous wastes from federal and other government facilities;
Discarded products, off-specification products, and containers;
Hazardous wastes from spills and abandoned sites;
-- PCBs;
State-designated hazardous wastes;
Industry-perceived hazardous wastes;
Mining wastes;
Wastes from small-quantity generators.
With the exception of PCBs (estimated at 15,000 to 35,000 wet metric tons per
year), we use the same exclusions in this report.
PHB's data indicated that 82 percent of the hazardous wastes are generated by
the industrial sectors shown in the following exhibit.
EXHIBIT 3-5
PUTNAM, HAYES & BARTLETT ESTIMATES:
INDUSTRIAL WASTE GENERATION BY STANDARD INDUSTRIAL CODE
i i i i
SIC Number Industry Percent of Total
28 Chemical 62%
33 Primary metals 10
29 Petroleum and coal products 5
34 Fabricated metals 5
Detailed examination of these sectors yielded limited data and only a few waste
streams. Thus, to provide representative coverage, we expanded our approach to
include all industrial waste streams as well as several non-industrial ones (PCB
wastes, dry cleaning solvents, and asbestos wastes). We collected waste
characterization data from a variety of sources (see References) which we used to
develop waste stream profiles.
We developed and applied a standard format for each waste profile:
• SIC number;
• EPA number (where applicable);
• quantity generated;
• number of generators;
• quantitative analysis of hazardous constituents;
• source of data. ICF Incorporated
-------
3-37
One limitation was the paucity of analytical data, which necessitated varying
degrees of approximation in order that the waste characterization process be
representative of the universe of hazardous wastes.
There is considerable uncertainty concerning the total amount of hazardous
waste generated per year. Some studies have suggested 23 million metric tons. In
order to minimize conflicts in the data, we have used PHB's estimates whenever
possible. The quantity represented by our initial 72 waste streams is about one-
quarter of PHB1s estimate of 41 million metric tons.
Wastes Are Grouped by Characteristics
The next steps in the process of characterizing wastes were to group the wastes
that had not been characterized with the 72 waste streams and to consolidate the 72
waste streams wherever common characteristics made this feasible. The latter task
proved considerably easier than the former and resulted in 63 waste streams,
grouped on the basis of the following characteristics:
• source of generation (SIC);
• mass flux (quantity generated per facility);
• constituent;
• concentration of constituent;
• physical state (for example, solid or sludge).
We chose these characteristics to reflect differences in waste streams in terms of
appropriate treatment and disposal technologies, industrial sector, and risk to
human health.
Assigning the additional wastes to the 63 waste streams proved extremely
difficult, largely because of the lack of data. We decided, in fact, that grouping
these wastes with the 63 waste streams would merely degrade the precision with
which the latter were defined. Thus, we adopted, instead, an alternative approach
that defined further waste categories by using highly approximate, artificial
characteristics. We assessed these characteristics by reviewing the literature,
and we estimated volumes and concentrations for waste streams by using best
engineering judgment. We were, therefore, able to add additional waste streams and
to characterize them in general terms such as "heavy metal sludge, medium
concentration, medium mass flux, not otherwise specified."
We define hazardous wastes generally on the basis of the guidelines set forth
under the Resource Conservation and Recovery Act (RCRA). We have, however, made
the following exceptions:
• Certain wastes that are not regulated under RCRA, but
that are of environmental concern, specifically PCBs and
asbestos, have been included in the characterization
process because they compete for hazardous waste disposal
capacity in the same way as RCRA-regulated wastes.
ICF Incorporated
-------
3-38
• Wastes that are recycled in an industrial process have
not been included in the characterization process unless
they have been specifically identified in a RCRA Background
Document as causing environmental concern because of their
storage (for example, certain wastewater treatment sludges
from nonferrous metal smelting). These wastes are
identified in Exhibit 3-6.
• Wastes that are hazardous solely because of their
ignitability, corrosivity, or reactivity are not included
in the waste characterization process. We assigned
approximately 12 million metric tons of hazardous waste of
this type to a category that is broadly defined as
"nontoxic hazardous waste."
Data Are Poor
Quality of the data on waste characteristics varies greatly, but is generally
poor. In some cases incomplete data are available for selected constituents or on
the basis of mass of a given constituent per mass of product. Such partial data
necessitate approximation to arrive at an overall composition. In other cases data
from a particular industry were particularly poor, forcing either the elimination
of that industry (for example, textiles and pulp and paper) from the classification
or the development of a highly approximate classification system such as for
pesticides. Some analytical data may be imprecise; thus, wastes containing
chromium are commonly characterized in terms of total chromium rather than
hexavalent chromium, which is the principal constituent of concern. Such
situations call for scientific judgment to determine the redox potential likely to
prevail in the waste stream and, correspondingly, the predominant oxidation state.
In many instances conflicts exist between different sources of information. The
results of several studies on national hazardous waste generation conducted in the
last A years have varied widely. PHB, for example, reports a total of about 41
million metric tons per year, in contrast to the 23 million metric tons reported by
A. D. Little in 1979. Our approach has been to use the most recent data, which
means that some of our data have been gathered from different sources and may yield
inconsistencies.
Data Base Includes Waste Streams
Representing 23 Million Metric Tons
Even with the limitations on the data, we developed 83 waste streams. A
detailed characterization for each is provided in Appendix 6. This information,
along with the source of the data, constitutes the input to the computer model. A
summary of this information appears in Exhibit 3-6, where we identify waste streams
by the principal constituents of concern or the source of the wastes and the EPA
number where applicable. We also include the SIC number for those wastes generated
from a predominant source, along with annual quantities, and we list concentration
and mass flux (kilograms of waste per day per facility). For the nonspecific
wastes, we selected an indicator constituent, and, as inputs to the computer model,
we estimated concentration and mass flux values using professional judgment.
ICF Incorporated
-------
3-39
Because the quality of the data varies, we also provide an indication of the
reliability of the information, The principal areas of uncertainty were mass flux
and concentration of constituent. In many cases we calculated these parameters by
approximation from sources that themselves had used approximations. The
uncertainty scores also appear in Exhibit 3-6.
We define three levels of uncertainty in terms of mass flux and concentration
of constituent. An uncertainty factor of 10 indicates that the estimate of the
parameter may be in error by a factor of 10 (up or down).
For the generalized waste streams designated as "not otherwise specified," this
uncertainty scheme does not take into account the additional uncertainty associated
with use of an indicator constituent.
Uncertainty Uncertainty
Factor Score
Mass flux: the uncertainty score reflects
likely inaccuracy of the quantity of the waste
generated and the distribution by size of the
generating facilities.
1. Actual reported data on quantity of
wastes generated and number of facilities. 10 1
2. Calculated quantities of waste from partial
data with known number of facilities. 20-50 1.5
3. Crude estimate using minimal data. 100 2
Concentration of constituent: the uncertainty
score reflects variation in waste-generating
operations and the availability of data.
1. Actual reported data on concentrations 10 1
of constitutents.
2. Calculated concentrations from partial 20-50 1.5
data.
3. Crude estimates using minimal data. 100 2
Of the 83 waste streams, 63 are characterized relatively well and the remainder
are highly approximate. The 63 streams account for approximately 10 million metric
tons per year, or approximately one-half to one-quarter of the total hazardous
waste generation. The 20 approximate waste streams account for another 20 million
metric tons.
The total of 23 million metric tons represented by the classification scheme
includes 65 of the 97 waste streams listed in the Federal Register. Four of the 97
streams were excluded from the classification because they were listed for
characteristics other than toxicity, and the remainder were omitted owing to lack
of data.
ICF Incorporated
-------
3-40
It is not clear what effect the 12 million metric tons of "nontoxic hazardous
waste" and the remaining 6 million metric tons of unassigned hazarous waste will
have on the matching of waste generation with existing disposal capacity. We can,
however, make some general comments. Of the corrosive, ignitable, and reactive
wastes, a significant portion of the corrosives and reactives may be processed in
such a way as to render the wastes nonhazardous. A large part of the ignitable
wastes may be recovered or used as fuel supplements. These practices remove the
wastes from the disposal cycle. While we are unable to present any quantitative
assessment of such practices, we suggest that they may be sufficiently widespread
for the capacity-matching process to be valid.
ICF Incorporated
-------
EXHIBIT 3-6
WASTE STREAM CHARACTERISTICS
WASTE STREAM
EPA
No.
SIC
No.
1 .000
TONS/
YEAR
CONSTITUENTS OF CONCERN
CONCENTRATION
QUANTITY
(ppm) (uncertainty) (kg/day) (uncertainty)
AQUEOUS INORGANICS
Metal Sludges
1. Mercury sludges K071,
from chloralkali K106
process
2. Lead sludges from n/a
battery production
3. Chromium sludges K002
from pigment
product ion
4. Zinc sludges n/a
from textile industry
5. Arsenic sludge K084,
from production of K101,
veterinary K102
pha rmaceut i caIs
6. Mixed metal n/a
sludges from
pa int product ion
7. Mixed metal K086
sludges from
ink formulation
8. Mixed metal K065,
sludges from K066,
nonferrous metal K067
product iona
9. Mixed metal F006,
sludges from F019
electroplating
10. Biosludge n/a
containing heavy metals
28 12 281
3691 150
2851 30
2893 10
3332, 69
3333
3471 400
2911 266
Mercury
Lead
2812 27 Chromium VI
2823 160 Zinc
2833 1.5 Arsenic
Lead,
Mercury,
ThaI Ii um
Lead,
Chromium VI
Lead,
Cadm i um
Chrom i um VI,
N ickeI
Chromium VI,
Lead
2,700
35,000
180
70,000
80,000
11,000
15,000
970
760
150
25,000
130
370
40
3
3
3
3
42,800
2,000
6,700
17,900
500
55
60
3
3
J 26,000
^ 110
J 3,500
3
i
3
3
3
-------
EXHIBIT 3-6 (cont.)
WASTE STREAM CHARACTERISTICS
WASTE STREAM
1 .000
EPA SIC TONS/
No. No. YEAR
CONSTITUENTS OF CONCERN
CONCENTRATION
QUANTITY
(pptn) (uncertainty) (kg/day) (uncertainty)
11. Cooling water, n/a
sludges from
petroleum industry
12. Spent Ii me n/a
bo iler feed from
petroleum refining
13.-16.
Metal sludges n/a
-- high toxicity,
n. o. s.
17.-20.
Metals and n/a
m i see I Ianeous
inorganics - low to
moderate toxicity,
n. o. s.
29 I I
2911
Acid Solutions Containing Heayy Metals
2 I. Pickle I i quo r
from steel finishing
22. Acid solutions
conta i n i ng heavy
metaIs, n.o.s.
Cyanide Sludge
23. Lime sludge from
coking operations
_ 24.
O
-n
Z3
o
o
T
"O
o
-J
Q>
r-t-
(D
Q.
25.
Heat treatment
wastes
Cyan ide waste
from electro-
plating
K062
n/a
K060
F010,
F011,
F012
F007,
F008,
F009
3398
I I 2
869
868
307
307
307
116
116
116
116
710
1U0
3312 963
3312
3471
12
Lead,
Chromium VI
Lead,
Chromium VI
Lead (i nd ica tor)
Zinc ( ind icator)
Lead,
Chromium VI
Lead ( i nd i ca to r)
Cyan i de,
Arsen i c,
PhenoI
Cyan i de
Cyanide
60
1
.04
31
31
310
310
31
31
310
310
98
870
31
740
110
100
50,000
60,000
3
i
1.5
^ 9,300
38,900
3, 100
310
310
3, 100
3, 100
310
310
3, 100
Jj 8,100
3, 100
1,800
1
3
'3
I
N3
-------
EXHIBIT 3-6 (cont.)
WASTE STREAM CHARACTERISTICS
WASTE STREAM
EPA
No.
SIC
No.
1 .000
TONS/
YEAR
CONSTITUENTS OF CONCERN
CONCENTRATION
QUANTITY
(ppm) (uncertainty) (kg/day) (uncertainty)
Mineral Wastes
26. Asbestos sep-
arator wastes
from the diaphragm cell
in chlorine production
AQUEOUS ORGAN ICS
Phenols
27.
28.
29.
30.
31,
32.
O
- 33.
o
o
T
¦U
o
Phenolie siudge
from plastics
p roduct i on
PhenoI -fo rmaIdehyde
solution from
plastics production
Wastewater treat-
ment sludge from
wood preserving
n/a
n/a
n/a
K001
2812 30
2821 53
2821 486
2491 40
Wastewater from Organic Chemical Production
D i st iI I a t i on
residues from
acetaIdehyde
product ion
St iI I bottoms
from acryI onitriIe
product ion
Agueous spent
antimony catalyst
from fluoromethane
product ion
Wastewater from
n i trobenzene**
aniline production
K009,
K010
K01 1,
K01 3,
K014
K021
K083,
K103,
K104
2869 100
2869 3,175
2869
.02
2869 150
Asbestos
PhenoI,
Forma Idehyde
PhenoI
Benzo(a Jpyrene,
Chrysene,
PentachIo rophenoI
ChloroacetaIdehyde,
Forma Idehyde,
Chloroform,
AcetaIdehyde
Ac ryI on i triIe,
Cyan ide,
Aceton i t r iIe
Carbon tetra-
chIo r i de.
Chloroform,
Ant i mony
Benzene,
N i trobenzene
84,000
50,000
20,000
50,000
.03
.5
30
6,600
600
600
9,600
400
6,000
3,000
4,000
4,000
1,000
75
30
1.5
2,200
3' 1
1,600
200
3' 3
300
538,800
145,000
27
1
51,400
3
32
3'
b
3"
W
•O
t-O
<0
Q.
-------
EXHIBIT 3-6 (cont.)
WASTE STREAM CHARACTERISTICS
WASTE STREAM
EPA
No.
SIC
No.
1 .000
TONS/
YEAR
CONSTITUENTS OF CONCERN
CONCENTRATION
QUANTITY
(ppm) (uncertainty) (kg/day) (uncertainty)
34. Wastewater from
chIo robenzene
p roduct i on
K105 2869 39 Benzene,
ChIo robenzene,
1,2-D i chIo robenzene,
1,4-D ichIo robenzene
Wastewater Treatment Sludge from Pesticide Production
n
3
O
O
-f
"O
o
-I
0)
r+
ID
Q.
35. Toxaphene waste
36. Pesticide wastes-
high toxicity,n.o.s.
37. Pesticide wastes-
moderate toxicity,n.o.s.
38.-41.
Aqueous organics,
n.o. s.
CONCENTRATED ORGANICS
Spent Solvents
42. TrichIoroethylene
43. 1,1,1-Tr i chIoro-
ethane
44. Methylene
chI or ide
45. Tetrachloro-
ethylene
46. Spent solvents
n. o.s.
K041,
K098
n/a
n/a
n/a
F001,
F002
F001,
F002
FOOT
F002
F004,
F005
2879 5
2879 500
2879 500
* 4,000
1,000
1,000
1,000
18
16
3
67
15
721
Toxaphene
Pa ra th i on
( ind icator)
ChIo rdane
( i nd i ca to r)
1,2-DichIoroe thane (i.e.,
Ethylene dichloride)
( i nd i ca to r)
ti
Tr i chIo roethene
(Tr i chIo roethy Iene)
1,1,1-Tri chIoro-
ethane
Dichloromethane
(Methylene chloride)
TetrachIoroethene
(TetrachIoroethy Iene)
To Iuene
(indicator)
1,800
80
5,000
5,000
10,000
30,000
30,000
310
31,000
310
310
800,000
800,000
800,000
600,000
800,000
•1.5
1.5
2
2
39,000
6,850
3, 100
3, 100
3, 100
310
310
3, 100
. 1
. 1
.15
10
20
1.5
2
2
1.5
1.5
1.5
1.5
2
-------
EXHIBIT 3-6 (cont.)
WASTE STREAM CHARACTERISTICS
WASTE STREAM
EPA
No.
SIC
No.
1 .000
TONS/
YEAR
CONSTITUENTS OF CONCERN
CONCENTRATION
QUANTITY
(ppm) (uncertainty) (kg/day) (uncertainty)
48.
50.
51.
52.
O
3
O
o
-J
•U
o
T
0>
i-h
n>
a
Still Bottoms From Solvent Recovery
47. TrichIoroethyIene
1,1,1-Tri chloro-
ethane
49. Methylene
chloride
Still bottoms
n.o.s.
Solvent/Metal Sludges
Pa int appIicat ion
sIudges
Organic/metaI Iic
sludges n.o.s.
F001,
F002
F001,
F002
F001
F004,
F005
n/a
n/a
18
15
2
15
29
73
Residues From Organic Chemical Production
K015 2869 12.5
53. Still bottoms from
benzyl chloride
d i st iI I a t i on
54. Heavy ends and
residues from carbon
tetrachloride production
K016
55. Heavy ends from
ep i chIorohyd rin
product ion
56. Heavy ends from
ethyl chloride
p roduct i on
KOI 7
KOI 8
2869
2869
2869
3.2
12.5
35
Tr i chIoroethene
(TrichIoroethyIene)
1,1,1-TrichIoro-
ethane
D i chIoroethane
(Methylene chloride)
Toluene (indicator)
Methyl ethyl ketone.
To Iuene,
Ch rom i um VI,
Mercury,
Lead
Chromium VI
(ind icator)
Chlorobenzene,
To Iuene
HexachIorobenzene,
HexachIorobutad i ene,
HexachIo rethane,
Tet rachIoroethene
(TetrachIoroethyIene)
Ep i chIorohyd r i n,
Bis (chIoromethyI
ether)
T ri chlorethene
(T r i chIo roethy Iene),
HexachIo robenzene,
HexachIorabutad i ene,
1,2-DichIoroethane
200,000
200,000
200,000
400,000
10,000
10,000
56
12
10
310
20,000
20,000
40,000
40,000
200,000
200,000
20,000
100,000
320,000
215,000
215,000
110,000
1.5
. 1
.15
20
500
7, 300
900
,600
-16,000
1.5
1.5
1.5
2
}
-P>
Ln
-------
EXHIBIT 3-6 (cont.)
WASTE STREAM CHARACTERISTICS
WASTE STREAM
EPA
No.
1 .000
SIC TONS/
NO. YEAR
CONSTITUENTS OF CONCERN
CONCENTRATION
QUANT ITY
(ppm) (uncertainty) (kg/day) (uncertainty)
57. Heavy ends from
ethyl dichloride
and vinyl chloride
product ion
58. Di st iI I at ion
residues from
phthaIic anhydride
product ion
59. St iI I bottoms
from nitrobenzene
product ion
60. St ri pp i ng still
ta iIs from methy I
ethyl pyridine
product i on
61. Organic residues
from toluene
diisoeyanate pro-
duct i on
62. St iII bottoms
f rom 1,1,1-
tr i chIo roethane
product i on
63. Distillation res-
idues from tri-
chIo roe thy Iene/
pe rchIo rethyIene
64. St. i I I bottoms
from chIorobenzene
product ion
65. D i st iI I a t i on
bottom tars from
production of
phenoI/acetone
KOI 9,
K020
K023,
K024,
K093,
K09U
K025
K026
K027
2869 135
2869
2869
2869
2869
1,1,1-TrichIoroethane, 220,000
1,1,2,2-Tetrachloroethane 80,000
7.8
Maleic anhydride
2,U-Dini trotoluene
0.7 Pyridine,
PhenoI
Toluene diisoeyanate.
Toluene diamine
K030 2869 15
K085 2869 7
K022 2869 72
HexachIorobenzene
PhenoI,
Benzo(a) anthracene,
Benzo(a) pyrene,
Chrysene
19,000
1,000
30,000
8
10
1
K095 2869 45 1,1,2,2-TetrachIoroethane 270,000
1,1,2,2-TetrachIoroethane, 200,000
HexachIorobenzene, 1,000
Hexachlorabutadiene 200,000
100,000
16,000
1,000
1,000
1,000
J' 3'
1.5
1.5
8,500
1,810
3, 100
3
,100
<41,600
]' 5
4, 100
2,700
¦18,100
3'
1.5
1.5
31
3-
-------
EXHIBIT 3-6 (cont.)
WASTE STREAM CHARACTERISTICS
WASTE STREAM
1 .000
EPA SIC TONS/
No. No. YEAR
CONSTITUENTS OF CONCERN
CONCENTRATION
QUANTITY
(dditiI (uncertainty) (kg/day) (uncertainty)
OTHER CONCENTRATED ORGAN ICS
66. Chlorinated hydro-
carbon waste from
chI oral kali process
K073
2812
6
67. Decanter tank tar
sludge from coking
K087
331
72
68. PCB fluids
n/a
#
4
69. Concentrated
organics n.o.s.
n/a
*
631
OILY WASTES
70.-73.
Heavy metaI
sludges from
petroleum re-
fining
K048
K049
K050
2911
2911
2911
198
48
2
K051
2911
316
74. MetaI fIuoride
sludge from
petroleum refining
n/a
2911
18
75. Spent clay from
oil re ref i n i ng
n/a
2992
69
76. Ac id ta r from o iI
re ref i ni ng
n/a
2992
89
77. Caustic sludge
n/a
2992
60
from re refining
Carbon tetrachloride, 108,000
Chloroform 737,000
Phenol 100,000
PCB-1254 (Aroclor-1254) 500,000
1,2-DichIoroethane 200,000
(Ethylene dichloride)
( indicator)
Lead, 300
Chromium VI 27
Lead, 290
Chromium VI 18
Lead, 78
Chromium VI 6
Lead, 3,100
Chromium VI 110
Nickel, 55
Lead, 7
Cyanide, 23
Fluoride 90
Lead, 40,000
Benzo(a) anthracene, 10,000
Benzoja) pyrene 10,000
Lead, 10,000
Benzo(a) anthracene, 10,000
Benzoja) pyrene 10,000
Lead, 20,000
Benzo(a) anthracene, 10,000
Benzo(a) pyrene 10,000
V 3
2,700
3,000
2
3,100
3
3
2,200
7,000
12,200
82,200
V
.3'
V
y
5
3
3
-------
EXHIBIT 3-6 (cont.)
WASTE STREAM CHARACTERISTICS
WASTE STREAM
EPA
No.
SIC
No.
1 .000
TONS/
YEAR
CONSTITUENTS OF CONCERN
CONCENTRATION
QUANTITY
(ppm) (uncertainty) (kq/davl (uncertainty)
SOLID RESIDUES
78. Scrap batteries n/a 3692 1.3
79. Arsenic salts K031 2879 17
from pesticide
product ion
80. Residues from the n/a 30 52
rubber industry
81. Emission control K061 331 337
dusts from i ron
and steel production
82. Emission control K069 3341 10
dust from secondary
Iead sme11 i ng**
83. Emission control n/a 33 1000
dusts, slags and
other residues, n.o.s.
Mercury,
Lead,
Ca d i urn
Arsen ic
Nickel,
Lead,
Ba r i urn.
Chromium VI .
Z i nc,
Vanad i urn
Lead,
Chromium VI
Lead,
Cadm i um.
Chromium VI
Cadmium (indicator)
O
3
O
o
-T
V
0
1
CD
r*
ID
a
10,000
10,000
10,000
63,000
100
100
1,000
200
10,000
150
1,400
4,500
120,000
900
150
30,000
V 3
100
15,000
700
17,200
400
3, 100
TOTAL 22,637
NOTES
[1] The following symbols or abbreviations are used in Exhibit 3-6:
o n/a = not applicable or waste not listed,
o n.o.s. = not otherwise specified
o *wastes generated by several industry types,
o **wastes may ultimately be recycled to process.
[2] Indicator parameters have been selected to characterize the 20 highly approximate waste streams as
pa renthes i s.
[3] Annual quantities generated per facility have been rounded to the nearest 100 -- actual calculated
compu te r mode I .
[4] 1,000 tons per year represents annual quantity generated in wet tons.
OJ
i
¦f-
Oo
indicated in
values are used in the
-------
CHAPTER 4
ENVIRONMENTS
In Chapter 3 we described our approach to adjusting inherent hazard scores for
exposure. As we noted, a number of site-specific factors affect our exposure
scores. This chapter discusses the adjustments we apply. Further background
concerning the exposure model itself is presented in Appendix 7.
ENVIRONMENTS CHANGE RISK
Our approach explicitly accounts for three factors in developing environmental
categories: potential population exposed, surface-water characterization
(hydrology), and ground-water contamination potential (hydrogeology). Each of
these factors has important implications for risk scoring; they may change total
exposure scores by tenfold, and sometimes more, depending upon the nature of the
environment.
We considered, but do not expressly use, other factors in developing
environmental categories: soil alkalinity or acidity; average precipitation;
average temperature; incidence of tornadoes, hurricanes, and thunderstorms; and
evaporation potential. The impact of such factors was examined in light of the
order-of-magnitude risk assessment methodology to be applied. We believe such
factors are less important than the broad environmental factors (population
density, hydrology, and hydrogeology) that we use explicitly. In Chapters 2 and 3
we considered such factors as meteorology and soil characteristics in developing
typical release rates and exposure scores.
We also considered the impact of environmental characteristics both on the cost
and risk of the technologies. One of the best examples of such an impact is the
higher cost of land in an urban area compared to a rural area. We did not, however,
find this type of impact to be significant when it was combined with other costs on
our technology cost scale, which distinguishes only between factors of two.
THIRTEEN ENVIRONMENTAL CATEGORIES ARE BASED ON
THREE MEDIA PLUS OCEANS
On the basis of differences in the potential population exposed, hydrology, and
hydrogeology, we defined 12 categories of environment. In addition, we include a
thirteenth category, deep ocean waters, because of its special nature. Exhibit 4-1
lists the 13 categories; the rest of this section explains the terms we used.
Air Medium Is Distinguished in Terms of
Three Levels of the Potential Population Exposed
We defined three ranges of populations:
• High (520 people/km2 and over)
• Medium (52-519 people/km2)
• Low (fewer than 52 people/km2)
ICF Incorporated
-------
4-2
EXHIBIT 4-1
ENVIRONMENT CATEGORIES
!
population density hydrology hydrogeology
(assimilation capacity) (contamination potential)
(1) High
Low
High
(2) High
Low
Low
(3) High
High
High
(4) High
High
Low
(5) Medium
Low
High
(6) Medium
Low
Low
(7) Medium
High
High
(8) Medium
High
Low
(9) Low
Low
High
(10) Low
Low
Low
(11) Low
High
High
(12) Low
High
Low
(13) Deep Ocean Waters
These ranges vary by approximately one order of magnitude. This implies that the
average difference in risk score between high and medium and between medium and low
densities will be one unit on the risk scale. Most areas of the country have
medium-level densities. The data base now includes population densities as a proxy
for the potential population exposed until better data are available. This variable
accounts for large differences in the potential number of people exposed, and it is
the best measure for defining differences in the air medium.
Model Uses Two Levels of Assimilative Capacity
for Surface Water (Hydrology)
We divided differences in surface water assimilative capacity (hydrology) into
two categories. Low assimilative capacity refers to areas where risk is relatively
high. Such risk is caused by discharge of wastes into relatively low-flow streams
(less than 3 x 108m3/day); location on larger streams where drinking-water intakes
are located within 30-3 years (approximately 6 hours) downstream at average flow;
or the possibility of frequent flooding, that is, location within a 100-year
floodplain.
High assimilative capacity describes locations in which risks are lower:
discharge to high-flow or high-volume surface waters (streams greater than
3xl08m3/day, estuaries greater than 108m3/day, and lakes greater than 3xl010m3).
ICF Incorporated
-------
4-3
Two Levels of Ground-water Contamination Potential
(Hydrogeology) Are Used
Characterizing hydrogeologic conditions is not simple. We have found no good
source of hydrogeologic information that covers the entire country. For selected
regions, good data exist on soil characteristics and depth to ground water, but the
extreme variability of these characteristics makes generalized mapping difficult.
We define low ground-water contamination potential (lower risk) locations to
include those over aquifers contaminated to 100 times the drinking-water standards.
Also included are locations with the following characteristics:
• low soil permeability (less than 10-6 cm/sec) and
moderate depth to ground-water saturation zone
(greater than 10m), or
• moderate soil permeability (less than 10-itcm/sec)
and large depth to ground-water saturation zone
(greater than 100m) .
We consider all other locations as high ground-water contamination potential areas.
RISK SCORES ADJUSTED FOR ENVIRONMENTAL CHARACTERISTICS
In Chapter 3, we explained how we assigned adjusted exposure scores to 140
chemicals for each release medium (air, surface water, and ground water). These
scores reflect, in most cases, the half-life of the substance in each medium. Where
appropriate, exposure scores also consider the tendency of the substance to
bioaccumulate and/or to be removed by conventional water treatment or such natural
processes as adsorption or washout.
We adjust the basic exposure scores for the environmental characteristics
discussed above as shown in Exhibits 4-2, 4-3, and 4-4. The values presented in the
exhibits represent orders-of-magnitude (base 10) adjustments and are, therefore,
simply added to the basic exposure scores by medium. Below we summarize our
approach to developing the exhibits; a more complete discussion of the exposure
model and environmental adjustments is presented in Appendix 7.
Modifications for Population Exposure
Other things being equal, risk is linearly proportional to the potential
population exposed. Hence, a factor is added to the score that reflects the
average population density (the model's approximation of potential exposure) for
the environmental category. This modification is itself modified according to the
half-life of the chemical. Since long-lived chemicals become widely dispersed in
the environment, the average density of the population exposed to them is
independent of the environment in which they are released. For this reason, we
make no modification for population density in the right-hand columns of the
exhibits. The scores that appear in the exhibits are derived from calculations
described in Appendix 7.
ICF Incorporated
-------
EXHIBIT 4-2
SCORING SYSTEM FOR EXPOSURE FROM AIR
EXPOSURE ADJUSTMENT SCORE (from Exhibit 3-2)
0 1 2 3 4 5 6+
a
FACTOR HALF-LIFE
b b~
3 min. 30 min. 6 hours 3 days 30 days 1 year 10 years
Modifications for Population Density
High (i 520 people/km^) +2
Medium 52 people/km^ and +1
< 520 people/km^)
Low (< 52 people/km^) -1
Modification for Atmospheric Behavior
If chemical is concentrated at +1
heights > 300 m
+1
+1
-1
+1
+1
0
-1
+1
-P-
i
aIncludes removal by atmospheric washout as well as degradation.
^Global atmospheric pollutants.
-------
EXHIBIT 4-3
SCORING SYSTEM FOR EXPOSURE FROM SURFACE WATER
EXPOSURE ADJUSTMENT SCORE (from Exhibit 3-2)
6+
FACTOR
30 min.
6 hours
HALF-LIFE
3 days
30 days
1 year
10 years
b,c
100 years
Modifications for Population Density
High ('Zr 520 people/km^) +1
Medium (2-52 people/km^ and -1
< 520 people/km^)
Low (< 52 people/km^) -2
Modification for Hydrography
Low assimilative capacity +1
High assimilative capacity -1
Discharge into ocean or estuary -2
+1
-1
-2
+1
-1
-2
+1
-1
-2
+1
-1
-2
+1
-1
-2
+1
-1
-2
+1
-1
-2
+1
-1
-2
i
aIncludes sedimentation and volatilization as well as degradation.
^Distribution outside North America.
cRisks may be discounted.
-------
EXHIBIT 4-4
SCORING SYSTEM FOR EXPOSURE FROM GROUND WATER
FACTOR
6 hours
3 days
EXPOSURE ADJUSTMENT SCORE (from Exhibit 3-2)
HALF-LIFE3
30 days
1 year
10 years
6+
100 years"1000 years6
Modifications for Population Density
High (1 520 people/km2)
Medium (^-52 people/km2 and
< 520 people/km2)
Low ( <. 52 people/km2)
Modification for Hydrology0
Low Contamination Potential
High Velocity (-20 km/yr)
Very Low Velocity (-20 m/yr)
High Contamination Potential
Low Velocity (-0.2 km/yr)
Very Low Velocity (-20 m/yr)
Modification for Vertical Extent of
Aquifer
If ground water aquifer
thickness > 300 m
+2
+1
-2
+2
+1
-2
-1
+2
+1
-2
0
-2
0
-2
-1
+2
+1
-2
+2
+1
-1
-1
+ 1
+ 1
-1
-1
0
+1
+1
-1
0
0
-1
-1
+1
+ 2
-1
4>
I
ON
O
3
O
o
-I
•U
o
-f
0>
rf
-------
4-7
Modifications for Hydrography
Exhibit 4-3 provides for variations in hydrographic factors that govern the
degree of dilution of the chemical after discharge. For discharge into fresh
waters (lakes and rivers) of low assimilative capacity (that is, low flow streams
of less than 3xl08m3/day, larger streams with drinking-water intakes located less
than approximately 6 hours downstream at average flow, or locations subject to
frequent flooding), exposure scores are adjusted upward by one unit because of the
greater likelihood of exposure. For discharge into estuaries and inshore coastal
waters, we assume further dilution by one or two orders of magnitude.
Modifications for Thickness of Dispersion Channel
The scoring rules for ground water are based on calculations for an aquifer
that is 30 meters thick (see Appendix 7). Exhibit 4-4 provides for modifying the
ground-water score if there is reason to assume a much thicker aquifer (that is
greater dilution).
ADJUSTMENTS REFLECT TENFOLD DIFFERENCES
The exposure adjustment values represent tenfold differences in the
probability of human exposure when a chemical is released in an environment. An
example best demonstrates the effect of adjusting exposure scores. Nickel has the
following exposure adjustment scores (from Exhibit 3-2): air, 4; surface water, 2;
ground water, 2. The air score is two orders of magnitude higher than the scores
for surface water and ground water. Thus, releases to air yield the highest risks,
all other things being equal. In some environments, the significance of persistent
releases to air diminishes, and the importance of other pathways increases.
Consider the adjustments for an environment with high population density, low
assimilative capacity, and high contamination potential.
air surface water ground water
Inherent hazard score 4 4 4
(from Exhibit 3-2)
Exposure adjustment score 4 2 2
(from Exhibit 3-2)
Environmental adjustments +0 +2 +2
(from Exhibits 4-2 to 4-4)
Total risk/unit release =8 =8 =8
In this example, the risks posed by the given release of a given quantity of
nickel in air, surface water, and ground water are roughly the same.
Site-Specific Issues Problems Must Be Examined
A number of site-specific factors determine whether a particular technology is
appropriate for a particular location. Land treatment of hazardous wastes, for
example, is not suitable in colder climates where the growing season is short, and
deep well injection is appropriate only where suitable deep geologic formations
exist. Just as volume-capacity considerations limit the use of incinerators and
landfills, these environmental constraints limit the use of land farming and deep
well injection. Exhibit 4-5 presents the environmental constraints that must be
considered when analyzing a strategy.
ICF Incorporated
-------
4-8
EXHIBIT 4-5
ENVIRONMENTAL CONSTRAINTS ON DISPOSAL TECHNOLOGIES
i 1
Disposal Technology Consideration
i i
Surface impoundment Not appropriate as ultimate disposal in
regions with low evaporation rates.
Land treatment Not appropriate in cold climates.
Deep well injection Not appropriate where geology unfavorable.
VOLUME OF WASTE AND CAPACITY OF FACILITIES MUST BE CONSIDERED
In order to determine the present limits on waste disposal, we must consider
the volume of waste generated in a particular environment and the capacity of the
disposal sites in a particular environment to handle that waste. We have touched
briefly on capacity in Chapter 2 and the volume of waste generated in Chapter 3. As
we noted, we limit the capacity of disposal technologies only for double-lined and
single-lined landfills and for 99.99% and 99.9% DRE incinerators. The missing
element is the manner in which we link the 13 environments that we use in the model
to these 2 sets of data. Below, we explain that link.
We derive 12 environments (excluding oceans) by combining the 3 basic
characteristics described earlier in this chapter. Each environment contains two
sets of data, one on the location of the generating facility and one on the location
of the waste management facility including zip code. The obvious link, then, is to
characterize each zip-code area in terms of the three basic characteristics. After
analyzing the zip-code patterns, we decided that a variant of the three-digit zip
code yields sufficient detail for the immediate purposes of the model.
We estimate the population density of each three-digit zip-code area using 1970
census data and apply the three parameters to sort out high, medium, and low
categories. By computing the ratio of the zip-code area to the area of the State in
which it appears, we estimated the area of the zip code.
We accounted for hydrologic factors by determining the proximity of the three-
digit zip-code areas to the coast, the Great Lakes, the lower Mississippi River,
and Chesapeake Bay. We simplified our determination for this iteration of the
model by assuming that, if a zip code borders on any of these geographic features
either completely or partially, it is an area of high assimilative capacity.
We accounted for hydrographic factors by examining the maps prepared for OSW by
E.A. Hickok, Inc. These maps indicate locations of aquifers and usable ground-
water areas. If a zip code overlaps these locations by one-third or more, we
designated it as an area of high contamination potential. We resolved questionable
situations conservatively by designating them as high contamination areas.
These estimates and the combination of three-digit zip-code areas (principally
in metropolitan areas) results in a list of 560 areas. Exhibit 4-6 shows the
distribution of total number of zip codes by basic characteristic. Exhibit 4-7
indicates their distribution among the 12 land environments.
ICF Incorporated
-------
4-9
EXHIBIT 4-6
DISTRIBUTION OF THREE-DIGIT ZIP CODES
BY THREE ENVIRONMENTAL CHARACTERISTICS
1
population density
1 1
assimilative
l
contamination
1
capacity
1 |
potential
1 i
category
i i
number
1 i 1
category number
i i i
i
category
i
l
number
i
low
117
low 442
low
296
medium
357
high 118
high
264
high
86
EXHIBIT 4-7
DISTRIBUTION OF THREE-DIGIT ZIP CODES
AMONG TWELVE LAND ENVIRONMENTS
assimilative
capacity
high
low
population
density
low
0
79
low
contamination
potential
5
33
high
med
19
150
low
49
139
high
high
19
29
low
26
12
high
ICF Incorporated
-------
CHAPTER 5
ALTERNATIVE STRATEGIES
In this chapter, we describe the purpose and components of a RCRA strategy in
the context of this project and provide an example derived from some of the test
runs of the model we have made. We conclude with a discussion of what additional
work must be done before a strategy developed by the model can be put into action.
This project develops a risk-cost policy model that produces sets of W-E-T
combinations on the basis of risk and cost trade-offs determined by EPA decision-
makers. The estimates of risk and cost that the model uses represent summaries of
the best available scientific and engineering information. The preceding
discussion in this document addresses the many steps required to generate the data
base on which trade-offs can be made.
For wastes, it is necessary to select chemical compounds; to determine a score
for the risk to human health; to modify the score to reflect exposure to humans
through three media: air, surface water and ground water; and to calibrate our
risk scores with real-life situations. We have selected 83 waste streams to
represent the universe of hazardous wastes.
\
For technologies, we have developed technology chains that describe the
processing of hazardous waste through treatment, transportation, and disposal. We
define costs for typical technologies and made engineering estimates of the release
rates. We compare the magnitude of the releases from the technologies with the
magnitude of the release on which we based the risk score to compute the risk for
the technology.
For environments, we define 13 separate types and adjust the risk scores to
reflect the characteristics of the environment. We differentiate between
environments using characteristics of the media: the potential population exposed
for air; the assimilative capacity of surface waters; and the contamination
potential of ground water.
The data we have developed up to this step represent the risks and costs of
managing specific units of waste, not their total effects. We assess total effects
by developing sets of W-E-T combinations that we refer to as strategies. A
strategy contains at least one W-E-T combination for each waste type. Thus, the
minimum number of W-E-T combinations in a strategy is 83, the same as the number of
waste streams now in the data base. Additional W-E-T combinations may be added if
waste is generated in different types of places or if there is limited capacity for
some technologies. A strategy need not include disposal in every environment or
use every technology under the current formulation of the model. For now, it is
conceivable that a strategy could dispose of all the waste in a single environment,
or use only one (or a few) technology chains.
ICF Incorporated
-------
5-2
The purpose of a strategy is to examine the effects that differing assumptions
about the levels of risk and cost and their trade-off would have on hazardous waste
disposal alternatives. For example, a strategy that identified single-lined
landfills and 99 percent DRE incinerators as meeting the constraints and goals of
that strategy would suggest that additional regulatory activity on double-lined
landfills and higher level DRE incinerators might not necessarily be useful for the
set of waste streams considered. Of course, for any strategy, a thorough
understanding of the basis on which the strategy is formulated is necessary. This
understanding will partly be gained through sensitivity analysis, where selected
risk and cost scores will be varied to reflect the uncertainty present in the score
and test the magnitude of those changes.
These are five steps to selecting the strategies that best represent the
decision-makers' desired trade-offs, each of which we discuss below:
1. Set Parameters
2. Choose Goals
3. Array Alternatives
4. Perform Sensitivity Analysis
5. Select Strategy
SET PARAMETERS
The development of a strategy depends on decisions made about a number of
parameters. These include determining the volume and location of waste stream
generation, estimating the capacity of potentially limited technologies, selecting
the number of treatment steps the model will investigate, and accounting for the
costs and risks of transporting hazardous wastes.
Some of the value of these parameters depend on data that may be improved with
additional research. Others reflect an informed but arbitrary choice. An example
of the latter case is the selection of up to three treatment steps for the current
model runs. We make this choice as a compromise between the increased cost of each
additional treatment step, the decreasing marginal reductions in risk, the addition
complexity of the technology chains the model would face, and our knowledge of
typical practices.
Additional research will be helpful in improving estimates on generation
volumes and disposal capacity. We use the best available sources but still find it
necessary to make a number of estimates, especially regarding the location and
capacity of disposal facilities by three-digit zip codes. Lacking detailed data by
zip code, we assume only double-lined and single-lined landfills have limited
capacity in the near term. We also assume treatment technologies could rapidly be
brought on stream except in two cases: 99.99 percent and 99.9 percent DRE
incinerators.
Constraining certain classes of landfills and incinerators requires including
transportation cost and risk from transporting the waste from the zip code of the
generator to a zip code with adequate capacity. Here, we assume typical trips of 25
miles within a zip code and 250 miles outside a zip code. The model fills the
capacity in one region with the most hazardous substances first and shifts less
risky substances to other regions or technologies if necessary.
ICF Incorporated
-------
5-3
Expanding the list of constrained technologies would require more detailed
information about the location and capacity of the technologies. Variations in the
transportation costs and risks would be introduced by redefining the 25/250 mile
trip parameters. Additional treatment steps could be added, but this very rapidly
expands the amount of possible combinations. Each treatment combination requires
careful engineering review regarding technical feasibility, which is somewhat
costly.
In any case, the model now uses the above parameters as being suitably
representative in order to let us proceed to the next step.
CHOOSE GOALS
The decision-maker has available many alternatives in choosing goals. Cost or
risk or a cost-risk combination may function as the objective of the analysis. The
usual approach would be to minimize either risk or cost. It is technically
impossible to minimize both cost and risk at the same time, without an explicit
weighting of each. Such a weighting requires a value judgment about the
appropriate relative values of risk and cost, and because of the inability to
prespecify those weights, we avoid using any combination of risk and cost as a
goal. We wish, instead, to eventually array the range of possible alternatives for
comparison. We only minimize, therefore, cost or risk as the explicit goal, but
set restrictions or constraints on the allowable levels of the other variables.
This procedure more explicitly reveals the effects of different choices.
For example, total costs could be constrained to never exceed a specified
amount and the model could then minimize total risks within that cost constraint.
To develop other alternatives, we could set the cost constraint one level lower or
higher and again have the model minimize risk. Repeating this procedure yields a
series of strategies that can compared and evaluated. As the total allowable cost
is set higher and higher, we would expect the resulting strategies to show that
greater levels of societal safety are attainable. Two other ways to choose goals
and establish constraints are (1) to constrain risk at alternative levels and
repeatedly minimize cost, and (2) to constrain the risk/cost ratio to a particular
value and minimize either cost or risk.
Although the model only minimizes total costs or risks, constraints can be set
at any three different levels. First, the total risk or cost for the entire system
can be constrained, as described above. This implies that the total effects on the
society is of concern. We anticipate using this type of constraint for every
alternative strategy the model generates. We can apply second level of constraint
on the total risk or cost of each W-E-T combination or variation. If a waste is
generated by a economically unstable industry, for example, it is possible to set a
limit on the total cost imposed on managing that waste to limit the financial
requirements on the sensitive industry. The third level of constraint the model
can incorporate is at the level of the risks and costs of individual W-E-T
combinations. The model could test, for example, the effects of limiting the risks
from certain types of technologies or wastes to more strict levels than others.
Wherever constraints are used, their most important function is to help us compare
situations rather than dictate levels of control or cost.
ICF Incorporated
-------
5-4
The example we use in this chapter limits total risk to no greater than 13, and
the risk from any individual W-E-T combination to a score of 9 or less. These
values are arbitrary and are only used to demonstrate the type of results possible
with the model. The example minimizes total costs within these risk constraints,
the parameter values discussed earlier, and the risk and cost information in the
data base.
ARRAY ALTERNATIVES
A realistic strategy must include all 83 waste types and account for all the
values of the waste. Because the 83 waste types are representative of all wastes, a
strategy incorporating all of them, we believe, is roughly representative of
handling of all wastes. A realistic strategy must also include as many as all 13
environments for each waste, depending on whether it is generated or disposed of in
that environment. Because waste volumes and disposal capacities are tied to
environments, a strategy that does not incorporate all relevant environments would
omit some disposal capacity and waste.
If each waste existed in each environment, a minimum of 1129 W-E-T combinations
would be necessary to constitute a realistic strategy (13 environments x 83 waste
streams). This number is reduced to the extent that not all wastes exist in each
environment and increased to the extent that many W-E-T combinations meet all the
constraints and equally satisfy the specific objective (minimizing cost or risk).
Results of Example Strategy
Our example strategy had as its goal the least cost W-E-T combinations at an
individual W-E-T risk of 9 or lower. For the universe of 83 waste streams, this
strategy selected W-E-T combinations which yielded a total risk score of roughly
12.6 and a total cost score of 33.6 (about $1.6 billion in direct capital and
operating costs). Total risk of managing the waste streams in all the environments
ranged from 6 to 12; total cost ranged from 11 to 32. The following exhibit
indicates the distribution of the W-E-Ts by risk and cost scores.
Exhibit 5-1 shows the range of total costs and total risks for handling each
waste stream in all possible environments. The cost score represents a doubling
scale, and the risk score represents a tenfold scale. There were no instances of
low risk, low cost combinations, a category that has been referred to as a "free
lunch." Nearly 85 percent of the total risks fall within one order of magnitude of
10 for this strategy. Nearly 10 percent of the occurrences fall in the high cost,
high risk area. This indicates the waste streams for which additional research on
handling technologies could reduce both the risk and the cost.
Three waste streams -- distillation residues from acetaldehyde production
(#30), still bottoms from acrylonitrile production (#31), and aqueous spent
antimony catalyst from fluoromethane production (#32) did not have any waste
management approach that enabled them to achieve a risk score of 9 or less. Two
possible explanations are that the model did not use the most effective management
techniques for these wastes or that managing these wastes may be especially
hazardous.
ICF Incorporated
-------
5-5
EXHIBIT 5-1
DISTRIBUTION OF WASTE STREAMS
BY TOTAL RISK AND TOTAL COST
RISK SCORE
6
7
8
9
10
11
12
TOTAL
10-14
0
0
2
0
4
1
0
7
c s
0 c
S 0
T R
E
15-19
0
0
3
5
9
5
0
22
20-24
1
1
1
6
17
10
1
37
25-29
0
1
0
1
5
3
1
11
30-35
0
0
0
0
0
2
1
3
TOTAL
1
2
6
12
35
21
3
80*
*Three waste streams had no W-E-T combination with a
risk of 9 or less.
ICF Incorporated
-------
5-6
Lined landfills appear as the low-cost choices infrequently; "no treatment" is
a frequent choice. Long-distance transportation of waste also is used relatively
infrequently. The pathway that generates the highest risk most frequently was
ground water. Air is seldom the pathway, and surface water, almost never.
Generally, the range of risks for a given waste in a given environment is very
similar. Some wastes, however, have scores that vary by about 5 orders of
magnitude. The environmental category can also produce differences of up to 5
orders of magnitude in risk for the same waste and technology; likewise, for the
same environment and technology, there can also be great variation in risk scores.
These tentative results derive from analysis of only one of many possible
strategies. They do tend to confirm our hypothesis that all three factors --
wastes, environments, and technologies -- are critical to comprehensive analysis of
the hazards posed by waste management situations.
These highly tentative results require much further testing and evaluation
before any trust can be placed in them. One step is to examine more strategies.
Another, as we discuss below, is sensitivity analysis.
PERFORM SENSITIVITY ANALYSIS
We noted above that understanding the basis for the results is essential to
beginning to implement a strategy. This understanding is partly gained through
sensitivity analysis.
We carry out sensitivity analysis by changing in a prescribed manner the
parameters input to the model. If we wish to test the effects of changes in risk,
for example, we could modify the risk scores by adding (and subsequently
subtracting) the uncertainty values we estimated for each score. We could make the
same test on the cost scores. In a similar manner, we could test other assumptions
such as the effect of misestimating generation volumes of certain wastes or the
capacity of technologies.
By identifying critical assumptions, we can point directions for the additional
analysis necessary before any strategy were adopted. Other relevant factors are
the reasonableness of the strategy in terms of the scientific and engineering
certainty necessary for regulatory changes and the administrative and legal
requirements to implement a potential strategy. Similarly, the costs must be
examined for sensitivity to site-specific environmental factors. A strategy that
used only evaporation technology might be suspect because its cost in cold climates
would be substantially increased.
SELECT STRATEGY
The primary objective of this project is to provide a framework that allows the
testing of alternative regulatory approaches on a coarse cost and risk scale.
Thus, useful strategies can be identified and inappropriate ones, screened out.
Any strategy depends on the value judgments used to set the parameters and choose
the goals. Thus, there is no "correct" strategy. There are merely sets of
strategies that represent particular risk-cost tradeoffs. What a strategy does
represent is a comprehensive method for considering the effects of a selected risk-
cost trade-off.
ICF Incorporated
-------
5-7
The risk-cost policy model provides a framework for debate over alternatives,
but requires restraint in application and interpretation. The major assumptions
and simplifications render detailed insight into the specifics of a regulation
impossible. Priorities and long-term goals, however, can be identified and
variations tested. If sufficient differentiation among the W-E-T combinations
exist to identify the potential extremes of under- and overregulation, the model
will have been a success.
ICF Incorporated
-------
REFERENCES
ICF Incorporated
-------
REFERENCES
RELATING TO CHAPTER 2 AND APPENDICES 1 THROUGH 6
1. Booz-Allen and Hamilton and Putnam, Hayes and Bartlett. 1980.
Hazardous Waste Generation and Commercial Hazardous Waste Management
Capacity -- An Assessment. U.S. Environmental Protection Agency, Office
of Solid Waste, Washington, D.C.
2. Putnam, Hayes and Bartlett. 1980. (Unpublished data in support of
Hazardous Waste Generation and Commercial Hazardous Waste Management
Capacity -- An Assessment.)
3. . 1981-1982. (Delisting petitions) U.S. Environmental
Protection Agency, Office of Solid Waste, Washington, D.C.
4. . 1981. Federal Register. Volume 46, No. 11, 1-16-81.
5. . 1980. Federal Register. Volume 45, No. 98, 5-19-80.
6. U.S. Environmental Protection Agency (USEPA). 1980. RCRA Background
Document, Identification and Listing of Hazardous Wastes, Finalization
of May 19, 1980 Hazardous Waste List. Office of Solid Waste,
Washington, D.C.
7. U.S. Environmental Protection Agency (USEPA). 1980. RCRA Background
Document, Identification and Listing of Hazardous Wastes, Finalization
of July 16, 1980 Hazardous Waste List. Office of'Solid Waste,
Washington, D.C.
8. SCS Engineers. 1976. Assessment of Industrial Hazardous Waste Practices,
Leather Tanning and Finishing Industry.
9. Wapora, Inc. 1977. Assessment of Industrial Hazardous Waste Practice,
Special Machinery Manufacturing Industries.
10. Battelle Columbus Laboratories. 1976. Assessment of Industrial Hazardous
Waste Practices, Electropolating and Metal Finishing Industries -- Job
Shops.
11. Versar, Inc. 1975. Assessment of Industrial Hazardous Waste Practices,
Storage, and Primary Batteries Industries.
12. Colspan Corporation. 1977. Assessment of Industrial Hazardous Waste
Practices in the Metal Smelting and Refining Industry.
13. Wapora, Inc. 1977. Assessment of Industrial Hazardous Waste Practices,
Electronic Components Manufacturing Industry.
ICF Incorporated
-------
R-2
14. Swain, John W., Jr. 1976. Assessment of Industrial Hazardous Waste
Management Practices, Petroleum Rerefining Industry.
15. Jacobs Engineering Company. 1976. Assessment of Industrial Hazardous Waste
Practices in the Petroleum Refining Industry.
16. Wapora, Inc. 1975. Assessment of Industrial Hazardous Waste Practices,
Paint and Allied Products Industry, contract Solvent Reclaiming
Operations and Factory Application of Coatings.
17. A.D. Little, Inc. 1976. Pharmaceutical Industrial Hazardous Waste
Generation, Treatment and Disposal.
18. Foster D. Snell, Inc. 1978. Assessment of Industrial Hazardous Waste
Practices, Rubber and Plastics Industry.
19. Versar, Inc. 1975. Assessment of Industrial Hazardous Waste Practices,
Inorganic Chemicals Industry.
20. TRW Systems Group. 1975. Assessment of Industrial Hazardous Waste
Practices,
Organic Chemicals, Pesticides, and Explosions Industries.
21. Versar, Inc. 1976. Assessment of Industrial Hazardous Waste Practices,
Textiles Industry.
22. Mitre Corp. 1979. Subtitle C, Resource Conservation and Recovery Act of
1976, Draft Environmental Impact Statement and Appendices.
23. TRW. 1979. Technical Environmental Impacts of Various Approaches for
Regulating Small Volume Hazardous Waste Generators, Volumes I and II.
24. A.D. Little, Inc. 1979. Draft Economic Impact Analysis, Subtitle C,
Resource Conservation and Recovery Act of 1976.
25. Versar, Inc. 1977. Alternatives for Hazardous Waste Management in the
Inorganic Chemical Industry.
26. Jacobs Engineering Co. 1979. Alternatives for Hazardous Waste Management
in the Petroleum Refining Industry.
27. K.W. Brown and Associates, Inc. 1980. Hazardous Waste Land Treatment.
28. SCS Engineers. 1979. Disposal of Polychlorinated Biphenyls (PCBS) and
PCB-Contaminated Materials, Volume I.
29. U.S. Environmental Protection Agency (USEPA). 1980. Disposal of
Hazardous Waste, Proceedings of the Sixth Annual Research Symposium.
EPA Municipal Environmental Research Laboratory.
30. Ontario Ministry of the Environment. 1979. Hazardous Contaminants
Programme, Polynuclear Aromatic Hydrocarbons.
ICF Incorporated
-------
R-3
RELATING TO CHAPTERS 3 AND 4 AND APPENDICES 7 THROUGH 9
Adamson, A.W. 1973. A Textbook of Physical Chemistry. Academic
Press, New York
Baird, R., Carmona, L., AND Jenkins, R.L. 1977. Behavior
of benzidine and other aromatic amines in aerobic wastewater
treatment. J. Water Pollution Control Federation 49:1609-1615
Catalytic, INC. 1980. Report to EPA, "Evaluation of Organic
Chemicals and Plastics and Synthetics," December 1980
Freeze, R.A., and Cherry, J.A. 1979. Groundwater. Prentice-Hall,
E~glewood Cliffs, N.J.
Herbes, S.E. 1981. Rates of microbial transformation of polycyclic
aromatic hydrocarbons in water and sediments in the vicinity
of a coal-coking wastewater discharge. Appl. Environ.
Microbiol. 41:20-28
Hwang, S.T. 1979. Treatability and Pathways of Priority Pollutants
in Biological Wastewater Treatment. U.S. Environmental
Protection Agency, Organic Chemicals Branch, Washington,
D.C.
Hwang, S.T. 1979. Treatability of the Priority Pollutants
by Activated Carbon. U.S. Environmental Projection Agency,
Organic Chemicals Branch, Washington, D.C.
Hwang, S.T. 1980. Priority Pollutants in Wastewater by Activated
Carbon. (Paper presented at AIChE Philadelphia Meeting,
June 1980). U.S. Environmental Projection Agency, Organic
Chemicals Branch, Washington, D.C.
Hwang, S.T., and Fahrenthold, P. 1979. Treatability of the
Organic Priority Pollutants by Steam Stripping. U.S. Environ-
mental Protection Agency, Organic Chemicals Branch, Washington,
D.C.
Lu, P.Y., and Metcalf, R.L. 1975. Environmental fate and
biodegradability of benzene derivatives as studied in
a model aquatic ecosystem. Environ. Health Perspect.
10:269-284
Maki, A., ed. 1980. Proceedings of the Workshop of Biotransfor-
mation and the Fate of Chemicals in the Aquatic Environment.
A-. Soc. Microbiol., Washington, D.C.
Mancini, J.L. 1978. Analysis framework for photodecomposition
in water. Environ. S~i. Technol. 12:1274-1276
ICF INCORPORATED
-------
R-4
Merck Index. 1976. 9th ed. Windholz, M., et al., eds. Merck
and Co., Rahway, N.J.
Metcalf and Eddy. 1972. Wastewater Engineering: Collection,
Treatment, Disposal. 1st ed. McGraw-Hill, New York
Miller, G.C., and Zepp, R.G. 1979. Photoreactivity of aquatic
pollutants sorbed on suspended sediments. Environ. Sci.
Technol. 12:806-863
Morel, F.M.M., and SHAFF, S.L. 1980. Geochemistry of Municipal
Waste in Coastal Waters. Massachusetts Institute of Technology,
Technical Report No. 259
Dialog Data Bases, OHMTADS Data. Lockheed Information Systems,
Palo Alto, Calif.
Oliver, B.G., Cosgrove, E.G., and Carey, J.H. 1979. Effect
of suspended sediments on the Photolysis of organics in
water. Environ. Sci. Technol. 13:1075-1077
Perry, R.H., and Chilton, C.H. 1973. Chemical Engineer's
Handbook. 5th ed. McGraw-Hill, New York
Shamat, N.A., and Maier, W.J. 1980. Kinetics of biodegradation
of chlorinated organics. J. Water Pollution Control Federation
52:2158-2166
Suffet, I.H., ed. 1977. Fate of Pollutants in the Air and
Water Environments. Wiley-Inter science, New York
Thibodeaux, J. 1979. Chemodynamics: Environmental Movement
of Chemicals in Air, Water and Soil. Wiley-Interscience,
New York
Tinsley, I.J. 1979. Chemical Concepts in Pollutant Behavior.
Wiley-Inter science, New York
T"cker, E.S., Saeger, V.W., and Hicks, 0. 1975. Activated
sludge primary biodegradation of polychlorinated biphenyls.
Bull. Environ. Contam. Toxicol. 14:705-713
U.S. Department of Commerce (USDC). 1981. Octanol/Water Partition
Coefficients and Aqueous Solubilities of Organic Compounds.
Washington, D.C. NBSIR 81-2406 (Wasik, Tewari, Miller
and Martire)
U.S. Environmental Protection Agency (USEPA). 1979. Water-
Related Environmental Fate of 129 Priority Pollutants.
Volumes I and II. Office of Water Planning and Standards,
Washington, D.C. , December 1979. EPA 440/4-79-029 A & B
ICF INCORPORATED
-------
R-5
U.S. Environmental Protection Agency (USEPA). 1979. Fate
of Priority Pollutants in Publicly Owned Treatment Works,
Pilot Study. Effluent Guidelines Division, Office of
Water and Waste Management, Washington, D.C. , October
1979. EPA 440/1-79-300
U.S. Environmental Protection Agency (USEPA). 1980. Fate
of Priority Pollutants in Publicly Owned Treatment Works.
Interim Report, Effluent Guidelines Division, Office of
Water and Waste Management, Washington, D.C. , October
1980. EPA 440/1-80-301
U.S. Environmental Protection Agency (USEPA). 1980. Ambient
Water Quality Criteria for Arsenic. Criteria and Standards
Division, Washington, D.C. , October 1980. EPA 440/5-80-
021
U.S. Environmental Protection Agency (USEPA). 1980. Ambient
Water Quality Criteria for Cadmium. Criteria and Standards
Division, Washington, D.C., October 1980. EPA 440/5-80-
025
U.S. Environmental Protection Agency (USEPA). 1980. Ambient
Water Quality Criteria for Chromium. Criteria and Standards
Division, Washington, D.C. , October 1980. EPA 440/5-80-
035
U.S. Environmental Protection Agency (USEPA). 1980. Ambient
Water Quality Criteria for Lead. Criteria and Standards
Division, Washington, D.C. , October 1980. EPA 440/5-80-
057
U.S. Environmental Protection Agency (USEPA). 1980. Ambient
Water Quality Criteria for Nickel. Criteria and Standards
Division, Washington, D.C., October 1980. EPA 440/5-80-
060
U.S. Environmental Protection Agency (USEPA). 1980. Ambient
"ater Quality Criteria for Zinc. Criteria and Standards
Division, Washington, D.C. , October 1980. EPA 440/5-80-
079
U.S. Environmental Protection Agency (USEPA). 1980. Attenuation
of Water-Soluble Polychlorinated Biphenyls by Earth Materials.
Municipal Environmental Research Laboratory, Cincinnati,
Ohio, May 1980. EPA 600/2-80-027
Walker, J.D., and Colwell, R.R. 1976. Biodegradation rates
of components of petroleum. Can. J. Microb. 22:1209-1213
Weast, R.C., ed. 1975. Handbook of Chemistry and Physics.
56th ed. Chemical Rubber Company Press
ICF INCORPORATED
-------
R-6
Weast, R.C., ed. 1977. Handbook of Chemistry and Physics.
57th ed. Chemical Rubber Co., Cleveland
Weber, W.J. 1972. Physico Chemical Processes for Water Quality
Control. Wiley-Inter science, New York
Zepp, R.G. 1978. Quantum yield for reaction of pollutants
in dilute aqueous solution. Environ. Sci. Technol. 12:327-329
ICF INCORPORATED
-------
R-7
Cox, R.A., R.G. Derwent, A.E.J. Eggleton, and J.E. Lovelock, "Photochemical
Oxidation of Halocarbons in the Troposphere," Atmospheric Environment 10:
305-308, 1976.
Crutzen, P.J., I. Isakesen, and J.R. McAffe, "The Impact of the Chlorocarbon
Industry on the Ozone Layer," Journal of Geophysical Research, 83(C1) :
345-363, 1978.
Dilling, W.L., N.B. Tefertiller, and G.J. Kallos, "Evaporation Rates and
Reactivities of Methylene Chloride, Chloroform, 1,1,1-Trichloroethane,
Trichloroethylene, Tetrachlorethylene, and Other Chlorinated Compounds in
Dilute Aqueous Solution," Environmental Science and Technology, 9: 833-838,
1975.
Mabey, W.R. and T. Mill (1976) Kinetics of Hydrolysis and Oxidation of Organic
Pollutants in the Aquatic Environment. Extended abstracts of Symposium on
Nonbiological Transport and Transformation of Pollutants on Land and Water:
Processes and Critical Data Required for Predictive Description, May 11-13,
1976. Gaithersburg, Maryland: National Bureau of Standards.
National Academy of Sciences, Chloroform, Carbon Tetrachloride, and
Halomethanes: An Environmental Assessment, National Academy of Sciences,
Washington, D.C., 1978.
Neely, W.B., G.E. Blau, and T. Alfrey, Jr., "Mathematical Models Predict
Concentration-Time Profiles from Chemical Spill in a River," Environmental
Science and Technology, 10: 72-76.
Yung, Y.L., M.B. McElroy, and S.C. Wofsy, "Atmospheric Halocarbons: A
Discussion with Emphasis on Chloroform," Geophysical Research Letters, 2(9):
397-399, 1975.
ICF Incorporated
-------
APPENDIX 1
TREATMENT TECHNOLOGIES
ICF Incorporated
-------
APPENDIX 1
TREATMENT TECHNOLOGIES
INTRODUCTION
Twenty-one hazardous waste treatment unit operations have been modeled for
incorporation into the risk-cost framework. For these unit operations,
treatment is effected through volume reduction and/or waste destruction
techniques.
As shown earlier in Exhibit 2-7, we define waste streams in terms of
eighteen characteristics. These characteristics change in response to the
treatment used. This appendix describes how the model makes these changes for
each of the treatment technologies included in our risk-cost framework.
The discussion of each technology below has four parts: a description of
the treatment process; a table indicating the quantity and routes by which the
constituent of concern (X) is released to the environment1 and the
parameters of the treatment operation (vessel size, time waste is retained in
vessel) determining the quantity of this release; a summary of the assumptions
or simplifications we made to simulate the treatment process; and an algorithm
presenting the mathematical basis for the changes in waste characteristics
brought about by the technology.
The terminology used in this appendix to describe the characteristics of
waste streams is somewhat more technical than that used in Exhibit 2-7. Below
is a key that equates the terminology used in this appendix to describe waste
characteristics with the terminology used in Exhibit 2-7. Values
corresponding to a waste characteristic as it leaves the treatment technolgy
are indicated by a prime (') following the characteristic code.
xAppendix 4 contains a detailed description of the four sources of
release (evaporation, aeration, routine spillage and accidental spillage) of
hazardous component X to the environment from the treatment processes.
ICF Incorporated
-------
Al-2
TERMINOLOGY USED IN APPENDIX 1
S(1) Total mass flux
S(2) Nonwater mass fraction of
the total mass flux
S(3) Suspended nonwater mass
fraction of the total mass
flux
S(4) Dissolved nonwater mass
fraction of the total mass
flux
S(5) Average specific gravity of
the suspended solids at 25°C
S(6) Relative volatility (P° /
P°H20) when P = vapor
pressure at 25°C
S(7) Mass fraction of X in the
total sludge flux
S(8) Dissolved mass fraction of X
S(9) Suspended mass fraction of X
S(10) Mass flux of X
S(ll) Mass flux of X in liquid
phase
S(12) Mass flux of X in suspended
solid phase
S(13) Mass flux of X in fixed solid
phase
S(14) Mass flux of nonwater species
S(15) Mass flux of water
S(16) Energy value of the bulk
waste stream
SC17) Solubility of X in water at
25°C
S(18) Molecular weight of species X
TERMINOLOGY USED IN EXHIBIT 2-7
Total weight including water
Percent nonwater by weight
Percent nonwater suspended by weight
Percent nonwater dissolved by weight
Same
Relative volatility (vapor pressure
of X)/(vapor pressure of water) at
25°C
Percent of X by weight
Percent of X dissolved by weight
Percent of X suspended by weight
Total X by weight
Total X dissolved by weight
Total X suspended by weight
Same
Total weight excluding water
Total water by weight
Net BTU content
Same
Molecular weight of X
ICF Incorporated
-------
Al-3
1. CHEMICAL COAGULATION
a. Description
Chemical coagulants are used to enhance settling of very small or
gelatinous suspended solid particles with densities similar to the density of
water. Such solids include viruses, bacteria, inorganic silts and powders,
and colloidal organics. Chemical coagulants are added to the sludge in a
flash mixing tank, where the coagulation reaction occurs rapidly. The slurry
is then fed to a flocculation basin. Slow mixing is maintained to encourage
coagulation of particles. Following flocculation, a conventional clarifier is
used to separate the effluent process return stream from the hazardous waste
stream underflow.
A summary of the important treatment vessel characteristics used in
estimating the individual and overall loss fractions to the environment is
given in Exhibit 1A-1.
b. Assumptions
We made the following assumptions and/or simplifications to simulate how
the chemical coagulation technology changes the characteristics of a waste
stream:
1. The effluent from the clarifier contains no suspended
solids. All influent suspended solids and all coagulant
chemicals are carried out with the underflow hazardous waste
stream.
2. The underflow hazardous waste stream contains 2 percent
solids and 98 percent retained liquor, by mass. If the
coagulated solids stream from the flocculation basin
contains more than 2 percent solids, then the clarifier
produces no effluent, and all of the flocculant stream
appears in the underflow as the hazardous waste stream.
3. The mass flux of chemical coagulant is calculated as a
function of the influent suspended solids.
4. This technology is inappropriate for influents in which
hazardous component X is in the dissolved phase. In this
case, the model advises the user to select a more
appropriate treatment technology.
c. Mass Balance Algorithm
1. The coagulant dosage is assumed to be 50 percent of the influent
sludge suspended solids mass:
COAG = .5 • S(3) • SCI)
ICF Incorporated
-------
EXHIBIT 1A-1
VESSEL CHARACTERISTICS AND RELEASE RATES
FOR CHEMICAL COAGULATION
Treatment Vessel Characteristics
vessel type
vessel dimensions
retention time (hours)
ag i tat ion
evaporative surface area (meters)
M (fraction of influent sludge
contained in vessel)
Flash mixer
open
H = 2D
. I
rap i d
-2
1.505 (10 )
-2
4. 17 ( 10 )
S( I )
2/3
F loccuI a tor
open
H = 0
.5
s low
-I 2/3
6.985 ( 10 ) • S( I )
-3
2.08 (10 )
C I a r i f i e r
open
H = . 5D
2
s I ow
N/A
-2
8.33(10 )
o
Tl
3
O
O
-I
TJ
O
"I
tl)
<-h
a>
a
Environmental Releases
aeration loss fraction
total aeration fraction
routine spillage fraction:
(a) routine maintenance
(b) m i xe r spI ash i ng
(c) materials handling
total routine spillage fraction
accidental spillage fraction:
best case, spills/year
worst case, spills/year
typical case fraction
total typical fraction
overall loss fraction
(sum of aeration, routine spill-
age, and accidental spillage
fractions)
10
10
-5
10
-U
-it
10
12
-5
1.28 (10 )
2.6(10 )
-1
3.7(10 )
-7
10
-6
10
.5
12
I .923 ( 10 )
-7
10
-6
10
.5
2
-U
I .997 (10 )
>
*
I
-F-
-------
Al-5
2. The total suspended solids mass fraction, considering influent
suspended solids plus coagulant, is calculated using the coagulant flow:
MASS FRACTION OF
SOLIDS IN FEED TO
CLARIFIER
TOTAL INFLUENT
SUSPENDED
SOLIDS FLOW
COAGULANT
DOSAGE
TOTAL INFLUENT + COAGULANT
SLUDGE FLOW DOSAGE
SOLIDS = [S(3)*S(1)+C0AG]/[S(1)+C0AG]
3. The influent to the clarifier is checked to determine whether or not a
clear effluent stream and an underflow with 2 percent solids can be obtained.
The result of this test determines some of the characteristics of the RCRA
stream underflow:
a. If SOLIDS < 0.0200, then the described effluent and underflow
can be obtained from the clarifier. The RCRA stream underflow is determined
as follows:
"the solids mass fraction is set equal to 2 percent
S(3)1 = 0.0200
*the total sludge flux is determined using the described solids
fraction, assuming that all of the suspended solids and coagulant appear
in the underflow:
INFLUENT SOLIDS _ SOLIDS IN
TO THE CLARIFIER UNDERFLOW
S(3) • SCI) + COAG = (.0200) • S(l)'
therefore,
S(l)' = (1/.02) • [S(3) • SCI) + COAG]
= 50 • [S(3) • SCI) + COAG]
*the dissolved nonwater mass fraction is determined as follows:
MASS FRACTION OF DISSOLVED
NONWATER SPECIES IN
RCRA STREAM LIQUOR
MASS OF DISSOLVED NONWATER
SPECIES IN RCRA
STREAM LIQUOR
MASS OF RCRA
STREAM LIQUOR
MASS FRACTION OF DISSOLVED
NONWATER SPECIES IN TOTAL
CLARIFIER LIQUOR
MASS OF DISSOLVED NONWATER
SPECIES IN TOTAL CLARIFIER
LIQUOR
MASS OF TOTAL CLARIFIER
LIQUOR
ICF Incorporated
-------
Al-6
[S(4)1 • S(l)']/[0.98 • S(1)'] = [S(4) • S(1)]/[S(4) • S(l)+S(15]
therefore,
S(4)1 = 0.98 • S(4) • S(1)/[S(4) • S(l)+S(15)]
"the total nonwater mass fraction is the sum of the mass fractions of
the dissolved and suspended species:
S(2)1 = S(3)1 + S(4)'
"the mass flux of X in the dissolved phase of the underflow is
determined as follows:
MASS FRACTION OF X IN THE MASS FRACTION OF X IN THE
DISSOLVED PHASE OF THE = DISSOLVED PHASE OF THE
RCRA STREAM CLARIFIER LIQUOR
MASS OF X IN THE DISSOLVED MASS OF X IN THE DISSOLVED PHASE
PHASE OF THE RCRA STREAM = OF THE CLARIFIER INFLUENT
MASS OF THE DISSOLVED PHASE MASS OF THE DISSOLVED PHASE OF
OF THE RCRA STREAM THE CLARIFIER INFLUENT
S(11)7[.98»S(1)'] = S(11)/[S(4)*S(1)+S(15)]
S(ll)' = [.98.S(11)#S(1)1]/[S(4)#S(l)+S(15)]
b. If SOLIDS > 0.0200, the clarifier cannot produce the desired
effluent and underflow. As a result, no process return stream is produced,
and the entire influent stream is retained as the RCRA stream.
"the solids mass fraction is set equal to the suspended solids mass
fraction of the clarifier influent, calculated earlier:
S(3)1 = SOLIDS
-the total RCRA stream flow is the sum of the influent sludge flow
and the coagulant dosage:
SCI)' = S(1) + COAG
*the total dissolved nonwater mass fraction is reduced:
S(4)1 = S(4) • S(l)/S(l)'
ICF Incorporated
-------
Al-7
*the total nonwater mass fraction is the sum of the suspended and the
dissolved mass fractions:
S(2)1 = S(3)1 + S(4)1
"the mass flow of X in the dissolved phase remains unchanged:
S(ll)1 = S(11)
4. The mass flow of X in the suspended phase remains unchanged:
S(12) ' = S(12)
5. The total mass flow of X in the RCRA stream is the dissolved and
suspended phase mass flows:
S(10)' = S(ll)' + S(12)1
6. The mass flow of the nonwater species is determined from the total
sludge flow and the appropriate mass fraction:
S( 14)1 = S (2)' • S(1)'
7. The water flow is found as the difference between the total sludge
flow and the flow of the nonwater species:
SC15)' = S(1) ' - S(14)1
8. The mass fraction of X and the proportion of X in the dissolved and
suspended solid phases are determined from the appropriate mass flows:
S(7)' = S(10)7S(1)'
S(8) ' = S(11)7S(10)'
S(9)' = S(12)'/S(10)1
ICF Incorporated
-------
Al-8
2. FILTER PRESS
a. Description
Filtration is used to dewater and concentrate the suspended solids in a
sludge stream. Pressure filtration removes suspended solids from the bulk
solution by passing the fluid through a cloth membrane under pressure. The
solids are retained on the membrane while the filtrate, free .of suspended
solids, passes through the membrane.
Filtration in a press is accomplished using alternating plates, the first
type grooved or perforated to permit filtrate drainage and the second type
hollow to collect the solids cake. The faces of each plate are covered with
filter cloth. The plates are hung face to face from horizontal parallel bars
and compressed to a relatively watertight seal. During filtration, the sludge
is fed into one end of the filter press under pressure. As the sludge
contacts the membrane, the liquid mixture passes through the membrane and
flows to the runoff channel as filtrate. Solids are deposited on one face of
the membrane, forming a cake containing suspended solids and residual liquor.
Backwashing with water may be used to remove residual liquor from the cake and
crevices in the filter press. The filtrate and backwash water are collected
in an open trough or a pipeline for recycling and/or disposal. The filter
press plates are dismantled at the end of the process to allow removal of the
solids cakes from the membrane, usually by scraping.
When the sludge contains extremely fine or gelatinous solids, the
filtration rate and efficiency may be improved by the addition of a filter aid
such as a metal hydroxide or polymer in a preconditioning step.
A summary of the important treatment vessel characteristics used in
estimating the individual and overall loss fractions to the environment is
given in Exhibit 1A-2.
b. Assumptions
We made the following assumptions and/or simplifications to simulate how
the filter press technology changes the characteristics of a waste stream:
1. Solids retention efficiency of the filter press during operation
is 100 percent. The filtrate leaving the filter press is free
of suspended solids. All suspended solids are retained in the
filter cake with some residual liquor. No backwashing step is
included.
2. The filtrate is returned to the industrial process. The solids
cake and residual filtrate are treated as the hazardous waste
stream.
ICF Incorporated
-------
Al-9
EXHIBIT 1A-2
VESSEL CHARACTERISTICS AND RELEASE RATES FOR FILTER PRESS
Treatment Vessel Characteristics
vessel type closed
vessel dimensions
retention time (hours)
agitation N/A
evaporative surface area (meters) N/A
M (fraction of influent sludge
contained in vessel)
Environmental Releases
aeration loss fraction N/A
total aeration fraction
routine spillage fraction:
(a) routine maintenance 8.82x10-''
(b) mixer splashing N/A
(c) materials handling 3.3xl0-l>
total routine spillage fraction 1.2xl0-u
accidental spillage fraction:
best case, spills/year 6
worst case, spills/year 100
typical case fraction 1.056xl0-4
total typical fraction
overall loss fraction 1.3xl0-3
(sum of aeration, routine spill-
age, and accidental spillage
fractions)
ICF Incorporated
-------
A1 -10
3. The addition of a filter aid as a preconditioning step is
optional:
a. If the sludge was not preconditioned, the resulting filter
cake is 20 percent solids and 80 percent filtrate by mass.
b. If the sludge was preconditioned in order to enhance
filterability, the solids content of the sludge is assumed
to increase owing to the presence of the filter aid. The
filter cake composition is 50 percent solids and 50 percent
filtrate by mass.
4. The hazardous component X need not be completely in the
suspended solids phase. If, however, less than 80 percent of X
is present as solids, the model advises the user to select a
more suitable process alternative.
5. No phase changes (solids precipitation or liquid vaporization)
are expected to occur during the filtration process.
c. Mass Balance Algorithm
1. The user must state whether or not the sludge was preconditioned by
the addition of a filter aid. The statement sets the following values for the
variables ALPHA and BETA:
ALPHA = solids mass fraction in filter cake
BETA = residual liquor mass fraction in filter cake
"if the sludge was not preconditioned:
ALPHA =0.2, BETA =0.8
"if the sludge was preconditioned:
ALPHA =0.5, BETA =0.5
Additionally, the influent sludge characteristics are modified to account
for the addition of the insoluble filter aid, 0.10 pounds of filter aid per
pound of suspended solids:
"the influent sludge flow is increased by the amount of filter aid
added:
SCI)' = SCI) + 0.10 • S(3) • S(l) = [1+0.10 • S(3)] • SCI)
"the influent suspended solids fraction is increased:
S(3)' = 1.10 • S(3) • S(l)/S(l)'
= 1.10 • S(3)/[1+0.10 • S(3)]
ICF Incorporated
-------
Al-11
*the influent dissolved nonwater fraction is decreased:
S(4)1 = S(4) • S(l)/S(l)'
"the total nonwater fraction is increased:
S(2)1 = S(4) + S(3)'
,vthe mass fraction of X in the sludge is decreased:
S(7) = S(7) • SCD/SC1)' = S(7) / [ 1+0 • 10 • S(13) ]
"the flow of the nonwater species is increased:
S(14)1 = S(l)' • S(2)1
2. The RCRA waste stream flow is calculated according to the described
solids mass fraction:
TOTAL MASS OF INFLUENT TOTAL MASS OF
SUSPENDED SOLIDS = SOLIDS IN CAKE
S(3) • S(l) = ALPHA • S(l)'
therefore,
S(1)' = (1/ALPHA) • S(3) • SCD
3. The solids mass fraction of the cake is set equal to the described
variable ALPHA:
S(3)' = ALPHA
4. The dissolved nonwater mass fraction retained in the cake is
calculated as a function of the described variable BETA:
MASS FRACTION OF MASS FRACTION OF
DISSOLVED NONWATER = DISSOLVED NONWATER
SPECIES IN TOTAL SPECIES IN THE RCRA
LIQUOR STREAM LIQUOR
MASS OF DISSOLVED NONWATER MASS OF DISSOLVED NONWATER
SPECIES IN TOTAL LIQUOR = SPECIES IN THE RCRA STREAM LIQUOR
MASS OF TOTAL LIQUOR MASS OF RCRA STREAM LIQUOR
MASS OF DISSOLVED NONWATER SPECIES IN THE
[S(4)*S(1)] = RCRA STREAM LIQUOR
[S(4)*S(1)+S(15)] BETA»S(1)'
ICF Incorporated
-------
Al-12
Rearranging and dividing by S(1)1 :
S(4)' = [Mass of dissolved nonwater species in RCRA stream]/S(l)1
= [BETA'S(1)1]•[S(4)»S(1)]*[S(4)»S(1)+S(15)]*S(1)1
therefore:
S(4)1 = BETA • S(4) • S(1)/[S(4) • S(l)+S(15)]
5. The total nonwater mass fraction is the sum of the suspended and the
dissolved nonwater fractions:
S(2)' = S(3)1 + S(4)1
6. The mass flow of X in the dissolved phase is calculated as a function
of the defined variable BETA:
MASS FRACTION OF X IN = MASS FRACTION OF X IN THE RCRA
TOTAL LIQUOR STREAM LIQUOR
MASS OF X IN TOTAL LIQUOR = MASS OF X IN RCRA STREAM LIQUOR
MASS OF TOTAL LIQUOR MASS OF RCRA STREAM LIQUOR
S(11)/[S(4) • S(l)+S(15)] = S(11)1/[BETA • S(l)']
Rearranging:
S(11)' = BETA • S(l)' • S(11)/[S(4) • S(l)+S(15)]
7. The mass flow of X in the suspended solids phase remains unchanged,
since all solids are retained. Thus, the total mass flow of X in the RCRA
stream is the sura of the dissolved phase and the suspended solids phase flows:
S(10)1 = S(ll)' + S(12)'
8. The mass flow of the nonwater species is determined from the total
RCRA stream flow and the nonwater species mass fraction found above:
S(14)' = S(l)' • S(2)1
9. The mass flow of water is the difference between the total RCRA stream
mass flow and the mass flow of the nonwater species:
S(15)1 = SCI)' - S(14)1
ICF Incorporated
-------
Al-13
10. The mass fraction of X in the RCRA stream and the proportions of X in
the suspended solids and dissolved phases are determined from the appropriate
mass flows:
S(7)' = S(10)'/S(l)'
S(8) ' = S(ll) 7S(10)1
S(9)1 = S(12)1/S(10)1
ICF Incorporated
-------
Al-14
3. CENTRIFUGE
a. Description
Filtration is used to dewater and concentrate the suspended solids in a
sludge stream by mechanical means. Centrifugal separators use gravity
settling to separate solids from the liquor, but improve process efficiency by
using a strong centripetal acceleration instead of the force of gravity. The
results are more rapid filtration and a solids cake containing less residual
liquor.
The most common centrifugal filter found in sludge treatment is the basket
centrifuge. The sludge is fed into the bowl near the bottom and flows axially
as the bowl is spun. Suspended solids collect at the wall, while clarified
liquor flows to the end and is discharged past an overflow lip. Operation is
continuous until the solids cake is thick enough to impede clarification. The
feed is then stopped, the speed of the bowl is reduced, and the solids cake is
removed with a blade for discharge from the bottom of the bowl. The combined
feed and discharge cycle lasts approximately 20 minutes.
When the sludge contains extremely fine or gelatinous solids, the
filtration rate and efficiency may be improved by the addition of a filter aid
such as a metal hydroxide or polymer in a preconditioning step.
A summary of the important treatment vessel characteristics used in
estimating the individual and overall loss fractions to the environment is
given in Exhibit 1A-3.
b. Assumptions
We made the following assumptions and/or simplifications to simulate how
the centrifuge technology changes the characteristics of a waste stream:
1. The solids retention efficiency of the centrifuge during
operation is 100 percent. The filtrate leaving
the centrifuge is free of suspended solids. All suspended
solids are retained in the filter cake with some residual liquor.
2. The filtrate is recycled to the industrial process from which it
came. The solids cake and residual liquor are treated as the
hazardous waste stream.
3. The addition of a filter aid as a preconditioning step is
optional:
a. If the sludge was not preconditioned, the resulting filter
cake is 10 percent solids and 90 percent residual liquor by
mass.
ICF Incorporated
-------
Al-15
EXHIBIT 1A-3
VESSEL CHARACTERISTICS RELEASE RATES FOR CENTRIFUGE
Treatment Vessel Characteristics
vessel type
characteristic dimensions
retention time (hours)
agitation
evaporative surface area (meters)
M (fraction of influent sludge
contained in vessel)
closed
N/A
.25 (batch cycle time)
N/A
N/A
2.19xl0-2
Environmental Releases
aeration loss fraction
total aeration fraction
routine spillage fraction:
(a) routine maintenance
(b) mixer splashing
(c) materials handling
total routine spillage fraction
accidental spillage fraction:
best case, spills/year
worst case, spills/year
typical case fraction
total typical fraction
overall loss fraction
(sum of aeration, routine spill-
age, and accidental spillage
fractions)
N/A
N/A
1
12
2.23x10-"
2.2x10-"
ICF Incorporated
-------
A1 -16
b. If the sludge was preconditioned in order to enhance
filterability, the solids content of the sludge is assumed
to increase owing to the presence of the filter aid. The
filter cake composition is 25 percent solids and 75 percent
residual liquor by mass.
4. The hazardous component X need not be completely in the
suspended solids phase. If, however, less than 80 percent of X
is present as solids, the model advises the user to select a
more suitable process alternative.
5. No phase changes (solids precipitation or liquid vaporization)
are expected to occur in the filtration process.
c. Mass Balance Algorithm
The user must state whether or not the sludge was preconditioned by the
addition of a filter aid. The statement sets the following values for the
variables ALPHA and BETA:
ALPHA = solids mass fraction in filter cake
BETA = residual liquor mass fraction in filter cake
"if the sludge was not preconditioned:
ALPHA = 0.10; BETA =0.90
*if the sludge was preconditioned:
ALPHA = 0.25; BETA =0.75
The user is referred to the discussion of the mass balance algorithm for
Filter Press for the remainder of the centrifugal filter algorithm. The
equations are identical except for the values of ALPHA and BETA described
above.
ICF Incorporated
-------
A1 -17
4. VACUUM FILTER
a. Description
Filtration is used to dewater and concentrate the suspended solids in a
sludge stream by mechanical means. A vacuum filter removes suspended solids
from the bulk solution by drawing the fluid through a membrane using vacuum.
The solids are retained on the membrane while the filtrate, free of suspended
solids, passes through and is discharged.
A common vacuum filter for industrial sludge treatment is the continuous
rotary drum filter. A perforated cylindrical drum is supported in an open
tank so that it can rotate axially in the horizontal plane. The lower part of
the drum is immersed in the tank, while the upper part is exposed. A vacuum
pump and receiving vessel are connected to the drum's interior. A filter
membrane is wrapped around the drum like a belt.
During operation, the drum is rotated slowly about its axis as the sludge
level is maintained in the tank. Vacuum is applied, and filtrate is drawn
through the filter medium while a layer of solids is deposited on the
submerged portion of the drum. As the drum continues to rotate, the solids
cake emerges from the tank and is dried by the vacuum beneath the belt. The
cake is dislodged at the end of its circuit with a knife blade. The cleaned
filter surface is then rotated back into the sludge tank for another cycle.
When the sludge contains extremely fine or gelatinous solids, the
filtration rate and efficiency may be improved by the addition of a filter aid
such as a metal hydroxide or polymer in a preconditioning step.
A summary of the important treatment vessel characteristics used in
estimating the individual and overall loss fractions to the environment is
given in Exhibit 1A-4.
b. Assumptions
We made the following assumptions and/or simplifications to simulate how
the vacuum filter technology changes the characteristics of a waste stream:
1. Solids retention efficiency of the vacuum filter during
operation is 100 percent. The filtrate drawn off by the vacuum
is free of suspended solids. All suspended solids are retained
in the filter cake with some residual liquor.
2. The filtrate is returned to the industrial process. The solids
cake and residual filtrate are treated as the hazardous waste
stream.
3. The addition of a filter aid as a preconditioning step is
optional:
ICF Incorporated
-------
EXHIBIT 1A-4
VESSEL CHARACTERISTICS AND RELEASE RATES
FOR VACUUM FILTER
Treatment Vessel Characteristics
vessel type
vessel dimensions
retention time (hours)
ag i ta t ion
open
drum diameter: lm
vat width: 2m
3
s low
drum submersion: 1/3 of drum
filtrate rate: 22,000 kg/day m
evaporative surface area (meters)
M (fraction of influent sludge
contained in vessel)
1.365 (10 ) • S(l) • (I — S(3 ))
-1
I .25 ( 10 )
O
3
O
o
"O
o
-I
O)
f+
(D
a
Environmental Releases
aeration loss fraction
total aeration fraction
routine spillage fraction:
(a) routine maintenance
(b) m ixer sp I ash i ng
(c) materials handling
total routine spillage fraction
accidental spillage fraction:
best case, spills/year
worst case, spills/year
typical case fraction
total typical fraction
overall loss fraction
(sum of aeration, routine spi
age, and accidental spillage
fract i ons)
>
i—*
i
00
10
-6
10
-6
3.3(10 )
-6
4.3(10 )
-4
1.11(10 )
-4
1.1(10 )
-------
Al-19
a. If the sludge was not preconditioned, the resulting filter
cake is 15 percent solids and 85 percent filtrate by mass.
b. If the sludge was preconditioned in order to enhance
filterability, the suspended solids content of the sludge is
assumed to be increased by 75 percent owing to the presence
of the filter aid. The filter cake composition is 30
percent solids and 70 percent filtrate by mass.
4. The hazardous component X need not be completely in the
suspended solids phase. If, however, less than 80 percent of X
is present as solids, the model advises the user to select a
more suitable process alternative.
5. No phase changes (solids precipitation or liquid vaporization)
are expected to occur during the filtration process.
c. Mass Balance Algorithm
The user must state whether or not the sludge was preconditioned by the
addition of a filter aid. The statement sets the following values for the
variables ALPHA and BETA:
ALPHA = solids mass fraction in filter cake
BETA = residual liquor mass fraction in filter cake
ieif the sludge was not preconditioned:
ALPHA = 0.15; BETA = 0.85
*if the sludge was preconditioned:
ALPHA =0.3; BETA =0.7
The user is referred to the discussion of the mass balance algorithm for
Filter Press for the remainder of the vacuum filter algorithm. The equations
are identical except for the values of ALPHA and BETA defined above.
ICF Incorporated
-------
Al-20
5. EVAPORATION/DRYING
a. Description
Evaporation and drying are used to dewater sludge streams that do not
contain volatile species, particularly those with a high concentration of
solids. The sludge is fed into the tank, where it is heated by steam coils
using relatively low-temperature steam. The underflow is drawn off
continuously from the side opposite the feed. Slow agitation is provided for
uniform evaporation.
A summary of the important treatment vessel characteristics used in
estimating the individual and overall loss fractions to the environment is
given in Exhibit 1A-5.
b. Assumptions
We made the following assumptions and/or simplifications to simulate how
the evaporation/drying technology changes the characteristics of a waste
stream:
1. All suspended solids and dissolved species are retained in the
hazardous waste stream. Only water is driven off during
treatment. The ratio of the vapor pressure of hazardous species
X to that of water is tested prior to execution of the unit
process. If this ratio is greater than or equal to 0.1, then X
could be evaporated, and the model instructs the user to choose
a more appropriate treatment technology.
2. The target for the dryer (underflow) product is 70 percent
solids by mass.
a. If this criterion can be met by evaporating only water, then
the product's solids mass fraction is .70. The residual
liquid is composed of all the dissolved nonwater species
(including X) and water.
b. If this criterion cannot be met by only evaporating water,
then the dryer product is assumed to be composed of all the
influent nonwater species in their original proportions. No
water is present.
c. If the influent solids concentration is greater than 70
percent, the model instructs the user to choose a more
appropriate treatment technology.
3. There is no change in phase (liquid vaporization or solids
precipitation) during drying.
ICF Incorporated
-------
Al-21
EXHIBIT 1A-5
VESSEL CHARACTERISTICS AND RELEASE RATES FOR EVAPORATION/DRYING
Treatment Vessel Characteristics
vessel type
characteristic dimensions
retention time (hours)
agitation
evaporative surface area (meters)
M (fraction of influent sludge
contained in vessel)
open
H = D
1
slow
1.109x10-* • s(l)2/3
4.17xl0-2
Environmental Releases
aeration loss fraction
total aeration fraction
routine spillage fraction:
(a) routine maintenance
(b) mixer splashing
(c) materials handling
total routine spillage fraction
accidental spillage fraction:
best case, spills/year
worst case, spills/year
typical case fraction
total typical fraction
overall loss fraction
(sum of aeration, routine spill-
age, and accidental spillage
fractions)
10-7
10-6
10-6
1
24
7.712x10-"
7.7x10-
ICF Incorporated
-------
Al-22
c. Mass Balance Algorithm
1. The program first tests the volatility of X relative to that of water
to see if evaporation/drying is appropriate for this waste stream:
"if S(6) > .1, hazardous component X is too volatile to be
treated by evaporation/drying. The user is instructed to choose a
more appropriate treatment technology.
"if S(6) < .1, hazardous component X can be safely treated by
evaporation/drying, and the program continues to the next step.
2. The program then tests the influent suspended solids to see if the
final treated solids/liquid criteria can be met:
*if S(3) ^ 0.7, the influent suspended solids content is too
high for evaporation/drying to be feasible. The user is instructed
to choose a more appropriate treatment technology.
*if S(3) < 0.7, evaporation/drying is appropriate, and the
program continues to the next step.
3. The proportions of suspended solids and dissolved nonwater species is
then tested to see if the 70 percent solids criteria can be reached without
evaporating the nonwater species. The value of a variable LIQ is determined:
LIQ = mass fraction of liquid that would be obtained in the
RCRA stream if all the water were evaporated
LIQ = [mass of influent dissolved nonwater species]
* [total mass of RCRA stream]
LIQ = [mass of influent dissolved nonwater species]
* [mass of influent suspended solids]/0.70
LIQ = S(4)/S(2)
a. If LIQ > 0.3, then all the water is driven off: S(15)' = 0
*the resulting RCRA stream is composed solely of all the
influent suspended solids and dissolved species: S(l)' = S(14)
"the mass fractions of the nonwater species are determined from
the appropriate flows:
S(4)1 = S(4) • S(l)/S(2) • SCD = S(4)/S(2)
S(3)1 = S(3)/S(2)
ICF Incorporated
-------
Al-23
b. if LIQ ^ 0.3, then the 70 percent solids criteria can be
attained without evaporating the nonwater species.
"the RCRA stream solids mass fraction is set equal to 0.70:
S(3)1 = 0.70
*the total RCRA stream flow is determined from the solids mass
fraction:
[mass fraction of solids] = [influent solids flow]/[flow of RCRA stream]
.70 = S(3) • S(l)/S(l)'
therefore,
S(l)* = S(3) • S(1)/.70
"the mass fraction of dissolved nonwater species is determined
from the appropriate flows:
S(4)1 = S(4) • S(l)/S(l)'
4. The total nonwater mass fraction is the sum of that for the suspended
solids and dissolved nonwater species:
S(2)1 = S(3)1 + S(4)1
5. The mass flow of X in each phase remains the same:
S(10)' = S(10)
S(ll)' = S(11)
S(12)1 = S(12)
6. The mass fraction of X and the proportions of X in the suspended
solids and dissolved phases are determined from the appropriate mass flows:
S(7) 1 = S(10)7S(1)'
S(8)1 = S(ll)'/S(10)'
S(9)1 = S(12)1/S(10)'
ICF Incorporated
-------
Al-24
6. AIR STRIPPING
a. Description
Air stripping is a mass transfer operation that can reduce the
concentration of dissolved gases or vapors in an industrial sludge stream.
The two most common uses are:
• removal of volatile organics for recovery or to decrease the
load on downstream waste treatment processes;
• removal of dissolved noxious gases to prevent air pollution
problems.
Packed towers are commonly used for air-stripping operations. The
filtered sludge is distributed evenly across the top packing surface and
allowed to trickle down around the packing. Air is bubbled up through the
packing countercurrent to the influent sludge. The large surface area
provided by the column packing allows prolonged contact between the air and
the sludge. Volatile compounds are driven from the sludge into the air by
virtue of the difference in the concentrations between the two phases.
A summary of the important treatment vessel characteristics used in
estimating the individual and overall loss fractions to the environment is
given in Exhibit 1A-6.
b. Assumptions
There are two major options for utilizing the air-stripping treatment
unit, depending upon the physical properties of hazardous component X. These
options and the corresponding assumptions are discussed below.
Hazardous Component X Is Volatile
When X is relatively volatile compared to the bulk of the influent sludge,
air stripping can be used to remove a portion of X from the sludge for
recycling to the industrial process. The remainder of X is retained in the
hazardous waste stream.
The packed column is not suited to solids loading and is therefore
preceded by a filter to remove suspended particles. The user may specify
whether the filtered solids are sent to the hazardous waste stream or the
process return stream. This technology is inappropriate for influents in
which hazardous component X is in the suspended solids phase. In this case,
the model advises the user to select a more appropriate treatment technology.
The assumptions made for this treatment unit are as follows:
ICF Incorporated
-------
Al-25
EXHIBIT 1A-6
VESSEL CHARACTERISTICS AND RELEASE RATES FOR AIR STRIPPING
Treatment Vessel Characteristics
vessel type
vessel dimensions
retention time (hours)
agitation
evaporative surface area (meters)
M (fraction of influent sludge
contained in vessel)
closed
N/A
.25
N/A
N/A
1.04xl0-2
Environmental Releases
aeration loss fraction
total aeration fraction
routine spillage fraction:
(a) routine maintenance
(b) mixer splashing
(c) materials handling
total routine spillage fraction
accidental spillage fraction:
best case, spills/year
worst case, spills/year
typical case fraction
total typical fraction
overall loss fraction
(sum of aeration, routine spill-
age, and accidental spillage
fractions)
N/A
N/A
2
12
1.319x10-"
1.3x10-'
ICF Incorporated
-------
Al-26
1. All of the influent X is in the dissolved phase. During
stripping, 95 percent of the influent X is removed by the air
stream while 5 percent remains in the hazardous waste stream.
Also stripped by the air is an additional fraction of the
dissolved nonwater species (not including X) specified by the
user.
2. Water is evaporated from the sludge in amounts governed by its
volatility relative to that of X.
3. The moisture content of the influent air is negligible.
Hazardous Component X Is Nonvolatile
When X is relatively nonvolatile compared to the bulk of the influent
sludge, air stripping can be used to reduce the concentration of other
compounds that could interfere with downstream treatment processes or present
air pollution problems. In either case, all of X is retained in the hazardous
waste stream. The assumptions made for this treatment unit are as follows:
1. A user-specified fraction of the dissolved nonwater species (not
including X) is removed by the air stream and returned to the
industrial process. ' All of X, all influent water, and all the
remaining dissolved species are retained in the hazardous waste
stream.
2. The moisture content of the influent air stream is negligible.
c. Mass Balance Algorithm
Hazardous Component X Is Volatile
1. If the influent sludge contains X in the suspended solids phase, this
technology is not feasible. The program advises the user to select a more
appropriate treatment. Otherwise, the user must state whether the prefiltered
solids are returned to the hazardous waste stream. This statement determines
the value of the variable SOLIDS:
-if the prefiltered solids are added to the process return
stream:
SOLIDS = 0
*if the prefiltered solids are added to the hazardous waste
stream:
SOLIDS = S(3) • SCI)
ICF incorporated
-------
Al-27
2. The user must input a value for BETA, the fraction of the dissolved
nonwater species (not including X) that is to be stripped by the air stream
with X. The mass flow of the nonwater species is determined using the
variables BETA and SOLIDS:
[Flow of [influent dissolved [95 percent of
nonwater species] = species] - influent X]
[Dissolved species [Filtered solids
stripped with X] + if added]
S(14)1 = S(4)*S(1) - .95*S(10) - BETA • [S(4) • S(l) - S(10)] + SOLIDS
3. The amount of water evaporated from the influent sludge is determined
by the partial pressure exerted by water relative to that of hazardous
component X. The amount of water remaining in the sludge stream is the
difference between the initial and evaporated quantities:
[Mass flow of water [influent water [amount of X
in hazardous waste stream] = in sludge] - evaporated]
[partial pressure of water]/[partial pressure of X]
S(15)1 = S(15) - .95 • S(10) [Y(H20)/Y(x)»[P(H20)/P(x)]
where P(i) = vapor pressure of species i
Y(i) = mass fraction of species i in influent sludge
S(15)1 = S(15) - .95»S(l)MS(15)/S(l)]t[S(10)/S(l)]-[l/S(6)]
= S(15) - .95 • S(15)/S(6)
= S(15) • [1 - .95/S(6)]
4. The total hazardous waste stream flow is the sum of the flows of water
and the nonwater species:
S(1)' = S(14) ' + S(15) '
5. The mass fractions of suspended solids and dissolved nonwater species
are determined from the appropriate mass flows:
S(3)* = S0LIDS/S(1)'
S(4)' = [S(14)* - SOLIDS]/S(1)1
ICF Incorporated
-------
Al-28
6. The mass fraction of the total nonwater species is the sum of the
fractions for the suspended and dissolved portions:
S(2)* = S(3)1 + S(4) 1
7. The mass flow of X in the dissolved portion and in the total hazardous
waste stream is 5 percent of that for the influent sludge. No X is present in
the solid phase.
SCIO) ' = .05 • S(10)
S(ll)' = .05 • S(10)
S(12)' = 0
8. The mass fraction of X and the proportions of X in the dissolved and
suspended phases are determined from the appropriate mass flows:
S(7)1 = S(10)'/S(l)'
S(8)' = 1
S(9) 1 = 0
Hazardous Component X Is Nonvolatile
1. If X is present in the suspended solids phase, the prefiltered solids
are automatically returned to the hazardous waste stream. Otherwise, the
program prompts the user to state whether the prefiltered solids are returned
to the hazardous waste stream. The answer to this question determines the
value of the variable SOLIDS:
"if the prefiltered solids are added to the process return
stream:
SOLIDS = 0
,vif the prefiltered solids are added to the hazardous waste
stream:
SOLIDS = S(3) • S(l)
2. The user must input a value for BETA, the fraction of the dissolved
nonwater species (not including X) that is to be stripped by the air stream.
The mass flow of the nonwater species is determined using the variables BETA
and SOLIDS:
[Flow of nonwater [influent dissolved [dissolved species [filtered
species] = species] - stripped] + solids, if
added]
S(14)' = S(4) • S(1) - BETA • [S(4) • SCD - S(ll)] + SOLIDS
ICF Incorporated
-------
Al-29
3. The flow of water is unaltered since none is evaporated by the air
stream. The total hazardous waste stream flow is the sum of the flows of
water and the nonwater species:
S(l)' = S(14)1 + S C15)*
4. The mass fractions of suspended solids and dissolved nonwater species
were determined from the appropriate mass flows:
S(3)' = S0LIDS/S(1)'
S(4)* = [S(14)' - SOLIDS]/S(l)1
5. The mass fraction of the total nonwater species is the sum of the
fractions for the suspended solids and dissolved nonwater species:
S(2)1 = S(3)' + S(4)1
6. The mass flows of X in both phases remain unchanged:
S(10) ' = S(10)
S(11) * = S(11)
S(12)1 = S(12)
7. The mass fraction of X and the proportions of X in the dissolved and
suspended phases are determined from the appropriate mass flows:
S(7)1 = SC10) 1 /S(l)'
S (8) * = S(ll)7s(10)'
S(9)1 = S(12)1/S(10)1
ICF Incorporated
-------
Al-30
7. STEAM STRIPPING
a. Description
Steam stripping is a mass transfer operation used to remove fairly
volatile organic liquids from a sludge for recovery or to decrease the load on
downstream waste treatment processes. Steam stripping is, in principle,
similar to air stripping, but can remove species with lower vapor pressures by
virtue of the heat and altered equilibrium provided by the superheated steam.
Sieve tray stripping columns with relatively large still pots are used for
industrial sludge steam-stripping operations. When the influent sludge
contains solids, it is fed directly to the still pot at the base of the
column; otherwise, it may be fed onto a tray further up the column for
improved recovery efficiency. As the steam bubbles up through perforations in
the trays, the volatile species are stripped from the sludge and sent to a
condenser for recovery. The heavier components are withdrawn from the still
pot.
A summary of the important treatment vessel characteristics used in
estimating the individual and overall loss fractions to the environment is
given in Exhibit 1A-7.
b. Assumptions
There are two major options for utilization of the steam-stripping
treatment unit, depending upon the physical properties of hazardous component
X. These options and the corresponding assumptions are discussed below.
Hazardous Component X Is Volatile
When X is relatively volatile compared to the bulk of the influent sludge,
steam stripping can be used to remove a portion of X from the sludge for
recycling to the industrial process. The remainder of X is retained in the
hazardous waste stream.
Steam-stripping system is inappropriate for influents in which hazardous
component X is in the suspended solids phase. In this case, the model advises
the user to select a more appropriate treatment technology. The assumptions
made for this treatment unit are as follows:
1. All of the influent X is in the dissolved phase. During
stripping, 90 percent of the influent X is removed by the steam
while 10 percent remains in the hazardous waste stream. Also
stripped by the steam is an additional fraction of the dissolved
nonwater species (not including X) specified by the user.
2. Water is not evaporated from the influent sludge owing to the
presence of the steam.
ICF Incorporated
-------
Al-31
EXHIBIT 1A-7
VESSEL CHARACTERISTICS AND RELEASE RATES FOR STEAM STRIPPING
Treatment Vessel Characteristics
vessel type
vessel dimensions
retention time (hours)
agitation
evaporative surface area (meters)
M (fraction of influent sludge
contained in vessel)
closed
N/A
.5
N/A
N/A
2.08xl0-2
Environmental Releases
aeration loss fraction
total aeration fraction
routine spillage fraction:
(a) routine maintenance
(b) mixer splashing
(c) materials handling
total routine spillage fraction
accidental spillage fraction:
best case, spills/year
worst case, spills/year
typical case fraction
total typical fraction
overall loss fraction
(sum of aeration, routine spill-
age, and accidental spillage
fractions)
N/A
N/A
1
12
2.14x10-''
2.1x10-
ICF Incorporated
-------
Al-32
Hazardous Component X Is Nonvolatile
When X is relatively nonvolatile compared to the bulk of the influent
sludge, steam stripping can be used to reduce the concentration of other
compounds that could interfere with downstream treatment processes. All of X
is retained in the hazardous waste stream. The assumptions made for this
treatment unit are as follows:
1. A user-specified fraction of the dissolved nonwater species (not
including X) is removed by steam stripping to the condenser and
returned to the industrial process. All of X, all influent
solids, and the remaining dissolved species are retained in the
hazardous waste stream drawn from the .still pot.
2. Water is not evaporated from the influent sludge because of the
presence of the steam.
c. Mass Balance Algorithm
Hazardous Component X Is Stripped by the Steam
1. The user must input a value for BETA, the fraction of the dissolved
nonwater species (not including X) that is to be stripped by the steam with X.
The mass flow of the nonwater species is determined using the value of BETA:
[Flow of [influent nonwater [90 percent of [dissolved species
nonwater species] = species] - influent X] - stripped with X]
S(14)1 = S(2) • S(l) - .90 • S(10) - BETA • [S(4) • S(l) - S(10)]
2. The flow of water remains unchanged:
S(15)' = S(15)
3. The total RCRA stream flow is the sum of the flows of water and
nonwater species:
S(1)' = S(14)1 + S(15)*
4. The mass fraction of dissolved species is determined using the value
of BETA:
ICF Incorporated
-------
Al-33
[Flow of [influent dissolved [90 percent of [dissolved species
dissolved species] = species] - influent X] - stripped with X]
= S(4) • S(l) " -90 • S(10) - BETA • [S(4) • S(l) - S(10)]
Dividing by S(1)1,
S(4)1 = [S(4)*S(1) - .90 • S(10) - BETA • [S(4) • S(l) - S(10)]]/S(1)1
5. The mass fraction of suspended solids is decreased:
S(3)1 = S(3) • S(l)/S(l)'
6. The mass fraction of nonwater species is the sum of the suspended
solids and dissolved portions:
S(2)' = S(2)* + S(4)*
7. The mass flow of X is decreased by 90 percent:
S(10)' = .10 • S(10)
S(ll)' = .10 • S(10)
S(12) ' = 0
8. The mass fraction of X and the proportion of X in the suspended solids
and dissolved phases are determined from the appropriate mass flows:
S(7) 1 = S(10)7S(1)'
S(8)1 = S(ll)'/S(10)'
S(9)1 = S(12)1/S(10)'
Hazardous Component X Is Not Stripped by the Steam
1. The user must input a value for BETA, the fraction of the dissolved
nonwater species (not including X) that is to be stripped by the steam. The
mass flow of the nonwater species is determined using the value of BETA:
Fflow of nonwater [influent nonwater [dissolved species
species] = species] - stripped by steam]
S(14)' = S(2) • S(l) - BETA • [S(4) • S(1) - S(10)]
2. The flow of water remains unchanged:
S(15) 1 = S C15)
ICF Incorporated
-------
Al-34
3. The total RCRA stream flow is the sum of the flows of water and
nonwater species:
S(l)' = S(14)1 + S(15)1
4. The mass fraction of dissolved species is determined using the value
of BETA:
[Flow of dissolved [influent dissolved [dissolved species
species] = species] - stripped by steam]
= S(4) • S(l) - BETA • [S(4) • S(1) - S(10)]
Dividing by S(l)',
S(4)1 = [S(4) • S(l) - BETA • [S(4) • S(l) - S(10)]]/S(1)'
5. The mass fraction of suspended solids is decreased:
S(3)' = S(3) • S(l)/S(l)'
6. The mass fraction'of nonwater species is the sum of the suspended
solids and dissolved portions:
S(2)' = S(3)1 + S(4) 1
7. The mass flow of X in both phases remains unchanged:
S(10)' = S(10)
S(11)' = S(ll)
S(12)' = S c12)
8. The mass fraction of X and the proportion of X in the suspended solids
and dissolved phases are determined from the appropriate mass flows:
S(7)' = S(10)'/S(1)'
S(8) 1 = S(11)7S(10)'
S(9)' = S(12)1 /S(10) 1
1CF Incorporated
-------
Al-35
8. SOLVENT EXTRACTION
a. Description
Solvent extraction achieves partial separation of the liquid constituents
of an industrial sludge by virtue of their different solubilities in two
insoluble liquid phases. It is used to extract specific compounds from the
sludge for recycling or to minimize their effect on treatment units downstream.
A mixer-settler combination is commonly used for industrial sludge
treatment operations. The baffled mixing vessel contains a rotating impeller
for mixing and for solvent dispersion. The vessel is first filled with
sludge. Agitation is begun, and the solvent is injected into the tank. When
a fine dispersion of solvent in sludge is achieved (usually at a ratio of 75
parts solvent to 100 parts sludge, or less), flow of the mixture to the
decanter is begun. The decanter allows the insoluble two phases to settle and
coalesce quietly. A nozzle is used periodically to remove dust and suspended
solids that settle at the interface and impede separation. The extraction
solvent and its dissolved components are called the extract stream, while the
remaining sludge is called the raffinate. The mixer-settler may be operated
continuously or as a batch process.
A summary of the important treatment vessel characteristics used in
estimating the individual and overall loss fractions to the environment is
given in Exhibit 1A-8.
b. Assumptions
There are two major options for utilization of the solvent extraction
treatment unit, depending upon the physical properties of hazardous component
X. These options and their corresponding assumptions are discussed below.
Hazardous Component X Is Soluble in Extraction Solvent
For some industrial sludges, it is feasible to dissolve and remove
selectively a portion of X with an extraction solvent for recycling to the
industrial process. The remainder of X is retained in the raffinate as the
hazardous waste stream. The assumptions made for this treatment unit are as
follows:
1. In the process return stream, 95 percent of both the suspended
solids and dissolved portions of X are extracted with solvent.
The remaining 5 percent of X is retained in the raffinate
hazardous waste stream. Since the extraction efficiency is
applied individually to both the suspended solids and dissolved
portions of X, the ratio of X in the suspended and dissolved
phases of the raffinate is the same as that for the influent
sludge.
ICF Incorporated
-------
EXHIBIT 1A-8
VESSEL CHARACTERISTICS AND RELEASE RATES
FOR SOLVENT EXTRACTION
Contact Tank
Coalescent tank
O
3
O
o
-I
o
-J
0)
r+
-------
Al-37
2. Also extracted with the solvent are additional fractions of the
suspended solids and dissolved nonwater species (not including
X) specified by the user.
3. None of the solvent is retained in the hazardous waste stream.
The user must specify whether the influent water is extracted or
retained in the raffinate.
Hazardous Component X Is Not Soluble in Extraction Solvent
For some industrial sludges, it is preferable to dissolve and remove a
portion of the influent material, leaving all of X behind in the raffinate, in
order to reduce the load on downstream treatment units. The assumptions made
for this treatment unit are as follows:
1. All of X is retained in the raffinate hazardous waste stream.
All of the extraction solvent is returned to the industrial
process. Removed with the solvent are fractions of the
suspended solids and dissolved nonwater species (not including
X) specified by the user.
2. All components retained in the hazardous waste stream are
present in the same phase (dissolved or suspended solids) in
which they entered the solvent extraction system.
3. The user must specify whether the influent water is extracted or
retained in the raffinate.
c. Mass Balance Algorithm
Hazardous Component X Is Soluble in the Extraction Solvent
1. The user must specify the values of two variables:
PSI1 = the fraction of the suspended solids (not including
X) that is extracted with the solvent
PSI2 = the fraction of the dissolved nonwater species (not
including X) that is extracted with the solvent
The total flow of the nonwater species is determined using these values of
S(14)' = S(14) - PSI1[S(3)*S(1)-S(12)]-PSI2*[S(4),S(1)-S(11)]-.95*S(10)
PSI1 and PSI2:
[Flow of
nonwater species]
[nonwater species,
[influent nonwater other than X, [95 percent of
species] - that are extracted] - influent X]
ICF Incorporated
-------
Al-38
2. The user must state whether the influent water follows the extract or
the raffinate. The answer to this question determines the flow of water in
the RCRA stream:
-if water follows the extract, S(15)1 = 0
-if water follows the raffinate, S(15)' = S(15)
3. The total RCRA stream flow is the sum of the flows of water and the
nonwater species:
S(l)* = S(14)1 + S(15) '
4. The mass fraction of suspended solids is determined using the value of
PSI1:
[95 percent of [solids
[Flow of [influent suspended suspended portion extracted
suspended solids] = solids] - of X] - with X]
= S(3)*S(1) - .95*S(12) - PSI1-[S(3)»S(1) - S(12)]
= (1 - PSI1) • S(3) • S(1) - (.95 - PSI1) • S(12)
Dividing by S (1) 1 ,
S(3)1 = [(1-PSI1) • S(3) • S(1) - (.95 - PSI1) • S(12)]/S(1)1
Likewise, the mass fraction of dissolved nonwater species is determined using
PSI2:
S(4)1 = [(1 - PSI2) • S(4) • S(l) - (.95 - PSI2) • S(ll)]/S(l)'
5. The mass fraction of the total nonwater species is the sum of the
fractions for the suspended and dissolved portions:
S(2)* = (S3)' + S(4)*
6. The mass flows of X in both phases are reduced by 95 percent:
S(10) = .05 • S(10)
S(11) = .05 • S(ll)
S(12) = .05 • S(12)
7. The mass fraction of X and the proportions of X in the dissolved and
suspended phases are determined from the appropriate mass flows:
S(7)1 = S(10)'/S(1) '
S(8)1 = S(11)'/S(10) '
S(9)1 = S(12)'/S(10)1
ICF Incorporated
-------
Al-39
Hazardous Component X Is Not Soluble in the Extraction Solvent
1. The user must specify the values of two variables:
PSI1 = the fraction of the suspended solids (not including
X) that is extracted with the solvent
PSI2 = the fraction of the dissolved nonwater species (not
including X) that is extracted with the solvent
The total flow of the nonwater species is determined using these values of
PSI1 and PSI2:
[Flow of nonwater [influent nonwater [nonwater species
species] = species] - extracted]
S(14)1 = S(14) - PSI1[S(3)*S(1)-S(12)] - PSI2[S(2)*S(1)-S(ll)]
2. The total RCRA stream flow is the sum of the flows of water and the
nonwater species:
S(l)' = S(14)1 + S(15)1
3. The mass fraction of suspended solids is determined using the value of
PSI1:
[Flow of [influent suspended [solids
suspended solids] = solids] - extracted]
= S(3) • S(l) - PSI1 • [S(3) • S(l) - S(12)]
Dividing by S(l)1,
S(3)1 = [(1 - PSI1) • S(3) • S(l) + PSI1 • S(12)]/S(l)'
Likewise, the mass fraction of dissolved nonwater species is determined using
PSI2:
S(4)1 = [(l-PSI2)*S(4),S(l)-(.95-PSI2)#S(11)]/S(1) 1
4. The mass fractions of the total nonwater species are the sum of the
fractions for the suspended and dissolved portions:
S(2)' = S(3)1 + S(4)'
ICF Incorporated
-------
Al-40
5. The mass flows of X in both phases remain unchanged:
s cio)1 = s(10)
S c11)' = S(11)
S c12)' = S(12)
6. The mass fraction of X and the proportions of X in the dissolved and
suspended phases are determined from the appropriate mass flows:
S(7) ' = SC10)'/S(1)1
S(8) ' = S(11)'/S(10) '
S(9)1 = S(12)*/S CIO) 1
ICF Incorporated
-------
Al-41
9. LEACHING
a. Description
Leaching is used in industrial sludge treatment for selectively dissolving
and separating a portion of the influent suspended solids from the bulk
sludge. The influent sludge is blended with solvent during rapid mixing in a
contact vessel, with retention times ranging from 30 minutes to 2 hours.
After dissolution, a conventional clarifier is used to separate the effluent
from the remaining solids. The underflow solids are washed with additional
solvent to remove the residual liquor.
This treatment technology should not be used for sludges in which the
influent X is not predominantly in the solids phase, as there are more
appropriate separation techniques for these streams. The solubility of X in
the leach solvent should be sufficiently different from that of the other
solids to promote the selectivity of the separation.
A summary of the important treatment vessel characteristics used is
estimating the individual and overall loss fractions to the environment is
given in Exhibit 1A-9.
b. Assumptions
There are two major options for utilization of the leaching unit,
depending on whether X is soluble or insoluble in the leach solvent. These
options and the corresponding assumptions are discussed below.
Hazardous Component X Is Soluble in Leach Solvent
For some industrial sludges with a high content of solids, it is feasible
to dissolve and remove a portion of X from the sludge retained in the
clarifier underflow as the hazardous waste stream. The assumptions made for
this treatment unit are as follows:
1. The clarifier effluent contains 98 percent of influent X, plus
an additional fraction of the influent suspended solids
specified by the user. All influent water and dissolved
nonwater species are removed in the effluent and subsequent
washings of solids and are included in the process return stream.
2. The hazardous waste stream underflow contains 2 percent of
influent X and the undissolved portion of the suspended solids.
The underflow is 2 percent solids by mass, and 98 percent
residual solvent from the washing process.
3. The solvent used may be either water or another solvent, as
specified by the user.
ICF Incorporated
-------
EXHIBIT 1A-9
VESSEL CHARACTERISTICS AND RELEASE RATES
FOR LEACHING
Contact Tank
CI a r i f i e r
Treatment Vessel Characteristics
vessel type
vessel dimensions
retention time (hours)
ag i tat ion
evaporative surface area (meters)
M (fraction of influent sludge
contained in vessel)
open
H = 2D
0.5
rap i d
-2
4.40 I( 10 )
-2
1.04(10 )
S( I )
2/3
open
H = . 5D
2
s I ow
- I
2.794(10 )
-2
8.33(10 )
S( I )
2/3
o
3
o
o
-f
•o
o
-J
ft)
r+*
(t>
Q.
Environmental Releases
aeration loss fraction
total aeration fraction
routine spillage fraction:
(a) routine maintenance
(b) mixer splashing
(c) materials handling
total routine spillage fraction
accidental spillage fraction:
best case, spills/year
worst case, spills/year
typical case fraction
total typical fraction
overall loss fraction
(sum of aeration, routine spill-
age, and accidental spillage
fract i ons)
to
10
10
-5
D
-5
-4
10
-4
( 10 )
.5
12
-5
9.616(10 )
-4
3.0(10 )
-4
4.1(10 )
10
-6
10
I.997(10 )
-------
Al-43
Hazardous Component X Is Not Soluble in the Leach Solvent
For some industrial sludges, it is necessary to dissolve and remove a
portion of the influent suspended solids, leaving all of X behind in the
clarifier underflow, in order to reduce the load on the downstream treatment
units. The assumptions made for this treatment unit are as follows:
1. The clarifier effluent contains a fraction of the influent
suspended solids (not including X) specified by the user. All
influent water and dissolved nonwater species are removed in the
effluent and subsequent washings of solids and are included in
the process return stream.
2. The hazardous waste stream underflow contains all of the
influent X and the undissolved portion of the suspended solids.
The underflow is 2 percent solids by mass and 98 percent
residual solvent from the washing process.
3. The solvent used may be either water or another solvent, as
specified by the user.
c. Mass Balance Algorithm
Hazardous Component X Is Soluble in Leach Solvent
1. The mass fraction of solids in the washed clarifier underflow is set
equal to 2 percent:
S(3)1 = 0.02
2. The user must state whether the leaching agent used is water or
another solvent. The answer to this question determines the water flow and
the mass fraction of dissolved nonwater species in the RCRA stream:
,vif the leaching agent is water, all residual liquid in the washed
clarifier underflow is water:
S(15)1 = 0.98 • S(l)'; S(14) = 0
*if the leaching agent is another solvent, all residual liquid in the
washed clarifier underflow is solvent:
S(15)* = 0; S(14)1 = 0.98
3. The user must specify the value of PSI, the fraction of the influent
suspended solids (not including X) that is extracted with the leaching agent.
The new RCRA stream flow is calculated according to the described solids mass
fraction:
ICF Incorporated
-------
Al-44
Total mass of total mass of
undissolved suspended = solids in cake
solids
[(1-PSI)-S(3).S(1)-S(12)] + 0.02 • S C12) = S (1) '
therefore,
S(1)' = 50 • [(1-PSI).(S(3).S(1)-S(12)] + 0.02 • SC12)
4. The mass fraction of nonwater species is equal to the sum of those for
the suspended solids and dissolved species:
S(2)' = S(3)' + S(4)1
5. The mass flow of nonwater species is determined from the appropriate
mass fraction and flow:
S(14)1 = S(2)' • S(l)'
6. The flow of X in the RCRA stream is 2 percent of the influent solid X,
all in the suspended solids phase:
S C10) ' = .02 • S (12)
SCH) ' = 0
S(12) 1 = .02 • Sc 12)
7. The mass fraction of X and the proportions of X in the dissolved and
suspended phases are determined from the appropriate mass flows:
S(7)' = S C10)'/S(1)'
S(8)' = S(ll)'/S(10)'
S(9)1 = S(12)'/S(10)'
Hazardous Component X Is Insoluble in Leach Solvent
The development for this treatment scenario is the same as for the
previous case in which X is soluble in the leaching solvent, with the
following exceptions:
1. The new RCRA stream flow includes all the influent X solids:
Total mass of total mass of
undissolved suspended = solids in cake
solids
[(1-PSI1)-S(3) • S(l)-S(12)] + S(12) = .02 • S(l)'
ICF Incorporated
-------
Al-45
therefore,
SCI)' = 50 • [(1-PSI1) • S(3) • S(1)-S(12)] + S(12)
2. The flow of X in the RCRA stream is equal to the influent X in the
solids phase. There is no dissolved X in the underflow:
S(10)1 = S(12)
S(11)' = 0
S(12)1 = S(12)
ICF Incorporated
-------
Al-46
10. DISTILLATION
a. Description
Fractional distillation is used to separate the constituents of a liquid
mixture, by virtue of their differences in vapor pressure, for the purpose of
removing and reclaiming a portion of the more volatile species. A properly
designed and operated distillation unit can yield nearly complete recovery of
the volatile components from the mixture.
The continuous countercurrent distillation column is commonly used in
industrial sludge treatment. A sludge containing solids or very heavy
(nonvolatile) liquids is fed directly to the reboiler; otherwise, the feed is
typically injected near the middle tray. The influent sludge is partially
vaporized upon contact with the contents of the column, yielding a vapor
containing higher concentrations of volatile species and a liquid residue
richer in less volatile components. As the vapor rises through the column, it
comes into continuous contact with a condensed portion of the vapor flowing
countercurrently down the column along the trays. This process secures a
vapor product enriched in volatile liquids. The liquid fraction of the feed
may be contacted countercurrently with a stream of vapor from the reboiler in
order to strip from it all traces of volatile components. The depleted heavy
liquid is then withdrawn from the reboiler.
A summary of the important treatment vessel characteristics used in
estimating the individual and overall loss fractions to the environment is
given in Exhibit 1A-10.
b. Assumptions
There are two major options for utilization of the distillation unit,
depending upon the vapor pressure of hazardous component X relative to those
of the other species in the sludge stream. These options and the
corresponding assumptions are discussed below.
Hazardous Component X Is Volatile
When X is relatively volatile compared to the bulk of the influent sludge,
distillation can be used to remove a portion of X from the sludge for
recycling to the industrial process. The remainder of X would be retained in
the hazardous waste stream.
When the influent sludge contains solids, it is fed into the reboiler;
otherwise, it is fed near the center tray. This technology is inappropriate
if hazardous component X is in the suspended solids phase. In this case, the
model advises the user to select a more appropriate treatment technology. The
assumptions made for this treatment unit are as follows:
ICF Incorporated
-------
Al-47
EXHIBIT 1A-10
VESSEL CHARACTERISTICS AND RELEASE RATES FOR DISTILLATION
Treatment Vessel Characteristics
vessel type
vessel dimensions
retention time (hours)
agitation
evaporative surface area (meters)
M (fraction of influent sludge
contained in vessel)
closed
N/A
.5
N/A
N/A
2.08xl0-2
Environmental Releases
aeration loss fraction
total aeration fraction
routine spillage fraction:
(a) routine maintenance
(b) mixer splashing
(c) materials handling
total routine spillage fraction
accidental spillage fraction:
best case, spills/year
worst case, spills/year
typical case fraction
total typical fraction
overall loss fraction
(sum of aeration, routine spill-
age, and accidental spillage
fractions)
N/A
N/A
1
12
2.37x10-"
2.1x10-"
ICF Incorporated
-------
Al-48
1. All of influent X is in the dissolved phase. During treatment,
95 percent of influent X is distilled to the condensate while 5
percent is withdrawn from the reboiler as the hazardous waste
stream. Also distilled is an additional fraction of the
dissolved nonwater species (not including X) specified by the
user. All solids and the remaining dissolved species are
retained in the hazardous waste stream drawn from the reboiler.
2. All influent water will be sent either to the condenser or to
the reboiler, depending on its vapor pressure relative to that
of X.
Hazardous Component X Is Nonvolatile
When X is relatively nonvolatile compared to the bulk of the influent
sludge, distillation may be used to remove a portion of the more volatile
species from the sludge for recycling to the industrial process. All of X is
retained in the hazardous waste stream.
When the influent sludge contains solids, it is fed into the reboiler;
otherwise, it is fed near the center tray. The assumptions made for this
treatment unit are as follows:
1. A user-specified fraction of the dissolved nonwater species (not
including X) is removed by distillation to the condenser and
returned to the industrial process. All of X, all influent
solids, and the remaining dissolved species are retained in the
hazardous waste stream drawn from the reboiler.
2. All influent water will be sent either to the condenser or the
reboiler, depending on its vapor pressure relative to that of X.
c. Mass Balance Algorithm
Hazardous Component X Is Volatile
1. The user must input a value for BETA, the fraction of the influent
dissolved nonwater species (not including X) that is distilled with X. The
mass flow of nonwater species in the RCRA stream is determined using the
variable BETA:
[Flow of [influent nonwater [95 percent of [dissolved species
nonwater species] = species] - influent X] - distilled with X]
S(14)1 = S(2) • S(l) - .95 • S(10) - BETA • [S(4).S(l)-S(10)]
ICF Incorporated
-------
Al-49
2. The flow of water in the RCRA stream is determined by comparing its
vapor pressure to that of X. Since S(6) =
[vapor pressure of X]/[vapor pressure of water]
••"if S(6) < 1.0, then water is as volatile as or more
volatile than X, and all water is distilled to the
condenser:
S(15 ) ' = 0
"if S(6) > 1.0, then water is less volatile than X,
and all water is withdrawn with the RCRA stream:
S(15) ' = S(15)
3. The total RCRA stream flow is the sum of the flows of water and the
nonwater species:
SCI)' = S(14)1 + S(15)'
4. The mass fractions of suspended solids and dissolved nonwater species
are determined from the appropriate mass flows:
S(3)1 = S(3) • S(l)/S(l)'
S(4)' = [S(4) • S(l) - .95 • S(10) - BETA • [S(4) • S(l) - S(10)]]/S(1)'
5. The mass fraction of the total nonwater species is the sum of the
fractions for the suspended and dissolved portions:
S(2)1 = S(3)' + S(4)1
6. The mass flow of X in the dissolved portion and in the total RCRA
stream is 5 percent of that for the influent sludge. No X is present in the
solid phase:
S(10)' = .05 • S(10)
S(ll)' = .05 • S(10)
S(12)1 = 0
7. The mass fraction of X and the proportions of X in the dissolved and
suspended phases are determined from the appropriate mass flows:
S(7)1 = S(10)*/S(l)'
S ( 8 ) ' = 1
S(9) ' = 0
ICF Incorporated
-------
A1 - 5 0
Hazardous Component X Is Nonvolatile
1. The user must input a value for BETA, the fraction of the influent
dissolved nonwater species (not including X) that is distilled. The mass flow
of nonwater species in the RCRA stream is determined using the variable BETA:
[flow of [influent nonwater [dissolved species
nonwater speciesl = species] - distilled with X]
S(14)1 = S(2) • SCI) " BETA • [S(4) • S(l) - S(10)]
2. The flow of water in the RCRA stream is determined by comparing its
vapor pressure to that of X. Since
S(6) = [vapor pressure of X]/[vapor pressure of water]
"if S(6) < 1.0, then water is more volatile than X,
and all the water is distilled to the condenser:
S(15)' = 0
"if S(6) > 1.0, then water is as volatile as or less
volatile than X, and all water is withdrawn with the RCRA
stream:
S(15) ' = S(15)
3. The total RCRA stream flow is the sum of the flows of water and the
nonwater species:
S(l)' = S(14)' + S(15) 1
4. The mass fractions of suspended solids and dissolved nonwater species
are determined from the appropriate mass flows:
S(3)1 = S(3) • S(l)/S(l) '
S(4)1 = [S(4) • S(l) - BETA • [S(4).S(1) - S(10)]]/S(l)'
5. The mass fraction of the total nonwater species is the sum of the
fractions for the suspended and dissolved portions:
S(2)1 = S(3) ' + S(4)1
6. The mass flow of X in both phases remains unchanged:
S(10)' = S(10)
S(11) ' = S(ll)
S (12) 1 = S(12)
ICF Incorporated
-------
A1 - 51
7. The mass fractions of X and the proportions of X in the dissolved and
suspended phases are determined from the appropriate mass flows:
S(7)1 = S(10)'/S(l)'
S(8)1 = SC11)*/S(10)'
S(9)' = S(12)1/S(10)'
ICF Incorporated
-------
Al-52
11. REVERSE OSMOSIS
a. Description
Reverse osmosis concentrates the dilute dissolved species in industrial
sludges by using a differential pressure across a selectively permeable
membrane to remove water. Water is transported through the membrane by
diffusion through vacancies in the molecular structure of the membrane
material. Even organic and ionic species of low molecular weight may be
concentrated by using reverse osmosis, provided that they do not damage the
structure of the membrane.
A tubular reverse osmosis unit is commonly used in industrial processes.
The reverse osmosis membrane is encased by a porous support tube for
structural strength. The membrane tubes are encased in a shell with an outlet
for product water. The influent sludge is fed into one end of the tube under
pressure. Pure product water diffuses through the tube walls and collects in
the shell side of the unit for withdrawal. The concentrated hazardous waste
brine flows out the end of the reverse osmosis tube.
A summary of the important treatment vessel characteristics used in
estimating the individual and overall loss fraction to the environment is
given in Exhibit 1A-11.
b. Assumptions
We made the following assumptions and/or simplifications to simulate how
the reverse osmosis technology changes the characteristics of a waste stream:
1. A total of 80 percent of the influent water is withdrawn from
the reverse osmosis unit as the process return stream. The
remaining water and all nonwater species compose the hazardous
waste stream.
2. No phase changes (solids precipitation or liquid vaporization)
occur during this process.
c. Mass Balance Algorithm
1. The flow of nonwater species remains unchanged. The flow of water is
reduced by 80 percent:
S(14)1 = S(14)
SC15) ' = .20 • S(15)
ICF Incorporated
-------
Al-53
EXHIBIT 1A-11
VESSEL CHARACTERISTICS AND RELEASE RATES FOR REVERSE OSMOSIS
Treatment Vessel Characteristics
vessel type
vessel dimensions
retention time (hours)
agitation
evaporative surface area (meters)
M (fraction of influent sludge
contained in vessel)
closed
N/A
.25
N/A
N/A
1.04xl0-2
Environmental Releases
aeration loss fraction
total aeration fraction
routine spillage fraction:
(a) routine maintenance
(b) mixer splashing
(c) materials handling
total routine spillage fraction
accidental spillage fraction:
best case, spills/year
worst case, spills/year
typical case fraction
total typical fraction
overall loss fraction
(sum of aeration, routine spill-
age, and accidental spillage
fractions)
N/A
1.04x10-g
l.OxlO-6
1
24
l .923x10-''
1.9xl0-u
ICF Incorporated
-------
Al-54
2. The total RCRA stream flow is the sum of that for water and the
nonwater species:
S(3)1 = S(3) • S(l)/S(l)'
S(4)1 = S(4) • S(l)/S(l)1
3. The mass fraction of the total nonwater species is the sum of the
suspended solids and dissolved fractions:
S(2)' = S(3) 1 + S(4)1
4. The mass flow of X in both phases remains unchanged:
S(10)' = S(10)
S(ll)' = S c11)
S c12)' = S(12)
5. The mass fraction of X and the proportions in the suspended solids and
dissolved phases are determined from the appropriate mass flows:
S(7)' = S(10)'/S(1)*; S(8) ' = S(ll)'/S(10)'; S(9)* = S(12)1/S(10)1
ICF Incorporated
-------
Al-55
12. CARBON ADSORPTION (PAC)
a. Description
Soluble trace and refractory organic compounds can be removed from
(pretreated) waste streams by adsorption on activated carbon. Two major types
of carbon adsorbants are used: powdered activated carbon (PAC) and granular
activated carbon (GAC); but PAC is the predominant treatment process used for
hazardous wastes.
The PAC is dispersed into the sludge stream in a flash mixing tank. The
slurry is then fed to a flocculation basin. Slow mixing is maintained to
encourage intimate contact of organic species with the adsorbent and to
promote the coagulation of the particles. Following adsorption and
flocculation, a conventional clarifier is used to separate the effluent
process return stream from the hazardous waste stream underflow.
A summary of the important treatment vessel characteristics used in
estimating the individual and overall loss fractions to the environment is
given in Exhibit 1A-12.
b. Assumptions
There are two options for the utilization of this treatment unit,
depending on the physical properties of hazardous component X. These options,
applications, and the corresponding assumptions are discussed below.
Hazardous Component X Is Organic
In those waste categories where hazardous component X is organic, all of X
will be adsorbed by the carbon and sent as the hazardous waste stream for
thermal degradation or final disposal. The dissolved inorganic species are
washed from the cake with water and recycled to the industrial process. The
assumptions made for this treatment unit are as follows:
1. All of the dissolved portion of hazardous component X and an
additional organic portion of the remaining dissolved nonwater
species (not including X) specified by the user are adsorbed by
the activated carbon.
2. All suspended solids, including the solid portion of X and the
carbon adsorbed, are filtered and retained with the hazardous
waste stream.
3. All dissolved inorganics and the wash water from filter bed
washing operations are recycled as the process return stream.
ICF Incorporated
-------
EXHIBIT 1A-12
VESSEL CHARACTERISTICS AND RELEASE RATES
FOR CARBON ADSORPTION (PAC)
Treatment Vessel Characteristics
vessel type
vessel dimensions
retention time (hours)
agitation
evaporative surface area (meters)
M (fraction of influent sludge
contained in vessel)
Flash Mixer
open
H = 2D
0. I
rap id
-2 2/3
I.505(10 )«S(I)
-3
4.17(10 )
Contact mixer
open
H = D
0.5
s I ow
-1 2/3
6.985(10 )«S( I )
-2
2.08(10 )
C I a r i f i e r
open
H = . 5D
2
s I ow
-1
2.794( 10 )»S( I )
-2
8.33(10 )
2/3
O
3
O
o
-I
¦U
o
T
0>
r+
(0
Q.
Environmental Releases
aeration loss fraction
total aeration fraction
routine spillage fraction:
(a) routine maintenance
(b) mixer splashing
(c) materials handling
total routine spillage fraction
accidental spillage fraction:
best case, spills/year
worst case, spills/year
typical case fraction
total typical fraction
overall loss fraction
(sum of aeration, routine spi
age, and accidental spillage
fract ions)
10
10
-5
)
-5
-4
10
-4
10
12
-5
4.284(10 )
1.3(10 )
-1+
5.3(10 )•
-7
10
10
.5
12
I.923(10 )
-7
10
-6
10
.5
2
I.997(10 )
Ln
On
-------
Al-57
Hazardous Component X Is Inorganic (Figure A-34)
In those waste categories where hazardous component X is inorganic, all of
the dissolved portion of X will be washed from the solids cake and retained in
the hazardous waste stream. All of the solid portion of X will be retained
with the solids cake and recycled in the process return stream. The
assumptions made for this treatment unit are as follows:
1. All the dissolved portion of X and an additional inorganic
portion of the remaining dissolved nonwater species (not
including X) specified by the user are washed from the filtered
solids. The amount of wash water used is assumed to be three
times the mass of the carbon adsorbent. The dissolved
inorganics and all wash water are returned as the hazardous
waste stream.
2. All solids, carbon, and adsorbed organics are retained in the
filter cake as the process return stream.
c. Mass Balance Algorithm
The user must first state whether hazardous component X is organic or
inorganic. The program will then select the appropriate algorithm of those
described below.
Hazardous Component X Is Organic
1. The user must input the values of two variables:
"GAMMA, the average molecular weight of the dissolved
organics in the liquor. If the user does not input
a value, the default value will be the molecular
weight of hazardous species X.
"DELTA, the fraction of the dissolved nonwater species (not
including X) that is organic and will be adsorbed
on the carbon. If the user does not input a value,
the default value is set at zero.
2. The mass of carbon adsorbent required is estimated using the values of
GAMMA and DELTA described above:
ADS = mass flow of carbon adsorbent
= [DELTA*[S(4)*S(1)-S(11)J + S(11)] * [GAMMA*10-3]
3. The mass of retained wash water in the carbon solids cake is estimated
to be 25 percent of the mass of the carbon used:
S(15)1 = .25*ADS
ICF Incorporated
-------
Al-58
4. The total mass flow of the nonwater species is the sum of the mass
flows of carbon, adsorbed organics, and influent suspended solids:
MASS FLOW MASS OF ADSORBED MASS OF MASS OF INFLUENT
OF = ORGANICS, X + CARBON + SUSPENDED
NONWATER PLUS OTHERS ADSORPTION SOLIDS
SPECIES
S(14) 1 = DELTA*[S(4)*S(1)-S(11)]+ADS+S(3)*S(1)
5. The total RCRA stream flow is the sum of the flows of water and
nonwater species:
S(1)' = S(14) 1 + S(15)'
6. The mass fraction of the nonwater species is calculated from the
appropriate mass flows:
S(2)' = S(14)1/S(1)1
Since all nonwater species present are either solids or are adsorbed onto the
solids, the above mass fraction is also equal to the mass fraction of
suspended solids. The dissolved nonwater fraction is equal to zero.
S(3)1 = S(2)'; S(4)1 = 0
7. The total mass flow of X remains unchanged. The flow of X in the
dissolved phase is zero, since all the X is either in solid form or adsorbed
onto the carbon.
S(10)' = S(10); S(ll)' = 0; S(12)' = S(10)
8. The mass fraction of X in the sludge stream is calculated from the
appropriate mass flows:
S(7)1 = S(10)'/S(l)'
The proportions of X in the dissolved and suspended solids phases are 0 and 1,
respectively.
Hazardous Component X Is Inorganic
1. The user must input the values of two variables:
*GAMMA, the average molecular weight of the dissolved organics in the
liquor. The default value for this variable is the molecular
weight of hazardous species X.
-'DELTA, the fraction of the dissolved nonwater species (not including
X) that is organic and will be adsorbed on the carbon. The
default value for this variable is 1.
ICF Incorporated
-------
Al-59
2. The mass of carbon adsorbent required is estimated using the constants
described below:
ADS = mass flow of carbon adsorbent
= [DELTA*[S(4)*S(1)-S(11)) * [GAMMA*10-3]
3. The mass of water required to wash the filtered solids is estimated to
be four times the mass of the carbon used:
S (15)1 = 4*ADS
4. The total mass flow of the nonwater species is the sum of the mass
flows of X and other unadsorbed species:
S(14)1 = S(ll) + (1-DELTA)*[S(4)*S(1)-S(11)]
5. The total RCRA stream flow is the sum of the flows of water and
nonwater species:
S(1)' = S(14)' + S(15)1
6. The mass fraction of the nonwater species is calculated from the
appropriate mass flows:
S(2)1 = S(14)1/S(1)1
Since all the solids were filtered from sludge, the proportions of nonwater
species in the dissolved and suspended phases are 1 and 0, respectively:
S(3)* = 0, S(4)1 = S(2) 1
7. The total mass flow of X remains unchanged, since all the X is washed
from the filtered solids:
s cio)' = s(io)
The proportion of X in the dissolved and suspended phases are 1 and 0,
respectively:
SC11)' = s(10), S(12)1 = 0
S(8)' = 1, S(9)* = 0
8. The mass fraction of X in the RCRA stream is calculated from the
appropriate mass flows:
S(7)1 = S(10)'/S(1)1
ICF Incorporated
-------
Al-60
13. ION EXCHANGE
a. Description
The ion exchange process concentrates dissolved ions from dilute streams.
The ions are adsorbed onto the solid ion exchange resin, then desorbed by
washing with a strong salt solution during a resin regeneration cycle. Common
applications for industrial sludges include demineralization and the treatment
of heavy metal wastes. Pretreatment is sometimes required to remove soluble
organics, which can foul the ion exchange resin. All suspended solids must be
removed by filtration to prevent resin plugging and interference with the
adsorption-desorption process.
A single-column ion exchange system is commonly used in industrial
processes. The sludge is distributed evenly across the top surface of the
resin, where it filters down through the particles during ion adsorption.
When the operating capacity of the bed has been approached, the feed is shut
off, and the bed is backwashed to remove suspended solids. A strong salt
solution is then filtered through the bed to regenerate the resin. A final
rinse is used to remove traces of the regenerate solution before the column is
ready for reuse. Multiple columns are often operated in parallel to prevent
interruptions in treatment during the regeneration portion of the ion exchange
cycle.
The resin column is not suited to solids loading and is, therefore,
preceded by a filter to remove suspended particles. If hazardous component X
is present in the solids phase, the filtered solids are automatically returned
to the hazardous waste stream. When the filtered solids do not contain X, the
user may specify whether they are sent to the hazardous waste stream or the
process return stream.
A summary of the important treatment vessel characteristics used in
estimating the individual and overall loss fractions to the environment is
given in Exhibit 1A-13.
b. Assumptions
We made the following assumptions and/or simplifications to simulate how
the ion exchange technology changes the characteristics of a waste streams:
1. All of the influent dissolved species are ionic inorganics that
are retained by the ion exchange resin and regenerated by an
acid or base solution in water.
2. The concentration of dissolved species in the regenerate stream
is six times greater than in the influent sludge. Salts of the
regenerate acid and/or base are neglected.
ICF Incorporated
-------
Al-61
EXHIBIT 1A-13
VESSEL CHARACTERISTICS AND RELEASE RATES
FOR ION EXCHANGE
Treatment Vessel Characteristics
vessel type
vessel dimensions
retention time (hours)
agitation
evaporative surface area (meters)
M (fraction of influent sludge
contained in vessel)
closed
N/A
N/A
N/A
7.793x10-
Environmental Releases
aeration loss fraction
total aeration fraction
routine spillage fraction:
(a) routine maintenance
(b) mixer splashing
(c) materials handling
total routine spillage fraction
accidental spillage fraction:
best case, spills/year
worst case, spills/year
typical case fraction
total typical fraction
overall loss fraction
(sum of aeration, routine spill-
age, and accidental spillage
fractions)
N/A
5.203x10-7
5.20x10-7
.1
.5
4.27x10-®
5.2x10-7
ICF Incorporated
-------
Al-62
c. Mass Balance Algorithm
1. If X is present in the influent suspended solid, then all prefiltered
solids are added to the RCRA stream automatically. Otherwise, the user must
specify the stream to which the solids are added.
o
"•'•'if the solids are added to the RCRA stream, the flow
f nonwater species remains unchanged:
S(14)1 = S(14)
SOLIDS = S(3) • S(1)
"if the solids are recycled with the process return
stream, the flow of nonwater components is equal to that of
the influent dissolved species:
S(14)1 = S(4) • S(1)
SOLIDS = 0
2. The concentration of dissolved species in the regenerant stream is six
times greater than in the influent sludge:
regenerant flow = [S(14)'-S0LIDS]/6*S(4)
The RCRA stream flow contains the regenerant as well as the filtered
solids, if they are added:
S(l)' = [[S(14)1-SOLIDS]/[6 • S(4)]+ SOLIDS
3. The flow of water is equal to the difference in the total flow and
that of the nonwater species:
S(15) 1 = S(l)' - SC14)'
4. The mass fraction of non-water species are determined from the
appropriate mass flows:
S(3)1 = SOLIDS/S(1)1
S(4)' = S(4)*S(1)/S(1)1
S(2)' = S(3)1 + S(4)1
5. The mass flow of X in each phase remains unchanged:
S(10)T = S(10)
S(11) ' = S(11)
S(12)1 = S(12)
6. The mass fraction of X and the proportions of X in the suspended
solids and dissolved phases, are determined from the appropriate mass flows:
S(7)' = S(10)'/S(l)'
S(8)' = S(ll)'/S(10)'
S(9)' = S(12)1/S(10)1
ICF Incorporated
-------
Al-63
14. CHEMICAL PRECIPITATION
a. Description
Chemical precipitation removes dissolved species from sludge by chemical
reaction to form insoluble compounds. The precipitants are added to the
sludge in a flash mixing tank, where precipitation reactions occur rapidly.
The resulting slurry is fed to a flocculation basin. Slow mixing is
maintained to encourage the coagulation of particles. Following flocculation,
a conventional clarifier is used to separate the effluent process return
stream from the hazardous waste stream underflow.
A summary of the important treatment vessel characteristics used in
estimating the individual and overall loss fractions to the environment is
given in Exhibit 1A-14.
b. Assumptions
We made the following assumptions and/or simplifications to simulate how
the chemical precipitation technology changes the characteristics of a waste
stream:
1. All of the dissolved portion of hazardous component X is removed
to the suspended solids phase. Also precipitated is an
additional fraction of the remaining dissolved nonwater species,
f, specified by the user.
2. The hazardous waste stream underflow contains 2 percent solids
and 98 percent retained liquor, by mass. If the precipitated
solids stream from the flocculation basin contains more than 2
percent solids, then the clarifier produces no effluent, and all
of the flocculant stream appears in the underflow as the
hazardous waste stream.
3. The precipitated species are presumed to be bivalent hydroxides
produced by the addition of lime [CaCOH)^] in stoichiometric
quantities. The mass of the coprecipitant hydroxide ion is
calculated using the average molecular weight of the
precipitated ions, which may be specified by the user. The mass
of the dissolved nonwater species is increased by the mass of
calcium associated with CaCOH)^ addition.
4. This technology is inappropriate for influents in which
hazardous component X is completely in the suspended solids
phase. In this case, the model advises the user to select a
more appropriate treatment technology.
ICF Incorporated
-------
EXHIBIT TA-14
VESSEL CHARACTERISTICS AND RELEASE RATES
FOR CHEMICAL PRECIPITATION
Flash Mixer
F I occuI a to r
C I a r i f i e r
Treatment Vessel Character!stics
vessel type
vessel dimensions
retention time (hours)
ag i tat i on
evaporative surface area (meters)
M (fraction of influent sludge
contained in vessel)
open
H = 2D
0. I
ra p i d
-2
1.51(10 )
-2
4. 17 ( 10 )
S( I )
2/3
open
H = D
0.5
s I ow
-1
6.985 (10 ) • S(I )
-3
2.08 (10 )
2/3
open
H = . 5D
2
s I ow
N/A
-2
8.33 ( 10 )
O
-n
3
O
o
-J
T3
O
T
Q>
«-»¦
(0
a
Environmental Releases
aeration loss fraction
total aeration fraction
routine spillage fraction:
(a) routine maintenance
(b) m i xer spI a sh i ng
(c) materials handling
total routine spillage fraction
accidental spillage fraction:
best case, spills/year
worst case, spills/year
typical case fraction
total typical fraction
overall loss fraction
(sum of aeration, routine spill-
age, and accidental spillage
fract i ons)
10
-5
)
-5
10
-4
10
10
12
-5
4.284 (10 )
-4
2.6(10 )
-4
3.7(10 )
-7
10
10
.5
12
-5
,923 ( 10 )
-7
10
-6
10
.5
2
I.997 (10 )
I
ON
-------
Al-65
c. Mass Balance Algorithm
The user is first given the option to supply three input values:
1. MW, the average molecular weight of the precipitated ions (the value of
MW is set equal to the molecular weight of hazardous component X, if no average
value is supplied).
2. f, the fraction of the dissolved nonwater species, not including X, that
is to be precipitated along with X by the addition of lime (the value of f is
set equal to zero if no value is supplied).
3. RHO, the average density of the precipitated species (if no value of RHO
is supplied, the influent sludge suspended solids density is not adjusted to
account for the new solids present).
The bulk characteristics of the influent sludge to the clarifier are
calculated using the constants described above:
influent OH ions
a. Total suspended = suspended + mass of ions + added with
solids flow solids precipitated lime
TSS = S(3)*S(1) + [1«(S(4)«S(1)S(11)] + 34/MW*[f*(S(4)*S(l)S(ll))+S(ll)]
TSS = S(3)*S(1) + [1 + 34/MW][f(S(4)*S(l)-S(ll))+S(ll)]
influent mass of mass of
b. Total dissolved = dissolved - ions + Ca++ ions
nonwater flow nonwater precipitated added
TDS = S(4)»S(l)-[f(S(4)»S(l)-S(ll))+S(ll)] + 40/MW[f•(S(4)»S(1)-S(11))+S(11)]
TDS = S(4)#S(1)-(1 - 40/MW)[f,(S(4)*S(l)-S(ll))+S(ll)]
Total mass flow
c. of sludge to = influent sludge + mass of + mass of water
clarifier flow lime added added with lime
TANK = S(l)+(740/MW)[f,(S(4),S(l)-S(ll))+S(ll)]+9*(74/MW)[f•(S(4)#S(1)-S(11))+SW]
TANK = SCI) + (740/MW)[f*(S(4)#S(l)-S(ll))+S(11)]
ICF Incorporated
-------
Al-66
d. Total water = influent + water added with
mass flow water addition of lime
S(15) = S(15)+9* (74/MW)[f*(S(4)«S(l)-S(ll))+S(ll)]
4. The influent to the .clarifier is checked to determine whether or not a
clear effluent stream and an underflow of 2 percent solids can be obtained.
The result of this test determines some of the characteristics of the RCRA
stream underflow:
mass fraction of = (total solids flow)/(total mass flow)
solids in clarifier influent
SOLIDS = (TSS/TANK)
a. If SOLIDS < 0.0200, then the described effluent and underflow
can be obtained from the clarifier. The RCRA stream underflow is determined
as follows:
-the solids mass fraction is set equal to 2 percent
S(3)1 = 0.0200
-the total sludge flow is determined using the described solids
fraction, assuming that all of the suspended solids appear in the
underflow:
INFLUENT SOLIDS = SOLIDS IN
TO THE CLARIFIER UNDERFLOW
TSS = ( . 0200)•(S(1)1
therefore,
S(1)1 = 50*TSS
ICF Incorporated
-------
Al-67
'•the dissolved nonwater mass fraction is determined using the
variable TDS calculated earlier:
MASS FRACTION OF THE MASS FRACTION OF DISSOLVED
DISSOLVED NONWATER = NONWATER SPECIES IN THE
SPECIES IN TOTAL CLARIFIER LIQUOR UNDERFLOW LIQUOR
MASS OF DISSOLVED MASS OF MASS OF DISSOLVED NONWATER
NONWATER SPECIES = UNDERFLOW • SPECIES IN TOTAL CLARIFIER LIQUOR
IN UNDERFLOW LIQUOR LIQUOR
MASS OF TOTAL CLARIFIER LIQUOR
Dividing both sides of the equation by hazardous waste stream flow S(l)',
S(4)1 = [49»TSS*S(1)'] • TDS/[TDS+S(15)] = [0.98*TDS]/[TDS+S(15)]
b. If SOLIDS ^ 0.0200, the clarifier cannot produce the desired effluent
and underflow. As a result, no effluent is produced, and the entire influent stream
is retained as the RCRA stream. The underflow is determined as follows:
*the clarifier influent and underflow flows are equal: S(l)1 TANK
,vthe mass fractions of suspended solids and dissolved nonwater species are
determined from their respective flows, calculated earlier:
S(3)1 = TSS/S(1)'
S(4)1 = TDS/S(1)1
5. The mass fraction of the nonwater species is calculated as the sum of the
suspended solids and dissolved nonwater fractions:
S(2)1 = S(3)1 + S(4)'
6. The mass flow of the nonwater species is determined from the appropriate
mass fraction and total clarifier influent sludge flow:
S(14)' = S(l)' • S(2)1
7. The mass flow of water is the difference between the total clarifier
influent sludge flow and the mass flow of the nonwater species:
S(15)1 = S(l)' - S(14)'
ICF Incorporated
-------
Al-68
8. The total mass flow of hazardous component X remains unchanged. All of X
appears in the suspended solids phase, however, and none appears in the dissolved
phase:
S(10)' = S(10)
s cli)1 = o
S(12) 1 = S(12)
9. The mass fraction of X in the clarifier effluent is determined from the
appropriate mass flows:
S(7)1 = S(10)'/S(l)'
10. The proportions of X in the dissolved and suspended solids phases, with
respect to the total mass flow of X, are set equal to 0 and 1, respectively:
S(8)1 = 0
S (9 ) 1 = 1
ICF Incorporated
-------
Al-69
15. CHEMICAL DESTRUCTION
a. Description
In a chemical destruction treatment system, the hazardous component in the
sludge is "destroyed" through chemical reaction to form one or more
nonhazardous species. This treatment category includes all oxidation,
reduction, and decomposition reactions in which the products remain in the
dissolved phase and are retained in the RCRA waste stream (such as ozonation
and wet oxidation).
Influent sludge and dissolved reaction chemicals are combined in a tank
and agitated mechanically to ensure proper contact. Excess reaction chemicals
are often used to force the destruction reaction as close as possible to
completion. The retention time in the reactor depends upon the kinetics of
the system involved, but values of 0.5 to 2 hours are typical.
A summary of the important treatment vessel characteristics used in
estimating the individual and overall loss fractions to the environment is
given in Exhibit 1A-15.
b. Assumptions
We made the following assumptions and/or simplifications to simulate how
the chemical destruction technology changes the characteristics of a waste
stream:
1. The reaction chemicals added provide 98 percent destruction of
hazardous component X from both the dissolved and suspended
portions of X. The reaction products are nonhazardous compounds
in the dissolved and/or suspended phase, as specified by the
user.
2. The flow of reaction chemicals added is determined by two of the
user's inputs:
CHEM NW = mass flow of reaction chemicals
CHEM W = mass flow of water in reaction chemicals stream
All of CHEM NW is assumed to react with X to form reaction
products.
3. No reactions or phase .changes occur in the reactor except for
those pertaining to the destruction of X.
ICF Incorporated
-------
Al-70
EXHIBIT 1A-15
VESSEL CHARACTERISTICS AND RELEASE RATES FOR CHEMICAL DESTRUCTION
Treatment Vessel Characteristics
vessel type
vessel dimensions
retention time (hours)
agitation
evaporative surface area (meters)
M (fraction of influent sludge
contained in vessel)
open
H = 2D
0.5
moderate
.sm2/3
4.401x10- *S(1)
2.08x10-"
Environmental Releases
aeration loss fraction
total aeration fraction
routine spillage fraction:
(a) routine maintenance
(b) mixer splashing
(c) materials handling
total routine spillage fraction
accidental spillage fraction:
best case, spills/year
worst case, spills/year
typical case fraction
total typical fraction
overall loss fraction
(sum of aeration, routine spill-
age, and accidental spillage
fractions)
10-'
10-6
10-6
.5
6
1.07 xlO-5
1.2 xlO-5
ICF Incorporated
-------
Al-71
c. Mass Balance Algorithm
1. The user must input values for the mass flow of reaction chemicals:
CHEM NW = mass flow of reaction chemicals
CHEM W = mass flow of water in reaction chemicals stream
The value of CHEM W is added to the influent water flow, while CHEM NW is
added to the flow of nonwater species:
S(14)1 = S(14) + CHEM NW
SC15)' = S(15) + CHEM W
2. The mass flows of hazardous component X in both phases is decreased by
98 percent:
S(10)' = .02 • S(10)
S(ll)' = .02 • S(ll)
S(12)' = .02 • S(12)
3. The total hazardous waste stream mass flow from the reactor is
determined by adding the flow of reaction chemicals to that of the influent
sludge:
S(l)' = S(l) + CHEM W + CHEM NW
4. The mass fraction of nonwater species is calculated from the
appropriate mass fractions:
S(2)1 = S(14)1/S(1) '
5. As stated earlier, all of CHEM NW is assumed to react with X to form
reaction products. The mass flow of reaction products is thus the sum of CHEM
NW and the mass of X that is reacted:
MASS OF MASS OF
REACTION X + CHEM NW
PRODUCTS REACTING
= (.98 • S(l)) + CHEM NW
The user must input the value of a variable PSI, the fraction of the reaction
products generated that goes to the suspended solids phase. (The value of PSI
is usually 0 or 1.) The mass flow of suspended solids in the hazardous waste
stream is the sum of the influent suspended solids (excluding 98 percent of
the X from the solids phase, which was reacted) and any new suspended solids
added by chemical reaction:
ICF Incorporated
-------
Al-72
MASS FLOW UNREACTED SUSPENDED SOLIDS
OF SUSPENDED = INFLUENT + GENERATED BY
SOLIDS SUSPENDED CHEMICAL REACTION
SOLIDS
= [S(3) • S(1) - .98 • S(12)] + [PSI • (.98 • S(10) + CHEM NW)]
Dividing by the total hazardous waste stream flow,
S(3)1 = [S(3) • S(l) - .98 • SC12) + PSI • (.98 • S(10) + CHEM NW)]/S(1)'
6. The mass fraction of dissolved nonwater species is the difference between
the total nonwater and suspended solids mass fractions:
S(4)' = S(2) ' - S(3) '
7. The mass fractions of X in the RCRA stream, and the proportions of X
in the dissolved and suspended phases, are determined from the appropriate
mass flows:
S(7)* = S(10)1/S(l)1
S(8)' = S C11)*/S(10)1
S(9)1 = S(12)'/S(10)'
ICF Incorporated
-------
Al-73
16. ELECTROLYTIC DECOMPOSITION
a. Description
Electrolytic decomposition is used to transform hazardous ionic species in
industrial sludges to nonhazardous forms via reactions of electrodes. This
technology is applied primarily in the treatment of spent plating solutions
where the decomposition of cyanides and the elimination of heavy metal ions
from solution by plating may be carried out simultaneously.
Almost any plating bath may be used for electrolytic decomposition
reactions as long as provisions are made to handle the gases produced from the
decomposition of cyanide.
A summary of the important treatment vessel characteristics used in
estimating the individual and overall loss fractions to the environment is
given in Exhibit 1A-16.
b. Assumptions
We made the following assumptions and/or simplifications to simulate how
the electrolytic decomposition technology changes the characteristics of a
waste stream:
1. Electrolytic decomposition to nonhazardous species "destroys" 98
percent of both the suspended solids and dissolved portions of
X. The remaining 2 percent of X is retained in the hazardous
waste stream where the proportion of X in the suspended and
dissolved phases is the same as that for the influent sludge.
2. No species other than X are destroyed during the reaction.
3. The decomposition products may be solids, gases, or dissolved
nonwater species, as specified by the user. If the products are
solids or gases, they are withdrawn from the reactor and
recycled to the industrial process. If the products are
liquids, they are included as a nonhazardous component in the
hazardous waste stream leaving the reactor.
c. Mass Balance Algorithm
1. The flow of hazardous component X in both phases is reduced by 98
percent:
S(10)' = .02 • S(10)
S(11)' = .02 • S(ll)
S(12)' = .02 • S(12)
ICF Incorporated
-------
Al-74
EXHIBIT 1A-16
VESSEL CHARACTERISTICS AND RELEASE RATES
FOR ELECTROLYTIC DECOMPOSITION
Treatment Vessel Characteristics
vessel type
vessel dimensions
retention time (hours)
agitation
evaporative surface area (meters)
M (fraction of influent sludge
contained in vessel)
open
width: 2
depth: 1.5
6
slow
1. 67x10-'* • S (1)
2.50xl0-w
Environmental Releases
aeration loss fraction
total aeration fraction
routine spillage fraction:
(a) routine maintenance
(b) mixer splashing
(c) materials handling
total routine spillage fraction
accidental spillage fraction:
best case, spills/year
worst case, spills/year
typical case fraction
total typical fraction
overall loss fraction
(sum of aeration, routine spill-
age, and accidental spillage
fractions)
10-7
10-6
2.5xl0-5
2.6xl0-5
2.276x10-"
2.4xl0-u
ICF Incorporated
-------
Al-75
2. The user must state whether the reaction products remain in the RCRA
stream as nonhazardous dissolved species or are removed from the sludge and
returned to the industrial process. This statement determines the flow of
nonwater species in the RCRA stream. The flow of water remains unchanged.
The total hazardous waste stream flow is the sum of the flows of water and
nonwater species:
-if the reaction products remain in the RCRA stream as
nonhazardous dissolved species, the flows of water and
nonwater species remain unchanged:
S(14)1 = S(14)
SC15) ' = S(15)
S(l)' = S(l)
The mass fractions of suspended solids and dissolved
species must be adjusted if any influent X was in the
solids phase:
S(4)' = [S(4) • SCI) + -98 • S(12)]/S(l)1
S(3)1 = [S(3) • S(1) - .98 • S(12)]/S(1)'
S(2)' = S(14)'/S(1)'
"if the reaction products are solids or gases and are
therefore removed from the sludge in a process return
stream, the flow of nonwater species in the RCRA stream is
decreased:
S(14)1 = S(14) - .98 • S(10)
S(15)1 = S(15)
S(1)' = S(14)' + S(15)*
The mass fractions of suspended solids and dissolved
species must be adjusted to account for the material
withdrawn:
S(4)1 = [S(4) • S(l) - .98 • S(ll)]/S(1)1
S (3) ' = [S (3) • S (1) - .98 • S (12) ] /S (1) '
S(2)' = S(14)1/S(1)'
/
3. The mass fraction of X in the sludge and the proportions of X in the
suspended solids and dissolved phase are determined from the appropriate mass
flows:
S(7)' = S(10)'/S(1)1
S(8)1 = S(11)'/S(10)'
S(9)' = S(12)1/S(10)1
ICF incorporated
-------
Al-76
17. CHEMICAL FIXATION/STABILIZATION
a. Description
Chemical fixation/stabilization is used to solidify a sludge stream by
adding chemicals that bind up all sludge components, including water, into a
solid matrix. The solid is then extruded into a form convenient for further
handling. The fixation/stabilization treatment is preferred when leaching or
leaking of hazardous wastes from storage vessels or disposal facilities could
present severe hazards.
A summary of the important treatment vessel characteristics used in
estimating the individual and overall loss fractions to the environment is
given in Exhibit 1A-17.
b. Assumptions
We made the following assumptions and/or simplifications to simulate how
the fixation stabilization technology changes the characteristics of a waste
stream:
1. The fixation/stabilization reaction yields 100 percent fixed
solids from the total influent sludge stream.
2. The mass of the influent sludge stream is increased 25 percent
by the addition of the stabilization chemicals.
3. The volume of the sludge stream is increased by 10 percent owing
to the chemical reactions during stabilization.
c. Mass Balance Algorithm
1. The mass flows of the influent sludge stream and the nonwater species
are increased 25 percent by the addition of the stabilization chemicals.
S(l)' = 1.25 • S(1); S(14)' = S(14) + .25 • S(l)
2. The values of S(3) and S(4) are set equal to 1.0 and 0.0 respectively,
since the waste is completely converted to the solids phase:
S(3)' = 1.0; S(4)1 = 0.0
The flow of water is zero, since it is bound up in the fixed solids:
S(15)' = 0
3. The mass fraction of the nonwater species in the fixed solids phase is
determined using the appropriate mass flows:
S(2)' = S(14)1/S(1)1
ICF Incorporated
-------
A1 - 7 7
EXHIBIT 1A-17
VESSEL CHARACTERISTICS AND RELEASE RATES
FOR CHEMICAL FIXATION/STABILIZATION
Treatment Vessel Characteristics
vessel type
vessel dimensions
retention time (hours)
agitation
evaporative surface area (meters)
M (fraction of influent sludge
contained in vessel)
open
H = 2D
2
moderate to rapid
1.11 xlO-w*[influent
sludge flow]^^
8.3 xlO-2
Environmental Releases
aeration loss fraction
total aeration fraction
routine spillage fraction:
(a) routine maintenance
(b) mixer splashing
(c) materials handling
total routine spillage fraction
accidental spillage fraction:
best case, spills/year
worst case, spills/year
typical case fraction
total typical fraction
overall loss fraction
(sum of aeration, routine spill-
age, and accidental spillage
fractions)
10-
10-u
10-"
1
24
1.54x10-3
1. 6x10-3
ICF Incorporated
-------
Al-78
4. The total mass flow of X does not change. All of X is present in the
fixed solids phase, while the flows in both the dissolved and suspended solids
phases are zero.
S(10)' = S(10); S(11)' = 0; S(12)' = 0; S(13)' = S(10)
5. The mass fractions of X in the dissolved phase, suspended solids
phase, and fixed solids phase are determined using the appropriate mass flows:
S(7)1 = S(11)S(10)/S(1)'
S(8)' = 0
S(9)1 = 0
6. The density of solids in the RCRA stream is adjusted to reflect the
mass and volume changes during the stabilization reactions:
S(5)1 = S(5) • 1.25/1.1
ICF Incorporated
-------
Al-79
18. INCINERATION
a. Description
Incineration destroys hazardous organic components by oxidation to form
carbon dioxide, water, and other combustion gases. Three major types of
incinerators are used in industrial sludge treatment: hearth, rotary kiln,
and liquid injection. Hearth and rotary kiln incinerators may burn solids,
liquid sludges, and containerized waste. Liquid injection incinerators are
used only with liquids or liquid sludges. All require auxiliary fuel when the
BTU content of the waste is less than 6000 BTU/kg. Although the treatment
efficiency is not necessarily dependent upon the type of incinerator used, the
capital and operating costs for each can vary significantly.
Total costs for hearth and liquid injection incinerators are about the
same; rotary kiln incineration is about twice as expensive. Liquid injection
incinerators are the most common, followed by hearth and then rotary kiln.
Recent trends, however, are toward increasing use of rotary kiln incinerators
for such difficult wastes as PCB capacitors and other containerized wastes.
We, therefore, narrowed the kinds of incinerators used in this model to liquid
injection (which we permit to accept wastes less than or equal to 10 percent
solids) and rotary kiln (which we permit to accept wastes greater than 10
percent solids).
Hazardous waste incinerators operate at a variety of temperatures and
waste feed rates. We selected two different temperatures -- 1600°F and 2200°F
-- as typical for all incinerator scenarios. In the usual case, a lower
operating temperature results in lower auxiliary fuel requirements and lower
scrubber waste stream volumes; however, for some organic waste streams, use of
a higher operating temperature may result in no need for air pollution control
scrubbers. For all incineration scenarios at both temperatures, we used a
typical waste feed rate of 2,300 kilograms per hour (kg/hr). The 2,300 kg/hr
feed rate is based upon results of a recent EPA-sponsored survey of
incinerator operators, and represents the approximate mean of facility feed
rates reported.
The amount of air pollution control equipment used on a particular
incinerator depends upon the regulations governing emissions to the
atmosphere, and the destruction and removal efficiency (DRE) sought to be
achieved. The most commonly used air pollution control device is a Venturi
scrubber to remove particulates and condense unburned organic species. The
size and cost of the scrubber, and the volume of scrubber water produced, are
functions of the degree of removal required.
Our model provides four levels of destruction and removal efficiency for
the constituent of concern (X) between the influent waste stream and the final
release of combustion products to the environment:
99.99 percent DRE implies
99.9 percent DRE implies
99 percent DRE implies
90.0 percent DRE implies
0.0 percent of X released
0.1 percent of X released
1 percent of X released
10.00 percent of X released
ICF Incorporated
-------
Al-80
These typical incinerators receive waste, excess air, and supplementary
fuel (if necessary). Two potentially hazardous waste streams (ash and
scrubber water) are discharged, along with stack emissions to air. The ash is
withdrawn from the bottom, while the gaseous and light particulate combustion
products rise to the top. The proportion of X present in each stream is
dependent upon the physical properties of X and the DRE to be achieved.
Scrubbers are added where needed to meet DRE requirements. The 90 percent
DRE incinerator scenario does not include a scrubber system, nor does the 99
percent DRE incinerator operating at 2200°F where X is an organic
constitutent. All other scenarios include the use of a scrubber system. As
DRE increases, so does the efficiency and cost of the scrubber system.
The scrubber systems used in this model are wet scrubber systems; that is
water is injected to adsorb pollutants and wash away particulates before the
gases are allowed to escape to the atmosphere. The quantity of water required
increases with the degree of removal of X demanded, as shown below:
Scrubber Water Requirements (kg/m3 of flue gas)
Incinerator Scenario Metal Constituents Organic Constituents
99.99 percent DRE 4.0 4.0
99.9 percent DRE 2.0 2.0
99 percent DRE 1.0 1.0 (1600°F)
0.0 (2200°F)
90 percent DRE 0.0 0.0
The scrubber water may be further treated for removal of X or sent directly to
disposal.
b. Assumptions
We made the following assumptions and/or simplifications to simulate how
the four incinerator technologies change the characteristics of a waste stream:
1. An incinerator's performance for a given waste stream is independent
of the final degree of treatment, but follows the standards set for the
emission of X in the ash and combustion gases stated above. The higher levels
of removal, if required, are attained through addition of Venturi scrubber
capability.
2. An incinerator's performance is highly dependent upon the
characteristics of the waste stream being incinerated. For example, most
organic substances are oxidized nearly completely, with the actual degree of
degradation increasing with temperature. Heavy metal constituents, on the
other hand, do not decompose when incinerated. The model provides for such
differences as follows:
ICF Incorporated
-------
Al-81
a. Organic constituents that are not destroyed during incineration
are completed vaporized under each of the scenarios; that is, there is no
residual constituent left in the ash stream). Depending upon the incineration
temperature, either 5 percent (1600°F) or 1 percent (2200°F) of any organic
constituent appears in the gas stream. If such a release does not meet the
appropriate DRE requirement, air pollution control devices must be added.
b. Heavy metal constituents do not decompose when incinerated.
During incineration, heavy metals are distributed between the ash stream and
the stack gas stream independent of incineration temperature. The fraction of
incoming metal constituents going to the stack gas stream is dependent on the
form and concentration of both the metal constituent and the waste stream. As
a general rule, the greater the concentration of incoming heavy metal, the
smaller the fraction of the metal going to the stack gas stream.
3. The ash stream (hazardous only where metal-contaminated wastes are
incinerated) contain 5 percent of the mass of the influent nonwater
constituents, plus 10 percent added water to assist in ash handling. All
nonwater species are assumed to be particulate solids.
c. Mass Balance Algorithm
The specific algorithms used to predict quantities and characteristics of
waste streams treated by incineration are presented below.
1. The model selects the type of incinerator to be used based upon the
characteristics of the waste being incinerated.
Value of Sludge (3)
(% of suspended solids) Incinerator Type
<10% Liquid Injection
^10% Rotary Kiln
2. The user must select the level of treatment required. The answer to
this question sets the values of two variables:
Destruction Removal
Efficiency (DRE) BETA W
99.99 percent
99.9 percent
99 percent
90 percent
.0001
.001
.01
4.0
2.0
1.0
0.0
(0.0 for organics
at 2200°F)
BETA = fraction of X released in stack gas to air
W = flow of scrubber water, kg/m3 flue gas
3. The model selects the most economic burn temperature, using cost as
its sole criterion. Except for high BTU content wastes and some organic
ICF Incorporated
-------
Al-82
wastes in the 99.9 percent DRE incinerator, the lower (1600°F) temperature is
the most economic, in that it requires less auxiliary fuel. The temperature
used determines the amount of stack gas (GAS) produced as follows:
Temperature GAS
1600°F 5.663x10-uM3/BTU
2200°F 8.495x10-*M3/BTU
The remainder of the mass balance algorithm is different for heavy metals
and organic species.
Heavy Metals
1. The concentration of X in the incoming waste stream is used to
determine the distribution of X between the fly ash (SUM) and bottom ash
(1-SUM) streams:
SUM = (LOG(SLUDGE(7))+1.58)/-4.05
where:
SUM = mass fraction of X in stack gas stream
SLUDGE(7) = concentration of X in influent waste (kg/kg)
If the computed value for SUM is greater than 1.0, its value is set to 1.0.
If the computed value of SUM is less than 0.1, it is set to 0.1.
2. A value for ALPHA, the mass fraction of X removed by the scrubber, is
determined:
ALPHA = SUM - BETA
Where the incinerator being used is the 90 percent DRE incinerator, ALPHA =
0.00, and BETA = SUM
3. The air emission of X is computed as:
AIRLOS = (BETA) • (SLUDGE(10))
4. The flow of scrubber water is determined from the variables GAS and W,
set earlier:
S(15)' = GAS • W • S(16)*S(1)
If some of the X remains in the bottom ash (i.e., SUM < 1.0), water mass
flux is increased for that used in bottom ash handling:
S(15)" = S(15)1 + (0.005) • (SLUDGE(14))
ICF Incorporated
-------
Al-83
The BTU value of the waste has been consumed, thus,
S(16)1 = 0.0
Nonwater mass flux is retained by the scrubber in the same proportion as is X,
S(14)' = (ALPHA) • (S(14))
If X remains in the bottom ash (SUM < 1.0), this value must be increased to
account for that stream. We assume 95 percent reduction, thus
S(14)" = S(14)1 + (0.05) • (S(14))
All X is present as suspended solids, thus
S(13) 1 = 0.0
S(11) ' = 0.0
S(9)' = 1.0
S(8)1 = 0.0
Except for X released to air, all X is retained in the RCRA stream.
S(10)' = (ALPHA) • (S(10)) + (1.0 - SUM) • (S(10))
(S(12)1 = S(10) '
The new mass flux is given by the sum of water and nonwater mass fluxes, as
computed above.
S(l)' = S(14)1 °r " + S(15)' °r "
All of the nonwater fraction is present as suspended solids.
S (2) 1 = S (14) ' °r 7S(1)', provided S(l)' does not = 0.0
S(3)1 = S(2)'
S(4)1 = 0
Specific gravity of suspended solids (SLUDGE(5)) is set to 2.5.
S(5)' = 2.5
Organics
For organics, the value of SUM (the fraction of X going to the stack gas
stream) is set as follows:
Burn Temperature SUM
1600°F 0.05
2200°F 0.01
ICF Incorporated
-------
Al-84
Air emissions of X are strictly a function of BETA, the DRE emission allowed.
Thus,
AIRLOS = (BETA) • (SLUDGE(10))
ALPHA, the fraction of X retained by the scrubber, is computed as
ALPHA = SUM - BETA; if ALPHA ^ 0.0
ALPHA is set to 0.0.
The flow of scrubber water is determined by the variables W and GAS (set
above):
S(15) 1 = (GAS) • (W) • (S(16)) • (S(1))
The BTU content of the waste is set to zero.
S(16) ' = 0.0
Nonwater mass flux is retained by the scrubber in the same ratio as is X,
except that 5 percent of the nonwater fraction is assumed to be consumed by
the process; thus
S(14)1 = (ALPHA) • (S(14)) • (0.95)
S(10)' = (ALPHA) • (S(10))
The fraction of X in dissolved and suspended forms is assumed to be the same
as in the influent waste.
S(8)1 = S(8)
S(9)' = S(9)
S(ll) ' = (ALPHA) • (S(11))
S(12)1 = (ALPHA) • (S(12)) and
S(13)* is set to zero
The new mass flux is the sum of water and nonwater fluxes.
S(1)' = S(15)' + S(14) 1
S(2)' = S(14)'/S(l)1
d. Cost Computations
Costs for the incinerator scenarios are based on a methodology suggested
in a recent EPA study of hazardous waste incineration facilities. Capital
cost elements include a waste storage and loading system, combustion chamber,
heat recovery devices (if any), and air pollution scrubber devices (if any).
We assume that each of these four major cost elements represent 25 percent of
total basic capital costs. Basic capital costs were estimated in the recent
study as follows:
ICF Incorporated
-------
Al-85
For liquid injection incinerators,
0 45
basic capital cost = (216) (heat input)
For rotary kiln incinerators,
0 52
basic capital cost = (225) (heat input)
Where heat input is the BTU input to the incinerator from waste and
auxiliary fuel, if any.
Basic capital costs are adjusted to delete the cost of: (1) heat recovery
devices if the waste has low BTU content, and (2) scrubber systems for the 90
percent DRE scenario and for the high temperature 99 percent DRE scenario
where only organics are burned. For the 99.9 and 99.99 DRE scenarios, the
cost for scrubbers is doubled and quadrupled, respectively.
To capital costs, we add operating costs of $520,000/year and auxiliary
fuel costs. The amount of auxiliary fuel required, if any, is a function of
the BTU content of the waste being incinerated and incinerator operator
temperature. We used natural gas as the auxiliary fuel in each of the
incinerator scenarios; a cost of $4 million BTU was used for 1600°F
incineration, and $5 million BTU for 2200°F incineration. Use of other
auxiliary fuels would result in higher fuel costs. We believe the use of the
lower cost of natural gas is appropriate because many hazardous waste
incinerators are able to mix high BTU content wastes with lower BTU content
wastes and avoid the use of auxiliary fuel all together.
ICF Incorporated
-------
APPENDIX 2
TRANSPORTATION TECHNOLOGIES
ICF Incorporated
-------
APPENDIX 2
TRANSPORTATION TECHNOLOGIES
In Chapter 2, we described the three types of transportation modules, (on-site,
local and long-distance) contained in the risk-cost framework. This appendix
provides further information on two aspects of these modules: (1) the method we
used to model the rates at which hazardous wastes are released during
transportation, and (2) the basis for the per-unit costs of the transportation
modules.
RELEASE RATE COMPUTATIONS
Transportation, as we use the term, includes loading and unloading hazardous
waste into and from the transport vehicle. Spills during loading and unloading
(called transfer spills below) are a significant release mechanism for the three
transportation technologies.
In a December 1980 study by F.G. Bercha and Associates, transfer spill rates
for hazardous wastes were estimated at 2.45xl0-u, divided almost evenly between
loading and unloading operations. The Bercha study noted that some 80 percent of
the waste spilled during transfer were small volume spills (less than 32 liters
each), and were unlikely to be cleaned up. Accordingly, we used a spill rate of
2x10-'' for transfer spills (0.8 x 2.45x10-'' = 2X10-11).
To transfer spills must be added releases owing to spills in transit (i.e.,
resulting from highway accidents, etc.). We used accident statistics compiled by
the Bureau of Motor Carrier Safety to estimate spills resulting from accidents.
Accident rates for trucks are about 2.5x10-® accidents per truck-mile. The Coast
Guard was involved in 569 incidents involving land vehicles in 1978. These
incidents resulted in a total release of 623,000 gallons of pollutants, or roughly
1,100 galIons/incident. 1,100 gallons represents approximately 20 percent of the
typical 6,000-gallon tank truck load. The rest, we assume, is either contained
and/or recovered. Thus, on a per-mile basis, releases during transit are computed
as 2.5xl0-6 x 0.2 or 5xl0-7 per truck-mile. This figure compares favorably with
the transit spill rate of 10-5 for a 60-mile trip reported by Bercha.
Exhibit A2-1 presents the spill release rates we used in this analysis. It is
interesting to note that, with the exception of long distance transportation,
spills due to handling hazardous wastes are much more significant than spills
during transit.
COST COMPUTATIONS
For this analysis, we assume transportation is by tank truck with a 6,000
gallon (about 23 cubic meters) capacity. Tank trucks are more expensive than most
other highway vehicles (about 2.5 times more expensive than a flatbed trailer, for
example), but tank trucks do not require as much costly loading equipment as other
vehicles. The 6,000 gallon capacity used for this analysis represents a tank truck
similar to large gasoline tractor-trailer combinations.
Exhibit A2-2 compares assumptions and costs for the local and long-distance
transportation modules. Since we assume that the equipment used for onsite
disposal is also used in transporting wastes onsite, we include the cost for onsite
transportation in the cost of the disposal technology.
ICF Incorporated
-------
A2-2
EXHIBIT A2-1
SPILL RATES FOR TRANSPORTATION TECHNOLOGIES
1 1
Total-
Technology
Transfer Spills Transit Spills
j i
Onsite (0.25 miles)
2x10-'
(0.25)(5xl0-7)
=1.25x10-7
2x10-'
= 10-'
Local transport
(25 miles)
2x10-"
(25)(5x10-7)
=1.25xl0-5
2.1x10-'
= 10-'
Long Distance 2x10-''
Transport (250 miles)
(250)(5x10-7)
=1.25x10-"
3.3x10-'
=10-3
"'•"Rounded to nearest order of magnitude on a logarithmic (Base 10) scale.
ICF Incorporated
-------
A2-3
EXHIBIT A2-2
ASSUMPTIONS AND COSTS FOR TRANSPORTATION MODULES
Local transportation
(50 miles roundtrip)
Long-distance transportation
(500 highway portion miles
roundtrip)
Assume 250 working days/year,
each roundtrip requires 1/2 day,
thus 500 trips/year
Labor (driver)
Cost of $20,000/year, or $40/trip
Operating (fuel and oil) and maintenance
expense of $0.40/mile, or $20/trip
Fixed capital expense of
$60,000 (tractor/truck)
20 ,000 (tank trailer)
$80,000 Total, with 10-year useful
life, or $8,000/year or
$16/trip
Assume 250 working days/year,
each roundtrip requires 2 days,
thus 125 trips/year
Labor (driver)
Cost of $20,000/year, or $160/trip
Operating (fuel and oil) and
maintenance expense of $0.40/mile,
or $200/trip
Fixed capital expense of
$60,000 (tractor/truck)
20,000 (tank trailer)
$80,000 Total, with 10-year
useful life, or
$8,000/year or $64/trip
Total per trip cost
$40 (labor)
20 (operation and maintenance)
16 (capital)
$76/trip
23 m3 capacity/trip
Cost = $3.30/m3
Total per trip cost
$160 (labor)
200 (operation and maintenance)
64 (capital)
$424/trip
23 m3 capacity/trip
Cost = $18.40/m3
REVIEW DRAFT: 4-27-82 (7533G)
-------
APPENDIX 3
DISPOSAL TECHNOLOGIES
ICF Incorporated
-------
APPENDIX 3
DISPOSAL TECHNOLOGIES
We consider nine typical disposal technologies for use in this project, as
follows:
• double-lined landfill;
• single-lined landfill;
• unlined landfill;
• double-lined surface impoundment;
• single-lined surface impoundment;
• unlined impoundment;
• land treatment (land farming);
• deep well injection; and
• ocean disposal.
After a general description of the disposal technology, we detail the alternate
levels of control in separate sections, e.g., double-lined, single-lined, unlined.
We also present assumptions used to develop costs and releases and any limitations
on the wastes each technology may accept.
LANDFILLS
a. Description
Landfilling -- the most common method of disposing of hazardous wastes --
usually involves placing the waste in a specially prepared excavation or trench and
then covering it with fill material. Modern landfills have a wide range of
specific features which depend upon the characteristics of the individual site.
We considered excluding wastes containing free liquids from all of our landfill
scenarios. However, to do so would prevent consideration of the effects of free
liquids in a landfill environment, as measured by risk and cost. Thus, for
purposes of this analysis, we have not limited acceptable wastes for landfill
disposal.
Proper design of landfills is highly site specific; thus no single landfill
configuration can be regarded as "typical." Geology, for example, influences the
depth of excavation, placement of liners, design of leachate collection system, and
the extent of the monitoring well system. In addition, the characteristics of the
waste itself often determines particular design features; thus, a landfill
containing wastes with free liquids requires a more extensive leachate collection
system than does one containing "dry" wastes.
For purposes of this project, we developed three different scenarios for
hazardous waste landfills: double-lined, single-lined, and unlined. These three
degrees of control reflect the wide range of available design and operating
practices of landfills.
ICF Incorporated
-------
A3-2
Exhibit 3A-1 presents the features of the three scenarios. All have the same
total waste capacity of about 674,000 m3 (550 acre-feet) and the same operating
area of 10 hectares (25 acres). These values represent the middle range of
landfills for which an EPA Part A permit application was filed. Reported capacities
varied, however, more than two orders of magnitude both higher and lower from the
typical value. We assume that the 25-acre operating area represents a small
commercial facility accepting wastes generated off-site or a large on-site
facility.
In all three scenarios waste is placed in excavations. Although almost all
existing hazardous waste landfills are excavated, some use combinations of above
grade and excavated disposal. We selected excavated landfills because they reflect
the trend in modern hazardous waste landfills.
b. Environmental Releases
Excavated landfills limit the available pathways for release (1) to ground
water through leachate migration, (2) to air by volatilization, and (3) to surface
waters from leachate collection system failure. Of course, accidental spills can
occur, but we account for those in the transportation technology. Each of the three
release mechanisms are discussed below.
Volatilization
Volatilization is a function of the volatility of the constituent of concern
and its exposure to the atmosphere. For this project, volatility has been
approximated using a function of molecular weight and vapor pressure. Exposure is a
function of waste concentration, area exposed, and the length of time the waste is
exposed. After the waste has been covered in .the landfill, some volatilization
will continue, although at a reduced rate.
Air emission rates are computed by assuming certain weather conditions
(temperature at 25 degrees Celsius, wind speed of 4 m/sec or 9 mph), and using
molecular weight, vapor pressure, and waste concentration to estimate volatility
and constituent exposure to the atmosphere. Our specific methodology is derived
from several existing models which predict hazardous air emissions from landfills.
The methodology is useful in determining order-of-magnitude release rates, but
contains a number of theoretical and practical weaknesses. First, the air release
rate computation is not completely compatible with the risk exposure scoring
methodology used in this framework. Air exposure scores are not adjusted for
adsorption potential in soil, and the air release rate computations do not take
adsorption into account. Thus, releases of some gases may be overstated by the air
release methodology used. On the other hand, there are some waste substances which
are not highly volatile but which are highly soluble in water. If the water/waste
mixture is exposed to air, simple gaseous diffusion could result in significant
release, contrary to the result obtained by using our methodology. Such special
cases must be identified using professional judgment during review of strategies,
and risk and cost scores adjusted accordingly.
ICF Incorporated
-------
EXHIBIT 3A-1
FEATURES OF THREE LANDFILL SCENARIOS
Landf iI I Scena ri o
Feature
Geology/hyd rogeoIogy
Area If
Dimens ion
Shape
DoubIe-L i ned
Closely scrutinized
25(50) acres
500 x 200 x 10m
3 stra ight waI Is and
one 3:1 slope waI I
S i ngIe-L i ned
Sc rut i n i zed
25(50) acres
500 x 200 x 10m
3 st ra ight waI Is and
one 3:I slope waI I
UnI i ned
Not scrutinized
25(50) acres
500 x 200 x 10m
3 straight walls and
one 3:1 slope waI I
Total void
962,267m
962,267m
962,267m
O
T|
3
O
O
"I
•O
o
T
0)
r+
(D
a
Capacity (70% of void)
Lining system
Leachate collection
Leachate treatment
Mon i tori ng
Control of surface runoff
Intermediate cover
F i naI cove r
Life span
Post-closure care
673,587m
DoubIe-Ii ne r (clay
and a rt i f ic iaI)
Extens i ve
Custom Plant
8 we I Is
Some
Twice da ily
Clay, artificial
I i ner, so i I, and
revegetat ion
20 years
Extens i ve
673,587m
In-place clay
Some
Package Plant
6 we I Is
Some
Once da i ly
CI ay, so iI, and
revegetation
20 years
Some
# 25 acres of operating area and 50 acres of total disposal facility.
673,587m
None
None
None
None
Little
At closure
Topsoil and revegetation
20 yea rs
Little
>
CO
I
LO
-------
A3-4
Both volatilization and leachate migration will occur beyond the 20-year
operating life of a landfill. The time-frame over which these releases are
computed is therefore important. Various time-frames were considered for use in
this analysis, ranging from 50 years (20-year operating life plus 30-year post-
closure care period) to 200 years. The period selected for this analysis is 100
years, based on the assumption that releases from landfills reach a steady state
after a 100 years.
We have made several other assumptions regarding landfill operations, each of
which we believe to fairly represent typical operations. These are:
Double-Lined
Single-Lined
Unlined
Maximum Waste Area Exposed
1/4 acre
1/2 acre
25 acres
Maximum Waste Exposure Time
4 hours/day
8 hours/day
20 years
Width of Waste Exposure Area
32m
/
32m
32m
200m
Length of Waste Exposure Area
64m
500m
Wind Speed (Average)
4m/sec
4m/sec
4m/sec
Using these assumptions, and using the 100-year period of reference
discussed above, two volatilization rates were calculated for each landfill as
follows.
I. When the waste is not covered, use the following equations:
a) dv = 2CeW*n/DLV/irFv' *Wi* (dv in ml/sec)
dt dt
Given: Ce = equilibrium vapor pressure (%) =
vapor pressure/760 = P/760 (P in mm Hg)
W = width = 3,200 cm double-lined and single-lined, 20,000
cm unlined
D = diffusion coefficient in cm2/sec
L = length = 3,200 cm double-lined, 6,400 single-lined, and
50,000 cm unlined
v = wind speed = 4 m/sec = 400 cm/sec
Fv = correction factor = 1 for Ce at 0 - 10%
Wi = (concentration in ppm)xl0-6
b) Emission rate is:
E = dv_ • (M/22.4) • 103 = y/sec
dt
ICF Incorporated
-------
A3-5
Procedure
Step I: Calculate Diffusion Coefficient
D = D'
where D' and M* are diffusion coefficient and molecular weight
respectively of a compound with similar molecular weight and
molecular diffusion to the constituent of interest having D
and M
Assume x = D* Vm7
Given:
Chemical
CC1,
D' @ IO°C
0.07500
D' @ 30°C
0.08451
M'
154
C2H2C14
PCB
-------
A3-6
Double-lined Landfill
L = 3,200 cm
W = 3,200 cm
E = 0.23 CPM"75
Single-lined Landfill
L = 6,400 cm
W = 3,200 cm
E = 0.32 CPM'75
Unlined Landfill
L = 50,000 cm
W = 20,000 cm
E = 5.59 CPM"75
II. When the waste is buried, use the following equation:
Ei = DCsAPt1,33 • Wi/L (Ei in yg/sec)
60 cm (Double-lined
Given: L = effective depth of final cover: and single-lined)
30 cm (Unlined)
D = diffusion coefficient in cm2/sec.
Cs = saturation vapor concentration in yg/cm3
A = exposed area: 5X101* x 2xl01' = 10s cm2
Pt = soil porosity: for double-lined and single-lined,
use 0.30; for unlined, use 0.60
Wi = (concentration of constituent in ppm)xl0-s
Procedure
Step I: Calculate Diffusion Coefficient
D = D'*(M'/M)^ = 0.9/VFT (see derivation earlier)
Step II: Calculate saturation vapor concentration
Cs = PM/RT
At 25°C (or 298°K)
Cs = (PMx106)t(6.23x10"*298)
= PM/18.57
ICF Incorporated
-------
A3-7
Step III: Calculate Pt
1.33
Pt^'33 = 0.30^"33 = 0.20 (Doubled-lined and single-lined)
= 0.601'33 = 0.51 (Unlined)
Step IV: Calculate Wi/L
Wi/L = (CxlO-6)/60 = 1.67x10-8C (Double-lined and
single-lined)
Wi/L = (CxlO-6)/30 = 3.33x10-8C (Unlined)
Step V: Calculate Ei
For double-lined and single-lined, with Pt = 0.30 and L = 60 cm
Ei = DCsAPt1'33*Wi/L
= 0.9/M^»PM/18.57xl09•0.20*1.67xl0-8C
= 0 i 30*10*PM^*C/18.57
Ei = 0.162*PM C
• PM^f
For unlined, with Pt = 0.60 and L = 30 cm
= 0.9/M^*PM/18.57x109*0.51*3.33x10-8C
= 15.28*PM^*C/18.57
Ei = 0.823*PM C
• PM^r
Example: Emission rate of a constituent when vapor pressure = 4x10-3
mm Hg @ 25°C, Concentration = 5xl03 ppm, MW = 258 g/mole
with a cover of 2' of compacted clay:
Double-Lined and Single-Lined:
Pt = 0.3
L = 60 cm
Ei = 0.162*4x10-3*258^*5x103
= 52 yg/sec
ICF Incorporated
-------
A3-8
Unlined:
Pt = 0.6
L = 30 cm
Ei = 0.823*4x10-3*258^*5x103
= 264 yg/sec
Leachate Migration
Leachate forms when waste is exposed to water which has infiltrated a
landfill. Some of the waste is dissolved in the water, which is then free to
migrate. As used here, the term leachate also refers to liquids placed in the
landfill and available to migrate (so-called "free liquids").
The most common technique used to predict quantities of leachate formed in
landfills is the water balance method. The method requires site-specific
information such as soil surface grading and permeability, annual rainfall, run-
off, evaporation and transpiration. Essentially, the method subtracts run-off from
rainfall to determine infiltration.
Annual rainfall varies widely in the United States. For purposes of this
analysis, we use a rainfall figure of 40 inches per year as typical. We believe the
40 inches/year assumption is representative, within one-half an order-of-
magnitude, of most regions in the nation.
The specific design and operating features for each landfill scenario
(principally the permeability of the cover) affect the amount of run-off from each;
the remaining infiltration will therefore depend on the specific landfill scenario.
The water balance method used for the analysis predicts the quantity, but not
the quality, of the leachate formed. Water extracts contaminants from disposed
wastes depending on the quantity of contaminant exposed, its solubility, and its
tendency to adsorb to soil or other media in the migration paths. In computing
releases, we assume solubility is the basic measure of leachate composition. We
assume that all infiltration becomes saturated with the constituent of concern, and
is available to form leachate. We are able to neglect the effects of adsorption in
computing release rates, because adsorption potential was a principal mechanism
used to determine exposure scores (see Chapter 3).
Free liquids pose a different problem. We based leachate generation on
engineering judgment, assuming that the waste material in the landfill cell will
retain up to 60 percent of its dry weight in liquids. The 60 percent figure is
based on reported maximum dewatering of sludges under gravity forces, and we
believe it is a conservative figure (i.e., free liquids may not form until waste is
80 to 90 percent liquid). Liquid in excess of 60 percent will be available as free
liquid. Depending on specific landfill features, free liquids will either be
collected in the leachate collection system or be available for ground water
release.
ICF Incorporated
-------
A3-9
Some materials are thought to have a significant adverse impact on the
performance of liner materials. Under almost all conditions, however, a clay liner
will both maintain its low permeability and protect the synthetic liner, if any,
from harm. A few substance (e.g., analine, acetone) are known to "quicken" clay
materials; in such cases permeability may be increased by about one order-of-
magnitude. To the extent such substances are placed in landfills as wastes (few
are), their effect must be determined during professional review, and releases
adjusted accordingly.
A number of assumptions were used to simplify the process of estimating
leachate generation rates. These assumptions are summarized below:
• Synthetic liner will last for 20 years, after which liquid
will move freely through the liner. The amount of liquid
passing through the liner in the first 20 years is negligible
due to low permeability (10-12 cm/sec).
• The only sources of liquids in the landfills are (1) rainfall
or infiltration from the landfill surface and (2) free liquids
(if any). We exclude lateral groundwater migration. Except
for operating impoundments, there will be only saturated flow
through the liner system.
• Free liquids (defined as liquids in excess of 60 percent of
waste, by weight) will form, if at all, during the 20-year
operating life for landfills. All free liquids will drain from
landfills within 3 years of closure (assume average
permeability of fill of 10-s cm/sec or 3.15 m/yr. and maximum
depth of fill of 10m; 10m/3.15 m/yr. - 3 years).
• Infiltration through the cover system after closure is less
than or equal to leachate movement through the liner system.
• Clay liner will retain its integrity indefinitely.
• Materials placed in the fill are more permeable than the
cover or liner systems.
• Synergistic effects are not considered.
The time required for leachate to first appear outside the clay liner is
approximated by the following equation.
t = ird2/4D*
where d = thickness of clay liner (60 cm) and
D* = linearlized diffusivity constant, assumed to be
10-5 cm2/sec (typical of clay). Note that
D* is not the same as the coefficient of
permeability (10-7 in this case).
ICF Incorporated
-------
A3-10
Thus, the time required for leachate to first appear outside the clay
liner is about:
Landfills: (3.14)(60 cm)2 * (4)(10-5cm2/sec) = 2.83x10-® sec,
which is about ten years.
We use ten years as the length of time the clay liner will prevent leachate release
in the single-lined and double-lined landfill scenarios. Note that for double-
lined landfills, the protection offered by the clay liner is redundant with the
assumed useful life of the synthetic liner.
During the period of protection offered by each liner system, we assume the
leachate collection system conveys all liquids to the leachate treatment system.
After the period of protection, we assume leachate releases are limited by the clay
liner, or by infiltration through the cover, whichever is less.
The total volume of leachate released over time is given by:
q = K G *A*At
^ s h
where Kg = saturated permeability coefficient
G^ = hydraulic gradient (use 1 for landfills)
A = area at base of landfill or surface
impoundment ^^SxlO^m2 for landfills)
At = length of time over which leachate is
released.
Applying this equation to the landfill scenarios yields the leachate releases shown
in Exhibit 3A-2.
Where the release quality for leachate is "contaminated infiltration only," we
assume that all such infiltration becomes saturated with the constituent of concern
(or that all the constituent available becomes dissolved in the infiltration,
whichever is less). This assumption overstates the groundwater release which would
actually be experienced. We use this conservative assumption for consistency
purposes and to reflect the principal characteristic (solubility) of waste
constituents affecting leachate quality. Ongoing EPA research may assist us in
later revising this assumption. Where the release quality shown is "liquid waste
only," we assume the dissolved constituent is released in proportion to its
concentration in the liquid waste.
Where release quality is shown as mixed "infiltration and free liquids," we
compute the quality in proportion to the quantity of each available. The operating
landfills receive 40 inches of rainfall over their areas, or about 96,000 m3/yr.
Since the landfills are excavated, this rainfall cannot run-off. Thus the total
annual quantity of leachate formed is the sum of 96,000 m3 and the annual quantity
of free liquids placed in the landfill. We assume a proportionate amount of each is
released to ground water (as computed above), with the balance removed by leachate
collection, if any.
IC F I ncorporated
-------
EXHIBIT 3A-2
QUANTITY AND QUALITY OF LEACHATE
GENERATED BY LANDFILLS
FACTORS
Leachate removed
K (m/yr)
s
T (yrs)
1 3
g (m /yr)
3
(m )
Re lease qua Ii ty
DOUBLE-LINED
All free liquids and pre-
cipitation during 20 yr
operating life and first
3 years of closure
.0315
100 yrs - 20 yr protection
by duaI Ii ne r -
- 20 year protection by
dual cover = 60 years
1,800
180,000
Contaminated infiltration
on ly
SINGLE-LINED
Most of free Iiquids and
precipitation during 20
yr operating life and
first 3 years of closure
.0315
100 yrs - 10 yr protec-
tion by clay liner =
90 yrs
2,700
270,000
Contaminated infiltra-
tion and free liquids
UNLINED
Dur i ng
Ope ra t i on
None
3.15
20
3
295,000
5,900,000
Contam i na ted
i nf i I ra t i on
and free
Iiqu i ds
Af te r
Closure 2
NA
3. 15
80
48,000
3,840,000
Contam i na ted
i nf i11 ra t i on
on ly
>
w
O
3
O
o
-J
TJ
O
"»
fi>
rt-
(D
Q.
I .
2.
K G A delta T; where G = hydraulic gradient = I.
Soil is highly permeable; thus permeability does not limit infiltration
4 2
coefficient of 0.5, as follows: q = (40 inches/yr.) (.0254
Infiltration was calculated using run-off
n/m) (9.43 x 10 in | (0.5 runoff).
3. Value shown is maximum value of leachate generation, including both rainfall ((40 inches/yr.) (.0254 in/m) (9.43 x
4 2 3
10 m ) = 96,000 m /yr) plus free liquids, if any.
-------
A3-12
The leachate collection and treatment system will be operated intermittently
after closure as required. During the first three years after closure (during
which time free liquids are flowing through the fill), operation of the leachate
collection system is assumed to result in removal of all free liquids (double-lined
landfill) or almost all free liquids (single-lined landfill). This assumption is
parallel to that used during landfill operations.
Long term performance data is not available for leachate collection sytems.
Thus we have conservatively assumed that the leachate collection system operates
after closure only for such a period of time (about 3 years) as to remove free
liquids, if any, from the fill.
Spillage from Leachate Collection System
The double-]ined and single-lined landfill scenarios have leachate collection
and treatment systems. Any leachate collection system is subject to occasional
accidental failure and inadvertent spillage. Releases due to such failure are
thought to be of secondary importance by comparison to losses due to leachate
migration or volatilization. Nevertheless, losses from the leachate collection
system are the only potential releases to surface waters from the double-lined
landfill, and therefore they should be estimated.
Review of all of the hazardous waste treatment technologies shows that most
have release rates of either 0.001 percent or 0.01 percent (see Exhibit 1A-4).
Because the leachate collection and treatment systems will be similar to our
hazardous waste treatment technologies, we choose the low side of typical -- 0.001
percent -- as the total release rate for the double-lined landfill leachate
collection and treatment system. We assigned the less expensive package plant
system used to treat leachate from single-lined landfills to the high side of the
typical release rate, 0.01 percent.
Leachate released from the treatment system is a mixture of rainfall saturated
with the constituent of concern and the dissolved constituent in free liquids, if
any.
c. Cost Computations
We developed costs for landfilling wastes by estimating unit costs for
components of the three scenarios. These cost figures for the three scenarios are
provided for in Exhibits 3A-3 through 3A-5. The most significant of these
components in terms of characterizing each of the three landfill scenarios is
discussed below.
d. Double-Lined Landfill
The double-lined landfill includes the use of a double liner system, leachate
collection and treatment, and a double cover system. The term "double" refers to a
combination of synthetic membrane and natural clay material.
The double-lined landfill also includes equipment and ancillary features
needed for proper site operation. Eight full-time personnel are used in the
scenario, including one foreman. Sufficient personnel and equipment are available
to provide intermediate cover twice daily over wastes disposed.
ICF Incorporated
-------
A3-13
EXHIBIT 3A-3
DOUBLE-LINED LANDFILL COST ESTIMATE
Immediate Capital Costs
Item
Land, 50 acres
Equipment
Front end loader, 1
Tractor, 1
Forklift, 1
Bulldozer, 1
Scraper, 1
Pick-up truck, 2
Building
Fencing, 2,500 m
Construction
Excavation, 963,000 m^
Synthetic Liner (36 mil)
95,000 m2
Clay liner 96,000m3
Drainage/Leachate
Collection System
Leachate Treatment
System
Collection Ditch 2,000m
Monitoring Wells, 8 ea
Discount (3%)
Unit
Cost
5,000/acre
Facto
Presen
12,000 each
10,000
40/m
2.50/m3
8/m2
5/m2
19/m
3,000 each
for
Value
SUBTOTAL, IMMEDIATE CAPITAL COSTS
Total ($)
250,000
33,000
30,000
15,000
100,000
180,000
24,000
10,000
100,000
2,407,500
760,000
480,000
398,000
100,000
38,000
24,000
4,949,500
Replacement Capital Costs
Item
Equipment @ 10 yrs.
Unit
Cost
382,000
Discount
Factor P/F
3%, 10 years
0.7441
Total ($)
284,000
ICF Incorporated
-------
A3-14
EXHIBIT 3A-3 (cont.)
DOUBLE-LINED LANDFILL COST ESTIMATE
Operating and Maintenance Costs
Item
Labor
Foreman/supervisor,
1 for 20 years
Laborers, 7 for 20 yrs.
Equipment Fuel and
Maintenance
Utilities
Security Service
Sample Collection/Analysis
Unit
Cost
40,000/yr
Discount
Factor P/A
3%, 20 years
14.82
20,000/yr/person 14.82
100,000/yr
2,500/yr
12,000/yr
12,800/yr
.14.82
14.82
14.82
14.82
SUBTOTAL, OPERATING AND
MAINTENANCE
Total ($)
592,800
2,074,800
1,482,000
37,050
177,840
189,696
4,554,186
Capital Costs at Closure
Item
Synthetic Cover (20 mil)
100,350m2
Clay Cover 100,000m2
Revegetation 100,000m2
Unit
Cost
4.5/m2
5/nr
1.25/m2
Discount
Factor P/F
3%, 20 years
0.5488
0.5488
0.5488
SUBTOTAL, CAPITAL COSTS
AT CLOSURE
Total ($)
247,824
274,400
68,600
590,824
ICF Incorporated
-------
A3-15
EXHIBIT 3A-3 (cont.)
DOUBLE-LINED LANDFILL COST ESTIMATE
Post Closure Care Costs
Item
Site Maintenance, perpetual
Monitoring, perpetual
Unit
Cost
2,000/yr
4,270/yr
Discount Factor
P/A 3%, perpetual
P/F 3%, 20 yrs
18.29
18.29
SUBTOTAL POST-CLOSURE CARE
COST
Total Cost for Stringent Landfill
Immediate Capital Costs
Replacement Capital Costs
Operating and Maintenance Costs
Capital Costs at Closure
Post Closure Care Costs
TOTAL COST
Total ($)
36,850
78,100
114,680
4,949,500
284,000
4,554,186
590,824
114,680
10,493,190
ICF Incorporated
-------
A3-16
EXHIBIT 3A-4
SINGLE-LINED LANDFILL COST ESTIMATE
Immediate Capital Costs
Item
Land, 50 acres
Equipment
Front end loader, 1
Forklift, 1
Bulldozer, 1
Scr aper, 1
Trailer
Fencing, 2,500 m
Construction
Excavation, 963,000 m^
Clay liner 96,000m^
Drainage/Leachate
Collection System
Leachate Treatment
System
Collection Ditch 2,000m
Monitoring Wells, 4 ea
5,000
20/m
2.50/m3
5/m2
Discount (3%)
Unit
Cost
5,000/acre
Facto
Presen
19/m
3,000 each
for
Value
Total ($)
250,000
33,000
15,000
100,000
180,000
5,000
50,000
2,407,500
480,000
200,000
50,000
38,000
12,000
SUBTOTAL, IMMEDIATE CAPITAL COSTS 3,820,500
Replacement Capital Costs
Item
Equipment @ 10 yrs.
Unit
Cost
328,000
Discount
Factor P/F
3%, 10 years
0.7441
Total ($)
244,000
ICF Incorporated
-------
A3-17
EXHIBIT 3A-4 (cont.)
SINGLE-LINED LANDFILL COST ESTIMATE
Operating and Maintenance Costs
Item
Labor
Foreman/supervisor,
1 for 20 years
Laborers, 5 for 20 yrs.
Equipment Fuel and
Maintenance
Utilities
Sample Collection/Analysis
Unit
Cost
40,000/yr
Discount
Factor P/A
3%, 20 years
14.82
20,000/yr/person 14.82
70,000/yr 14.82
2,000/yr
6,400/yr
14.82
14.82
SUBTOTAL, OPERATING AND
MAINTENANCE
Total ($)
592,800
1,482,800
1,037,400
29,640
94,850
3,236,690
D. Capital Costs at Closure
Item
Clay Cover 100,000m^
Revegetation 100,000m^
Unit
Cost
5/m^
1.25/m2
Discount
Factor P/F
3%, 20 years
0.5488
0.5488
SUBTOTAL, CAPITAL COSTS
AT CLOSURE
Total ($)
274,400
68,600
343,000
ICF Incorporated
-------
A3-18
EXHIBIT 3A-4 (cont.)
SINGLE-LINED LANDFILL COST ESTIMATE
Post Closure Care Costs
Discount Factor
Unit P/A 3%, perpetual
Item Cost P/F 3%, 20 yrs Total ($)
Site Maintenance, perpetual 1,000/yr 18.29 18,290
Monitoring, perpetual 1,070/yr 18.29 19,570
SUBTOTAL POST-CLOSURE CARE 37,860
COST
Total Cost for Moderate Landfill
Immediate Capital Costs 3,820,500
Replacement Capital Costs 244,000
Operating and Maintenance Costs 3,236,690
Capital Costs at Closure 343,000
Post Closure Care Costs 37,860
TOTAL COST 7,682,050
ICF Incorporated
-------
A3-19
EXHIBIT 3A-5
UNLINED LANDFILL COST ESTIMATE
Immediate Capital Costs
Item
Land, 50 acres
Equipment
Front end loader, 1
Forklift, 1
Bulldozer, 1
Scraper, 1
Trailer
Construction
Excavation, 963,000 m^
Storm Sewer (Drainage)
Unit
Cost
5,000/acre
Discount (3%)
Factor for
Present Value
1
1
1
1
1
1
Total ($)
250,000
33,000
15,000
100,000
180,000
5,000
5,000
2.50/m3
SUBTOTAL, IMMEDIATE CAPITAL COSTS 3,028,560
2,407,500
38,000
Replacement Capital Costs
Item
Equipment @ 10 yrs.
Unit
Cost
328,000
Discount
Factor P/F
3%, 10 years
0.7441
Total ($)
244,000
ICF Incorporated
-------
A3-20
EXHIBIT 3A-5 (cont.)
UNLINED LANDFILL COST ESTIMATE
Operating and Maintenance Costs
Item
Labor
Foreman/supervisor,
1 for 20 years
Laborers, 3 for 20 yrs.
Equipment Fuel and
Maintenance
Utilities
Unit
Cost
Discount
Factor P/A
3%, 20 years
40,000/yr 14.82
20/000/yr/person 14.82
50,000/yr 14.82
2,000/yr
14.82
SUBTOTAL, OPERATING AND
MAINTENANCE
Total ($)
592,800
889,200
741,000
29,640
2,252,640
D. Capital Costs at Closure
Item
Revegetation 100,000m2
Unit
Cost
0.625/m2
Discount
Factor P/F
3%, 20 years
0.5488
Total ($)
34,300
ICF Incorporated
-------
A3-21
EXHIBIT 3A-5 (cont.)
UNLINED LANDFILL COST ESTIMATE
Post Closure Care Costs
Item
Site Maintenance, perpetual
Unit
Cost
500
Discount Factor
P/A 3%, perpetual
P/F 3%, 20 yrs
18.29
Total ($)
9,145
Total Cost for Uncontrolled Landfill
Immediate Capital Costs
Replacement Capital Costs
Operating and Maintenance Costs
Capital Costs at Closure
Post Closure Care Costs
3,028,500
244,000
2,252,640
34,300
9,145
TOTAL COST
5,568,585
ICF Incorporated
-------
A3-22
Post-closure care for this scenario includes site inspection, maintenance of
the dual cover to prevent premature failure, and monitoring of the eight ground
water wells.
For ground water releases, the liner and cover systems have a significant
impact and are the subject of several major assumptions. First, the synthetic
membrane is assumed to retain its integrity for 20 years after installation, after
which it fails completely. The 20 year life is based on manufacturer's literature,
and is, in our opinion, a conservative assumption for most applications. However,
due to lack of long-term data, the 20 year life assumption is thought prudent for
this analysis.
We assume that both liner and cover are installed properly; i.e., no rips or
tears are left unrepaired during installation. In fact, undetected mistakes can be
made when installing large liner systems. We do not consider releases due to
faulty installation of liner materials. They may be roughly balanced by the
conservative assumption of material life, discussed above.
During the 20-year operating life of the landfill, we assume that the
synthetic/clay liner system retains all leachate formed, and that the leachate
collection system removes all accumulated leachate from the landfill. On closure
of the landfill, the synthetic/clay cover prevents infiltration of rainfall for 20
years after closure. Thus, during the 100-year period of reference, only the final
60 years have leachate release under our scenario.
Since the only period of leachate generation (years 40 to 100 in our 100-year
period of reference) comes well after site operations have ceased, leachate from
the double-lined scenario contains no free liquids, whether or not free liquids
have been placed in the fill. Leachate is composed entirely of infiltrated
rainwater, which we assume to be saturated with the constituent of concern.
Leachate formed during site operations is a mixture of rainfall and free
liquids, if any. As discussed above, we assume all leachate during operations is
removed by the leachate collection system. Using the 0.001 percent release rate
discussed earlier, we calculate the amount of air, ground water and surface water
losses due to accidential spillage, etc.
e. Single-Lined Landfill
The single-lined landfill scenario has the same general size and configuration
as the double-lined scenario. Important differences in site features include the
use of single-layer, in-place clay liner and cover systems. The leachate
collection and treatment system used is less extensive (and less expensive) than
that assumed for the double-lined scenario.
Operation of the single-lined landfill includes once-daily cover. This means
that a larger waste area lies exposed during operations for a longer period of
time, with a resulting increase in air emissions. Operation of the landfill
requires fewer personnel and pieces of equipment than does the double-lined
scenario with a resulting lower cost.
ICF Incorporated
-------
A3-23
Post-closure care is adequate to maintain the integrity of the clay cover
system, but is less extensive than that required to maintain the cover for the
double-lined landfill. Groundwater monitoring is also less extensive than for the
double-lined landfill.
The absence of the synthetic liner yields slightly greater leachate
generation. However, leachate quality may be much different, depending upon the
presence of free liquids. Over the 100-year period of reference, an average of
2700m3/year of leachate is released. During operations, leachate is composed of
both infiltration and free liquids, if any, in proportion to the amounts present in
the fill. After site closure, leachate is comprised only of infiltration.
As noted above, we assigned a spill rate of .01 percent for losses due to
leachate collection system failure, which is a mid-range value for hazardous waste
treatment technologies generally. The leachate system spill rate used for single-
lined landfills is one order of magnitude higher than that used for double-lined
landfills, due to the lower capital and operating expense incurred by the single-
lined landfill system.
f. Unlined Landfill
The unlined landfill scenario, although the same size and capacity as the
other two, differs in almost every other important respect. It does not use a liner
system or an intermediate cover, final cover is simply native soil, and post-
closure care is limited to minimal grading.
These differences have a major impact on both releases and unit costs. Since
the waste is exposed for the full 20-year period of operations, volatilization is
higher for this scenario. The lack of a liner system permits a two order-of-
magnitude increase in leachate generation rates as compared to other scenarios,
provided that sufficient free liquids are placed in the fill. During operations,
leachate results from infiltration and all free liquids placed in the fill; after
closure, leachate is just infiltration. No leachate collection system is present
to interrupt leachate flow.
The lack of a leachate collection system means that we do not consider system
spills and resulting releases. The only releases on the surface of the ground are
those associated with waste handling during transportation, because even the
unlined scenario is completely below grade. This means that surface water releases
are lower for the unlined scenario than for either of the other landfill scenarios.
SURFACE IMPOUNDMENTS
a. Description
Surface impoundments are not usually considered ultimate disposal facilities,
but rather merely long-term storage facilities. In regions of the country where
evaporation rates are high, surface impoundments can serve as disposal sites.
Evaporation reduces the volume of waste disposed, and the site can be eventually
closed in a fashion similar to landfill closure. Other surface impoundments serve
as disposal facilities by serving as sink-holes, allowing waste to pass into the
soil.
ICF Incorporated
-------
A3-24
Our approach emphasizes the use of surface impoundments as disposal
facilities, rather than long-term storage areas. Thus, evaporation and ground
water migration are the mechanisms which determine disposal capacities. The unit
costs and release rates used in this approach are also useful in assessing risks
and costs associated with surface impoundments used for long term storage.
There are a number of differences in size, features, and configurations for
surface impoundments, depending on their purpose. For this project, we developed
three "typical" scenarios for hazardous waste surface impoundments, representing
three different levels of control.
The three scenarios are presented in Exhibit 3A-6. All three have the same
surface area (1 hectare or 2\ acres), volume (16,660 m3), and evaporation rate
(5,000 m3/yr). However, waste capacity differs widely among the scenarios,
depending on the quantity of the waste liquids lost through the liner and/or soil.
All three impoundment scenarios are excavated; waste liquids are placed below
surface grade. While some hazardous waste surface impoundments are only partially
excavated, most are completely below grade.
Use of excavated surface impoundments limits pathways available for release.
Up to the time the waste is placed in the impoundment, losses are attributed to the
transportation technology, and not the disposal technology. After the waste has
been placed in the impoundment, it cannot be released to surface waters unless the
impoundment is overfilled, or unless the leachate collection and treatment system
fails (double-lined impoundment only).
b. Environmental Releases
Releases from impoundments are due to volatilization of waste (to air)
subsurface liquid waste migration and leachate migration (both to ground water),
and spills due to overfilling or leachate collection system failure (both divided
among air, ground water, and surface water). These releases are discussed below.
Volatilization
All three surface impoundment scenarios have the same area exposed to the
atmosphere for the same period of time (20 years). Thus for surface impoundments,
air losses depend only on the characteristics of the waste materials. (Surface
impoundments expose liquid waste directly to air, which means that the methodology
used for predicting air emissions from landfills would understate emissions for
surface impoundments.)
The following procedure was used for computing air emissions from the surface
impoundment scenarios.
ICF Incorporated
-------
EXHIBIT 3A-6
FEATURES OF THREE SURFACE IMPOUNDMENT SCENARIOS
Surface Impoundment Scenario
Feature
Doub1e-Li ned
S i ng1e-
Li ned
Un1i ned
Geo 1ogy/hyd rogeo1ogy
Closely scrutinized
Scrut i n i zed
Not scrutinized
Area*
2.5(5) Acres
2.5(5)
ac res
2.5(5)
ac res
D i mens i on
100 x 100 x 3m
100 x 1
00 x 3m
100 x 1
00 x 3m
Side slope
3: 1
3: 1
3: 1
Total void
3
23,800 m
23,800
3
m
23,800
3
m
Capacity (70% vo'd)
3
16,660 m
16,660
3
m
16,660
3
m
20 year waste capacity**
3
100,000 m
1 1 3,400
3
m
268,000
3
i m
Liner system
Double 1iner
(clay and artificial)
In-place clay
None
>
O)
i
fo
Ln
Leachate collection and
treatment
Yes
No
None
Berms
Extensive
Some
Little
Surface run-off and
contro1
Mon i to ring
Some
6 we 11s
Some
U we 11s
Little
None
F i na1 Cove r
Clay, artificial liner,
soil, and revegetation
Clay, soi 1, and
revegeta t i on
Topso i1
and revegetation
Life span
20 years
20 years
20 years
Post-closure care
Some
Little
None
* 2.5 acres of surface impoundment and 5 acres of total disposal facility.
3 3
** About 30 percent volume (5,000 m ) of the liquid waste per year is lost through evaporation. Thus, 5,000 m of liquid
3 3
waste can be added annually, totaling 100,000 m for 20 years. An additional 8,U00 m /year is added to the unlined
3
impoundment, and 670 m /year to the single-lined impoundment, due to ground-water releases (see Appendix 6).
-------
A3-26
Ei = (Koa)A XiMi
Where Ei = emissions rate, g/sec
A = lagoon surface area, cm2
Xi = concentration of waste in pond, mole fraction
Mi = molecular weight of waste, g/g-mole
Koa = overall mass transfer coefficient, g-mole/cm2-sec
Xi = Ci/105»(l/Mi)T(l/18) = 18xl0-6-Ci/Mi
Ci is concentration of waste in mg/1
Ei = 18x10-6•(Koa)A#Ci
1/Koa = 1/Kt + 1/KKG
Li
Kt = liquid phase mass transfer coefficient, g-mol/cm2-sec
Li
K_, = gas phase mass transfer coefficient, g-mol/cm2-sec
K = equilibrium constant of liquid and gas phases
The mass transfer rate for most low solubility gases is liquid phase
controlled, whereas for more soluble gases, such as SO^, the gas phase may
control. For Hi (Henry's law constant) 10J, Koa = K^, thus
Ei = 18xl0-6*K ACi
1j
Where K = 5.78*1.024t"2°V°'67H"0'85Di/Do
ij
t = temperature, °C
V = windspeed fps
H = depth of lagoon, ft
Di = diffusion of waste, cm2/sec
Do = diffusion of oxygen (0.18 lb. mol/ft2/hr)
Di/Do = (Mo/Mi)* and Mo = 32
= (32/Mi)*
K*l/7350 g-mol/cm2-sec
Ij
Kt = 4.45xl0"3M~^1.024t"20V0,67H_0'85 g-mol/cm2-sec
Li
ICF Incorporated
-------
A3-27
Procedure
Step I: Calculate K , mass transfer coefficient
Jj
Kt = 4.45xl0"3M"^1.024t"20V°'67H"0,85
Ll
For the surface impoundment of our design features,
t = 25°C
V =4 m/sec = 12 ft/sec
H = depth of lagoon = 2.5m = 7.5 ft.
Thus, Kt = 4.45x10"3m"^(1.024)s12°"67(7.5)"0"85
Li
g-mol/cm2-sec
Step II: Calculate Ei, emission rate
Ei = 18x10-6K AC
Li
Where A = 100 x lOOm2 = 108cm2
C = waste concentration in mg/1
Thus, Ei = 18x10-6KlAC
= 18x10-6•(4.77x10-3M~*)10sC
Ei = 8.59M~*C
Applies to all wastes with low water
solubility and low vapor pressure
Example: A waste with C = 200 mg/1 and M = 258. What is the emission
rate from the lagoon?
Ei = 8.59x258"* x 200
= 107 g/sec
Assuming the lagoon is in operation for 20 years the emission rate is:
Ei = 8.59M~*C#3 .15# lO11 *20, in kg/20 years
ICF Incorporated
-------
A3-28
This calculation assumes that non-volatilization occurs after site closure due to
waste removal, low concentration, and placement of final cover. Also note that the
same quantity of waste will emit from the impoundment regardless of scenario or
site features.
Subsurface Liquid Waste and Leachate Migration
Releases through the liner or soil base of the impoundment scenarios are
computed in much the same way as for the landfill scenarios, with one important
difference. For landfills we assumed there was no hydraulic gradient (actually a
hydraulic gradient of one). This means we assume no "ponding" in landfills. Since
surface impoundments do have ponding, a hydraulic gradient must be added to the
analysis for migration during operation of the impoundment.
The hydraulic gradient measures the hydraulic pressure added to gravitational
forces which together cause liquids to move through soil. It is computed by
dividing the elevation of the free standing liquid by the elevation of the top of
the liner system, using the bottom of the liner system as a baseline elevation. In
all three scenarios the liquid elevation is about 3 meters and the elevation of the
top of the liner is about 1 meter, yielding a hydraulic gradient of about three.
This means that all other things being equal (permeability of soil, etc.),
three times as much migration through the soil will occur per unit area for surface
impoundments during operations as compared to landfills.
After the impoundment is closed, the hydraulic gradient becomes one and
leachate release for surface impoundments are computed in the same manner as for
landfills. Applying the equations used to derive ground water releases for
landfills yields the leachate releases.for surface impoundments shown in Exhibit
3A-7.
Spillage
Spills caused by accidental overfilling of an impoundment or by failure of the
leachate collection and treatment system for the double-lined impoundment are
divided between surface water, ground water, and air releases.
No good source of information or modelling technique has been discovered to
assist in predicting the frequency of overfilling surface impoundments. A
properly-run surface impoundment should never be overfilled. This is especially
true for the scenarios used here, in which only 70 percent of the impoundment
capacity is used. Nevertheless, there is at least some remote chance that
overfilling may occur and we assigned our smallest spill interval, 0.00001 percent,
to the possibility of overfilling. Such a possibility translates into one day of
overfilling every 20 years for one of one thousand surface impoundments.
The leachate collection and treatment system for the double-lined surface
impoundment is subject to failure and resulting spillage. The leachate treatment
system for the double-lined impoundment might be any one or a combination of the
hazardous waste treatment technologies discussed earlier. We selected a spill
interval of 0.001 percent as a typical spill rate for leachate treatment systems.
ICF Incorporated
-------
EXHIBIT 3A-7
QUANTITY AND QUALITY OF LEACHATE
GENERATED BY SURFACE IMPOUNDMENTS
DOUBLE-LINED
SINGLE-LINED
FACTORS
During
Operat ion
After
Closure
Duri ng
Operat ion
Af te r
C1osure
Du r i ng
Ooe ra t i on
Percent leachate
removed
100%
NA
0%
NA
0%
K (m/yr)
s
NA
.0315
.0315
.0315
3. 15
T (yrs)
1 3
g (m /yr)
20 - 20 yrs.
protect ion
by dua1
1 i ner = 0
yrs.
0
80 yrs. - 20 yrs.
protection by
dua1 cove r =
60 yrs.
280
20 yrs. - 10 y"rs.
protection by
c1 ay 1i ne r =
10 yrs.
840
80 yrs. - 10 yrs.
protection by
clay cover =
70 yrs.
280
20
8,400
(m )
0
16,800
8,400
19,600
168,000
Re 1ea se qua 1i ty
NA
if residual waste
present after
closure, contami-
nated infiltration;
i f none present,
no contamination
1i qu i d wa ste
on ly
if residual waste
present after
closure, contami-
na ted i nf i11 ra t ion;
if none present,
no contamination
1 i q u i d
waste
on ly
UNLINED
After
NA
3.15
80
4,500
360,000
>
w
ro
vo
o
res i duaI
wastes
present
a fte r
cIosure,
contam i na t
ed
i n f i 11 ra t i
on; if
none
present,
no
contam i-
na t i on.
3
O
O
¦U
0
1
d>
r+
(T>
a.
q = K
A delta T; where G = 3 during operations; G = I after closure.
2. Soil is highly permeable; thus permeability does not limit infiltration. Infiltration was calculated using run-off
2
coefficient of 0.5, as follows: q = (40 inches/yr.) (.0254 in/m) (8,900 m ) (0.5 runoff).
-------
A3-30
c. Cost Computations
We developed costs for using surface impoundments by estimating unit costs for
components of the three scenarios. These cost figures are presented in Exhibits
3A-8 through 3A-10. The most significant of these components in terms of
characterizing each of the three scenarios is discussed below.
d. Double-Lined Surface Impoundment
The double-lined surface impoundment includes the use of a double liner
system, leachate collection and treatment, and a double cover system. Cover and
liner systems are combination synthetic and clay materials.
We assume the synthetic liner provides complete protection against subsurface
waste migration during the 20 year operation of the impoundment. Upon closure of
the site, all liquid waste remaining in the impoundment is removed, and a double
cover system placed over the site. Again, we assume the synthetic cover prevents
any infiltration through the closed site for 20 years, after which time
infiltration passes through the clay cover at its permeability rate (10-7 cm/sec or
3.15 cm/yr).
If all of the constituent of concern is in liquid form during operations, then
at closure all remaining constituent of concern is assumed to be removed and
subsequent infiltration cannot become contaminated. In this case there would be no
programmed release to ground water for the double-lined impoundment.
Where some part of the constituent of concern is in suspended solid form in
the liquid waste, then we assume the constituent is available to contaminate
infiltration. During the 100 year period of reference, infiltration occurs during
60 years (20 years of liner protection plus 20 years of cover protection). We
assume all of this infiltration is saturated, provided that some constituent of
concern is left in the impoundment after closure.
e. Single-Lined Surface Impoundment
Our single-lined surface impoundment scenario is the same size and
configuration as the double-lined scenario. Important differences include the use
of single-layer, in-place clay liner and final cover systems as compared to the
dual clay/synthetic liner systems for double-lined impoundments. No leachate
collection system is included.
This scenario includes fewer personnel and pieces of equipment than the
double-lined impoundment, with a resulting lower cost. Post-closure care is also
less extensive than that for the double-lined scenario, due to reduced effort in
site maintenance.
Ground water releases include those during site operations and those after
site closure. During operations, the clay liner provides complete protection for
ten years, after which liquid waste passes in saturated flow through the liner. At
closure, we assume all remaining liquid waste is removed from the impoundment.
Where some part of the constituent of concern is present in suspended solid form,
we assume a residue is available after site closure which contaminates
infiltration.
ICF Incorporated
-------
A3-31
EXHIBIT 3A-8
DOUBLE-LINED SURFACE IMPOUNDMENT COST ESTIMATE
A. Immediate Capital Costs
Item
Land, 50 acres
Equipment
Building
Fencing, 650 m
Construction
Excavation, 23,800 m3
Synthetic Liner, 10,400m2
Clay liner, 10,000m2
Sand Layers, 4,760m3
Berms, 10,000m3
Leachate Collection, 600m
Leachate Treatment
Monitoring Wells, 6 ea
Unit
Cost
5,000/acre
100,000
10,000
40/m
2.50/m3
8/m2
5/m2
8.5/m3
15/m3
2 5/m
100,000
3,000 each
Discount (3%)
Factor for
Present Value
1
1
1
1
1
1
1
1
1
1
1
1
SUBTOTAL, IMMEDIATE CAPITAL COSTS
Total ($)
250,000
100,000
10,000
@S*,000
59,500
83,200
50,000
40,460
150,000
15,000
100,000
18,000
677,160
B. Replacement Capital Costs
Item
Equipment @ 10 yrs.
Unit
Cost
100,000
Discount
Factor P/F
3%, 10 years
0.7441
Total ($)
74,410
ICF Incorporated
-------
A3-32
EXHIBIT 3A-8 (cont.)
DOUBLE-LINED SURFACE IMPOUNDMENT COST ESTIMATE
Operating and Maintenance Costs
Item
Personnel
Fuel and Maintenance
Utilities
Security
Sample Collection/Analysis
Capital Costs at Closure
Item
Synthetic Cover 10,500m2
Clay Cover 5,000m^
Revegetation 10,000m2
Unit
Cost
70,000/yr
40,000/yr
2,500/yr
6,000/yr
9,600/yr
Discount
Factor P/A
3%, 20 years
14.82
14.82
14.82
14.82
14.82
SUBTOTAL, OPERATING AND
MAINTENANCE
Unit
Cost
4.5/m2
3/m2
1.25/m2
Discount
Factor P/F
3%, 20 years
0.5488
0.5488
0.5488
Total ($)
1,037,400
592,800
37,050
88,920
142,270
1,898,440
SUBTOTAL, CAPITAL COSTS
AT CLOSURE
Total ($)
25,930
13,720
6,860
46,510
ICF Incorporated
-------
A3-33
EXHIBIT 3A-8 (cont.)
DOUBLE-LINED SURFACE IMPOUNDMENT COST ESTIMATE
Post Closure Care Costs
Item
Site Maintenance, perpetual
Monitoring, perpetual
Discount Factor
Unit P/A 3%, perpetual
Cost P/F 3%, 20 yrs
1,000/yr 18.29
3,200/yr 18.29
SUBTOTAL POST-CLOSURE CARE
COST
Total ($)
18,290
58,528
76,818
Total Cost for Stringent Surface Impoundment
Immediate Capital Costs
Replacement Capital Costs
Operating and Maintenance Costs
Capital Costs at Closure
Post Closure Care Costs
677,160
74,410
1,898,440
46,510
76,818
TOTAL COST 2,773,338
WASTE CAPACITY 100,000m3
UNIT COST 27.73/m3
ICF Incorporated
-------
A3-34
EXHIBIT 3A-9
SINGLE-LINED SURFACE IMPOUNDMENT COST ESTIMATE
Immediate Capital Costs
Item
Land, 5 acres
Equipment, LS
Trailer
Fencing, 650 m
Construction
Excavation, 23,800m3
Clay liner 10,000m^
Berms 10,000m3
Monitoring Wells, 4 ea
Unit
Cost
5,000/acre
80,000
5,000
20/m
2.50/m3
5/m^
10/m3
3,000 each
Discount (3%)
Factor for
Present Value
1
1
1
1
1
1
Total ($)
25,000
80,000
5,000
13,000
59,500
50,000
100,000
12,000
SUBTOTAL, IMMEDIATE CAPITAL COSTS
344,500
Replacement Capital Costs
Item
Equipment § 10 yrs.
Unit
Cost
80,000
Discount
Factor P/F
3%, 10 years
0.7441
Total ($)
59,530
ICF Incorporated
-------
A3-35
EXHIBIT 3A-9 (cont.)
SINGLE-LINED SURFACE IMPOUNDMENT COST ESTIMATE
C. Operating and Maintenance Costs
Discount
Item
Unit
Cost
Factor P/A
3%, 20 years
Total ($)
Personnel
50,000/yr
14.82
741,400
Fuel and Maintenance
32,000/yr
14.82
474,240
Utilities
2,000/yr
14.82
29,640
Security
5,000/yr
14.82
74,100
Sample Collection/Analysis
5,200/yr
14.82
77,065
SUBTOTAL, OPERATING AND
MAINTENANCE
1,396,045
Capital Costs at Closure
Item
Unit
Cost
Discount
Factor P/F
3%, 20 years
Total ($)
Clay Cover 5,000m3
CM
m
0.5488
13,720
Revegetation 10,000m^
1.25/m2
0.5488
6,860
SUBTOTAL, CAPITAL COSTS
AT CLOSURE
20,580
ICF Incorporated
-------
A3-36
EXHIBIT 3A-9 (cont.)
SINGLE-LINED SURFACE IMPOUNDMENT COST ESTIMATE
Post Closure Care Costs
Item
Unit
Cost
Site Maintenance, perpetual 500/yr
Monitoring, perpetual 70/yr
Discount Factor
P/A 3%, perpetual
P/F 3%, 20 yrs
18.29
18.29
SUBTOTAL POST-CLOSURE CARE
COST
Total ($)
9,145
15,912
25,057
Total Cost for Moderate Surface Impoundment
Immediate Capital Costs
Replacement Capital Costs
Operating and Maintenance Costs
Capital Costs at Closure
Post Closure Care Costs
TOTAL COST
344,500
59,530
1,396,045
20,580
25,057
1,845,712
WASTE CAPACITY 114,300m"
UNIT COST
$16.05/m3
ICF Incorporated
-------
A3-37
EXHIBIT 3A-10
UNLINED SURFACE IMPOUNDMENT COST ESTIMATE
Immediate Capital Costs
Item
Land, 5 acres
Equipment, LS
Trailer
Construction
Excavation, 23,800m3
Berms 10,000m3
Unit
Cost
5,000/acre
50,000
5,000
2.50/m3
15/m3
Discount (3%)
Factor for
Present Value
1
1
TOTAL IMMEDIATE CAPITAL COSTS
Total ($)
25,000
50,000
5,000
59,500
150,000
289,500
Replacement Capital Costs
Item
Equipment 0 10 yrs.
Unit
Cost
50,000
Discount
Factor P/F
3%, 10 years
0.7441
Total ($)
37,205
ICF Incorporated
-------
A3-38
EXHIBIT 3A-10 (cont.)
UNLINED SURFACE IMPOUNDMENT COST ESTIMATE
Operating and Maintenance Costs
Item
Personnel
Fuel and Maintenance
Utilities
Security
Unit
Cost
40,000/yr
20,000/yr
1,500/yr
3,500/yr
Discount
Factor P/A
3%, 20 years
14.82
14.82
14.82
14.82
SUBTOTAL, OPERATING AND
MAINTENANCE
Total ($)
592,800
296,400
22,230
51,870
963,300
Capital Costs at Closure
Item
Revegetation 10,000m^
Unit
Cost
0.60/m2
Discount
Factor P/F
3%, 20 years
0.5488
Total ($)
3,295
Total Cost for Uncontrolled Surface Impoundment
Immediate Capital Costs
Replacement Capital Costs
Operating and Maintenance Costs
Capital Costs at Closure
Post Closure Care Costs
289,500
37,205
963,300
3,295
0
TOTAL COST 1,293,300
WASTE CAPACITY 268,000m3
UNIT COST 4.83/m3
ICF Incorporated
-------
A3-39
f. Unlined Surface Impoundment
The unlined surface impoundment scenario is a poor hazardous waste management
facility. It more resembles a sink hole than a disposal impoundment. This
scenario has no liner system at all; waste is placed in an unlined hole and allowed
to seep into the ground.
Fewer personnel and pieces of equipment are required to operate our unlined
impoundment, with resulting lower costs. Final closure includes only minimal
grading. There is no post-closure care for the unlined scenario.
Losses to the ground water are much higher in the unlined scenario than in the
other scenarios. This has the effect of giving the unlined surface impoundment
much more waste capacity than the other impoundment scenarios, since waste which
leaks from the unlined impoundment can be replaced with more waste.
After site closure, any remaining liquid waste is removed, and the area
filled. If the waste placed in the impoundment contained any of the constituent of
concern in solid form, infiltration may become contaminated. Contaminated
infiltration is released to the ground water throughout the post-closure period.
LAND TREATMENT
a. Description
Land treatment is a process in which industrial wastes are mixed with surface
soils and allowed to decompose. Land farming, land cultivation and soil
incorporation are other names for land treatment, as the term is used here.
The objective in land treatment is to assure that chemical constituents in the
waste are retained and/or decomposed in the surface layer. Microbial degradation
is the principal decomposition mechanism; volatilization and chemical and
photochemical degradation are also important processes in land treatment.
Since microbial degradation is the principal treatment mechanism in land
treatment, the process is not suited to wastes too toxic or too persistent to be
degraded. In addition, wastes accepted for land treatment should not adversely
affect soil structure, resulting in impaired internal drainage and/or oxygen
transfer.
Hazardous or potentially hazardous wastes from the following industries are
candidates for land treatment:
Textile finishing
Wood preserving
Paper and allied products
Organic fibers (non-cellulosic)
Drugs and pharmaceuticals
Organic chemicals
Petroleum refining
Leather tanning and finishing
ICF Incorporated
-------
A3-40
For our model, only those wastes which are biodegradeable are candidate wastes for
land treatment.
Land treatment sites come in many sizes and configurations. Based upon a
review of the Part A Permit Application data for land treatment facilities and our
professional judgment, we selected a 10-acre operating area as typical for a land
treatment facility. Such a facility would require a 20-acre site, and is further
described in Exhibit 3A-11.
EXHIBIT 3A-11
LAND TREATMENT SITE FEATURES
• Facility 20 acres
• Operating area 10 acres
• Life Span 20 years
• Waste Capacity 4,000 m3/year; 80,000 m3 total
• Storage Lagoon
- Capacity 1,000 m3 (1/4 of annual waste
- 2:1 slope with berm quantity received)
- Lined with 36-mil Hypalon with
sand subgrade and protective
soil cover
• Application rate/year 400 m3/acre
• Runon diversion berm (perimeter berm)
• Runoff collection ditches
• Runoff containment basin
• All-weather access road
• Monitoring of soil, soil-water (leachate), surface runoff, and ground
water
• Site Closure
- Removal of contaminated soil and waste residue from lagoon and basin
- Application of material removed to land treatment site
- Filling and revegetation of lagoon and basin
- Operation of land treatment site for three additional years or until
waste residue has been degraded
- Revegetation of land treatment site
• Post-Closure care
- Maintaining soil pH>6 by liming
- Maintaining vegetative cover
- Monitoring ground water at reduced frequency
b. Environmental Releases
Releases from land treatment facilities include those to air, ground
water, and surface water. Estimates were made for each of these releases
based upon waste characteristics (vapor pressure, solubility, etc.) and
assumed operating conditions.
Volatilization
Wastes in the land treatment facility are exposed to the atmosphere for
extended time periods. At any given time, roughly half the waste at a given
facility will lie open; the other half will have been mechanically
incorporated into the soil to a depth of 6-12 inches.
ICF Incorporated
-------
A3-41
Thus, during the 20-year operating life, some 5 acres will be continuously
exposed to the atmosphere. Using vapor pressure, concentration and molecular
weight, we determined the emission rate for uncovered waste. Once incorporated with
the soil, the waste may continue to volatilize, although at a reduced rate. At site
closure, all of the waste will be incorporated, and emissions will continue for a
short time (until degradation is complete). This post-closure emission period was
conservatively assumed to be 5 years, although most wastes will degrade in a
shorter time period. The specific derivation of the air emission rates for land
treatment facilities is as follows:
I. When the waste is exposed, use the following equation (see Landfill
section):
dv = 2*CeW •¦j/DLV/irFv'Wi
dt
Ce = P/760
W = 100m = 10"cm
D = 0.9/ M
L = 200m = 2 x 10"cm
v = 4 m/sec = 400 cm/sec
Fv = 1
Wi = (concentration in ppm) x 10 ^
Thus,
dv = 2* (P/760) • 10u-j^0.9,2xl0'*,4xl02) / (M^*3.14# 1 CxlO-6
dt
= 2.63x10-5•1.51xl03 *PC/M^
dv = 3.97x10-2*PC/M*
dt
Ei = dv*M/22.4xl03
St
Ei = 1.77PCM'^^ in yg/sec
ICF Incorporated
-------
A3-42
II. When the waste is not exposed, use the following equation (see
Landfill section):
Ei = DCsAPt1'33,Wi/L
L = 15 cm
A = 2 ha = 2 x lO^m2 = 2 x 108cm2
Pt = use 0.50
Wi = (concentration in ppm) x 10 ^
Ei = 0.9/M^,PM/18.57,2x108,0.501'33,(CxlO-6/15
= 0.7146x102/278.6*PM^C
i
Ei = 0.256 PM2C
For a given time, use E^ + E^ for total emissions.
In the 20 years of operation,
E. = (1.77 PCM-75 + 0.256PM*C) x 20 years
75 i
= (1.77 PCM + 0.256PM C) x 0.63
(answer in kg/20 years)
Assuming emission exists for 5 or more years after site closure, then
E2 = 1/4 E1 .
(answer in kg/5 years)
Total air emissions are computed over 20 years of operation plus 5 years after
closure.
Ground Water and Surface Water Releases
During the 25 year period of interest (20 years of operations plus 5 years for
total degradation), some 1,000 inches of precipitation will have fallen upon the
land treatment facility. This translates to just over one million cubic meters
(40,000 m3/year) of precipitation over the 10 acre operating area of the facility.
Runoff and infiltration resulting from this rainfall will be the sources of surface
water and ground water releases, respectively.
Land treatment sites normally have permeable surface soils to promote
microbial growth. Thus infiltration of rainfall at such sites is often high, and
runoff, low. We have assumed a 30 percent runoff coefficient (typical of some
pasture or farmland) and that 30 percent of the annual precipitation evaporates.
Thus of the 40,000 m3 of rainfall striking the operating area,
30% or 12,000 m3 runs off
30% or 12,000 m3 evaporates
40% or 16,000 m3 infiltrates.
ICF Incorporated
-------
A3-43
All of the infiltration is exposed to waste being treated, regardless of
whether the waste is incorporated in the soil. Only half of the runoff is exposed
to waste, because only half the operating area contains exposed waste.
Furthermore, a significant portion of the runoff (we have assumed 50 percent) is
collected in the runoff collection system.
To compute releases to ground water and to surface water from the land
treatment site, we have assumed all runoff from exposed waste areas and all
infiltration will be saturated with the constituent of concern. Thus, 6,000 m3/yr
(16.5 m3/day) of saturated runoff is released to surface waters and 16,000 m3/yr
(44 m3/day) of saturated infiltration is released to ground water.
c. Cost Computations
Up-front capital expenses (about $4.50/m3 of waste) are lower for land
treatment than for most other land disposal alternatives. This is because land
treatment does not require the extensive site preparation expense (i.e. excavation)
required for landfills or surface impoundments. Taken together, all capital costs
for land treatment total about $500,000, or $6.20/m3 of waste.
Land values noticeably affect the total capital costs for land treatment, as
shown below:
Land Value Total Capital Costs Capital Costs $/m3 of waste
$1,000/acre $420,000 $ 5.20
$5,000/acre $500,000 $ 6.20
$20,000/acre $800,000 $10.00
However, the impact of the land values on the total cost of land treatment is
not as significant.
Land Value Total Costs Unit Costs
$1,000/acre $1,770,000 $22/m3 of waste
$5,000/acre $1,850,000 $23/m3 of waste
$20,000/acre $2,150,000 $27/m3 of waste
The unit costs of the components of our land treatment scenario are shown
in Exhibit 3A-12.
DEEP WELL INJECTION
a. Description
Deep well injection is a process whereby wastes are pumped under pressure into
an area below the earth's surface. A number of issues surround the practice,
including the possibility of system failure (resulting in contamination of a usable
drinking water aquifer), possible effects of the practice on deep geologic
formations (resulting in seismic activity) and the possibility that future
generations may wish to use the deep disposal areas for another (incompatible)
purpose. Of these, only the first--the possibility of system failure--is amenable
to formal consideration in this analysis.
ICF Incorporated
-------
A3-44
EXHIBIT 3A-12
LAND TREATMENT COST ESTIMATE
Immediate Capital Costs
Item
Land, 20 acres
Equipment
Track-dozer
Rototiller
Tank Wagaon
Miscellaneous
Construction
Site clearing/grading
Storage lagoon
Runoff Collection Ditches
Runoff Containment Basin
Perimeter Berms
Fending, gates, signs
Building
Monitoring wells, 5 ea
Unit
Cost
5,000/acre
119,000
Discount (3%)
Factor for
Present Value
1
1
250/ac
3000/each
TOTAL IMMEDIATE CAPITAL COSTS
Total ($)
100,000
119,000
2,500
17,000
4,500
28,000
25,000
41,000
10,000
18,000
365,000
Replacement Capital Costs
Item
Equipment @ 10 yrs.
Unit
Cost
119,000
Discount
Factor P/F
3%, 10 years
0.7441
Total ($)
88,500
1CF Incorporated
-------
A3-45
EXHIBIT 3A-12 (cont.)
LAND TREATMENT COST ESTIMATE
C. Operating and Maintenance Costs
Item
Labor for 20 years
Equipment Fuel/
Maintenance
Utilities and Security
Sample Collection/Analysis
Unit
Cost
50,000/yr
12,000/yr
15,000/yr
10,000/yr
Discount
Factor P/A
3%, 20 years
14.82
14.82
14.82
14.82
SUBTOTAL, OPERATING AND
MAINTENANCE
Total ($)
741,000
177,840
222,300
148,200
1,289,340
D. Capital Costs at Closure
Item
Construction, LS
Unit
Cost
77,000
Discount
Factor P/F
3%, 20 years
0.5488
Total ($)
42,250
ICF Incorporated
-------
A3-46
EXHIBIT 3A-12 (cont.)
LAND TREATMENT COST ESTIMATE
Post Closure Care Costs
Item
Site Maintenance
grading pH adjustment
Monitoring
Unit
Cost
4,500/yr
3,170/yr
Discount Factor
P/A 3%, 30 yrs
P/F 3%, 20 yrs
8.238
8.238
SUBTOTAL POST-CLOSURE CARE
COST
Total ($)
37,076
25,540
62,616
Total Cost for Land Treatment
Immediate Capital Costs
Replacement Capital Costs
Operating and Maintenance Costs
Capital Costs at Closure
Post Closure Care Costs
365,660
88,500
1,289,340
42,250
62,616
TOTAL COST
1,847,700
ICF Incorporated
-------
A3-47
To be suitable for deep well injection, the waste must not bind or clog the
deep formation receiving the waste. Suspended solids content must generally be
low, and certain dissolved constituents may react with the natural fluids in the
deep formation and cause precipitation and clogging. We have set the condition
that to be injectable, the suspended nonwater fraction of the waste must be less
than or equal to 0.005, i.e., 5,000 ppm. Such a figure is too high for some
geologic formations (sandstone) but is acceptable for others (cavernous,
fractured, or limestone). Wastes with suspended solids content of 50 to 5,000 ppm
are allowed to be injected in this analysis with the caveat that injection may not
be feasible.
Candidate sites for deep well injection facilities must include a geological
formation which can accept liquids located below useable aquifers, and which is
separated from such aquifers by an impermeable layer.
Depths of deep wells used for injecting hazardous wastes vary widely depending
on geologic characteristics. Deep wells for hazardous wastes are generally deeper
than other waste injection wells. We think most hazardous waste deep well
injection systems are between 3,000 and 7,000 feet deep. We selected a one-mile
deep well as typical for use in this project. Other assumptions for deep well
injection include:
• EPA Class I Well
injection tubing placed inside protective casing
casing cemented to well bore
annular space between tubing and casing filled with
solution of proper density
• Pumping rate of 140 gpm
• Useful life of 10 years
• Total capacity of 700 million gallons (2.65 million cubic
meters)
• Enclosed pressure/surge tanks
The total capacity used for our typical deep well injection facility is about
one standard deviation lower than the 5.5 billion gallon capacity reported in Part
A permit applications. We suspect that many operators of deep well facilities
either overstated their capacity or are using the same formation for disposal.
Wells located elsewhere in the nation could not take advantage of the same large-
capacity formation.
b. Environmental Releases
A deep well injection facility which is properly designed, installed and
operated should not experience significant release to any medium. By monitoring
injection pressures, the facility operator can readily determine problems in the
well shaft or disposal area. Once problems are detected, the well operation can be
quickly halted.
ICF Incorporated
-------
A3-48
For purposes of estimating release rates, we have assumed that some part of the
system might fail and not be detected for some period of time. This approach was
used because no good source of information or modelling technique for deep well
releases was available. Two kinds of system failures were considered: failure at
or before the wellhead, and failure in the well shaft
Losses due to the first kind of failure would be spills on the surface of the
ground. The algorithm used for treatment process spills is used here to divide
spilled material into releases to air, surface water, and ground water. Wellhead
valves, fittings and pumps are subject to catastrophic failure. We assumed that
one hour's waste flow per year (140 gpm x 60 minutes = 8,400 gallons) would be lost
over the 10-year life of the facility. This corresponds to a spill rate of
8,400/70,000,000 = 1.2X10-11 due to the wellhead failure.
The second kind of failure is harder to detect and more significant than
surface spills. A failure in the well shaft might not be detected for some time,
even where daily logs are kept of injection pressures, etc. We have assumed that
one full day's flow (140 gpm x 1,400 minutes = 201,600 gallons of waste) could go
undetected before repairs are made over the 10 year life of the facility. All of
these losses are releases to ground water. This corresponds to a release rate of
201,600/700,000,000 = 2.88x10-''.
c. Cost Computations
Costs for deep well injection are mostly for operations (utilities) and
maintenance. Deep well injections facilities do not require much surface area;
most of the capital expense involved is in construction of the facility. The
derivation of unit costs for deep well injection is shown in Exhibit 3A-13.
EXHIBIT 3A-13
COST ESTIMATE FOR DEEP WELL INJECTION
(Total Capacity = 2.65 million m3)
Discount Factor
Element
Unit Cost
P/A 3%, 10 yrs Total $
Immediate Capital Expense $1 million
Land
Construction
$ 1 million
($0.40/m3)
Operating Expense (utilities) $1.4 million/year
8.511
$11.9 million
($4.50/m3)
Maintenance Costs
Repairs
Monitoring
$70,000/year
8.511
$ 0.6 million
($0.20/m3)
Total Cost, Deep Well Injection
$13.5 million
Unit Cost
$ 5.10/m3
ICF Incorporated
-------
A3-49
OCEAN DISPOSAL
a. Description
The oceans have historically been used as a repository for a variety of
wastes. The practice has come under increasing scrutiny within the last
decade, and is now regulated to some extent under international as well as
United States law.
Any waste which does not float is a technically-feasible candidate for
ocean disposal. In ocean disposal, wastes are either dumped in bulk from
special barges or placed in containers (typically, weighted or perforated
55-galIon drums) and dumped. Disposal areas for hazardous wastes are normally
well off-shore, in areas without significant aquatic life which would be
immediately affected by the practice.
b. Environmental Releases
We consider deep ocean waters a special case of surface waters. Thus,
releases from ocean disposal are releases to surface waters. Losses during
land transportation and loading of wastes are considered transportation
losses, and are discussed in Appendix 2.
Once waste is placed in the ocean, the physical characteristics of the
waste govern the release rate to surface waters. We assume that all of the
liquid-phase constituent of concern is immediately released to surface
waters. Solid phase constituents present a different problem.
Long-term data on the durability of solids placed in the ocean have not
been developed. We believe deterioration and release of solid phase
components begins almost immediately, and continues over a 5-50 year period
for most wastes. We set a 10-year period as typical for complete release of
solid-phase constituents, an assumption which we believe to be conservative
(i.e., many solid wastes will remain intact after 10 years in the ocean
environment). Thus, 1/10 of the solid phase constituents are released on an
annual basis, together with all liquid-phase constituents.
c. Cost Computations
We approached unit costs for ocean disposal somewhat differently than
other disposal technologies, due to several major differences. Capital
expenditures for ocean disposal include not only the equipment (work ship,
barges) actually used in ocean disposal, but wharfage as well. Wharfage need
not be solely dedicated to hazardous waste disposal, but is normally used for
other ocean shipments as well. This co-existing use of capital facilities is
inconsistent with other disposal technology land uses, which are
single-purpose sites. We, therefore, rely upon reported costs from the
literature to determine unit costs for ocean disposal.
A wide range of costs ($1.70/ton - $24/ton +) is reported for ocean
disposal,- depending primarily upon the distance from shore to the disposal
area. A typical range for hazardous waste ocean disposal is $8.00/ton -
$11.00/ton. We used the mid-point in this range, $9.50/ton or about
(10.50/metric ton), as the cost of ocean disposal.
ICF Incorporated
-------
APPENDIX 4
RELEASE RATES FOR TREATMENT TECHNOLOGIES
ICF Incorporated
-------
APPENDIX 4
RELEASE RATES FOR TREATMENT TECHNOLOGIES
In order to obtain an order-of-magnitude approximation of the uncontrolled
release of hazardous component X to the environment for each treatment technology,
the major sources of emission were grouped into four categories:
Evaporation
Aeration
Routine Spillage
Accidental Spillage
The evaporative loss of X from open process-vessels is a function primarily of
the physical properties of X. Methods for estimating this emission are explained
in Section 2 of this appendix.
The aeration, routine spillage, and accidental spillage losses are more heavily
dependent upon the processes and equipment used during treatment and the care with
which the system is operated than on the physical properties of X. These three
losses are estimated individually and added to give an overall loss fraction as a
function of the influent sludge flow to the treatment unit. These overall loss
fractions are grouped into categories and assigned a release rate code (RCODE)
representing the magnitude of the loss inherent to the utilization of each
treatment option. The estimation of the aeration and spillage losses, and their
incorporation into the RCODE parameter are described in the following section.
SECTION 1
DETERMINATION OF RELEASE CODES FOR TREATMENT TECHNOLOGIES
AERATION
When open vessels are mechanically agitated, a portion of the sludge may be
lost in the form of fine droplets which are dispersed into the atmosphere. This
aspiration loss is a function of the vessel size, the type of mixing used, and the
rate of agitation. It is determined for treatment units which typically utilize
open vessels.
Mixing rates for open vessels used in the sludge treatment technologies are
listed qualitatively in Exhibit A4-1. The fraction of the influent sludge aerated
during mixing is estimated according to this mixing speed for each vessel in the
treatment train. The total aeration loss fraction is the sum of the loss fractions
for the individual vessels in each technology.
ROUTINE SPILLAGE
A portion of a sludge stream may be lost on a regular basis due to poor
materials handling techniques. Such routine spillage is primarily a function of
equipment design, maintenance, and operation.
ICF Incorporated
-------
A4-2
EXH
AERATION LOSSES
I BIT A4-1
FOR OPEN TREATMENT VESSELS
T reatment
Unit/Subunit
Chemical Coagulation
Flash Mixer
Flocculator
Clarifier
Vacuum Filter
Sludge Vat
Evaporation/Drying
Evaporator
Solvent Extraction
Contact Tanks
Coalescent Tank
Qualitative
Mixing Rate
Rapid
Slow
Slow
Slow
Slow
Fraction of
Sludge Aerated
10-5
10-7
10-7
total
Rapid
None
10-5
10-7
10-7
10-5
Leaching
Contact Tank
Clarifier
Carbon Adsorption, PAC
Flash Mixer
Contact Mixer
Clarifier
Chemical Precipitation
Flash Mixer
Flocculator
Clarifier
Chemical Destruction
Reactor
Chemical Fix./Stabilization
Reactor
Electrolytic Decomp.
Electrolysis Tank
Rapid
Slow
Rapid
Slow
Slow
Rapid
Slow
Slow
Intermediate
Rapid/Intermediate
Slow
total
total
total
10-5
10-7
10-5
10-5
10-7
10-7
10-5
10-5
10-7
10-7
10-5
10-s
10-
10-7
ICF Incorporated
-------
A4-3
The estimated losses due to routine spillage are listed in Exhibit A4-2 as
fractions of the influent sludge flow. Losses due to routine maintenance and
materials handling were listed for batch treatment units or those which require
periodic shutdown for cleaning or unloading. These losses were estimated as
functions of vessel size and the expected length of time between shutdowns. The
losses due to splashing in open agitated tanks were estimated as a function of
qualitative mixing speed similar to the method used for estimating aeration losses.
The total routine spillage values listed were used in the calculation of
overall release rates.
ACCIDENTAL SPILLAGE
Large volumes of sludge may be spilled when catastrophic equipment failure
occurs, forcing unscheduled equipment shutdown and loss of vessel contents. This is
usually caused by pipeline plugging, instrument malfunction, or equipment
breakdown. Such accidental spillage is a function of the properties of the sludge,
the age and maintenance history of the treatment equipment, and the care with which
the system is operated. It is estimated for all treatment units.
The estimated losses due to accidental spillage are listed in Exhibit A4-3 as
fractions of the influent sludge flow to each technology. "Best case," "worst
case," and "typical" values are given.
RELEASE CODE CALCULATIONS
The "typical" loss fractions for routine and accidental spillage are added to
the aeration loss to give an overall loss fraction for each treatment module. The
overall loss fractions are grouped into categories according to their order of
magnitude and assigned a release rate code (RCODE) representing the magnitude of
the loss. The resulting RCODE values for all treatment units are shown in Exhibit
A4-4. The value of RCODE is an indicator of the magnitude of the losses inherent to
a given treatment technology and is used to estimate the emission of hazardous
component X to the environment during treatment.
SECTION 2
DETERMINATION OF EVAPORATIVE EMISSIONS FROM TREATMENT TECHNOLOGIES
The emission of hazardous component X to the environment due to the selective
volatilization of X will occur from open process vessels and from sludge that
spills onto the ground. In order to predict the total evaporative loss of X from a
treatment unit, the rate of evaporation of X from an open tank is first predicted.
The results for this case are then generalized and applied to open sludge treatment
vessels and to waste spilled during sludge treatment.
DERIVATION OF EVAPORATION RATE EQUATION
In order to predict the loss of component X due to evaporation from a sludge
stream, the rate of evaporation of X must be predicted. This evaporation rate is a
function of many parameters:
ICF Incorporated
-------
A4-4
EXHIBIT A4-2
ROUTINE SPILLAGE FRACTIONS OF INFLUENT SLUDGE FLOW
(kg/day)
Technology
Chem. Coagulation
Filter Press
Centrifuge
Vacuum Filter
Evaporation/Drying
Air Stripping
Steam Stripping
Solvent Extraction
Leaching
Distillation
Reverse Osmosis
Carbon Adsorp., PAC
Ion Exchange
Chem. Precipitation
Chem. Destruction
Electrolytic Decomp.
Chem. Fix./Stabilization
Routine
Maintenance
8.82x10-
1.04xl0-6
5.20x10-3
Materials
Handling
3.3x10-"
3.3xl0-6
2.50xl0-5
Mixer
Splashing
1.0x10-"
l.OxlO-6
l.OxlO-6
1.0x10-
1.0x10-
1.0x10-"
l.OxlO-6
l.OxlO-6
1.0x10-"
Total
Routine
Spillage
1.0x10
1.2x10
1.2x10-
4. 3xl0-s
1.0x10-"
1.0x10-"
l.OxlO-6
5.2xl0-3
1.0x10-"
l.OxlO-6
2.6xl0-5
1.0x10-"
ICF Incorporated
-------
A4-5
EXHIBIT A4-3
ACCIDENTAL SPILLAGE FRACTIONS OF INFLUENT SLUDGE
Technology
Unit/Subunit
Best Worst
Case Case
(spills/yr) (spills/yr)
Typical
Case
(kg/day)
Chemical Coagulation
Flash Mixer
Flocculator
Clarifier
Total
Filter Press
Centrifuge
Vacuum Filter
Evaporation/Drying
Evaporator •
Air Stripping
Steam Stripping
Solvent Extraction
Coalescent Tank
Coalescent Tank
Total
Leaching
Contact Tank
Clarifier
Distillation
Reverse Osmosis
Carbon Adsorption (PAC)
Flash Mixer
Contact Mixer
Clarifier
Total
Ion Exchange
Chemical Precipitation
Flash Mixer
Flocculator
Clarifier
Total
Chemical Destruction
Reactor
Electrolytic Decomp.
Chemical Fix./Stabilization
Reactor
Extruder
1
1
0.5
0.5
6
1
0.1
1
2
1
0.5
0.5
0.5
0.5
1
1
1
0.5
0.5
0.1
1
0.5
0.5
0.5
0.1
1
12
12
12
2
100
12
1
24
12
12
6
6
12
2
12
24
12
12
2
12
12
2
6
1
24
4.284xl0-5
1.923xl0-5
1.997x10-*
2.6x10-"
1.056x10-"
2.229x10-"
1.110x10-"
7.212x10-"
1.319x10-"
2.137x10-"
1.068x10-"
5.342xl0-5
1.6x10-"
9.616xl0-5
1.997x10-"
2.137x10-"
1.923x10-"
. 284xl0-5
.923x10-"
.997x10-"
.3x10-"
4.270xl0-9
4.284xl0-5
1.923xl0-5
1.997x10-"
2.6x10-"
1.068xl0-5
2.226x10-"
1.54xl0-3
ICF Incorporated
-------
A4-6
EXHIBIT A4-4
RELEASE CODE DETERMINATION
Unit Name
A—1 Chemical Stabilization
A-2 Chemical Precipitation
A-3 Chemical Destruction
A-4 Chemical Coagulation
A-5 Filter Press
A-6 Centrifuge
A-7 vacuum Filter
A-8 Evaporation
A-9 Air Stripping
A-10 Steam Stripping
A-ll Solvent Extraction
A-12 Leaching
A-13 Distillation
A-14 Electrolytic Decomposition
A-15 Reverse Osmosis
A-16 Carbon Adsorption, PAC
A-17 Ion Exchange
A-18 Incineration, 99.99 % DRE
A-19 Incineration, 99.90 * DRE
A-20 Incineration, 99.00 % DRE
A-21 Incineration, 90.00 % DRE
Aeration
Fraction
1.0x10
1.0x10
-5
1.0x10
1.0x10
-5
1.0x10
-7
1.0x10
-7
1.0x10
-5
1.0x10
-5
1.0x10
-7
1.0x10
-5
Typical
Routine
Spillage
Fraction
1.0x10
1.0x10
-4
1.0x10
.-6
1.0x10
-4
1.2x10
-3
4.3x10
-6
1.0x10
-6
1.0x10
-4
1.0x10
-4
2.6x10
-5
1.0x10
-6
1.0x10
5.2x10
-4
Typical
Accidental
Spillage
Fraction
1.5x10
2. 6xl0~
1.1x10"
2.6x10"
1.1x10"
2. 2xl0~
1.1x10"
7. 7xl0~
1.3x10"
2. lxlo"
1. 6x10*
3.Oxlo"
2. lxlo"
2.2x10"
1.9x10*
4.3x10"
4.3x10
-9
Overall
Loss
Fract ion
1.6x10
3.7x10"
1. 2xlo"
3. 7x10*
1. 3x10*
2.2x10"
1.1x10*
7.7x10*
1.3x10"
2. lxlo"
2.7x10*
4.1x10"
2.lxlo"
2.4x10"
1. 9x10*
5. 3x10*
5.2x10"
RCODE
Values
-------
A4-7
the chemical properties of the sludge (volatility of
liquid species, chemical interactions between
components, solids present, etc.)
the physical configuration of the equipment handling
the sludge stream (tank size and shape, degree of
mixing of the tank contents, etc.)
the properties of the treatment unit's environment
(temperature, pressure, wind velocity, relative
humidity, etc.)
To simplify the prediction of evaporative losses while maintaining the goal of
order-of-magnitude results, we developed the model described below.
The vessel used in the development of this model is 5 meters in diameter and
open to the atmosphere. Agitation is vigorous enough to provide a uniform liquid
composition throughout the tank including the surface. Air flow across the liquid
surface is fully turbulent due to the perturbation induced by the tank rim. Wind
velocity is constant at five miles per hour, and the relative humidity is zero. The
temperature of both the vessel contents and the ambient air is 25°C.
The vessel contains a mixture of liquid and solid species, two of which are
hazardous components and water, labeled X and W, respectively. The air surrounding
the tanks is called species A. For low rates of mass transfer (low enough so that
the air's velocity and concentration profiles are undisturbed by the evaporating
species), the average rate of evaporation of a species i into species A may be
defined as follows:
Rate of Evaporation Mass transfer due Mass transfer of i
of species i into A = to diffusion of i + and A away from liquid
into A at surface surface due to bulk
liquid flow
W. = k.A (X. - X. ) + X. (W. - W.)
1 1 10 100 io i A
where VL = flow of species i into A across the interface (kg/L)
X_^o = concentration of species i in A at the tank surface (kg/kg)
X^qo = concentration of species i in the bulk A (kg/kg)
2
A = surface area of liquid (m )
2
k^ = mass transfer coefficient (kg/m t)
WA = flow of species A into i liquid across the interface (kg/t)
Two assumptions are made:
(1) the concentration of species i in the bulk of A is zero, therefore
X. =0
100
ICF Incorporated
-------
A4-8
(2) the solubility of A in i is negligible compared to that of i in A;
therefore
W. - W = W.
1 A a
The resulting equation is:
W. = k.A X. + X. W.
1 1 10 10 1
Rearranging, W. =k. AX. /(1-X. )
® & 1 1 10 10
This is a general expression for the rate of evaporation of pure species i
into species A. For the evaporation of pure Water and pure X into air, then,
the following equations apply:
W = k A • X / (1-X ) W = k A • X / (1-X )
w w wo WO X X xo xo
The individual mass transfer coefficients k and k are usually measured
w x
experimentally, since the number of variables upon which they depend is large
and theoretical prediction is difficult. Their ratio is more easily estimated.
The equation for is therefore expressed as follows:
W /A = k • (k /k ) • X /(1-X )
X w X w xo xo
The equation contains three major terms, estimated as follows:
Estimation of k
W
According to the Chilton-Colburn analogy for heat and mass transfer,
kw . fpV/2 • PDWA/y
where f = the total Fanning coefficient of friction for the flow of air
over the liquid surface (dimensionless)
p = density of air
V = velocity of air stream
y = viscosity of air
= diffusivity of water in air
ICF Incorporated
-------
A4-9
The friction factor f is a function of the Reynolds number Rg, for the
liquid surface:
rel = Lvp/U
where R_T = Reynolds number for flow along the liquid surface
EL
(dimensionless)
L = length of air flow
v = air velocity
y = air viscosity
p = air density
For the tank upon which this model was based,
L = 5m
v = 5 miles/hr. = 2.235m/s
y = .018cp = .018xl0-3 Ns/m2
p = 1.1795 kg/m3
therefore,
"el'7-32*10'
An empirical correlation reported by Schlicting describes the relationship
between f and R for turbulent flow past a flat plate (2 surfaces):
£iL
f = 0.455/(log R£L)2'58
For the single liquid surface in the evaporation model, half of this value may
be used. A value of f for the flow regime may thus be obtained:
f = 4.742xl0-3
Using this value for f, a value for k^ may be obtained:
= .220 cm2/sec
kw = i(0.4742x10-3) • 1.1795 kg/m3 • 2.235 m/sec
• 1.1795 kg/m3 • .220 cm2/sec ? ,018xl0-3 Nsec/m2
• Nsec2/kgm • (m2/10'*cm2)2/'3
= 8.41xl0-3 kg/m2sec
ICF Incorporated
-------
A4-10
Ratio of Mass Transfer Coefficients
The ratio of the mass transfer coefficients for X and water may be found
by considering a Chilton-Colburn analogy for the flow regime. As stated
earlier, the mass transfer coefficient is related to the Fanning friction
factor as follows:
k. = (fpV/2 • pDiA/v)2/3
k /k = (Dv./D„.)2/3
x w XA WA
There are three methods to obtain the ratio of the diffusivities of X and
water in air. Listed in order of accuracy, they are:
(1) Values for D^A and D^A at 25°C may be obtained from technical
literature.
(2) The ratio may be approximated using an empirical correlation of the
diffusivity of a vapor in a gas. According to the correlation by
Fuller et al.:
d.a =.00100 t1-75»(i/mx)+(i/ma) * [P(ivx)y3+(ivA)1/3]2
where T = temperature, K
P = pressure
MV,M. = molecular weights of X and air, respectively
A A
IVv, XV. = the sum of partial molecular volume terms for species
A A
X and A, respectively
The partial molecular volume terms contributing to ZV^ for different
species are tabulated in the literature.
(3) If tables for IVx are not available, or cannot be used for a
particular species x, the ratio of may approximated
using the same correlation given above, assuming the partial molar
volume terms to be equal:
DXA/DWA = [(1/Mx + 1/MA} * (1/Mw + 1/MA)]i
This ratio is calculated by the computer as a default value for the
diffusivity ratio if a value for D^A is not supplied.
ICF Incorporated
-------
A4-11
Estimation of Driving Force for Evaporation
This terra represents a measure of the driving force for the evaporation of
X. It may be approximated using Raoult's law of partial pressures:
X = X • P° /P = X P
XO X X XX
where X^ = mass fraction of X in the liquid
P°x = equilibrium vapor pressure of pure X, in mmHg
P = total pressure, 760 mmHg
P^ = normalized equilibrium vapor pressure of pure X
(dimensionless)
therefore,
X /(1-X ) = X P /(1-X P )
XO XO XX XX
In summary, the evaporative loss of X to air is expressed as:
W = (8.41 (10-3) kg/m2s) • k /k • X P /(1-X P ) • A
X a X W X X X X
where A is the liquid surface area exposed to the air
X^ = the mass fraction of X in the liquid (kg/kg sludge)
P^ = the ratio of the vapor pressure of X to atmospheric pressure
(k /k ) = the ratio of mass transfer coefficients for X and water,
x w '
determined by one of the following methods (listed in order of
accuracy):
(a) using a value for the diffusivity of X in air at
25°C, either from literature values or empirical
correlations. In this case,
Vkx " "WW2'' = I«xA/<'220cra>2/3l2/3
(b) from the molecular weights of X, water, and air, using a
relationship derived from an empirical correlation:
kx/kw = [(1/MX + 1/MA)/(1/Mw +
ICF Incorporated
-------
A4-12
Application of Evaporation Rate Equation on Open Vessels and Sludge Spills
The area of liquid surface exposed to air can be determined for each open
process vessel using the relationship between tank volume and retention time:
V = A • H = S(l) • C/p
where
V = tank volume, m3
A = area of liquid surface exposed to the air, m2
H = tank height, m
SC1) = influent sludge flow, kg/day
p = sludge density, kg/m3
C = retention time in process vessel, day ^
The treatment technologies which require evaporative loss determinations
are the same as those listed in Exhibit A4-1. The characteristic dimensions
and retention times for each piece of equipment and the evaporative surface
area of each open vessel are listed in the separate discussion of each
technology in Appendix 1.
The rate of evaporation of X from an open vessel in a given treatment
technology for a particular sludge flow rate is thus calculated by the
equation:
W = (8.41 CIO-3) kg/m2 s) • k /k • X P /(1-X P ) • A
X X W XX XX
The rate of evaporation of X from a sludge spill may be estimated using
the same equation, assuming that the spill area is approximately equal to the
surface area of the vessel.
Certain qualifications should be noted regarding the application of the
above equation.
(1) The equation gives the rate of evaporation of X from a particular
sludge stream assuming that the concentration of X in the sludge does
not change with time. For a component X of high volatility or sludge
vessels with relatively long retention times, the concentration of X
could drop and the rate of evaporation could decrease significantly.
The prediction given by the equation for W should be considered a
maximum possible rate of evaporation for X.
(2) Because it is assumed that the evaporation rate does not slow down as
the concentration of X decreases, it is possible for the model to
predict a loss of X greater than the amount of X in the influent
sludge to a treatment module. The computer program therefore
includes default values to restrict the possible emission of X due to
evaporation.
ICF Incorporated
-------
A4-13
Chemical interactions between components can alter the evaporation
rates significantly, usually by decreasing the volatility of all
components. The assumption that the hazardous component X acts
independently of the other sludge components should predict
evaporation rates higher than those actually observed.
ICF Incorporated
-------
APPENDIX 5
DISTRIBUTION OF RELEASES OF X
AIR, SURFACE WATER, AND GROUND WATER
ICF INCORPORATED
-------
APPENDIX 5
DISTRIBUTION OF RELEASES OF X
TO AIR, SURFACE WATER, AND GROUND WATER
Uncontrolled emissions of hazardous component X to the environment can occur
during industrial waste treatment. For the purposes of this analysis, the
global environment was divided into three subenvironments:
(1) air
(2) ground water (wells, aquifers)
(3) surface water (streams, rivers)
Releases of X from the treatment process will enter any or all of these three
environments, depending upon physical properties of X and on climatic,
physiographic, and geomorphic characteristics of the area where the spill
occurs. Portions of X entering each environment are determined in subroutine
ACC0UNT. Values for emissions of X to air, ground water, and surface water,
as well as the total release of X, are tabulated in the LOSS vector:
L0SS(1) = loss to air environment
LOSS(2) = loss to ground water environment
L0SS(3) = loss to surface water environment
LOSS(4) = total loss to global environment
In the remainder of this section, subroutine ACCOUNT is explained and an
example of its operation given.
Determination of Loss of X to the Air Environment
The total emission of X to air is the sum of three terms:
(1) evaporative loss from open tank processes
(2) mechanical aeration loss from open tank processes
(3) evaporative loss from routine and accidental sludge spills from
all processes.
The variables in ACC0UNT that apply to these calculations are:
WX = Rate of evaporation of X from open tank
(WX=0 for closed vessels)
NVPX = Normalized vapor pressure of X, the ratio of the vapor
pressure at X at 25° to barometric pressure (760 mmHg)
DXA = Diffusivity of X in air at 25°C, cm2/sec
ICF INCORPORATED
-------
A5-2
SPILL = Amount of X spilled from the treatment process
WXS = Amount of X spilled that evaporates
RMULT = RCODE, release rate code for each technology (see
Exhibit A4-4) for values of RCODE)
AS = Liquid surface area exposed to air for evaporation
KRAT = Ratio of the mass transfer coefficient of X in air to that of
water in air, both at 25°C (dimensionless)
The loss of X from open tanks due to evaporation is determined according to
the equation derived in Section 2 of this appendix:
WX=S(1) • A • (8.41x10-3) • KRAT • S(7) • NVPX/(1-S(7) • NVPX)
The upper limit for WX is the total influent flow of X to the treatment system.
The value of WX is zero for closed tank processes.
The loss of X, due to mechanical aeration of sludge is determined as a
function of THETA1 and RCODE estimated for each treatment technology (see
Section 1 of this Appendix for details).
aeration loss = THETA • RMULT • (S(IO)-WX)
The third contribution to the total emission of X to the air is that portion
of X which evaporates from the sludge lost to the environment through routine
and accidental spills. This is determined using the evaporation rate equation
derived in Section 2 of this appendix, applied to the combined routine and
accidental spills:
SPILL = (1-THETA) • RMULT • (S(IO)-WX)
WXS = SPILL • A • (8.41x10-3) • KRAT • NVPX
• (S(10)-WX)/(1-NVPX • (S(IO)-WX))
The upper limit for WXS is the total X contained in routine and accidental
spills. Therefore, the total air loss is the sum of these three terms:
L0SS(1) = WX + THETA • RMULT • (S(IO)-WX) + WXS
Division of Remaining Spilled X into Liquid and Solid Portions
The amount of released X left after possible evaporation from the combined
routine and accidental spills is determined from variables previously calculated:
SPILL + (1-THETA) • RMULT • (S(10)-WX)-WXS
1Theta = fraction of overall loss due to aeration = aeration loss fraction
divided by the overall loss fraction.
ICF INCORPORATED
-------
A5-3
This remaining release of X may enter the ground water and/or surface water
environments, depending on physical properties of X and on climatic,
physiographic, and geomorphic characteristics of the area where spillage occurs.
In order to estimate the amount of X which ultimately reaches each
environment, the total spill of X to the land is split into liquid and solid
phase releases according to the corresponding mass fractions in the influent
sludge stream to the treatment unit:
SPILXL = liquid X spilled = SPILL • S(8)
SPILXS = solid X spilled = SPILL-SPILXL
The movements of the liquid and solid phase portions of the spilled X are
modeled separately.
Fate of Liquid Portion of Spilled X
The liquid portion of X spilled (SPILXL) can either run off or infiltrate
the ground cover. In order to determine the proportions of SPILXL which reach
pervious and impervious surfaces, the characteristics of the spill area are
estimated using a run-off coefficient of 0.90, typical of industrial areas.
The portion of liquid X that infiltrates is computed as:
LIQUID LOSS (2) = SOAK (Ground-water Release) = (1.0 - 0.9) • (SPILXL)
It is assumed that the amount of SPILXL which does not infiltrate the soil
either runoffs the ground directly to surface water or is later washed there as
stormwater runoff. The loss to the surface water environment is:
LOSS (3) = Surface-water Release = SPILXL-SOAK.
Fate of Solid Portion of Spilled X
The solid portion of X spilled (SPILXS) will be available to erode to
surface waters.
We assume that 10% of the solid X spilled is eroded and washed by stormwater
to the surface water environment. None of the solid X spilled will percolate
down to the ground water environment. Thus, the loss to surface waters
(L0SS(3)) is increased as follows:
LOSS (3) = LOSS (3) +0.10 • SPILXS
Total Emissions of X to All Environments
The emissions of X to the three major environments are the sums of the
vapor, liquid, and solids portions discussed previously.
The total emission of X to the global environment is the sum of the four losses
above:
L0SS(4) = L0SS(1) + LOSS(2) + L0SS(3)
ICF INCORPORATED
-------
APPENDIX 6
WASTE STREAM DESCRIPTIONS
ICF Incorporated
-------
APPENDIX 6
WASTE STREAM DESCRIPTIONS
This appendix presents the detailed descriptions of the 83 waste streams that
we defined for the model. The descriptions list all the input parameters
necessary to define a waste stream. The reference sources that we used to develop
the information are indicated by number under the appropriate heading and appear
in the Reference chapter. The letters E and C indicate estimated and calculated,
respectively.
The data appear in a standard format for each waste stream. The following
should be noted:
a. EPA # refers to EPA Waste Identification Number (40 CFR 261), if any.
b. Sources lists two sets of information: reference data for waste
generation rate followed by reference data for waste characterization. Numbers
correspond to numbered.entries in the Reference chapter.
c. Percent Non-water refers to the fraction of the waste which is non-water,
i.e., .20 = 20 percent.
d. Percent Dissolved refers to the fraction of waste which is in the solid
phase, i.e., .20 = 20 percent suspended solids; 80 percent liquid.
ICF Incorporated
-------
waste stream
£ name
MERCURY SLUDGES FROM CHLORALKALI PROCESS
tota I
we i ght
42770.
const i tutent
of concern
MERCURY
percent
nonwater
.260
percent specific net BTU
d i sso I ved g rav i t.v content
.060
concen-
t ra t i on
.002727
2.5
percent
d i ssoIved
.25
waste
type
01
molecular
we iaht
201 .
applicable treatment technologies:
evaporation/drying and leaching
evaporation/drying and vacuum filter
I each i ng
no treatment
waste stream
£ name
2 LEAD SLUDGES FROM BATTERY PRODUCTION
tota I
we i ght
2000.0
const i tutent
of concern
percent
nonwater
. 110
percent specific net BTU
d i ssoIved g rav i ty content
. 100
concen-
tration
3.5
percent
d i ssoIved
waste
type
molecular
weight
EPA
JL.
K071, K106
vapor
pressure
.0018
so Iub i I i ty
Q 25 C.
20.
EPA
JL-
N/A
vapor
pressure
so Iub iIi ty
@ 25 C.
SIC
No.
2812
SIC
No.
3691
Sou rces
19 2,6
Sou rces
27 2,11
LEAD
0.035
.01
01
207.
150.
applicable treatment technologies:
chemical fixation/stabiIization
evapo ra t i on/d ry i ng
evaporation/drying and leaching
I each i ng
no treatment
vacuum f iIter
-------
waste stream
& name
CHROMIUM SLUDGES FROM PIGMENT PRODUCTION
tota I
we i qht
6724.0
percent
nonwater
. 100
percent specific net BTU
d i ssolved qravi t.v content
const i tutent
of concern
CHROMIUM (VI)
.09
concen-
tration
.00018
2.0
pe rcent
d i ssoIved
.02
waste
t.vpe
02
mo I ecu I a r
we iqht
52.
applicable treatment technologies:
chemical destruction and vacuum filter
chem i caI f i xa t i on/stab iIi za t i on
evaporation/drying and chemical fixation/stabilization
filter press
no treatment
vacuum f iIter
waste stream
U name
ZINC SLUDGES FROM TEXTILES INDUSTRY
tota I
we i qht
17900.
const i tutent
of concern
ZINC
pe rcent
nonwater
.200
percent specific
d i ssoIved q rav i tv
net BTU
content
.200
concen-
trat ion
.07
2.0
percent
d i ssoIved
00
waste
type
03
molecular
we i qht
65.
applicable treatment technologies:
chemical fixation/stabilization
f iIter press
filter press and chemical fixation/stabilization
no treatment
vacuum f iIter
vacuum filter and chemical fixation/stabilization
EPA
JL
K002, K008
vapor
pressure
0.
so I ub iIi ty
Q 25 C.
50.
EPA
JL
N/A
vapor
pressure
so Iub iIi ty
Q 25 C.
2000.
SIC
No. Sources
2812
SIC
No.
2823
Sou rces
18 2,18
-------
waste stream
# name
EPA SIC
H No. Sou rces
ARSENIC SLUDGE FROM PROD. OF VETERINARY PHARMACEUTICALS
K08U, K101, K102
2833
17
tota I
we i ght
500.00
const i tutent
or concern
ARSENIC
percent
nonwater
.300
pe rcent
d i ssoIved
.290
concen-
trat i on
.08
spec i f i c
qravi tv
2.0
net BTU
content
percent
d i ssoIved
.001
waste
type
01
mo I ecu I a r
we i g h t
75.
applicable treatment technologies:
chemical fixation/stabilization
no treatment
waste stream
name
6 MIXED METAL SLUDGES FROM PAINT PRODUCTION
tota I
we i ght
55.000
const i tutent
of concern
THALLIUM
LEAD
MERCURY
percent
nonwater
. 110
percent
d i ssoIved
. 100
concen-
t rat i on
.00097
0.011
.015
spec i f ic
qravi t .v
2.5
net BTU
content
percent
d i ssoIved
.95
.01
.01
waste
type
01
01
01
applicable treatment technologies:
chem i caI f i xa t i on/stab iIi za ti on
chemical precipitation and carbon adsorption (PAC)
I each ing
no treatment
steam stripping and leaching
vapor
pressure
mo I ecu I a r
weight
20b.
207.
201.
so Iub iIi ty
Q 25 C.
500000.
EPA
JL
N/A
vapor
pressure
0.
0.
.0018
so Iub iIi ty
@ 25 C.
250000.
150.
20.
SIC
No. Sources
2851
27 6,24
-------
waste stream
£ name
MIXED METAL SLUDGES FROM INK FORMULATION
tota I
we i qht
60.000
const i tutent
of concern
CHROMIUM (VI)
LEAD
percent
nonwater
.07
percent
di ssolved
.06
concen-
trat ion
.00015
.00076
spec i f i c
qravi t .v
2.0
net BTU
content
percent
d i ssoIved
.01
.01
waste
type
01
01
molecular
we i g h t
52.
207.
vapor
pressure
0.
0.
applicable treatment technologies:
chemical destruction and chemical fixation/stabilization
chemical destruction and chemical precipitation and vacuum filter
chem i caI f i xat ion/stab iIi za t ion
no treatment
waste stream
B. name
8 MIXED METAL SLUDGES FROM NON-FERROUS METAL PROD.
tota I
we i qht
2600.0
const i tutent
of concern
CADMIUM
LEAD
percent
nonwater
.200
percent
d i ssoIved
. 180
concen-
tration
.000134
.0255
spec i f i c
qravi tv
6.0
net BTU
content
pe rcent
d i ssoIved
.20
00
waste
type
01
01
mo I ecu I a r
we ight
112.
207.
applicable treatment technologies:
chem i caI f i xa t ion/stab iIi za t ion
evaporat i on/d ry i ng
I each i ng
no treatment
vacuum f iIter
vacuum filter and chemical fixation/stabilization
vapor
pressure
0.
0.
EPA SIC
U No. Sources
K086 2893 27 7
so I ub iI i ty
Q 25 C.
50.
150.
EPA SIC
U No. Sources
K065, K066, K067 3332, 3333 27 2,6
so Iub iIity
@ 25 C.
2.6
150.
-------
waste stream
# name
EPA
SIC
No. Sources
MIXED METAL SLUDGES FROM ELECTROPLATING
F006, F019
3471
1,6
2C
tota I
we i qht
110.00
const i tutent
oT concern
De rcent
nonwa te r
.02
CHROMIUM
CYANIDE
(VI )
percent specific
d i sso I ved q ra v i t.v
.001
concen-
trat ion
.00037
.0000429
2.5
percent
d i ssoIved
.20
.95
net BTU
content
waste
t.YPe
02
02
mo I ecu I a r
we i qht
52.
27.
vapor
pressure
0.
0.
so I lib i I
Q 25
«ty
C.
50.
480000.
applicable treatment technologies:
air stripping and chemical coagulation
air stripping and chemical coagulation and chemical fixation/stabilization
chemical destruction and chemical coagulation
chemical destruction and chemical coagulation and chemical fixation/stabilization
no treatment
waste stream
£ name
10 BIO SLUDGE CONTAINING HEAVY METALS
tota I
we i qht
3452.0
const i tutent
of concern
CHROMIUM (VI)
percent
nonwater
.130
percent specific net BTU
d i sso I ved g rav i t.v content
.110
concen-
t ra t i on
.000006
2.5
percent
d i ssoIved
00
9000.0
waste
type
02
52.
applicable treatment technologies:
eva po ra t i on/d ry i ng
filter press
filter press and incineration at all four DRE's
no treatment
vacuum f iIter
vacuum filter and incineration at all four DRE's
EPA
N/A
mo I ecu I a r
we i ght
vapor
pressure
0.
so Iub iIi ty
Q 25 C.
50.
SIC
No.
2911
Sources
15 15
-------
waste stream
name
11 COOLING WATER, SLUDGES FROM PETROLEUM INDUSTRY
tota I
we i g h t
9260.0
percent
nonwate r
.250
percent specific net BTU
d i ssoIved gravity content
const i tutent
of concern
CHROMIUM (VI)
LEAD
.200
concen-
t rat ion
.0000013
.000057
2.5
percent
d i ssoIved
.02
.02
wa ste
type
01
01
mo I ecu I a r
we i ght
52.
207.
applicable treatment technologies:
chem i caI f i xa t ion/stab iIi zat i on
no treatment
EPA
JL.
N/A
vapor
pressure
0.
0.
so Iub iIi ty
Q 25 C.
50.
150.
SIC
No.
2911
Sources
15 15
waste stream
£ name
12 SPENT LIME BOILER FEED FROM PETROLEUM REFINING
tota I
weight
38900.
percent
nonwater
.410
percent specific net BTU
d i sso I ved gravi t.v content
const i tutent
of concern
CHROMIUM (VI )
LEAD
.410
concen-
t rat ion
.00000004
.0000038
1.5
percent
d i ssoIved
.02
.98
waste
type
01
01
mo I ecu I a r
we i ght
52.
207.
applicable treatment technologies:
chem i caI f i xa t ion/stab iIi za t i on
evapora t i on/d ry i ng
no treatment
EPA
N/A
vapor
pressure
0.
0.
so I ub iIity
@ 25 C.
50.
150.
SIC
No.
2911
Sources
15 15
-------
waste stream
£ name
13 METAL SLUDGES - HIGH TOXICITY, N.O.S.
tota I
we i ght
3100.0
const i tutent
of concern
LEAD
pe rcent
nonwa te r
.200
percent specific net BTU
d i sso I ved g ray i t.v content
.180
concen-
trat ion
.000031
2.5
pe rcent
d i ssoIved
.50
waste molecular
type we i ght
01
207.
applicable treatment technologies:
chem i caI f i xat i on/stab iIi za t i on
evapora t i on/dry i ng
evaporation/drying and chemical fixation/stabilization
f i I ter press
filter press and chemical fixation/stabilization
no treatment
waste stream
# name
14 METAL SLUDGES - HIGH TOXICITY, N.O.S.
tota I
we i g h t
310.00
const i tutent
of concern
LEAD
pe rcent
nonwa te r
.200
percent specific
d i sso I ved g ra v i t.v
net BTU
content
.180
concen-
t ra t i on
2.5
percent
d i ssoIved
waste molecular
type we i gh t
01
207.
.000031 .50
applicable treatment technologies:
cliein ical fixation/stabilization
evapo ra t i on/d ry i ng
evaporation/drying and chemical fixation/stabilization
fiIter press
filter press and chemical fixation/stabiIization
no treatment
EPA SIC
H No. Sources
N/A * EE
vapor solubi Ii ty
pressure Q 25 C.
0. 150.
EPA SIC
ti No. Sou rces
N/A » EE
vapor so Iub iIi ty
pressure Q 25 C.
0.
150.
-------
waste stream
ft name
15 METAL SLUDGES - HIGH TOXICITY, N.O.S.
tota I
we i ght
310.00
const i tutent
of concern
LEAD
percent
nonwater
.200
percent specific net BTU
d i ssoIved gravi ty content
. 180
concen-
trat i on
.00031
2.5
percent
d i ssoIved
. 10
waste molecular
type . weight
01
207.
applicable treatment technologies:
chemical fixation/stabilization
evapora t i on/d ry i ng
evaporation/drying and chemical fixation/stabilization
fiIter press
filter press and chemical fixation/stabilization
no treatment
waste stream
name
16 METAL SLUDGES - HIGH TOXICITY, N.O.S.
tota I
we i ght
3100.0
const i tutent
of concern
LEAD
percent
nonwater
.200
percent specific
d i ssoIved g rav i tv
net BTU
content
. 180
concen-
trat ion
.00031
2.5
pe rcent
d i ssoIved
. 10
waste molecular
t.vpe we i ght
01
207.
applicable treatment technologies:
chemical fixation/stabilization
eva po ra t i on/d ry i ng
evaporation/drying and chemical fixation/stabilization
fiIter press
filter press and chemical fixation/stabilization
no treatment
EPA SIC
U No. Sources
N/A * E E
vapo r so I ub iIi ty
pressure Q 25 C.
0. 150.
EPA SIC
M No. Sources
N/A * E E
vapor
pressure
0.
so Iub iIi ty
Q 25 C.
150.
-------
waste stream
# name
EPA
JL.
17 METALS AND MISC. INORGANICS - LOW TO MODERATE TOXICITY, N.O.S.
N/A
tota I
we i g h t
pe rcent
nonwa ter
percent specific
d i sso I ved g ra v i t.v
net BTU
content
3100.0
.200
. 180
2.5
constitutent concen- percent waste molecular vapor solubility
of concern trat ion d i ssoIved type we iqht pressure ® 25 C.
ZINC .000031 .50 03 65. 0. 2000.
applicable treatment technologies:
chemical fixation/stabilization
evaporat ion/drying
evaporation/drying and chemical fixation/stabilization
filter press
filter press and chemical fixation/stabi I ization
no treatment
waste stream EPA
I name #
18 METALS AND MISC. INORGANICS - LOW TO MODERATE TOXICITY, N.O.S. N/A
total percent percent specific net BTU
we i ght nonwa te r d i sso I ved g rav i t.v content
310.00 .200 .180 2.5 0
constitutent concen- percent waste molecular vapor solubility
of concern trat ion d i ssoIved type we iqht pressure ® 25 C.
ZINC .000031 .50 03 65. 0. 2000.
applicable treatment technologies:
chemical fixation/stabilization
e va po ra t i on/d ry i ng
evaporation/drying and chemical fixation/stabilization
f i I ter press
filter press and chemical fixation/stabiIization
no treatment
-------
waste stream
name
19 METALS AND MISC. INORGANICS - LOW TO MODERATE TOXICITY, N.O.S.
tota I
we i qht
310.00
const i tutent
of concern
ZINC
pe rcent
nonwater
.200
percent specific net BTU
d i ssoIved gravi tv content
. 180
concen-
trat ion
.00031
2.5
percent
d i ssoIved
. 10
wa ste
type
03
mo I ecu I a r
we i oht
65.
vapor
pressure
0.
applicable treatment technologies:
chem i caI f i xa t i on/stab iIi za t i on
eva po ra t i on/d ry i ng
evaporation/drying and chemical fixation/stabilization
fiIter press
filter press and chemical fixation/stabilization
no treatment
waste stream
name
20
METALS AND MISC. INORGANICS - LOW TO MODERATE TOXICITY, N.O.S.
tota I
we iqht
3100.0
const i tutent
of concern
ZINC
percent
nonwater
.200
percent specific net BTU
d i ssoIved q rav i tv content
. 180
concen-
t ra t i on
.00031
2.5
percent
d i ssoIved
10
waste
t.vpe
03
molecular
we ight
65.
vapor
pressure
0.
applicable treatment technologies:
chemical fixation/stabilization
evapo ra t i on/d ry i ng
evaporation/drying and chemical fixation/stabilization
fiIter press
filter press and cher»ica! f! xa t • on/stab i I i za t i on
no treatment
EPA SIC
H No. Sources
N/A * EE
so Iub iIi ty
Q 25 C.
2000.
EPA SIC
tt No. Sources
N/A * EE
so Iub I I i ty
Q 25 C.
2000.
-------
waste stream
£ name
EPA
JL.
SIC
No. Sources
21
PICKLE LIQUOR FROM STEEL FINISHING
K062
3312
2,6
tota I
we i ght
8100.0
pe rcent
nonwater
.200
const i tutent
of concern
CHROMIUM (VI)
LEAD
percent specific net BTU
d i sso I ved gravi t.v content
.02
concen-
trat ion
.000867
.000098
1.3
pe rcent
d i ssoIved
1.0
1.0
waste molecular
type we ight
02
02
52.
207.
vapo r
pressure
0.
0.
so Iub iIi ty
@ 25 C.
50.
150.
applicable treatment technologies:
chemical precipitation
chem i ca I
chem i ca I
no treatment
p rec i p i ta t i on
precipitation and chemical coagulation and chemical fixation/stabilization
precipitation and chemical fixation/stabilization
waste stream
& name
22 ACID SOLUTIONS CONTAINING HEAVY METALS N.O.S.
EPA
N/A
SIC
No.
Sou rces
E E
tota I
we i ght
3100.0
const i tutent
of concern
LEAD
percent
nonwa te r
.100
percent specific net BTU
d i ssoIved g rav i tv content
.01
concen-
t ra t i on
.000031
2.5
N
pe rcent
d i ssoIved
1.0
waste molecular
type we ight
vapor
pressure
01
207.
so Iub iIi ty
@ 25 C.
150.
applicable treatment technologies:
chemical precipitation
chemical precipitation and chemical
chemical precipitation and chemical fixation/stabilization
chemical precipitation and chemical coagulation and chemical fixation/stabilization
no treatment
-------
waste stream EPA SIC
£ name # No. Sources
23 LIME SLUDGE FROM COKING OPERATIONS K060 3312 6 6
tota I
weight
67650.
const i tutent
of concern
ARSENIC
PHENOL
CYAN IDE
percent
nonwater
.05
percent
d i ssoIved
.04
concen-
trat ion
.0011
.001
¦ 0074
spec i f i c
qravi tv
2.2
percent
d i ssoIved
.02
.98
.98
net BTU
content
waste
type
12
12
12
mo I ecu la r
weight
75.
94.
27.
vapor
pressure
.8
0.
so I ub iIi ty
6 25 C.
500000.
84000.
480000.
applicable treatment technologies:
chemical destruction and vacuum filter and chemical fixation/stabilization
chemicaI fixation/stabiIization
no treatment
waste stream
& name
24 HEAT TREATMENT WASTES
tota I
we ight
percent
nonwater
percent
d i ssoIved
spec i fic
gravi tv
net BTU
content
EPA SIC
U No.
F010, F011, F012 3398
Sources
170.00
.200
. 100
1.5
2200.0
const i tutent
of concern
concen-
trat ion
percent
d i ssoIved
waste
type
mo I ecu la r
weight
vapo r
pressure
so I ub iIi ty
® 25 C.
CYAN IDE
.05
.98
12
27
480000.
applicable treatment technologies:
chemical destruction
chem i caI f ixati on/stab iIi zat i on
incineration at all four DRE's
no treatment
-------
waste stream
£ name
EPA SIC
U No. Sources
25 CYANIDE WASTE FROM ELECTROPLATING
F007, F008, F009
3471
6C
6E
tota I
we i ght
1750.0
const i tutent
of concern
CYANIDE
pe rcent
nonwater
.200
percent specific net BTU
d i ssoIved qravi tv content
. 100
concen-
t ra t i on
.06
1.5
percent
d i ssoIved
.98
5300.0
waste molecular
type we i ght
vapo r
p ressu re
12
27.
so Iub iIi ty
@ 25 C.
480000.
applicable treatment technologies:
chemical fixation/stabiIization
chemical destruction
no treatment
waste stream
# name
26 ASBESTOS SEPARATOR WASTES FROM THE DIAPHRAGM CELL IN CHLORINE PROD. N/A
EPA
tota I
we i aht
2200.0
const i tutent
of concern
ASBESTOS
pe rcent
nonwater
.700
percent specific net BTU
d i ssoIved q rav i tv content
1.00
concen-
t ra t i on
.084
2.5
percent
d i ssoIved
00
waste molecular
t.vpe we i ght
14
60.
vapor
pressure
0.
so I ub i I i ty
Q 25 C.
0.
SIC
No. Sources
2812
29
25
applicable treatment technologies:
chem i caI f i xa t i on/stab iIi za t i on
no treatment
-------
waste stream
# name
27
PHENOLIC SLUDGE FROM PLASTICS PRODUCTION
tota I
we i ght
1620.0
percent
nonwater
. 100
percent
dIssoIved
const i tutent
of concern
PHENOL
FORMALDEHYDE
.02
concen-
tration
.05
.02
spec i f i c
gravi tv
1.1
net BTU
content
2000.0
percent
d i ssoIved
.98
.98
waste
type
11
11
applicable treatment technologies:
incineration at all four DRE's
no treatment
solvent extraction
steam stripping and evaporation/drying
waste stream
name
28 PHENOL-FORMALDEHYOE SOLUTION FROM PLASTICS PRODUCTION
EPA
JL.
N/A
mo I ecu I a r
we i ght
94.
30.
vapor
pressure
.8
3880.
so Iub i I i ty
Q 25 C.
84000.
300000.
EPA
JL-
N/A
SIC
No. Sources
2821 18,27 2,18
x
x
SIC
NO.
2821
Sources
E 2,18
tota I
we i ght
180.00
const i tutent
of concern
percent
nonwater
.250
percent specific
d i sso I ved q rav i t.v
.200
concen-
trat ion
1.5
pe rcent
d i ssoIved
net BTU
content
2000.0
waste molecular vapor solubility
type we ight pressure Q 25 C.
PHENOL
.05
.98
11
94.
.8
84000.
applicable treatment technologies:
chemical rixation/stabiIization and solvent extraction
evapo rat i on/d ry i ng
evaporation/drying and solvent extraction
incineration at all four DRE's
no treatment
-------
waste stream
£ name
29 WASTEWATER TREATMENT SLUDGE FROM WOOD PRESERVING
tota I
we i ght
260.00
percent
nonwa te r
.100
pe rcent
d i ssoIved
. 100
spec i f ic
q ra v i ty
1.5
net BTU
content
25000.
const i tutent
of concern
concen-
t ra t i on
percent
d i ssoIved
waste
type
molecular
we i ght
PENTACHLOROPHENOL
BENZO(A)PYRENE
CHRYSENE
.0000302
.000000031
.00000045
.10
.01
.01
04
04
OU
266.
252.
228.
applicable treatment technologies:
evaporation/drying and leaching
incineration at all four DRE's
I each i ng
no treatment
steam stripping
waste stream
H name
30 DISTILLATION RESIDUES FROM ACETALDEHYDE PRODUCTION
tota I
pe rcent
percent
spec i f i c
net
BTU
we i ght
nonwater
d i ssoIved
q ra v i t.v
content
538800
.06
.001
1.0
0
const i tutent
concen-
pe rcent
waste
mo 1 ecu 1,
of concern
tration dissolved
type
we i ght
ACETALDEHYDE
.0096
.99
04
44.
CHLOROACETALDEHYDE
.0066
.99
04
79.
CHLOROFORM
.0006
.99
04
119.
FORMALDEHYDE
.0006
.99
04
30.
applicable treatment technologies:
chemical destruction
evapo ra t i on/d ry i ng
incineration at all four DRE's
no treatment
EPA
K001
SIC
No. Sou rces
2491 27 27E
vapor solubiIi ty
pressure Q 25 C.
.0005 20.
0. .004
0. .004
EPA SIC
H No. Sou rces
K009, K010 2869 6E 6
vapor
pressure
36.
100.
150.5
3880.
so Iub iIi ty
Q 25 C.
9999999.
400000.
8000.
300000.
-------
waste stream
name
31 STILL BOTTOMS FROM ACRYLONITRILE PRODUCTION
tota I
we i ght
145000.
const i tutent
or concern
ACETONITRILE
CYAN IDE
ACRYLONITRILE
percent
nonwa te r
.100
pe rcent
d i ssoIved
.01
concen-
t ra t i on
.006
.003
.0004
spec i Tic
q rav i tv
1.2
net BTU
content
percent
d i ssoIved
.99
.98
.95
waste molecular vapor
type weight pressure
09
09
09
41.
27.
53.
100.
0.
120.
applicable treatment technologies:
chemical destruction
electrolytic decomposition and air stripping and reverse osmosis
incineration at all four DRE's
no treatment
waste stream
ff name
32 AQUEOUS SPENT ANTIMONY CATALYST FROM FLUOROMETHANE PRODUCTION
tota I
we i ght
27.000
percent
nonwater
.02
percent
d i ssoIved
const i tutent
of concern
CARBON TETRACHLORID
CHLOROFORM
ANTIMONY
.001
concen-
tration
.004
.004
.99
spec i f ic
g rav itv
1.0
net BTU
content
percent
d i ssoIved
.99
.99
.01
waste
type
04
04
04
mo I ecu I ar
we i ght
154.
119.
122.
vapor
pressure
90.
150.5
0.
applicable treatment technologies:
no treatment
steam stripping and chemical precipitation
steam stripping and chemical precipitation and vacuum filter
EPA SIC
If No. Sources
K011, K013, K014 2869 6,27C 6
so I ub iIi ty
Q 25 C.
9999999.
480000.
80000.
EPA SIC
H No. Sources
K021 2869 2E 2,6C
so Iub iIi ty
Q 25 C.
800.
8000.
4750000.
-------
waste stream
name
33 WASTEWATER FROM NITROBENZENE / ANILINE PRODUCTION
EPA SIC
U No. Sources
K083, K103, K104 2869 7E 7E
tota I
we i ght
51400.
percent
nonwater
.05
const i tutent
of concern
NITROBENZENE
BENZENE
percent specific net BTU
d i ssoIved q rav i ty content
.01
concen-
t ra t i on
.00003
.000075
1. 1
percent
d i ssoIved
.95
.95
waste
type
07
07
molecular
we i oht
123.
78.
vapor
pressure
.26
95.2
so Iub iIity
@ 25 C.
1900.
1800.
applicable treatment technologies:
centrifuge and carbon adsorption (PAC)
incineration at all four DRE's
no treatment
steam stripping
waste stream
name
34 WASTEWATER FROM CHLOROBENZENE PRODUCTION
EPA
K105
SIC
No. Sources
2869
7C
tota I
percent
percent
spec i f i c
net
BTU
we ight
nonwater
d i ssoIved
a rav i ty
content
39000.
.02
.01
1.1
0
const i tutent
concen-
percent
wa ste
mo 1 ecu 1 a r
vapor
so I ub i I i ty
of concern
tration dissolved
tVDe
we i ght
pressure
@ 25 C.
BENZENE
.0018
1.0
07
78.
95.2
1800.
CHLOROBENZENE
.00008
1.0
07
113.
8.8
500000.
1,2-DICHLOROBENZENE
.005
1.0
07
147.
1.8
150.
1,4-DICHLOROBENZENE
.005
1.0
07
147.
0.4
80.
applicable treatment technologies:
filter press and carbon adsorption (PAC)
incineration at all four DRE's
no treatment
-------
waste stream
H name
35
TOXAPHENE WASTE
tota I
we i ght
6850.0
const i tutent
of concern
TOXAPHENE
percent
nonwater
.250
percent specific net BTU
d i sso I ved qrav i t.v content
.22
concen-
trat ion
.01
1.3
percent
d i ssoIved
.02
applicable treatment technologies:
incineration at all four DRE's
no treatment
solvent extraction
9000.0
waste
type
0t
mo I ecu I a r
we i ght
UU».
waste stream
# name
36
PESTICIDE WASTES - HIGH TOXICITY, N.O.S.
tota I
we i ght
percent percent specific net BTU
nonwater d i sso I ved q rav i t.v content
3100.0
1.00
1.0000
1.5
35000.
constitutent concen- percent waste molecular
of concern trat ion d i ssoIved type we ight
PARATHI ON .03 00 09 291.
applicable treatment technologies:
incineration at all four DRE's
no treatment
solvent extraction
EPA SIC
tf No. Sources
K041, K098 2879 6C 6C
vapor so Iub iIi ty
pressure ® 25 C.
.3 3.
EPA SIC
tf No. Sources
N/A 2879 E E
so Iub i I i ty
@ 25 C.
vapor
pressure
.00002 2t».
-------
waste stream
£ name
EPA
JL_
SIC
No. Sou rces
37 PESTICIDE WASTES - MODERATE TOXICITY, N.O.S. N/A 2879
total percent percent specific net BTU
we ight nonwater d i sso I ved aravi t.v content
3100.0 1.00 1.0 1.5 30000.
constitutent concen- percent waste molecular vapor solubility
of concern t ra t ion d i ssoIved type we ight pressure Q 25 C.
CHL0RDANE .03 00 09 410. .00001 1.
applicable treatment technologies:
incineration at all four DRE's
no treatment
solvent extraction
waste stream
it name
EPA SIC
tt No. Sou rces
38 AQUEOUS ORGAN ICS N.O.S.
N/A
tota I
we i g h t
3100.0
pe rcent
nonwater
.06
const i tutent
of concern
1,2-DICHLOROETHANE
percent specific
d i ssoIved gravity
concen-
trat i on
.00031
1.0
percent
d i ssoIved
1.0
applicable treatment techno Iogies:
carbon adsorption (PAC)
chemical destruction
no treatment
net BTU
content
waste molecular
type we i gh t
04
99.
vapor
pressure
61.
so Iub iIi ty
@ 25 C.
8600.
-------
waste stream
£ name
39 AQUEOUS ORGAN ICS N.O.S.
tota I
we i ght
310.00
pe rcent
nonwater
.06
pe rcent
d i ssoIved
spec i f ic
g rav i tv
1.0
const i tutent
of concern -
1,2-DICHL0R0ETHANE
concen-
trat ion
.031
percent
d i ssoIved
1.0
applicable treatment technologies:
incineration at all Tour DRE's
no treatment
steam stripping
waste stream
name
net BTU
content
1000.0
waste
type
Ot
mo I ecu I a r
we i q h t
99.
EPA
N/A
vapor
pressure
61.
so Iub iIi ty
Q 25 C.
8600.
EPA
JL
SIC
No.
Sou rces
E E
SIC
No. Sources
ttO AQUEOUS ORGAN ICS N.O.S.
N/A
total percent percent specific net BTU
we ight nonwater d i sso I ved qravi t.v content
310.00 .06 0 1.0 0
const i tutent
of concern
concen-
tration
percent
d i ssoIved
waste
type
molecular
we ight
vapor
pressure
so Iub iIi ty
@ 25 C.
1,2-DICHLOROETHANE
.00031
1.0
04
99.
61.
8600.
applicable treatment technologies:
caroon adsorption (PAC)
chemical destruction
no treatment
-------
waste stream
& name
41
AQUEOUS ORGAN ICS N.O.S.
total percent percent specific net BTU
we ight nonwater d i ssoIved g rav i tv content
3100.0 .06 0 1.0 0
constitutent concen- percent
of concern t ra t ion d i ssoIved
1,2-DICHLOROETHANE .00031 1.0
applicable treatment technologies:
carbon adsorption (PAC)
chemical destruction
no treatment
waste molecular
type we i ght
04
99.
waste stream
# name
42 TRICHLOROETHYLENE SPENT SOLVENT WASTE
percent
nonwa te r
1.00
tota I
we i ght
.10000
const i tutent
of concern
TRICHLOROETHYLENE
percent
d i ssoIved
.02
concen-
tration
.8
spec i f i c
g ra v i t.v
2.5
percent
dissoIved
1.0
applicable treatment technologies;
d i st iIlat ion
incineration at all four DRE's
no treatment
net BTU
content
38000.
waste
type
04
mo I ecu I a r
weight
131.
EPA SIC
ti No. Sources
N/A * EE
vapor
pressure
61.
so!ub i I i ty
Q 25 C.
8600.
EPA SIC
ti No. Sources
F001, F002 * 6 6C
vapor so Iub iIi ty
pressure Q 25 C.
78.
160.
-------
waste stream
# name
*43 1, 1,1-TR ICHLOROETHANE SPENT SOLVENT WASTE
tota I
we i q h t
. 10000
percent
nonwater
1.00
percent
dissolved
const i tutent
of concern
1,1,1-TRICHL0R0ETHA
.02
concen-
tration
.8
spec i f i c
q rav i tv
2.5
pe rcent
d i ssoIved
1.0
applicable treatment technologies:
d i st iI I a t i on
incineration at all four DRE's
no treatment
waste stream
# name
net BTU
content
39000.
waste
type
04
mo I ecu I a r
we i ght
133.
EPA
F001, F002
vapor
pressure
129.
so Iub iI i ty
@ 25 C.
2000.
EPA
SIC
No. Sources
* 6 6C
SIC
No. Sources
HH METHYLENE CHLORIDE SPENT SOLVENT WASTE F001 * 6 6C
total percent percent specific net BTU
we ight nonwater d i ssoIved gravi tv content
.15000 1.00 .02 2.5 35000.
constitutent concen- percent waste molecular vapor solubility
of concern trat ion d i ssoIved type we i ght pressure Q 25 C.
DICHL0R0METHANE .8 1.0 04 85. 465. 1380000.
applicable treatment technologies:
d i st iI I a t i on
incineration at all four DRE's
no treatment
-------
waste stream
# name
45 TETRACHLOROETHYLENE SPENT SOLVENT WASTE
tota I
we i g h t
10.200
pe rcent
nonwa te r
1.00
const i tutent
of concern
TETRACHLOROETHENE
percent specific
d i sso I ved q rav i tv
.2
concen-
t ra t i on
.6
2.5
percent
d i ssoIved
1.0
net BTU
content
33000.
waste molecular
type weight
04
166.
EPA
F002
vapor
pressure
18.
so Iub iIi ty
@ 25 C.
200.
SIC
No. Sou rces
721
6 6,23C
applicable treatment technologies:
d i st i I I a t i on
incineration at all four DRE's
no treatment
waste stream
name
46
SPENT SOLVENTS N.O.S.
tota I
we i oht
20.000
const i tutent
of concern
TOLUENE
percent
nonwater
1.00
percent specific net BTU
d i ssoIved g rav i tv content
. 1
concen-
t ra t i on
8
2.5
percent
d i ssoIved
30000.
1.0
wa ste
type
08
mo I ecu I a r
we i ght
92.
applicable treatment technologies:
distiI I ation:no treatment
incineration at all four DRE's
EPA
_iSL
F004, F005
vapor
pressure
28.7
so I ub i I i ty
@ 25 C.
500.
SIC
No.
Sources
-------
waste stream
# name
EPA
SIC
No. Sources
47
TRICHLOROETHYLENE STILL BOTTOMS
F001, F002
6C
tota I
we i qht
const i tutent
of concern
percent
nonwa te r
1.00
percent specific
d i ssoIved gravi tv
. 1
concen-
t ra t i on
2.5
percent
d i ssoIved
net BTU
content
19000.
waste molecular vapor solubility
t.vpe we iqht pressure @ 25 C.
TRICHLOROETHYLENE
.2
1.0
04
131.
78.
160.
applicable treatment technologies:
incineration at all four DRE's
no treatment
waste stream
£ name
48
1,1,1-TRICHLOROETHANE ST ILL BOTTOMS
EPA
JL.
F001, F002
SIC
No.
Sources
6 6C
tota I
we i ght
const i tutent
of concern
percent
nonwater
1.00
1,1,1-TRICHLOROETHA
percent specific
d i ssoIved gravi tv
. 1
concen-
tration
.2
2.5
percent
d i ssoIved
1.0
net BTU
content
19000.
waste molecular
type we i ght
04
133.
vapor
pressure
129.
so Iub iIi ty
Q 25 C.
2000.
applicable treatment technologies:
incineration at all four DRE's
no treatment
-------
waste stream
& name
EPA
49
METHYLENE CHLORIDE STILL BOTTOMS
FOOT
total percent percent specific net BTU
we i ght nonwa ter d i ssoIved g rav i tv content
.15000
1.00
2.5
17000.
constitutent concen- percent waste molecular vapor solubility
of concern trat ion d i ssoIved type we ight pressure ® 25 C.
DICHLOROMETHANE .2 1.0 01 85. *465. 1380000.
applicable treatment technologies:
incineration at all four DRE's
no treatment
wa ste
I
st ream
name
50
STILL BOTTOMS N.O.S.
EPA
JL_
F004, F005
tota I
we i ght
20.000
pe rcent
nonwa te r
1.00
pe rcent
d i ssoIved
.!»
spec i f i c
g ra v i tv
2.5
net BTU
content
13000.
const i tutent
of concern
TOLUENE
concen-
t ra t i on
pe rcent
d i ssoIved
1.0
wa ste
type
08
mo I ecu I a r
we i ght
92.
vapo r
p ressu re
28.7
so I ub iIity
® 25 C.
500.
applicable treatment technologies:
incineration at all four DRE's
no treatment
-------
waste stream
& name
EPA
SIC
No. Sources
51
PAINT APPLICATION SLUDGES
N/A
27C
tota I
we i g h t
1.1000
pe rcent
nonwater
.750
percent
d i ssoIved
spec i fIc
gravi tv
7.5
net BTU
content
2U000.
const i tutent
concen-
pe rcent
waste
molecular
vapo r
so 1ub i1i ty
of concern
trat ion
d i sso1ved
type
we iqht
pressure
@ 25 C.
TOLUENE
.01
1.0
08
92.
28.7
500.
CHROMIUM (VI)
.000056
00
08
52.
0.
50.
LEAD
.00001
00
08
207.
0.
150.
MERCURY
.000012
00
08
201.
.0018
20.
METHYL ETHYL KETONE
.01
1.0
08
72.
91.2
226000.
applicable treatment technologies:
chemical fixation/stabilization
incineration at all four DRE's
no treatment
waste stream
name
52 ORGANIC / METALLIC SLUDGES N.O.S.
tota I
we i ght
500.00
percent
nonwater
1.00
percent
d i ssoIved
spec i f i c
gravi tv
2.5
const i tutent
of concern
CHROMIUM (VI)
concen-
t rat i on
.00031
pe rcent
d i ssoIved
.02
net BTU
content
2000.0
waste
type
08
mo I ecu I a r
we i ght
52.
applicable treatment technologies:
chem i caI f i xa t i on/stab iIi zat i on
incineration at all four DRE's
no treatment
steam stripping
steam stripping and chemical fixation/stabilization
EPA
N/A
vapo r
pressure
so Iub iIi ty
Q 25 C.
50.
SIC
No.
Sources
6E
-------
waste stream
$ name
53
STILL BOTTOMS FROM BENZYL CHLORIDE DISTILLATION
tota I
we i ght
7260.0
const i tutent
of concern
TOLUENE
CHLOROBENZENE
pe rcent
nonwater
1.00
percent
d i ssoIved
.05
concen-
tration
.02
.02
spec i f i c
q ra v i tv
1.3
net BTU
content
37000.
pe rcent
d i ssoIved
.80
.80
waste
type
05
05
mo I ecu I a r
we i ght
92.
113.
vapor
pressure
28. 7
8.8
applicable treatment technologies:
incineration at all four DRE's
no treatment
waste stream
name
54 HEAVY ENDS & RESIDUES FROM CARBON TETRACHLORIDE PRODUCTION
tota I
we i ght
877.00
percent
nonwa te r
1.00
pe rcent
d i ssoIved
.05
spec i f i c
gravi tv
2.0
net BTU
content
33000.
EPA
_l2_
K015
so I Lib i I i ty
@ 25 C.
500.
500000.
EPA
JL.
K016 2869
SIC
No.
2869
SIC
No.
Sources
6 6
Sou rces
const i tutent
concen-
pe rcent
waste
mo 1 ecu 1 a r
vapor
so Iub i1ity
of concern
trat ion
d i ssoIved
type
we i aht
pressure
@ 25 C.
TETRACHLOROETHENE
.2
1.0
01
166.
18.
200.
HEXACHLOROBENZENE
.04
1.0
on
285.
.00002
.01
HEXACHLOROBUTADIENE
.on
1.0
on
261 .
. 3
2.
HEXACHLOROETHANE
.2
1.0
04
237.
.4
50.
applicable treatment technologies:
incineration at all four DRE's
no treatment
-------
waste stream
£ name
55 HEAVY ENDS FROM EPICHLOROHYDRIN PRODUCTION
tota I
we i ght
11600.
percent
nonwater
1.00
pe rcent
d i ssoIved
const i tutent
of concern
1,2-DICHL0R0PR0PANE
EPICHLOROHYDRIN
.05
concen-
tration
.1
.02
spec i f ic
qravi t.v
1.5
pe rcent
d i ssoIved
1.0
1.0
applicable treatment technologies:
incineration at all four DRE's
no treatment
net BTU
content
40000.
wa ste
type
04
04
mo I ecu I a r
we i ght
113.
93.
EPA
JL-
K017
vapor
pressure
52.
17.
so Iub iIi ty
@ 25 C.
2700.
640000.
SIC
No. Sou rces
2869 6,27 6
waste
stream
name
56
HEAVY ENDS FROM ETHYL CHLORIDE PRODUCTION
EPA
K018
tota I
percent
percent
spec i f i c
net
BTU
we i qht
nonwa te r
d i sso1ved
aravi tv
content
15980.
1.00
.05
1.5
33000.
const i tutent
concen-
pe rcent
wa ste
mo I ecu I a r
vapor
so 1ub i1ity
of concern
tration dissolved
t.voe
we iaht
pressure
@ 25 C.
TRICHL0R0ETHYLENE
.32
1.0
04
131.
78.
160.
1,2-DICHLOROETHANE
.11
1.0
04
99.
61.
8600.
HEXACHLOROBENZENE
.215
1.0
04
285.
.00002
.01
HEXACHLOROBUTAD1ENE
.215
1.0
04
261.
.3
2.
applicable treatment technologies:
incineration at all four DRE's
no treatment
SIC
No.
2869
Sources
6 6
-------
waste stream
£ name
57
HEAVY ENDS FROM ETHYL DI CHLORIDE & VINYL CHLORIDE PROD.
tota I
we ight
18500.
pe rcent
nonwater
1.00
percent specific net BTU
d i ssoIved q rav i tv content
const i tutent
of concern
1,1,2,2-TETRACHLORO
1,1,1-TRICHLOROETHA
1,2 — DICHLOROETHANE
.05
concen-
trat ion
.08
.22000
.20000
2.0
pe rcent
d i ssoIved
1.0
1.0
1.0
33000.
waste
type
04
04
04
mo I ecu I a r
we i ght
168.
133.
99.
applicable treatment technologies:
incineration at all four DRE's
no treatment
waste stream
# name
58 DISTILLATION RESIDUES FROM PHTHALIC ANHYDRIDE PROD.
tota I
we ight
1810.0
pe rcent
nonwater
1.00
percent specific
d i ssoIved grav i tv
constitutent
of concern
MALE IC ANHYDRIDE
.05
concen-
tration
.019
1.5
pe rcent
d i ssoIved
net BTU
content
31000.
waste molecular
t.vpe we i ght
.90
11
98.
applicable treatment technologies:
incineration at all four DRE's
no treatment
EPA
J-
K019, K020
SIC
No.
2869
vapor
pressure
10.
129.
61.
so Iub iI i ty
@ 25 C.
2900.
2000.
8600.
EPA
SIC
No.
K023, K024, K093, K094 2869
vapor
pressure
.502
so IubI Ii ty
@ 25 C.
1000.
Sources
27 2,6
Sou rces
6C 6C
-------
waste stream
g_ name
59 STILL BOTTOMS FROM NITROBENZENE PRODUCTION
tota I
we i qht
3050.0
percent
nonwater
1.00
percent specific net BTU
d issoIved gravi tv content
const i tutent
of concern
2, t»-D I N I TROTOLUENE
.05
concen-
trat i on
.001
1.5
percent
d i ssoIved
1.0
09
182.
applicable treatment technologies:
incineration at all four DRE's
no treatment
waste stream
name
EPA
K025
40000.
waste molecular
type we ioht
vapor
pressure
.0001
so Iub iIi ty
Q 25 C.
300.
60
STRIPPING STILL TAILS FROM METHYL ETHYL PYRIDINE PRODUCTION
EPA
JL.
K026
SIC
No. Sources
2869
SIC
No.
2869
6C
Sources
6C 6
tota I
we i qht
282.00
const i tutent
of concern
PHENOL
PYRIDINE
percent
nonwater
1.00
percent specific
d i sso I ved gravi tv
.05
concen-
trat ion
.000008
.03
1.5
pe rcent
d i ssoIved
1.0
1.0
net BTU
content
33000.
waste molecular
tvoe we i ght
09
09
94.
79.
vapor
pressure
20.
so Iub iIi ty
Q 25 C.
84000.
9999999.
applicable treatment technologies:
incineration at all four DRE's
no treatment
-------
waste stream
# name
61
ORGANIC RESIDUES FROM TOLUENE DI ISOCYANATE PRODUCTION
tota I
we i qht
2055.0
pe rcent
nonwa te r
.990
percent specific
d i ssoIved q rav i tv
const i tutent
of concern
TOLUENE DIAMINE
TOLUENE DlISOCYANAT
.9
concen-
trat ion
.000005
.00004
3.0
pe rcent
d i ssoIved
1.0
1.0
net BTU
content
31000.
waste molecular
t.vpe weight
09
09
122.
174.
applicable treatment technologies:
incineration at all four DRE's
no treatment
waste stream
& name
62 STILL BOTTOMS FROM 1,1,1-TRICHLOROETHANE PRODUCTION
tota I
we i ght
41600.
pe rcent
nonwate r
1.00
percent specific net BTU
d i sso I ved g rav i t.v content
const i tutent
of concern
1, 1,2,2-TETRACHLORO
.05
concen-
t ra t i on
.27
1.5
percent
d i ssoIved
42000.
waste molecular
type we i ght
1.0
04
168.
applicable treatment technologies:
incineration at all four DRE's
no treatment
EPA SIC
ff No. Sources
K027 2869 3E 6C
vapo r so Iub i I i ty
pressure @ 25 C.
.001 500000.
.03 500.
EPA SIC
ft No. Sou rces
K095 2869 6 6
vapor so Iub iI i ty
pressure ® 25 C.
10.
2900.
-------
waste stream EPA
S_ name H
63 DISTILL. RESIDUES FROM TRICHLOROETHYLENE/PERCHLORETHYLENE PROD. K030
SIC
No.
2869
Sources
tota I
we i qht
4110.0
percent
nonwater
1.00
const i tutent
of concern
1,1,2,2-TETRACHL0R0
HEXACHLOROBENZENE
HEXACHLOROBUTADIENE
percent specific net BTU
d i ssoIved q rav i tv content
.05
concen-
t ra t i on
.23
.2
.338
1.5
percent
d i ssoIved
1.0
1.0
1.0
39000.
waste
t.vpe
04
04
04
mo I ecu I a r
we i ght
168.
285.
261.
vapor
pressure
10.
so Iub iIi ty
Q 25 C.
.00002
3
2900.
.01
applicable treatment technologies:
incineration at all four DRE's
no treatment
waste stream
& name
64 STILL BOTTOMS FROM CHLOROBENZENE PRODUCTION
EPA
K085
SIC
No.
2869
Sources
7 2,7
tota I
we i q h t
2740.0
percent
nonwater
1.00
const i tutent
of concern
HEXACHLOROBENZENE
percent
d i ssoIved
.05
concen-
t rat ion
. 1
spec i f i c
q rav i tv
1.5
percent
d i ssoIved
1.0
net BTU
content
35000.
waste
type
04
molecular
we Ight
285.
vapo r
pressure
.00002
so Iub iIi ty
@ 25 C.
.01
applicable treatment technologies;
incineration at all four DRE's
no treatment
-------
waste stream
# name
65 DISTILLATION BOTTOM TARS FROM PHENOL / ACETONE PRODUCTION
tota I
percent
pe rcent
spec i f i c
net
BTU
we i q h t
nonwater
d i sso1ved
g rav i tv
content
18080.
o
o
. 1
1.5
37000.
const i tutent
concen-
percent
waste
mo 1 ecu 1;
of concern
trat ion d
i sso1ved
type
we i qht
PHENOL
.016
1.0
06
94.
BENZ0(A)ANTHRACENE
.001
1.0
06
228.
BENZ0(A)PYRENE
.001
1.0
06
252.
CHRYSENE
.001
1.0
06
228.
applicable treatment technologies:
incineration at all four DRE's
no treatment
waste stream
£ name
66
CHLORINATED HYDROCARBON WASTE FROM CHLORALKALI PROCESS
tota I
we i qht
2740.0
percent
nonwa te r
1.00
percent specific net BTU
d i sso I ved q ra v i t.v content
const i tutent
of concern
CARBON TETRACHLORID
CHLOROFORM
.05
concen-
t ra t i on
. 108
.737
1.5
percent
d i ssoIved
1.0
1.0
40000.
wa ste
type
OU
04
mo I ecu I a r
weight
154.
119.
vapo r
pressure
.8
0.
0.
0.
vapor
pressure
90.
150.5
applicable treatment technologies:
incineration at all four DRE's
no treatment
EPA SIC
ff No. Sou rces
K022 2869 30E 6
so Iub iIi ty
Q 25 C.
84000.
.004
.004
.004
EPA
JL_
K073 2812 7 1,7
SIC
No. Sou rces
so Iub iIi ty
Q 25 C.
800.
8000.
-------
waste stream
& name
67
DECANTER TANK TAR SLUDGE FROM COKING
tota I
we iqht
3000.0
const i tutent
of concern
PHENOL
percent
nonwater
1.00
percent specific
d i sso I ved a ray i t.v
.97
concen-
trat ion
.00003
1.0
percent
d i ssoIved
net BTU
content
31000.
waste molecular
type we ight
1.0
06
94.
applicable treatment technologies:
chemical fixation/stabilization
no treatment
solvent extraction
EPA
K087
vapor
pressure
.8
so Iub iIi ty
@ 25 c.
81000.
SIC
No. Sources
331
waste stream
A name
68
PCB FLUIDS
tota I
weight
const i tutent
of concern
percent
nonwater
1.00
percent specific net BTU
d i sso I ved oravi t.v content
concen-
trat i on
1.0
percent
d i ssoIved
EPA
_j2_
N/A
SIC
No. Sources
waste molecular
t.vpe we i ght
vapor
pressure
so Iub iIi ty
@ 25 C.
28
28
PCB-1254
1.0
04
328.
.00008
.03
applicable treatment technologies:
chemical destruction
distillation and incineration at all four DRE's
incineration at all four DRE's
no treatment
-------
waste stream
g name
EPA SIC
U No. Sou rces
69
CONCENTRATED ORGANICS N.O.S.
N/A
tota I
we iqht
3100.0
const i tutent
of concern
percent
nonwa te r
1.00
percent specific
d i ssoIved q rav i tv
.05
concen-
t ra t i on
.2
2.5
percent
d i ssoIved
net BTU
content
7000.0
waste molecular
type we i ah t
04
1,1-DICHLOROETHENE .2 1.0
applicable treatment technologies:
incineration at all four DRE's
no treatment
waste stream
£ name
70 HEAVY METAL SLUDGES FROM PETROLEUM REFINING
99.
tota I
we i ght
36246.
pe rcent
nonwater
. 180
percent specific
d i ssoIved gravi tv
net BTU
content
const i tutent
of concern
CHROMIUM (VI)
LEAD
.055
concen-
trat i on
.00003
.00027
2.5
percent
d i sso I ved
.02
.60
4000.0
waste
type
08
08
applicable treatment technologies:
incineration at all four DRE's
no treatment
solvent extraction and centrifuge
vapor
p ressu re
230.
mo I ecu I a r
we i ght
52.
207.
so I ub i I i ty
® 25 C.
400.
EPA
JL_
SIC
No.
K048, K049, K050, K051 2911
Sources
2,6
vapor
pressure
0.
0.
so I ub iIi ty
@ 25 C.
50.
150.
-------
waste stream
£ name
71
HEAVY METAL SLUDGES FROM PETROLEUM REFINING
tota I
we i g h t
6301.0
const i tutent
of concern
CHROMIUM (VI )
LEAD
pe rcent
nonwater
.600
percent specific
d i sso I ved q rav i t.v
.12000
concen-
trat ion
.000018
.00029
2.5
percent
d i ssoIved
.02
.50
net BTU
content
16000.
waste molecular
type . we ight
08
08
52.
207.
applicable treatment technologies:
incineration at all four DRE's
no treatment
solvent extraction and centrifuge
waste stream
£ name
72 HEAVY METAL SLUDGES FROM PETROLEUM REFINING
tota I
we i ght
499.00
const i tutent
of concern
SELENIUM
CHROMIUM (VI)
LEAD
pe rcent
nonwater
.470
percent specific net BTU
d i sso I ved gravi t.v content
.36000
concen-
t ra t i on
.000027
.000006
.000078
2.5
percent
d i ssoIved
.02
.02
.50
3600.0
waste molecular
t.vpe we i ght
08
08
08
632.
52.
207.
applicable treatment technologies:
incineration at all four DRE's
no treatment
solvent extraction and centrifuge
EPA SIC
U No. Sources
6 2,6
K048, K049, K050, K051 2911
vapor
pressure
0.
0.
so Iub iIity
® 25 C.
50.
150.
EPA
JL.
SIC
No.
K048, K049, K050, K051 2911
Sou rces
6 2,6
vapor
pressure
so I ub iIity
Q 25 C.
0.
0.
50.
150.
-------
waste stream
g_ name
73
HEAVY METAL SLUDGES FROM PETROLEUM REFINING
tota I
we i q h t
4940.0
const i tutent
of concern
pe rcent
nonwa te r
.470
pe rcent
d i ssolved
.24000
spec i fic
q rav i ty
2.5
net BTU
content
concen-
tration
pe rcent
d i ssoIved
7600.0
wa ste
type
mo I ecu I a r
we ight
vapo r
p ressu re
PHENOL
CHROMIUM (VI)
LEAD
.000014
.00011
.0031
.98
.02
.50
applicable treatment technologies:
incineration at all four DRE's
no treatment
solvent extraction and centrifuge
waste stream
£ name
08
08
08
94.
52.
207.
74
METAL FLUORIDE SLUDGE FROM PETROLEUM REFINING
tota I
we i ght
2200.0
percent
nonwa te r
.460
percent specific net BTU
d i ssoIved gravi ty content
.4
2.5
2300.0
.8
0.
0.
constitutent concen- percent waste molecular vapor
of concern trat ion d i ssoIved type weight pressure
CYANIDE .000023 .98 12 27. 0.
FLUORINE (FLUORIDES .00009 .15 12 38. 0.
LEAD .0000071 .98 12 207. 0.
NICKEL .000055 .98 12 59. 0.
applicable treatment technologies:
no treatment
solvent extraction and chemical precipitation
solvent extraction and chemical precipitation and vacuum filter
EPA
K048, K049, K050, K051 2911 6 2,6
SIC
No. Sou rces
so Iub iIi ty
@ 25 C.
84000.
50.
150.
EPA SIC
ft No. Sou rces
N/A 2911 15 15
so Iub i I i ty
@ 25 C.
480000.
40000.
150.
100.
-------
waste stream
£ name
75
SPENT CLAY FROM OIL REFINING
tota I
we iqht
7000.0
const i tutent
of concern
percent
nonwater
1.00
percent specific
d i sso I ved gravi t.v
BENZ0(A)ANTHRACENE
BENZ0(A)PYRENE
LEAD
concen-
trat ion
.01
.01
.00004
2.5
pe rcent
d i ssoIved
.02
.02
.02
applicable treatment technologies:
chemical fixation/stabilization
incineration at all four DRE's
no treatment
solvent extraction
waste stream
# name
net BTU
content
6600.0
waste
type
08
08
08
mo I ecu I a r
we i qht
228.
252.
207.
76
ACID TAR FROM OIL RE-REFINING
tota I
we i ght
12190.
percent
nonwater
.900
percent
d i ssoIved
spec i f ic
gravi t .v
2.5
net BTU
content
18000.
constitutent concen- percent waste molecular
of concern trat ion d i ssoIved type we i ght
BENZ0(A)ANTHRACENE .01 1.0 06 228.
BENZO(A)PYRENE .01 1.0 06 252.
LEAD .01 .98 06 207.
applicable treatment technologies:
chem i caI f i xat i on/stab iIi za t i on
incineration at ail four DRE's
no treatment
solvent extraction
EPA SIC
H No. Sources
N/A 2992 14E 6
vapor so Iub iIi ty
pressure Q 25 C.
0. .004
0. .004
0. 150.
EPA SIC
U No. Sou rces
N/A 2992 14E 2,14
vapor so Iub iIi ty
pressure Q 25 C.
.004
.004
150.
0.
0.
0.
-------
waste stream
£ name
EPA
-JL-
SIC
No. Sources
77 CAUSTIC SLUDGE FROM RE-REFINING N/A 2992 14E 2
tota I
we i qht
pe rcent
nonwa te r
pe rcent
d i ssoIved
spec i f i c
a rav i tv
net BTU
content
82190.
.400
. 1
2.5
20000.
const i tutent
of concern
concen- percent
tration dissolved
waste
type
mo I ecu I a r
we i qht
vapo r
pressure
so Iub iIi ty
@ 25 C.
BENZ0(A)ANTHRACENE
BENZO(A)PYRENE
LEAD
.01
.01
.02
1.0
1.0
.02
06
06
06
228.
252.
207.
0.
0.
0.
.0C4
.OOU
150.
applicable treatment techno 1oqies:
chemical fixation/stabilization
incineration at all four DRE's
no treatment
solvent extraction
waste stream
H name
EPA
78 SCRAP BATTERIES
N/A
tota 1
we i qht
pe rcent
nonwater
percent
d i sso1ved
spec i f i c
q rav i tv
net BTU
content
82.000
1.00
1.
6.0
0
const i tutent
of concern
concen- percent
tration dissolved
waste
type
mo I ecu I a r
we iqht
vapor
pressure
so Iub iIity
@ 25 C.
CADMIUM
LEAD
MERCURY
.01
.01
.01
00
00
00
01
01
01
112.
207.
201 .
0.
0.
.0018
2.6
150.
20.
applicable treatment technologies:
I each i rig
no treatment
-------
waste
I
st ream
name
EPA
JL-
SIC
No. Sources
79 ARSENIC SALTS FROM PESTICIDE PRODUCTION K031 2879 6 6E
total percent percent specific net BTU
ve i qht nonwater d i ssoIved q rav i tv content
15000. 1.00 1. 5.0 0
const i tutent
of concern
concen-
t ra t i on
percent
d i ssoIved
wa ste
type
molecular
ve i aht
vapor
pressure
so Iub iIi ty
@ 25 C.
ARSENIC
.0063
00
02
75.
0.
500000.
applicable treatment technologies:
chemical fixation/stabilization
I each i ng
no treatment
waste stream
ft name
80 RESIDUES FROM THE RUBBER INDUSTRY
EPA
.1—
N/A
SIC
No.
30
Sources
18 18,24
total percent percent specific net BTU
we i ght nonwater d i sso I ved g ra v i t.v content
712.00 1.00 1. 5.0 0
const i tutent
concen-
percent
waste
molecular
vapor
so Iub i I i ty
of concern
trat ion
d i ssoIved
tvpe
we i qht
pressure
@ 25 C.
HEXACHLOROETHANE
.001
00
01
237.
.4
50.
VANADIUM
.00015
00
01
51.
0.
10000.
ZINC
.01
00
01
65.
0.
2000.
CHROMIUM (VI)
.00015
00
01
52.
0.
50.
LEAD
.0001
00
01
207.
0.
150.
NICKEL
.0001
00
01
59.
0.
100.
applicable treatment technologies:
chem i c-a I f i xa t i on/stab i I i za t i on
no treatment
-------
waste stream
£ name
81
EMISSION CONTROL DUSTS FROM IRON & STEEL PROD.
tota I
we i ght
17150.
percent
nonwater
1.00
percent specific net BTU
d i sso I ved q ra v i t.v content
const i tutent
of concern
CHROMIUM (VI)
LEAD
concen-
tration
. 0045
.0014
5.0
pe rcent
d i ssoIved
00
00
wa ste
type
02
02
mo I ecu I a r
we iqht
52.
207.
applicable treatment technologies:
chemical fixation/stabilization
I each i ng
no treatment
waste stream
U name
82
EMISSION CONTROL DUST FROM SECONDARY LEAD SMELTING
tota I
we i g h t
350.00
const i tutent
of concern
pe rcent
nonwa te r
1.00
percent specific net BTU
d i ssoIved q rav itv content
1.
concen-
t ra t i on
6.0
pe rcent
d i ssoIved
waste
type
mo I ecu I a r
weight
EPA
K061
vapor
pressure
0.
0.
so Iub iIi ty
Q 25 C.
50.
150.
EPA
K069
vapor
pressure
so Iub i I i ty
@ 25 C.
SIC
No.
331
Sou rces
6 6
SIC
No. Sou rces
3341 6,27 2,6C
CADMIUM
.0009
00
01
112.
0.
2.6
CHROMIUM (VI)
.00015
00
01
52.
0.
50.
LEAD
.12000
00
01
207.
0.
150.
applicable treatment technologies:
chemical fixation/stabilization
I each ing
no treatment
-------
waste stream
& name
83 EMISSION CONTOL DUSTS, SLAGS & OTHER RESIDUES, N.O.S.
tota I
we i q h t
3100.0
percent percent specific
nonwater d i ssoIved gravi tv
1.00 1. 2.5
net BTU
content
const i tutent
of concern
CADMIUM
concen-
tration
.03
percent
d i ssoIved
00
waste molecular
type we ight
02
112.
applicable treatment technologies:
chemi caI f i xat ion/stabiIi za t i on
I each i ng
no treatment
EPA SIC
U No. Sources
N/A 33 EE
vapor so Iub iIi ty
pressure Q 25 C.
0. 2.6
-------
APPENDIX 7
MODEL FOR ENVIRONMENTAL EXPOSURE
ICF INCORPORATED
-------
APPENDIX 7
MODEL FOR ENVIRONMENTAL EXPOSURE
We developed the following model for the limited purpose
of comparing risks among the three media and for considering
the effects of differing time scales. It is used to establish
the framework in two ways: to establish the mathematical form
of the dependence of exposure on the factors that are scored;
and to match the scoring systems in the various media so that
the same score in different media will correspond to similar
potential for exposure. The model is highly schematized and
is not intended to yield precise estimates of exposure or risk
in any one medium.
The dispersion of toxic chemicals in the environment and
the potential for human exposure vary greatly from one medium
to another. If comparisons of risk are to be made between
chemicals released into different media, such explicit models
of dispersion need to be developed. Should time be an important
factor in considering risks, it is necessary to take explicit
account of the great differences in the speed with which pollutants
are dispersed in the three media.
Our model works in the following way.
Consider a point source that emits a material at constant
rate R from time 0 to time T. The material diffuses outward
in one medium in a channel of depth D, within a radial angle
0, at constant velocity v. At time t, material released at
1
-------
time 0 will have traveled a radial distance r=vt. [Note:
D may be constant or proportional to rly/^.\]
The material is removed from the medium with time constant
X""1". [Note: X includes degradation, intermedia transfer,
irreversible adsorption, or other irreversible removal processes.
Reversible adsorption is included in the definition of v.]
ft T ft ^
The population density is assumed to be Ne ep per unit
area. [Note: a may be positive if the material is released
in an area of low population density, negative if in an area
of high density. 8 is the rate of population growth.]
The concentration of the material is denoted by c(r),
and one person takes in a quantity Ic(r) per unit time. [Note:
I is an intake factor and is the product of the amount of the
medium taken in and the absorption coefficient.]
If one person is exposed to dose rate B for time dt- the
incremental risk to that person is pBdx/L, where L is the average
lifetime and p is a risk coefficient. For this simple model
we assume that p is constant; thus, if b is constant, the lifetime
risk = pB.
The total population risk is defined to be the sum of
all risks to individuals within the exposed population. Hence,
if ni is the number of people exposed to dose rate B^, the
incremental risk to the population is n^pB^dt/L. The total
population risk is then
2
-------
T
J (ppn^/LjJdt
[Note that the time integral can extend over more than one
generation.]
1. MODEL
2. Calculations from the Model
The quantity emitted from the source between times t'
and t'+dt is Rdt. At time t'+t, this material will have been
3
-------
reduced in quantity to Rdte~^t and will be distributed in an
2
annulus with volume r0Ddr = v t0Ddt (see diagram).
Also, Mt.-» - M£e-(Wv>
Hence c(r) = concentration at radius r = 1 e~r^/v
Number of people exposed = NrSdre(ar+&t)
= Nr0drer(a+&/v)
Hence, the incremental risk in the annular area
= RicpdT Nr0drer (ot+^V*
L
= RIpN -r(X/v-a-B/v), .
VLD T
[Note: This quantity is independent of 0; if the material
is dispersed through a wider angle, the reduced concentration
is offset by a larger number of people exposed.]
To calculate the total risks presented by a facility,
this equation is integrated over all relevant values of r and t,
Depending on the way in which risks are discounted and on the
timing of release, several distinct possibilities arise. The
next three sections present calculations for the most important
situations.
3. Model 1: Discounted Risks
People in the annular area are exposed between times (r/v)
and (T+r/v) . If, therefore, risks are discounted at rate 6"^",
4
-------
total population risk accruing in this area
T+r/v .
= RIpN -r(X/v-a-B/v)»drJ e"0T
VLD e dr r/v
= RIpN Q c- )c-g(X/v-a-B/v+6/v)dr
vLD6
Thus, total population risk
= RIpNfi c-01)f -r(X/v-o-B/v+6/v)d
VLD v J O
= RIpN(l-e~^T)
LD(X-av-B+6)6
Model 2: Undiscounted Risks, Infinite Time
If the material is released for a time long compared with
X"1, a steady state is established Then people in the annular
area are exposed throughout their lifetimes. Thus, risk per
lifetime in the annular area
L
= /^incremental risk
J o
= I^N -r (X/v-ot-B/v),
vD
Hence, total risk per lifetime
00
= JQ l£N -r(X/v-o-B/v)d
vD
= RIpN
D(X-av-B)
This is equivalent to setting 6L = 1 and T>>L in the first formula,
5
-------
Model 3; Undiscounted Risks, Emission for Finite Time T
People in the annular area are exposed between times (r/v)
and (T+r/v). Therefore, the population risk experienced in
the annular area
= RIpN -r(X/v-a-B/v)d fr/v+T -6t,
vLD e arJr/v € aT
= RIPNT -r(X/v-o-B/v)d
VLD
Hence, total population risk
= RIpNT r°° -r (X/v-a-B/v) -
vLD Jo
= RIpNT
LD(X-av-B)
Model 4: Risks Fully Discounted Beyond Time t=T*
Total risk between time t1 and t'+dx over all space
= RIpNdt / e~r (^/v~a~B/v) (jr, provided that T*>T,
vLD
= RIpNdx
LD (X-av-B)
Thus, risk over all time
= RIpNT*
LD(X-av-B)
-------
In all models, the case a>0 corresponds to that in which
the chemical is released in an area of low pop^ation density
and diffuses outward into an area of higher density; the increased
density partially offsets the degradation. This model is only
meaningful, however, for a fairly short-lived chemical. If
av>X, the chemical becomes widely dispersed relative to the
scale of variability in population density. If (av-X) is sufficiently
large, the density of the popu'ation exposed to the chemical
becomes statistically uniform, and we can replace N by its
large-scale spatial average and set a = 0. This case is referred
to below as "Widely dispersed chemical.n
General Comments
The total population risk in all models is equal to
x a time factor
[Note: The total risk is independent of 0 and v (except that
v enters into the time factor if the population is nonuniformly
distributed). 0 and v otherwise influence only the distribution
of the risk and not its magnitude.]
If risks are not discounted, the time factor is
T
L(X-
-------
If T is large,
this for large 6 (6>>X) (discounting dominated),
^ 1
L6X for large X (X>>6) (degradation dominated).
In either case, the time factor can be written as the product
of T (or 1/6) and T* = l/(X-av-B). The former is the time
scale over which risks are counted (either the time of release
or the discounting time, whichever is smaller). T* is the
time scale for removal of the chemical from the human environment
(either the time scale for degradation or the time scale for
transport away from the human population, whichever is smaller).
Modeling of Scores in Different Media
To calibrate the system, we match the scoring systems
for the various media so that they give approximately the same
score for otherwise similar conditions. We consider the case
in which the following parameters are held constant:
p (lifetime risk per unit increment of dose—that is,
per unit of dose rate for unit time);
R (release rate in mass per unit time);
8 = 0.01 (population doubles in 69 years);
T = 35 years (35=year lifetime of facility) = 1/2 L; and
T* = 35 years (6 = 0.029, or risks discounted to half after
24 years);
Then relative risks in the various media are proportional to
IN
D(X-av-8)
8
-------
We assume the following as representative values for population
density:
Short-lived chemical (otv<\) , emission in areas of medium density
_ c o 2
N = 6x10 /m density 60 persons/km , decreasing
a = -lxlO'^nT^ by a factor of 10 within 207 km
Short-lived chemical (ctv<\) , emissions in areas of high density
_ •> 2
N = 10 /m city with central density
_ c _ i 2
a = -3x10 m ^ 1,000 persons/km , falling
off by a factor of 10 within 69 km
Short-lived chemical (av<\), emissions in areas of low density
— 52 2
N = 10 /m density 10 persons/km , increasing
a = +3xl0~®m~^ by a factor of 10 within 690 km
Widely dispersed chemical (av>X)
— 5 2
N = 3x10 /m area with average population density
a = 0 of U.S.
We also present data for a representative case in which
the bioaccumulation factor 10 is sufficiently great that inges-
tion via fish is a more important route of exposure than direct
ingestion of the chemical in drinking water. Specifically,
we present calculations for the case k = 3, so that exposure
via fish is about 10 times greater than that via drinking water
(see text) .
9
-------
We assume the following values for the media-specific
parameters:
Air
Surface
Water
Surface Water
and Fish
Ground
Water
v (m/year)
2xl08
(14 mph)
2xl07
(1.4 mph)
2xl07
(1.4 mph)
2xl03
(2 km/yr)
D (m)
3xl03
10
10
30
I (m3/year)
5xl03
7xl0_1
7xl0-1
7xl0_1
f (bioaccumu-
lation factor)
1
1
1,000
1
X (year-1)
i—1
o
u>
102
102
1
v/X(km)
2xl02
2xl03
2xl03
2
To illustrate the way in which the risk depends on the environ-
mental lifetime of the chemical, we also calculate risks for values
of X greater by a factor of 10 and smaller by factors of 10 and
100. Smaller values of X apply to chemicals with larger environment;
half-life (and correspondingly wide environmental distribution).
The values used in this calculation range over three orders of
magnitude and are sufficient to illustrate the behavior of the
risk function, although a wider range of values of X is considered
in Tables 4-1 to 4-3.
10
-------
The media-specific risk factors [IN/D(X-av-3)] are then as
follows:
Air
Surface
Water
Fish
(f=1000)
Ground
Water
Short-lived chemical
High density
(N=10 )
Medium density
(N=6xl0 )
Low density
(N=10 )
High density
(N=10 , XxlO)
2xl0"7
3xl0"8
4xl0"8
10
-7
10"7
2xl0"8
2xl0'8
Widely dispersed chemical (N=3xl0~"\ a=0]
X/10
X/100
5x10
-7
2x10
,-7
5x10 6 2xl0"6
10
-6
2x10
-7
2x10
-7
4xl0~8 4xl0"7
2x10
-6
2x10
-5
2x10
10"6
2x10
2x10
7x10
7x10
-5
-7
-6
-6
-5
These figures were used to match the scoring systems in Tables
4-1 to 4-3/ by assigning a score of 2 to entries of approximately
— 8
3x10 in the table above.
Simplifications and Implicit Assumptions
The following simplifications and implicit assumptions
are made in the model:
1. Population density is assumed to be uniform throughout
the area affected by the chemical, varying only as ear and
independent of direction from the site of release. This is
justifiable as a statistical (ensemble) average, but omits
important variability owing to local patchiness.
11
-------
2. Widely distributed chemicals are assumed to affect population
with the same average density throughout the whole United States.
This ignores important east-west and upstrearn-downstream differences.
In subsequent developments of the model, we might assume a>0
for all chemicals discharged into surface waters since they
move downstream into more populated areas.
3. All people in the affected area are assumed to drink ground-
water and surface water and to eat fish. The fraction of the
population that does so will, in fact, always be less than 1.
In subsequent developments of the model, we might make assumptions
about this fraction in various environmental circumstances.
4. The chemical is uniformly distributed in a plume of constant
angle 6 and constant thickness D. It might be more realistic
to assume that 0 varies as r"^^ and D varies as r^^, but
this gives exactly the same mathematical result.
5. Risk is linearly proportional to dose. As explained in
the text, this assumption can be relaxed and generalized by
If
assuming that risk varies as dose , weighting the exposure
score by k and the toxicity score by k~*.
Finally, the model runs into several difficulties for
long-lived chemicals. In air and surface water, chemicals
with more than moderate half-lives move east over the Atlantic
Ocean or down the rivers to the sea. Because of their longer
half-lives, those moving east (at least) become global pollutants
and return to the United States. The prototype is carbon tetra-
chloride (CC14), every parcel of which stays in the air
12
-------
and returns past every individual in the United States every
20 days or so. It is logical to assume that every pass contributes
to the long-term population risk. After a few passes, however,
the "plume" no longer widens, and the model greatly underestimates
exposure. Another problem with long-lived chemicals is the
assumption of low-dose linearity. The model counts risks for
a very large number of people exposed to very low concentrations
and hence would overestimate risks if the assumption of low-
dose linearity is wrong. Finally, for long-lived chemicals
the problem of discounting risks becomes important. The model
allows for discounting, but an explicit decision has to be
made about the discount rate.
13
-------
APPENDIX 8
DOSE CONVERSION CALCULATIONS
ICF Incorporated
-------
APPENDIX 8
DOSE CONVERSION CALCULATIONS
We made the following dose conversion calculations for
our model:
1. Feeding Studies^-
Variable = Quantity
a
b
w
concentration of
substance in food by
weight
percentage of substance
in food by weight
weight of substance
weight of food
x (10~6 g/g) "(103 mg/g)
x 10"3
x' 10
weight of food consumed
body weight
Units
ppm
(10 b g/g)
percent
(10~2 g/g)
mg/g
g/day
kg
thus
dose rate
a b
w
mg/kg/day
1
-------
2
2. Drinking Water Studies
Variable
Quantity
Units
concentration of substance
in drinking water by weight
per unit volume
weight of substance
volume of water
ppm
(mg/liter H^0)
mg/liter
b
w
thus
volume of water consumed
body weight
dose rate
a b
w
g/day
kg
mg/kg/day
2
-------
3
3. Inhalation Studies
Variable
M
w
t
thus
Quantity
molecular weight of
substance
concentration of substance
in air by number of
molecules
concentration of substance
in air by weight per unit
volume
weight of substance
volume of air
x m
24450
22
using 2.463 x 10 molecules
of air/liter at 1 atm and 25°C
x' 10~3
minute volume (air breathed
per unit time)
body weight
length of exposure
dose rate
c v t
w
Units
g/mole
ppm
mg/m3
mg/liter
liter/min
kg
min/day
mg/kg/day
When lifetime average body weights and food consumption were
not given in a particular study, the values for various species
as given in Table A-l were used.
2
When lifetime average daily water consumption and body weight
were not given in a study, the values in Table A-l were used.
3It was assumed that all of the inhaled dose was absorbed into
the body. The minute volumes for various species are given
in Table A-l.
3
-------
TABLE 8A-1
SELECTED NORMATIVE VALUES
Species
Average
Average Daily Food
Body Weight3 Consumption
Average
Daily Water
Consumption
Average Respiratory
Minute Volume
(Liters)
Surface Area
Conversion.
K-Value
Human
M 70 kg
49 ml/kga
7.4 <5.8-10.3)d,(R)2 ,
28.6 {27. 3-30 .9) j (LW) "J
42.9 (39.3-45.2)° (HW)4
10.6
F 54 kg
49 ml/kga
4.5 (4.0-7.0)d tR)
16.3 (15.9-16.8)j (LW)
24.5 (17.3-31.8) (HW)
Guinea
Pig
M 899 ge 30 gc
F 837 g
0.16 (0.09-0.38)a
8.6 (8.4-8.9)
Hamster
120 g 10-14 ga
0.054 (0.025-0.083}a
Fat
M 364 (26 3-465)g 100 g/kgb
F 243 (196-290}g
80-110 ml/kgb
0.100 (0.075-0.I30)a
9.0
Mouse
M 33.1 (25.7-40.5)g 3 g°
F 32.3 (25.5-39 .1) g
180 ml/kga
0.023 (0.011-0.036)a
9.0 (8.4-9.4)
aSpector, 1956
bBaker et al., 1979
cLehman, 1959
dTaylor, 1941
eAltinan and Dittmer, 1972
*The K-value, although nominally a constant, is empirically derived and hence varies depending on the
the method used to measure the surface area of animals.
2 (R) Resting
3(LW) Light work
4 -HVT 'let wo '
-------
REFERENCES
Altman, P.L., and Dittmer, eds. 1972. Biology Data Book.
Vol. I. 2d ed. Bethesda, Md.: Federation of American
Society for Experimental Biology.
Baker, H.J., Lindsey, J.R., and Weisbroth, S.H. 1980. The
Laboratory Rat. Vol. II Academic Press, New York: Appendix 1.
Food Safety Council. 1980. Proposed System for Food Safety
Assessment. Final Report of the Scientific Committee
of the Food Safety Council. June 1980. Washington, D.C.:
The Council.
Freireich, E.J., et al. 1966. "Quantitative Comparison of
Toxicity of Anti-cancer Agent in Mouse, Rat, Hamster,
Dog, Monkey, and Man." Cancer Chemother. Rep. 50:219-244.
Lehman, A.J. 1959. Appraisal of the Safety of Chemicals in
Foods, Dru^s, and Cosmetics. Association of Food and
Drug Officials of the United States.
Spector, W.S. 1956. Handbook of Biological Data. Philadelphia:
W.B. Saunders, p. 175.
Taylor, L. 1941. Am. J. Physiol. 135:27 (as cited in Holtman,
Gibson, Wang, Dittmer, and Crebe. 1958. Handbook of
Respiration)
5
-------
APPENDIX 9
CALIBRATION OF RISK SCORING SYSTEM
USING CASE STUDIES
ICF Incorporated
-------
APPENDIX 9
CALIBRATION OF RISK SCORING SYSTEM
USING CASE STUDIES
In order to relate our relative risk scores to absolute
scores, we selected four case studies of actual occurrences:
cadmium emissions to air from a smelter; leakage of chloroform
from barge terminals into the Mississippi River; discharge
of PCBs into the Hudson River; and road salt (sodium) runoff
to groundwater. The calibration procedure consists of three
parts: calculating the relative risk score; calculating an
estimated risk to the population based on the case study; and
correlating the relative risk score with the population risk.
Calibration of Risk Scoring System for Air Emissions
We based our calibration of the scoring system for air
on a study of air emissions of cadmium from a lead smelter
and slag processing facility in East Helena, Montana (see Rupp
et al.f 1978). The environmental assessment model used in
that study is a site specific, and rapid deposition of cadmium
(corresponding to highly localized pollution) is assumed.
Rupp et al. presented data indicating that significant exposure
to cadmium (>1.2 yg/100 m^ in air) has been found within a
12.9-km radius of the plant, an area of 520 km . By estimating
human population within this area to be 5,000, we calculated
a population density of 10 persons/km . This population density
puts the smelter area in the low density category of the scoring
scheme. We assigned an exposure score of 3.5 on the assumption
1
-------
that the principal mechanism of cadmium removal is washout
from the atmosphere. Because exposure scores greater than
3 are unmodified by the population density, the score remains
3.5.
The toxicity score for cadmium is based on a critical
review by Fleischer et al. (1974). For human exposure by in-
halation, an intake of 26 yg/day, or 0.4 yg/kg/day, appears
to lead to a critical accumulation in the kidney over 50 years
of exposure (Fleischer et al. 1974, p. 298). Assuming this dose
rate to be an MED, the MED/10 is estimated to be 0.04 yg/kg/day,
corresponding to an inherent hazard score of 6.
The score for the release rate is calculated by taking
the logarithmic (base 10) of the number of metric tons released
annually. Using the data of Rupp et al. (1978, Table 3), we
estimated cadmium release from the smelter to be 4 x 10^ grams/year
or approximately 4 tons/year. The overall score is computed
as the sum of the exposure score (as modified by population
density), the inherent hazard score, and the release score:
3.5 + 6 + 0.6 = 10.1
To derive an independent estimate of the population risk,
we proceed as follows. The contribution of the Helena smelters
to cadmium levels in air near the facility has been measured
at 0.06-0.29 yg/m3 (Rupp et al. 1978, Table 1). We assume
that exposed persons breathe 15 m of air per day; thus, the
daily intake is 0.9-4.3 yg/day. This range of intakes is centered
on the value of 2.6 yg/day, which we previously estimated to
2
-------
be the MED/10, and hence would correspond to a 1 percent response
rate in the population. If the population of the area affected
is assumed to be 5,000, this response rate would result in
50 cases over a lifetime, or an average of 0.7 cases per year.
By comparing the risk score with the estimated population
risk, we find that a score of 10.1 corresponds to 0.7 cases
per year. A score of 10.3 would, therefore, correspond to
a risk of 1 case per year.
Calibration of Risk Scoring System for Fresh Water Releases
Chloroform in Water
The first calibration of the scoring system for aquatic
discharges is based upon a well-documented spill of chloroform
into the Mississippi River when two tank barges ruptured at
Baton Rouge, Louisiana. Neely et al. (1976) presented measure-
ments of the time-concentration profile of chloroform in the
water at two points downstream. According to their data, 1.75 mil-
lion pounds (about 800 tons) of chloroform were spilled. This
led to an average concentration of about 170 ppb over a 24-hour
period at a point 25 km downstream, and to an average concen-
tration of about 60 ppb over a 48-hour period at a point 200 km
downstream. The mean velocity of the river was about 2 km/hour.
For the purposes of this calibration, we will consider a case
in which the discharge of 800 tons of chloroform is distributed
throughout a year. By assuming that the relationship between
release and downstream concentrations is linear and independent
of release rate, we calculate the average concentrations to
3
-------
be about 0.5 ppb (170/365) 25 km downstream and 0.33 ppb (60x2/365)
200 km downstream.
To calculate the risk score, we note that the data of
Neely et al. suggest that the average concentration would fall
by somewhat less than half in 87 hours. This corresponds to
a basic score of 2. Neely et al. also cite laboratory data
that suggests that the half-life for evaporation of chloroform
from water solution is less than 1 hour. For the purpose of
this calculation, we, therefore, assign a basic score of 1.
This is modified for high population density to yield an exposure
score of 2 (chloroform is not significantly bioaccumulated—
log PQct = 2). EPA's Carcinogen Assessment Group (CAG) has
estimated the unit risk for carcinogenic effects of chloroform
via ingestion to be 0.167 (mg/kg/day)_1. This corresponds
to an inherent hazard score of 3. The release score is the
logiO of the number of metric tons released per year, or 2.9.
Hence the overall risk score is 2 + 3 + 2.9 = 7.9.
To calculate the population risk, we assume that the average
exposed person would consume 2 liters of water containing 0.4 ppb
chloroform—that is, about 0.8 yg/day, or about 0.012 yg/kg/day.
Combining this with the unit risk for chloroform, we estimate
the lifetime risk to an individual to be 2 x 10~®. On the
assumption that the exposed population is 2.5 million persons,
i
we estimate the number of cases to be 5 per lifetime, or about
0.07 cases per year.
4
-------
Comparing the risk score with the estimated population
risk, a score of 7.9 corresponds to 0.07 cases per year. Hence
a score of 9.1 would correspond to 1 case per year.
PCBs in Fish
The second calibration of the scoring system for aquatic
discharges is based upon the contamination of the Hudson River
with PCBs. According to Turk (1980), about 225,000 kg of PCBs
had been released into the Hudson River over a period of 22 years.
Turk estimated (p. 180) that current deposits of PCBs in bottom
sediments were sufficient to maintain current levels of water
contamination for about 100 years. Hence, current levels would
be maintained indefinitely in a steady state by a constant
discharge of about 2,250 kg/year, or 2.2 tons/year. Other
sources indicate that PCB levels in fish in the Hudson River
range from 2 ppm to 800 ppm. We assume a mean value of 10 ppm
in edible tissue.
To calculate the risk score, we assume the 100-year lifetime
to correspond to a basic score of 6. This is modified by adding
2.2 points for the bioaccumulation factor of 3.7 x 104, to
yield an exposure score of 8.2 (no modification is required
for population density). This score corresponds to the middle
of the range listed for PCBs in Table 4=4. The inherent hazard
score is 5 (see Table 4=4), which is based on a MED of about
0.006 mg/kg/day for reproductive and behavioral toxicity in
rhesus monkeys. In this case, the dose corresponding to a
1 percent response rate in humans is estimated to be MED/10
5
-------
since only a small scaling factor is needed between humans
and monkeys. The release score is the l°g1Q of the number
of metric tons released per year—that isf 0.3 under the assumed
conditions of steady state discharge. Hence the overall risk
score is 8.2 + 5 + 0.3 = 13.5.
In order to calculate the risk to the population, we estimate
the average daily intake of PCBs by persons eating fish from
the river to be 160 yg/day, based on the consumption of fish
by the average person of about 16 g/day. This corresponds
to about 2.5 yg/kg/day, or about 0.4 times the MED on which
the risk score was based. This would correspond to a response
rate of about 0.04. The most difficult factor to estimate
is the number of persons exposed. The total population living
near the river and estuary exceeds 10 million, but most of
the fish consumed by these people comes from elsewhere. Commercial
catches in the river would have supplied a daily intake of
16 g/day to only about 50,000 people. Sport fishing is, however,
probably more important than commercial fishing. Since the
model assumes high utilization of fish from affected rivers,
we will assume a total of 300,000 consumers. Then the predicted
number of cases is 12,000/lifetime or about 170 cases per year.
Comparing the risk score to the estimated population risk
yields a score of 13.5, which corresponds to about 170 cases
per year. Hence a score of about 11.3 would correspond to
1 case per year.
6
-------
Calibration of Risk Scoring System for Groundwater Contamination
We based the calibration of the scoring system for groundwater
on statistics on the contamination of groundwater with sodium
resulting from the use of sodium chloride on roads to melt
snow and ice. Although this is a dispersed release rather
than a point discharge, we believe that it provides a useful
average measure of the relationship between release of a chemical
into groundwater and human exposure to it via drinking water.
According to the Massachusetts Department of Public Works
(1978), about 200,000 tons of NaCl were used by State agencies
on roads in Massachusetts each year, primarily in the eastern
half of the State. The quantity used by private agencies was
not recorded precisely, but was probably roughly comparable
to that used by State agencies about 200,000 tons. The area
2
concerned is about 8,000 km and contains about 2 million people
(excluding the Boston metropolitan area whose water is drawn
from an uncontaminated supply). Also according to the same
source, about 35 percent of the salt applied to roads penetrates
into groundwater. Concentrations of chloride and sodium have
risen steadily in areas where salt is heavily used, parallel
with the rate of application, and chlorides have reached average
levels of about 80 mg/liter. This would correspond to about
50 mg/liter of Na+.
To calculate the risk score, we assume a 10-year half-
life for sodium in groundwater (based on the general geologic
characteristics of the area, with shallow aquifers and usually
7
-------
rapid recharge rates). This corresponds to an exposure score
of 5, which does not require modification for population density
(mean density about 300/km^). We note that the average intake
of NaCl by the U.S. population is about 2 g/day, and that this
is believed to aggravate hypertension (and hence to increase
mortality from related causes) in at least 10 percent of the
population. Persons on low salt diets are limited to about
200 mg/day. Accordingly, we will assume that the MED is about
2,000 mg/day and that 200 mg/day would correspond to about a
1 percent response rate. This is a MED/10 of about 3.0 mg/kg/day
of Na+, that is, an inherent hazard score of 2. Of the 400,000 tons
of road salt, 160,000 tons are NA+. The number of metric tons
of Na+ entering groundwater annually is about 35 percent of
160,000 tons, that is, 56,000 tons. The score for release
rate is the log^g of this number 4.8.
The overall risk score is the sum of the scores for expo-
sure, inherent hazard, and release rate:
5 + 2 + 4.8 = 11.8
To estimate the risk to the population, we assume the
average intake by an individual in the area is 2 liters/day
x 50 mg/liter = 100 mg/day Na+. Because this is slightly higher
than our estimate of the MED/10, we estimate the response rate
as 0.012. In a population of 2 x 10^ persons, this corresponds
to 24,000 cases per lifetime, or 330 cases per year.
Comparing the risk score with the estimated risk to the
population, we find that a score of 11.8 corresponds to about
8
-------
330 cases per year. Accordingly, a score of about 9.3 wo\ild
correspond to 1 case per year.
Conclusions
The above independent estimates vary between 9.1 and 11.3.
The relatively close agreement between these estimates (within
about two orders of magnitude) confirms that our exposure models
succeeded in matching the scoring systems for all of the media.
Where PCBs have been used to calibrate risk resulting from
contamination of surface water where the contaminant is bioaccumu-
lated, our estimate of risk becomes a worst case in that the
total population is assumed to eat contaminated fish. Similarly,
the persistence of chloroform is likely to be somewhat greater
than we have estimated if a chemical is discharged at concentra-
tions greater than its solubility limit (that is, the rate
of dissolution at the interface may be rate limiting rather
than the volatilization rate).
9
-------
20
16
16
14
12
E
10
8
6
4
2
0
FIGURE 9A-1
RISK SCORE CALIBRATION BY MEDIA
———• Groundwater (Na+)
mimmmmm. Surface Water PCB
—— Surface Water (CHCIj )
¦ Alr(Cd)
to
-8
-6
-4
-2
—i
A
LOG10 OF CASES PER YEAR
-------
REFERENCES
The Habitat School of Environment. 1972. De-icing Salts and
the Environment. Prepared for Massachusetts and National
Audubon Societies (prepared by Habitat, Inc., Belmont,
Mass. February 1972)
Heisher, M., Sarofim, A.F., Fassett, D.W., Hammond, P.,
Shacklette, H.T., Nisbet, I.C.T., and Epstein, S. 1974.
Environmental impact of cadmium: A review by the panel
on hazardous trace substances. Environ. Health Perspect.
5:253-323
Neely, W.B., Blau, G.E., and Turner, A., Jr. 1976. Mathematical
models predict concentration-time profiles resulting from
chemical spill in a river. Environ. Sci. Technol. 10:72-76
Rupp, E.M., Parzyck, D.C., Walsh, P.J., Booth, R.S., Raridon, R.J.,
and Whitfield, B.L. 1978. Composite hazard index for
assessing limiting exposures to environmental pollutants:
Application through a case study. Environ. Sci. Technol.
12:802-807
Turk, J.T. 1980. Applications of Hudson River Basin PCB-transport
studies. Contam. Sediments 1:171-183
-------
APPENDIX 10 - SUMMARY
INHERENT HAZARD SCORING DATA BASE
ICF Incorporated
-------
APPENDIX 10 - SUMMARY
INHERENT HAZARD SCORING DATA BASE
This appendix summarizes the data appearing in Appendix 10. The references
listed in Table 1 are cited in full in the complete appendix. See the note below
for additional explanation of the contents of Table 1.
NOTE TO TABLE 1
The table lists results from the study that yielded the highest inherent hazard
score. For effects other than carcinogenesis, the score is based on the minimum
effective dose (MED) -- i.e., the lowest dose that yielded an effect judged to be
significant. For carcinogenic effects, the score is based on an estimate of the
slope of the dose-response relationship at low doses. Where a published estimate
of the unit risk q(1)was readily available, this is listed. In other cases, the
listed figure is I/MED, where MED is the lowest dose that yielded a significant
increase in tumor incidence I. This is a close approximation of q(l)" provided
that I is less than 0.5, but falls below it for greater values of I. All doses are
converted to units of mg/kg/day, using the conversion factors listed in Appendix 8.
ICF Incorporated
-------
TABLE 1
BASIS FOR INHERENT HAZARD SCORE
NON-CARCINOGENS
CARCINOGENS
Compound
MED qi* or I/MED
mq/kq/day Species/Route Target Organ(s) Primary Effect(s) (rog/kg/day)
-1
Cancer Type/
Target Organ
Reference
Acenaphthene
Acetaldehyde
Acetonitrile
Acrolein
Acrylamide
Acrylonitrile
4-Aminobiphenyl
Antimony
Arsenic
Asbestos
«<-BHC
ft -BHC
y-BHC
279 (subchronic)
hamster/
inhalation
Trachea, kidney
127 (subchronic) rat/inhalation Lung, kidney
0.016
7.5 (chronic)
human/inhalation Lung, eye
rat/oral
CNS
0.17 (subchronic) rat/inhalation Cardiovascular
carcinogenicity
ascribed by analogy
to other PAHs
Tracheal irritation;
increased kidney
weight
Congestion and edema
of the lungs; cloudy
swelling of the renal
tubules
Eye and respira-
tory irritation
Peripherial
neuropathy
Degeneration of the
myocardium
6.9
0.7
27.8
0.36
0.011
0.038
0.019
Astrocytoma
of CNS (rat)
Bladder
carcinoma (dog)
Respiratory
cancer (human)
Lung cancer and
mesothelioma
(human)
Hepatocellular
carcinoma (mouse)
Hepatocellular
tumor (mouse)
Liver tumor
(mouse)
U.S. EPA 1980
Kruysse et al.
1975
Pozzani et al.
1959
AIHA 1968
Fullerton and
Barnes 1966
Ouast et al.
1980
Deichmann et al.
1965
Brieger et al.
1954
Clement 1980
Schnelderman
et al. 1981
Ito et al. 1973
Thorpe and
Walker 1973
Thorpe and
Walker 1973
-------
TABLE 1 (cont.)
BASIS FOR INHERENT HAZARD SCORE
NON-CARCINOGENS
CARCINOGENS
Compound
MED
mg/kg/day
Species/Route
Target Organ(s)
Primary Effect(s)
qjL* or I/MED
(mg/kq/day)"1
Cancer Type/
Target Organ
Reference
Barium
Benzene
Benzidine
Benzo(a)anthracene
3,4-Benzofluor-
anthene
Benzo(a)pyrene
Beryllium
10 (chronic)
17 (chronic)
rabbit/injection CNS
rat/oral
Liver
Not specified
Fatty degeneration
Fazekus et al.
1953
0.0216
Incidence of hepa-
tomas not statis-
tically significant
0.015
0.0375
67
111.1
Leukemia (human) CAG 1978
Boyland et al.
1954
Hepatoma (mouse) Klein 1963
Injection site
sarcoma (mouse)
Skin tumor
(mouse)
Alveolar
adenocarcinoma
(rat)
Lacassagne
et al. 1963
Roe et al.
1970
Reeves et al.
1967
Bis(chloromethyl)-
ether
Bis(2-ethylhexyl) -
phthalate
Butane•
Cadmium
Carbon disulfide
Carbon tetra-
chlor ide
Chlordane
11.8
900 (subchronic)
233
0.1 (chronic)
0.396 (sub-
chronic)
15.9 (chronic)
rat/oral
human/
inhalation
rat/oral
rat/inhalation
guinea pig/
inhalation
Testes
CNS
Liver
Lung
CNS
Degeneration, tubu-
lar atrophy
Changes in VER
wave amplitude
Liver enzyme changes
Inflammation of the
bronchi
Optic nerve damage
Squamous cell
carcinoma of
lung (rat)
0.092
Hepatocellular
carcinoma (mouse)
Laskin et al.
1971
Shaffer et al.
1945
Stewart et al.
1977
IARC 1976
Misiakiewicz
et al. 1972
Adams et al.
1952
NCI 1977
-------
TABLE 1 (cont.)
BASIS FOR INHERENT HAZARD SCORE
NON-CARCINOGENS
CARCINOGENS
Compound
MED
mg/kq/day
Species/Route
Target Orqan(s)
Primary Effect(s)
qi* or I/MED
(mg/kg/day)-1
Cancer Type/
Target Organ
Reference
Chloroacetal-
dehyde
4-Choroaniline
Chlorobenzene
0.82 (sub-
chronic)
25
54.6 (subchronic)
rat/injection Hematopoietic,
lung
rat/diet
dog/oral
Spleen
Liver
Decrease in RBC and
lymphocyte count;
increase BSP dis-
appearance; histo-
pathologic changes
of lung
Nonneoplastic lesions
Not specified
Lawrence et al.
1972
NCI 1979
Monsanto Co.
1967
Chloroform
Chromium
Chrysene
Copper
human/oral
Not specified
MED based on NAS
(1973) maximum
tolerated dose
estimate
0.167
43.34
4
Liver tumor
Lung cancer
Skin tumor
(mouse)
CAG
CAG
IARC 1973
NAS 1973
Creosote
Cyanide
Cyclohexane
Dibenzo(a,h)¦
anthracene
Di-n-butyl
phthalate
1,2-Dichloro-
benzene
0.04
0.2
417 (subchronic)
318 (subchronic)
270 (chronic)
human/oral
rabbit/
inhalation
rat/injection
rat/oral
Death
Kidney, liver
Teratogenic
Liver
MED approximated
as LD]_o
Histopathologic
changes
Skeletal abnormali-
ties, reduced fetal
weight
Liver swelling
0.577
Epidermal cancer Poel and
(mouse) Krammer 1957
Gosselin et al.
1976
Treon 1943
Local sarcoma
(mouse)
Bryan and
Shimkin 1943
Singh et al.
1972
Hollingsworth
et al. 1958
-------
TABLE 1 (cont.)
BASIS FOR INHERENT HAZARD SCORE
Compound
MED
mq/kq/day
NON-CARCINOGENS
CARCINOGENS
Species/Route
Target Organ(s)
qi* or I/MED
Primary Effect(s)
(mq/kq/day)
-1
Cancer Type/
Target Organ
Reference
1,3-Dichloro-
benzene
800 (subchronic)
rat/oral
Liver
Hepatic porphyria
Poland et al.
1971
1,4-Dichloro-
benzene
3,3-Dichloro-
benzidine
265 (subchronic)
rat/inhalation Liver
Cloudy swelling,
centrilobular necrosis
0.0052
Mammary adeno-
carcinoma
Hollingsworth
et al. 1956
Stula et al.
1975
1,2-Dichloro-
ethane
44.6 (subchronic) rat/inhalation Lung, kidney
1,1-Dichloroethene 5 (chronic)
Dichloromethane 349 (subchronic)
2,4-Dichlorophenol
2,4-Dichlorophen-
oxyacetic acid
100 (subchronic)
1,2-Dichloropropane 441 (subchronic)
1,3-Dichloropropene 0.35 (chronic)
0,0-Diethyl S-(2- 6.75
(ethylthio)ethyl)
ester of phosphoro-
thioic acid
p-Dlmethylamino-
azobenzene
rat/oral Kidney
mice/inhalation Liver, kidney
rat/oral
human/oral
Teratogenic
rat/inhalation Liver
rat/inhalation Kidney
Hematopoietic
Mortality, pulmonary
congestion, fatty
degeneration of kidney
tubules
Chronic inflammation
Changed liver enzyme
levels, fatty infil-
tration) non-specific
kidney tubule changes
Decreased fetal sur-
vival and weight;
skeletal abnormalities
Changes in liver
enzyme activity
Cloudy swelling of
renal tubular
epithelium
Decreased plasma and
RBC cholinesterase
activity
0.0157
0.073
Skin papilloma
(mouse)
Liver tumor
(rat)
Heppel et al.
1946
NCI 1981
Haun 1972
Boutwell and
Bosh 1959
Khera and
McKinley 1972
Sidorenko
et al. 1976
Torkelson and
Oyen 1977
Rider et al.
1969
Druckrey and
Kupfmuller 1948
-------
TABLE 1 (cont.)
BASIS FOR INHERENT HAZARD SCORE
NON-CARCINOGENS
CARCINOGENS
Compound
MED qx* or I/MED
mg/kq/day Species/Route Target Organ(s) Primary Effect (s) (mg/kg/day)
-1
Cancer Type/
Target Organ
Reference
1,2-Dimethyl
hydrazine
Dimethyl phthalate 8,000 (chronic) rat/oral
Dimethyl sulfate
Dinitro-ortho- 1.2
cresol (DNOC)
2,4-Dinitrophenol 2
2,4-Dinitrotoluene 10 (chronic)
1,4-Dioxane
Endrin
Epichlorohydrin
Ethylbenzene
Ethylenebis-
d ithiocarbamate
Fluorides
94 (chronic)
human/oral
human/oral
dog/diet
rat/oral
1.25 (chronic) rat/oral
Kidney
CNS
Skin, eye
Hematopoietic
Kidney, liver
Brain, liver,
kidney, adrenal
glands
Depressed growth,
kidney effects
Lassitude, malaise,
headache
Skin irritation,
cataract formation
Changes in erythro-
cyte counts, reticulo-
cytes, Heinz bodies,
hemoglobin and methe-
moglobin levels
Degenerative histo-
pathologic changes
Diffuse degeneration
95.9 (subchronic) rat/inhalation Liver, kidney Organ weight changes
0. 24
human/
inhalation
Bone
Increase in bone
density
0.52
1.16
0.12
0.00306
Ang iosarcoma
(mouse)
Nasal cell
carcinoma (rat)
Toth and Wilson
1971
Draize 1948
Druckrey et al.
1970
Harvey et al.
1951
U.S. EPA 1980
Ellis et al.
1979
Kociba et al.
1974
Treon et al.
1955
Squamous cell
carcinoma of
nasal cavity (rat)
Laskin et al.
1980
Malignant liver
tumor (mouse)
Wolf et al.
1956
Bionetics 1968
Derryberry
et al. 1963
-------
TABLE 1 (cont.)
BASIS FOR INHERENT HAZARD SCORE
NON-CARCINOGENS
CARCINOGENS
Compound
MED
mg/kq/day
Species/Route
Target Orqan(s)
Primary Effect(s)
Formaldehyde
Hexachlorobenzene
Hexachlorobutadiene
Hexachlorocyclo-
pentadiene
Hexachloroethane
Hexachlorophene
n-Hexane
Hydrazine
0.396 (subchronic) rat/inhalation Lung, liver,
kidney
Isophorone
Lead
0.547 (subchronic) Guinea pig. Lung, liver,
rabbit, rat, kidney
mouse/inhalation
10 (subchronic)
630 (chronic)
rat/oral
mouse/
inhalation
Brain
CNS
human/
inhalation
Hydrogen chloride 0.64
Hydrogen sulfide 1.41 (subchronic) rat/inhalation CNS
Mucous membrane
Indenol1,2,3-c,d) 10 (chronic)
pyrene
4.85
0.0107
mouse/dermal
human/
inhalation
human/
inhalation
Skin
Not specified
Hematopoietic,
reproductive,
kidney
Histopathologic
changes
Degenerative changes
of liver and kidney;
pneumonia, edema,
bronchitis
Focal brain lesions
Effects on peripheral
nerve motor branches
at 'neuromuscular
junction
Irritation
Effects on motor
chronaxie
Carcinoma (no
incidence data)
Fatigue and malaise
Not specified
qi* or I/MED
(mg/kq/day)
-1
0.12
0.008
0.0005
2.8
Cancer Type/
Target Organ
Reference
Hepatoma
(hamster)
Renal tubular
neoplasm (rat)
Fel'dman and
Bonashevskaya
1972
Cabral et al.
1977
Kociba et al.
1977
Treon et al.
1955
Hepatocellular NCI 1978
carcinoma (mouse)
Nasal tumor
(rat)
Kimbrough 1973
Miyagaki 1967
MacEwen et al.
1980
ACGIH 1980
Duan 1959
Hoffman and
Wynder 1966
ACGIH 1980
U.S.E EPA 1977
OSHA 1978
NIOSH 1978
-------
TABLE 1 (cont.)
BASIS FOR INHERENT HAZARD SCORE
NON-CARCINOGENS
CARCINOGENS
Compound
MED q^* or I/MED Cancer Type/
mq/kg/day Species/Route Target Organ(s) Primary Effect(s) (mg/kq/day) Target Organ
Reference
Maleic anhydride
Mercury
Methomy1
0.26
human/
inhalation
0.03 (subchronic) monkey/oral
20 (chronic)
rat/oral
Methyl bromide
Methyl chloride 427 (chronic)
Methyl ethyl ketone 595
24.2 (subchronic) rabbit/
inhalation
mouse/
inhalation
Lung
CNS
Spleen
"Typical
poisoning"
Liver
rat/inhalation Teratogenic
Upper respiratory
tract infection
Histopathological
changes of the cal-
carine cortex
Histopathologic
changes
Not specified
Hepatocellular
degeneration
Gross abnormalities,
sternal abnormalities
ACGIH 1980
Sato and Ikuta
1977
Kaplan and
Sherman 1977
U.S. EPA 1980
U.S. EPA 1980
Schwetz et al.
1974
Methylhydrazine
Naphthalene
2-Naphthylamine
0.032 (sub-
chronic)
rat/inhalation Hematopoietic
Methyl parathion 0.5 (subchronic) dog/oral
CNS
1,000 (subchronic) rabbit/oral Eye
Decreased hematocrit,
hemoglobin conc.,
erythrocyte counts
Decreased cholines-
terase activity
Lens and retinal
changes
0.0146
Liver tumor
Darmer and
MacEwen 1973
Williams et al.
1979
Van Hegnigen
and Pirie 1976
Bonser et al.
1952
Nickel
0.5 (subchronic) rat/oral
Reproductive
Increased fetal
mortality, runts,
decreased litter size
Schroeder and
Mitchner 1971
Nitrobenzene
125 (subchronic) rat/injection Teratogenic
Delayed embryogenesis,
altered placentation,
gross fetal abnor-
malities
Kazanina 1968
Nitrogen oxides
0.4
human/ Two times the ambient air quality
inhalation standard was assumed to be the
MED/10.
-------
TABLE 1 (cont.)
BASIS FOR INHERENT HAZARD SCORE
NON-CARCINOGENS
CARCINOGENS
Compound
MED qi* or I/MED
mq/kq/day Species/Route Target Orqan(s) Primary Effect(s) (mg/kq/day)
-1
Cancer Type/
Target Organ
Reference
2-Nitrophenol
4-Nitrophenol
N-Nitrosodimethyl-
amine
N-N i t rosod iphenyl-
amine
PCBs
Parathion
225 (subchronic) rat/oral
23.8 (subchronic) rat/oral
0.4 (chronic)
Phenol
Pyridine
Selenium
Silver
0.006
0.0077 (sub-
chronic)
Pentachlorophenol 30 (subchronic)
14.4
0.5 (chronic)
mice/oral
rat/oral
0.7 (subchronic) rat/oral
human/
inhalation
rat/oral
Lung
Lung
Tumors
monkey/oral Teratogenic
rat/inhalation CNS
63.5 (chronic) rat/oral
Teratogenic
Reproduction
CNS, GI tract,
abdominal dis-
comfort
growth inhibi-
tion
Carciovascular
Increased respiratory
volume
Increased respiratory
volume
Unspecified (no
incidence data)
Reduced body weight
at birth, decreased
weight gain of off-
spring; hyperactivity
Decreased RBC
cholinesterase
activity
Increased incidence
of resorptions,
decreased fetal body
weight and crown-
rump length
Depressed neonatal
growth
Headache, dizziness,
nervousness, insomnia,
mental dullness,
nausea, anorexia
Hypertrophy of the
left ventricle
0.0018
Transitional cell
carcinoma of the
bladder (rat)
Grant 1959
Grant 1959
Clapp and Toya
1970
NCI 1979
Bowman et al.
1981
Edgewood
Arsenal 1975
Schwetz et al.
1974
Heller and
Pursell 1938
Fassett and
Roudabush 1953
NAS 1976
Olcott 1950
Sod ium
14.3
human/oral
Cardiovascular Hypertension
NAS 1977
-------
TABLE 1 (cont.)
BASIS FOR INHERENT HAZARD SCORE
NON-CARCINOGENS
CARCINOGENS
Compound
MED
mg/kg/day
Species/Route Target Organ(s)
Primary Effect(s)
qi* or I/MED
(mg/kg/day)"1
Cancer Type/
Target Organ
Reference
Sulphur dioxide
2,3,7,8-Tet ra-
chlorodibenzo-p-
dioxin
1,1,2,2-Tetra-
chloroethane
Tet rachloroethene
Thallium
7.23 (chronic) dog/inhalation Lung
0.00001 (chronic) rat/oral
0.588 (chronic) rat/inhalation
109 (subchronic) guinea piq/
inhalation
0.45 (subchronic) rat/oral
Liver, kidney,
lung
Hematopoietic,
hypophysis
Increased pulmonary
flow resistance,
decreased lung
compliance
Increased urinary
porphyrin excretion,
liver and lung lesions
Decreased weight
gain, increased WBC
counts, increased ACTH
content of hypophysis
Decreased growth,
decreased liver
weight, increased
liver lipid and
esterified choles-
terol, fatty
degeneration
Death, hair loss
Lewis et al.
1969
Kociba et al.
1978
Schmidt et al.
1972
Rowe et al.
1952
Hanzlik et al.
1928
Thiourea
Toluene
0.0032
Toluene-2,4-dlamine
Toluene diiso-
cyanate
53.2
0.0010
human/
inhalation
human/
inhalation
CNS
Lung
CNS depression
Chronic loss of
pulmonary function
0.18
Epidermoid car-
cinoma of ear
duct and orbit
(rat)
Mammary tumor
(rat)
Rosin and Ungar
1957
van Ottinger
1942
NCI 1979
Wegman et al.
1977
O-Toluidine
hydrochloride
0.0025
Sarcoma (rat)
NCI 1979
-------
TABLE 1 (cont.)
BASIS FOR INHERENT HAZARD SCORE
Compound
NON-CARCINOGENS
CARCINOGENS
MED
mq/kq/day
Species/Route
Target Organ(s) Primary Effect(s)
qj* or I/MED
(mq/kq/day)
-1
Cancer Type/
Target Organ
Reference
Toxaphene
0.062
Hepatocellular NCI 1979
carcinoma (mouse)
1,2,4-Trichloro-
benzene
20.5 (chronic) rat/inhalation Liver
Hepatocytomegaly
Coate et al.
1977
1,1,1-Trichloro-
ethane
48.2
human/ CNS
inhalation
Impairment of psycho-
physiological func-
tions
Gamberale and
Hultengren 1973
1,1,2-Trichloro-
ethane
0.0017
Hepatocellular NCI 1978
carcinoma (mouse)
Trichloroethylene
29.6 (subchronic) rat/inhalation
CNS
Behavioral
alterations
Goldberg et al.
1964
2,4,6-Trichloro-
phenol
Vanadium
Vinyl chloride
0.0071
human/
inhalation
Lung
Mild delayed pul-
monary irritation
0.001
0.04
Hepatocellular NCI 1979
carcinoma (mice)
Zenz and Berg
1967
Mammary adeno- Maltoni 1977
carcinoma (mouse)
Zinc 90 (chronic) rat/oral Adrenal, Hypertrophy of Aughey et al.
pituitary adrenal cortext, 1977
changes in pancreatic
islets and pituitary
gland
------- |