5613
905R81108
REPORT OF THE
TOXICS INTEGRATION PROJECT
Co-chairs: Marilyn Bracken
Roy Gamse
Project Director: Dan Beardslev
September 11, 1981
-------
Envi.-cr.msnt-l detection Agenqg
-------
TABLE OF CONTENTS
Executive Summary
Chapter One: Statement of the Problem
Chapter Two: Managing Toxics Integration
Chapter Three: Chemical, Industry, and Geographic
Work Groups
Chapter Four: Intergovernmental and Regional Issues
Chapter Five: Science
Chapter Six: Contracting
List of Supporting Documentation
-------
EXECUTIVE SUMMARY
The Environmental Protection Agency was assembled from
program elements in other Agencies, in part for the purpose
of creating a fully integrated approach to pollution control.
The objective was to have a single Agency that could analyze
the health and ecological risks of pollutants in a compre-
hensive fashion and regulate in a sensible, consistent,
and efficient manner across all sources and pathways of
concern.
The performance of the Agency to date falls short of
achieving that objective. Today's organization is dominated
by semi-autonomous programs, each built on its own series
of legislative additions and expansions. Each program measures
the health and environmental consequences of pollution with
its own set of legislatively imposed values. Each competes
for scarce resources with a uniquely defined view of priorities
based on this individual perspective. The Agency's agenda
has historically been little more than the sum of many loosely
associated parts.
Attempts to develop a management structure that promotes
centralized, or at least consistent, decision-making have
so far met with only partial success. EPA still lacks an
independent forum for analyzing environmental problems
prior to the point when the programs commit to developing
regulations. It does not conduct comprehensive evalutions
of the overall impacts of specific industrial activities or
products to see what controls, if any, the individual programs
should design. As a consequence, programs make major regulatory
commitments through the routine application of their separate
priorities. The Agency is controlling pollution with a
management structure that fails to maintain a balance of costs
against benefits across program boundaries, and often per-
mits the inadvertant transfer of pollution problems from one
environmental medium to another.
The present Toxics Integration Project (TIP) responds
to a directive from the Office of Management and Budget to
design a more aggressive strategy for integrating the Agency's
control over toxic substances. Toxics-related activities
account for the major part of the Agency's unfinished business,
and represent the most complex management task EPA has yet
attempted. More clearly than any other issue, toxics dramatize
the need for central policy coordination of individual program
priorities: ways must be found to review clearly articulated
policy options programs commit themselves to action. In
addition, the past actions of the Agency must be reviewed for
consistency along the same dimension and remedial action
considered where appropriate.
-------
-11-
Integration requires that the Agency avoid inconsistency
and duplication of effort, avoid letting its programs merely
tranfer problems among environmental media, be consistent in
interpreting ambiguous data, and base its actions on consistent
policy which would guide the balance to be struck between
human health protection and the long-term stability of ecosystems,
Most important, it must set overall priorities such that
they are legitimately defensible both to a worried public
and to industries under financial stress.
With this in mind, since January of this year TIP has
pursued three major strategic approaches for integration: by
chemical, by industry, and by geographic area. All of them
offer potential for making Agency actions more cost-effective,
minimizing the economic burdens it imposes on the private
sector, while protecting public health and the environment.
All of them provide the kind of forum for prior analysis of
problems that has been lacking in the past.
To illustrate just one of these approaches as an example,
through the use of cross-program industry-oriented work
groups, the Agency could for the first time be able to review
simultaneously all the burdens it places on specific economic
subsectors. All environmental and health impacts of the
industry would be evaluated simultaneously. The cumulative
impacts of all existing and proposed regulations would be
calculated. The overall pattern of controls could then be
revised to reflect a centrally defined balance between costs
and benefits. Such an analysis may suggest the occasional
need for revision of existing legislative requirements, if
regulatory changes cannot alone rationalize the Agency's
requirements.
To accomplish this the work groups would present the
Administrator with a range of policy options affecting the
subject industry. One option could be the recommendation to
promulgate no further controls at all. Another might be to
modify or rescind certain existing regulations whose benefits
to health and the environment fail to exceed the costs of
control. Any recommendations for additional controls would
be presented in the full context of the cumulative economic
impact they would have on the industry, and the incremental
public benefits they would yield beyond those already gained
by existing controls. Programs would be instructed to write
additional regulations only if the Administrator concluded,
on the basis of the evidence presented, that significant
uncontrolled risks from the industry in fact exist, and
that the incremental costs of control are justifiable.
The principal recommendations of the TIP report are
summarized below, followed by a summary of the studies that
led to them.
-------
-111-
A. Summary of Major Recommendations
1. The Associate Administrator for Policy and Resource
Management should manage all aspects of the inte^"
gration process, including making recommendations
to the Administrator on Agency priorities affecting
key chemicals and industries.
The Associate Administrator for Policy and Resource Man-
agement should manage the integration process. After consul-
tation with the relevant Assistant Administrators and Regional
Administrators, he should make suitable recommendations to
the Administrator on the Agency's toxics priorities including
establishment of specific chemical and industry (and perhaps
geographic) work groups (see Recommendations 2 and 3.)
In FY 1982 the Agency should set up, as carefully monitored
pilots, three or four chemical work groups, two or three
industry work groups, and perhaps one geographic work group
(whose selection would be made in consultation with the
regional offices and the affected states). These would in-
volve OPRM staff and representatives detailed from various
programs, and they would report directly to the Associate
Administrator for Policy and Resource Management.
The AA for OPRM should have staff support to further
test an analytical approach to assist the priority-setting
function. That approach, if successful, might be useful in
setting priorities across a range of activities not limited
solely to toxics problems, such as trade-offs between attention
to toxic and conventional pollutants, and information
gathering for toxics.
2. EPA should establish cross-Agency chemical work groups.
Chemical work groups would develop consistent Agencywide
strategies for specific chemicals or chemical groups. They
would focus on a limited number of chemicals whose special
toxicity problems, high levels of production or discharge,
and multi-media/multi-industry attributes require such
strategies to ensure Agencywide consistency and efficiency.
Chemical work groups would require a budget, as well as
staff commitments from the relevant program offices and
ORD.
3. SPA should establish cross-Agency industry work groups,
Industry work groups would comprehensively analyze
environmental and economic factors and develop options for
-------
-iv-
regulatory rescission, no action, or additional regulations.
Each work group would review all Agency activities affecting
the subject industry to recommend a consistent Agency strategy
which (1) reduces undue actual or potential cumulative economic
impacts in a way that maximizes resulting health/environment
protection, (2) eliminates duplicative studies, and (3)
resolves inter-media issues. Their authority would be similar
to that of the chemical work groups (with which they would
have to work quite closely). Industry work groups would
also require resources and staff.
4. All hazard assessments used as the basis for restrictive
regulations should be given formal peer review.
Currently several groups conduct peer review on Agency
hazard assessments, while some hazard assessments may be
used without any formal peer review at all. The Science
Advisory Board should be assigned responsibility for reviewing
all health-related hazard assessments used to support new
regulations, or amendments to existing regulations, that
limit the production, use, or disposal of toxic substances,
except for those required by FIFRA to be reviewed by a separate
Science Advisory Panel. The SAB should be given the resources
to accomplish these peer reviews without undue delay. The
Deputy Administrator/Science Advisor should have the respons-
ibility for seeking consistent peer review across all SAB
and SAP groups.
5. SPA needs more centralized oversight of risk
assessment procedures"!
Risk assessments are currently undertaken by various groups
in ORD and the program offices. Some inconsistencies occur.
The Agency should give a central office the necessary authority
to coordinate all risk assessment procedures and review
inconsistencies. Several controversial alternatives, not
recommended at this point, include centralizing hazard assessment
or possibly all risk, benefit, and economic assessment in one
office.
6. The Agency should continue methodological development
to support the chemical, industry, and geographic
approaches to integration.
Over the next year, the Toxics Integration Staff should
complete the Agency strategy for chlorinated solvents and
further refine methodological issues relating to consideration
of substitutes and cross-program cost-effectiveness analysis.
-------
-v-
In addition, further work is necessary to develop the analytic
tools required to refine industry-specific strategies, as
well as strategies for specific regional/local areas-. Primary-
issues for both include further development, testing, and
validation of the models to be used, comprehensive review
of data needs, assessment of costs involved, and analysis of^
other management support needs. For the industry approach,
this includes testing a novel way to analyze health protection
cost-effectiveness issues related to a particular industry
or industry subsector. For the geographic approach, we
need to do additional work to determine if available data
and resources allow the approach we have been testing to be
practically applied to local situations by regional/state/local
organizations with Federal assistance.
7. The Agency should expand the ORDIS contracts tracking
system and develop practical mechanisms for using it
to prevent unnecessary contract duplication.
Several studies revealed obvious duplication and other
inefficiencies in EPA's use of contract resources. Procedural
reforms in this area will assist in better integration and
more efficient use of the Agency's information base. In FY 1982
OPRM should evaluate the feasibility of developing similar track-
ing systems for Agency grants and cooperative agreements.
8. The Agency should merge the staff and contractual
resources devoted to toxics integration which are
now split between the Office of Toxics Integration
in OPTS and the Toxics Integration Project in the
Office of Policy and Resource Management.
The existence of two separate toxics integration staffs
leads to inefficiencies and to a less focused approach to
integration. As the FY 1983 budget reflects, resources can
be conserved (and, most likely, stronger direction provided-)-
with one single toxics integration staff.
3. Managing Priorities for Toxics Integration
Creating a management mechanism for centralized priority-
setting will assist EPA in achieving the goals of addressing
the worst toxics problems first and achieving the maximum
amount of public risk reduction for a given cost.
The chemical, industry, and geographic approaches
developed in this project separately address almost all
basic contingencies for toxics management, at least within
-------
-VI-
the constraints of SPA's mandates. The strategic goal of
the project has been to find an organizationally effective
way to build a network among the three approaches, setting
each in a context that plays to its strengths, but defining
interlocking responsibilities in such a way as to implement
an integrated set of Agency priorities.
Specifically, the chemical approach snpuld be used to
cover problem chemicals with widely dispersed uses, like
solvents and asbestos. It can also be used for very high
hazard compounds like dioxins. The industry approach would
be used to evaluate our approach to major industrial sectors,
such as parts of the metals or chemicals industries. The
geographic approach would be used to assess multi-media
exposure problems in areas where populations may be exposed
to unusual risk and determine what mix of local, state, and
Federal solutions is most appropriate.
OPRM should design the network of these approaches
using a priority-setting process. The Toxics Integration
Staff will test the utility of an analytical approach for
identifying priorities. The scheme would not substitute for
management judgment, but may be helpful in identifying sub-
sequent work group priorities. If the analytical approach
proves useful, it could be applied more broadly to rank
Agency functions not covered by the work groups. For example,
it might be used to rank regulatory activities affecting
toxic and conventional pollutants. It might also prove
useful as an aid to Agency budget decisions.
C. Chemical, Industry., and Geographic Work Groups
This report recommends the use of cross-program work groups,
but with much stronger authority than they have had in the past.
Work groups should be formed as soon as possible for a small
number of priority chemicals and industries and, eventually,
to address pollution problems particular to specific geographic
areas. They must have formal charters, have sufficient resources
and strong leaders, have members who truly represent the AAs or
RAs, and be accountable to the Associate Administrator for Policy
and Resource Management.
Chemical work groups: Looking at toxics on a chemical-
by-chemical basis is the most direct integrating approach.
For any chemical (or group of chemicals) in commerce, or which
might be a byproduct or contaminant of a commercial process,
-------
-vii-
one can analyze its health and environmental hazards in
multiple environmental media and other routes of exposure
(such as in food or consumer products), quantify exposure
patterns of humans and other biota, and arrive at a detailed
understanding of the risk that chemical poses nationwide.
One can then review and select among options for control in
each of the routes that might represent unreasonable levels
of risk, being sure to address the economic impacts of those
controls and the risks attached to chemicals that would be
used as substitutes. The main advantage of the chemical
perspective is its ability to track exposures to a chemical
throughout its "life cycle", including all occupational and
consumer product exposures.
Industry work groups; The direct complement to the
chemical approach is an industry-oriented approach., one
where industry sectors rather than chemicals are the organizing
principle. The clear advantage here is that one can analyze
all effluents from an industry simultaneously, taking into
account cross-media shifts of pollution as controls are
implemented in various media, and looking at the total environ-
mental and health risks of the industry simultaneously.
In addition, the total economic impact of all controls can
be assessed at once, and it becomes practical to conceive of
producing integrated packages of regulations that are inter-
nally consistent economically as well as environmentally.
Geographic work groups; The last principal mode of
integration is geographic. Here one seeks to identify and
diagnose areas where industrial density, population concentration,
and perhaps unusual environmental conditions conspire to
raise local toxics exposures significantly beyond the conditions
on which national regulations are based. By using geographic
analysis, responses can be tailored to local situations,
using Federal, state, and local powers as appropriate.
The purpose of all work groups should be to develop
strategic options for addressing the health and environmental
impacts of the chemical, industry, or geographic area under
review. When completed, the options would encompass specific
program office assignments for regulatory action (or revision,
and they may well include recommendations that no control is
needed). The chemical approach is ready for full implement-
ation by'work groups, and three or four should be started in
FY 1982. The industry approach should be pilot tested on
two or three selected industries in FY 1982, and the results
assessed before this approach is broadly applied. The method-
ology for the geographic approach needs further refinement,
'out at least one pilot work group can probably be launched
in FY 82.
-------
-viii-
Resources for work groups; Resource needs (and their
distribution among the program offices) of the various work
groups will vary according to the chemicals and industries
chosen, but average resource estimates for several model
chemical and industrial work groups and a geographic project
have been estimated. In FY 82, the total contract costs for
an analytic program consisting of two industrial work groups,
three chemical work groups, and one geographic project would
be about $1.5 million? an additional $1.5 million may be needed
for data development for a geographic study, but this money
might be available partially through grant, reprogrammings and
partly through in-kind contributions by the state involved.
The basic $1.5 million would be provided out of the combined
TIP/OTI budget, though support from the individual programs
will also be necessary, particularly in such areas as data
retrieval and data development. Each program would have to
supply approximately three to four staff members on a half-time
basis to participate in the various work groups.
D. Specific Technical Methodologies
EPA should also attempt to bring the best available
analytic tools into effective service for the proposed work
groups. This report describes innovative analytic methodologies
to support the three approaches. The end result of analysis
of this type would be a set of options for solving a particular
pollution problem. The broadest array of choice consistent
with the available information would be offered to Agency
leadership. All these methodologies should be seen as con-
tinually evolving as the science behind them improves. None
is represented here as being more than an attempt to package
the best available data in the most useful possible form?
they are not mechanical systems for defining control needs.
This year's work on work group methodologies has emphasized
analysis of human health effects; next year's work will
expand the methodologies to fully address environmental
effects as well.
The chemical methodology; The heart of the chemical meth-
odology is the materials balance concept, through which major
routes of release and media affected are identified for all
stages of a chemical's production, use, and disposal. This
is coupled with an evaluation of all scientific data available
on that chemical, rationalization of Agencywide efforts to
collect data and do analysis, preparation of a complete risk
assessment profile of all releases of the chemical found to
be significant in the materials balance analysis, and a
complete control options analysis. This analysis reviews
all regulations affecting the chemical not only in EPA, but
-------
-IX-
elsewhere in the Federal government. It also examines risks
associated with expected chemical substitutions under
various options for control. The process is being tested
this year by the Solvents Work Group, which is analyzing
six chlorinated solvents used widely as metal degreasers.
The industry methodology; The Toxics Integration Project
is attempting to develop a scheme that organizes everything
EPA knows about a given industry's emissions and pathways of
exposure, establishes priorities for risk assessment and
control technology development, and tests possible inter-media
regulatory strategies to determine which would achieve the
most satisfactory results for a given level of cost. Thus
far, the Toxics Integration Project has been able to incorporate
risk, exposure, and cost information to compute cost-effectiveness
of each of the alternative control options for the indus-try.
A preliminary test of this approach has been prepared for
chlorinated solvents plants and for copper smelters.
The methodology attempts to consolidate all significant
pollution and cost factors at an initial analytic phase of
the regulatory development process. The scheme is promising
enough to warrant further testing by industry work groups.
Naturally, work groups will choose the methodologies sufficient
to handle their particular subject industries, and in many
cases simpler, less ambitious, and less costly techniques
will be satisfactory.
Though attractive conceptually, the approach developed
this past year is subject to data limitations and other-
analytical shortcomings. For one thing, there is no generally
accepted means of quantifying health risks, so as to compare
them to one another and to the costs of reducing them. The
techniques of quantification used in this study will therefore
be controversial. Although sensitivity analysis has shown
that the cost-effectiveness curves developed during the
pilot studies are insignificantly affected by widely varying
assumptions about health effects, additional sensitivity
analysis is needed to determine whether the relationships
suggested by the studies are reliable. At this point, no
final conclusion can be reached about how useful or how
broadly significant the approach will be.
The geographic methodology; The geographic approach
requires application of a series of pathway analyses similar
to those used for the industry approach. This project has
created a formal system of steps to (1) gather and evaluate
existing toxics data on a given region, and fill in gaps as
necessary, (2) conduct an integrated multi-media pathway
-------
-X-
analysis that tracks a series of key pollutants from all
sources through all environmental routes to all final biological
compartments, and (3) link this analysis to a comprehensive
exposure assessment of all populations locally at risk. The
procedure gathers and arrays the information necessary to
allow eventual Federal/state/local work groups to define
a comprehensive area-wide toxics strategy that utilizes the
most cost-effective regulatory and non-regulatory controls.
Even more than the industry methodology, this methodology
is a conceptually attractive but untested approach. It may
be quite limited in applicability because of heavy data needs
and large expense. But it may also be highly beneficial for
areas where data are available and the magnitude of toxic
problems warrant significant analytical expenditures. These
potential benefits justify a commitment to further refine-
ment of the methodology.
E. General_Scientific Issues _
The task of the project regarding EPA scientific work
was to locate and describe inconsistencies of approach and
duplications of effort that might have adverse effects on
Agency toxics control policy or on the use of its scientific
resources. Changes in the way science policy is made at
SPA, or judgments about the relative importance of various
health or ecological effects, or decisions about how health •
end points are to be construed, were not considered to be
part of this task and are not addressed.
The most important scientific tool used for evaluating
toxics problems is "risk assessment", a three part process
that combines (1) an evaluation of the inherent hazard of a
given compound through a. given route of exposure (air, drinking
water, surface water, food), at various levels, with (2) a
quantification of current exposure levels (prevailing levels
times the population exposed to each level), to produce (3)
an estimate of total risk related to current exposures to
that chemical. Human health risk assessments in EPA have
tended to stress cancer effects, but teratogenic and other
chronic health effects can also be accommodated. EPA also
has considerable experience at estimating environmental risks,
such as to fish, crops, and other biota.
The Agency has some difficulty in achieving consistency
and uniform quality in the risk assessments it produces.
This project analyzed the technical and procedural issues
of risk assessment and makes several recommendations, including
(1) a call for multi-media risk assessment and (2) a proposal
for submitting a number of identified technical and scientific
-------
-XI-
inconsistencies between program offices to a single review
process for resolution. A central policy/review office
should be assigned the responsibility for issuing guidance
resolving the inconsistencies identified in this review,
and it should review future program office risk assessments
to ensure consistent analysis across the Agency. Finally,
the report recommends that all hazard assessments supporting
restrictive regulations be subject to peer review, and that
some attention be given to making these peer reviews consis-tent
to the extent possible.
In the course of evaluating the present system for risk
assessments, TIP considered various options for modifying
the organization structure through which risk assessments
are conducted. The option of centralizing part or all of
the actual risk assessment work of the Agency (or perhaps
the benefits, cost, and economic impact work as well) was
considered and rejected by the Office Directors involved for
several reasons.
First, a centralized group may not be sufficiently
responsive to the individual priorities of program offices,
and the risk assessment function might become a bottleneck
that would prevent programs from meeting their externally
imposed deadlines. Second, centralizing exposure assessment
would offer little additional efficiency, since exposure
analysis is performed differently across the various media
(for example, the meteorological analysis done in the air
program is irrelevant to exposure estimates done by the
water program). Third, there seems little point in centralizing
cost and economic analysis since consistency of these across
programs has not been raised as a major issue. Fourth,
centralization could undermine the Agency's checks and bal-
ances against mistaken risk judgments and could seriously
impair the Agency's ability to bring technically competent
presentations of divergent points of view to the Administrator
for decision. Finally, a substantial amount of disruption
would probably accompany centralization of these functions
and the hundreds of staff who currently perform them.
-------
CHAPTER ONE: THE PROBLEM
This report proposes a strategy for toxics integration—
that is, integrating the responses of EPA's individual
programs to the presence of man-made toxic chemicals in the
environment. It is the result of the first year of a two
year study project begun in accordance with a directive from
the Office of Management and Budget issued as part of the FY
81 budget passback.
Although the workplan for this year, and in consequence,
this report, deals mainly with regulatory issues, it was
recognized from the start that other facets of a complete
integration strategy—problem identification and assessment,
monitoring, enforcement, legislative reform, and alternatives
to regulation, for example—would receive increased attention
next year. Also, this year's work concentrated on headquarters
activities and organization. Rethinking the current division
of labor between EPA, its regional offices, and state and
local governments will be a major priority for the further
development of the strategy during the second year.
A. Previous Integration Efforts
Integrating the national effort to control pollution
was a central part of President Nixon's charge to EPA when
he created it in 1970, but subsequent legislation addressed
to pollution has reinforced the original programmatic organi-
zation of the Agency. Developing an Agency-wide toxics strategy
should therefore be undertaken with an understanding of past
attempts to mesh disparate elements of the Agency.
In particular, in 1977 the Integrated Toxics Strategy
Work Group submitted a report which led to the creation of
an Office of Toxics Integration (OTI) in the Office of Pesti-
cides and Toxic Substances and a Toxic Substances Priority
Committee (TSPC). These arrangements provided the Agency
with a focal point for addressing issues related to toxics,
and both of these organizations have been able to increase
integration within their relatively narrow mandates. Experience
has demonstrated, however, that this narrowness defeats the
purposes of integration, at least in the sense that the term
is being used in this report. For example, the TSPC is
advisory only. It does not dispose of its own budget, nor
does it have any formal control over the priorities of its
members (see Supporting Document 1 for a more complete
discussion of current Agency integration efforts). Similarly,
-------
1-2
OTI is limited largely to work within the purview of the
Toxic Substances Control Act as it has been implemented by
the Office of Pesticides and Toxic Substances. It does not
exercise authority over toxics control or information gather-
ing activities of offices outside of OPTS.
This latest attempt at Integrating toxics policy benefits
from previous experience. TTfye following assumptions are
culled from this experience a'nd represent, in part, the
basis for the present work:
o An integrated toxics strategy needs the active support
and political guidance of the Administrator and the
political leadership of the Agency.
o Toxics integration must be recognized as a priority
for the entire Agency. If it is perceived as the
primary responsibility of an individual program office,
the effort is bound to fail.
o Toxics integration must provide for decision-making
structures that can link policy judgments and decisions
to technical assessments of scientific findings and
economic impacts.
o Technical integration, such as integration of data
bases or various administrative functions, is necessary
but not sufficient. Integration efforts should
focus on Agency policy as well.
o Integration should take place as early as possible
in the analytic process on which Agency decisions
rest. Integrative devices or forums that function at
stages where programs have invested substantial resources
in policy positions tend not to work.
o Previous integration efforts have not given sufficient
attention to the role of the states and of EPA Regional
Offices in dealing with toxic substances.
B. Definition of the Problem
The following section discusses the problem posed by
the nature of toxic substances and how this problem is
aggravated by the organization of the Agency into programs
exercising different mandates. Toxics integration, as we
define it, is unusually difficult because it represents the
intersection of two problems, the toxic effects problem and
the integration problem, each of which is difficult enough
alone.
-------
1-3
The Toxic Effects Problem
This arises from our profound ignorance of the long-term
effects on both people and the natural world of the chemicals
we have released, and continue to release, into the environment.
We do not have the resources, nor will we ever have the
resources, to reliably determine the action of all the chemicals
of concern, under all the conditions where they are encount-
ered, on all the organisms that are vital to us, for a long
enough time. The toxic effects problem has at least the
following six major components:
o Scientific conclusions about toxic effects are often
indefinite.
The chronic toxicology of most chemicals (unlike their
acute toxicology) is not far advanced. Most data come from
short-term, high-concentration bioassays (animal tests),
which are difficult to relate to possible effects of long-term
exposure to low concentrations. Even elucidating the exposure
of organisms to toxics is a costly and difficult business.
We know next to nothing about the synergistic effects of
chemicals in combination, although we would expect that most
actual exposure is to such complex mixtures.
o Many toxic chemicals may have an effect in extremely
low concentrations.
In the past, EPA has worked hard to establish "safe"
levels of pollutants, as various laws demanded. However,
most current scientific opinion has it that certain toxic
substances, mainly carcinogens, have no safe levels; that
is, any amount imposes some risk, however small. Where
statutes oblige us to eliminate such risks, and as our ability
to detect minute amounts improves, we may be under increasing
pressure to impose high control cost burdens on industry to
produce (substantially) very small benefits.
o Many toxic chemicals are useful products.
Much of EPA's work has concentrated on control of the
waste products of technology, or their side effects. Sulfur
dioxide and particulates are examples, and the aim of control
has been to reduce them to some standard presumed safe.
Many toxic substances are, in distinction, useful products.
They are manufactured, processed, used, and disposed of
throughout the economy, and at every stage of this "toxic
cycle" they (and the chemicals they may turn into through
-------
1-4
breakdown or combination) may represent some threat to health
or the environment. A related problem: toxic chemicals
typically retain their toxicity when transferred from one
environmental medium to another via control technology. It
is essential to keep track of toxics shifted in this way to
ensure that control technology really reduces the net amount
of significant exposure.
o There are very large numbers of possibly dangerous
substances.
Scale is an important constituent of the toxic effects
problem. There are around 55,000 chemicals in commercial
use; the number of contaminants, process intermediaries,
and combination/breakdown products is unknown, but undoubtedly
large. The National Institute of Occupational Safety and
Health register lists about 40,000 chemicals as having some
inherent hazard determined by research. But arriving at
a valid estimate of the risk represented by a specific
chemical is tedious and expensive; we have not even begun to
do so for more than a few hundred chemicals. It is therefore
likely that, for the immediate future, unpleasant surprises
will be inevitable.
o Public fear and the resultant publicity have led to
substantial political pressure affecting decisions
having to do with the control of toxics.
Our lack of knowledge and the possibility of catastrophe,
however remote, have been transmitted to the public by the
media, often in an exaggerated way, and public fear has
become an important influence on policy. For example, it is
almost impossible to obtain a new disposal site for hazardous
chemical waste in this country. Many people expect the
Agency to immediately eliminate all chemicals risks, no
matter how small. As this is clearly impossible, the typical
response to fear-generated pressure is to shift the focus of
Agency activities to each new "threat" revealed in the press.
This so-called "chemical of the month" response interferes
with the ordered considerations of relative hazard and cost
that ought to be the foundation of environmental policy-making.
The Integration Problem
EPA is organized primarily along environmental media lines,
and this can produce a failure to integrate its actions as
they relate to problems that cut across other important
dimensions. Thus, we may understand the source of all the
-------
1-5
contaminants in, say, the air, and may be able to judge the
effect of this mixture on people, and may be able to specify
a set of controls that will maintain air at a certain quality.
This is integration with respect to air and is considered
highly desirable. But such integration within the various
programs has necessarily meant that the actions of the Agency
with respect to particular chemicals, or particular industries,
have often been "dis-integrated"; duplicative, uncoordinated,
conflicting, or inconsistent. '
The toxic effects problem and the integration problem
exacerbate one another. It is clear that integration problems
"pervade all parts of the Agency, not just those that dear ~
with what have been called "toxics". The Agency has informally
divided substances of concern into "conventional" pollutants
and "toxics", although all pollutants have, at least potent-
ially, some toxic effect on some form of life. The division
between toxics and conventionals was practical when the
Agency was making its original effort to stop the grossest
and most familiar forms of air and water pollution. At that
time it seemed useful to have a short list of substances
which could be limited by scientifically supported standards
to meet the diverse purposes of the various pieces of legis-
lation. Also, these targets appeared to be of concern in
only, or mainly, one environmental medium. But as more and
more substances came to the attention of the Agency as potential
threats, and as it became clear that efforts to remove substances
from one medium may have caused problems in another, the
division between toxics and "conventionals" has become less
useful in an operational sense.
This is particularly relevant to our regulation of
specific industries. Industry has made substantial investment
in the technology we have required to control conventional
pollutants. The Agency is still requiring improvements in
these controls. If we now require firms to control large
numbers of toxics also, the question of relative importance
will arise, since controlling everything at once is not
always technically and economically feasible. Is adopting ~
technology to control the last increment of some conventional
pollutant worth more than controlling some toxic chemical?
How well do available technologies control conventional
pollutants and toxics simultaneously? We are often hard
pressed to answer questions of this type now, and at some
point the integration efforts of the Agency should involve
conventional pollutants as well.
Meanwhile, the limitations on our ability to integrate our
activities around the control of toxicity, broadly defined,
-------
1-6
have caused serious problems for the Agency. The following are
the most important of these.
o The present approach to toxics control allows
significant duplication of effort..
Because toxic compounds can be dangerous in more than
one medium, they may be of interest to more than one EPA
program office. Particularly in the gathering and analysis
of scientific information, multiple interests can produce
duplication and waste. For example, between 1978 and 1980,
six different offices of the Agency ordered or conducted
nine different studies of ethylene dichloride, which contained
duplicative information, on health effects, production and
use, sources of release, and exposure. Similarly, the
Agency's investigation of arsenic comprises eleven separate
studies. Arsenic release data have been studied on four
different occasions recently, by different offices and
contractors. While it does not necessarily follow that
a single study of a given chemical would be sufficient for
all users, the present situation is clearly inefficient.
o Incompatible results of inquiries into toxic hazard and
toxic'exposure may lead to differing policies toward the
same substance.
Risk assessment as a discipline is imprecise enough to
guarantee that the duplication noted above will often produce
conflicting results. For example, material balances represent
important starting data for chemical risk assessments; five
different materials balances for ethylene dichloride now
exist. There is at present no routine way of reaching or
enforcing a consensus on the risks posed by toxic substances.
To continue with the example of arsenic: the arsenic release
studies conflict to such an extent that the offices using
them as bases for policy would set widely dissimilar priorities
for control actions against the releasing industries. Differing
arsenic risk assessments have also led to the publication of
a Water Quality Criterion that was considerably more stringent
than the level that the National Drinking Water Standard
regarded as safe.
o Control technology may transfer toxic material from one
environmental medium to another.
The division of control responsibilities into environmental
media-based offices has often led to policies that remove
toxics from one medium and put them in another. Examples
are numerous: in cleaning the air, stack gas scrubbers produce
-------
1-7
a toxic sludge, which often must be burned or disposed of on
land; in cleaning the water, sewage treatment systems have
produced vast quantities of sludge, often rich in toxic
materials, which must be disposed of on land; volatile organic
compounds are removed from water by processes that release
them into the air. Though cross-media shifts of residuals
may often be desirable, and might at times be a deliberate
objective of the Agency, the transfers that now occur are
usually haphazard and inadvertant.
o Each program sets priorities for toxics investigation
and control independently.
The combined effects of all the programs' activities may
not lower total risk from all media as much as if the programs
coordinated their efforts. Line offices at EPA are typically
constrained in the way they set priorities, but rarely by
the senior management of the Agency. The statutes are often
strictly written; consent decrees may name specific substances
that must be regulated within strict time limits; and each
program office is under the supervision of Congressional
subcommittees, all of which have their own priorities. Each "
line unit is thus obliged to strike a reasonable balance
among the pressures placed upon it.
But from time to time the different balances so struck
turn out to produce, in combination, unreasonable impacts on
the regulated community. Because priorities are not set
across media, a series of new controls on particular chemicals
or industrial processes may hit industries successively,
increasing total costs and discouraging intelligent planning.
It is hard to see how we could reduce regulatory burden
while fulfilling our responsibilities to protect public
health and the environment, without comparing (whether
analytically or judgmentally) risk among substances and
across media and adjusting Agency budgets and actions accord-
ingly. If we accept that there are large numbers of toxic
substances that may need attention and that the Agency's
budget will be level or declining in the near future, this
problem becomes more urgent than it has been in the past.
o The actions of EPA aimed at controlling toxics, considered
across all media, may cause unnecessary economic problems.
Measures to control toxics are paid for largely by in-
dustry and government, to begin with, and ultimately by the
consumer and taxpayer. Each office that imposes a control
cost on industry is obliged by Agency policy and Executive
-------
1-8
Order, if not by statute, to identify the costs of compliance.
What we do not ordinarily see is the cumulative effect of
controls imposed by all the media offices on a specific
industry.
Thus, some industries or segments of industries may be
severely affected by the combined effects without EPA awareness?
no one has the responsibility for keeping track of all the
straws on the camel as they pile up. Also, although the
costing out of control technology is a well-developed skill,
the analysis of what we buy, in terms of health/environmental
risk reduction, has rarely been performed. We may be imposing
economic hardship for little gain in those terms.
If we assume that, in practice, many industries or firms
can only afford a certain amount for pollution control, in
the absence of a rational system for determining what pollutants
are "worst" and organizing to control them first, we may find
that those industries, firms, or plants are so committed to
the control of less dangerous substances that they cannot
afford to deal with others we have now found to be more
dangerous.
Finally, because industrial plant is not distributed
evenly about the country, there may be circumstances where
nationwide controls on some pollutant or process do not
make sense economically. We do not analyze particular indus-
trial concentrations to determine actual all-media exposures
of populations (and the risks these involve). As a result,
some specific areas may be over-regulated in terms of a
desired reduction of risk, while others may not be regulated
sufficiently.
-------
CHAPTER TWO: MANAGING TOXICS INTEGRATION
A. The structure of the toxics integration strategy
As described in Chapter I, EPA was assembled from separate
Federal program units and expanded by new legislative author-
ities, resulting in an organization oriented towards decisions
by program or by environmental medium. The Agency's structure
does not now sufficiently encourage the kind of coordinated
analysis and action with regard to toxics-related risks
that is necessary. It therefore cannot easily set po-1-lution
control priorities in a carefully reasoned, economically— -
efficient manner across all sources and pathways of concern.
So far/ attempts to move toward a more coordinated management
system have met with only partial success. The Agency needs
a more aggressive strategy with which to integrate the analysis
of toxic problems and the development of options.
An integrating strategy needs integrating concepts—
approaches that subsume all relevant activities within
understandable central ideas. The toxics problem lends
itself to three such approaches: toxic risks can be viewed
as arising from specific chemicals or chemical groups, from
separate industrial sectors, or from individual geographic
areas. Each has its administrative advantages. Each leads
to the development of a new management approach for integrating
toxics. Together these approaches support an umbrella
strategy for integrating all the Agency's toxics-related
activities.
The chemical approach: Looking at toxics on a chemical-
by-chemical basis is the most direct integrating idea. For
any chemical or group of chemicals employed in commerce, or
which might be a byproduct or contaminant of an industrial or
commercial process, one can analyze its health and environ-
mental hazards in multiple environmental media and other
routes of exposure (such as in food or consumer products),
quantify exposure patterns of humans and other biota, and
arrive at a detailed understanding of the risk that chemical
poses nationwide. One can then decide whether any significant
risks exist and, if so, design a strategy to abate those
risks, taking into account available options for control,.
the economic impacts of those controls, and the risks-attached
to chemicals that would be used as substitutes were the
first chemical to be restricted in some way. The main advantage
of the chemical perspective is its ability to track exposures
to a chemical throughout its "life cycle", including all
occupational and consumer product exposures. Its drawback is
-------
II-2
that it does not look at chemicals in their most convenient
regulatory context under most EPA statutes: industry subsectors.
The industry approach; The direct complement to the
chemical approach is therefore an industry-oriented approach.
The clear advantage here is that one can analyze all effluents
from the industry simultaneously, taking into account cross-
media shifts of pollution as controls are enacted in various
media, and looking at the total environmental and health risks
of the industry. In addition, one can address the cumulative
costs faced by the industry for all existing controls, setting
additional controls in the light of industry's ability to pay,
and considering rollbacks of previous controls where necessary.
The drawback of the industry approach is that it does not
adequately handle problems of widely dispersed chemical use,
as the chemical approach can; and neither the industry nor
the chemical approach can treat geographically-specific pockets
of pollution, arising from concentrations of industry or unusual
patterns of chemical use.
The geographic approach; The last principal mode of
integration is therefore that of the geographic perspective.
Here one seeks to identify and diagnose areas where multi-media
industrial density, population concentration, and perhaps
unusual environmental conditions conspire to raise local
toxics exposures significantly beyond the conditions and
assumptions on which national regulations are based. By
using geographic analysis, controls can be enacted locally,
using Federal, state, and local powers as appropriate.
The strategy is a network of the three approaches
The chemical, industrial, and geographic approaches
collectively exhaust all basic contingencies for toxics
management, at least within the constraints of EPA's mandates.
They are weakest in dealing in detail with occupational and
consumer routes of toxics exposures, but they are not in-
consistent with either of these and can be made to mesh with
programs under other Federal agencies, such as those of the
Occupational Safety and Health Administration and the Consumer
Product Safety Commission.
The chemical approach should be used to analyze problems
involving widely dispersed chemicals, like solvents, asbestos,
or PCBs; it can also be used for very high hazard compounds
like dioxin. The industry approach should be used to evaluate
program regulations for major industrial sectors for multimedia
effectiveness and economic efficiency. The geographic approach
-------
II-3
can be used to analyze areas where populations are at unusual
risk in dense and heterogeneous industrial areas and to pre-
scribe appropriate actions integrated vertically through all
levels of government, with maximum emphasis on locally-oriented
measures for problems that are not national in scope.
The strategic goal of this project's first year (1981)
has been to find an organizationally effective way to build
a network to link the three, setting each in a context that plays
to its strengths, but defining interlocking responsibilities
in such a way that they support an integrated set of Agency
priorities. One approach for accomplishing this would be to
reorganize the Agency along chemical and industrial lines,
supported by resources now allocated to the current program
offices. The regional offices are suitable for carrying out
the geographic approach, but would need substantial internal
rearrangements to be run exclusively this way. But a
reorganization of this scale would create many more problems
than it would solve.
A more realistic and practical option is to establish
cross-Agency work groups with sufficient authority and
resources to ensure that their goals are met. The Agency
has long used work groups to solve problems of inter-program
significance. Work groups resemble what industry calls matrix
management, but-their history here has been disappointing.
They have seldom had clear objectives or schedules, and
usually lack appropriate resources and analytic support.
Membership selection and internal discipline are typically
loose, and access of the groups to senior program directors
has been inadequate. Yet the concept itself is a good one
if the problems can be overcome.
This report recommends a pilot application of the_
work group concept along each of the three integratingnhbdes"
described above. At headquarters there should be chemical
and industry work groups charged with designing comprehensive
multimedia strategies for each problem assigned. In the
regions there should be geographic work groups composed of
teams of EPA, state, and local staffs to analyze complex
problem areas develop options for both EPA and the states
pertaining to local toxics risks.
To be effective, work groups must become a central, rather
than an ancillary, part of each program's operations. They
must have formal charters, their own budgets, strict membership
criteria, and accountable management. Organizing these
integrating groups, selecting their targets for action, provid-
ing them with adequate resources and proper scientific support,
and managing their performance—these are the central
toxics integration responsibilities. The remainder of this
chapter is devoted to a discussion of these issues. Ultimately,
-------
II-4
the approaches applied to toxics integration, if successful,
could be extended to integration and priority setting across
a broader range of Agency activities.
B. Managing the strategy
We recommend that the Administrator designate the Associate
Administrator for Policy and Resource Management as the official
responsible for coordinating all matters relating to toxics
integration. He would be responsible, in consultation with
the relevent Assistant Administrators and Regional Administrators,
for recommending priorities for work groups to the Administrator
and for establishing work groups in response to her decisions. He
would assign each work group an allocation of staff and a
budget, and he would determine schedules and milestones.
He would be responsible for recommending Agency strategies
based on the work groups' suggestions to the Administrator,
after consultation with the relavent Assistant Administrators
and Regional Administrators.
Staff to "support these functions would be assembled in the
Office of Policy and Resource Management through a merger
of two existing groups: the Toxics Integration Project, set
up in reponse to OMB requirements in the FY 82 passback, and
the Office of Toxics Integration within the Office of Pesticides
and Toxic Substances.
The major responsibilities of the AA for OPRM and the
supporting Toxics Integration Staff are described below. A
detailed description of work group functions, organized by
type of group, is presented in Chapter III.
1. Targeting problems to work groups
As mentioned above, chemical work groups will focus on
ubiquitous compounds associated with significant risks through
multiple media. They will emphasize highly toxic compounds
not closely associated with individual industrial sectors.
Classic examples of such chemicals are PCBs, asbestos, organic
solvents, and chlorinated fluorocarbons. All have strong
potential adverse effects; several have been designated as
carcinogens. Their releases are distributed widely throughout
the environment. Regulatory response is therefore likely to
focus (as it has in contemplating action, if needed, on
chlorinated fluorocarbons) on limitations on manufacture,
distribution, and use, not on industry-oriented effluent
-------
II-5
limitations. In general, criteria for candidates for chemical
work groups include:
o evidence that the chemical or group of chemicals is
of high inherent hazard in multiple environmental
media
o high volumes of production, with substantial
releases nationwide (multi-media releases would
be considered especially important)
o patterns of releases widely dispersed, not closely
associated with a small number of industrial
subsectors or processes
By contrast, industry work groups will be formed around
industrial subsectors that pose unusual toxics control problems
on their own. The priority-setting process should emphasize
work on those industries with complex multi-media impacts from a
mix of toxic emissions, especially those where the uncoordinated
application of single media controls would result in excess risks
through cross-media transfer. Of equal importance will be consi-
deration of cumulative economic impacts. In general, criteria
for selecting candidate industry work groups include:
o evidence that the industry is associated with toxics
exposures for multiple chemicals in multiple media
o evidence that the industry already faces a heavy
compliance burden for environmental controls, and
may in fact not be able to comply with existing
and/or future requirements in all media
o likelihood of adverse impacts from inadvertent
cross-media shifts of pollution due to uncoordinated
application of controls on separate effluent streams
Priority setting for the geographic approach is essen-
tially site selection, and here direct state involvement is
probably necessary. To identify candidate sites for which
EPA and the states might agree to set up work groups, several
criteria can be used. These include:
o state and local interest
o high concentrations of toxics-related industrial
discharges, especially from plants defined as
"major" under the various national permitting
programs
-------
II-6
o evidence of above average levels of exposures in
multiple environmental media, as shown by ambient
environmental monitoring
o evidence of actual adverse health effects relatable
to environmental exposures to toxics, or (a more
practical indicator) evidence of human uptake of
toxics into body fluids and tissues as shown by
direct sampling
o history of toxics incidents, such as spills or
other emergencies
o availability of appropriate data
2. Review and approval of strategies developed by
work groups ~__
The goal of each work group is to prepare a comprehensive
analysis that identifies the most cost-effective opportunities
for reducing health and environmental effects connected with
the target chemical, industry, or geographic area. These
strategies will consider all relevant routes of exposure and
media of discharges to determine, in the first place, whether
any action at all is required. If action may be warranted,
they will prepare for the Administrator a set of options
covering all feasible levels of available environmental
controls, focusing wherever possible on mechanisms other
than conventional command and control regulation.
The Administrator can select from these options the
elements of complete strategy, consisting of a series of
actions to be carried out by the program offices, regions,
and ORD. Before presenting the recommendations of the work
group to the Administrator, however, the Associate Admin-
istrator for OPRM will review and evaluate all the strategy
elements together as a package, and confer with the affected
Assistant Administrators and Regional Administrators, The
Administrator will resolve any strategic policy disputes
that the work group itself cannot resolve, or which cannot
be resolved among the Assistant Administrators.
While each work group is in progress, any rule under
development by a program office that affects the work group's
subject area will require a specific sign-off by the work group
itself. Other regulatory activity will proceed unaffected.
If there is a dispute as to whether or not a given rule
should go ahead, the Associate Administrator for OPRM would
refer the issue to the Administrator for resolution.
-------
II-7
The presumption here is that the Agency would not issue
any rules within the purview of the work group until its
analytic work is completed and the Administrator has made
her decisions on the options presented. There will, however,
be instances where that presumption should be set aside:
exceptions would be made to allow EPA to comply with court
orders and statutory mandates or to respond to imminent
hazards. Exceptions would also be appropriate in cases
where there is little danger of inconsistency between the
rule and the ultimate strategy.
3. Supervising strategy implementation
At this stage, following approval of a strategy by the
Administrator, EPA would implement the strategy recommendations.
The Associate Administrator for OPRM would issue necessary
guidance and directives to the appropriate offices to:
o develop and promulgate chemical-specific and
industry-specific regulationsy
o modify or rescind existing regulations as
required by the strategy;
o develop and implement non-regulatory control
options (such as market incentives) required by
the strategy; and
o develop and implement the intergovernmental
components of the strategy (such as revisions or
additions to State/EPA agreements).
Through the Toxics Integration Staff, the Associate
Administrator for OPRM would supervise the implementation of
the strategy components, assuring that necessary adjustments
are made to the Agency budget and accountability systems.
4. Monitoring and evaluation of strategies
The Associate Administrator, through the Toxics Integration
Staff, would then monitor the implementation of the strategy
components, including the development and promulgation of
regulations and other actions. In addition, the Office of
Policy and Resource Management would assess the implementation
of the strategy and its components, to determine:
o whether in practice the strategy actually results
in the attainment of specified health and environ-
mental goals, and
-------
II-8
o whether the actual economic burdens imposed by the
strategy on industry and government are as
anticipated by the work group.
5. Assigning multi-media risk assessments in support
of work groups and resolving related issues
A major function of the Associate Administrator for OPRM
will be to review and approve recommendations on priorities
for multi-media risk assessments, as presented by the Toxics
Integration Staff. A fuller discussion of these issues is
contained in Chapter V.
C. Setting Agencywide priorities for toxics integration
The Associate Administrator for OPRM will have the
responsibility for selecting targets for chemical and
industry work groups. In the short term he will have to
accomplish this, after discussion with the other affected
offices, based on the criteria listed above (pp. II 5-6).
In the longer term, the Toxics Integration Staff should test
a more objective and analytically defensible means by which
to set these priorities..
In a regulatory Agency as complex as EPA, where the
steps to accomplish products (e.g., regulations) are so numerous,
internally complex, expensive, and time consuming, priority
setting itself becomes a complex process. Insofar as is
reasonably achievable, priority setting should organize
work such that the Agency's actions achieve the greatest pos-
sible degree of environmental and health-related risk reduction
for a given cost. Generally, the concept of cost involves
both internal cost (cost, to the Agency or to government at
all levels) and external cost (the compliance cost to industry
and the total economic effect throughout the economy of all
control actions). Total costs can rarely be optimized, but
such optimization should be a goal. Where toxics are concerned,
the priority-setting process works by combining estimates of
two parameters of a given problem:
1. What is the risk associated with the problem?
2. What is the cost of reducing the risk?
What priority setting should look for in relation to toxics
is the ability to rank problems by these two factors, such that
the Agency controls problems with the maximum possible public
payoff in terms of risk reduction per unit of cost. For example,
-------
II-9
one might not place a high priority on regulating an industry
whose wastes are intrinsically most toxic, because (1) that
industry's wastes might be low in total tonnage, (2) populations
or environmentally sensitive areas might generally be located
far from the industry's plants, or (3) there might be no prac-
tical technical way to reduce emissions very much without
putting the industry out of business. In such a situation
it might be more effective to put more restrictive controls
on an industry whose wastes are perhaps intrinsically less
toxic, but which has a higher volume of waste or a larger
exposed population, or can afford to abate a great deal more
of its effluent at lower cost than the first industry.
In order to make the concept of priority setting work,
the Toxics Integration Staff will have to develop analytic-
approaches that are acceptable to the rest of the Agency as
fair, unbiased, and reasonable in their conclusions. There
is no foreseeable way to make them entirely mechanical and
"objective." The best that can be done is to organize exist-
ing information in the most useful possible way for analysis
and evaluation. Recommendations made by the Toxics Integ-
ration Staff on which of the candidates for work groups should .
be highest priority would be reviewed by the Associate
Administrator for OPRM, in consultation with the Assistant
Administrators and Regional Administrators, and sent by
him for action by the Administrator.
An approach to setting priorities
The OPTS Office of Toxics Integration has done consider-
able work in the area of designing a generally applicable
approach to setting Agencywide priorities. OTI has outlined
an approach that should be tested in the coming fiscal year,
and its results considered for use in selecting future candidates
for work groups.
Step 1; General policy guidance from senior management:
The senior management of the Agency would have to provide
guidance on criteria to be used for ranking the relative
importance of toxics problems. To be most helpful, that
guidance should give some indication to senior management of
the relative importance of ecological versus health values,
cancer versus other diseases, reversible versus irreversible
environmental effects, etc. Understandably, it is difficult
to be precise in providing such general guidance, but the
more detailed the guidance the easier it will be for staff
to reflect top management priorities in its recommendations.
-------
11-10
Step 2; Chemical Identification: The least difficult
and currently most information rich step in dealing with toxics
is to identify major chemicals of concern. Given that there
are over one million known organic compounds and over 55,000
chemicals in commerce, a short cut is needed to screen down
to a more manageable number. This proposal uses a subset of
chemicals (about 130) that were designated by the Agency's
program offices in late 1980 to develop the list of Intermedia
Priority Pollutants (discussed later in this chapter). These
130 chemicals are the ones on which EPA is expending the bulk
of its investigative resources. This assumes that these
chemicals are the ones about which the programs should be
most concerned. This list would be continually modified as
other major chemicals of concern are identified, whether by
new scientific studies or field toxics incidents. The list
will be circulated to the program offices with references to
what Agency reports are available on such chemicals.
Step 3; Selecting chemicals of highest concern and
assocTated use/industry categoriesTTo ensure wide-ranging
screening of chemicals for their hazardous properties, regard-
less of release to the environment, some evidence of biological
activity is needed. For purposes of early evaluation, literature
citations contained in automated data systems might be consulted.
The staff would consider categories of health studies conducted
(acute or chronic bioassay versus epidemiological studies;
types of effects studied, such as carcinogenicity versus
other health end points), the exposure routes studied (occup-
ational, ambient environmental through air or drinking water,
food or consumer product use), the length of time the chemical
has been studied and the currency of the most recent studies,
and the availability of studies on non-human effects. Even
the total number of studies available could give some indication
of medical interest in the hazards presented. Chemicals
could then be grouped into bands representing levels of
apparent "intrinsic hazard." Industry categories (based on
use or production of the chemicals) could then be associated
with chemicals on the top quarter of the list.
Step 4; Identification and screening of major industry
and chemical use categories: In parallel with step 3, EPA
and other (i.e., Census) source, emissions, and exposure
documents are used to identify the major exposure routes for
all chemicals listed in steps 2 and 3 (with particular atten-
tion to the top half of the list). Total chemical releases
are calculated for the major chemical uses (e.g., degreasing)
and production industries (e.g., steel). The major potential
-------
11-11
problem sectors are then ranked for more resource-intensive
screening (e.g., the top quarter of this list).
Step 5; Detailed screening: As a result of steps 2 and
3, a list of 20-40 chemical uses and industry categories exists
based on hazard concerns. Similarly, a list of 20-40 uses/
categories results from the step 4 examination of releases.
These two are examined together to obtain a consolidated list
that covers the principal problems of each, presuming there
will be some overlap. At this point, a more detailed analysis
is conducted that further utilizes the policy guidance provided
by senior Agency management. As suggested in Supporting
Document 2, this step may be conducted as a weighting of
information about these policy concerns followed by scoring
and a ranking of uses and categories, or it can be conducted
in an unweighted, multiple listing format. The factors to
be considered include: quality of toxicity data, detail of
exposure/release data (e.g., geographic concentrations,
worker exposure as it relates to total release), degree of
current regulation, and economic burden information. In
either case, since the policy guidance will probably be
flexible rather than precise, several scenarios of chemical
use and industry categories will likely be provided for the
Associate Administrator's review.
Use of this priority-setting approach
This analytical approach to priority-setting is conceptually
attractive as a means of translating technical and policy
criteria into priorities for strategy development and action.
The approach must be tried to see if it can indeed be useful
to the Agency. Assuming that it does prove useful, might
then be expanded to cover a broader range of Agency priority-
setting, such as among toxic and conventional pollutants.
Ultimately, if the approach is found to be a useful way to
reflect senior management values in Agency decisions, it
could be applied to the Agency budget/resource planning
processes to help set Agencywide priorities in that context.
-------
CHAPTER THREE: CHEMICAL, INDUSTRY, AND GEOGRAPHIC WORK GROUPS
This chapter describes in detail the role of work groups
in analyzing cross-program toxics problems, using chemical,
industry, and geographic approaches. These work groups_would be
established by the Administrator on recommendations from the
Associate Administrator for OPRM, as described in the previous
chapter.
We recommend establishing' a small number of work groups
in the coming year—perhaps three or four chemical work
groups and two or three industry work groups—as a test of
this concept. One geographic project might also be started
in about six months.
A. Work Group Roles and Structure
Work groups would have responsibilities for the. follow-
ing functions:
o Strategic Option Development: Work groups analyze
toxics problems, and where necessary would develop
the Agencywide comprehensive strategic options for the
chemicals, industries, and geographic areas of concern.
o Data-Gathering and Analysis; Work groups should
gather data and perform intensive analyses needed
to develop strategic options for the chemicals and
industries of concern, for example, doing materials
balances or relating risk reduction to control costs,
or calculating cumulative economic burden.
o Strategy Support and Tracking; While the individual
program offices would have responsibility for putting
the strategy into action, work groups would be
responsible for tracking the development or relax-
ation of regulatory and non-regulatory controls by the
program offices, and reporting any deviations from the
comprehensive strategy to OPRM for corrective action.
Effective accomplishment of these tasks demands work
group management of a much stronger type than EPA has required
in the past. We propose that an individual senior manager
be assigned the management of each chemical and industry
work group. No individual should be the manager of more
-------
III-2
than one work group, and a single work group should never
have more than one manager. Work group managers should be
recruited from the Branch Chief or Senior Project Manager
level, or above. In some cases it may be appropriate to
relieve work group managers of their normal line management
responsibilities and allow them to devote all their time to
work group management. Naturally, arrangement will have to be
made to ensure that the removal of these personnel does not
adversely affect normal Agency operations/ but this should
not be a major problem, as there will be few work groups in
operation at any one time.
The work groups should consist both of Toxic Integration
staff members who report directly to the work group manager
and of members drawn from the relevant program and functional
offices. The size of the two contingents will vary, as will
the size of the entire work group, but the core staff size
should not normally represent less than one-fourth of the
FTEs devoted to a work group. TIP explicitly addressed the
problems sometimes associated with cross-program work groups;
the solutions proposed for major problems are shown in Table
III-l. For example, we propose that each work group should
have these features:
o a manager who reports directly to the Associate
Administrator "?o"r OPRM the duration of the work group
o a formal detailed charter, with a scheduling and
planning process under the control of the work group
manager
The charter must be: an expression of Agency policy, pre-
pared by the Toxics Integration Staff and approved by the
Associate Administrator for Policy, Resource, and Management,
with appropriate consultation,
o its own budget, with substantial authority over
resources vested in the work group manager
Each work group should be allocated a share of the
toxics integration budget.
o strict membership and performance criteria
Members must be senior enough to represent the manage-
ment of programs. Work group managers should be allowed
to refuse individuals proposed as members, or dismiss them
for poor performance, and should also have the authority
to recommend the award of cash bonuses for outstanding
wo r k.
-------
III-3
o the authority to sign off on all regulatory activity
within its narrow purview
This is necessary to ensure consistency with the ultimate
strategy. Regulations not related to the subject chemical or
industry of the work group would obviously not require any
such sign-off.
The life span of work groups will, of course, vary
according to the difficulty of their tasks, but based on this
year's experience we should expect each to represent a one
to three year project. Typically, the work group would
complete its analytic work during this time and present its
strategy options to the Associate Administrator for OPRM and
all affected Assistant Administrators. The AA for OPRM
would then make a recommendation to the Administrator on
what course to follow. Program staff would then put the
approved recommendations into effect in the form of guidance,
regulations, etc. Responsibility for tracking implementation
of work group recommendations would reside in the OPRM Toxics
Integration Staff. Chemical work groups will probably terminate
shortly after decision is reached on their recommendations,
while industry groups would be revived from time to time as
particular industries required increased attention.
Resource requirements (and possible distributions of
resource taps on programs) for several model chemical and
industrial work groups and a geographic project have been
estimated. The total contract costs for an analytic
program consisting of, for example, two industrial work
groups, three chemical work groups, and one geographic project
would be about $1.5 million? an additional $1.5 million
may be needed for data development for a geographic study,
but this money might be available partially through grant
reprogrammings and contributions from the states. Each
program AA would have to supply on average approximately
three work years to support this level of work. Specific
program office contributions will vary with the particular
industries or chemicals chosen. A summary of costs for each
type of work group will be found after the more detailed
discussions of each approach presented below.
B. Chemical Work Groups
Purpose
The objective of the chemical approach is to develop an
Agency strategy for a chemical or class of chemicals that may
threaten human health or the environment in more than one
environmental medium. Such a strategy would indicate which
specific releases of a particular chemical represent the
greatest threat, and would recommend cost-effective controls .
-------
III-4
on such releases. In this approach, a variety of cross-media
analyses—of chemical release, exposures, health and environmental
hazards, risk, control technologies, and related control
costs—are integrated to produce a set of options from which
such a strategy could be drawn.
Products
Each chemical work group will be responsible for:
o proposing a strategy that includes all major
research, data gathering, regulatory, and non-
regulatory actions the Agency should take regarding
the chemical and indicates the offices responsible.
These should eliminate any overlapping data
collection or analysis requirements and suggest
how to fill any voids.
o certifying that proposed or pending rules
concerning the chemical are consistent with the
strategy (even if it is still under development)
or otherwise appropriate for promulgation.
Rules that are inconsistent should not be issued
unless they are required by statute or court
order, or have de minimis effect.
What chemical work groups would do
The chemical approach examines the potential threats
posed by a group of chemicals throughout their life cycle
from production to disposal. At each stage in this cycle
chemicals can be released to the environment, resulting
in some exposure to man and the biosphere. (See illustration,
next page) These exposures can in turn result in adverse
health and environmental effects. The chemical approach
evaluates the risk posed at each stage in the cycle; the
result tells the Agency which are the worst risks—in
terms of industrial use categories or chemical release
points. Then, using this information and economic analysis,
the approach can determine whether any stages in the cycle
require EPA intervention, and, if so, which problems
should be dealt with first, and at which stage in the
cycle it is most cost-effective to do so.
(Much of this method is based on the experience of the pilot
organic solvents work group, whose findings are covered in
greater detail later in this discussion.)
-------
PROCESSING-
i
TRI
I
DISTRIBUTION-
JSE-
DISPOSAL-
The Toxics "Cycle*
PRODUCTION > RELEASE »>» EXPOSURE >»» HEALTH EFFECT
-------
III-5
Step 1 - Identification, review, and critique of agency
scientific support documents; baseline docu-
ments developed"
The first step of the chemical work group is to identify
and gather the Agency's scientific assessment data on
these chemicals (materials balances, exposure, health and
environmental effects studies, risk assessments). The
solvents work group found that many reports are not published,
and key Agency experts are often the best guides for judging
the value of the data. Generally, Agency support documents
are developed under contract for each major program office.
Different contractors may use different primary sources,
monitoring, or modelling data to come to contradictory
conclusions regarding major exposures or health effects.
As long as program offices act independently, such dis-
crepancies will remain. (See Chapter 6 for recommendations
aimed at addressing this problem).
The chemical work group should prepare baseline documents
on the health effects, exposures, and risks from each chemical.
Where plans for further data collection or research overlap
or leave important gaps, they should be resolved or~trrought~
to the attention of OPRM and the relevent other offices for
resolution, in some cases through assignment to the appropriate
program office. Field monitoring may be necessary to fill
in the gaps on exposure pathways. . - ~ —
Step 2 - Determine overall risk
Work group members will combine the information on the
hazard demonstrated by the chemicals of concern with data
derived from materials balances, and with estimates of the
various population groups exposed. From this they can determine
the amount of the chemical to which people would be exposed
via all environmental media, and the risk to health this
represents. This summary document provides the scientific
basis for comparing the risks resulting from each chemical's
major uses.
Step 3 - Develop a control options analysis and recommend
a strategy
The cross-media risk assessments then must be reviewed
in light of economic and technological constraints. The
work group will examine available control technologies capable
of dealing with the significant exposures revealed in the
cross-media risk analysis. The group must determine the capital
-------
III-6
and operating costs of these technologies, and relate the costs
to various levels of risk reduction. The most economic pattern
of controls in all media that reduces total "full-cycle" risk
to an acceptable level can then be identified and incorporated
in a strategy. Inter-media transfers, where unavoidable,
could also be examined to make sure that risk is actually
reduced by the transfer. Potential impact of chemical controls
on the use of chemical substitutes will also be assessed to
ensure that such control will not result in the use of even
more hazardous chemicals.
Review of all existing and proposed Federal regulations
(EPA, OSHA, C?SC, etc.) is critical to the development of
the Control Options Analysis. This review provides an overall
perspective on the regulatory framework that has been imposed
on the use of these chemicals. Regulations are reviewed for
possible inconsistencies (e.g., are standards set by DOT
compatible with EPA's?); potential intermedia transfers (e.g.,
do water pollution controls merely transfer the pollutants
to the air?); for regulatory duplication (e.g., are two
programs or agencies regulating the same emission?); and for
major gaps (e.g., is a major exposure pathway uncontrolled?).
Finally, industrial users of the chemicals will be asked
to identify their problems with existing and proposed EPA
regulations. For example, the solvents work group, with
assistance from the Halogenated Cleaning Solvents Association,
has visited several local plants that reclaim solvents,
produce urethane foam, distribute chemicals, etc., to determine
the effects of EPA actions. This EPA-industry dialogue has
been helpful in identifying problems, such as potential
EPA/OSHA regulatory conflicts on venting versus controlling
volatile organics emissions. Such problems will be addressed
in the final strategy.
Step 4 - Present control options to the AA for OPRM
The final control options analysis summarizes the results
of all the above steps into one brief document for OPRM.
All major actions that might need to be taken to deal with
these chemicals are listed as options, and recommendations
are made. Such options would include: asking one or more
program offices to work on regulations or alternative controls
for specific emissions; referring some or all action to the
States or other Federal agencies; conducting additional,
focused research, monitoring, or risk assessment; and revising
existing or proposed rules to be consistent with the strategy.
-------
III-7
The Associate Administrator for OPRM, with appropriate
consultation, will present the Administrator with a set of
options for dealing with the chemical. Should the Administ-
rator decide that the strategy for EPA regarding the chemicals
involves multi-program (and possibly inter-Agency) action, the
OPPM Toxics Integration Staff will monitor the actions approved
by the Administrator. Program offices will still be responsible
for developing regulations, gathering and analyzing data,t:
conducting monitoring/ etc.
The Solvents Work Group—a Case Study
In FY 1981, the Agency undertook the first prototype of
an Agencywide toxics strategy work group. The solvents work
group, dealing with six halogenated organic solvents used as
metal degreasers, will be finished with its initial analysis
in December 1981. It represents the first effort to develop
a cross-media analytic and control strategy for a class of
chemicals and its major industrial uses. Although the work
group experienced procedural problems, the pilot effort
suggests that with increases in authority and resource control,
as proposed earlier, chemical work groups can demonstrably
integrate Agency toxics activities. The Office of Toxics
Integration is evaluating the progress of the work group to
(1) assess its strengths and weaknesses; and (2) refine the
chemical work group approach methodology discussed here so
that it can be applied more effectively next year. However,
the solvents effort has already had the following results:
o The solvents work group's initial data gathering
showed that air is the major source of and fate
for these volatile chemicals (Table III-2). Health
and environmental effects data, now being developed,
will determine whether the air exposure route also
presents the greatest risk. Table III-3 shows that
the use of solvents in metal degreasing operations
may be of greater concern than their use as inter-
mediates, since all six solvents are used in this
way. Finally, Table III-4, which displays all
industrial uses with the media they affect, indicates,.
that solvents production is of minimal concern,
while use in degreasing, paint removers, and dry
cleaning results in the greatest releases.
o The work group examined the existing data on these
chemicals to assess the need for Agency controls.
In doing so, it discovered areas where further
-------
TABLE III-2
RELEASE OF TSPC SOLVENTS kkg (PERCENT)
»J
AIR _ LAND WATER
1,1,1-trichloroethane 213,327 (88) 26,340 (10) 2,305 (2)
tetrachloroethene 142,000 (64) 87,400 (35) 900 (1)
dichloromethane 157,200 (83) 17,700 ( 9) 15,033 (8)
tetrachloromethane 19,253 (96) 1,207 (4) ?
trichloroethene 103,100 (93) 7,500 (6) 131 (1)
SOURCE: EPA/OTS Materials Balances
-------
en
o>
I -u
t-l ID
0) -H
•fJ "O
C 0)
M £
cn
E-
z
U
o
cn
c
&
CQ
fi-
fe
C
CO
M
CO
ID
09
(U
•O
• C
a ~i-t
-------
TABLE III-4. RELEASE OF TSPC SOLVENTS TO THE ENVIRONMENT
AIR LAND WATER
1, 1, 1-trichloroethane
production 1/771 1 1
metal degreasing 159,500 22,200 975
aerosols 18,100 1,980 ----
adhesives 1,37400 919 329
textiles 2,920 175 194
paints 4,950 100 1
inks 2,780 57 ----
-misc. 7,500 155 997
Total 213,327 26,340 2,305
Tetrachloroethene
dry cleaning 93,000 72,000 ----
metal degreasing 39,700 6,700 ----
vapor degreasing 30,000 5,000 ----
cold degreasing 9,700 1,700 ----
__ grain fumigant 9, 500 _ ? - ---
Total 142,200 87,400 900
Dichloromethane
paint remover 51,000 7,000 33
metal degreasing 44,000 4,900 15,000
aerosols 41,000 5,000 ?
20, OOP — — — — -
Total 157,200 17,700 15,033
Tetrachloromethane
production 4,453 1,188 3
intermediate 485 5 ?
metal degreasing 700 13 ?
fumigant 12,092 66 ?
other 1,523 _ ? _ ?
Total 19,253 1,207 3
Trichloroethene
production 725 2 ?
metal degreasing 98,300 7,180 123
paints " 1,260 245 3
textiles 119 136 ?
adhesives 2,484 _ 68 _ ?
Total 103,100 7,500 131
Trichlorotrif luorethane
refrigerant 1,181* 3 ----
solvent 30,454 609 ----
other 4,318 _ 86 ----
Total 35,953 6~9S~ ----
*Assumes 2% of production
SOURCE: OTS Materials Balance
-------
III-8
research is needed, where data are lacking, or
where existing data are not conclusive. For example,
the work group is examining the need for more
health effects research on these chemicals for popula-
tions exposed to low level concentrations through
inhalation or ingestion. Such recommendations
will help focus research, monitoring, and other
Agertjty information-gathering efforts on the
major toxics needs.
o At least three offices (OAQPS, ODW, and OWRS) are
actively developing solvents-related regulations.
As the result of the work group, these and other
involved offices have had to approach solvents
from a multi-media environmental perspective,
taking greater consideration of each regulation's
impact on other media. The work group has identified
potential intermedia transfers of toxics as a
result of EPA control (e.g., aeration is a recom-
mended treatment technology for removing solvents
from water, perhaps creating problems in the air)
and will attempt to reconcile these in the overall
strategy. The work group has also identified some
unexplained inconsistencies between programs (e.g.,
differences between ODW's Suggested No Adverse
Response Levels and Water Quality Criteria) and
some apparently extreme divergence among the control
technology costs imposed on various industries.
Implications
The solvents work group pilot methodology, with modifica-
tions to allow it more staff availability, resources, and
more impact on budgetary, planning, and regulation decisions,
could be a major improvement on the status quo. When the
solvents strategy is completed in December, a refined version
of the same methodology could be applied to several chemical
work groups in FY 82. The work group spent $250,000 and
required 4 workyears over nine months to do the assessments
and technical analyses not already available. The solvents
studies were already in EPA's "pipeline," and so these costs
may be representative only of work groups on well-studied
chemical classes. Work groups on chemicals that have not
been extensively assessed may be much more expensive and
take a good deal longer.
At the same time, it is important to note several limita-
tions of the chemical methodology which should be addressed
in FY 82.
-------
III-9
Significant data gaps, particularly those related to
exposure routes, chronic/subchronic health effects, and
environmental effects of chemicals of concern, need to be
quickly identified early in the work group process. Then,
an operational mechanism or the guidance needs to be developed
for determining: (1) the required level of data needed for
Agency decision-making; and (2) the most appropriate tool
(ORD, existing contract, or legal information-gathering
authority) for getting the necessary data to expedite Agency
and work group analysis/decision-making. The decision on
the minimum data needed for decision-making involves such
issues as the utility of exposure modeling versus field
monitoring and the form and quality of health effects data
needed for Agency peer review.
Another methodological issue is the extent to which
chemical substitutes for work group chemicals should be
evaluated. Work groups must consider what effect EPA actions
will have on the use of substitutes for the chemicals being
addressed. This will ensure that work group chemicals are
not inadvertently being replaced by potentially more toxic
chemicals. Standard methods for identifying the range of
potential substitutes need to be established. Furthermore,
procedures must be developed for estimating potential effects
of the use of substitutes and comparing these with currently
known effects of the chemicals of concern.
A final area of methodological refinement for next year
is the analysis of intentional and unintentional transfer of
toxics. Such intermedia transfers are generally due to:
(1) basic physical-chemical properties (e.g., voia-t-iiity), or
(2) one-medium oriented regulations that do not fully assess
the impact of controls on the total environment.
To fully address existing and potential intermedia transfer
of toxics, the work group will have to devise Agency procedures
for determining and apportioning among the environmental media
the ultimate fate of pollutants (i.e., air, land, or water).
This will present a challenge for each chemical work group,
involving evaluating known and suspected effects (e.g., ingestion
vs. inhalation) in light of the present environmental distribution
of the toxic chemical. Agency controls will then be devised
(with cost considerations) to optimally reduce overall risk.
Ultimately, the Agency may be able to make decisions more generi-
cally, along chemical "class" lines, than on chemical-by-chemical
basis.
-------
111-10
Resources and possible work groups for FY 82
Experience with the solvents work group suggests that a
chemical work group on a class of chemicals of similar scope
launched in FY 82 would require approximately seven work
years and §450,000 in contract money for that year. This
figure does not include research expenditures by ORD related
•fib" such an effort. At least $10,000 would have to be set
aside as travel budget for the work group. Approximately two
work years and $1,200 in travel would be required for imple-
mentation tasks in FY 83. Program offices should expect to
contribute from one-half to one work year each for a project
such as this.
For purposes of comparison, we have also estimated the
cost of a work group on asbestos and similar fibrous materials
Such a group would need about five work years, $250,000 in
contract support, and $3,000 in travel in FY 82. As before,
ORD research support is not included. The asbestos group
would cost less than the group on a new class of chemicals
because of the substantial amount of work that has been done
on asbestos over the last few years.
These estimates are restrictive and pertain only to
the sort of pilot-scale projects planned for FY 82. Any
more elaborate use of this technique would, of course,
be more expensive. Also, the methodology as proposed here
is workable only if TIP receives full support in data
gathering from programs and regions.
Three or four chemical work groups could be started in
FY 82. Examples of subjects for such groups might include:
asbestos and related fibrous materials
DEHP and related plasticizers
chlorobenzenes
phenols and chlorinated phenols
polycyclic organic matter
1, 4 dioxane and tetrahydrofuran
formaldehyde
-------
III-ll
C. Industry Work Groups
Purpose
The objective of the industry-specific approach is to
analyze those waste streams from a particular industry that
might be expected to have the- greatest effect on health and
the environment, and to present options for reducing risks
from such waste, where necessary, at an affordable cost.
More specifically, the strategy should designate which
waste streams should not be controlled, which should be
controllad, when they should be controlled, and the amount
by which each discharge should be reduced. It should provide
industrial managers with evidence of the Agency's long-terra
goals for controlling various waste streams, so they can
plan intelligently, and it should examine intermedia transfers
that may result from the application of control technology
to ensure that such transfers result in net risk reduction.
Products
Each industry work group will be responsible for:
o proposing a strategy that includes all major research,
data gathering, regulatory, and non-regulatory
actions the Agency should take regarding the industry
and indicates the offices responsible. The strategy
should eliminate any overlapping data collection
or analysis requirements and suggest how to fill
any voids. It should also synchronize data collection
and regulatory efforts wherever feasible and
worthwhile.
o certifying that proposed or pending rules concerning
the industry are consistent with the strategy (even
if it is still under development) or otherwise
appropriate for promulgation. Rules that are incon-
sistent should not be issued unless they are required
by statute or court order, or have de minimus effect.
Where statutory requirements are inconsistent with
the strategy (and particularly when the requirements
would have significant economic impacts on the
regulated parties), the Agency should consider
seeking amendments to the legislation. This suggestion
applies in the instance of chemical strategies as
well.
-------
111-12
What industry work groups would do
Step 1 - Assessing regulation costs
The work group begins by estimating the regulatory burden
on a particular industry. Existing, pending, and contemplated
regulations are considered, to generate estimates of what con-
trols have already been paid for, what the industry is currently
under obligation to install, and what would be the cost of
new regulations under consideration. This is done for the
regulations in all media, yielding an estimate of the cumulative
economic impact on the industry. The result establishes a
context for the analytical steps described below.
Step 2 - Estimating release, exposure, and risk
The work group must now find out what production processes
occur in the industry, and what pollutant streams arise from
each of them. It must tie this knowledge to the real world
by studying the annual production cycles of the target
industry, and the pollution control mechanisms the industry
now uses, or could potentially use, to minimize its
pollution control costs. The fate of the pollution streams
must be determined, as well as the fate of the wastes
accumulated by the control technologies in operation. EPA
has developed models, of varying precision, in each of the
environmental media to predict the fate of waste streams. These
will be used as needed. Using the exposure patterns produced
by the models, the effects of the emitted pollutants on
human health and the environment can be estimated, assuming
that we understand the degree of hazard associated with
each pollutant. Where we do not, industrial work groups,
like their chemical counterparts, should propose research pri-
orities to OPRM.
Step 3 - Synthesis
At this point, the work group has collected a substantial
amount of data in each component of the framework. It must
now devise some way of representing the interactions among
these components, to determine the most cost-effective way
of reducing risk. Where an industry is simple and presents
few obvious trade-offs, this task can be done informally by
knowledgeable people. Where an industry has many variable
processes and produces important emissions in all media, a
more complex model would be required. The main point is that
the sythesis represented by this step, whether complex or
simple, and whether accomplished judgmentally or with sig-
nificant analytic input, is essential if the industry work
group is to answer basic: policy questions.
-------
111-13
Step 4 - Control options analysis
The synthesis described above will generate a number of
outputs. At a minimum, an industry work group will calculate
the cumulative economic impact on an industry from all regula-
tions, and give some rough estimate of whether it can afford
contemplated controls. It will also rationalize the regulatory
pipeline for that industry across all programs-, rescheduling
where necessary and feasible, so that any adverse effects of
successive regulations are mitigated.
Ideally, industry work groups would also have outputs
such as:
o cost-effectiveness curves, which plot the costs of
pollution abatement as a function of reduction in
health risk (or environmental risk, or both) and
identify the pollution control options necessary to
achieve each level of risk reduction;
o measures of the ability of individual firms to remain
financially sound—to pass substantial portions of
pollution control costs forward to consumers, to
maintain acceptable coverage ratios, to continue to
raise funds on acceptable terms, and to avoid plant
closures or significant layoffs of workers.
Other questions that would be examined at this step in a
a fully developed industrial approach are:
o Are some pollution control options more cost-
effective than others in reducing health risk?
o Would regulations based on. controlling certain pro-
cesses actually reduce risk more cheaply than
those based on reducing certain pollutants to a
specific level?
o Is it possible to move from current controls to a
more cost-effective strategy (e.g., reduce health
risk significantly at little additional cost)?
o What is the cumulative cost of all controls in all
media on a particular industry?
o From the standpoint of health protection, which
environmental media should receive additional
regulatory attention?
-------
111-14
o To what extent do inter-media transfers influence
the choice of regulatory approach?
The analysis and strategic options are given to OPRM, which
proceeds as it does in the case of the chemical work groups
already described.
Case Studies _
The industry-specific approach should ideally be supported
by an analytical methodology that links industry economics,
production processes and control technologies, the transport
and ultimate fate of pollutants, and the risk to exposed
populations. Such analyses have been carried out in the past
for specific pollutants and specific environmental media.
The methodology used for this year's case studies is an
attempt to accurately specify the linkages among hazardous
waste streams in all media simultaneously. That is, we
wished to determine whether an integrated analytical tool
is feasible and sufficiently promising to justify further
development within EPA. Specifically, it was important to
assess three things:
o whether it was possible to develop sufficiently
accurate models of facility operations, pollution
production functions, and the life cycles of waste
streams;
o the volume and quality of data needed to conduct
analyses that are sufficiently reliable to act
as a basis for industry strategies;
o the cost, level of effort, and time period needed to
conduct the analysis and formulate a strategy.
As already mentioned, industry work groups could use more
informal estimates of risk and combine these with the same
cumulative economic impact measurement proposed here, if some
parts of the methodology prove impractical.
In FY 81, two industries were selected for pilot work.
Copper smelting was chosen because this industry produces
a relatively simple set of pollutants (hence an "easy" test
for the methodology) and has been under heavy financial
pressure because of general market conditions and international
competition. Manufacture of chlorinated solvents was
selected because much scientific work has been performed on
-------
111-15
many of the pollutants this industry creates; many of..the_.
pollutants appear in all three environmental media (air,
water, and hazardous waste); and, the industry is so complex
that controls on a single environmental medium may cause
industry to adopt changes of process, product mix, and
control technology that allow serious residual risk to health
and environment.*
Copper Smelting
Copper is smelted in fifteen U.S. plants, mostly in the
West. Several are located near population centers, while
others are in relatively uninhabited areas. The plants are
mostly old and face stiff competition from newer foreign
plants.
Without pollution controls, the copper smelting industry
would emit large amounts of SO2, particulates, lead, and
arsenic. (SO2 and particulates are found in stack emissions,
while the two metals are found in both stack and fugitive
emissions.) Because stack emissions are better understood,
and because they account for over ninety per cent of total
emissions, current controls have tended to focus on stack
rather than fugitive emissions. At most smelters, electrostatic
precipitators and acid plants have been employed to abate
TSP and 302 respectively. While these controls have
significantly decreased emissions, they have been very costly.
Using the methodology described, the study team plotted,
for each piant, cost-effectiveness curves relating potential
reduction in health risk to the annualized costs of achieving
that reduction. The curve for one sample plant ( located
near residential areas) is shown in Figure III-l. If we look
at health risk reduction alone, fugitive emission controls
are far more effective than stack controls. That plant's
current situation is shown by point A: it has stack emission
controls and some fugitive controls in place, acheiving about
88 per cent health risk reduction, but at substantial cost.
Installing further fugitive emission controls would bring
it to point B, acheiving an additional six per cent health
risk reduction, at relatively little cost. The situation
for plants located far from residential areas is quite different.
lyIt was understood from the outset that most of the health
risks associated with these solvents are the result of use,
not manufacture. Despite this, the wealth of data on solvents
made this industry a good choice for a short-term pilot project.
-------
03
I
HI
M C
2S <M
01
u!
CJ
0)
3 =
(0 0) 91 C 03 93
jJ N O 0 ON iJ
O --4 CJ •~4 —I fl
V
3
O
M
03
-------
111-16
In those circumstances, fugitive controls bring about little
improvement for the general population but may still be
effective in improving workers' health.
This study focused exclusively on risks to human health.
Virtually all threats to health from copper smelters are
transmitted through the air. As a result, there are no
significant inter-media trade-offs in this study. This
does not imply that the same would be true if we factored
environmental effects into the model with some attached
value. We plan to attempt that next year. ~~
Chlorinated Organic Solvents
Thirty-one U.S. organic chemical plants produce a mix of
about a dozen major industrial chlorinated solvents and inter-
mediates. The team studied fifteen plants in three regions,
accounting for 70% of U.S. chlorinated solvents production.
In the absence of mandated controls, these plants would
discharge many substances to the atmosphere, the waters, and
the land, exposing about four million people to respiratory,
carcinogenic, renal, hepatotoxic, and reproductive toxicity
effects. The overall magnitude of these risks, however,
appears thus far to be small. Considering cancer, for
instance, the unregulated plants would cause about an additional
0.1 to 10 cancer cases per year. Even with no regulations,
the plants would employ some air pollution and hazardous
waste controls because of their valuable by-products.
Cost-effectiveness curves were developed for groups of
plants. The curve for one such group is shown in Figure-III-2.
As with the copper industry, the first 90 percent of reduction
in health risk is relatively inexpensive. In this range,
the plants would invest in more air and hazardous waste
control and would install sophisticated water pollution
controls as the knee of the curve is approached. Beyond
that point, the plants would need to employ fugitive air
emission controls and eventually shut down the more polluting
processes.
Actual plants currently employ a fair degree of air
pollution control and relatively sophisticated water treatment.
The group studied in Figure III-2 operates at point A. It
achieves a reduction in health risk of approximately 28
percent and possibly significant environmental improvements
not reflected on the graph. These plants could move from
-------
EIGOHE m-2.'
total
Amuaiiaad
(million*
0* IMS
doLlacs)
200
too
CHLORINATED- ORGANIC SOLVENTS
(Seven Plants on tower Mississippi River)
More- Intense Controls Onr Air
Hazardous waste- Provide
Improvements- lit
Public Health
Current
Control
Point
Potential
Controls to
90% RisJc
Reduction
Effectiveness Carve)
:<*
P«rowt ^ducziort irt SWitti
Different Segments Of TI» Cost-
Effectiveness Curve Require.
Different Cbaoinarions
0£ PoLLuticn Control
Equipment
Am
«*
Iff
33
-------
111-17
their current state (point A) to a 90 percent health risk
reduction (point B) with a moderate upgrading of both air
and water treatment controls in place. If forced to adopt
these controls, none of these plants would sustain serious
financial damage.
Remember, however, that the overall baseline health risks
for this sector appear rather low, and a reduction from
the current state might not buy much actual health improvement.
Before firmly deciding whether to push for improved control,
one should examine any health risk "hot spots" in those
areas where plants may be concentrated.
A very rough estimate indicates that, with present controls,
the plants contribute relatively high annualized cancer risks -
(between 10~6 and 10~4) to a few hundred people living close
by. If this type of analysis were used to select regulatory
directions, the Administrator would need to exercise her judg-
ment as to whether this level of risk, faced by a small number
of individuals, warrants regulation. If so, then moving to
point B would reduce the number of people at this risk level to
nea-rly zero. Other, nonregulatory approaches might be worth
exploring.
Sample Policy Conclusions
The study team conducted a number of policy analyses to
illustrate the kinds of insights that could be obtained through
the industry methodology. From the standpoint of health
risk reduction, these two cases yielded these tentative
policy conclusions.
o For copper smelters, fugitive controls are far
more effective in reducing health risks than stack
controls. Similarly, for chlorinated organic solvents,
incinerators and gas absorbers are more cost-effective
than advanced water treatment.
o For the specific situation of chlorinated organic
solvents, either performance standards (a "bubble") or
an EPA strategy based on process control appears
better than one based on pollutant control, over a
limited range of health risk reduction. See Figure III-3
o In both case studies the current state of control can
be modified by certain feasible steps to bring about
more cost-effective operating strategies.
-------
*
a
•o
3
o
TO
a
I
•wfcZ
U
t
ulcj-K
iaioJ
jj
of*»
u ~
«"o
J=CU
op
0
oto
:T
-1
". i
s
3 =
-J
3
1
8
8
--•saw »n-~
J2 S
10
o a
-225 to
^•^ *** ^^
e°-8
-------
111-18
o In the case of copper smelters, if we focus on health
protection alone, regulatory attention on air pollution
dominates the other two media.
o For the solvents industry, air stripping from some
water treatment technologies can cause a significant
contribution to air pollution.
Limitations of this approach
The study team encountered serious data and theoretical
problems. As a result it was not possible to estimate precisely
the costs and health benefits of alternative regulations-
Thus, at this time, the methodology presented in this report
is not sufficient to support individual regulatory development.
However, it may be useful for the purpose of evaluating the
tradeoffs inherent in alternative abatement strategies and
to suggest the waste streams and controls on which Agency
activities should focus.
There is some difficulty in establishing the relative
risk of each disease associated with various chemicals. The
only pollution-related disease for which the scientific com-
munity has developed generally accepted incidence models is
cancer. As part of the two pilot studies, toxicologists---
developed estimates of risk for other classes of diseases
and a weighting system for comparing health effects that
would be controversial if widely used in policy-making.
Although preliminary sensitivity analysis indicate that the
cost-effectiveness curves are fairly insensitive to changes
in health effects scores or in the weighting, additional
analytic work of this type is needed before we understand
to what degree this methodology has actually related control
costs to risk reduction.
Moreover, there is a substantial body of scientific opinion
holding that efforts to quantify health effects in this fashion
are erroneous, in the absence of well-accepted incidence models
for environmentally-related disease. In other words, the
lack of acceptable incidence models prevents the development
of an approach that predicts absolute levels of health effects
and the exact benefits of pollution control.
Data and model reliability are problems for this model,
but they are problems for EPA in general. In fact, we do make
decisions based on data and models similar to those on
which the project methodology is constructed. Substantial
effort to improve capabilities in these areas will be required
-------
111-19
regardless of the degree to which the Agency makes a commitment
to this particular methodology.
The current effort focused exclusively on human effects
and omitted other environmental considerations. The decision
to focus on human health was made because of time and resource
limitations and is not meant to imply that other effects of
industrial pollution are unimportant. Thus, on the basis of
these studies, one cannot reasonably generalize about the value
of specific regulations, except in terms of human health.
If this approach proves feasible, it would have substantrierir—
advantages over the existing system of regulation-specific ana-
lytic efforts. For example, it might lead to:
o development of risk assessments that would reveal the
health risks imposed by an industry by individual
media and by all combined;
o more reasonable priorities for data collection, research,
and enforcement;
o establishment of public health protection levels in an
industry on an internally consistent basis across
environmental media;
o consideration of tradeoffs between control of different
pollutants and media; and
o issuance of regulations more sensitive to the economic
health of an industry and specific firms.
This is not to say that the proposed approach is perfected
or ready for widespread application. More effort needs to
be invested in reviewing the proposed methodology, particularly
its treatment of health effects and movement of toxic substances
from hazardous waste disposal sites to groundwater; accounting
for ecological and other environmental effects; and considering
alternative ways to display results and consider tradeoffs
among health effects and other environmental impacts. However,
the results of the work performed this year seem sufficiently
promising to justify further development and testing of the
industry analytical methodology next year.
The costs of performing industry studies are reasonable
in comparison with the costs EPA typically incurs in analyzing
problems and developing regulatory strategies, and would be
-------
111-20
a small percentage of the savings that may be realized by
industry from a more integrated control approach.
Resources and possible work groups for FY 82
Resource requirements for industrial work groups have
been estimated, using an organic chemical industry subsector,
of approximately the same scope as the present chlorinated
solvents study, as a model. About six work years and $400,000
in contract support would be required to bring such a group
through the implementation phase, which would probably extend
into FY 83. Travel requirements would be around $10,000.
The estimates also assume that eVironmental as well as health
modelling will be done. They include ORD's contribution.
Media programs would have to supply somewhat under two work
years altogether, with the actual distribution depending on
the industry sector chosen. ORD would have to contribute
about one work year.
Two or three industrial work groups could be started
next year. Examples of industries that could become work
groups subjects in FY 82 include:
iron and steel
foundries
one or more organic chemical industry subsectors
-------
111-21
D. Geographic work groups
Purpose
National environmental regulations are necessarily
based on estimates of typical or average local conditions:
typical p%nt configurations, typical relations of popula-
tions to pollution sources, typical risks and benefits of
control. There are benefits to this—and equity across
industries and states—but there are also drawbacks.
Uniform national controls may significantly over- or under-
regulate actual health or environmental risks at the local
level. This is simply because few areas are truly typical
in the sense national regulations must assume.
If one looks at the way in which pollution sources are
distributed across the country, it appears that a relatively
small number of areas have unusually high concentrations of
industry, hazardous waste disposal sites, populations, and
other factors indicating potential environmental problems.
Figure II1-4, for example, shows chemical production volume
by county for the 100 counties where the most chemicals
are produced: the first ten counties alone account for 30%
of the total national chemical production. Making uniform
national standards stringent enough to control conditions in
a few unusual areas is certainly not economically efficient.
Some other approach is needed for dealing with the outliers.
One obvious remedy is some sort of "geographic approach"
for environmental control, in which local areas are sep-
arately analyzed and their problems individually regulated.
This is in fact the oldest concept of pollution control,
on a single program basis at least (SIP planning, for example).
But as regulations become more complex, controlling greater
nuntoers of substances in many environmental media simultaneously,
a more comprehensive, multi-media type of geographic analysis
seems indicated. The concept has intuitive appeal. Its
benefits include:
1. Efficiency of control of health and environmental
effects' Heavy local concentrations of industry or
population, particularly if combined with unusually
sensitive environmental conditions, can lead to
pollution levels signficantly higher than desirable.
Multimedia geographic analysis can analyze this,
and can also demonstrate which media might be
overregulated locally in relation to actual risks.
-------
i:
8
s
Ul
I
o
§
oc
ai
u
Ul
o
Ul
OC
§
i
s
s
Ul
(0
O
cc
2
Ul
u
cc
Ul
s
Ul
Qo
in
to
3
S
S
10
o
o
O
OC
u.
o
cc
Ul
a
OC
o
*
e
z
«/»
u>
1
Q
ui
cc
a
Ul
O
10
10
o
(5011 iwnnoA NOiionaoHd IVOIWHHO do sawnod
-------
111-22
2. Ability to increase the cost-effectiveness..of-
controls Detailed analysis of a single area will
reveal opportunities for more cost-effective
control actions than might otherwise_.J>e~ applied
under routine application of national standards.
The costs of complying with nationally-imposed
pollution controls can endanger industry and employ-
ment within a local area. Geographic analysis can
show how to minimize costs of controls by using
approaches tailored to the specific situations
(including use of economic incentives, ambient-based
standards, and state and local regulation).
3. Encouragement for states, industry, and the public
to participate directly in the analysis and control
of pollution problems and the selection of controls.
Geographic analysis can function as a participatory
process. As such it can increase local under-
standing of pollution issues, bring local interests
into decision making, and improve the possibility
of controlling problems that national programs do
not adequately address{
Multi-media geographic analysis can be applied to any •
class of pollutants, toxic or conventional. For various
reasons, however, geographic studies appear particularly
promising for toxics. First, as we have noted earlier,
toxics often tend to be highly mobile between media; geographic
studies can analyze intermedia effects in detail and design
cost-effective controls accordingly. Second, monitoring
toxics is still many times as expensive as monitoring for
conventionals; geographic studies can focus scarce toxics
monitoring resources for best effect.
The geographic approach could support this Administra-
tion's goals:This Administration is seeking to shift more
environmental control responsibility to state governments,
improve the scientific basis on which regulatory decisions
are made, and reduce compliance costs to industry. Although
some important technical and institutional issues remain to
be resolved, the methodology developed this year suggests that
this old approach to pollution control might offer some fresh
benefits in relation to our toxics control responsibilities.
-------
111-23
First/ it is clear as a practical matter that the states
would play a key role in a full scale geographic study, though
the actual division of responsibilities—and limits of state
authority in the process—are not clear. States could in
theory use the process to reassess all toxics-related controls,
existing or planned, though this might require legislative
changes. In any event, the process can be designed to cast
EPA in a technical support role, supporting states in sampling,
lab work, modeling, engineering, and economics. This would
increase state and local participation in decision making,
without removing EPA from a key role within that process.
Second, the geographic approach advances the Administration's
goal to increase the level of scientific analysis brought to
bear on pollution control decisions. In those (relatively
few) complex areas where toxics exposures are potentially
greatest, the geographic approach can focus EPA and state
analytic resources to produce much more detailed information
on toxic discharges under actual local conditions than is
usually available. The depth of this technical analysis,
bringing to bear the best scientific tools available for use
under field conditions, permits design of a comprehensive
local control strategy that could minimize compliance costs
without sacrificing environmental goals.
This detailed diagnostic approach could have particular
benefits in the older industrial centers of the Great Lakes,
East Coast, and Gulf Coast regions. These areas are already
hard pressed to comply with existing regulatory requirements
and may find the next round of toxics controls especially
difficult and expensive to implement. By creating a com-
prehensive data base to support innovative control techniques
in these areas, and by helping to rank exposure routes based on
order of actual health and environmental impacts, all varieties
of controls could be more accurately and economically designed.
Products
At the conclusion of a geographic project, the local
government, state, and EPA will have available the following
data and analytic studies on the area studiedr
o a complete inventory of all significant sources of
toxics chemicals, including all current emitters
and all other direct or indirect sources of emission?
-------
111-24
o directly sampled or mathematically estimated values
on current flows of selected chemicals throughout
the local environment, in all significant environ-
mental pa thway s;
o directly sampled or mathematically estimated values
on human and environmental exposures to those selected
chemicals, summed across all contributing pathways; and
o a strategy for reducing risk in the way that is
least costly.
These data will be a permanent resource for future permit
and compliance activities and, since toxics trend monitoring
is still seriously inad.equate, especially in air and groundwater,
the data could help EPA plot toxics trends nationwide. It
could also support an agency-wide regulatory priority-setting
process, as well as contribute to the annual updating of the
Intermedia Priority Pollutant list.
The most important product of the work group, however,
will be a strategy for controlling those toxics exposures
that are significant problems within the local area. The
strategy will be a comprehensive plan for integrating environ-
mental controls vertically across all levels of government,
and will be cooperatively implemented, enforced, and updated
as necessary.
What a geographic work group will do
Various tools and procedures already exist to handle
virtually every technical element of a comprehensive
geographic study, but these have never been tied together
into a single package that could be tested under realistic
field conditions. We therefore analyzed the overall feasib-
ility of a prototype geographic approach based on currently
available technical components.
It is important to note that our approach is based on
exposures and estimated uptakes, not observed dose or any
comprehensive attempt to quantify health and environmental
risk. For instance, at present the methodology does not
emphasize direct sampling of populations to detect actual
toxics uptake in the body (sampling of breath, hair, blood,
urine, adipose tissue, etc.). Such sampling could, and
should, be undertaken to verify estimated exposures where
they could be unusually severe. At this point, however, we
feel that it is not cost-effective to use a "bottom-up"
-------
111-25
methodology. We have designed a top-down methodology that
begins with source identification, models environmental
pathways, and verifies predictions with various kinds of
monitoring—monitoring which could well involve tests for
actual doses received by the population. The process
would take from 24 to 36 months.
We have not yet dealt extensively with the crucial
issue of state involvement, but it appears clear from past
experience that state participation in geographic studies
is crucial, and that states should take an equal if not lead-
ing role in the eventual use of the process should EPA
choose to adopt it widely.^ What we have been able to define
are general technical steps that the project must conduct in
order to complete an integrated exposure study.
Step 1 - Data acquisition, study boundary definition,
initial scan of significant environmental
pathways, and selection of pollutants
The purpose of the first step is to develop a rapid
understanding of the potential environmental problems within the
geographic area through scanning the.available data concerning
emissions, ambient levels, and exposures. Relevant data must be
collected and interpreted in order to determine what the signifi-
cant exposure pathways might be, and to identify data gaps that
need to be resolved? this could involve major data gathering
efforts lasting a year or more. This step simultaneously
considers all pollutants, source categories, receiving media,
environmental compartments, and receptor categories in order to
identify potentially important pathways. Those chemicals and
pathways that are considered most significant on the basis of the
available data will then be subjected to more careful scrutiny in
step 2.
Step 2 - Detailed exposure assessment: quantification
of source emissions, modelling of chemical
fate and transport pathways, and^determination
of receptor exposure routes in various media
During this step, detailed data are developed on the sources,
pathways, and receptor exposures selected in the first step.
Most of the data required for the exposure assessment will
have already been gathered, so that the exposure assessment
I/ This has been most clearly shown in Region IV's projects
on the cities of Memphis and Louisville. In the first case
the project was seriously undermined by local political factors;
in the second the Region has gone to considerable length to
devise improved methods for local participation.
-------
111-26
work will involve mainly analytic efforts and computer model-
ling. Additional data collection may be desirable for
verifying the predictions of models in sub-areas.
Step 3 - Identification of potential toxic substance
problems, development of strategy options, and
evaluation of costs and benefits of alternative
control strategies J
The emphasis here on exposure rather than risk is based
on the belief that it is not crucial to assign risk values
to exposures in order to make control decisions. Observed
exposures can be compared to existing standards for toxics
(such as MCLs in drinking water or water quality standards)
or, where these do not yet exist, to typical levels of exposure
elsewhere in the country. As our ability to quantify health
and environmental risks of toxics improves, exposure data
could be transformed into those terms.
Several courses of action are available in Step 3.
o Where the work group detects clear violations of
existing standards or regulations, states may
initiate enforcement actions.
o Where potential problems are of sufficient gravity,
the work group may consult the Centers for Disease
Control regarding the development of suitable public
health protection plans for imminent hazards.
o Where the data suggest the possibility of long-term
human health effects, epidemiological studies may be
initiated for specific pollutants; the Centers for
Disease Control can also be helpful here.
o In the light of all of the above, the work group
will consider all available controls to achieve the
most cost-effective local strategy.
The Kanawha Valley—a case study
The pilot project focused on a major industrialized sector
of the Kanawha River Valley in West Virginia, centered around
the city of Charleston. The analytic work was directed by
Region III staff with contractor support. The State of West
Virginia was extremely cooperative in all aspects of the
study, supplying new data to the analytic effort and working
effectively with local industries to explain the study's
(limited) goals and gain their full cooperation. West Virginia's
-------
TABLE in-5
FINAL POLLUTANT LIST FOR KANAWHA
VALLEY GEOGRAPHIC STUDY
Acrylonitrile
Arsenic
Benzene
CARSOK TSTRACHLORIDS-t-
Ch-lorine*
CHLORQFORM-*-
Fo-Raaldeiivde *
LEBD+
Nlclcel*
Mercury
PAH*
PC3
Perchloroethylene
1^1,1-trichloroetiiane
VINYL CHLORIDE +•
+• Considered in exposure analysis modelling efforts of" Phase 2
* Level Z pollutants: collection of data on sources, loadings
and ambient levels was not. as> intensive as fr others on the
list. . , • • .. - . ....... -
Underlining'indicates'that'the'chemical'appears on the TI?
Intermedia Priority Pollutant (IPP) liar.
-------
I
rTGCES ni-6-
KANAWEA V]\LL2Z INITIAL SCAN"
1
1
I
1
1
1
1
I
1
1
MOTE: The dots indicate categories
selected for cansideration
ire this study.
-------
ZIGUSE HI-5.
GZOG2ASSIC BOUOTABISS OF' TEST KANAWEa. PILOT STUDY
Ffgur« 3-3
West Virginia Hatar Quality Manacjenent Plan
(20S) Sufcwa'tarsneds in )Canav*ha Valley Study Area
-------
111-27
participation confirmed the crucial importance of state par-
ticipation in studies of this type. Two major products have
come out of this effort: .(I) a report on the Kanawha area
itself, and (2) recommendations for a general methodology
that could fully handle all the issues raised in the pilot.
The Kanawha area was used as a pilot site for several
reasons. First, it is a major organic chemical producer
whose products were also under study within other integra-
tion projects (the chemical and industrial). Second, it is
geographically and ecologically quite isolated, such that
the study would not be unnecessarily complicated by imported
background pollution levels. Lastly and most important,
however, toxics related data for the area are plentiful. They
have been generated not only by Region III, but also by the
State, NEIC, and various special research projects. We did
not select the Kanawha in expectation of finding unusual
toxics exposure problems: indeed, the extensive data available
are reflective of the attention the area has received over
the past several years, and the progress that has been made.
The project proceeded using the three steps described
above.
Step 1
Boundary definition; To define the study boundaries we
included all known industrial sources of toxics in the general
area, as well as pertinent hydrological and geographic con-
siderations. As a consequence, two slightly different study
boundaries were chosen:: a primary source-oriented one, and a
secondary one that followed subwatershed outlines (see
Figure III-5).
Source identification; The Kanawha has a number of
possible sources of toxics. Chemicals plants and coal raining
are the primary ones, but toxic waste dumps, commercial and
residential development, and others are also significant:.
The study categorized these and researched all available
data at all levels of government.
Selection of pollutants; Drawing on all available data
sources, some 59 toxic pollutants were identified as poten-
tially significant. Out of this number the study had resources
to focus on only 16, and could conduct full exposure modeling
analysis on only 4. Criteria for selecting these subsets of
pollutants included such factors as their presence in industrial
sources, the amount released to the environment, availability
-------
111-28
of ambient monitoring data, listing of the substances on
various Agency priority lists, and so forth. The final set
selected is shown in Table III-5.
Results of initial scan: Major pathways in the valley for
each of the 4 selected primary chemicals are shown in Figure
III-6. This simplified diagram demonstrates that humans,
fish, and wildlife are exposed to all of the chemicals but
through different (usually multiple) pathways. Actual
quantification of activity in each of the cells became the
study agenda of step 2.
Step 2
Pathways analysis; The two major modes of discharge in
the Kanawha Valley are to air by stack emissions and to
water by plant outfalls. Two environmental models were chosen
to chart the effects of these discharges: (1) an air model for
the source-to-air pathway (standard EPA "ISC" package), and
(2) a water body model for the source-to-water and water-to-
other-media pathways.
An example of the output derived from these models is
shown in Figure III-7. Here chloroform levels in the river
are shown for both average and low flow conditions. Not
surprisingly, concentrations during low flow are many times
higher than they are during average conditions. During low
flow the values fall within the range of concern indicated
by SPA's water quality criterion for chloroform.
Exposure assessment; The objective of the exposure
assessment was to identify human subpopulations and aquatic
biota residing in the Kanawha River Valley that are exposed
to toxics levels in environmental media greater than those
typical for the general valley area. In addition, some
exposure levels were quantified for the general population
as well as for selected subpopulations that were potentially
sensitive or exposed to higher levels.
Only humans and aquatic organisms were considered in this
project, a limitation arising from constraints on the overall
scope of the study. This limitation does not imply that we
expected exposure to be unimportant for other potential
receptors—plants, microbial communities, etc. Some con-
clusions of the exposure assessments are shown on Table III-6.
-------
UI
1—
2
*-+
s
0
M
i— tn
2:2:
UK
ox
ZW *
0 J
cro 5
om t.
U_i— «
C A L
— lit I
o a. i~~~
t— ^
o
o
c
a;
r— 1 -^
re V) rO t. to
4-> tOTj 4-> O ••-
C i OJ
O ^ 'i- r— >
S_ OS-1*- C •<-
O 4- O O 0] 1-
™ flj 4J ^J 4-i 1^
^j ^f S *^ * *
s "r" *f" 3k (/) (/I
C i- +-1 +•> •>- ai
M U f8 0) C". ^ l/l
S J= 4- C 1— 0
"* >)4-> -r- T- Q.
*J i— ^i S-
•r- to C • 3
^ D re f- i- a
,T3 +•> S- C
3 U 4J •<- O
s_ '-a '5 ^ *" H
"D C Q^ QJ fQ
*j ^- -a o j= s-
ro O> 4) +-> --J
3- — . +J a. to
a) IT: s. 3
C C -i- S. O i—
J= -r- 0 O S- r-
1_ t— O '*- 4- i-
F-
t
P
I
1 .... 1 .... 1 ....)...
"'•""["""'"I"'
^^ 00
0 »
o c*
0 0
to
•^
.c
*->
c
•f—
O)
u
i-
3
0
to
^
0)
^— *
ra
2
c:
•^
^y
C
•r-
i.
"O
tO
Ul
fO
T3
OJ
to
3
^_)
a
c
1 1 1
r
•
*
f
/
•
* .
/
;
A z
o
• *~"
• LL
/ 1-
/ 5
/ >
! i--
ra I >~
*• i 2
t a
* t
/ ?
/
1
"""*"*""* — — •«
J
f
""""""?
i
4
/
\
i i i i i » r i i
iii|iMiliiii{Mitliiii{ttiHiiti(imtiiii[iiiiltiitj
r^» us uo ^f co c
o o o o o <:
o o o o o c
o o o o o c
^-
T
—
1
J
1
"
1
1
1
r* ««
1
* I "
1 *
j I
• 1
• /
J t
1 «
* 1 «
• J*
.
• i -
f :
f -
t :
\ :
T :
( :
> ;
/ :
T '
1 ;
ii
«
-
j
i .«»i • • > . t . . • • i . . . . i
''"'""I""'""1:
J *-H
3 ^
» 0 C
J 0 C
.0
^^
o
C\J%
.0
LU
HH
.0 31
LU
cr
-S
•s
.0
o
^*
»
>
CWdd)NOIlVH13NONOO
4-35
-------
b
9
a
«
b
a
a
o
flu
a
o
'o
b
O
J=
U
o»
to
1
s
ca
b b b b
X X X X
90 ao ao ao
a J* o a.
h4 U ~4 «4 i-t
A 2 « »1 flU JS U
a
90 90 90
Jtf .* Jtf
I
** X
** a «•
• Ob
e -H « jrf
1-4 M a u
« « o
• b b
TJ • O f
x Q. » i
J3 O * J
b U 00
o q
•n B
i
u
i
»
<
QZAX2S90
• ^HB« ^HM* ^^** «•«• ^MHB ^IM
ss^VHOSia
b
1
i
O
w
a
i
!
•1110
1
K«
Ri
ii
ao a*
O M 90
-4 O "*
i
"S
in ^
42 3
b M
-j
b u
a b
3 M
U
O O O c*
•* O
O O
* s 11
3 A OiTI
X «* Xao -4
M «4 W +4 •
«4 4» »4 • 4« <*4
a • a « b
ft
b
Q *^«^
lh*
3 b b
<§
b
U
0 O O
-------
a
o
o
SU
u
o
VU
o
M
O
en
o
i—i
en
1
H
Z
Cd
cn
vt
Cd
M
cn
Cd
CA
2
X
Cd
C
0
o
vo
I
06
Ed
H
3t
act
^J
c
«
4)
B
^
u
^
*^»
oo
a
CO
o
o*
c
o
4J
a.
e
o
u
ft
a
•H
• • hi
U X J=
.C « ^
X ofl
~* 3
^^
• fcri
U C J=
j= a o
OX'S
CM ^"^
** ac
a
^^
iH
C . U
O M X
• S "*"
a x ac
4) 8
cn NB*'
^4 ^^
a u
e . x
o ao —
a > OQ
a •< a
41 ^
CO
41
I
^
a
ad
J
a
x
a
a
v- '
U hi
X X
oo ao
a a
o\ >o
•» en
e .-»
c e
••4 o
U *H
Ok U
hi a
0 4)
OB hi
^3 O
•< 41
hi
e ^
CO
o sr o o
O "•> O O
«y
^^ c^ —4 c^
O*
•H o jt a o
u S U Q J3 T3
« a o a. a -^
C -1 U 4«
M Q a -i . -H
a a o «-i •
• . js w x i a.
U U CO 41 41 hi O
•H -H c •— ' a a.
> > O fib
asA^ssao
x
a
a
3
at to
4>
1 §
u **
9
M
a
4) w
a •* e
K y M
u to
^j
^^ 44 ^4
a o 3
e -<
0 X X,
•H a
^ V
a tt
z « X-N
u u
41 -v
u ^4 ae
8 « 3
•H a
u z
«
41
•• 4J
« a
^ 8
41 -4
> 0 CM ON xO hi
fj*
^
^4
a
3
O*
±4
^. 4)
fO C1^ U
O CO O r>. O O a
• • • * • • 3
•"•^ *-^ *^ ^">
«* CM *O *» «*>
(^ (^ ^^
« u
« s
a
B -i a a a -* a <
§« 3 -< 393 SU
u e ^ sue cd
v4 ^4 *^ CQ ^"4 ^4 ^*4
c a. x td c a. x a
•^ X eg «4 x S °
X H X ^ X H 2 hi
Cxi ^^**
e e
o o
a a
•4 ^"4 ^^
a a
J= ^ U
e co
*« »H Q
C* fiT ^4
a a b
a a 4i
.5 3 "
^
1 1 w
X
hi hi hi U U •*•!
XX XXX -<
ao ao ao ao ap 3
a a a a a cr
— u
CM O -* O O 4)
• • • • • u
in -T o ^a- a
!•» «M 0 3
•* CM
9 G*\
» — 1
JJ U
a =
^ s a a -* a u
« 3 44 3 « 3
y B •" 8 U S •<
•H
-------
111-29
Step 3
Two control problems were treated during the study. The
first was an analysis of options available to limit carbon
tetrachloride and chloroform exposure from two plants along
the river/ considering both air and water pathways. Various
"bubble" solutions were examined, given different environmental
and health policy goals. Figure III-8 summarizes the analysis.
The second was an evaluation of lead contamination in the
study area from three sources: urban runoff, industrial
point sources, and drainage from inactive coal mines.
Conclusions
Generally, the geographic methodology appears highly
promising. The Kanawha study has yielded, within the short
period of six months, a methodology to create exposure
assessments of the depth and detail necessary for designing
a comprehensive local toxics control strategy (though this
will not be done in this pilot effort). At the regional
level, cross-media analytic work has progressed very smoothly.
At the state level, cooperation with the study has been
exemplary; without it the study could not have been attempted.
The Kanawha study did not find evidence of widespread
elevated toxics exposures affecting the local population —
at least for the four chemicals fully modeled by the analysis.
While this was expected (as discussed above), it tends to
confirm that existing control mechanisms put in place through
state programs are adequate to protect the local population,
even when the cumulative effects of multi-media exposure
are taken into account. The more important general finding,
however, is that the analytic process itself is workable.
The geographic approach is, however, highly data dependent,
Relatively good toxics data exist for the Kanawha Valley, but
even here there are significant gaps. In order to launch a
full-scale geographic work group in FY 82, three to five months
of advance planning will be necessary to plan data needs in
detail. ORD will participate extensively in this process.
Geographic work groups appear to have the potential to be
an effective response to the needs of a relatively small
number of highly significant sites. The necessary analytical
techniques to conduct the studies are available today. While
considerable work may be needed to increase their accuracy and
precision, they have been shown to be workable under field
conditions in the Kanawha study. In the long run we expect
-------
FIGUSE III-8
Cost-Effectiveness Evaluation for
Chloroform Control in the Kanawha Valley
PERCENTAGE OF REACHES EXCEEDING
1.9 mfr VS. MINIMUM TOTAL COST"
FOR EACH STRATEGY - MEAN FLOW CASE
SO
X
ui
20
13
Interior Hull
8 9 5 14 11 12
15
I
1.0
1.5
I
2JO
Minimum Coft (Millions of $)
'"Percentage of reaches exceeding 1.9 jig/?* is literally the number of reaches in which the 1.9 jjg/2 target
concentration is exceeded divided by the total number of reaches and multiplied by 100%.
"•"Minimum total cost" is the minimum total annual dollar cost to plants A and 3 combined, and is not
necessarily a measure of value to the plants.
*** Points on the interior hull of the graph comprise the tffidmt an of strategies.
"his rigure arrays costs of various packaaes of control options
(0 through 20) against water Quality tene^its orojected fron-
their irole^entation (nurfcers of river reaches f^at would attain
water quality criterion associated with 10~- cancer risk). The
diagrar illustrates that only options ?. (no controls), 1, 13,
and 2. l^e on the cost-effectiveness frontier (the "interior hull"
i1lustrated).
-------
111-30
that the Agency could conduct at most three to four such
studies per year.
In the coming year the most important methodological
considerations will be designing data-gathering protocols,
identifying site selection criteria in cooperation with the
states and EPA regions, and working out institutional rela-
tionships with state and local governments to permit the
start-up of a full scale project. Obviously, site selection
and many other aspects of the geographic approach will depend
on policy decisions regarding the roles of regional offices
and states in the exercize of environmental protection
responsibilities.
Resources and possible geographic project
One geographic project could be started in FY 82. The
choice of site would depend on consultation with state and
local government and with Regional office personnel. Costs
for data development are estimated at from $1.5 to $ 2 million
over two to three years, most of which would be spent via
contract or through state and local government. Sources of
funds would be state grants, state in-kind contributions, and
the Agency's toxics integration budget. Control strategy
development would require 10 to 12 work years, from regional,
state, and contractor personnel. Reprogramming of state
monitoring resources under existing grant agreements will be
required, together with extra grant assistance, in the neighbor-
hood of $75,000.
-------
as £
O M
2 = cn
S a o
HH i-i as
2 O cj
C < iS
Cd 25
588
u^
cj cn >
O c as
cn O =H
cn = cn
Cd
a o cd
E-i cj 3 a
a- cn <
icid
a z £
J < Cd
CQ ~
O cn cj
c: i
1
as
o
c
z en
a a
M J
O CQ
> O
< 2
cu
as
o ca
cu cn
cn s
cu
M cj
CJ Z
Z M
O N
CJ M
" £
CJ M
M Z
cn M
< £
03
1
|
cn
2
Cd
J
a
O
Eu
><
Cd
^
i
CQ
CU
f\
i_l
3
O
CQ
CU
Ll
Ll
CU
Ll >
CU CQ O
JJ CQ Li
Ll CU >i CU
(0 y JJ cn
y LI u c
Cu o to
*U -C £
CU Cn -ti
.H C 3 .*
•1-4 1-1 (0 Li
o o o
^
i-4
CU CQ
£ CU
•<-t >
JJ -^
JJ
U U
O cu
\-l-l
T3 J2
C 0
tO
c
CU CJ
> *""
•^ 3
CQ
•« CQ
0 C
CU O
•O -H
CQ JJ
cu cu y
> J^ CU
•>H 03 Li
J-> £ Ll
0 0
CJ 0 U
•n jj
n cu
O >i CQ
JJ Ll
Ll -H 3
<0 — ' 0 O
JJ
cu
T3
3
JO
C
3
0
ca
jj
•i-t
cu
>
(0
i
T!
f-^
3
O
f"
CQ
a
3
O
Ll
cn
v>
Li
O
2
O
C
3
Ll
I
cn
C
o
'O
CQ C
CU I0 CQ
y cu
LI £ y
3 3 Ll
0 •-• 3
CQ t3 O
CU CU CQ
Ll £ CU
Ll
CU Ll
JJ CU U-f
to > O
3 0
S > jj
T3 JJ i™*
(0 C -^
•fH -i-l
«-i to 13
O JJ (O
T3
CU
— I Cn
•t-4 £
Ifl i-l
JJ JJ
CU (0
*O 'O
a
tO 3
C Li
•i-i 0
(0 U-l
JJ
C CQ
O CQ
o cu
0
tJ O
c-1 Ll
3 C4
O
CQ <0
£
Li U
CU 0
jj ua
^j
(0 <0
J^
O T3
C
Cu (0
3
O cu
Ll C
cn-i-i
—4
.* CU
Ll £ •
O -i JJ
2 JJ -rH
O
JJ
U-I
•f-4
Ll
"U
U— I
o
CU
CQ
C
cu
CQ
1
1
CU
— 1
3
"Q
CU
!*•
"o
CQ
£
U
cu --H u cn u — i — '
— 1 .2 1 C .ii CU — i U-!
1 y to 'C
i— • Z3 (H
0 0
O
CQ
U
O
i [^1
e o
CU CO CU
£ y LI i
cn -i-i a, -t^ jj
c LI x 3 a c
o cu o o cu
J^ EH Li CU CQ
JJ 10 C7i U CU
CQ £ CU 01 iH Li
•f-t «C J^ C U.4 CU
co >i jj LI o u-i cu
c y o LI o s-i
1-4 -^ M-I 3 jj
p-i O CQ CU Lt
T3 O CU u •«•) « t:
— «^ Q) ^J j^j CU 3 CU CQ
3 (0 ^J — ^ 'O ^ CX
O O CD <0 C JJ 3
.C JJ JC T3 JJ -»-4 0
CQ C Ll JS Ll
CQ 0) <0 CO 0 J-> CP
CQ CQ .C T3 JJ •<-!
U CU JJ - C 3 J*
CU U U-I 3 CQ Ll
cn y » *J JJ £ O
(0 (0 £ (0 'O Li y 3
C OS JJ M O 3
(Q d) Qa C.O 3 *^ O ^J
E > o O *w JJ J=
(0 C £ CU JJ
Oij: LI o CQ c
a 0 -H = "t c
O 0 u-i jj cn cr> o
Li JS (0 Ll C CU
Cn3 <
Ll Ll CU JJ C CU — I
OCU JICtOCQO-^
S-° E- *- S s jj us
0 0
a
3
o
Li (0 1
O^ CP C-
O CO
&. JJ —1 ..---4
LI QJ rj3 cn
O CQ CU CQ--H
jj 3 CO CU CU CJ
C 3 ^ U CU
CU >i CQ CnTJ
CQ J2 CQ £ O
0) --I O U— i
u c LI a cu
a o CQ '4-1 >
CU **— * CU ^ . CJ
LI -u y cn 3 1—
3 Ll -10
JJ .— 1 0
0 O '•w Cu cn y
c • cn — i —
ca cu >i y ^ i— '
0 Ll Ll —( C Ll 0
T: O cu •«* o n
JJ i
Ll CU CQ i-l C 10
cuu cau^ cu o — '
.Q -1-4 -^ di CU — V
EC O > •"-» CU T3
0) >i Ll CU JJ
S cu to Oi~- cn (0 c
U -H Gu JJ ^ 03
Cj-»^ cutOLi ciycn
3^-1 'OCCU CUOC
O JJ CQ-1-1
Li O C ^ D 1 -*-^
Cn fO 0 "^ Li 2j— iI3
U CJJJJC i~iJS
^ •-»
Li CU
O -C 1 1
Z ' 2 JJ 1 1
O
O
-------
1
1
i
1
-— »
T
^
"
C
U
_
^
J
5
r*
Ci
o
CJ
2 cn
M s;
O !i2
M 1-3
o a
> i
Cl^
Qi
O Cs3
su cn
u
cn s
s-i £_
fti
Z2 CJ
CJ 2
2 t-t
O N
CJ i-"
CJ M
M 2
cn *"^
i
u
^ ,
a-.
^ CO
C jz cn
S4 c
cn o O
r- •'-'
_^ tj
i-l • -H
O tn tn
3 ^ O
CO Q.
* cn
JJ 03 JJ
C C CU
CO 03 CO
£ £ U
CO O
cn s-i to
03 O
C -H 0
CO CO CU
jj OJ O
O t3 C Li
co ^ 03 cn
u-i 3 £
U-i O X
tn u O
CO 33
54 cn tn
35 co co c
C cn U — t
(U tj -3 T3
C 'O (U
0 03 C CO
E- E -i ->
0
2
5
E
tn u
CO OJ
c T:
•-H C
CO
JJ 03 JJ
U -4
JJ CO 3
•-^ cn cn
54 ro CO
CO C W
c "C
£ 03
JJ
O U 33
C O 03
o "c -c
-COO)
tn cn
03 0!
Cu >i C
3 O 0!
! C - S
1 !-i CO
cn E >i
•—• — i
*J w
C 04 0
L|
o
Uj ^J
o
'U co cn
C CO O CO
o o -^ -^
•i-l JJ U-I Jj
jj C U-l •<-*
a •- 03 c c
o cn ij 3
CO OS JJ JJ
a •-* 3 c >4
£ KJ u cn co 0
JJ O -i-4 S U d
.— t a] pr
Cy jj .M CO CU 0
> 3 .Q >
OJ O •"-! 05 CO ^4
£ ffl £. £. 03
U C JJ CO
*O co 0 *U ^^
-i U-I Q, -HO O
3 tn tn 3 jj
o c co o co
JT oj u js = j2
cn u tn cn
jj CO •*-* *^
S-I C i»l JC >H
co cn*H co cn 3
cn c 'H cn-i-i o
03 *-^ 03 ^4 j«
c JJ -u c tn
0! 03 C 03 C
£ CO CO £ l-i CO
3 i-l 3 - ti
CO CT ^ CO JJ 13 CO
J= CO 3 jC CO C C.
jj U U JJ U 03 JJ
1 1 1
1 1 1
cn
c
.,H
Cn
5
03
£
0)
iH
JZ
3
4J
C
£
U
>—
03
>
'O •
03 cn
p!
TZ 3
CO 0
3 U
c cn
•H
jj .***
c u
0 0
0 3
te
01
i-j
CJ
"U /^
E t
O CO
tn S
CO CU
> 3
03 O
J= U
T3
•— ' ii
3 S-l
0 0
£ 3
cn
54
« a
^ >
CO O
cn
03 CU
c cn
03 rtj
E s-i
0)
CU ^
3 CO
O -I
S-i
cn >-^
oj
^ £
U U
0 0
3 u-j
0
s
_>J
U
O
3
jj
O
C
o
^
1
cn
u
CO
i
CO
£
cu
w
0
*4
cn
S-|
0
2
cn
c
O
•^4
JJ
03
3
OJ
>
co
-H
cu
c
c
0
cn •
w tn
CO CO
CU >
S-i JJ
1- 'J
> CO
O •(-!
Q
co O
o
c >
CO 03
3 CU
-4
,
JJ 03
•i-4 U
S4 CO
o ^
jj O
3 £
03
a
CO 3
> 0
03 Sa
jc cn
TJ X
-4 Sj
a o .
O 3 CO
j= O
tn O c
JJ 0}
03 £
u cn s-i
co co O
cn cn
03
CO
CO
U
CT
cn
03
CO
JJ
03
3
e
CO
1-1
co
cn
03
C
03
£
tn
u
CO
cn JJ
— jj
•-I CO
cn —i
3
C
C O
•fi *-H
JJ
T! 03
03 C
CO
tn £
cO" 0
> u -
CO
^ cn
03
-a
03 .C
£ U
iJ 3
O 01
U-I
^
CO 03
03 C
CO 03
J= CO
CUJJ £
3
O
14
cn
u
0
3
0
cn—4
C 03
•- ^ £
tn u
3 0
U-I
c c
•r4 —I
c
o
•u
C CT
03 C
03 C
i-l -r-l
CO OJ
cn s-i
03 JJ
C
-03 .-4
£ (0
s
CU i-i
3 O
O U4
jj
cn cu
\^> f _j
U CO
0 0
3 CO
14
•
CO 'C
JJ -4
OJ 3
•^ C
54 —
Cu cn
O
u cn
CU ^J
cu a
OJ C
cn co
<; £
C
C
o
-— i
JJ
03
CU
y
• r-t
JJ
L|
03
tl
O
U4
O
jj
^
CO
gj
o
cn •
OJ 03
c cu
-------
T3
-
jJ
C
O
u
—
f— 1
1
M
f^
I— (
a
j-O
<
ta
CS
O
o
z en
w £
Q U
M J
o ca
> 0
< as
&
OS
O u
&4 M
U
03 =
I:*4
Ui*
Cd O
CJ Z
Z M
O N
u *•«
s
u >-l
M Z
U3 M
£*
w
M
^J
03
§
o.
>H
CJ
Si
'
1
c
o
•*H
-U
• PH
T;
c
o
u
(0
0)
.a
T3
^-4
D
O
£.
01
JJ
C
fl\
V
JJ
•rH •
£ a
= .-i
o .c
O 01
u
£ 0)
U 0
0) S
jj 0)
1 £
o<
c u
o o
>j: *j
o
01
u
0
2
es
0)
a
3
O
u
o
^
u
c
3
C1
c
o
&
5
it
M
0)
O
u
2
jj
o
JJ
C4
3
O
M
C3
^
U •
O 4)
3 Oj ,~
E JJ
O •-*
U £
U £
0
0) CJ
•l-l
JJ C
0
jj ••-(
U JJ
0) rtj
!a U
•1-4 a>
T3 4)
JJ
J= C
01 M
•r-l
— 1 01
J2 U
nj -H
JJ X
01 O
U E-
O
jj
O
c
4)
U
it3
S-i
o
41
>
»H
o
03
4)
U
0
JJ
O1
c
o
,— 1
0
0
JJ
4)
J«i
iT3
JJ TJ
O
03 >
CJ i—*
£ 1=0
c1 ! en M
1 1 -^1 ! 'fl 0)
1 i —
—
Q
M U
0
•
-------
CHAPTER FOUR: REGIONAL AND INTERGOVERNMENTAL ISSUES
A. Introduction
Given the extent to which EPA exercises its mandates
through state and local governments, we should expect that
any major change in the way the Agency develops strategies
for the control of toxics would have important effects on
those governments. Three such effects can be distinguished:
o Analysis may determine that a particular toxic
problem is restricted to one state, or a small
group of states, and that state action is more
appropriate than Federal action.
o An area might be selected for analysis under the
geographic approach, requiring a realignment of
staff and resources by both the region and the
state(s) concerned.
o Strategic changes based on work group recommendations
may suggest changes in state programs.
Analysis of these potential effects is, of course, only
in the preliminary stage, as we await policy guidance on the
extent to which the major reforms recommended elsewhere in
this report are to be carried out. The intergovernmental
work this year focused on identifying those broad problems
of communication, procedure, and funding that may cause
difficulties when we attempt to put into practice the inte-
grated strategies described here. The results of this work
are summarized in Supporting Document 15.
The basic problem stems from the fact that EPA's regional
offices reflect the same program-by-program structure and
approach as EPA's headquarters; most states have similar
organizations. Where programs are split among different
state agencies, no regional resources are dedicated to assur-
ing that crossprogram considerations are addressed when
carrying out the several toxic control programs. Regional
offices and states are able to obtain chemical-related
information only on a single program, single medium basis.
There is no EPA mechanism in the field to disseminate
multi-media information (e.g., priority multi-media pollutants)
or multiple program directives on the same chemical; similarly,
there is no mechanism for states to share multi-media concerns
or information with EPA. An examole of what could be done
-------
rv-2
in this area is the recent development of the Intermedia Priority
Pollutant (IPP) list, by the Office of Toxics Integration for
the Toxics Integration Strategy Committee.
B. The Intermedia Priority Pollutant (IPP) List; A First
Step Toward Improving Intergovernmental Integration
Regional responsibilities—encompass both the active
management or oversight of planning, permitting, and enforce-
ment programs, and the responsibility for handling all sorts
of inquiries and emergencies. The latter range from simple
requests for information to full-scale response to toxics
waste and other spills.
Over the last few years, toxics issues have reached high
priority in regional activities, just as they have everywhere
else in the Agency. But the nature of regional responsibilities
imposes burdens not felt in headquarters. Regions have to
take daily actions—such as making permit decisions or making
commitments to deal (or not to deal) with alleged emergencies.
Where toxics are concerned, taking these actions in an intelligent
and responsible way is exceedingly difficult. Regions
must set rational priorities for action despite the enormous
number of compounds at issue, with only limited understanding
of almost all of them, and in the face of public fear.
One of the first recommendations of the Toxics Integration
Strategy Committee was to assemble a list of priority chemicals
around which to organize regional priorities. The principal
benefit of such a list was to focus regional priorities on a
manageably short list of substances that, as far as could be
determined, present the greatest exposure problems nationwide.
This list would be backed up by comprehensive and consistent
scientific and policy statements covering all programs and
media. Its proposed major feature was that it be periodically
updated to reflect current program priorities: chemicals
were to be retired from the list before new ones could be
added.
This last point is important. ?To mechanisms now exist for
removing problems from Agency priority lists. The result
has been a steady accretion of lists of "priorities" whose
mere length invalidates the term. An important concept
behind the IPP list is that once the regions have largely-
discharged Agency responsibilities in relation to a specific
compound (e.g., have cleaned up most PCB wastes or have
written appropriate permits to deal with a particular chemical),
the compound should be replaced on the list with another
whose net risks have by then become greater.
-------
IV-3
Action this year
Project staff have moved quickly to put such a list into
place. The chemicals for the initial list have been selected
and the first support documents written, and guidance for
applying the list has been written into the FY 82 Operating
Plan.
Selecting the list
For a chemical to be included in the list, the principle
of "objectivity" implied that (1) it would have to exhibit
significant toxicity along one or more routes of exposure,
(2) it would have to be widely distributed in the environment
in sufficient concentration to cause excess exposure, and
(3) the Agency would have to have authority to effect controls,
As a practical matter, the selection process weighed three
factors:
o the extent of pre-regulatory and regulatory interest
in EPA
o total volume of national production (an admittedly
crude proxy for exposure)
o current regional interest
The final list appears in Table IV-1 (at the end of this
chapter) along with a brief justification for the selection
of aach entry. This list was necessarily assembled quickly,
using imperfect information. It will be reviewed and amended
over time to improve its objectivity.
Supporting documentation
So far, complete background documents are available on
seven chemicals: arsenic, asbestos, benzene, cadmium, lead
mercury and toluene. The remaining ones will be completed
by October, 1981.
For each chemical, the support document contains succinc-t
and complete information on the following:
o physical/chemical properties of all species of
significance,
o health effects information at acute and chronic
dose levels,
-------
IV-4
o environmental effects information in all media
and biological compartments of concern,
o environmental release data by source, use, and
medium (kkg/year),
o complete references to existing data bases,
o -caction level recommendations for all known con-
tingencies, and
o personal contacts (name, FTS *) for interpretation
and expansion of all of the above.
Assembling the documentation for the IPP chemicals has
also contributed to integration activities at headquarters.
Information gaps, such as the unknown precision and accuracy
of certain monitoring technologies, have come to light.
Inconsistencies among program offices, such as the different
standards set for arsenic in surface water versus drinking
water, remain to be resolved. We conclude that the creation
of authoritative documents on these chemicals will, over
time, be a useful discipline in itself.
Implementing the list
Guidance on the use of the IPP list within each of the
programs is contained in the FY 1982 Operating Plan. Staff
from the Office of Toxics Integration worked with each of
the program offices to be sure that references to the list
exist in each of the areas where it will be useful. In
addition, staff visited two regions to evaluate the quality
of the initial guidance packages.
These two regions indicated that the materials as prepared
will be particularly useful in responding to toxic contamina-
tion incidents involving either hazardous waste sites or
drinking water (including groundwater). They will also be
of use in designing regional compliance programs, surveillance
and analysis (monitoring) programs, and in implementing
consolidated permits. Lastly, they will be beneficial in
responding to inquiries from the public, news media, and
environmental groups.
C. Recommendations
1. An integration function should be established in each
region in the Office of the Regional Administrator to (1) help
-------
IV-5
coordinate integrated control efforts (e.g., coordination
of the regional program on Intermedia Priority Pollutants)
and (2) manage the exchange of multi-media information (e.g.,
especially the Chemical Substances Information Network [CSIN])
capability. This function should also develop and monitor
integration plans for each regional office, and assist the
states in developing toxics integration plans.
•cT~ 2. The Toxics Integration Staff should work with the
Office of Intergovernmental Liaison and the regions on a
joint effort to improve technical assistance to state
and local government on toxic integration issues (please
note the recent National Governors' Association Report, in
which this was the major request/recommendation, for further
detail regarding perceived state needs). Regional offices
should have the lead role in developing and operating geographic
projects, though the headquarters Toxics Integration Staff
should continue to develop the methodology for this type of
analysis. In addition, regional offices should be represented
on chemical and industry work groups.
-------
TABLE IV-1
INTERMEDIA PRIORITY POLLUTANTS
Name
INORGANICS
Arsenic
Asbestos
Cadmium
Lead
Mercury
ORGANIC CHEMICALS
Acrylcmitrile
Justification
Arsenic is associated with adverse effects on the skin,
liver, kidney, blood, and reproductive, nervous, and
respiratory systems. It is designated a carcinogen
by GAG. Arsenic is a semi-metallic solid that sublimes
on heating. It is present as a contaminant in many
nonferrous and coal ore bodies.
Asbestos is a mineral that is known to cause a variety
of respiratory and gastrointestinal cancers. Because
of its widespread use as a fire retardant and because
asbestos is present in a number of ore bodies, human
exposiure can be significant.
Cadmium is a soft metal that occurs in small quantities
with zinc and lead. It is associated with kidney
dysfunction and is implicated as a carcinogen. It is
of ma.jor concern because of its tendency to concentrate
in foodstuffs.
Although most major sources are now regulated, lead
is so stable that a considerable quantity still
exists in the environment from earlier dispersive
uses. Numerous adverse health effects have been
identified, including hematologic, neurologic, and
renal effects. For these reasons, and because its
health effects tend to concentrate in children,
lead is still of considerable concern.
Mercury is the only liquid metal; it is relatively
volatile and accumulates in aquatic systems, where it
can be converted into extremely toxic organic forms.
Mercury causes both central nervous system and kidney
damage. Its mobility and tendency toward bioaccumulation
combine with its health effects to result in continuing
concern.
AN is principally a raw material used in the manufacture
of various widely used polymers and resins. Production
volume (1.5 billion #) is only moderate compared to many
organics, but it is volatile and soluble in water. Re-
leases to the environment can therefore be substantial.
AN is associated with cancer deaths among exposed workers,
and it has caused significant increases in tumors in lab-
oratory animals; CAG has designated AN as a carcinogen.
-------
TABLE IV-1 (cont'd)
INTERMEDIA PRIORITY POLLUTANTS
Carbon
Tetrachloride
Chloroform
Chlorinated
Solvents
para-
Dichlorobenzene
Dioxin
Justification
Benzene is a high volume chemical that is an increasing
constituent in gasoline and is used as a raw material
in the manufacture of several important chemicals.
Benzene is a volatile liquid but only slightly soluble
in water. Benzene is a known leukemogen and causes
other human blood disorders. It is designated a car-
cinogen by GAG.
Carbon tetrachloride is a moderate volume chemical that
is principally used as a raw material in the production
of fluorocarbons. It is a volatile liquid but only
slightly soluble in water. It is designated a carcinogen
by CAG and is also known to result in liver, kidney and
neurological damage in animals.
Chloroform is also a moderate volume chemical used
principally in the manufacture of fluorocarbons, but
also as a solvent. Chloroform is a volatile liquid but
only slightly soluble in water. It is associated
with an elevated incidence of bladder cancer in people
drinking water with elevated levels, and it has been shown
to induce tumors in animals. It is designated a carcino-
gen by CAG.
Over 2 billion pounds of chlorinated solvents are manu-
factured each year, virtually all for dispersive use
These solvents are volatile but not very soluble in
water. They exhibit a range of adverse health and
environmental effects and their potential for inter-
changeability lead to their inclusion in this group.
para-Dichlorobenzene is a low volume chemical principally
used for moth control and as a space odorant. As such,
human exposure can be high. This chemical is a solid
at standard conditions but is volatile, as evidenced by
its use. It is also slightly soluble in water. It
is a leukemogen and is metabolized to substances that
are carcinogenic as a class.
A contaminant and combustion product in some systems,
dioxin is one of the most toxic substances known. It
is also a carcinogen and a teratogen. Although its
total release to the environment is snail, it is so
potent as to be of considerable concern.
-------
TABLE IV-1 (cont'd)
INTERMEDIA PRIORITY POLLUTANTS
Name
Justification
Ethylene
Dichloride
Formaldehyde
Polychlorinated
Biphenyls
Polycyclic
Organic Matter
Toluene
EDO is a very high volume (11 billion #) chemical that
is used both as a raw material and an end product* As
such, environmental release and human exposure can be
high. EDC is a volatile liquid that is not very soluble
in water. It is designated a carcinogen by GAG.
Formaldehyde is a very high volume chemical that is used
principally as a raw material for plastics and
adhesives, but also has a number of consumer uses.
Is is also a product of combustion and results from
some natural processes. It has been implicated in
various instances of consumer illness, and has been
designated a carcinogen by GAG.
Although production in the USA has ceased and non- •
contained uses tightly regulated, some products con-
taining PCBs are still in use and accidental releases
can occur. PCBs are associated with several well-
known health and environmental effects.
l>
POM consists of a large number of different compounds,
many of which are known carcinogens. They are produced
during the combustion of most complex organics, most
notably during inefficient combustion processes.
Total production volume is not precisely known but
certainly exceeds many millions of pounds per year.
Toluene is the largest-volume organic chemical produced
(67 billion #). It is a volatile liquid that is only
slightly soluble in water. Although there is no direct
adverse health effects data at this time, toluene is
of concern owing to its high volume and because it is
structurally very similar to benzene, a known leukemogen.
-------
CHAPTER FIVE: SCIENCE AND RISK ASSESSMENT
A. Introduction
As noted in Chapter I, EPA has had difficulty in achieving
consistency and uniform quality in the risk assessments it
produces, a difficulty that stems directly from the historic
division of responsibilities among offices carrying out
specific statutes. Since scientific inquiry at EPA is largely
the handmaiden of regulation, differences have appeared over
time in the scientific or technical approaches used by the
different parts of the Agency.
Although the use of scientific data is contingent
on policy, this chapter does not address policy-level issues.
It is concerned instead only with identifying variation that
may have some effect on the kind and quality of information
given to policy makers, and with offering options for controlling
such variation and eliminating it where it is found to be
indefensible.
Differences in approach between offices
The approach any organization takes to risk assessment
varies according to the intended use of the results, as
established by statute or policy. In the case of determinations
of carcinogenicity, for example, some program offices are
supposed to take regulatory actions or develop criteria and
guidelines as a result of the determination of a hazard
finding. The control of carcinogenic air pollutants under
Section 112 of the Clean Air Act, the listing of carcinogens
for regulation as hazardous wastes, and the development of
water quality criteria are good examples. Other program
offices, in contrast, are authorized to take into account
the strength of the evidence for carcinogenicity and balance
this against the costs of controlling some substance; the
pesticide program, for example, uses this approach.
In general, where qualitative characterization of a
substance as a carcinogen triggers regulatory machinery,
there is substantial pressure to make conservative assumptions
in conducting assessments. On the other hand, where action
may be scaled to the strength of evidence, and quantitative
degrees of risk can be established and related to control
costs, assessments will usually include data from studies of
varying quality and relate more or less intuitive valuation
of risk to the regulatory decision maker. Program offices
-------
V-2
that operate under different policies in regard to what a
finding of carcinogenicity means may develop quite different
evidentiary criteria for such findings.
Multimedia risk assessment
The total risk imposed by a substance that moves between
environmental media has typically not been studied by program
offices focused on a single medium. For example, organic
compounds released to water may have their most serious
effect after volatilizing to the air.
The approaches used by the Office of Toxic Substances (OTS)
and (to a lesser extent:) by the Office of Pesticides Programs
(OPP) are exceptions. OTS primarily evaluates chemicals in
commerce (or associated by-products and contaminants) and
their effects, considering all routes of exposure. OPP does
not consider all environmental releases of pesticides but
only those related to use. However, multimedia assessments
are conducted for the releases considered. The Toxic Substances
Control Act (TSCA) and the Federal Insecticide, Fungicide,
and Rodenticide Act (FIFRA) give these two offices substantial
authority to deal with toxic substances in an integrated
fashion. Other efforts to conduct multimedia assessments,
such as that of the Solvents Work Group (see Chapter III),
ara the exception rather than the rule.
Although multimedia risk assessments provide insight and
information related to pollution control and the upcoming
year's guidance directs SPA to perform them, current legislative
mandates restrict the courses of action that could result
from using such work. For example, a court order to
control a substance in a given medium would be compelling,
even if we knew, through multi-media assessment, that control
in another medium would reduce risk more or cost less for
the same risk reduction. On the other hand, if the Agency
had convincing multi-media assessments prepared, it might
be far less vulnerable to the kind of suit that produces
such orders. Multi-media analyses of risk could also contribute
to the debate around reform of environmental legislation,
since they provide a broader understanding of intermedia
pollutant transfers (and the impacts of regulatory controls
on these transfers). For the present, they could suggest
more economically sensible control strategies within current
statutory mandates.
-------
V-3
Previous and current coordination activities
Despite inherent differences in approaches, the Agency
has attempted to achieve general consensus on the technique of
risk assessment. In 1976 the Carcinogen Assessment Group
(GAG) developed guidelines for cancer risk assessments. In
1979, GAG was expanded to form the Office of Health and
Environmental Assessment (OHEA). In the same year, the
Interagency Regulatory Liaison Group (IRLG) published carcinogen
assessment guidelines, which describe criteria for the evaluation
of carcinogenic evidence from human populations, bioassays,
and short-term tests. In 1980, OHEA proposed Agencywide
exposure guidelines to estimate exposure to toxic substances
and mutagenicity risk. Today the Toxic Substances Priorities
Committee (TSPC) provides coordination among offices to
minimize duplication of major efforts related to assessment
of toxic substances.
But none of the coordinating bodies established in the
Agency has the authority to require that Agencywide standards
or procedures be applied to individual scientific assessments.
Managers and staff scientists often are reluctant or unable
to criticize work done in peer organizations, which are
typically designed to treat only a narrow range of issues.
Also, different programs are geared up to work on particular
substances at different times. A request by one office for
peer review will often not solicit a helpful response unless
the reviewing office happens to have the requisite expertise
at that moment. When two offices work on the same chemical
successively, the usual course is for the second office to
repeat the work with a different emphasis. Occasionally,
this results in two or more varying assessments of risk or
hazard on the same substance; at minimum there is duplication
of expensive work.
In addition, the Agency's extramural peer review policies
vary from program office to program office, largely because
of differences in legislative mandates. For example, OPP is
required to submit all major scientific studies to a Scientific
Advisory Panel. In contrast, OTS is under no legislative
obligation to do so, although it is now developing a review
process for its own use.
This year, the Toxics Integration Project engaged a
contractor to examine risk assessment in the various offices
that perform it to determine what significant difference in
approaches exist. (Supporting Document 12 is a summary of the
contractor's report.) What follows is an evaluation of the
meaning and consequences of the variations discovered, with a
discussion of several options for increasing the consistency
of risk assessment conducted by the Agency.
-------
V-4
B. Survey findings
The information presented below summarizes the findings
of the contractor's study, as interpreted by the EPA staff
that reviewed the data the contractor collected. As with
any survey based on written responses to subjective questions,
the results may be skewed by differences in respondents'
interpretations. We have attempted to eliminate the worst
errors through of program office reviews, but subtle mis-
statements undoubtedly remain.
Where real differences between programs were found, the
timetable of this year's TIP study did not permit elaborate
analysis. Only limited consideration of the causes of differ-
ences was possible, and in areas where we encountered many
divergent views, more study is warranted in the coming year.
1. Variance in risk estimates among offices
Carcinogenic effects
Of the nine offices responding) most conduct assessments
of both carcinogenic hazard and risk. The Office of Air
Quality Standards and Planning (OAQPS) performs risk assessments
only, while the Office of Solid Waste (OSW) does not conduct
cancer assessments of any kind.
All of the offices that perform cancer hazard and/or
risk assessment are aware of EPA or IRLG guidelines. The Office
of Drinking Water (ODW) reports that most of its cancer
assessments are performed by the National Academy of Sciences
(NAS) or by GAG.
Despite awareness of EPA and IRLG guidelines in all the
offices studied, there are still apparent differences in
practices. For example, IRLG guidelines recommend that, in
the absence of definitive evidence that a chemical is incapable
of inducing malignant tumors, tests that show statistically
significant increases in benign tumors be taken as positive
evidence of carcinogenicity. The Office of Water Regulations
and Standards (OWRS) and OPP follow this approach, while
OTS generally considers such data only to indicate the need
for further testing. ODW examines such results on a case-by-
case basis. Clearly all of the offices interpret such findings
as indicative of carcinogenic potential? differences appear
to be in whether further confirmatory data should be sought.
Most of the deviations from IRLG guidelines appear to be of
this sort, and appear to be mainly influenced by the program-
matic context of the assessment.
-------
V-5
Another difference in carcinogen assessment procedures
is the use of correction factors applied to animal test
results for animals exposed for less than a full lifetime.
All offices apply such factors except ODW, since NAS, which
conducts many of ODWs risk assessments, does not use them.
A final area of apparent difference among the offices
relates to the pooling of data from separate studies in
making findings of carcinogenicity. QHEA does not pool
data, while OWRS, OPP, and OTS occasionally do. Further study
in this area appears necessary.
Mutagenicity and reproductive effects
Several differences exist with regard to application of
mutagenicity risk assessment. One is the manner in which the
various offices define the role of such assessment. Although
most agree with the underlying concepts presented in the
Proposed Mutagenicity Risk Assessment Guidelines, and thus
emphasize both somatic cell mutagenicity and germ cell
mutagenesis, Certain offices (ODW, OSW, OAQPS) place a much
greater emphasis on mutagenicity as supportive evidence for
carcinogenicity and have n'ot used heritable mutagenic risk
as a separate toxicological endpoint for regulatory purposes.
Variations also exist among programs regarding testing
strategies used, approaches for presentation and categorization
of evidence, and so on, but the key finding of the contractor's
study is that there are no Agencywide criteria for judging
the severity of reproductive and developmental effects.
Evaluations of individual studies are presented in various
levels of detail, and ultimate conclusions are stated in
various ways, or sometimes not at all.
Other chronic health effects
As far as we know, only a few offices have recently been
active in other chronic health effects (e.g., liver, kidney, or
respiratory system damage) risk assessments. OHEA conducts
chronic non-cancer risk assessments to support OAQPS in
developing National Ambient Air Quality Standards (NAAQS). The
focus is on defining populations that are unusually sensitive
to certain problems such as asthma or bronchitis, and deter-
mining the pollution levels that produce adverse effects.
OPP conducts chronic health effects risk assessments to
support the registration or reregistration of pesticides,
and to set tolerances for pesticide residues on food and
animal feeds. ODW examines non-cancer risk when developing
Suggested No Adverse Response Levels (SNARLs) and Maximum
-------
V-6
Contaminant Levels (MCLs) for drinking water. No Agencywide
guidance is available to support any of these efforts.
Summary! Implications of differences in risk assessments
It appears that cancer assessment is the most consistent
type of health effect assessment. Few mutagenicity assessments
have been performed, but Agencywide guidelines have at
least been proposed. Agencywide standards for other types
of effects are absent.
Different models used to estimate carcinogenic risk as a
function of dose level can lead to different conclusions
regarding cancer risk. Although conflicting results are
rare, interpretation of these results may cause problems in
policy making 1 Differences in statistical analytic techniques
may lead to different regulatory decisions, while acceptance
of different test results can still lead one program office
to regard a compound as a carcinogen while another does not.
"The same issues ap>ply to mutagenicity testing, only more
so. Here the Agency's relative lack of experience, coupled
with incomplete implementation of the new testing guidelines,
are likely to lead to conflicting technical findings and
inconsistent regulatory policy. The situation is similar with
regard to chronic health effects not related to cancer or
mutagenicity.
2. Variance in exposure assessments among offices
Most of the progrson offices use the Exposure Assessment
Guidelines recently developed by OHEA. The exceptions are
ODW and OSW: these two do not follow any written guidelines,
but do observe basic principles inherent in the OHEA docu-
ments .
Most of the program offices consider all sources of
exposure in the assessment, though they tend to emphasize
the pathway they control. There are exceptions: OP? con-
siders only sources of exposure related to pesticides use;
ODW is concerned primarily with the occurence, rather than
the sources, of contaminants in drinking water; and OSW is
primarily concerned with the release of toxics from sludge.
In these cases a more global consideration of toxics sources
appears to be inconsistent with individual program mandates.
All program office's consider ambient data and data on
contaminant levels in humans as important components of
-------
V-7
exposure assessment. Most also rely on environmental models
to predict ambient concentrations: ODW is the exception,
relying mainly on drinking water data drawn from a national
statistical sample. The use of surrogates to estimate exposure
is universally considered a last resort.
Most offices calculate both average and worst case
exposure, and separately estimate exposure for high risk
population groups. Program offices also tend to use the same
values for commonly used parameters; where there are differ-
ences, they appear to stem from differences in policy rather
than scientific disagreements. A number of potentially
significant differences do, however, appear to exist among
the more difficult parameters to estimate, such as absorption
rates through skin, lungs, and GI tract, as well as types of
food consumed. Further work is needed before we can determine
if such differences have a scientific basis.
The underlying principles used by program offices to
conduct exposure estimates appear to be uniform, although
there are rtiany differences in specific procedures. The
greatest potential for inconsistency is in the values of
factors (including body weight) used to estimate whole body
dose to humans.
3. Variance of review procedures among offices
Procedures for reviewing assessments differ depending upon
the nature of the assessment and the scope of responsibility
of the assessment office. Assessments related to priority-
setting receive less review than those related to regulations,
though this varies. Assessments supporting regulations
receive extensive review within the program office, within the
Agency, and in some cases by external groups.
Regulatory requirements for external peer review vary
among programs. Under the 1980 FIFHA amendments, all major
scientific studies conducted by OPP must be reviewed by a
Science Advisory Panel. The Safe Drinking Water Act requires
EPA to use NAS studies to identify drinking water contaminants
and toxicological effects, but the NAS work does not undergo
peer review. TSCA does not explicitly require EPA to conduct
peer reviews of scientific studies, but OPTS, as already
noted, is in the process of establishing certain peer reviews
anyway.
In general, however, Agency policies toward peer reviews
are inconsistent and fragmented. Although it was not reported
-------
V-8
in the survey itself, we have found that programs are experien-
cing serious delays in obtaining reviews from panels such
as the Science Advisory Board, and a number of offices believe
that the Agency assigns too few staff to this important
function. The variability of existing review procedures opens
the Agency to criticisms related to the quality of exposure and
hazard assessments. Given the large number of test methods
and models EPA employs, adequate technical review for health
effects assessment is extremely important.
These assessments rely heavily on scientific judgment.
Without full consensus review by professional peers, policy
conclusions and regulatory policy are easily challenged.
C. Options and Recommendations
The above section identified a number of differences in
various aspects of scientific assessments conducted by the
program offices in the Agency, as well as problems that
arise from the distribution of scientific inquiry among
independent program offices. However, the Scientific
Assessment workgroup established for this Project could not
reach agreement on a number of aspects of this study in the
limited time available. Supporting Document 12 contains
statements by the program offices that address these issues.
Consequently, the resolution of what the differences mean
requires an analysis beyond the scope of this study.
The options offered here and displayed in Table 7-1
(at the end of the chapter) are designed to provide three
increasingly centralized ways of addressing the same set of
objectives: (1) avoiding duplication of effort and concentrating
scientific resources on problems of Agencywide high priority;
(2) establishing and applying Agencywide policies for conduct
of all scientific assessments; and (3) increasing efficiency
and, where appropriate, uniformity of approach in conducting
all assessments.
Under all options, ensuring that individual assessments
are consistent with Agency policy and guidelines would be
accomplished by giving the Science Advisory Board (SAB) the
authority and the resources to provide consistent peer
review for significant Agency assessments and by establishing
stronger guidelines and procedures governing scientific
review of assessment documents. (This step has already been
initiated by the Deputy Administrator in his August 1981
memo.) Increased EPA staff support for the SAB, and possibly
some expansion of SAEi 's membership, would be essential to
accomplish this. It is necessary to reconcile the SAB's
-------
V-9
role with that of the Science Advisory Panel, whose review
is required under FIFRA, and with the statutory role of
the NA~S in establishing drinking water standards, so that
the Agency gets consistent peer review to support consistent
approaches to risk or hazard assessment.
Option 1; Modified decentralized approach
Under this option, preparation of hazard, exposure, and
risk assessments would remain decentralized. The most signi-
ficant changes under this option would be (1) the assignment
of responsibility for preparing recommendations on priorities
for multi-media assessments to the Toxics Integration Staff
within OPRM, and (2) the use of a central policy/review
office to review all significant Agency hazard assessments,
to provide assessment guidelines, and to enforce compliance
with scientific policy guidelines. The central policy/review
office would also make recommendations on such Agencywide
policy issues as: what safety factors should be used; what
criteria decide when a problem must be handled via a multimedia
assessment; and which end-points should be used to define
various health effects.
The Toxics Integration Staff office would, under this
option, rank multi-media toxics problems; make recommendations
on which assessments should be done first-and which program
offices ne-ed to take part in them; get commitments from the
program offices to perform them; and act as secretariat for
multi-media assessment teams performing the work and report
on progress to the AA for OPRM.
Advantages
o Little disruption of Agency organization and
procedures would occur.
o Scientific assessments would be based on a single
set of policies.
o The existence of a Toxics Integration Staff would
help insure that priority multi-media assessments
-------
V-10
were given appropriate attention.
o A single uniform review procedure would enforce
adherence to Agency guidelines.
Disadvantages
o It will be difficult to adopt a single set of policies
that reflect the specific program requirements of
each office, and handling exceptions would consume
substantial staff time.
o The length of time required to perform reviews could
slow the production of assessments, particularly if
inadequate resources are allocated to the function.
Option 2: A single organization for hazard assessment
This option would implement Option 1, but give the central
policy/review office the responsibility for performing to
perform all hazard assessments for the Agency. Exposure
and risk assessments would remain decentralized, as they are
now. The central office would support laboratory studies
for hazard assessments and would maintain associated data
bases. It would also review risk" assessments and set
guidelines, as in Option 1.
The toxics integration office in OPRM would be responsible
for coordinating preparation of multimedia exposure and risk
assessments, as in Option 1.
Advantages
o Centralized hazard assessment activities would
increase consistency in assessments.
o Duplication of effort, would be minimized.
Disadvantages
o Responsiveness to program office needs would be
reduced. This is particularly true in cases of
emergency responses encountered by the Office of
Drinking Water and Office of Emergency Response.
o Substantial disruption and delays would result.
-------
V-ll
o A centralized hazard assessment group would be
isolated from the practical impacts of its
conclusions.
Option 3; Combine all hazard, exposure, and risk assessment
with all economic analysis in a single organizational unit
This option proposes placing in a single organization
all Agency resources capable of assessing hazard, exposure,
and risk and relating reduction of risk (or other benefits
of pollution control) to the economic impact of regulation.
Such an organization would have to subsume the responsibilities
for calculating risk, benefits, and economic impact now in
OTS, OPP, OAQPS, OWRS, ODW, and OSWER as well as those lodged
in ORD and OPRM.
Advantages
o A single organizational unit would be responsible for
the entire cost-benefit analytic procedure, ensuring
consistency of like functions Agencywide, as well
as a good fit between risk analysis, benefits assess-
ment, and cost and economic impact analysis.
o Duplication of effort would be reduced or eliminated.
o The problem of conflicting risk, benefit, or economic
impact assessments would be eliminated.
o Some economies of scale might be realized if various
data collection efforts and data bases were centralized,
Disadvantages
o Responsiveness to program office priorities and schedules
would be more of a problem than under Option 2.
o Exposure analysis involves different skills and
disciplines in each of the environmental media.
As coordination is not a noticeable problem
here, no apparent benefits would result from the
consolidation of this activity. Such a move
could, however, weaken the program offices by
removing personnel that contribute to such
activities as monitoring policy and control
analysis, which are dependent on knowledge of
exposure assessment.
-------
V-12
There seems little point in centralizing cost and
economic impact analysis across the Agency since
consistency across programs has not been raised
as a major issue.
This option risks repeating the experience of the
Agency's earlier years, when program managers made
regulatory decisions without due consideration of
economic impacts because such analysis was not
performed by their own offices. For this reason,
economic staffs were consciously placed within
program offices in recent years to force program
managers to take responsibility for considering
economic factors as early as possible in the
development of regulatory approaches.
This option would require major reorganization —
there are 160 Agency work years devoted to cost
and economic analysis alone. Disruption and
delay in many vital Agency functions would be
almost inevitable.
Recommendation II; Adopt Option 1 — modified decentralized
approach — on a trial basis.
The inconsistencies and duplications detected in EPA
scientific enterprise and described in this report are quite
troublesome, but not crippling. A combination of enhanced
peer review and enforced guidelines should go far towards
eliminating them. The selection of this option will pro-
duce more uniform and consistent scientific assessments
without the major organizational disruptions and possible
delays or bottlenecks in the central analytic units
established by the other options. Should the trial prove
unsatisfactory, the more extensive reorganizations called
for in the other options could be adopted.
The concept of work groups advanced elsewhere in this
report is a way of getting the parts of a basically decentral-
ized organization to work together effectively. We have
stressed the problems attendent on decentralization; we
should remember that decentralization has some advantages
in the field of scientific research. Science thrives on
controversy, and having a number of centers of scientific
-------
V-13
expertise within the Agency is a positive good, as long as
they agree on standards of quality and as long as there is
some mechanism for reviewing all work in a particular area
and settling differences before any scientific conclusion is
used in policy-making. These conditions should be met through
the adoption of Option 1 and the other recommendations that
follow.
Recommendation f2; The central risk assessment policy/
review office should resolve major scientific differences
between offices"!
The central policy/review office recommended in Option 1
should, as part of its review and oversight function and
under the guidance of the Deputy Administrator/Science
Advisor, resolve the major differences between program offices
presented in the contractor's report within six months of
approval of Option 2. These would include recommendation
on: what constitutes a finding of carcinogeneity; correction
factors applied to animals exposed to carcinogens for less
than a full lifetime; policy on pooling of data from different
cancer studies; policy on heritable mutagenic risk as a
toxic endpoint; and other, more technical, differences revealed
in the report.
Recommendation »3; The 'central risk assessment policy/
review office should develop chronic health effects
guidelines and certify program methodologies.
The central policy/review office should develop guidelines
to fill gaps in scientific assessment procedures. Specifically,
within 30 days of adoption of Option 1, the central policy/review
office should submit a plan and schedule for developing and
transmitting chronic health effects assessment guidelines
to the Associate Administrator for Policy and Resource Management.
Within 90 days of the adoption of Option 1, the central policy/
review office should also submit a .program plan for certifying
methods and procedures that meet minimum criteria established
in all accepted Agency assessment guidelines. The program
plan will include policies for acceptance of new methods and
procedures and deletion of obsolete ones.
Recommendation |4; The central risk assessment policy/
review office should design review procedures. "~
Within 90 days of the adoption of Option 1, the central
-------
V-14
in all accepted Agency assessment guidelines. The program
plan will include policies for acceptance of new methods and
procedures and deletion of obsolete ones.
Recommendation #4; The central risk assessment policy/
review office should design review procedures.
Within 90 days of the adoption of Option 1, the central
policy/review office should submit to the Associate Administrator
for Policy and Resource Management procedures for reviewing
and processing scientific assessments conducted by the Agency
to ensure that applicable guidelines are followed in all
Agencywide assessments.
-------
Function/Option
TABLE V-l
OPTION MATRIX
Option 1
Option 2
Option 3
Sets priorities &
policies
Conducts exposure
assessments
Conducts hazard
assessements
Conducts risk
assessments
Reviews assessments
Develops
assessment
guidelines
Conducts lab studies
for exposure and
risk assessments
Maintains data base
for exposure and
risk assessments
Toxic Integra-
tion Staff
(OPSM)
Program
Program
Program
1. Central
office
2. SAB/SAP
Central
office
Program
Program
Toxic Int.
Staff
(OPSM)
Program
Central
office
Program
1. Central
office
(other
than
hazard
assessments)
2. SAB /SAP
Central
office
Program
Program
Central
office
Central
office
Central
office
Central
office
SAB /SAP
Central
office
Central
Office
Central
office
-------
CHAPTER SIX: CONTRACTING
A. Problem Statement
The Agency does not now have an effective system for
preventing the funding of overlapping and duplicative studies.
Not only is this wasteful of resources, it also encourages
contradictory and ineffective decisions later in the regulatory
process, decisions that are all the harder to reverse because
of the scale of the investments made in them.
A few of the more specific problems are:
o The Agency does not keep track of the information
it has purchased under contract, nor does it
effectively monitor the content of contracts in
progress to avoid unnecessary duplication of effort
across programs.
o There is no central Agency file of existing contractor
reports arranged by subject or topic. This also
results in contract duplication. Contract results
are rarely used outside of the initiating program
office, and contracts that use the same original
sources frequently reach inconsistent conclusions.
(See Supporting Document 14 a for case s-cudy review
of how arsenic and 1,2 Dichlorothane have been the
subject of overlapping Agency contract studies.)
o Because programs cannot determine routinely what
information is available on particular chemicals
and industries and what conclusions have been
reached previously on the basis of that information,
new initiatives cannot build on existing reports,
hence are often constrained to repeat the same
preliminary groundwork.
o No mechanism makes it easy to add unanticipated
"emergency" research needs to the most appropriate
and efficient in-place contract vehicle, especially
if the most convenient contract vehicle is in a
different program office.
B. Discussion
To correct these management concerns, the Agency should:
o Keep track of the content of current and completed
contract studies. Provide Agency staff responsible
-------
VI-2
for contracting with extramural activity reports that
describe the nature and scope of current and planned
contracts. Organize contract information in a
system accessible by subject category.
o Evaluate the feasibility of similar procedures for
grants and cooperative agreements.
o Develop administrative procedures ensuring this contract
information is used to prevent wasteful overlap or
duplication in proposed contracts.
The Office of Research and Development (ORD) has
already instituted a major project tracking system around
which the required Agency system can be built. Implemented
in 1977, the ORD Information System (ORDIS) includes technical
and contract data for most ORD projects. The data base is
generated from 2-page standardized project summaries submitted
by ORD project officers and reviewed for quality control
purposes by a small ORDIS technical staff.
ORD contracts represent 50% of the Agency1s contract
budget. Extending ORDIS to the remaining portion of EPA's
contracting budget would have several advantages:
o The Agency would incur minimal additional development
costs. ORD has already invested $600,000 to design
the ORDIS tracking system including software, forms,
and procedures. The system would require only slight
modifications for use Agencywide.
o Merging ORD project information with other programs'
contract data should increase program and ORD
interaction.
o EPA would, as a result, possess a substantial inventory
of information generated by contract, across all
programs.
Whether this contracts inventory must be complete remains a
subject for further consideration. The decision involves an
obvious tradeoff between economy of effort and completeness
of information. An analysis will be completed very shortly
to determine (1) whether there is a practical way to eliminate
those contracts from inclusion in the system that are not
relevant to toxics integration and (2) whether so doing will
save enough money to be worth the bother.
-------
VI-3
Once this system is developed, whether for all or a subset
of Agency contracts, project officers should search it before
submitting a Request for Proposal (RFP) . Whether to actually
require project officers to use the system, and how such a
requirement could be enforced, also need additional review.
The potential burden which such a system could represent for
program officers must be weighed against the potential savings
to the Agency which might result.
Searching this expanded ORDIS system should reveal
whether entire projects or their sub-elements (literature
reviews, particular tests, etc.) have already been done or
are under way. Such a search would also reveal opportunities
to modify existing contracts to perform needed new work
without initiating whole new contracts with their attendant
start-up and overhead costs.
Finally, once a contract is completed the tracking
system should reflect the project's completion, and the
library should receive a copy of the final report. All
contract reports, after some specified date, should include an
abstract. Some subset of reports already on file in the
library may need to be abstracted as well.
C. Recommendation
The Toxics Integration Staff should work with staff from
ORD, Management Information and Data Systems Division (MIDSD),
Contracts Management Division (CMD), and the library to
develop an efficient integrated tracking system for current
and completed contracts. This system should use ORDIS as
its base. 3y November 15, the group should submit a report
to the Associate Administrator for OPRM which:
o considers the feasibility and cost of modifying the
ORDIS system to include those Agencywide contracts
generating information about chemicals and industries
only vs. all contracts, including a recommendation
for action;
o specifies Project Officer, Library, CMD, MIDSD and
Toxics Integration Staff responsibilities in designing,
operating, and evaluating this system, including
abstracting responsibilities as necessary:
o specifies whether criteria are necessary to establish
when a report is final and must be released to the
library;
-------
VI-4
o specifies which, if any, subset, of Agency contracts on
file with the Library should be abstracted, at what
cost, and how that information, should be related to
the ORDIS tracking system; and
o outlines the steps necessary to implement the system,
with submission to the Associate Administrator for
OPRM for review by early January, 1982.
o outlines the steps necessary to implement the system,
in early January, 1982.
Finally, we recommend that final reports from all new
contracts should be provided to the library with abstracts.
'-s : ' ; .-;-:• -tion Agency, j
2" - ,.* . X
/""
.£.,.,
-------
SUPPORTING DOCUMENTATION
1. History of prior integration efforts
2. Priority setting analysis proposal
3. Control options analysis manual
4. Case study on copper smelting
5. Case study on organic solvents industry
6. Geographic site selection methodology
7. General methodology for geographic studies
8. Geographic pilot study: Kanawha Valley
9, The regional perspective on geographic analysis: Report
from Region III
10. Innovative and alternative controls for toxic chemicals
11. Federal laws applicable to geographic studies
12. Descriptive analysis of differences in approach to risk
assessments between program offices
13. Regulating tracking recommendations: defining system
capabilities
14. Analysis of contracting process and recommendations for
improvements
15. Regional and intergovernmental relations
-------
------- |