ANALYSIS OF THE ECONOMIC EFFECTS
OF REGULATORY STRATEGIES FOR
SECTIONS 102 AND 103 OF THE
COMPREHENSIVE ENVIRONMENTAL RESPONSE,
COMPENSATION, AND LIABILITY ACT
A Report to the
Emergency Response Division
Office of the Emergency
and Remedial Response
United States Environmental Protection Agency
November 2, 1981
ICF INCORPORATED 1850 K Street. Northwest,
Suite 950, Washington. D C. 20006
-------
SS.Ssu
Rcsoufca
ANALYSIS OF THE ECONOMIC EFFECTS OF
REGULATORY STRATEGIES FOR
SECTIONS 102 AND 103 OF THE
COMPREHENSIVE ENVIRONMENTAL RESPONSE,'
COMPENSATION, AND LIABILITY ACT
Emergency Response Division
Office of Emergency and Remedial Response
United States Environmental Protection Agency
A Report to the
November 2, 1981
ICF INCORPORATED
-------
PREFACE
The ICF staff for this project have prepared this draft report for
discussion purposes only. It does not represent the policy of the U.S. EPA
and has not been reviewed by EPA or by Rockwell International, the prime
contractor. Please do not quote or cite this draft.
The ICF project staff are pleased to acknowledge the assistance and
support of the Rockwell International staff, the EPA Edison, N.J. ORD
Laboratory staff, and the EPA headquarters staff who have contributed to our
understanding of the issues raised in this analysis. We are especially
grateful to Dr. Bart Tuffly, Dr. Milt Kirsch, Dr. Wayne Stone, Dr. Jack
Kooyoomjian, Mr. James Lounsbury, Mr. Leo McCarthy, Mr. John Riley, and
Mr. John Cross.
The ICF project staff for this report includes Jim Janis, Project Manager;
Gilah Langner, Task Manager; and Carol Andress, Michael Barth, David Bruce,
Frank Lerman, Peter O'Connor and Mary Sexton.
ICF INCORPORATED
-------
TABLE OF CONTENTS
PAGE
EXECUTIVE SUMMARY ES-1
CHAPTER 1: INTRODUCTION
1.1 Purpose of the Report 1-1
1.2 Organization of the Report 1-2
CHAPTER 2: UNDERSTANDING
2.1 Introduction 2-1
2.2 Statutory and Regulatory Background 2-1
2.2.1 Designation of Hazardous Substances - Section 102(a) 2-2
2.2.2 Revision of Reportable Quantities - Section 102(b) 2-3
2.2.3 Notification - Section 103 2-3
2.3 Regulatory Impact Analysis Requirements 2-4
2.3.1 Executive Order 12291 2-4
2.3.2 The Regulatory Flexibility Act 2-6
2.3.3 Significant Impact Determination 2-6
2.3.4 0MB Draft Guidance on Executive Order 12291 2-7
2.3.5 EPA Draft Regulatory Impact Analysis Guidelines 2-8
2.3.6 Application of EPA Guidelines to Sections 102
and 103 Regulations 2-9
2.3.7 EPA Guidance on Reporting Burdens 2-12
2.4 Special Topics 2-13
2.4.1 De-Listing 2-13
2.4.2 Radionuclides 2-13
2.4.3 RCRA Wastes 2-14
2.4.4. Mixtures and Pesticides 2-14
2.5 Conclusion 2-15
CHAPTER 3: BASELINE
3.1 Introduction 3-1
3.2 Alternative Baselines.. 3-1
ICF INCORPORATED
-------
TABLE OF CONTENTS
(continued)
PAGE
3.3 Specification of the Baseline 3-3
3.3.1 Statutory Language of CERCLA 3-4
3.3.2 Sections 103(b), 106, 107, 108 and 109 3-4
3.3.3 Sections 104 and 105 3-4
3.4 Characterization of Baseline Activities 3-5
3.4.1 Designation 3-5
3.4.2 Revision of Reportable Quantities 3-5
.-3.4.3, Notification 3-7
3.5 Conclusion 3-7
CHAPTER 4: STRATEGIES AND REGULATORY ALTERNATIVES; EVALUATIVE FRAMEWORK
4.1 Introduction to Framework 4-1
4.2 Guidance on Developing Regulatory Alternatives 4-2
4.3 Strategies and Regulatory Alternatives for Designation 4-4
4.3.1 Rationale for Designation Regulations 4-4
4.3.2 Designation Strategies 4-5
4.3.3 Designation and Reportable Quantities 4-10
4.4 Strategies and Regulatory Alternatives for Revising
Reportable Quantities 4-10
4.4.1 Rationale for Revising Reportable Quantities 4-11
4.4.2 Issues in Designing Reportable Quantity Regulations 4-12
4.4.3 Strategies for Revising Reportable Quantities 4-14
4.5 Strategies for Notification 4-19
4.6 Non-Regulatory Approaches 4-20
4.6.1 Possible Alternatives 4-21
4.7 Conclusion 4-23
CHAPTER 5: ECONOMIC EFFECTS
5.1 Int roduct ion 5-1
ICF INCORPORATED
-------
TABLE OF CONTENTS
(continued)
PAGE
5.2 Guidance on Identifying Effects 5-1
5.3 Effects of Designation Regulations 5-2
5.4 Effects of Revising Reportable Quantities 5-8
5.5 Effects of Notification Regulations 5-13
5.6 Conclusion _ 5-14
CHAPTER 6: MEASUREMENT OF ECONOMIC EFFECTS
6.1 Introduction 6-1
6.2 Approach to Measurement . 6-1
6.2.1 Methods of Measuring Benefits and Costs 6-3
6.2.2 Methods of Measuring the Number of Actions 6-7
6.3 Findings and Analysis 6-13
6.3.1 Information on Benefits 6-13
6.3.1.1 Benefits of Federal Government Involvement 6-13
6.3.1.2 Benefits of Increased Cleanup of Releases 6-16
6.3.1.3 Benefits Resulting From Preventive Activities... 6-17
6.3.2 Information on Costs 6-18
6.3.2.1 Cleanup Costs 6-18
6.3.2.2 Costs of Preventive Activities 6-23
6.3.2.3 Further Investigation of Costs 6-27
6.3.3 Information on Numbers of Actions 6-28
6.3.3.1 Total Number of Releases and
Distribution by Class of Substance 6-29
6.3.3.2 Number of Releases Above and Below
Reportable Quantities 6-29
6.3.3.3 Size Distribution of Releases 6-33
6.3.3.4 Further Investigations 6-40
6.4 Summary and Conclusions 6-41
APPENDIX A: SUMMARIES OF PUBLISHED REPORTS ON CLEANUP COSTS A-l
ICF INCORPORATED
-------
TABLE OF CONTENTS
(continued)
PAGE
APPENDIX B: A PRELIMINARY ANALYSIS OF THE PENALTY, LIABILITY,
AND FINANCIAL RESPONSIBILITY PROVISIONS OF CERCLA
1. Introduction B-l
2. Effect of Liability and Financial Responsibility
in Three Cases B-2
2.1 Hazardous Substance Releases Below
Reportable Quantities B-3
2.2 Hazardous Substance Releases At or Above
the Reportable Quantity B-4
2.3 Releases of Pollutants and Contaminants B-5
3. Section 107 — Liability Provisions B-6
3.1 Section 107(a) and (b) B-7
3.2 Section 107(c)(3) B-14
3.3 Section 107(d).. B-16
4. Limits to Liability B-16
4.1 Section 107(c)(1) B-16
4.2 Section 107(c)(2) B-18
4.3 Section 107(k)(l) B-21
5. Section 103 — Penalties Related to Notification Requirements... B-22
5.1 Section 103(b) B-23
6. Financial Responsibility Requirements B-28
6.1 Section 108(a) B-28
6.2 Section 108(b) B-31
7. Other Sources of Liability B-32
8. Summary B-35
iCF INCORPORATED
-------
LIST OF EXHIBITS
EXHIBIT PAGE
4-1 Evaluation of Designation Strategies 4-6
4-2 Evaluation of Strategies for Revising Reportable Quantities.... 4-16
4-3 Market-Oriented Approaches to Sections 102 and 103
Regulations 4-22
5-1 Major Components of a Benefits and Costs Analysis 5-3
5-2 : Effects of Designation 5-7
5-3 Effects of Revising Reportable Quantities 5-9
6-1 The Calculation of the Costs and Benefits of
Regulatory Alternatives 6-2
6-2 Summary of Published Reports on Cleanup Costs 6-19
6-3 Cost of Cleanup of Releases from PIRS Data Base, 1974-1980 6-20
6-4 Relationship Between Cost of Cleanup and Quantity Released 6-22
6-5 Relationship Between the Logarithms of Cost of Cleanup
and Quantity Released 6-23
6-6 Number of Releases Reported to EPA and to PIRS, FY 79 6-30
6-7 Distribution of Releases by Specific Chemical Identity,
PIRS 1979-1980 and EPA Region IV, FY 79 6-31
6-8 Percentage of Releases Reported at or Above Reportable
Quantities for Selected Substances, PIRS 1979-1980 and
EPA Region IV, FY 79 6-32
6-9 Size Distribution of Releases from PIRS .Data Base, 1980 6-33
6-10 Concentration of Reports at Certain Quantities, PIRS Data
Base, 1980 6-34
6-11 Size Distribution of Releases, PIRS Data Base, 1974-1980 6-35
6-12 Size Distribution of Releases, PIRS Data Base, 1980 6-36
ICF INCORPORATED
-------
LIST OF EXHIBITS
(continued)
EXHIBIT PAGE
6-13 Size Distribution of Non-PCB Releases, EPA
Region VII, 1981 6-36
6-14 Size Distribution of PCB Releases, EPA Region VII, 1981 6-37
6-15 Size Distribution of Releases, PIRS Data Base (1980) and
Non-PCB Releases, EPA Region VII (1981) 6-38
A-l ' Release Categories A-3
A-2 Response Cost Summary A-4
B-l Liability Provisions in CERCLA Section 107 B-8
.B-2 Liability Limitation Provisions in CERCLA Section 107 B-17
B-3 Penalty Provisions in CERCLA Section 103 B-24
B-4 Penalty Provisions in CERCLA Section 103--Exceptions B-26
B-5 Financial Responsibility Requirements in CERCLA Section 108.... B-29
B-6 Applicability of Liability Rules to Different
Categories of Costs of Releases... * B-34
ICF INCORPORATED
-------
EXECUTIVE SUMMARY
1. PURPOSE
This report reviews the potential economic effects associated with the
designation of hazardous substances and establishment of reportable quantities
under the authority of Sections 102 and 103 of the Comprehensive Environmental
Response, Compensation, and Liability Act of 1980 (CERCLA or "Superfund").
The economic effects are analyzed according to an evaluation framework which
was developed by ICF for this report. The framework includes key criteria
that EPA must meet in proposing or promulgating a regulation. This economic
report supplements a technical report, prepared by Rockwell International,
which discusses criteria and strategies for the designation of hazardous
substances, and strategies for the assignment of reportable quantities to
those substances.
2. BACKGROUND ON SECTIONS 102 and 103
Regulation tinder Sections 102 and 103, in the broadest sense, is expected
to correct the imperfections of the market by requiring potential releasers to
internalize the costs to society of releasing hazardous substances into the
environment. These regulations would increase the liability and responsibility
of releasers for cleanup costs and natural resource damages, encouraging them
to seek preventive measures to reduce the number or severity of releases.
Sections 102 and 103 of CERCLA contain provisions on designating
substances as hazardous, setting reportable quantities, and notifying the
appropriate authority after a release. These provisions broaden EPA's
previous authority to monitor and clean up releases of designated hazardous
substances by expanding the number of hazardous substances that must be
reported when released and requiring reports of releases into all
environmental media, rather than from navigable waters alone.
Section 102(a) specifies that the EPA Administrator may designate a
substance (or element, compound, mixture, or solution) as hazardous if the
substance, when released into the environment, may present a substantial
threat to public health, welfare, or the environment. In addition, substances-
designated as hazardous under certain sections of the Clean Water Act (CWA),
the Clean Air Act, the Solid Waste Disposal Act, and the Toxic Substances
Control Act (TSCA) are covered by CERCLA's response, notification, liability,
and financial responsibility provisions.
In promulgating regulations that designate additional substances as
hazardous, EPA must also establish reportable quantities for them. Reportable
quantities (RQs) are those quantities of hazardous substances which, if
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
ES-2
released, trigger the notification requirements of Section 103; this section,
among other things, requires an immediate report from a responsible party who
releases an amount of a hazardous substance equal to or greater than its RQ.
CERCLA permits EPA to establish a single RQ for each,hazardous substance,
regardless of the medium into which the substance is released.
In addition to establishing RQs for newly-designated substances, two
categories of existing reportable quantities will be subject to regulation:
• Reportable quantities for hazardous substances
designated under Section 311 of the CWA. These
substances already have reportable quantities, based
primarily on their level of aquatic toxicity.
However, EPA may decide to revise the reportable
quantities for some or all of these substances.
• Reportable quantities for all other hazardous
substances designated under CERCLA. For substances
previously designated as hazardous under other laws
(besides Section 311 of CWA) and defined as hazardous
under Section 101(14) of CERCLA, the statute sets an
interim reportable quantity of one pound unless or
until superseded by regulation. Because the one pound
limit may be too low for some substances and too high
for others, EPA is developing strategies for adjusting
the one pound reportable quantity limit.
The reportable quantity adjustments for substances already designated
under CERCLA will be established in rulemaking scheduled for the Spring of
1982. EPA also is developing regulations to determine reportable quantities
for all hazardous substances. These regulations will establish final
reporting requirements and fully implement EPA's approach to reportable
quantities.
Notification provisions discussed in this report are contained in Section
103(f) of CERCLA. These provisions relate to continuous releases of hazardous
substances that are stable in quantity and rate. CERCLA requires annual
notification or notification at the time of a statistically significant
increase in the release, as long as the continuity, regularity, and quantity
of the release have been established for a sufficient period of time through
notification under Section 103(a). EPA is currently engaged in an effort to
develop and assess strategies for notification of continuous releases.
3. STRATEGIES FOR REGULATION
This report elaborates on several of the regulatory strategies for
designation and RQs developed by Rockwell International. Where possible, the
scope of several regulatory alternatives within a strategy is discussed.
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
ES-3
Designation Strategies
Three results of the designation of a substance can be projected:
(a) the required reporting of releases of the substance; (b) the application
of the liability and financial responsibility provisions of Sections 107 and
108 of CERCLA; and (c) an increased likelihood of cost recovery of funds
expended in response actions. The language of CERCLA implies that designation
should be based on the potential harmfulness of the substance.
Three designation strategies derived from Rockwell's report are discussed"
in this report.
Strategy A: Designate all chemical substances found at
hazardous waste sites or commonly involved in accidental
releases that are considered potentially harmful. (Derived
from Rockwell s Designation Strategy #2.)
This strategy serves two primary purposes: it brings a large number of
substances muter the reporting requirement and increases the likelihood of
successful cost recovery of funds expended in response actions, thereby
conserving the Superfund and stretching its resources for use in the most
urgent cases.
The effects of this strategy depend on how many and which substances are
designated. If RQs are low, both industry and government would be burdened by
making and processing many additional reports of releases. If many substances
are designated, releasers might incur large insurance costs due to the CERCLA
liability and financial responsibility provisions. On the other hand,
designation of many substances may be necessary to activate industry's
investment in preventive measures. Since this strategy would designate many
substances, it will be necessary to investigate the economic effects in more
detail to determine if this strategy is appropriate.
Strategy B: Establish a Hazard Index and rate the degree of
hazard of all candidate substances. Based on this initial
screening, use a panel of experts to select substances for
designation. (Derived Rockwell s Designation Strategies #3 and
#4.)
This strategy explicitly relates designation to the potential
hazardousness of the substance.1 Because many different hazard indices
could be constructed, and because only arbitrary cut-off points could be
established to determine when a substance is hazardous or not, the
defensibility of the hazard index would be enhanced if it were accompanied by
something like a panel of experts to make the actual selection of substances
for designation.
1Refer to Rockwell's report, "Technical Base for Designation of
Hazardous Substances and Assignment of Reportable Quantities," (November 2,
1981) for a discussion of possible hazard indices.
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
ES-4
The attractiveness of this strategy lies in its coupling of relatively
objective assessments of hazard and subjective assessments based on expertise
and experience. Both these assessments, however, may be subject to challenge,
since a different hazard index or group of experts could produce different
results. A further consideration in the use of this strategy is the number of
substances that might ultimately be designated. Experience with other EPA
regulatory programs has shown that with limited resources, relatively few
substances can be investigated in depth at any one time. If only a handful of
substances may need to be designated or if the regulations may be phased in,
this approach could be useful. If, however, hundreds of substances may
eventually need to be designated, this process may prove administratively
unsuitable.
Strategy C: Establish a checklist of criteria for
designation. Substances meeting the most stringent criteria are
subsequently designated. Substances meeting less stringent
criteria or lacking sufficient data remain "candidates for
designation. (Derived from Rockwell s Designation Strategy #5.)
Under this strategy, EPA would develop a series of "on/off" criteria
(rather than a ranking system), based on factors such as degree of hazard,
release history, and possibly economic factors. ("On" in this context means
"on" the designation list.) EPA would then designate those relatively few
substances that meet the strictest criteria, such as proven carcinogenicity or
extreme flammability.
EPA would consider a second tier of substances as candidates for
designation if they met less strict criteria or if data were not available to
determine their effects based on the strictest scientific peer review. These
substances would still be of concern to EPA, and reporting of releases of
these substances would be encouraged oh a voluntary basis or perhaps required
under the authority of TSCA, Section 8. If further evidence, based on release
history or health effects data, shows that the substance can meet the most
stringent criteria, EPA would then designate the candidate substance under
CERCLA.
This strategy has several positive elements. It would permit a large
number of substances to be evaluated for designation. It also would establish
stringent and defensible criteria for designating substances. In this
respect, this strategy differs from Strategy B, by providing a "yes or no"
definition of whether the hazard warrants designation, rather than a ranking
of the relative hazard of substances. Finally, and perhaps most important,
this strategy could permit EPA to attain the reporting goals of Sections 102
and 103 for a large number of candidate substances, while relieving industry
of the potential burden posed by CERCLA's liability and financial
responsibility provisions.
The net effects of these three strategies cannot be determined in advance
of a full analysis. However, one likely source of effects will be the number
of substances designated. Some regulatory processes may be less, suitable for
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
ES-5
achieving the designation of a large number of substances. In addition, the
designation of several hundred substances may have very different economic
effects than the designation of a handful of substances. Although we cannot
indicate whether the net benefits of designating many substances would be
larger than the net benefits of designating few substances, we hypothesize
that this question will be an important factor in the selection of designation
strategies, especially in view of the potential liability which accompanies
designation.
Strategies for Reportable Quantities (RQs)
Required reporting for releases of hazardous substances ensures that
response will be made to the release, if a field assessment by an On-Scene
Coordinator indicates that response is necessary. It also encourages
releasers to respond, increases the speed of response, and provides for expert
government assistance when needed. The objective in setting reportable
quantities is to find the appropriate trade-offs among four key factors: (1)
the probability of harm to the public or damage to the environment arising if
a release is not reported; (2) the cost of cleanup if the release is not
reported in a timely fashion; (3) the cost to industry of making the report;
and (4) the cost to the government of receiving and acting on the report.
The legislative history of CERCLA makes clear that EPA may establish RQs
solely on the basis of administrative feasibility and practicality. Thus, a
rationale for the RQ revisions may simply be that at an RQ of one pound, EPA
would be overwhelmed by so many reports that its response program would be
impaired.
This report presents five strategies derived from Rockwell's report for
establishing reportable quantities. These strategies are based on either
administrative feasibility, hazard, or a combination of hazard and feasibility.
EPA may also wish to consider in future analyses additional alternatives based
on the specific design of the reportable quantity levels. For example,
regulatory alternatives could be formulated on the basis of different
reportable quantities for different environmental media.
Strategy A: No Prior Screening. (Derived from Rockwell's RQ
Strategy #1.)
This strategy is based on an aspect of administrative feasibility —
the amount of resources available to screen reports of releases. The analogy
here is to a police or fire department, each of which, like the National
Response Center (NRC), receives numerous "false alarms." The decision EPA
must make is whether to screen out false alarms before they are reported (by
setting fairly high RQs), or whether to encourage many reports and screen the
reports as they are received. Since reports of releases are screened at the
NRC anyway before response is undertaken by a federal On-Scene Coordinator, a
strategy of "no prior screening" would simply require additional resources to
handle the calls. EPA could leave the statutorily-set RQs at one pound or
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
ES-6
raise them to a higher level on the basis of the resources available for
receiving reports. This process would be iterative, as EPA gained experience
with the amount of resources needed to handle incoming calls. Under this
strategy, EPA would be weighing the industry costs of reporting many low-level
releases, the benefits of collecting the information, and the benefits of
being able to respond when necessary to potentially harmful releases of small
quantities of hazardous substances.
One fact that lends support to this strategy is the low level of reports
currently received by the NRC. We have been told that the NRC at present is
receiving about 35 to 50 reports of releases of CERCLA-designated substances
per week. Given this level of reporting, it may not be necessary to develop
elaborate prior screening regulations.
Strategy B: Hazard Index. (Rockwell's RQ Strategy #4.)
As discussed previously under designation strategies, a hazard index
could be devised to evaluate the relative toxicity and other adverse effects
of substances. The hazardousness of substances would be rated based on
physical and chemical properties and other characteristics.
This is a viable approach that directly links the RQ regulations to the
concept of hazard. As discussed, this linkage appears to be permissible and
to some degree desirable, since response to releases should be conditioned on
minimizing or eliminating harmful exposure. There are, however, some
difficult problems associated with this strategy. First, as Rockwell points
out, there are severe data limitations that make relative rankings difficult
to obtain. Second, there is no general agreement in the scientific community
on how hazard factors are to be combined to form an acceptable hazard index.
Rockwell has formulated four different algorithms that illustrate this
problem. It is especially unclear how to incorporate aggravating factors,
such as persistence and bioaccumulation. Finally, this approach does not
provide a direct link between an RQ and the need for response, relying instead
on hazard as a measure of the need for response. It may, however, be possible
to construct a hazard index that will incorporate additional factors relating
to the need for respons.e.
Strategy C; Hazard Index Combined With Fine-Tuning To Incorporate
Administrative Feasibility. (Rockwell s Strategy #5.)
This strategy would combine the initial use of a hazard index with
subsequent "fine-tuning" to ensure that the RQs assigned are reasonable. One
method of fine-tuning is to rely on a panel of experts with the experience
needed to evaluate the RQ assignments. A second method would be to construct
an additional set of criteria, based on economic and other factors, to adjust
the RQ assignments.
One problem with the latter approach is that it is not clear what criteria
should be used to revise the RQs. A criterion commonly mentioned is the
production volume of a substance, which is intended as a proxy for the
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
ES-7
potential for exposure to a substance. The volume of production, however, is
an inexact measure which does not consider methods of handling, transporta-
tion, and types of use — in short, those factors that make up release
potential. Even the use of a criterion such as release potential is
problematic. It is not clear that a highly toxic substance with a very low
release potential (or production volume) should not be reported if it is
released in low quantities. If the substance is infrequently released, there
will be little reporting burden on industry and the government. If, however,
it is highly toxic, even a small quantity should perhaps be reported and
cleaned up.
Strategy D: Empirical Evidence (Derived from Rockwell's
Strategy #5.)
This strategy adopts an empirical approach to revising RQs. Since
the reporting requirement is already in place under CERCLA, over time EPA will
presumably be developing an excellent data base. Under this strategy, EPA
would defer assigning new RQs until it had accumulated enough experience with
actual reports of releases. Then, EPA. would develop decision rules that
relate RQs to the frequency of response to releases of different sizes.
As an example, assume a substance had an RQ of one pound and that over a
specified period of time, the following data were recorded:
For releases under 10 lbs., responses were made 5% of
the time.
For releases between 10 and 100 lbs., responses were
made 25% of the time.
For releases between 100 and 1000 lbs., responses were
made 50% of the time.
Using a specified cut-off point (say, 50 percent response rate), the RQ
for this substance would be set at 100 pounds.
The attractiveness of this strategy is that it is based almost entirely on
actual experience and can easily be modified as further experience develops.
This strategy does not need to rely on models, assumptions, or incomplete past
data to obtain estimates of release potential. In addition, it goes beyond
the estimation of release potential to get at the heart of the regulations:
the need for response by the federal government.
The most unattractive aspect of this strategy is that it could cause some
confusion in the regulated community because of the potential lack of finality
in the RQs established for designated substances. However, since most RQs
would be raised rather than lowered over time if this strategy were followed,
releasers would be unlikely to violate the regulations inadvertently.
November 2 Draft * * * Do not quote or cite
1CF INCORPORATED
-------
ES-8
Strategy E: Two-Tiered Approach. (Derived from Rockwell's
Strategy #5.)
This strategy again combines considerations of hazard and
administrative feasibility, in perhaps the simplest approach of all. Under
this strategy, EPA would choose a single reportable quantity level for the
majority of substances, based on estimates of reporting burden, probability of
need for response, or the Agency's experience to date. However, for those
substances which are known to be extremely hazardous, a lower RQ would be
set. The assignment of these substances to the low RQ level could be done on
the basis of Agency experience and best scientific judgment or by constructing
a checklist of stringent hazard criteria.
This strategy can best be envisioned as the "big box/little box" approach
to RQs. If a substance is very hazardous, it fits into the "little box" with
the other (relative few) substances that have one pound RQs. If, on the other
hand, the substance is harmful only in large quantities, it fits into the "big
box" with the many other harmful, but not highly hazardous, substances.
The attractiveness of this strategy lies in its comparative simplicity; of
all the strategies, regulatory costs would probably be least for this one.
The strategy is straightforward in dealing with the uncertainty that pervades
the RQ issue. It provides for low RQs (one pound, for example) for the most
hazardous substances and partially relieves the reporting burden on releasers
for all other substances. This strategy would implicitly recognize that the
major effects of the Sections 102 and 103 regulations arise from the
designation process and that assignment of RQs plays more of a safeguarding
role for releases of the most highly hazardous substances. But the usefulness
of this strategy could probably only be realized if only two RQ levels are
used, despite the temptation to establish additional RQs to account for the
relative hazard of different substances.
4. PROPOSED EVALUATION CRITERIA
A key element of this economic analysis is the use of an evaluative
framework for assessing and selecting strategies and regulatory alternatives.
This report contains five proposed criteria which can be used to evaluate the
merits and impacts of regulatory approaches. This list is not exhaustive and
EPA will want to weigh some criteria more heavily than others in selecting its
regulatory approach. The proposed criteria are:
(1) Scientific Validity and Protection of Public Health:
Is the regulatory strategy defensible on scientific
grounds? Does it adequately protect public health,
welfare, and the environment? Does it yield
unambiguous results? Is it easily complied with and
consistent with other Agency findings and approaches?
.Does it rely as much as possible on the experience
gained from existing governmental programs that
regulate toxic substances?
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
ES-9
(2) Executive Order 12291; Do the strategy's potential
real resource benefits outweigh its potential costs?
Does the regulatory strategy maximize net benefits to
society?
(3) Administrative Ease: Will the strategy result in an
excessively resource-intensive regulatory development
process? Will the regulations resulting from the
strategy be enforceable? Will EPA's responses to
hazardous substance releases be improved?
(4) Regulatory Flexibility Act: Does the strategy
minimize significant impacts on small businesses, local
governments, and other small entities?
(5) Distributional Effects: Will the strategy result in
inequitably distributed costs or benefits?
Each of the proposed regulatory strategies is reviewed in this report
according to the above criteria. The report also contains a review of the
factors that will affect the direction of the economic effects, i.e., those
elements of a given regulatory strategy which could increase or decrease the
costs or benefits. These factors are discussed in detail in the main body of
the report (Chapters 4, 5 and 6) and summarized here.
5. EFFECTS OF DESIGNATION REGULATIONS
The designation of a substance as "hazardous" is the key action-initiating
factor in the hazardous substance response program. Designation is the
essential first step that brings a chemical substance under the purview of the
regulations. This section discussers the economic effects of designation.
A. Reporting and Response
A direct effect of designation is that releases of the substance at
or above the reportable quantity level must be reported. The act of reporting
a release presumes that the release will be cleaned up, if necessary.
Therefore, the source of major real resource benefits and costs of designation
is that some releases may be cleaned up which would not have been cleaned up
without the regulations or which would have received an incomplete or untimely
response, resulting in potentially higher long term costs. These cleanup
benefits and costs depend crucially on how designation affects the behavior of
parties responsible for the release of a substance. If responsible parties
diligently clean up their releases of any substance (whether designated or
not), designation of a substance would have few effects. On the other hand,
if releases of non-designated substances are largely ignored by responsible
parties, the benefits and costs of designation might be very great.
B. Liability
The largest potential distributional effect of designating a
substance as hazardous may arise from the fact that releasers of the
November 2 Draft * * * Do not quote or cite
1CF INCORPORATED
-------
ES-10
substance become subject to the liability and financial responsibility
provisions of CERCLA. (See Appendix B for a full discussion.) This would
likely translate into increased costs for potential releasers of designated
substances because they would seek to augment their insurance coverage.
In addition, because of potential liability Tinder CERCLA, releasers may
be encouraged to improve handling and management of newly-designated
substances and to clean up more of these releases to reduce their potential
liability. These actions could result in indirect benefits to society
through a reduction in the number or severity of releases and a reduction in
exposure to releases, and in indirect costs to releasers of instituting
preventive measures. In many cases, however, newly designated substances may
already be subject to DOT or OSHA regulations, so that industry practices
might not change substantially under CERCLA regulations.
Most of the potential costs of designation would be incurred in the first
year or so following designation. Releasers would presumably act within that
time to increase their insurance, or to install preventive practices. Upon
promulgation of regulations, releasers would immediately be responsible for
reporting costs and applicable cleanup costs for releases of the designated
substances.
The benefits of designation, on the other hand, have a more uncertain time
frame. Benefits accruing from more certain and more complete compensation for
damages and cleanup costs will be felt in the short term. The benefits of
avoiding economic losses and acute adverse health effects would also accrue
immediately if there were a reduction in the number or severity of releases.
Long-term benefits might result from a reduction in the risk of exposure to
substances with chronic adverse health effects.
6. EFFECTS OF REVISING REPORTABLE QUANTITIES
Regulations revising reportable quantities will have costs and benefits
depending on whether RQs are raised or lowered. If the RQ for a substance is
lowered, for example, the regulations would probably result in added costs to
industry and added benefits to public health and the environment. Another key
point is that all of the costs and benefits enumerated below relate only to
the incremental change in reported releases. (That is, if RQs are lowered,
benefits and costs are associated only with the additional releases that would
become subject to the reporting requirement;)
A. Reporting and Response
The value of the reporting requirement is in encouraging the act of
reporting and the benefits that flow from that act. There may, of course, be
releasers that will not report their releases, either because they remain
unaware of their obligations, because they do not accurately estimate the true
quantity released, because they have intentionally released the substance
(e.g., "midnight dumpers"), or because they do not want to incur the costs
attendant to reporting.
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
ES-11
On the other hand, the fact that the law exists and will be enforced will
probably prompt compliance by most releasers. The penalties for failure to
report releases and increased public awareness also act as incentives to
report. Finally, releasers may wish to avoid liability for delayed cleanup
that may exceed the cost of cleaning up when 'the release first occurs.
It is important to note that releasers currently do report some releases
below existing reportable quantity levels. In fact, some releases of
non-designated substances are reported. Even if the release is not reported,
releasers may still clean up in order to protect their reputations, limit
their liability, and salvage released materials.
If the release is reported, it must be assumed that the releaser
understands (or will soon be made to understand) his responsibility to clean
up if cleanup is needed, or, if the releaser is not capable of taking response
action, to pay for government cleanup. This is the primary potential source
of benefits and costs from the reportable quantity regulations. That is,
depending on the direction of the change in reportable quantities, there will
be some change in the number of releases cleaned up. If RQs are lowered, more
releases will likely be cleaned up. Additional cleanups result in benefits to
society of reduced risk of exposure to hazardous substances and in costs to
releasers of cleaning up releases they might not have otherwise responded to.
The magnitude of the effects noted here depend on industry and government
behavior without the Section 102 and 103 regulations. If releases of
hazardous substances below the reportable quantity are responded to as
diligently as releases above the reportable quantity, there may be no
appreciably different effects from revising reportable quantities.
Whether or not response is affected, changes in RQs and reporting habits
may have another major impact on releasers and those affected by releases. To
the extent that releasers report their releases (whether above or below RQs),
it could be easier to prove the releaser liable for any possible damages and
cleanup costs. Indirectly, therefore, a change in RQs may change the
likelihood of assigning liability to the releaser.
A lowering of RQs may also result in releasers initiating preventive
actions, to avoid the costs associated with releases. These actions may have
indirect benefits of reducing the number or severity of releases, thereby
reducing the risk of exposing people and the environment to hazardous
substances. Further, since most of the material released in a typical
incident is not recovered or reused by the releaser, actions which reduce the
number of releases save valuable resources from being wasted.
Finally, reporting of releases has informational value. Depending on how
EPA uses this information, data on releases may prove useful in designing
programs to effectively contain releases, in alerting releasers to ways of
preventing releases, and in educating people to mitigate the effects of
releases by exercising cautionary measures. Indirectly, benefits could result
from fewer or smaller releases, with reduced exposure to people and the
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
ES-12
environment. The direct costs to industry of producing the information
consists of the nominal costs of a telephone call to.the National Response
Center. Indirect costs of reporting include the costs of recordkeeping,
training of personnel, labeling or special marking of containers for hazardous
substances, increased insurance costs, monitoring and measurement costs for
evaluating compliance, and the costs of whatever preventive action might be
taken to reduce the likelihood of a reportable release.
B. Effects of Federal Involvement
Other benefits and costs may result from the involvement of the
federal government in the response process. For example:
(1) Immediate reporting of releases can reduce the delay in response
action. A releaser may not realize that response is actually necessary or
that quick action is important. The National Response Center expert on the
telephone line can inform the releaser of that need. Faster response can have
two benefits: (a) it may decrease exposure to hazardous substances, and (b)
it may reduce the ultimate costs of cleanup and potential liability for
damages by containing the spread of hazardous substances in the environment.
(2) Reported releases permit the government to contribute additional
expertise and experience. Federal involvement often contributes a wide range
of experience with hazardous substances and cleanup methods and a wide variety
of resources from support agencies (such as the Food and Drug Administration,
Center for Disease Control, Army explosives teams, etc.). In general, this
expertise should have beneficial effects in limiting the damage of a release.
The effects on releasers' costs Qf cleanup are uncertain: government
officials may require a higher degree of cleanup, a concern for environmental
damage, and greater precautions (e.g., in evacuating area residents or
providing alternate water supplies) than the releaser might have provided. On
the other hand, the releaser might reduce his ultimate liability for damages
and might reduce cleanup costs by using more efficient cleanup methods
suggested by the government.
(3) Reporting of releases allows for continued government monitoring and
assessments of the effects of a release, whether or not immediate removal, is
necessary, and provides for continued protection of public health and the
environment.
\
7. ALTERNATIVE BASELINES
In order to analyze fully the effects of regulatory alternatives, it is
necessary first to project public and private sector actions in the absence of
the Sections 102 and 103 regulations. This projection -- the baseline --
plays a crucial role in any evaluation of economic effects, since the effects
of the regulatory alternatives are measured relative to the baseline. Thus,
if both the baseline and a regulatory alternative represent similar or
identical situations, the analysis will show few or no economic effects. The
November 2 Draft • * * * Do not quote or cite
ICF INCORPORATED
-------
ES-13
magnitude and nature of the effects to be evaluated are therefore conditioned
on the baseline.
To clarify what the baseline represents:
1. The baseline is a projection of what would happen in
the absence of something eventually expected to happen
(i.e., the regulation). As such, the baseline
typically represents a hypothetical situation.
2. The baseline does not necessarily represent existing
practices or behavior, although the current state of
the world might be reflected in the baseline to some
extent.
3. The baseline should be constructed for purposes of
evaluating only the regulatory alternatives, not the
enabling statute. In this case, CERCLA Sections 102
and 103 provide a stringent level of control in the
absence of regulations. The baseline cannot,
therefore, stipulate a world without any notification
provisions; that would in effect ignore Congress1
imposition of such provisions.
4. Because it is difficult to determine "what the world
would.be like" in the absence of the regulations, many
different baselines are conceivable. The Sections 102
and 103 regulations form only one part of a complex
set of statutory provisions and associated
regulations, all of which will interact to affect the
behavior of the regulated parties.
In developing this report, we have considered three alternative baselines:
Baseline 1: Pre-CERCLA Behavior: The behavior of
parties prior to CERCLA*s enactment would be projected to
continue after passage of CERCLA.
Baseline 2: Statutory Language of CERCLA: Behavior
would be governed by the statutory language of CERCLA, the
legislative history, and any judicial interpretations.
However, behavior would be projected to be unmodified by
regulations interpreting the statute.
Baseline 3: Current Practices: Behavior as it currently
exists would be projected to continue under the baseline.
This report recommends that alternative 2 be selected as the baseline for
calculating the benefits and costs of the regulatory alternatives. The pros
and cons of each alternative arei presented in Chapter 3 of the report.
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
ES-14
8. SUMMARY OF CONCLUSIONS FROMANALYSIS OF RELEASE DATA
Chapter 6 contains an analysis of the data currently available concerning
the size distribution, frequency, and cost of cleanup of releases of hazardous
substances into the environment. The analysis of available data indicates:
1. With the exception of PCBs, the substances most frequently reported
released are produced in billions of pounds per year. This finding tends to
confirm the intuitive conclusion that the substances produced in the largest
quantities are released most often. However, the number of releases of PCBs,
which are no longer used in new products, is very high (about 35 percent of
Region VII*s reported releases). This indicates that the candidates for
designation should include not only those substances produced in large
quantities today, but also those which were previously produced and widely
used. Some pesticides, for example, whose registration for most or all uses
has been cancelled, should still be candidates for designation.
2. . About .half of all reports of releases are for quantities below the
reportable quantity. This implies that at least some firms have been
reporting releases without regard to the reportable quantity established by
EPA. Estimating the extent of such behavior is important for assessing the
effects of changes in reportable quantities.
3. In general, there appears to be substantial under-reporting of
releases. This finding was obtained by a statistical analysis of reports of
PCB releases and reports of releases of other substances in the EPA region in
which reporting appears to be most complete (Region VII). Thus, the data
obtained so far indicate that some releases are reported whether or not the RQ
has been reached. Other releasers seem not to be reporting releases
regardless of the RQ.
4. The relationship between the cleanup costs of releases and
quantities released is not linear, but the logarithms of cleanup cost and
quantity released appear to be linearly related. The relationship,
furthermore, is significantly affected by environmental conditions at the
release site. Analysis of data on the relationship of cleanup cost to
quantity released should be useful in establishing reportable quantities; EPA
may want to establish the RQ for some substances slightly below that level at
which the cost of cleanup of a release increases dramatically for each
additional pound released. If this were the case, the government's use of
Superfund monies for immediate removals would have a higher probability of
being cost-effective.
5. The information available on the consequences of cleaning up
releases that would not otherwise be cleaned up is probably insufficient to
generate quantitative estimates of the benefits of Sections 102 and 103
regulations. Several anecdotal case studies of the costs of delayed
notification have been included in Chapter 6, but data do not appear to exist
for a comprehensive analysis of benefits as envisioned by Executive Order
12291.
November 2 Draft * * * Do not quote or cite ,
ICF INCORPORATED
-------
ES-15
9. SUMMARY OF BENEFITS AND COSTS OF REGULATORY STRATEGIES
Since the overall EPA strategy for Sections 102 and 103 and the exact
regulatory alternatives have not yet been selected by EPA, this report cannot
discuss the benefits and costs of the Agency's decisions. Rather, this report
presents a methodology for analyzing the potential costs and benefits, and a
review of the data available for the analysis. The methodology proposed here
is intended to comply fully with the requirements of E.O. 12291, 0MB guidance
on E.O. 12291, EPA guidance on E.O. 12291, the Regulatory Flexibility Act, and
EPA guidance on calculating reporting burdens. It is hoped that, in addition
to their other comments, the reviewers of this report will comment on the
methodology proposed here and perhaps suggest other data sources that might be
useful — especially in calculating the costs and benefits to industry and to
state and local governments from the designation and RQ regulations.
Despite the limitations stated above, the report does contain several
tentative conclusions on the benefits and costs of the proposed strategies and
an analysis of several key economic and policy issues that should be useful to
EPA management in deciding upon an overall regulatory strategy.
1. The degree of similarity between the list of CERCLA hazardous
substances and the hazardous substances designated by other legislation will
influence the costs and benefits of designation strategies. All substances
that have already been designated as hazardous under CERCLA have thus far
been designated as hazardous under other laws. However, substances being
considered for designation under CERCLA may not yet be regulated under any
other legislation. The designation of a substance as "hazardous" or "harmful"
under any legislation requires an individual or firm handling that substance
to single it out for special treatment. Firms must design special operating
procedures and purchase special equipment. If a substance is designated under
CERCLA that is also designated under other legislation, the incremental cost
of designation under CERCLA is likely to be relatively small, because the
substance is already being specially treated by regulated parties. If a
CERCLA substance is not considered hazardous under other laws, special
treatment must be initiated at increased cost. On the other hand, the
benefits of designation under CERCLA are also likely to be lower if the
substance is considered hazardous under other legislation. In that case, the
special treatment it already receives may have reduced the possibility of a
harmful release.
2. The designation of a substance under CERCLA may affect the insurance
coverage of regulated parties, thereby increasing their costs. As noted
above, the designation of a substance under CERCLA may be a trigger that
causes insurance companies to label a substance as "dangerous," thereby
causing them to raise their rates for coverage of damage resulting from a
release of that substance. This could prove to be the major source of
distributional costs to releasers resulting from designation regulations.
3. The designation of a substance under CERCLA triggers action by the
Department of Transportation, thereby indirectly causing additional costs and
November 2 Draft * * * Do not quote or cite
IUr INCORPORATED
-------
ES-16
benefits. Under Section 306(a) of CERCLA, each hazardous substance
designated under CERCLA must be designated as hazardous under the Hazardous
Materials Transportation Act (HMTA). Designation under HMTA may generate
additional costs and benefits, and they may be partially attributable to
CERCLA.
4. More of the costs and benefits of the Sections 102 and 103
regulations may be generated by the designation of additional substances than
by the establishment of reportable quantities. Designation invokes the
liability provisions of CERCLA. The establishment of a reportable quantity
only sets the level at which a report must be made to the government. The
liability provisions may be more important in defining the actions of
regulated parties than the necessity of reporting a release to the government.
5. A key element in the magnitude of benefits and costs resulting from
designation regulations is the number of substances designated. This in turn
will likely be determined by the length of the master list of substances
considered for.designation and the stringency of the designation criteria.
Considered in light of the existing designation of over 1,100 substances,
additional designation of a handful of substances may result in relatively few
benefits and costs; designation of many more substances may, however, change
industry practices substantially.
6. The primary source of real resource costs and benefits from both
designation and reportable quantity regulations occurs in .the change in the
number of releases responded to in a timely and effective manner.. Thus, in
order to be cost-effective, the designation process should result ill the
cleanup of additional releases (beyond the baseline) by the responsible
parties or by the government. Any such releases, furthermore, should be
cleaned up sooner than they would have been without the regulations. Finally,
in order to capture most of the preventive benefits of this rulemaking, the
regulations should address those substances most likely to be released.
7. The extent to which responsible parties will respond more diligently
to a release of a given size if it is at or above the RQ than if it is below
the RQ is an important determinant of the costs and benefits of the
reportable quantity regulations. As discussed in Appendix B, liability for
response costs and damages does not depend on the size of the release. Any
release of a designated substance triggers the liability provisions. However,
it is plausible to expect that a responsible party will respond more
diligently to a release if required to report it to the NRC (assuming the
party obeys the law and does report it) than if not so required. In order to
conduct a regulatory impact analysis, some assumption must be made about how
much more diligent the response would be.
CONCLUSION
It is appropriate to state again that the data analyses and conclusions on
potential economic effects contained in this report are necessarily
tentative. They have not been discussed with interested parties outside of
EPA. That key step begins with the completion of this draft report. After
November 2 Draft * * * Do not quote or cite | __
ICF INCORPORATED
-------
ES-17
completing the appropriate review of the technical and economic reports by
internal agency and external interested parties, EPA management intends to
begin the rulemaking process in the Spring of 1982. The rulemaking will
reflect the concerns raised during the extensive reviews planned for the
contractor reports.
November 2 Draft * *
Do not quote or cite .
IGF INCORPORATED
-------
CHAPTER 1
INTRODUCTION
This report has been prepared by ICF as a subcontractor to Rockwell
International, the prime contractor charged with helping EPA design strategies
for implementing Sections 102 and 103 of the Comprehensive Environmental
Response, Compensation, and Liability Act (CERCLA or "Superfund"). Rockwell
is focusing on the scientific and technical components of the strategies while
ICF is analyzing the potential benefits and costs of the strategies, and
developing a framework for evaluating the effects of the strategies.
This first chapter discusses the purpose of the report and outlines the
contents of the remaining chapters.
1.1 PURPOSE OF THE REPORT
EPA expects to begin the rulemaking process for regulations implementing
CERCLA, Sections 102 and 103 in the Spring of 1982. These regulations will
cover the following key CERCLA provisions:
(1) Designation of hazardous substances (Section 102(a));
(2) Setting reportable quantities for reporting releases
of designated hazardous substances (Section 102(a) and
(b)); and
(3) Exemption from notification (in certain circumstances)
for continuous releases of hazardous substances
(Section 103(f)).
The regulations implementing Sections 102 and 103 of CERCLA — and any
other new regulations that EPA prepares -- must meet the requirements of a
variety of laws, policies, and guidelines issued by Congress, the President,
and the EPA Administrator concerning the potential economic effects of new
regulations. Foremost among these is CERCLA itself, which requires in Section
305 that regulations be submitted to the Congress for possible veto before
they take effect. The Regulatory Flexibility Act (RFA) provides that agencies
must fit regulatory and information requirements to the scale of those
businesses, organizations, and governmental jurisdictions being regulated.
Executive Order 12291 stipulates that agencies must evaluate the benefits and
costs of regulatory alternatives, and choose regulatory objectives that
maximize the net benefits to society. Finally, the Office of Management and
Budget (0MB) and EPA have developed Regulatory Impact Analysis guidelines that
establish policies for identifying and measuring the effects of regulatory
alternatives.
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
1-2
To fulfill these requirements, the EPA Office of Emergency and Remedial
Response may need to prepare a Regulatory Impact Analysis (RIA) for the
regulations it will propose. This report is not an RIA; instead, it provides
analytic support to the Office and prepares the groundwork for the eventual
performance of RIAs if they are required.
This report has two main purposes:
(1) To review various strategies and regulatory
alternatives that may be presented in the rulemaking,
and to present an evaluative framework by which these
strategies may be assessed. This report also suggests
what additional information may be needed in
evaluating the strategies and choosing appropriate
strategies for regulation.
(2) To examine in some detail the potential economic
effects of Sections 102 and 103 regulations and to
alert EPA decision makers to major economic issues
arising from the regulations.
In addition, this report attempts to structure the policy and economic
analyses so that any RIAs that may be prepared will fully meet the
Congressional and Executive Branch requirements. The report discusses the
analytical requirements that will need to be addressed in preparing
regulations and present methods for accomplishing the required analyses. The
report aiso lists the data sources available for conducting the RIAs and
estimates the impact of data gaps on the analyses.
1.2 ORGANIZATION OF THE REPORT
The remainder of the report is organized into five chapters and two
appendices as follows:
• Chapter 2 provides background on the Superfund
legislation and discusses the requirements of Sections
102 and 103, Executive Order 12291, and 0MB and EPA
guidance on analyzing regulations.
• Chapter 3 proposes a baseline for the analysis of
economic effects. The baseline stipulates the
activities that would take place in the absence of
Sections 102 and 103 regulations, against which the
incremental effects of the regulations can be compared. .
• Chapter 4 establishes a proposed framework for
identifying and evaluating Sections 102 and 103
regulatory alternatives. It explores several
strategies and regulatory alternatives for
designation, reportable quantities, and notification.
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
1-3
Chapter 5 discusses the nature of the economic
effects deriving from the Sections 102 and 103
regulations. It explores the potential benefits and
costs of the regulations covering designation,
reportable quantities, and notification.
Chapter 6 discusses the measurement of the economic
effects resulting from the Sections 102 and 103
regulations. It explains the research performed,
findings, and the overall approach to gathering and
using additional data.
Appendix A presents the cost data collected for this
report and analyzed in Chapter 6.
Appendix B analyzes the penalty, liability, and
financial responsibility provisions that are
triggered by Sections 102 and 103 reporting and
des ignat ion regu1at ions.
November 2 Draft * * * Do not quote or cite
UCF INCORPORATED
-------
CHAPTER 2
UNDERSTANDING
2.1 INTRODUCTION
The purpose of this chapter is to present the essential background
information that would be needed in the development of a Regulatory Impact
Analysis (RIA) for the CERCLA Sections 102 and 103 regulations. Section 2.2
provides an understanding of the Superfund legislation, describes the
designation, reportable quantity, and notification provisions of Sections 102
and 103 of CERCLA, and explains EPA's mandate for regulation. Section 2.3
reviews the Congressional, Presidential, Office of Management and Budget
(0MB), and EPA guidelines for the analysis of regulatory programs, and
discussers whether the Sections 102 and 103 regulations can be expected to have
major economic impacts. Section 2.4 reviews a number of "special topics" --
issues relating primarily to the coverage of the regulations.
2.2 STATUTORY AND REGULATORY BACKGROUND
This section discusses key provisions of Sections 102 and 103 of CERCLA
and EPA's mandate for regulation.
CERCLA gives the federal government authority to respond to releases or
threatened releases of hazardous substances and certain other pollutants or
contaminants that threaten public health and welfare. CERCLA establishes a
$1.6 billion reimbursable fund (the Superfund) to finance response actions,
and charges EPA with administering critical portions of the response program.
CERCLA also imposes liability for the costs of certain response actions on
owners or operators of vessels or facilities from which a release of hazardous
substances has occurred, as well as on other persons (such as transporters of
the substances) who have contributed to the problem. Such liability extends
to: (1) the costs of removal or remedial action incurred by the Federal
Government or State Governments that are consistent with the National
Contingency Plan (NCP); (2) necessary response costs, consistent with the NCP,
incurred by any other person; and (3) damages for injury to, destruction of,
or loss of natural resources. The Act establishes certain limitations to
liability as well as limited defenses to such liability. The liability,
penalty, and financial responsibility provisions of CERCLA are discussed in
detail in Appendix B.
Sections 102 and 103 of CERCLA contain provisions on designation of
hazardous substances, reportable quantities, and notification of releases.
These provisions expand EPA's authority to monitor and cleanup releases of
designated hazardous substances beyond that authorized by Section 311 of the
Clean Water Act (CWA). Specifically, CERCLA expands the number of hazardous
November 2 Draft * * * j)0 not quote or cite
ICF INCORPORATED
-------
2-2
substances that must be reported when released and requires the reporting of
releases into all environmental media, rather than solely to navigable waters.
Regulation under Sections 102 and 103, in the broadest sense, is expected
to correct the imperfections of the market by requiring potential releasers to
internalize the costs to society of releasing hazardous substances into the
environment. These regulations would increase the liability and responsibility
of releasers for cleanup costs and natural resource damages, encouraging them
to seek preventive measures to reduce the number or severity of releases.
The following three subsections describe the requirements of Sections 102
and 103 under the three headings that will be used throughout this report:
designation, reportable quantities, and notification.
2.2.1 Designation of Hazardous Substances - Section 102(a)
Section 102(a) specifies that the EPA Administrator may designate a
substance (or element, compound, mixture or solution) as hazardous if the
substance, when released into the environment, may present a substantial
threat to public health, welfare, or the environment. In addition, substances
designated as hazardous under Sections 307(a) and 311(b)(2)(A) of the Clean
Water Act, Section 112 of the Clean Air Act, Section 3001 of the Solid Waste
Disposal Act, and Section 7 of the Toxic Substances Control Act, are also
covered by CERCLA1s response, notification, liability, and financial
responsibility provisions. Substances designated under Sections 102 or
101(14) must also be listed as hazardous materials pursuant to CERCLA Section
306(a) and are subject to regulations under the Hazardous Materials
Transportation Act, administered by the Department of Transportation. For any
substance designated as hazardous, releases of the substance must be reported
and releasers are liable for cleanup costs and natural resource damages, as
well as the financial responsibility provisions of CERCLA Section 108.
In addition to defining and providing for the designation of hazardous
substances, CERCLA also defines "pollutants and contaminants". The term
includes, but is not limited to, substances which, after release into the
environment and upon exposure, ingestion, inhalation, or assimilation into any
organism, may cause death, disease, behavioral abnormalities, cancer, genetic
mutation, physiological malfunctions, or physical deformations. (See CERCLA,
Section 104(a)(1) and (2).) It appears that the notification, liability, and
financial responsibility provisions of CERCLA may not apply to "pollutants or
contaminants" (i.e., substances not designated as hazardous), although
responses to releases of both "pollutants and contaminants" and hazardous
substances may be taken and financed under Superfund.
Neither the hazardous substance category nor the pollutants or
contaminants category may include petroleum, natural gas or synthetic fuels.
In promulgating regulations that designate additional substances as
hazardous, EPA must also establish reportable quantities for these
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
2-3
substances. Under CERCLA, EPA may establish a single reportable quantity for
a hazardous substance, regardless of the medium into which the substance is
released.
2.2.2 Revision of Reportable Quantities - Section 102(b)
Reportable quantities are those quantities of hazardous substances
which, if released, trigger the notification requirements of Section 103.
(There are no reportable quantity provisions for pollutants or contaminants.)
Aside from reportable quantities for newly-designated substances (considered
as part of the designation process), two categories of reportable quantities
will be subject to regulation:
(1) Reportable quantities for hazardous substances
designated under Section 311 of the Clean Water Act.
These substances already have reportable quantities,
based largely on their level of aquatic toxicity.
However, EPA may decide to revise the reportable
quantities for some or all of these substances.
(2) Reportable quantities for all other hazardous
substances designated under CERCLA. For substances
previously designated as hazardous under other laws
(besides Section 311 of CWA) and defined as hazardous
under Section 101(14) of CERCLA, the statute sets an
interim reportable quantity of one pound unless or
until superseded by regulation. Because the one pound
limit may be too low for some substances and too high
for others, EPA is developing strategies for adjusting
the one pound reportable quantity limit.
The reportable quantity adjustments for substances already designated
under CERCLA will be established by EPA in rulemaking scheduled for the Spring
of 1982. EPA is also developing regulations to determine reportable
quantities for all hazardous substances. These regulations, which will be
proposed in the Spring of 1983, will establish final reporting requirements
and fully implement EPA's approach to reportable quantities in a final
regulation scheduled for the Fall of 1983. This schedule is the best estimate
that was available at the time this report was completed.
2.2.3 Notification - Section 103
There are three types of notification requirements in Section 103
of CERCLA. First, Section 103 provides that the person in charge of a vessel
or facility must immediately notify the National Response Center (NRC) when
there is a release of a designated hazardous substance equal to or greater
than the reportable quantity set by EPA for this purpose. The notification
requirement is similar to the requirement of Section 311 of the Clean Water
Act, but releases into all media, not just U.S. navigable waters, must be
reported. As under the CWA, failure to notify under CERCLA carries a criminal
penalty of up to $10,000 and/or one year imprisonment.
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
2-4
The immediate notification requirement is expected to be fulfilled by
telephone reporting to the NRC. EPA may decide to require written reports of
these releases or record-keeping in addition to telephone reporting. However,
the possibility of additional requirements will not be explored in this
report. This analysis considers immediate reporting to be an element of the
designation and reportable quantity regulations rather than a notification
regulation.
Second, Section 103(c) requires written notification of known hazardous
waste sites that do not have permits or interim status under RCRA's hazardous
waste regulations. EPA has already established this program, and as of June
1981, had received approximately 10,000 notifications. Because the site
notification program has, for the most part, been completed, this economic
analysis will exclude the impacts of the site notification provisions.
Finally, Section 103(f) contains notification provisions for continuous
releases of hazardous substances that are stable in quantity and rate. CERCLA
requires either annual notification or notification at the time of a
statistically significant increase in the release, as long as the continuity,
regularity, and quantity of the release have been established for a sufficient
period of time through notification under Section 103(a). (No notification
requirements apply to federally permitted releases or to the application,
handling, and storage of a registered pesticide by an agricultural producer.)
This third type of notification will be analyzed in this report and
referred to under the heading of "notification."
The next section of this chapter outlines the Congressional, Presidential,
and Executive Branch guidelines that are applicable to the development of
interim and final regulations for the implementation of Sections 102 and 103.
The remainder of the report analyzes these requirements in detail.
2.3 REGULATORY IMPACT ANALYSIS REQUIREMENTS
This section reflects our understanding of the need for a regulatory
impact analysis and of the requirements detailed in Congressional,
Presidential, and 0MB guidelines for developing new regulatory programs. It
also describes EPA's draft guidance on the implementation of Congressional and
0MB directives.
2.3.1 Executive Order 12291
Executive Order 12291, issued in February 1981, establishes agency
requirements for the issuance of new regulations, the review of existing
regulations, and the development of legislative regulatory proposals. The
purposes of E.O. 12291 are: to reduce the burden of existing and future
regulations; to increase agency accountability for and Presidential oversight
of regulatory actions; to minimize regulatory duplication and conflict; and to
ensure well-reasoned regulations.
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
2-5
The major requirements of the Order that appear applicable to the
development of Sections 102 and ,103 regulations are the following:
• The potential benefits to society from the
regulation must outweigh the potential costs to
society.
• An agency (in this case EPA) must choose regulatory
objectives that maximize the net benefits to society.
Under the Order, EPA must determine if the Sections 102 and 103
regulations are "major" rules, that is, if they result in: (1) an annual
effect on the economy of $100 million or more; (2) a major increase in costs
or prices for consumers, industries, or Federal, State, or local government
agencies; or (3) significant adverse effects on competition, employment,
investment, productivity, innovation, or the ability of U.S.-based enterprises
to compete with foreign-based enterprises in domestic or export markets. The
requirement for a Regulatory Impact Analysis (RIA) flows from the
determination that a regulation is a major rule.
If EPA determines that the Sections 102 and 103 regulations will have a
major impact, then EPA must perform the following major tasks in order to
comply with the Order's broad guidelines:
• Prepare a summary of the regulations'
objectives, legal bases, and schedule, to be
included in the annual October regulatory agenda.
• Prepare a rationale for the determination that
the regulation will have major impacts, to be
included in the notice of proposed rulemaking.
• Prepare a preliminary RIA to be transmitted,
along with the notice of proposed rulemaking, to
the Director of 0MB at least 60 days prior to the
publication of a notice of proposed rulemaking.
• Prepare a final RIA to be transmitted to 0MB
along with the final rule at least 30 days prior
to the final rule's publication.
Both the preliminary and final RIA must: describe and analyze alternative
regulatory approaches; describe the potential benefits and costs of the
regulatory alternatives, including those effects that cannot be quantified;
identify the parties receiving the benefits and incurring the costs; and
determine the net benefits, including net benefits that cannot be quantified.
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
2-6
2.3.2 The Regulatory Flexibility Act
The Regulatory Flexibility Act (RFA), enacted by Congress in
September 1980, provides that agencies must fit regulatory and information
requirements to the scale of those businesses, organizations, and governmental
jurisdictions being regulated. The Act requires that agencies solicit the
ideas and comments of small entities and specifically examine the impact of
proposed and existing regulations on small entities. The RFA appears to apply
to EPA's regulatory program for Sections 102 and 103, as well as to all rules
for which agencies publish proposed rulemakings and which are not defined as
wage, rate, service, or price regulations.
A number of the Act's provisions may be applicable to EPA's regulatory
program for Sections 102 and 103. First, if EPA determines that the Sections
102 and 103 regulations are likely to have a significant economic impact on a
substantial number of small entities, then EPA must prepare:
i
(1) A brief description of those regulations' objectives, legal bases, and
schedules. The description must be included in EPA's regulatory flexibility
agenda, and published in the Federal Register.
(2) An initial regulatory flexibility analysis that describes the impact
of the proposed rule on small entities. If feasible, the analysis must
contain an estimate of the number of affected small entities. The analysis
must also include a description of any significant alternative regulations
that meet the objectives of Sections 102 and 103.
(3) A final regulatory flexibility analysis to accompany the publication
of the final Sections 102 and 103 regulations. This analysis must describe:
the need for and objectives of the rule; how public comments affected the
rule's development; and the regulatory alternatives designed to minimize any
significant economic impacts on small entities.
The Act allows agencies to conduct the above analyses in conjunction with
other kinds of regulatory analyses as long as the analyses include all of the
factors mentioned above.
2.3.3 Significant Impact Determination
As can be seen from the two preceding sections, under both the
Regulatory Flexibility Act and E.O. 12291, the determination of a regulation's
impact to some degree is expected to precede an agency's determination that a
regulatory impact analysis is required. This determination has a circular
quality: agencies must first assess a proposed regulation's effects in order
to require an RIA, yet agencies often need to perform something like the RIA
in order to assess the regulation's micro and macro effects.. In this case,
the Office of Emergency and Remedial Response has instructed ICF to analyze
the potential economic effects as if a formal RIA were needed -- although it
may not be — since the analysis is considered useful to the program managers
in understanding the behavior of regulated parties and the effects of the
regulations.
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
2-7
This analysis is intended to help ensure that all of the viable regulatory
alternatives are identified and evaluated. In addition, it is intended to
lead to an understanding of the expected economic effects on all affected
parties. The analyses required by the Regulatory Flexibility Act will be
integrated into the overall economic analysis. This approach will strengthen
the quality of the economic analysis and help fulfill the requirements of the
Regulatory Flexibility Act. In sum, the analyses outlined here will provide
EPA with sufficient background information and a structural framework to
conduct RIAs if they are required.
2.3.4 OMB Draft Guidance on Executive Order 12291'
0MB has developed draft guidance on agency compliance with E.O.
12291, and specifically on the necessary contents of an RIA. To meet the OMB
draft guidance, EPA must include the following information and analyses in its
preliminary and final RIAs for the Sections 102 and 103 regulations:
• A statement of the need for and the effects of the
regulation.
• An examination of alternative regulatory approaches,
including the effects of no regulation, major
alternatives consistent with CERCLA's provisions,
major alternatives outside the scope of CERCLA's
provisions. Agencies must also evaluate alternative
stringency levels, effective dates, and compliance
methods.
• An examination of alternative, market-oriented ways
of regulating.
• An analysis of the net present value of potential
real incremental costs and benefits. OMB further
specifies the use of a 10 percent discount rate in the
analysis (with other discount rates used in the
sensitivity analyses); the clear presentation of all
assumptions and background materials; and the
presentation of a schedule showing the type of effect
and to whom and when it accrues.
• Net benefit estimates. These include any
nonmonetary but quantifiable cost and benefit
information, and the results of sensitivity analyses.
• Rationale for choosing the proposed regulatory
action.
• Statutory authority for the regulatory action.
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
2-8
2.3.5 EPA Draft Regulatory Impact Analysis Guidelines
EPA has prepared specific Agency guidelines for the implementation
of the RIA requirements of E.O. 12291. A portion of the draft EPA guidelines
restates the analytical standards and framework laid out in the Order and in
OMB's follow-up guidance. However, the EPA directives, which will
specifically guide the format and content of any RIAs prepared for the
Sections 102 and 103 regulations, go beyond the 0MB guidance in several
important respects. This section of the chapter reviews several areas for
which EPA has clarified or expanded the requirements for an RIA.
EPA's September 4, 1981 draft guidelines state the following general
directives on the contents and conduct of an RIA:
• Where feasible to use, benefit-cost analysis should
be the principal methodology employed in the RIA.
• Where benefits are not easily quantified, the RIA
should delineate cost-effective regulatory
alternatives.
• The statement of the need for regulation should
concisely state the nature and severity of the problem.
• The RIA must identify regulatory alternatives. EPA
may reject some of these alternatives prior to
detailed analysis because of impact, cost, or
implementation problems. However, the RIA must record
the identified advantages and disadvantages of these
alternatives.
• The RIA must evaluate major alternative regulatory
approaches. This evaluation should contain, where
feasible:
~ An evaluation of potential future changes in
governing statutes;
~ The feasibility of tailoring the regulatory
stringency to stages of processing, particular
industries, or other pertinent groups;
~ The feasibility of issuing joint regulations;
~ The feasibility of different enforcement
options or use of different compliance methods
for different industry segments or types of
activity;
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
2-9
~ An examination of existing or potential market,
judicial, or state or local regulatory mechanisms
that could be used to resolve the problem; and
~ The applicability and feasibility of market-
oriented incentives.
* Incremental costs and benefits should be measured
relative to what would occur without the regulation.
The draft EPA guidelines also contain specific directives on how to
measure the benefits, costs, and equity effects of regulatory alternatives.
The applicability of these guidelines are discussed in the next subsection.
Finally, the RIA should contain a section on net benefits that summarizes
all benefits and costs, including those that are monetized, quantified, and
qualified. The RIA should also explicitly account for nonquantifiable
benefits and costs. In addition, the RIA section on net benefits should
present the distributional effects of the regulation, including the effects on
businesses, on employment, and on communities.
2.3.6 Application of EPA Guidelines to Sections 102 and 103 Regulations
This subsection discusses in more detail the applicability of the
EPA draft guidelines to estimation of benefits and costs of the Sections 102
and 103 regulations.
Benefits. EPA defines benefits as the values that individuals place on
the adverse effects of pollution. Ideally, in order to measure benefits, the
following chain of events must be traced:
• release of hazardous substance;
* changes in ambient quality;
• exposures to people and the ecosystem;
• adverse effects on people and the ecosystem;
* the valuation of those effects.1
Following the first four steps in the benefits measurement chain enables
one to quantify the adverse effects of pollution. The fifth step is the
valuation of those effects.
Willingness-to-pay is one among several methods of evaluating effects.
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
2-10
The EPA guidance acknowledges that this chain is complex to model.
Obtaining the extensive data needed to model this process and developing such
a model are difficult enough when the regulation involves a single substance
or a small group of substances. But the Sections 102 and 103 regulations
encompass over a thousand hazardous substances. Even if the requisite data
were available for all of these substances, the resources available for the
Sections 102 and 103 regulations simply do not allow the detailed
investigation of the effects of the regulation of these hazardous substances.
Instead, we will have to select certain substances to represent all of those
regulated by CERCLA, or use broad aggregates instead of data on individual
hazardous substances. Perhaps the detailed estimation of effects envisioned
in the EPA guidance is feasible for analyzing the effects of designating one
or a small number of substances (assuming, of course, that adequate data are
available). But the analysis for establishing reportable quantities for the
101(14) substances requires a more aggregated approach.
*
The first step in benefits measurement is the compilation of a set of case
studies of releases of hazardous substances which were not promptly reported
to competent government authorities. For each case, the consequences of each
release should be expressed in as much detail as possible. For example, the
consequences of a particular release which was not promptly reported might be
a day of illness for 12 people (after which recovery is complete), the
destruction of 10 acres of wheat, the death of 20,000 fish, etc.
The estimation of the consequences through these case studies will hot
permit the estimation of the relationship between the substance released and
the consequences of the release. Over a thousand substances are already
listed as hazardous under Section 101(14) of CERCLA. In order to develop a
relationship between the identity of the substance and the consequences of the
release it would be necessary to examine thousands of case studies. The
problem is made even more complicated by the fact that the consequences of a
release are a function not only of the identity of the substance, but also of
the volume released, the medium into which the release is made, a large number
of variables idiosyncratic to each release (e.g., weather, the composition of
soil around the release, population density near the release, etc.) and
interactions among these variables. To accurately sort out the relationships
among all of these factors would require the analysis of additional thousands
of cases. Given the complexity of these relationships, the best that can be
done is to treat the consequences of a relatively small number of cases as
representative of a much larger universe.
Costs. EPA's draft guidance defines total social costs as the value of
goods and services lost by society as the result of the use of resources to
comply with a regulation, the use of resources to implement a regulation, and
reductions in output in lieu of compliance. According to the EPA guidance,
total social costs are the sum of five components:
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
2-11
• Real resource costs - the net present value of the
resources used to comply with the regulations.
• Government regulatory costs - the costs incurred by
federal, state, and local governments in implementing
and administering the regulations.
• Deadweight welfare loss - the net loss in producer
and consumer surplus resulting from a change in
production.2
• Adjustment costs for unemployed resources3 - this
includes (1) the value of the resources required to
produce any output lost due to the regulations; (2)
resource allocation costs; and (3) the resources spent
by society to operate programs to help the unemployed.
• Other costs - the costs of changes in innovation,
market structure, or productivity caused by the
regulations, which are not included in the other
categories.
As mentioned in the discussion of benefits measurement, the Sections 102
and 103 regulations encompass over a thousand substances. These hazardous
substances are generated in dozens of industries, each with its own special
characteristics. Because some of the costs of the regulations may vary by
industry, a complete analysis of the' costs of regulations may require a
separate analysis for each of dozens of industries. For example, a recent
analysis of the costs of hazardous substance pollution prevention analyzed 18
major industry groups.4 As with benefits, some of the regulations issued
2This definition of deadweight welfare loss appears to include both
deadweight welfare loss and real resource costs. This is undoubtedly only a
problem with the wording of the guidance.
3Note that the value of the resources required to produce any output
lost due to the regulations is usually not considered part of social cost; the
costs of this lost output are fully represented deadweight welfare loss. The
resources spent by society to operate programs to help the unemployed should
also include resources spent to aid owners of all unemployed factors of
production, not just unemployed workers. In addition, one must be careful not
to mistake the cost of operating these programs for the total cost of these
programs, which includes the operating costs plus any transfer payments made
under these programs. Only the operating costs represent net costs to society.
''Midwest Research Institute, Cost Analysis for Proposed Hazardous
Substance Pollution Prevention Regulation, August 7, 1980.
November 2 Draft * * . * Do not quote or cite
ICF INCORPORATED
-------
2-12
under Sections 102 and 103 may affect only a few industries. But the
reportable quantity regulations are likely to affect a wide range of
industries. For the latter types of regulations, it may be necessary to use a
more aggregated approach, or to make some simplifying assumptions about the
similarities among industries.
As EPA states in its draft RIA guidelines, for many regulations, real
resource costs account for practically the entire total social costs. This
appears to be the case here, and we will concentrate our efforts on the
estimation of real resource costs. The calculation of deadweight welfare
loss, the adjustment costs for unemployed resources, and other costs flow from
the real resource costs and their further effects on the economy. Before
these other costs can be calculated, good estimates of real resource costs
must be obtained.
2.3.7 EPA.Guidance on Reporting Burdens
In addition to guidance on E.O. 12291, EPA has developed guidance
on criteria and formats for estimating reporting burdens. The person-hours
used to gather, compute, record, report, and maintain information comprise the
burden of reporting or recordkeeping requirements. The time that must be
estimated for compliance includes:
• Time to read instructions;
• Time to plan activities;
• Time to create information, including sampling,
testing, or analysis;
• Time to gather information, including assembly,
retrieval, and calculations;
• Time to transcribe information;
• Time to complete forms and applications;
• Time to develop and maintain a recordkeeping
system;
• Time to train personnel; and
• Time for audits.
This kind of information, when applied to a particular regulatory option,
may offer important insights into the costs of alternatives. It would be
discussed in detail in the economic impact analysis prepared for the Sections
102 and 103 regulations.
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
2-13
2.4 SPECIAL TOPICS
This section examines several issues that will likely need to be addressed
in the Sections 102 and 103 regulations. These topics involve a variety of
legal, technical, and economic considerations and primarily pertain to the
applicability of the regulations to different types of substances. While the
resolution of these issues will certainly have economic implications, the
impacts will likely not differ among the regulatory strategies and
alternatives discussed in the remainder of this report.
2.4.1 De-Listing
It appears clear from the language of Section 101(14) of CERCLA
that automatic designation occurs for any substance designated as hazardous
under the statutes mentioned therein. Thus, if EPA designates additional
wastes under Section 3001 of RCRA in the futuire, those wastes automatically
become designated under CERCLA as well. However, if a RCRA-designated waste
is "de-listed" under the mechanism established in the RCRA regulations, it is
likely that the waste will also be delisted under CERCLA, since no independent
determination of hazard will have been made under CERCLA.
In some cases, though, EPA may want to "block" delisting under CERCLA or
redesignate the substance under Section 102. For example, although a
substance may no longer be considered hazardous for the purposes of an air
pollution regulation, it may still be considered hazardous under CERCLA
because of its effects when released into water or on land. Or, EPA may
decide that the applicable water pollution regulations no longer need apply to
a particular substance, but its designation under CERCLA is still necessary to
ensure reporting and liability for accidental releases of the substance. It
is not clear how these possibilities will be treated.
There is also no statutory guidance on delisting substances designated
under Section: 102 of CERCLA. The opportunity to have a substance delisted
would presumably benefit firms handling the substance and the provision of a
delisting mechanism may be important to ensure consistency in EPA's
regulations. These types of issues would need to be addressed in establishing
any rules for delisting.
2.4.2 Radionuclides
Radionuclides are designated hazardous substances under Section 112
of the Clean Air Act and therefore defined as hazardous under CERCLA. Until
regulations are promulgated otherwise, the reportable quantity for releases of
radionuclides is set by CERCLA at one pound.
There are a number of problems involved in adjusting reportable quantities
for radionuclides, which EPA will need to address in its regulations.
Radionuclides are measured in curies rather than pounds, and a one pound
release of some radionuclides is considerably larger than that amount which
should require response. This suggests that the one pound reportable quantity
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
2-14
is inappropriate, both in magnitude and in unit of measurement, for
radionuclides. On the other hand, reporting of radionuclide releases may be
adequately regulated by the Nuclear Regulatory Commission; thus reporting to
the National Response Center may not be imperative. Rockwell is performing an
in-depth analysis of the appropriate reportable quantity for radionuclides and
the applicable response mechanisms.
2.4.3 RCRA Wastes
Section 101(14) of CERCLA includes in the definition of hazardous
substance "any hazardous waste having the characteristics identified tinder or
listed pursuant to section 3001 of the Solid Vaste Disposal Act. . There
are two categories of wastes designated under the Resource Conservation and
Recovery Act (RCRA) for which designation under CERCLA is problematic. The
first category is wastes with the designated RCRA "characteristics." These
characteristics — ignitability, corrosivity, reactivity, and toxicity -- were
defined and identified in the RCRA Subtitle C regulations (May 19, 1980). A
close reading of the section of CERCLA quoted above suggests that all wastes
with these characteristics would be included as designated substances under
CERCLA.
From a programmatic viewpoint, this blanket designation mechanism is a
mixed blessing. On the one hand, it permits EPA to determine that many
substances found at sites, meeting the waste characteristics, are in fact
hazardous. By virtue of designation, EPA can proceed to clean up a site and
recover the costs of cleanup. (CERCLA may not authorize cost recovery for
responses to releases of pollutants and contaminants.)
On the other hand, the number of wastes with RCRA characteristics is
enormous; in fact, no list of such wastes was published when the RCRA
regulations were proposed. Without such a list identifying these wastes, it
would be difficult to set reportable quantities, and the effectiveness of a
program designed to respond to releases of the wastes would be diminished.
Furthermore, although under RCRA, these wastes are supposed to be tested and
identified as hazardous by the generator, if EPA simply declares that
"characteristic wastes" are included in the Section 101(14) list without
naming these wastes, it would create some confusion and place an economic
burden on the releaser.
The second problematic category of substances designated under RCRA is the
list of hazardous waste constitutents (published as Appendix VIII in the May
19, 1980 regulations). The language of CERCLA does not specifically include
these substances. EPA counsel will need to rule on this issue, and the issue
will need to be explained in the regulations.
2.4.4 Mixtures and Pesticides
Hazardous substances may be released as part of a mixture that con-
tains non-hazardous substances. It is unclear whether the RQ determination
should consider the overall mixture quantity or only the quantity of the parti-
cular hazardous constituent of the mixture. This problem may be especially
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
2-15
pronounced with pesticides which are shipped as mixtures. A determination
based on the overall mixture quantity is probably easier from an administrative
standpoint, but such an approach might trigger the Section 103 reporting
requirements when the actual quantity of the released hazardous substance was
below its reportable quantity. It may seem anomalous to require reporting in
such a situation.
If overall mixture quantity is the basis for the RQ determination, greater
costs will likely be imposed on both EPA and the regulated community. The
regulated community may have to report releases it otherwise would not have
reported; EPA may have to process additional reports for releases that are
unlikely to require a response effort or to prompt enforcement action by the
Agency. EPA's current approach is to base the reportable quantity
determination on the actual amount of hazardous substance released rather than
the overall quantity of the overall mixture. Where more than one hazardous
substance is involved in such a release, the reporting requirements are
triggered as soon as the reportable quantity for any one hazardous substance
is reached. This approach can also be problematic because many releasers will
have difficulty determining the quantities of the hazardous substances
contained in mixtures.
2.5 CONCLUSION
This chapter has explained the statutory background for the Sections 102
and 103 regulations and reviewed the available guidance and requirements for
performing a Regulatory Impact Analysis. The chapters which follow have four
major elements: (1) the selection and characterization of a baseline; (2) the
specification of regulatory strategies and alternatives; (3) the identifica-
tion of potential effects of the Sections 102 and 103 regulations; and (4) the
methods of estimating the magnitude and direction of these effects (benefits
and costs).
The chapters devoted to each of these tasks present the relevant
directives on performing the analysis contained in 0MB and EPA guidance. In
each case, we show how our analysis fulfills the requirements for an RIA (if
an RIA should be required) or how the appropriate information can be generated
to fulfill these requirements.
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
CHAPTER 3
BASELINE
3.1 INTRODUCTION
This chapter discusses the baseline for the analysis of the Sections 102
and 103 regulations. The term "baseline" is shorthand for the situation that
would occur in the absence of regulation, or, a base of comparison for
measuring the incremental effects of regulation. Section 3.2 contains a
recommendation on the selection of a baseline. Section 3.3 specifies the
baseline by exploring the rules governing activities during the baseline
period. Section 3.4 characterizes the baseline activities to be assessed and
briefly .describes methods of measuring baseline activities.
3.2 ALTERNATIVE BASELINES
The 0MB draft guidance on E.O. 12291 requires that agencies identify and
evaluate alternative regulatory approaches when performing an RIA. Within
this regulatory alternatives analysis, 0MB stipulates that agencies evaluate
the effects of not implementing the regulation. The EPA draft guidance on the
Order further requires that this regulatory alternative -- the absence of the
regulation — serve as the standard against which incremental benefits and
costs are measured.
Therefore, to analyze the impact of regulatory alternatives, it is
necessary to project public and private sector actions in the absence of the
Sections 102 and 103 regulations. This projection — the baseline -- plays a
crucial role in any evaluation of economic effects, since the effects of the
regulatory alternatives are measured relative to the baseline. Thus, if both
the baseline and a regulatory alternative represent similar or identical
situations, the analysis will show few or no economic effects. The magnitude
and nature of the effects to be evaluated are therefore conditioned on the
baseline.
To clarify what the baseline represents:
1. The baseline does not necessarily represent existing
practices or behavior, although the current state of
the world might be reflected in the baseline to some
extent.
2. The baseline is a projection of what would happen in
the absence of something eventually expected to happen
(i.e., the regulation). As such, the baseline
typically represents a hypothetical situation.
November 2 Draft * * * Do not quote or cite
IGF INCORPORATED
-------
3-2
3. The baseline should provide a base of comparison for
estimating effects. The baseline should not subsume
the effects of regulation; this would render the
evaluation of regulatory alternatives meaningless and
lead to an assessment of the regulations as having no
positive or negative effects, and thus probably lead
to rejection of the regulation.
4. The baseline should be constructed for purposes of
evaluating only the regulatory alternatives, not the
enabling statute. In this case, CERCLA Sections 102
and 103 provide a stringent level of control in the
absence of regulations. The baseline cannot,
therefore, stipulate a world without any notification
provisions; that would in effect ignore Congress1
imposition of such provisions.
5. Because it is difficult to determine "what the world
would be like" in the absence of the regulations, many
different baselines are conceivable. The Sections 102
and 103 regulations form only one part of a complex
set of statutory provisions and associated
regulations, all of which will interact to affect the
behavior of the regulated parties.
In developing this report, we have considered three alternative baselines:
Baseline 1: Pre-CERCLA Behavior: The behavior of
parties prior to CERCLA s enactment would be projected to
continue after passage of CERCLA.
Baseline 2: Statutory Language of CERCLA: Behavior
would be governed by the statutory language of CERCLA, the
legislative history, and any judicial interpretations.
However, behavior would be projected to be unmodified by
regulations interpreting the statute.
Baseline 3: Current Practices: Behavior as it currently
exists would be projected to continue under the baseline.
From these three, we must select one baseline to be used in the analysis that
follows. To ensure that the baseline represents identifiable and realistic
behavior in the absence of Sections 102 and 103.regulations, we evaluated the
baseline alternatives using the following criteria:
(1) The identification of future private and public sector
behavior in the absence of regulation must be based upon
reasonable assumptions about behavior.
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
3-3
(2) The components of the baseline must be relatively easy to
identify and project.
The first baseline assumes that the private and public sectors will
operate as they did prior to CERCLA's passage. Behavior under this baseline
is relatively easy to identify and project. However, because industry and
government behavior has already changed in response to CERCLA, the baseline
assumptions are not realistic. Therefore Baseline 1 has been rejected for
this analysis.
The comparison of Baselines 2 and 3 is more complicated. Baseline 2
assumes that in the absence of regulations, the statutory language of CERCLA
Sections 102 and 103 will prevail. Baseline 3 recognizes that the statutory
requirements are in effect but also recognizes that full compliance with those
requirements may not yet be achieved. Many releasers, for example, do not
know the exact provisions of CERCLA, may be unable to compile a list of
designated substances regulated under the various statutes referenced in
Section 101(14), and are probably uncertain how to comply with notification
requirements for continuous releases.
For these reasons, the current practices baseline (3) appears to be more
accurate as a projection of behavior in the absence of regulations. It would
also generate more data about releases and reporting activities. The problem
with the current practices baseline, however, is that in the absence of
regulations, it is unlikely that current practices would be permitted to-
continue indefinitely. If EPA never promulgated regulations under Section 102
and 103, the statutory language would still prevail; releasers would become
more familiar with the statutory requirements as time passes through industry
trade group or EPA-sponsored publicity. Therefore, current practices are
likely to be only temporary and can be expected to change, in the direction of
greater compliance with the statutory language. In short, behavior observed
over the last year may not be a good predictor of longer-run behavior; it is
the longer-term projection of behavior in the absence of any regulation that
the baseline, should capture. Since most of the provisions of Sections 102 and
103 can only be changed by regulation, the statutory language of CERCLA would
by definition prevail in the baseline period. Therefore the baseline selected
for the economic analysis should be Baseline 2.
3.3 SPECIFICATION OF THE BASELINE
The baseline to be used in this analysis consists of the statutory
language of Sections 102 and 103 of CERCLA. This section of the report
discusses the statutory language and specifies additional elements of the
baseline that include other provisions of CERCLA affecting the behavior of
parties under Sections 102 and 103. In RIAs for the Sections 102 and 103
regulations, it would be important as well to specify additional mechanisms,
such as state and local government programs and judicial institutions that
might serve the same purposes as the Sections 102 and 103 regulations, in
order to determine whether the regulations duplicate or complement -existing
mechanisms.
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
3-4
3.3.1 Statutory Language of CERCLA
The statutory language of Sections 102 and 103 of CERCLA provides
for the following:
(1) Designation. All substances defined as hazardous under
Section 101(14) of CERCLA are designated substances,
subject to all relevant CERCLA provisions. However, new
designations of substances probably will be added by
regulation; therefore, no new designation appears in the
baseline.
(2) Reportable quantities. For designated substances, a
reportable quantity of one pound is established by statute,
unless the substance already has a reportable quantity
under.Section 311 of the Clean Vater Act. These reportable
quantities can only be superseded by regulation; therefore,
they apply in the baseline situation. The requirements
that releasers notify the NRC immediately upon a release of
a designated substance at or above the reportable quantity
is also applicable during the baseline period.
(3) Notification. The statutory language of Section 103(f)
gives releasers an exemption from.immediate reporting of
continuous releases, provided that the nature of the
release is sufficiently established through notification
and that the releasers notify EPA annually (or if there is
a statistically significant increase in the release). The
baseline assumes that releasers will take advantage of this
exemption and fulfill the conditions required by the
statutory language.
3.3.2 Sections 103(b). 106. 107. 108. and 109
The baseline must also include a projection of behavior of parties
under these sections of CERCLA, that deal With penalties relating to notifica-
tion requirements (103(b)), enforcement policy (106), liability for hazardous
substance releases (107), and financial responsibility (108 and 109).
Assumptions about the extent and effects of these activities must be made in
order to characterize the behavior of parties under Sections 102 and 103. The
activities undertaken by potential releasers as a result of the liability
provisions of Section 107 are particularly important to characterize with
respect to designation regulations. Appendix B contains a discussion of the
provisions of Sections 103(b), 107, 108, and 109.
3.3.3 Sections 104 and 105
The provisions of CERCLA Sections 104 and 105 also affect the
projection of public and private sector behavior under Sections 102 and 103.
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
3-5
Section 104 of CERCLA (response authorities) and Section 105 (National
Contingency Plan) authorize the removal and remedial actions that will be
taken for responding to releases of pollutants and contaminants, as well as
hazardous substances, and the methods and procedures under which they will be
carried out. These are important elements of the baseline, because they
define the types of response actions that the government may undertake.
The recommended baseline assumes that both the NCF and the emergency
response program are fully effective. Based on ICF's earlier economic impact
analysis of the NCP and additional discussions with EPA staff, we will
investigate further how the NCP will be implemented, the level of expected •
funding for emergency responses, and the extent of cleanup required for
government removals and for private removals. As the NCP is further revised
and published, we will gain a better understanding of these factors.
3.4 CHARACTERIZATION OF BASELINE ACTIVITIES
Having recommended a baseline and specified the "rules" of the baseline,
this section characterizes baseline activities and discusses their measurement.
The characterization of baseline activities should be linked to the
development of regulatory alternatives. Because the purpose of this study is
to identify potential regulatory impacts and assess their magnitude, it is
necessary to characterize the activities that take place in the baseline and
that would be significantly different under the regulatory alternatives.
3.4.1 Designation
The costs and benefits of the statutory designation of hazardous
substances in Section 101(14) are generally assumed to be part of the baseline
and are therefore not estimated for this analysis. Included here too will be
the effects of designation of new substances under other statutes and their
subsequent automatic designation under CERCLA. The only reason for estimating
costs and benefits from statutory designation is if substances are "delisted"
by regulation. It is also reasonable to assume that the incorporation of
Section 101(14) substances into the Department of Transportation's hazardous
materials regulations is also already part of the baseline. Designation of
substances under Section 102 requires regulation; therefore, no activities
associated with designation are projected to occur in the baseline.
3.4.2 Revision of Reportable Quantities
Several categories of activities associated with reportable
quantity regulations may be undertaken by the regulated community in the
baseline. These include:
Preventive Activities. Firms wishing to avoid the costs of releases may
incur costs to reduce the number or mitigate the effects of releases. These
activities could include:
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
3-6
• purchase of equipment or other capital expenditures;
• improved operation and maintenance of equipment;
• increased inspections;
• improved personnel training;
• improved spill contingency and countermeasure planning;
• modification of operating procedures; and
• labeling of substances.
Information on these activities will be obtained from discussions with
industry members and anecdotal evidence.
Protection Activities. Releasers may incur costs to protect themselves
in the event of a release. This might involve:
• obtaining or revising permits; and
obtaining or increasing insurance coverage.
Discussions with industry and insurance offices and government permitting
officials will provide estimates of the extent of these activities during the
baseline period.
Removal Activities. To analyze the cleanup actions parties undertake in
the baseline period, the following three categories of releases have been
developed:
• releases of pollutants or contaminants (i.e.,
substances not designated as hazardous but which may
be harmful);
• releases of designated substances below reportable
quantity levels; and
• releases of designated substances at or above
reportable quantity levels.
For each of these categories, estimates or assumptions will be needed for:
(1) the response of releasers to these releases;
(2) the extent and cost of cleanup provided; and
(3) the numbers of such releases.
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
3-7
This information will be obtained from interviews with On-Scene. Coordina-
tors and industry personnel, analysis of available data on releases, and
analysis of data on cost of cleanup.
Reporting Activities. A nominal cost is probably associated with the
actual reporting of a release to the NRC since this can be accomplished by
means of a toll-free telephone call. This analysis will instead focus on the
uncertainty facing industry personnel in the event of a release and the
expected probabilities that reporting will occur for different types of
releases. EPA intends to contact industry officials for their assessments of
the statutory impacts. EPA also expects to evaluate industry's perceptions of
the risks of not reporting releases. The major source of information,
however, would be the extent and type of reporting currently received by EPA.
In addition, the incidence of costs on affected parties will be examined
for each of the activities listed above. This enables an estimation of
distributional effects of baseline activities by industry, size of firm, and
geographical area.
Government Activities. The next phase of this effort will attempt to
specify a model that characterizes government activities during the baseline
period, including (a) the availability of Fund monies for removal actions, (b)
the frequency of government response and extent of cleanup provided, (c) the
level of enforcement action and cost recovery actions, and (d) government
permitting processes. It will also be necessary to characterize state and
local reporting requirements and response programs.
3.4.3 Notification
Although the baseline for notification assumes that releasers will
fulfill the conditions for the exemption for immediate notification required
in Section 103(f), it is expected that releasers will interpret these
conditions in different ways, in the absence of regulations. Ve would
therefore need to make .assumptions, based on data from incoming reports, about
the frequency of notification, the amount of monitoring or testing done for
notification purposes, and the level of effort and cost involved in
notification for continuous releases in the baseline period.
3.5 CONCLUSION
This chapter has evaluated three alternative baselines, specified the
recommended baseline, and outlined the activities in the baseline period that
will need to be characterized and measured. The next chapter discusses
strategies and alternatives for regulation under Sections 102 and 103 and
presents a framework for evaluating the effects of the strategies.
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
CHAPTER 4
STRATEGIES AND REGULATORY ALTERNATIVES; EVALUATIVE FRAMEWORK
4.1 INTRODUCTION TO FRAMEWORK
This chapter discusses the development of regulatory alternatives for the
Sections 102 and 103 regulations and explores the underlying strategies that
are needed to generate the regulatory alternatives. The analysis reviews the
strategies proposed by Rockwell and presents several new strategies that
elaborate on ideas and themes developed originally in the Rockwell report.
This chapter also presents an evaluative framework by which the various
strategies may be assessed.
Each, of the three elements of the Sections 102 and 103 regulations --
designation, reportable quantities, and notification -- is considered
separately in this chapter. Section 4.2 discusses EPA and 0MB guidance on
developing regulatory alternatives. Section 4.3 presents strategies for
designating new hazardous substances and setting reportable quantities for
them. Section 4.4 discusses the issues that arise in setting reportable
quantities for chemicals already designated as hazardous substances under
CERCLA. Section 4.5 discusses notification requirements. Finally, Section
4.6 contains a discussion of "non-regulatory" approaches and their
applicability to this analysis.
This chapter concentrates on a discussion of strategies for regulation.
In this context, a "strategy" is an underlying rationale or approach to
regulation. For example, one strategy for revising reportable quantities is
to consider the relative hazard of the substances. Another strategy could be
based on the practicality of administering the RQ regulations. Within each
strategy, several different regulatory alternatives may be developed. These
regulatory alternatives would be based on the same underlying approach, but
would vary the criteria or decision rules used to designate substances or
assign RQs. In other words, regulatory alternatives provide specific
formulations of the broad approach contained in a strategy. The value of
having regulatory alternatives based on the same general strategy is in
allowing EPA to "fine-tune" the regulation by considering variations in the
criteria and decision rules used.
EPA may want to evaluate different strategies (and their different
regulatory alternatives) or to select one strategy and concentrate on its
regulatory alternatives. Both approaches are possible, and by evaluating
different strategies, EPA can ensure that the analysis does not unduly
constrain the available options. Thus, for example, by examining reportable
quantity strategies based on both administrative feasibility and on hazard,
EPA can select the appropriate strategy with a better understanding of the
effects of each approach.
November 2 Draft * * * Do not quote or cite
8CF INCORPORATED
-------
4-2
On the other hand, neither strategies nor regulatory alternatives can be
assessed in the abstract. That is, the final economic analysis can be
conducted only after the regulatory alternatives are specified and the actual
lists of designated substances and RQ levels can be compared. Therefore, to
the extent that EPA can narrow the field of strategies and regulatory
alternatives, the analysis will be more feasible to perform.
A key element of this analysis is the use of an evaluative framework for
assessing and selecting strategies and regulatory alternatives. We present
below a list of five criteria which can be used to evaluate the merits and
impacts of regulatory approaches. Naturally, this list is not exhaustive and
EPA will want to weigh particular criteria more heavily than others in
selecting its regulatory approach. This framework is, however, a useful
analytical tool for organizing the various requirements and factors that must
be considered in development of regulations.
The five criteria proposed in this chapter include:
(1) Scientific Validity: Is the regulatory strategy
defensible on scientific and legal grounds? Does it
adequately protect public health, welfare, and the
environment? Does it yield unambiguous reisults? Is it
easily complied with and consistent with other Agency
findings and approaches and existing governmental
programs?
(2) Executive Order 12291: Do the strategy's potential
real resource benefits outweigh its potential costs?
Does the regulatory strategy maximize net benefits to
society?
(3) Administrative Ease: Will the strategy result in an
excessively resource-intensive regulatory development
process? Will the regulations resulting from the
strategy be enforceable? Will EPA's response to
hazardous substance releases be improved?
(4) Regulatory Flexibility Act: Does the strategy
minimize significant impacts on small entities?
(5) Distributional Effects: Will the strategy result in
inequitably distributed costs or benefits?
4.2 GUIDANCE ON DEVELOPING REGULATORY ALTERNATIVES
A major component of any RIA that may be prepared is the specification of
alternative regulatory approaches. There are a number of Congressional,
Presidential, OMB, and internal EPA guidelines applicable to this task:
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
4-3
• EPA must describe in the RIA the most important and
feasible alternatives that may be able to achieve the
objectives of the Sections 102 and 103 statutory
provisions at lower cost than the proposed
alternative. EPA's draft regulatory impact guidance
further suggests that, in the initial stages of the
evaluation, an agency may eliminate alternatives from
further scrutiny on the basis of costs or
implementation problems.
• EPA must consider regulatory alternatives that
minimize adverse economic effects on small entities
under the Regulatory Flexibility Act (RFA). Examples
of possible alternatives include: different
compliance or reporting requirements for small
entities; the consolidation or simplification of small
entity reporting requirements; the use of performance
.rather than design standards; and small entity
exemption from certain requirements.
• EPA must identify alternatives within CERCLA's
scope. The 0MB draft guidelines state that these
alternatives can include: alternative stringency
levels; alternative effective dates; and alternative
methods of compliance. The EPA draft guidelines
further suggest that this section of the RIA can
examine the feasibility of: tailoring the degree of
stringency to stages of processing, particular
industries, or other piertinent groups; or issuing
regulations jointly and using different compliance
methods for different industry segments or types of
activity.
• . EPA must identify whether there are major
alternatives that lie beyond the scope of the
proposed regulation s statutory provisions. The EPA
draft guidance suggests that this review can include a
look at any needed statutory changes.
• EPA must examine the feasibility of implementing
market-oriented ways of regulating (i.e., non-
regulatory options). The 0MB draft guidelines require
that an RIA include an analysis of the feasibility of
information or labeling alternatives, performance
rather than design standards, and economic
incentives. The EPA draft guidelines suggest
examining the transition costs of moving from the
existing control system to any of the market-oriented
alternatives.
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
4-4
• Identified alternatives must include the
alternative of having no regulation. As discussed in
Chapter 3, this alternative would serve as the
baseline.
4.3 STRATEGIES AND REGULATORY ALTERNATIVES FOR DESIGNATION
This section first discusses the rationale for designation of hazardous
substances, an^ then presents several designation strategies and a discussion
of the linkage between designation and assignment of reportable quantities.
4.3.1 Rationale for Designation Regulations
EPA is permitted to designate as many substances under Section 102 of
CERCLA as can be justified. Even without regulatory action, over 1,100
substances are .already defined as hazardous under Section 101(14), and more
would be added to the list whenever designation occurs under the statutes
listed in 101(14). At least three reasons for designation regulations can be
identified:
(1) Designation of a substance is a prerequisite for
required reporting of releases of the substance.
(2) Designation activates the liability and financial
responsibility provisions of Sections 107 and 108 of
CERCLA. Thus releasers of designated substances can.
be liable for cleanup costs and damages to natural
resources, generally up to $50 million. (See Appendix
B for a detailed discussion of Sections 107 and 108.)
Designation also entails the listing of a substance as
a hazardous material under the Hazardous Materials
Transportation Act (Section 306(a) of CERCLA).
Designation also triggers liability insurance
requirements for carriers and shippers of hazardous
substances, under Section 30(B) of the Motor Carrier
Safety Act of 1980.
(3) Designation of substances may increase the likelihood
of recovering the costs of Superfund-financed
responses.
In the abstract, strategies for designating hazardous substances can be
based on any or all of these rationales. The main source of guidance on
designation is the language of Section 102, which mandates designation for
substances which, "when released into the environment may present substantial
danger to the public health or welfare or the environment." This suggests
that the potential for harm or hazard may be the primary criterion for
designation.
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
4-5
4.3.2 Designation Strategies
Rockwell International has identified five strategies for the designation
of hazardous substances:
1. Retention of substances listed in Section 101(14),
with no further designation. (This strategy
represents the baseline in this analysis, discussed in
Chapter 3.)
2. Use of a comprehensive list for designation, such as
the Toxic Substances Control Act (TSCA) inventory.
3. Use of a "Blue Ribbon Committee" to select substances
for designation.
4. Use of a Hazard Index to rank all candidate hazardous
substances by their potential degree of harm, and
selection of a cutoff level for designation.
5. Derivation of candidates for designation from a master
list based on preliminary screening criteria (i.e.,
"on/off" switches or gates).
The discussion that follows examines in more detail three strategies that
represent extensions of Rockwell's proposed strategies. For each strategy, we
suggest how to determine regulatory alternatives and briefly discuss the
economic and regulatory implications. An evaluative framework for these three
strategies is presented in Exhibit 4-1.
Strategy A: Designate all chemical substances found at
hazardous waste sites or commonly involved in accidental
releases which have a clear potential for harm. (This strategy
is derived from Rockwell s strategy 2.)
This strategy serves two primary purposes. First, it brings a fairly
large number of substances under the reporting requirement. If the reportable
quantities for these substances are set fairly low, this strategy would almost
eliminate the possibility of a release of a hazardous substance without EPA
notification. Second, this strategy could perhaps increase the likelihood of
successful cost recovery of funds expended in response actions, thereby
conserving the Superfund and stretching its resources. The determination of
which substances are potentially harmful is a difficult problem, and is
discussed in Strategies B and C below. This strategy, however, focuses
attention on substances that are actually being released into the environment.
Regulatory alternatives based on this strategy would differ in the
coverage of substances designated. For example, EPA could designate all
harmful substances currently found at hazardous waste sites or only those
found at a particular site before remedial action is started. Alternatively,
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
EXHIBIT 4-1
EVALUATION OF DESIGNATION STRATEGIES
Strategy A
Designate all substances
commonly released which
have a clear potential
for harm (Rockwell's
strategy 12.)
Strategy B
Hazard index and panel
of experts (Bbckwell's
strategies 13 and 14.)
Basic
Premises
Near total coverage
of hazardous sub-
stances by regula-
tions; Increased
potential for cost
recovery.
Designation on the
basis of relative
hazardousness of
substances.
Scientific
Validity
Not strictly related
to hazard.
E.O.
12291
Undetermined
Hay be hard to
defend.technical
merits of index,
choice of cut-off
point.
Undetermined
Administrative
Ease
RFA
May be easiest for regu- Could exclude
latory process, but could
impair response program
if too many reports re-
ceived; may be hard to
enforce.
Process may limit desig-
nation to relatively few
substances.
substances
handled mostly
by small busi-
nesses.
Undetermined
Distributional
Effects
Potentially high
costs associated
with liability
it*
o>
Number of substances
designated, and
hence, effects,
would depend on
criteria and
decision rules.
Strategy C
"On/Off" criteria
(Rockwell's
strategy 15.)
Designation on the
basis of simple
absolute standard
of hazardousness.
May be most defen-
sible if very strict
criteria were used.
Undetermined
May be difficult to
limit regulatory pro-
cess on an on-going
basis.
Criteria could
incorporate con-
cerns for
effects on small
business.
Number of substances
designated, and
hence, effects,
would depend on
stringency of
criteria.
November 2 Draft * * * Do not quote or cite
-------
4-7
EPA could consider designating many harmful chemicals from the TSCA inventory,
depending on their release histories. Or, if EPA is more interested in
receiving reports of releases than in imposing liability under CERCLA, EPA
could consider requiring reporting for the majority of these substances under
Section 8 of TSCA and designating relatively few substances under CERCLA.
The effects of this strategy depend on how many and which substances are
designated. If reportable quantities are low, both industry and government
could be burdened by making and processing many additional reports of
releases. If many substances are designated, releasers might incur large
insurance costs due to the CERCLA liability and financial responsibility
provisions. (This effect would depend on how insurance mechanisms for release
liability operate, and whether the substances designated are handled by
releasers who are already substantially covered by insurance.) On the other
hand, designation of many substances may be necessary to activate industry's
investment in preventive measures. Since many substances would probably be
designated under this strategy, it will be necessary to investigate the
economic effects in more detail to determine if this strategy is appropriate.
Strategy B: Establish a Hazard Index and rate the degree of
hazard of all candidate hazardous substances. Based on this
initial screening, use a panel of experts to select substances
for designation. (This strategy combines Rockwell's strategies
3 and 4 — hazard index and blue ribbon committee.)
This strategy explicitly relates designation to the potential
hazardousness of the substance,1 but does not begin with the same premise as
Strategy A above, that the candidates for designation should be derived from
experience at sites and spills. Rather, this strategy encompasses far more
potentially hazardous substances and proposes a method for ranking them
according to their degree of hazard if they were released at a site or spill.
Because many different hazard indices could be constructed, and because the
cut-off points would necessarily be somewhat arbitrary to determine when a
substance is hazardous or not, the defensibility of the hazard index would be
enhanced if it represented only the first step in the designation process and
if it were accompanied by something like a panel of experts to make the actual
selection of substances for designation -- as suggested by Rockwell.
Regulatory alternatives based on this strategy could vary the criteria
used in constructing the hazard index or the criteria used in final
assessments by the panel of experts.
A simple set of regulatory alternatives could be based on the cut-off
ratings used in screening substances. For example, under one regulatory
lRefer to Rockwell's report, "Technical Base for Designation of
Hazardous Substances and Assignment of Reportable Quantities," (November 2,
1981) for a discussion of possible hazard indices.
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
4-8
alternative, only substances with a "90" rating or higher on the hazard index
would be considered for designation. Another alternative could use a rating
of "50" as the cut-off point, and so on.
The attractiveness of this strategy lies in its coupling of relatively
objective assessments of hazard and subjective assessments based on expertise
and experience. Both these assessments, however, may be subject to challenge,
since a different hazard index or group of experts would likely produce
different results. A further consideration in the use of this strategy is the
number of substances that might ultimately be designated. Experience with
other EPA regulatory programs has shown that with limited resources,
relatively few substances can be investigated in depth at any one time. If
only a handful of substances may need to be designated or if the regulations
are phased in, this approach could be useful. If, however, hundreds of
substances may eventually need to be designated, this process may prove
administratively unsuitable.
# Strategy C: Establish a checklist of "on/off" criteria for
designation. Substances meeting the most stringent criteria are
subsequently designated. Substances meeting less stringent
criteria or lacking sufficient data remain "candidates" for
designation. (This strategy is derived from Rockwell s
strategy 5 -- use of a master list and screening criteria.)
Under this strategy, EPA would develop a series of "on/off" criteria or
gates (rather than a ranking system), based on factors such as degree of
hazard, release history, and possibly economic factors. "On/Off" in this
context means that a substance would be "on" the designation list if it met
certain criteria and "off" if it did liot. Under this option, EPA would
actually designate only those relatively few substances that meet the
strictest criteria (such as proven carcinogenicity or extreme flammability,
regulation tinder Section 6 of TSCA, or cancelled registrations for almost all
uses of a pesticide).
EPA would consider a second tier of substances as candidates for
designation if they met less strict criteria or if data were not available to
determine their effects based on the strictest scientific review. These
substances would still be of concern to EPA and reporting of releases of these
substances would be considered "important. EPA could encourage industries to
report these releases voluntarily and to take additional precautions in
handling the substances. Or, EPA could provide for mandatory reporting of
these releases under Section 8 of TSCA which grants such authority. If
further evidence, based on release history or health effects data, shows that
the substance can meet the most stringent criteria, EPA would then designate
the candidate substance.
Regulatory alternatives based on this strategy would vary by using
different on/off criteria or by varying the decision rules that trigger the
designation of a candidate substance.
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
4-9
This strategy has several positive elements. It would permit a large
number of substances to be evaluated for designation. It would also establish
stringent and defensible criteria for designating substances. In this
respect, this strategy differs from Strategy B, by providing a "yes or no"
definition of whether the hazard warrants designation, rather than a ranking
of the relative hazard of substances. Third, and perhaps most important, this
strategy could permit EPA to attain the reporting goals of Sections 102 and
103 for a large number of candidate substances, while relieving industry of
the burden posed by CERCLA's liability and financial responsibility provisions
by not designating these candidates.
One potential problem in using this strategy is that once the set of
criteria has been established, many substances not originally considered by
EPA might qualify for designation. EPA might need to refine its criteria or
develop another method by which to limit the number of substances considered
for designation on an on-going basis.
To summarize, all three of these strategies (and others not considered
here) may be evaluated for their regulatory and economic impacts (as we have
shown, in a preliminary manner, in Exhibit 4-1). Or, EPA may select one or
two of the strategies and analyze regulatory alternatives within the
strategies. The use of more than one strategy would increase the complexity
of the analysis considerably, because of the need to develop separate sets of
criteria and decision rules for each regulatory alternative within each
strategy.
Whatever strategies are ultimately selected by EPA, they will likely need
to address explicitly the following elements in the designation process:
(1) Selection of a master list of substances to be
evaluated.
(2) Criteria for evaluating substances on the master list.
(3) Decision rules for selecting substances to be
des ignated.
Although we have dwelt mainly on the second and third elements, the
selection of a master list is a crucial step in any strategy. (In fact,
Strategy A gives a great deal of attention to this step.) Clearly, the
selection of the master list will be a major determinant of the economic
effects of the designation regulations. Along with the stringency of the
criteria and decision rules used for designation, the number and type of
substances included on the master list will determine the number and type of
substances eventually designated. However the master list is constructed,
there will likely be some pressure on EPA to include or exclude some
substances. Therefore, a clear statement of EPA's approach in constructing
and revising the master list will be essential.
As noted earlier, some regulatory processes may be less suitable for
achieving the designation of a large number of substances. In addition, the
designation of several hundred additional substances may have very different
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
4-10
economic effects than the designation of an added handful of substances.
Although we cannot indicate whether the net benefits of designating many
substances would be larger than the net benefits of designating few
substances, we hypothesize that this question will be an important factor in
the selection of designation strategies.
4.3.3 Designation and Reportable Quantities
For each substance designated under Section 102, EPA must also establish a
reportable quantity (RQ). This subsection discusses the relationship between
the designation process and the establishment of an RQ.
As mentioned earlier, the basis for designation should probably rest on
the potential harmfulness of a release of a substance. As will be discussed
in the next section, there is no such guidance for reportable quantity
regulations. In fact, the only guidance in the legislative history of CERCLA
suggests that administrative feasibility and practicality should be used in
setting reportable quantities. On the other hand, EPA can presumably
establish RQs solely on the basis of the relative hazard of the substance, or
any other justifiable criteria.
EPA should perhaps keep two points in mind in selecting a method of
setting reportable quantities for newly-designated substances. First, the
establishment of reportable quantities for newly-designated substances should
ideally follow the same basic and logical approach as the revisions of
reportable quantities for substances already designated by statute. Second,
the designation and RQ regulations have overlapping but different effects.
Designation activates not only reporting of releases but also the liability
and financial responsibility provisions of CERCLA. On the other hand, RQs
relate largely to the reporting of releases. Because of this difference,
designation and RQs do not have to be strongly linked. Thus, it would be
entirely possible to designate substances as hazardous and to then assign them
high reportable quantities. For example, EPA might designate a substance
because its known adverse health effects meet the designation criteria.
However, based on its release history and limited applications, EPA might
decide to set a relatively high RQ for the substance. In other words, the
intuitive conclusion that highly hazardous substances should have low RQs may
not be correct in all cases.
4.4 STRATEGIES AND REGULATORY ALTERNATIVES FOR REVISING REPORTABLE QUANTITIES
The regulatory alternatives that EPA will consider for setting reportable
quantities can be directed at the two categories of designated substances
listed in Section 101(14) — those with a statutorily-established RQ of one
pound and the Section 311 substances with previously established RQs. Because
the RQs for the Section 311 substances have already been established by
regulation under the Clean Water Act, EPA will likely want to concentrate its
efforts on the substances with the statutory RQs of one pound. However, EPA
may also want to change RQs for some or most of the Section 311 substances,
particularly since those RQs, based on levels of aquatic toxicity, were set in
the context of releases into navigable waters, rather than the broader,
multi-media scope of CERCLA.
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
4-11
This section describes the rationale for revising reportable quantities,
certain important issues in designing the reportable quantity regulations, and
the strategies and regulatory alternatives for revising RQs.
4.4.1 Rationale for Revising Reportable Quantities
The rationale for revising reportable quantities stems from the rationales
for required reporting of releases at or above a given triggering quantity.
Required reporting ensures that response will be made to the release, if a
field assessment by an On-Scene Coordinator indicates that response is
necessary. It also encourages releasers to respond, increases the speed of
response, and provides for expert government assistance when needed. The
objective in setting reportable quantities is to find an appropriate trade-off
between the probability of harm to the public or damage to the environment
arising if a release is not reported, the cost of cleanup if the release is
not reported in a timely fashion, the cost to industry of making the report,
and the cost to the government of receiving the report.
The legislative history of CERCLA makes clear that EPA does not have to
consider the probability of harm arising from an unreported release.2 EPA
2The Senate report of the Committee on Environment and Public Works on a
predecessor bill to Superfund, states:
The provision intentionally omits from the requirement to determine
"reporting" quantities any reference to harm or hazard. A single quantity
is to be determined for each hazardous substance, and this single quantity
requires notification upon release into any environmental medium. It
would be virtually impossible to determine a single quantity applicable to
all media while at the same time linking such quantity to any subjective ,,
concept of harm.
It is essential that such quantities be relatively simple for those
subject to notification requirements to understand and comply with. Since
releases in such quantities trigger notification requirements, but do not,
in and of themselves, give rise to other liabilities under this Act, the
President's broad discretion to select quantities will not unfairly burden
those persons subject to the Act.
In determining reportable quantities under this paragraph [S3(a)(2)], the
President may consider any factors deemed relevant to administering the
reporting requirement or the President's other responsibilities under this
Act. Administrative feasibility and paracticality should be primary
factors. In addition, the President may revise such regulations from time
to time if under-reporting or over-reporting is occurring under existing
regulations.
U.S. Congress. Senate. Environmental Emergency Response Act. S. Rept.
96-848. To accompany S. 1480. 96th Cong., 2d sess., 1980.
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
4-12
may establish the RQs solely on the basis of administrative feasibility and
practicality. Thus, a rationale for the RQ revisions may simply be that at an
RQ of one pound, EPA would be overwhelmed by so many reports that its response
program would be impaired.
4.4.2 Issues in Designing Reportable Quantity Regulations
Practically speaking, only a relatively small range of reportable
quantities can be considered in any analysis. For example, on the low side,
it would be infeasible to require reporting of releases below the level of
detectability. Thus, it makes little*.sense to set a reportable quantity of
one gram, if a responsible party is unlikely to notice or to be able to detect
a release of one gram for most substances. Moreover, enforcement officials
may be unable to detect violations of the regulations if reportable quantities
are set at very low levels.
Numerical Value of Reportable Quantities: An important issue in
designing reportable quantity regulations is the numerical value and number of
levels of RQs. This issue can be examined by investigating two options for
setting reportable quantities:
1. A reportable quantity can be any integral number of
pounds; or
2. A reportable quantity must be one of a restricted set
of values, such as the 1, 10, 100, 1,000 or 5,000
pounds intervals used in the Section 311 CWA
regulations.
In the first case, EPA would need to justify the choice of, say, 10 pounds
rather than 11 pounds as a reportable quantity on the basis of benefits and
costs. In the second case, EPA would need to justify the choice of 10 pounds
rather than 100 pounds. Given the uncertainties associated with the
measurement of benefits and costs, justifying the choice of a reportable
quantity is much easier in the second case than in the first case.
The use of a restricted set of reportable quantities also provides
operational advantages. In theory, each regulated party would instruct
employees on how to detect releases of hazardous substances in reportable
quantities, and would label the container holding the substance so that the
reportable quantity could be quickly determined. Those labels and
instructions would be much easier to comprehend if employees used only a
restricted set of reportable quantities rather than a very large set of
integers. Morever, unless the reportable quantity levels are reasonably
distinguishable, releasers may find it impossible to determine whether a
release at or above the RQ level has in fact occurred.
The decision on the number of reportable quantity levels and the value of
the RQs is an important one, but not necessarily germane to the development of
regulatory alternatives. Thus, a set of reportable quantity levels may
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
4-13
include 4 such levels, or 5 or 3, or any other reasonable number. The levels
can proceed by order of magnitude (such as 1, 10, 100, 1000 pounds or 5, 50,
500, 5000 pounds) or by another reasonable progression (such as 1, 10, 100,
500, 1000, 5000).
The current set of RQs for Section 311 substances (1, 10, 100, 1000, 5000
pounds) are proportional to the ranges established for the toxicity
categories, with the smallest commercial container size (1 pound) defined as
the harmful quantity for the most highly toxic category of substances. EPA
may, however, wish to change the RQ levels under CERCLA. Such a decision
could be made on the basis of three factors: (1) the ease to releasers of
distinguishing between RQ levels; (2) the consistency of the RQ levels with
current regulations with which industry is familiar; and (3) the prevailing
methods of containerizing and transporting hazardous substances. For example,
EPA may find that RQs should be set in terms of gallons released rather than
pounds, since releases are most commonly reported in units of gallons. EPA
may also want to consider a category of reportable releases corresponding to
the category of "sheen" used in oil discharge reporting for releases of
undeterminable quantities but visible effects.
In this report, we have formulated regulatory alternatives for revising
RQs on the basis of the assignment of RQs to hazardous substances. It is
possible, however, that in an RIA for revising reportable quantities, EPA may
want to evaluate more formally regulatory alternatives based on the numerical
value of the reportable quantity levels.
Varying Reportable Quantities by Medium. Section 102(a) of CERCLA gives
the Administrator the authority to set a single reportable quantity for any
hazardous substance, regardless of the medium into which that substance is
released. The legislative history clearly prefers, although CERCLA does not
mandate, that a single reportable quantity be set for all media (see footnote
2 on page 4-11). If a single reportable quantity is chosen, that single
quantity could be constructed from the reportable quantities for the
individual media in several ways. Some of these ways are:
• Establish a single reportable quantity at the lowest
of the reportable quantities developed for each medium;
• Establish a single reportable quantity at the
reportable quantity for the medium which is most in
need of response;
• Establish a single reportable quantity at the
reportable quantity for the medium into which releases
most frequently occur.
Whatever method is chosen, some justification will have to be provided in
the regulations for how the RQs take into account differences between releases
into different media. One simple way to account for this in strategies based
on hazard, is to base the hazard criteria on the medium in which releases are
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
4-14
most likely to occur or on the most sensitive medium. For example, if a
release of a substance into the air is considered most hazardous, inhalation
toxicity levels might be used in the criteria. Although data may not be
available to do a full analysis o-f the effects of a substance in all media,
EPA will likely want to check that the RQ for each substance does not result
in obvious errors. At a minimum, EPA will want to ensure that the RQs are
consistent with the guidance provided On-Scene Coordinators on responding to
releases of substances into different media.
4.4.3 Strategies for Revising Reportable Quantities
Rockwell has identified the following seven strategies as potentially
worthwhile approaches to revision of reportable quantities:
1. Strategies based on Administrative Feasibility and
Practicality, using factors such as container size or
production volume to assign RQs.
2. "Worst-Case" Scenarios to model the behavior of
substances released into different media.
3. Fate and Effects Research to determine the ultimate
fate of released substances and their effects on all
exposed organisms.
4. Use of a hazard index to assign RQs.
5. Combined approach of a hazard index and "fine-tuning"
on a case-by-case basis to incorporate administrative
feasibility factors.
6. Selective criteria processing, using one or more
criteria of hazard (such as human or mammalian
toxicity) to assign RQs in a manner similar to the RQ
regulations for Section 311 substances.
7. Chemical Manufacturers' Association strategy, based on
an adjustment of the Section 311 RQ framework using
technical criteria suitable for CERCLA hazardous
substances.
The remainder of this subsection elaborates on these strategies and
assesses them using the evaluative framework proposed earlier. Where
possible, regulatory alternatives are suggested. The strategies considered
here are by no means exhaustive. The primary purposes of this subsection are
to complement Rockwell's report by showing the relative advantages and
disadvantages of the three types of strategies discussed here: those based on
administrative feasibility, those based on hazard potential, and those
combining both hazard and feasibility factors. Strategy A below is a
variation on Rockwell's strategy 1, Administrative Feasibility. Strategy B is
November 2 Draft * * * Do not quote or cite ,__
ICF INCORPORATED
-------
4-15
Rockwell's hazard index strategy. Strategies C, D, and E elaborate on the
strategy developed by Rockwell of combining hazard and administrative
feasibility criteria in an overall approach. Rockwell's strategies 2, 3, 6,
and 7 are not discussed here because at this stage, their effects cannot be
distinguished from related strategies. Exhibit 4-2 displays a preliminary
assessment of the strategies using our evaluative framework.
Strategy A: No Prior Screening (Administrative Feasibility).
(Derived from Rockwell's Strategy #1).
This strategy represents one formulation of Rockwell's strategy #1.
It is based on an aspect of administrative feasibility -- ,the amount of
resources available to screen reports of releases. The analogy here is to a
police or fire department, each of which, like the NRC, receives numerous
"false alarms." The decision EPA must make is whether to screen out false
alarms before they are reported by assigning fairly high RQs, or whether to
screen the reports as they are received. Since reports of releases are
screened, anyway at the NRC before response is undertaken by a federal On-Scene
Coordinator, a strategy of "no prior screening" would simply require
additional resources to handle the calls at the NRC and EPA Regions. EPA
could leave the statutorily-set RQs at one pound or raise them to a higher
level on the basis of the resources available for receiving reports. This
process would be iterative, as EPA gained experience with the amount of
resources needed to handle incoming calls. Under this strategy, EPA would be
weighing the costs of reporting to industry many low-level releases, the
benefits of collecting the information, and the benefits of being able to
respond when necessary to potentially harmful releases of small quantities of
hazardous substances.
One fact that lends support to this strategy is the low level of reports
currently received by the NRC. We have been told that the NRC at present is
receiving about 35 to 50 reports per week of CERCLA-designated hazardous
substances. Given this level of reporting, it may not be necessary to develop
elaborate prior screening regulations.
Strategy B: Hazard Index (Rockwell's Strategy #4).
As discussed previously under designation strategies, a hazard index
could be devised to evaluate the relative toxicity and other potential adverse
effects of substances. The hazardousness of substances would be rated based
on physical and chemical properties and other characteristics.
Regulatory alternatives based on the hazard index strategy could vary the
criteria for evaluating hazard or could vary the relationship between hazard
category and RQ by using different cut-off points.
This is a viable approach that directly links, as much as is technically
possible, the RQ regulations to the concept of hazard. As discussed, this
linkage appears to be permissible and to some degree desirable, since response
to releases should be conditioned on minimizing or eliminating harmful
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
EXHIBIT 4-2
EVALUATION OF STRATEGIES FOR REVISING REPORTABLE QUANTITIES
Strategy A
No prior screening
(Rockwell's strategy |1
Administrative
Feasibility)
Basic
Premises
Current experience
(and analogy to
fire department)
suggest that high
PQs may not be
needed to screen
out reports.
Scientific
Validity
May be justified by
current reporting
experience.
E.O.
12291
Hay result in
high reporting
costs.
Administrative
Ease
RFA
Would be based on avail- Undetermined
ability of administra-
tive resources.
Distributional
Effects
May required exces-
sive reporting by
some releasers.
Strategy B
Hazard Index
(Rockwell's strategy t4)
Strategy C
Hazard Index and Fine-
Tuning
(Rockwell's strategy #5)
Assumes that RQs
should be directly
linked to the re-
lative hazardous
ness of substances.
Would adjust
hazard-based RQs
by incorporating
administrative and
economic concerns.
Index and relation-
ship to RQs may
be controversial.
Difficulties in for-
mulating "fine-
tuning* criteria.
Undetermined
May require intensive
regulatory efforts.
Undetermined
Effects would depend
on rating criteria.
I
cn
Undetermined
May require large regu-
latory development costs.
Could incor-
porate concerns
for small
business.
Distributional
effects could be
mitigated by
fine-tuning.
Strategy D
Empirical Evidence
(Rockwell's strategy #5
— Combined Approach)
Strategy E
Two-Tiered Approach
(Rockwell's strategy <5
— Combined Approach)
o
By waiting until
enough experience
has accumulated,
EPA can base RQs on
actual record of
need for response.
Assumes that RQs
have less effects
than designation
regulations! uses
RQs in safeguarding
role for most
highly hazardous
substances.
Relies on experience
in the field; may be
most defensible.
Undetermined
Acknowledges uncer-
tainty in setting
RQs.
Undetermined
May encounter problems
with insufficient sample
size, disincentive to
report.
May be easiest to develop
and administer.
Undetermined
Undetermined
Targets RQ regula-
tions appropriately.
Effects would depend
on criteria for
assigning low RQ
level.
November 2 Draft
Do not quote or cite
nt
o
-------
4-17
exposure. There are, however, some difficult problems associated with this
strategy. First, as Rockwell points out, there are severe data limitations
that make relative rankings difficult to obtain. Second, there is no general
agreement in the scientific community on how hazard factors are to be combined
to form an acceptable hazard index. Rockwell has formulated four different
algorithms that illustrate this precise problem. It is especially unclear how
to incorporate aggravating factors, such as persistence and bioaccumulation.
Finally, this approach does not provide a direct or conclusive link
between an RQ and the need for response, relying instead on potential hazard
as a measure of the need for response. It may, however, be possible to
construct a hazard index that will incorporate additional factors relating to
the need for response.
Strategy C: Hazard Index Combined With Fine-Tuning to Incorporate
Administrative Feasibility (Rockwell s Strategy #5).
_This strategy would combine the initial use of a hazard index with
subsequent "fine-tuning" to ensure that the RQs assigned are reasonable. One
method of fine-tuning is to rely on a panel of experts with the experience
needed to evaluate the RQ assignments. A second method would be to construct
an additional set of criteria, based on economic and other factors, to adjust
the RQ assignments.
One problem with the latter approach is that it is not clear what criteria
should be used to revise the RQs. A criterion commonly mentioned is the
production volume of a substance, which is intended as a proxy for the
potential for exposure to a substance. The volume of production, however, is
an inexact measure which does not consider methods of handling, transporta-
tion, and types of use — in short, those factors that constitute release
potential. Even the use of a criterion such as release potential is
problematic. It is not clear that a highly toxic substance with a very low
release potential (or production volume) should not be reported if it is
released in low quantities. If the substance is infrequently released, there
will be little reporting burden on industry and the government. If, however,
it is highly toxic, even a small quantity should perhaps be reported and
cleaned up.
Strategy D: Empirical Evidence (Hazard and Administrative
Feasibility) (Derived from Rockwell s Strategy #5 —
Combined Approach.)
This strategy adopts an approach to revising RQs based on the data
collected in the response program. Since the reporting requirement is already
in place under CERCLA, over time EPA will be acquiring data on reporting under
existing RQs. Under this strategy, EPA would defer assigning RQs until it had
accumulated enough experience with actual reports of releases. Then, EPA
would develop decision rules that relate RQs to the frequency of response to
releases of different sizes.
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
4-18
As an example, assume a substance had an RQ of one pound and that over a
specified period of time (say, six months or one year), the following data
were recorded:
For releases under 10 lbs., responses were made 5% of
the time.
For releases between 10 and 100 lbs., responses were
made 25% of the time.
For releases between 100 and 1000 lbs., responses were
made 50% of the time.
Using a specified cut-off point (say, 50 percent response rate), the RQ
for this substance would be set at 100 pounds.
The attractiveness of this strategy is that it is based almost entirely on
actual experience and can easily be modified as further experience develops.
This strategy does not need to rely on models, assumptions, or incomplete past
data to obtain estimates of release potential. In addition, it goes beyond
the estimation of release potential to get at the heart of the regulations:
the need for response. By relating the RQ directly to the response history of
a substance, this strategy also incorporates a concept of hazard.
Some variations and refinements can be developed to counter this
strategy's faults. For example, the regulations could be delayed until enough
of a record had accumulated to propose the actual RQs (as well as the
approach). This would minimize any disincentive to report releases. Or,
while a public "trial" period was in progress, EPA could raise most RQs to 10
pounds to reduce the interim reporting burden on releasers.
The main problems with this strategy are that it works well only when a
low RQ is in effect; it also relies on appropriate response decisions being
made by On-Scene Coordinators; it does not recognize that response history may
depend in part on funds availability; and it is possible that an insufficient
sample size would be generated within a specified time period to perform a
full analysis on a given substance. Furthermore, this strategy could cause
some confusion in the regulated community because of the potential lack of
finality in the assignment of RQs.
Regulatory alternatives based on this strategy could vary the cut-off
points or decision rules used to assign RQs to frequency of response, or they
could differ in whether industry response or government response is considered
to be the criterion.
Strategy E: Two-Tiered Approach (Hazard and Administrative
Feasibility) (Derived from Rockwell's Strategy #5 —
Combined Approach).
This strategy again combines considerations of hazard and
administrative feasibility, in perhaps the simplest approach of all. Under
this strategy, EPA would choose a single reportable quantity level for the
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
4-19
majority of substances, based on estimates of reporting burden, probability of
need for response, or the Agency's experience to date. However, for thosfe
substances which are known to be extremely hazardous, a lower RQ would be
set. The assignment of these substances to the low RQ level could be done on
the basis of Agency experience and best scientific judgment or by constructing
a checklist of stringent hazard criteria.
This strategy can best be envisioned as the "big box/little box" approach
to RQs. If a substance is very hazardous, it fits into the "little box" with
the other (relative few) substances that have one pound RQs. If, on the other
hand, the substance is harmful only in large quantities, it fits into the "big
box" with the many other harmful, but not highly hazardous, substances.
The attractiveness of this strategy lies in its relative simplicity; of
all the strategies, regulatory costs would probably be least for this one.
The strategy is straightforward in dealing directly with the uncertainty that
pervades the RQ issue. It provides for low RQs (one pound, for example) for
the most hazardous substances and partially relieves the reporting burden on
releasers for all other substances. This strategy would implicitly recognize
that the major effects of the Sections 102 and 103 regulations arise from the
designation process and that assignment of RQs plays more of a safeguarding
role for releases of the most highly hazardous substances. But the usefulness
of this strategy can probably only be realized if only two RQ levels are used,
despite the temptation to establish additional RQs to account for the relative
hazard of different substances.
4.5 STRATEGIES FOR NOTIFICATION
Notification requirements for continuous releases are contained in CERCLA
Section 103(f). There are two important elements in this section. First,
CERCLA provides an exemption from notification under Section 103(a) for
continuous releases. Second, CERCLA conditions this exemption on two
requirements: (a) that the continuity of the release is sufficiently well
established, and (b) that annual notification occurs, or notification occurs
in the event of a statistically significant increase in the release.
CERCLA does not define the rationale for continuous release notification.
We assume that the rationales for such notification are to give EPA a better
information base, to allow EPA to take remedial action or require releasers to
reduce or eliminate the releases, and, possibly, to bring continuous releases
of hazardous substances under the purview of existing permit systems. The
rationale for non-immediate reporting of continuous releases is self-evident:
immediate reporting would excessively burden the releaser, the National
Response Center, and the On-Scene Coordinator with huge volumes of reports of
usually non-emergency releases. Furthermore, it is not clear how releasers
would "immediately" report continuous releases.
ICF is currently assisting EPA in a larger effort of developing and
assessing strategies for notification of continuous releases. That effort
will elaborate on the material discussed here, as well as the economic effects
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
4-20
of notification strategies, in a report due to be submitted in 1982. The
report will also incorporate the contributions of EPA's program offices as
well as outside groups interested in the regulation, and will contribute to
the deliberations of the EPA work group established for this purpose.
4.6 NON-REGULATORY APPROACHES
As discussed earlier, an examination of "alternative approaches" to a
proposed regulation is required by Executive Order 12291. In addition to
alternative criteria for designating substances, varying levels of reportable
quantities, and variations of notification requirements, 0MB guidance on E.O.
12291 further requires the consideration of:
(1) the consequences of having no regulation, and
(2) major alternatives beyond the scope of the specific
statutory provisions under consideration (i.e., CERCLA
Sections 102 and 103).
The "no regulation" option is the "base case," defined by the incentives
of the marketplace and existing statutory and regulatory requirements.
Section 102 regulation is not required for companies to be subject to (1)
liability for natural resource damages and costs of responding to releases of
hazardous substances and (2) penalties for failure to report such releases.
In the absence of designation by regulation, the definition of hazardous
substances contained in Section 101(14) of CERCLA would be controlling. This
definition incorporates, lists of chemical substances developed under other
laws that may not be entirely appropriate for CERCLA -- some lists may be
over-inclusive, others may be under-inclusive. Benefits and costs of Sections
102 and 103 regulations must be measured in light of this statutory baseline.
Other options beyond the scope of CERCLA Sections 102 and 103 include both
regulatory and non-regulatory alternatives. For purposes of this analysis, a
regulatory option is one involving some direct "command and control" over
business activities and operations; a "non-regulatory" option is one that
affects industry activities indirectly by altering incentives and economic
consequences of actions. For example:
" A regulatory alternative might be the required
purchase of liability insurance by certain classes of
businesses dealing in (or with) hazardous substances.
* A non-regulatory alternative might be a required
waiver of certain legal defenses to damage claims, or
financial penalties for releases, or rewards for good
performance.
In most cases, either approach will require some official action or
rule-making. This means that even "non-regulatory" alternatives will usually
need to be implemented through the issuance of regulations.
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
4-21
According to OMB, a regulatory impact analysis must show that, of all the
alternative approaches to the given regulatory objective(s), the proposed
action will maximize the benefits to society.
4.6.1 Possible Alternatives
To satisfy Executive Order 12291, alternatives beyond the scope of
Sections 102 and 103 need to be considered. Some market-oriented solutions
are shown in Exhibit 4-3.
Analogies and precedents can be found for these alternatives in other
programs. For example, special tax rules apply to pollution control
equipment. Financial responsibility is required under the Clean Water Act
(Section 311), the Motor Carrier Safety Act (Section 30), and the Resource
Conservation and Recovery Act (Section 3004(6)) for carriers of hazardous
substances and wastes. (See also CERCLA Section 108.) The Nuclear Regulatory
Commission requires licensees to waive certain common law defenses to third
party claims; this requirement establishes strict liability. As directed by
OMB, such alternatives would, in many cases, be beyond the scope of the
Sections 102 and 103 authority under CERCLA; however, some, such as financial
responsibility assurances, can be required by other provisions of CERCLA
(Section 108).
For each of the goals of the Sections 102 and 103 regulations, EPA would
need to evaluate an array of reasonable alternatives. Consider, as a simple
example, alternatives to reporting requirements as a means of developing an
adequate data base regarding the frequency and characteristics of releases:
* Contractors (private or state and local governments)
could be engaged to monitor sites and transport routes
and report on releases;
* A voluntary statistically-based "sampling" program
could be developed similar in many respects to the
system developed by the Bureau of Labor Statistics to
estimate rates of workplace injuries and fatalities;
* Rewards or bounties could be offered for reports,
analogous to the system developed by the National
Weather Service for collecting meteorological data and
the U.S. Coast Guard rewards system, where one-half of
the penalty assessed against a releaser may be paid to
a third-party reporter of the release.
Each alternative will have different costs and may encourage over- or
under-reporting. Outcomes will need to be analyzed carefully to permit
comparison with the regulatory options developed under CERCLA.
For the goal of providing incentives for better handling and care of
hazardous substances and waste facilities, another array of alternatives will
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
4-22
EXHIBIT 4-3
MARKET-ORIENTED APPROACHES TO SECTIONS 102 AND 103 REGULATION
Approach
Positive economic incentives
Examples
Tax breaks for added investment in
containment and release prevention
hardware
Rewards for exemplary operating
experience
Payment for reports of incidents
submitted to EPA.
Negative economic incentives Automatic financial penalties for
releases, in addition to liability for
response costs
Operating permit fees that escalate
based on the release experience of the
facility
Required disclosures of liability and
experience under Securities and
Exchange Commission (SEC) laws or other
information disclosure strategies
Financial responsibility requirements
(e.g., insurance, bonds) as a condition
of owning or operating a facility
handling hazardous substances
Establishment of strict liability
for third party claims arising
from spills or releases.
Fostering of voluntary response Existing programs operated by
and clean-up programs individual companies and industry
groups.
Implementation of a quasi-market Operating rights or marketable permits,
system
November 2 Draft
* * Do not quote or cite
ICF INCORPORATED
-------
4-23
need to be evaluated. In this case, the "quasi-market approach" may prove to
be an innovative alternative. Most fully developed as a regulatory reform for
air pollution control, this alternative involves establishing "rights" or
"quotas" for handling hazardous substances. .A facility that had a significant
release would be penalized by losing some or all of its rights (e.g., disposal
rights). Alternatively, "good actors" could be permitted to purchase more
"rights" or could be rewarded by the government with more permits.
• For example, transporters of hazardous substances or
wastes could be allocated a certain quota (in terms of
tons per year) that would be adjusted periodically (up
or down) based on that transporter's "track record."
This alternative would provide a strong financial incentive for better
handling of hazardous substances even in the absence of liability for response
costs.
In performing this aspect of the analysis, it would be necessary to work
closely with EPA to:
(1) define the specific objectives of the proposed
regulations in a way that facilitates the analysis of
alternatives;
(2) develop a list of candidate non-regulatory
alternatives; and
(3) perform a preliminary assessment of the relevance of
non-regulatory alternatives to the analysis.
4.7 CONCLUSION
This chapter has presented a framework for considering strategies and
regulatory alternatives for the major elements of the Sections 102 and 103
regulations. In addition, we have reviewed non-regulatory approaches for
achieving the same objectives. In the next chapter, we identify the economic
effects that the Sections 102 and 103 regulations are expected to have.
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
CHAPTER 5
ECONOMIC EFFECTS
5.1 INTRODUCTION
This chapter discusses the nature of the economic effects resulting from
the Sections 102 and 103 regulations. It contains a step-by-step analysis of
the kinds of economic effects which could result from the designation,
reportable quantity, and notification regulations. Chapter 6 contains the
initial attempts to measure these effects.
For the most part, the effects described here represent the effects of
regulation as compared to the baseline. The emphasis in this chapter is on
establishing the linkage between regulations, resulting actions, and the
expected costs and benefits.
In discussing benefits and costs, it is useful to distinguish among the
levels of causality between the regulation and the effect. Here, two such
levels are used: direct and indirect effects. Assigning a particular effect
to either of these categories is largely a matter of judgment, and there is no
one "correct" category for each effect. Instead, the use of this typology is
a convenient.way of denoting the uncertainty of the two level of effects.
That is to say, we are less certain that indirect effects will occur than
direct effects. Furthermore, indirect effects are more "removed" from the
regulatory source than are direct effects; we may not be able to attribute
indirect effects entirely to the regulations under investigation.
Many potential releasers, for example, will take preventive action to
reduce the number and frequency of certain releases whether or not EPA
designates the substance. Other potential releasers, however, will be
encouraged by these regulations to take preventive action. Since EPA cannot
under CERCLA require preventive actions, any that are taken as a result of
the Sections 102 and 103 regulations are probably "indirect" effects of the
regulation. Whether or not they can be attributed to the regulations, such
actions cost money and must be estimated.
Another categorization that will be used both in this and the next chapter
is the differentiation between real resource effects and distributional
effects, as required by the 0MB and EPA guidelines. Real resource effects are
those that change the total resources available to the economy; they represent
enlarged or foregone opportunities. Distributional effects represent
transfers of available resources from one economic entity to another.
5.2 GUIDANCE ON IDENTIFYING EFFECTS
OMB's draft regulatory guidelines state that a fundamental test of a
satisfactory RIA is whether it enables reviewers to make informed judgments
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
5-2
about whether the RIA satisfies the E.O. 12291's objectives. For a reviewer
to understand why the net benefits of a proposed regulatory option are greater
than the alternative options, the reviewer must first be able to understand
and trace the effects of each option. This will also facilitate EPA's
identification of all of the impacts of the regulatory alternatives and the
subsequent measurement of benefits and costs.
An analysis of benefits and costs meeting OMB and EPA requirements would
have the major components listed in Exhibit 5-1.
5.3 EFFECTS OF DESIGNATION REGULATIONS
The designation of a substance as hazardous is the starting point of the
hazardous substance response program. For a whole variety of actions,
designation is the essential first step that brings a chemical substance under
the purview of the regulations.
Reporting and Response. A direct effect of designation is the required
reporting of releases at or above the reportable quantity level. All of the
potential benefits and costs resulting from the CERCLA reporting requirements
are applicable to a designation regulation. The act of reporting a release
presumes that the release will be cleaned up, if necessary. Therefore, the
source of major benefits and costs of designation is that some releases may
be cleaned up which would not have been cleaned up without the regulations or
which would have been responded to incompletely or in an untimely fashion.
These benefits and costs of designation depend crucially on how designation
affects the behavior of parties responsible for the release of a substance.
If responsible parties diligently clean up their releases of any substance --
whether designated or not -- then the designation of substances would not have
much impact. On the other hand, if releases of non-designated substances are
completely ignored by responsible parties, the benefits and costs of
designation might be very great. (The reality is probably somewhere in
between the two extremes, but more research is needed before we can reach a
conclusion on this key point.)
Liability. The largest effect of designating a substance as hazardous,
however, may arise from the fact that releasers of the substance become
subject to the liability and financial responsibility provisions of CERCLA.
For example, a releaser of a designated hazardous substance may be liable for
$50 million in natural resources damages to compensate for a release. (See
Appendix 6 for a full discussion.) This increased liability would likely
translate into increased costs for potential releasers of designated
substances because most releasers would augment their insurance coverage to
cover the added liability. Any increased costs of insurance would represent a
distributional transfer,1 but could result in real, macroeconomic effects if
the resulting increased costs of handling hazardous substances is large enough.
1Most of the cost of insurance is returned to insured parties via
payments for claims. This portion of the premium is not a real resource cost,
but a transfer among insured parties. Only the cost incurred in administering
the insurance company is a real resource cost.
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
5-3
EXHIBIT 5-1
MAJOR COMPONENTS OF A BENEFITS AND COSTS ANALYSIS
Section
1. Definition of benefits and costs
2. Identification of benefits and costs
3. Measurement of effects
4. Quantification and description
of nonmonetizable effects
Contents
Explain that benefits and costs of
regulatory alternatives are those
gains accruing to and costs
imposed on society as a result of
implementation of regulations.
Identify actions that yield
benefits and costs and describe
direct and indirect effects of
regulatory alternatives that lead
to net social gains or losses.
The RIA must monetize effects
wherever possible. For example,
where possible, EPA should attempt
to assign values to any
improvements in human health or
reductions in non-health damages
that result from the
implementation of regulatory
provisions. The RIA should
include a statement of all
assumptions used and an
explanation of all data sources.
Discuss any effects that cannot be
monetized. If possible, they
should be quantified to some
extent — such as the amount of
new information expected from the
notification requirements. Where
it is not possible to quantify
these effects, the RIA should
present case study or anecdotal
evidence.
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
5-4
EXHIBIT 5-1
MAJOR COMPONENTS OF A BENEFITS AND COST ANALYSIS
(continued)
Section
5. Identification and analysis of
small entity effects
6. Distributional effects
7. Macroeconomic effects
Contents
Describe the alternatives' effects
on firms of different sizes or on
industries. This effort may
require the use of firm,
industrial, or total economy
models that can be used to predict
the distributional effects of
compliance costs.
Specify those benefits and costs
accruing to particular parties,
even if those effects are
counterbalanced by impacts on
Other parties. The distributional
analysis should include the
benefits accruing to certain
businesses, to labor, and to
communities. The RIA should also
analyze tax effects, insurance
premium effects, and other kinds
of transfer payments.
Building up from analysis of
effects on individual sectors of
the economy, estimate effects of
regulation on aggregate output,
employment, and the price level.
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
5-5
EXHIBIT 5-1
MAJOR COMPONENTS OF A BENEFITS AND COSTS ANALYSIS
(continued)
Section
Contents
8. Sensitivity analysis
Focus on the areas of the benefits
analysis that are the most
critical and the most uncertain.
For each regulatory alternative,
modify any key assumptions that
may influence outcomes, and
reconstruct the analysis. Employ
discount rates other than the OMB
mandated 10 percent discount rate,
to reflect the opportunity costs
associated with the financing of
public investments, the
differences in investment
riskiness, and the nature of the
benefits and costs. Project
macroeconomic affects under the
different assumptions specified.
9. Schedule of benefits and costs
Include in the RIA a schedule of
all benefits and costs for each
regulatory alternative that
details the nature of the benefit
or cost, to whom it accrues, and
when it accrues. State the
monetizable benefits and costs for
each regulatory alternative in
constant annualized dollars.
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
5-6
On the benefits side, the application of strict liability and federal
jurisdiction, and the provisions for treble damages for failure to respond
when ordered to do so under Section 106 provide an added measure of protection
to society from damages due to a release. Protection to individuals affected
by releases might arise as a result of a spillover effect in state courts from
the stringency of CERCLA's provisions (and, for example, from CERCLA's
legislative history, which refers to designated substances as
"ultra-hazardous").
In addition, because of potential liability under CERCLA, releasers may
be encouraged to improve handling and management of newly-designated
substances and to clean up more of these releases to avoid liability. These
actions could result in indirect benefits to society through a reduction in
the number or severity of releases and a reduction in exposure to releases.
Indirect costs to releasers could result from instituting preventive
measures. However, in many cases, newly-designated substances may already be
subject to DOT, Coast Guard, or OSHA regulations, so that industry practices
might not change substantially under CERCLA regulations.
A different reaction potential releasers might have to the liability
provisions is to seek to incorporate releases into the existing permit
systems, since federally-permitted releases are excluded from reporting and
liability provisions. This approach, however, can have only limited
usefulness; first because permit standards for many hazardous substances have
not been developed, and second, because many of these releases (accidental or
transportation-related) are not applicable to permit systems.
Exhibit 5-2 contains a display of the benefits and costs of designation
regulations as compared to the baseline of no designation. Each potential
effect is related, in a probabilistic manner, to an action or event resulting
from regulation. In each case, we also identify the parties who would
experience the effects and whether the effects represent primarily real
resource effects or distributional effects. In general, most of the
potential costs of designation would be incurred in the short time period
following designation — probably the first year or so. Releasers would
presumably act within a short time to increase their insurance, or to install
preventive practices.. Upon promulgation of regulations, releasers would
immediately be responsible for reporting costs and applicable cleanup costs
for releases of the designated substances.
The benefits of designation, on the other hand, have a more uncertain
time frame. Benefits accruing from more certain and more complete
compensation for damages and cleanup costs will be felt in the short term.
The benefits of avoiding economic losses and acute adverse health effects
would also accrue immediately if there were a reduction in the number or
severity of releases. Long-term benefits might result from a reduction in the
risk of exposure to substances with chronic adverse health effects.
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
5-7
EXHIBIT 5-2
EFFECTS OF DESIGNATION
Regulation
Actions
Costs
Benefits
Designation of
hazardous
substances and
assignment of RQs.
1. Required re-
porting of
releases at or
above RQ.
1. Direct:
Industry: Cost
of making the report
(real). Higher
probability of
assuming cost of
cleanup (real).
Government: Cost
of receiving and
responding to
report (real). .
1. Direct:
Industry: Higher
probability of
saving released
materials (real).
Potential for
saving on costs
of cleanup and
liability (real).
Society: Reduc-
tion in exposure
to hazardous sub-
stances (real).
2. Application of
limited liability
and financial
respons ibi1ity
provisions.
2. Direct:
Industry: Cost of
obtaining or in-
creasing insurance
2. Direct:
Society: In-
creased and more
certain compensa-
or higher probability tion for damages
of paying damages (distributional),
(real and
distributional).
3. Possible
initiation of
preventive
practices.
3. Indirect:
Industry: Costs of
initiating and main-
taining preventive
practices (real).
3. Indirect:
Society: Reduced
risk of exposure
to releases
through fewer
and/or less
severe releases
(real).
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
5-8
5.4 EFFECTS OF REVISING REPORTABLE QUANTITIES
Regulations revising reportable quantities will have costs and benefits
depending on the direction of the changes (up or down) in reportable
quantities. If the reportable quantity (RQ) for a substance is lowered, for
example, the regulations would probably result in added costs to industry and
added benefits to public health and the environment -- as explained in detail
in the next section. However, the regulations establishing new reportable
quantities for the Section 104(14) substances will likely be directed at
raising some of the statutorily-established one-pound RQ levels. On the
other hand, EPA may raise some and lower other RQs for the Section 311
substances. In discussing benefits and costs, therefore, it is necessary to
keep in mind the direction of the change in reportable quantity levels.
Another key point is that all of the costs and benefits enumerated below
relate only to the incremental change in reported releases. That is, if RQs
are lowered, benefits and costs are associated only with the additional
releases that would become subject to the reporting requirement at the new
lower leyel, and not all of the releases reported at that level.
It may be useful to view the results of a change in RQs by means of a
decision tree. Exhibit 5-3 shows such a decision tree and the associated
costs and benefits for each course of events. Note that the exhibit
represents the effects of using or not using any particular RQ for any given
substance. It does not show what the aggregate effects are of a change in
many different RQs nor does it indicate which of many possible RQs may be
appropriate for a given substance.
The usefulness of the decision tree lies in its portrayal of the complex
chains of events resulting from a release. All of the decisions and events
shown on the tree have associated probabilities.2 In the following
discussion, we trace through, in narrative form, these actions and effects.
Reporting and Response. First, it must be noted that whether or not
reporting is required, releasers may or may not report a release. The value
of the reporting requirement is in encouraging the act of reporting and the
benefits that flow from that act. There will, of course, be releasers that
will not report their releases, either because they remain unaware of their
2These probabilities will change both as a result of requiring reporting
at a given RQ level and as a result of the particular level of the RQ. For
example, consider two cases, one in which the decision is whether to change an
RQ of 10 pounds, the other concerning an RQ of 1000 pounds. If no reporting
is required, the releaser may have a higher probability of reporting in the
second case than in the first case, either because the releaser feels the
release is more serious and help is needed or because he feels that it will be
discovered anyway. For similar reasons, a release of 1000 pounds may be more
likely to be cleaned up, even if unreported, than a release of 10 pounds.
November 2 Draft * * * Do not quote or cite
I CP INCORPORATED
-------
5-9
EXHIBIT 5-3
EFFECTS OF REVISING REPORTABLE QUANTITIES
Release of Designated
Substance in Specified*
Amount
Reporting^
Required
Reports
•No eleanup needed
eleaser cleans up
k Releaser doesa
not report
» Government cleans up
•Releaser cleans up
¦"•Re lease
discovered,
traced
.Releaser does ,
not cleanup
_ No discovery
or tracing
No
„ Reporting.
Required
•Releaser*
Reports
"No cleanup needed
¦ Releaser cleans up
Releaser does
""not report
¦ Government cleans 19
¦Releaser cleans 19
Release
discovered,
traced
, Releaser does .
not cleanup
Release not
^discovered,
traced
Cost of reporting
Liability for damages
may be increased
Cost of reporting
Cost of cleanup
Liability for damages
may be increased
Cost of reporting
Releaser liability
for cost of cleanup
and damages may be
increased
Costs of cleanup
Liability for damages
Potential penalty
Penalty for failure
to report
Liability for cleanup
costs and damages
(possibly unlimited)
Oosts to society if
eleanup delayed
Costs to Fund and
releaser if cleanup
more costly
•Oost. to society if
release has adverse
effects
Cost of reporting
Liability for damages
may be increased
Cost of reporting
Cost of cleanup
Liability for damages
may be Increased
Cost of reporting
Releaser liability
for cost of cleanup
and damages may be
increased
Costs of cleanup
Liability for damages
Releaser liability for
cleanup and damages
Oosts to society Fund
and releaser if cleanup
delayed and more
costly
Cost to society if
release has adverse
effects
More effective and
timely response
Continued monitoring
possible
More effective and
timely response;
Continued monitoring
possible
More effective and
timely response
Continued monitoring
possible
More effective and
timely response
Continued monitoring
possible
November 2 Draft
Do not quote or cite
ICF INCORPORATED
-------
5-10
obligations, because they do not accurately estimate the true quantity
released, because they have intentionally released the substance (e.g.,
"midnight dumpers"), or because they do not want to incur the costs and
possible liabilities attendant to reporting, (even though liabilities if the
release is not reported may be greater).
On the other hand, the fact that the law exists and will be enforced means
that most releasers will probably attempt to comply with it. The penalties
for failure to report releases also act as incentives to report. Finally,
releasers may wish to avoid liability for cleanup costs that may be greater
than the cost of cleaning up in the first place. This could result from the
government discovering the release and removing it at greater cost than the
releaser would have spent, or because a delay in response results in far
greater damage than if the release had been contained and cleaned up initially.
In general, reporting of releases is a necessary element in the design of
the CERCLA system for controlling hazardous substance releases. Discovery of
a release may be made independently of releaser reporting; nevertheless,
responsibility for reporting must be assigned to the releaser. It is
important to note that releasers currently do report some releases below the
reportable quantity. (See Chapter 6 for data.) In fact, some releases of
non-designated substances are reported.
Even if the release is not reported, releasers may still clean up in order
to protect their reputations, limit their liability, and salvage released
materials. Of recent releases into water,, for example, EPA staff estimate
that only 10 percent are "mystery spills" or other situations that have-
required government removal -- thus EPA assumes that 90% of all common spills
are cleaned up by the responsible party. (The same high percentage, of
course, may not apply to long-term low-level releases from hazardous waste
sites.)
If releases are reported, it must be assumed that the releasers understand
(or will soon be made to understand) their responsibility to clean up if
cleanup is needed, or, if the releasers are not capable of taking response
action, to pay for government cleanup. This is the primary potential source
of benefits and costs from the reportable quantity regulations. That is,
depending on the direction of the change in reportable quantities, there will
be some change in the number of releases cleaned up. If RQs are lowered, more
releases will likely be cleaned up, resulting in benefits to society of
reduced risk of exposure to hazardous substances and costs to releasers of
cleaning up releases they might not have otherwise responded to. (These costs
may be offset by benefits to firms offering cleanup services in the form of
increased demand for their services. Also, to the extent that releasers
finance more cleanups, Superfund monies are freed up to remedy other releases
and States need to spend less of their resources on responding to releases.)
The magnitude of this effect depends on industry and government behavior
in the baseline. If releases of hazardous substances below the reportable
quantity are responded to as diligently as releases above the reportable
quantity, there may be no appreciably different effects from revising
reportable quantities.
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
5-11
Whether or not response is affected, changes in RQs and reporting habits
will have another major impact on releasers and victims of releases. To the
extent that releasers report their releases (whether above or below RQs), it
will probably be easier to prove the releaser liable for damages and cleanup
costs. Indirectly, therefore, a change in RQs affects, through the reporting
requirement and probability of discovery of the release, increase the
likelihood of assigning liability to the releaser.
Effects of Federal Involvement. Other benefits and costs may result from
the involvement of the federal government in the response process. For
example:
(1) Immediate reporting of releases can reduce the delay in response
action. A releaser may not realize that response is actually necessary or
that quick action is important. Faster response can have two benefits: (1)
it may decrease exposure to hazardous substances, and (2) it may reduce the
ultimate costs of cleanup and potential liability for damages by containing
the spread of hazardous substances in the environment.
(2) Reported releases permit the government to contribute additional
expertise and experience. This is not to say that a federal On-Scene
Coordinator always makes the correct decisions or that state officials may not
be equally capable. Nevertheless, federal involvement contributes, in most
cases, a wide range of experience with hazardous substances and cleanup
methods and a wide variety of resources from support agencies (such as the
Food and Drug Administration, Center for Diseasie Control, Army explosives
teams, etc.). In general, this expertise should have beneficial effects in
limiting the damage of a release.
The effects on releasers' costs of cleanup are uncertain: government
officials may require a higher degree of cleanup, a concern for environmental
damage, and greater precautions (e.g., in evacuating area residents or
providing alternate water supplies) than the releaser might have provided. On
the other hand, the releaser might reduce his ultimate liability for damages
and might reduce cleanup costs by using cheaper cleanup methods suggested by
the government.
(3) Reporting of releases allows for continued government monitoring and
assessments of the effects of a release, whether or not immediate removal is
necessary, and provides for continued protection of public health and the
environment.
Finally, there are three additional areas of benefits and costs which do
not show up in the decision tree but that might result from changes in RQs.
Preventive Actions. First, a lowering of RQs may result in releasers
initiating preventive actions, to avoid the costs associated with releases.
Further, since most of the material released in a typical incident is not
November 2 Draft * * * Do not quote or cite
IGF INCORPORATED
-------
5-12
recovered or reused by the releaser, actions which reduce the number of
releases save valuable resources from being wasted.3
To some extent, potential releasers may also take protective actions (such
as obtaining permits or increasing insurance) to mitigate the effects of a
release, but these are more likely to result from designation regulations,
rather than regulations that revise reportable quantities.
Useful Information from Reports. Second, reporting of releases has
informational value. Depending on how EPA uses this information, data on
releases may prove useful in designing programs to contain releases
effectively, in alerting releasers to ways of preventing releases, and in
educating people to mitigate the effects of releases by exercising cautionary
measures. Indirectly, benefits could result from fewer or smaller releases,
with reduced exposure to people and the environment. The direct costs to
industry of producing the information consists of the nominal costs of a
toll-free telephone call to the National Response Center.
It is important to recognize that just because the RQ for a substance is
lowered, that does not mean that all of the new reportable releases must
receive response by the government or by the releaser or that they are all
harmful. Rather, the RQ functions as an alarm bell; a lowering of RQ levels
increases the information available to government, permitting EPA to choose to
respond to smaller releases if the particular conditions of the release appear
to warrant response.
On the other hand, there are costs involved in increasing information as
well. Aside from the costs of processing many reports, there is also the
possibility that low RQs could lead EPA more often to take response action and
spend valuable resources on releases of lower priority. Higher RQs, to some
extent, would focus EPA's attention on releases where response may be more
often needed.
Compliance with RQs. Third, effects may arise from the degree to which a
set of RQs considered as a whole facilitates compliance. For example, to the
extent that RQs are related more closely to the sizes of typical containers,
or to the extent that substances commonly used or transported together have
similar RQs, the ease of complying with the regulations will benefit releasers
and ultimately the public. Similarly, a restricted set of reportable quantity
levels, a single RQ for all media, and reportable quantities that are far
enough apart in value so as to be easily distinguishable, would all help to
increase compliance with the regulations. Higher RQs would also relieve the
burden imposed on many releasers in the transportation industry who experience
real difficulties in measuring the amount of substances released (e.g., in
3EPA staff members, for example, have been informed that the
establishment of Section 311 reporting requirements resulted in positive
savings by some industrial plants because the required monitoring of releases
alerted plant managers early to spills of valuable materials.
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
5-13
chronic, low-level leaks of pressurized gases). Similar measurement problems
caused by mixtures of substances with very different RQs would also be
alleviated. In general, the benefits of reportable quantity regulations are
increased when the regulations are designed to facilitate reporting.
It is useful to note that both designation regulations and revisions of
reportable quantities may have similar types of effects in terms of
information, prevention, and clean up activities and the assumption of
liability for costs and damages. It cannot be determined, at this point,
whether designation or RQ regulations will have the greater impact. However,
because designation imposes certain costs (of liability) in the event of a
damaging release, while the reportable quantity mostly affects the likelihood
that liability will be assigned to the releaser, we suspect that designation
regulations will have the larger impacts.
5.5 EFFECTS OF NOTIFICATION REGULATIONS
The most obvious effect of a notification requirement for continuous
releases (either annual reporting or some other mechanism) is that it produces
the specific information desired by EPA. The costs to industry of generating
the information include monitoring of releases, testing to determine
concentrations, record-keeping, and preparation of any required forms. Firms
may need to buy or lease monitoring equipment and train or hire additional
employees. The magnitude of these costs depends on the exact type of
notification required, the number of firms involved, and the degree to which
the requirements correspond to current practices within firms.
For the government, notification regulations can impose a large burden of
processing and reviewing many reports, and increased inspection and
enforcement efforts.
Although the costs of a notification program can be more or less easily
tallied, the benefits of notification depend on what the government does with
the information received. It must be assumed that no specific notification
regulations would be promulgated unless a corresponding program to use the
information were planned; without such a program, there would be no benefits
from notification regulations. Although notification itself provides
essential information in determining whether continuous re-leases are hazardous
~to health and the environment, EPA should be able to predict roughly the need
for a control program for these releases.
The primary benefits of notification would, therefore, be: (a) a more
accurate understanding of the magnitude of the continuous release problem, and
(b) a control or prevention program to reduce the effects of hazardous
substances in the environment. The program could target small reductions in
releases for many industries or large reductions in a few releases; it could
concentrate on prevention, reductions, pretreatment, or cleanup; it could
emphasize incorporation into permits, voluntary compliance, inspections and
audits, or enforcement. The resulting benefits, in any case, should be viewed
in terms of reduced risk of exposure to hazardous substances.
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
5-14
It is worth noting, too, that the act of notification increases the
likelihood of assigning liability to the releaser for cleanup costs and
damages to natural resources. This produces an incentive not to report
continuous releases or to misrepresent information. Any notification
strategy, therefore, must contain provisions to counter this effect.
It may be the case that releasers are not entirely aware of their
continuous releases of hazardous substances. Notification requirements could
induce releasers, either voluntarily or as a result of public pressure or the
threat of enforcement, to reduce or eliminate their continuous releases.
5.6 CONCLUSION
This chapter has discussed the actions and effects resulting from
designation, reportable quantity, and notification regulations. In the next
chapter, we examine whether and how these effects can be measured and present
the data and findings accumulated to date.
November 2 Draft * * *
Do not quote or cite
ICF INCORPORATED
-------
CHAPTER 6
MEASUREMENT OF ECONOMIC EFFECTS
6.1 INTRODUCTION
This chapter builds on Chapter 5's discussion of the economic effects of
the Sections 102 and 103 regulations. The purpose of this chapter is to
suggest methodologies for the estimation, in qualitative and quantitative
terms, of benefits and costs. Section 6.2 provides a description of the basic
approach to measurement issues, the selection of important benefits and costs
which can be estimated, and a description of data sources and data avail-
ability. Section 6.3 presents the data and findings from research to date and
discusses the next steps in performing the analysis. Section 6.4 contains a
summary of the major conclusions drawn from the data.
Because this study has not involved a major research effort, and because
no regulatory alternatives have been selected, this chapter does not present
estimates of benefits and costs. Instead, it is intended as a compilation of
analysis and research performed thus far and as a starting-point if RIAs are
prepared for designation, reportable quantity, and notification regulations.
The ultimate use of the information discussed in this chapter is to
estimate the net social benefits of each regulatory alternative evaluated in a
regulatory impact analysis. To see how this information would be used,
suppose that one impact of a regulatory alternative is to induce firms to take
preventive action which reduces the number and severity of releases. This in
turn, reduces the number of response actions which must be taken, and benefits
society by reducing the number of releases that occur and the number that go
unreported. The information discussed in this chapter would be used to
calculate the costs and benefits of that regulatory alternative. Exhibit 6-1
displays the elements in that calculation, along with the sections of this
chapter in which each element is discussed.
6.2 APPROACH TO MEASUREMENT
This section presents our approach to the measurement of benefits and
costs of the Sections 102 and 103 regulations. As can be deduced from the
discussion in Chapter 5, each of the Sections 102 and 103 regulations can
result in a variety of benefits and costs, some of which are highly
speculative. For many of the benefits and costs, we expect to be able only to
identify the direction of the effect, and, based on anecdotal evidence, to
indicate whether the effect is likely to be realized or is merely speculative.
This discussion, therefore, concentrates on a few benefits and costs which
are expected to account for the bulk of the impacts of regulation and which
may be most readily estimated.
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
6-2
EXHIBIT 6-1
THE CALCULATION OF THE COSTS AND
BENEFITS OF REGULATORY ALTERNATIVES
Sections 6.2.1,
6.3.1
November 2 Draft
Do not quote or cite
ICF INCORPORATED
-------
6-3
The selection of these variables and the general approach used to measure
them are discussed in the next subsections, which also provide a brief
description of the available sources of data.
6.2.1 Methods of Measuring Benefits and Costs
This analysis focuses on a limited number of benefits and costs applicable
to designation and reportable quantity regulations. Because the costs and
benefits of the continuous release notification program depend so crucially on
the formulation of the program, we cannot make much progress on the
notification issue until the regulatory alternatives are better specified.
The only variable relevant to notification for continuous releases that will
be discussed here is the number of releasers potentially affected by the
notification regulations (see Section 6.2.2).
Benefits and costs arising from the designation and RQ regulations are of
essentially the same nature for both regulations,1 although some effects may
be more attributable to one regulation than the other. (For example, although
releasers may augment their insurance coverage because of lower RQs, they are
much more likely to do so as a result of a designation regulation.)
The benefits and costs concentrated on are the following:
* Benefits and costs of changes in the number of
releases cleaned up;
• Benefits of EPA involvement in responses to releases
(less delay, added expertise, etc.);
* Benefits and costs of preventive activities; and
• Costs resulting from application of liability
provisions.
1In the baseline, many hazardous substances are assigned reportable
quantities of one pound. Thus, the regulatory alternatives for the reportable
quantity regulations are likely to increase the reportable quantities for a
large number of substances. Therefore, both the costs and benefits are likely
to be negative rather than positive. This might cause some confusion in
keeping track of the costs and benefits, and in assessing the net benefits of
the regulatory alternatives. However, even if both costs and benefits are
negative for all regulatory alternatives, they are still measured as described
here, and the optimum policy is still determined by choosing the regulatory
alternative with the highest net benefit to society. For a positive net
benefit under these circumstances, the reduction in costs must exceed the
reduction in benefits.
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
6-4
The first step in analyzing these potential effects is to determine
exactly how they might arise. For example, how do insurance mechanisms handle
potential liability for releases of designated substances? Exactly what types
of preventive measures might be taken and with what effectiveness? (For the
most part, we will have to rely for this information on evidence from case
studies, interviews, and industry surveys.)
An important issue that arises here is the "level of estimation," that is,
the magnitude of the changes that the analysis will be able to uncover. This
issue can be seen more clearly by examining the extremes of a regulatory
alternative. For example, consider what would happen if only one substance
were designated as hazardous. Most likely, the effect of all the statutory
provisions on a single substance would induce firms to take out insurance,
reduce by a large amount their use of the substance, or perhaps reduce
entirely the use of this substance by substituting other substances for the
designated one. On the other hand, if over 1,100 substances are already
designated, the effect of designating one additional substance may be
undetectable in certain respects. Companies may have already instituted
applicable prevention measures and their insurance premiums may remain the
same.
In short, the marginal costs and benefits of a designation or RQ
regulation will depend in large part on existing regulations and practices.
At this point, we do not yet know the details of the regulatory alternatives,
for example, whether one substance will be designated or several hundred.
Even more important, it is not clear whether we will be able to estimate the
point at which marginal costs and benefits will become significant, (i.e.,
does the 1,200th designation produce a major impact or does the 1,500th
designation have a major impact?). This type of information, if it can be
collected, could be important in guiding decision-makers in designating the
optimal number (and type) of substances that will produce positive net
benefits without increasing costs.
The remainder of this subsection addresses the overall approach to
measurement of benefits and costs.
Benefits. The primary issue in measuring benefits is how to relate the
evidence provided by case studies and anecdotes to a standard valuation of
benefits. This is a familiar problem in public policy economics. This
chapter uses a readily understandable common denominator for valuing benefits
— the number of people whose risk of exposure to hazardous substances is
reduced. This expression admittedly begs the question of how harmful the
substances are, whether the risk of exposure relates to acute or chronic
adverse effects, and how long an exposure is averted. It is therefore
impossible to convert the number of people whose exposure is reduced into an
expression of the health benefits of the Sections 102 and 103 regulations,
although this approach could also eventually be expanded to incorporate
different types or lengths of exposure.
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
6-5
On the other hand, this approximation of benefits has the advantages of
being intuitively understandable and less controversial than a monetary
valuation of health benefits. Monetary valuation may be more appropriate for
non-health benefits (property damage avoided, environmental amenities) and
will be estimated to the extent permitted by available data.
Another measure of the benefit of more timely and/or effective cleanup is
the cost avoided if cleanup is delayed or performed inefficiently. For
example, if the timely response to a release in a feeder stream can prevent
the contamination of a river, the potential cost of the cleanup of the river
would be a benefit of the timely response to the release.
Costs. The discussion of cost measurement necessarily concentrates on
real resource costs to the regulated parties, which are usually the bulk of
the costs associated with a regulation (See Chapter 2). Costs are incurred by
regulated parties by the actions they take, which are grouped into three major
categories:
• Cleanup activities -- since the regulations are
intended to encourage responsible parties to cleanup
their own releases;
• Preventive activities -- since the regulations are
also intended to encourage the prevention of releases;
• Protection activities — since the regulations are
likely to cause the purchase of additional insurance.
The next section discusses methods of analyzing and data to be used in
measuring the cost of each of these categories of actions.
• Cleanup Activities. There are two fairly useful
sources of data on the cost of these activities. The
first is published reports on the cost of cleanup.
These include both reports which estimate average
cleanup costs for a wide range of substances and
reports which discuss the costs of cleanup of a
particular release of a hazardous substance. The
second source is the compilations of releases reported
to various government agencies (e.g., EPA Regions, the
NRC), some of which contain data on the cost of cleanup.
This information is not sufficient to estimate the
relationship between cleanup costs and the identity of
the substance. Cleanup costs are a function not only
of the identity of the substance, but also of the
volume released, the medium into which the release is
made, the idiosyncratic characteristics of each release
(such as weather conditions), and the interactions
among these variables. Because of the complexity of
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
6-6
the relationship between the characteristics of the
release and the cost of cleanup, we will be forced to
estimate the average cost of cleanup, or the cost of
cleanup for some representative substances, rather than
the complex relationship outlined above.
Preventive Activities. Preventive activities
include a whole range of activities, from capital
investments to changes in operating procedures,
designed to reduce the number and severity of releases.
We hope to obtain information on the cost of preventive
activities from two sources. One source is published
estimates of the costs of preventive actions taken by
various types of establishments. These include costs
of preventive activities which have been taken in the
past, and projections of preventive activities that
might be taken.under certain conditions. The second
source is interviews with officials of chemical
companies and other firms that may incur costs of
preventive action as a result of the Section 102 and
103 regulations. Ve hope these officials will supply
information on the cost actions they took before the
passage of CERCLA, how the passage of CERCLA has
changed their costs, and how the implementation of
regulations might further change their costs.
An additional source of information is the experience
of industries under the Department of Transportation's
hazardous materials regulations. We shall investigate
the extent of preventive measures taken in response to
those regulations to determine if the experience is
comparable.
Preventive activities may vary by the type of industry
and the volume and type of hazardous substances used.
Where possible, our cost estimates will accomodate
these differences. In other cases, it may be necessary
to develop average estimates, or estimates for some
representative industries and firms.
It is necessary to estimate the cost of preventive
activities not only because of their direct effects on
the costs of regulation, but because of their indirect
effects as well. Preventive actions increase costs
directly, but decrease costs in the long run because
they reduce the number and severity of releases, which
in turn reduce the costs of cleanup. The extent to
which cleanup costs are reduced depends on the amount
and effectiveness .of preventive actions taken, which in
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
6-7
turn depends on the costs of preventive actions. These
linkages cannot be measured precisely, but they can be
roughly estimated through interviews with officials
from chemical companies and other firms.
* Protection Activities. Protection activities are
undertaken by a regulated party to protect itself from
incurring costs in the event a release does occur. The
primary type of protective activity anticipated is the
purchase of insurance or increase in insurance coverage
to cover the costs of releases to the responsible
parties. It is important to note that the total social
cost of insurance is not the sum of all insurance
premiums. Instead, it is the administrative costs of
the insurance program, i.e., the sum of all insurance
premiums minus the expected payments. The
non-administrative costs simply represent transfers
across time.
6.2.2 Methods of Measuring the Number of Actions
To find the total real resource costs and benefits of a regulatory
alternative, it is necessary to multiply the costs and benefits of each action
by the number of such actions expected under the regulatory alternative (see
Exhibit 6-1). For example, the total cost of cleanup for each regulatory
alternative is the cost per cleanup, times the total number of releases
cleaned up because of the regulatory alternative, measured with respect to the
baseline. Correspondingly, total benefits would be equal to the average
benefits per release cleaned up because of the regulatory alternative, times
the number of releases cleaned up because of the regulatory alternative. The
number of releases cleaned up because of that regulatory alternative is in
turn a function of the number of releases, and the way in which regulated
parties and government officials respond to those releases under the
regulatory alternatives. As mentioned earlier, the number of releases will be
affected by the preventive actions taken in response to the Sections 102 and
103 regulations.
To estimate total real resource costs and benefits, in addition to the
cost and benefit information discussed in the previous section, it is
therefore necessary to measure three other variables:
* The total number of releases and their distribution;
* The total number of facilities covered by CERCLA;
* The response of regulated parties to the regulatory
alternatives.
Below we discuss methods of measuring each of these variables.
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
6-8
The Total Number of Releases and Their Size Distribution. The reason for
estimating the total number of releases is apparent from the discussion of
total costs above, but the reason for estimating the size distribution of
releases is less obvious. The regulatory alternatives for the reportable-
quantity regulations will alter the reportable quantities from the' baseline
quantities, thereby either including or excluding an incremental set of
releases. The number of releases in this incremental set depends not only on
the total number of releases, but on the distribution of those releases by
quantity released.
For example, assume that the size distribution of releases is as shown
below (which is, in fact, a fair representation of the EPA data analyzed in
this chapter). If the reportable quantity changed from A to B, the percentage
of releases between A and B can be determined mathematically. That percentage
can then be multiplied by the total number of releases to find the number of
releases affected by the regulation.
Size of release
Information on the size distribution of releases may be useful in
estimating the number of releases which are not currently reported. It is
very possible that not all releases are currently being reported. Because the
benefits and costs of the regulations are a function of the total number of
releases, not just those that are currently reported, it is necessary to
estimate the total number of releases based on reported releases. To estimate
the total number of releases and their size distribution from the available
data bases, three steps must be taken:
• Total reported releases must be obtained by
eliminating releases reported in more than one data
base.
• Total releases must be estimated from total reported
releases.
• The size distribution of total reported releases of
CERCLA substances must be estimated.
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
6-9
The first step, estimating total reported releases by eliminating
duplicates, is relatively straightforward, though time-consuming. The
releases in each data base must be compared to those in all other data bases
in order to eliminate doublecounting. To estimate total releases from total
reported releases, the number and distribution of releases from other EPA
regions can be compared to the number from Region VII, the region with the
most complete reports of releases. By assuming that other regions are really
like Region VII in terms of total releases, and that the only difference is
that reporting is more extensive in Region VII, the rate of reporting in other
regions can be extrapolated to be consistent with the rate of reporting in
Region VII. Of course, by this method any underreporting in Region VII cannot
be captured.
In order to be able to make the projections described above, it is
necessary to make some assumptions about the rate at which releases occur. If
we assume that the rate at which releases occur in a given geographical area
is roughly proportional to the volume of chemicals manufactured in that area,
information on chemical production in each state, available from the Census of
Manufacturers, can be used to extrapolate from Region VII to the rest of the
country. In this formulation, the ratio of releases to chemical production is
assumed to be the same in Region VII as in the rest of the country. If the
number of releases in Region VII, chemical production in Region VII, and
chemical production in the rest of the country are known, total releases in
the rest of the country can be calculated.
Finally, the size distribution of total releases can be estimated by
examining the size distribution in Region VII, and the size distribution in
the rest of the country, determining in which size categories the
underreporting occurs, and adjusting the distribution for the entire country
to fit the distribution in Region VII.
To estimate the number and distribution of releases that will occur under
a regulatory alternative, it is necessary to estimate the .number that occur
under the baseline, then determine how the regulatory alternatives will change
the number and distribution of the releases. If the baseline is in any way
different than the policy prevailing during the period for which data are
available, the number and distribution of releases under the baseline must be
projected from the number and distribution of releases for the period in which
data are available.
In estimating the total number and distribution of releases, we intend to
rely on the data on hazardous materials releases that have been gathered by
others. The data bases that we have used thus far include:
• The Pollution Incident Reporting System (PIRS), which
includes all releases of substances designated as
hazardous under Section 311 of the Clean Water Act,
which are reported to the National Response Center
(NRC), and which are referred from the NRC to the Coast
Guard for action. Only releases to water are covered.
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
6-10
• A compilation of all hazardous substance incidents
reported to the EPA Regional Offices from October, 1977
through September, 1979. These primarily include
substances designated as hazardous under Section 311 of
the CWA, although releases of other substances are also
included. Releases to land, air, and water are
included in this compilation.
• Releases reported to EPA Region VII since May, 1981.
Perhaps more than any other region, Region VII has
encouraged reporting of releases. In particular, PCB
releases have been extensively reported in Region VII.
In addition to these data bases, we intend to examine several other data
bases which we have not yet acquired. These include:
• The compilation of all releases reported to the NRC.
These data are not computerized, existing in the form
of hard copy reports. The data required for this
analysis must be extracted from the reports before the
analysis can begin.
• A computerized data base of detailed information on
seventy cases developed by Rockwell International under
contract to EPA.
• The Hazardous Materials Information. System (HMIS), a
compilation of transportation-related releases reported
to the Department of Transportation. Only releases of
substances designated as hazardous by DOT are included.
• A compilation of releases of six substances
transported through pipelines that are reported to the
Office of Pipeline Safety.
• Spill Prevention Control and Countermeasures (SPCC),
a compilation of releases that have occurred at NPDES
facilities. Most of the releases included are Section
311 hazardous substances, although some releases of
other substances are included. Releases included in
SPCC are supposed to be those releases that reach
water. Most of the releases are into water, although
some land releases are included.
Each of these data bases contains important information on releases.
However, each is incomplete and all must be used in order to estimate the
total number of releases of all substances designated as hazardous under
CERCLA, and to estimate the distribution of those releases.
November 2 Draft. * * * Do not quote or cite
ICF INCORPORATED
-------
6-11
The Total Number of Facilities Covered by CERCLA. All facilities
(including both fixed facilities and moving vehicles) that handle chemicals
designated as hazardous under CERCLA in quantities greater than the reportable
quantity are potentially subject to the Sections 102 and 103 regulations. We
have no way of directly determining the number of such facilities. However,
this number can be approximated from information on certain categories of
facilities which are likely to contain high percentages of facilities subject
to CERCLA. Some of these categories of facilities, along with the sources of
information about them, are listed below:
• Facilities which have applied for or have been
granted NPDES permits. Information on these facilities
is contained in EPA's Permit Compliance System (PCS).
• Facilities subject to federal and state emission
requirements as stated in State Implementation Plans
(SIP), New Source Performance Standards (NSPS), and
National Emission Standards for Hazardous Air
Pollutants (NESHAP). Information on these facilities
is contained in EPA's Compliance Data System.
• Facilities involved in the RCRA hazardous waste
program. Information on these facilities is contained
in the Hazardous Waste Data Management System.
• Hazardous waste disposal sites which are reported to
EPA under Section 103(c) of CERCLA.
• Establishments in industries which use hazardous
chemicals, such as the chemical industry. Information
on the number of facilities in particular industries
can be found in the Census of Manufacturers.
• Transporters subject to DOT regulations.
These different listings of facilities are likely to overlap
considerably. However, because of the large numbers of facilities in each
data base it would be too costly and time-consuming to check each data base
against the others for duplicates. It might be possible to check for
duplication within one state or region, and extrapolate the results to the
rest of the country. However, there may be an easier way to eliminate
duplicates. EPA is currently developing the Facility Index System (FINDS), a
collection of information for all those facilities which have an environmental
impact. FINDS is designed to integrate the information contained in many of
the above-mentioned data bases, insuring a consistent source of information
for EPA. FINDS is not currently operational. The software should be ready by
the end of October, but it will take some time for the regions to enter data
into the system. Until FINDS is operational, we will be forced to use
procedures such as those outlined above to determine the number of facilities
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
6-12
subject to the Sections 102 and 103 regulations. Note that information on the
total number of facilities is also important for estimating impacts from
continuous release notification regulations.
The Response of Regulated Parties to the Regulatory Alternatives. A
major factor in determining the costs and benefits of the regulatory
alternatives under Sections 102 and 103 is the response of the regulated
parties to the regulatory alternatives. All of the costs and benefits of the
regulations are generated because of actions taken by somebody, either
regulated parties or government officials. In the extreme, if regulated
parties did nothing differently as a result of the Sections 102 and 103
regulations, these regulations would have few costs and benefits.
In discussing the response of regulated parties to the regulatory
alternatives, it is useful to distinguish among three different categories
into which a release may fall:
. (1) The release of a "pollutant of contaminant" [defined
in Section 104(a)(2) of CERCLA], which has not been
designated as hazardous under CERCLA;
(2) The release of a substance designated as hazardous
under CERCLA in an amount below the reportable quantity;
(3) The release of a substance designated as hazardous
under CERCLA in an amount at or above the reportable
quantity.
The obligations of responsible parties for each of these types of releases are
discussed in detail in Appendix B. Here, suffice it to say that the
differences in the responses of regulated parties to these three types of
releases have an important bearing on the costs and benefits of the regulatory
alternatives. For example, if a responsible party would respond no
differently to a release of a given substance and quantity if it was of Type 2
than if it was of Type 3, adjusting reportable quantities up or down would
cause few costs and benefits. If a responsible party would respond no
differently to a Type 1 release than to a Type 3 release, designation of a
hazardous substance would cause few costs and benefits.
We look forward to obtaining information on the potential response of
regulated parties to the Section 102 and 103 regulations from the regulated
parties themselves, and other individuals and firms associated with them. By
and large, we expect only qualitative, rather than quantitative, information
and insights into the behavior of these parties in terms of cleanup actions,
preventive actions, and protection actions. We hope to gather information on
the extent to which regulated parties engaged in each of these activities
during the period before the passage of CERCLA and since the passage of
CERCLA. We also will ask regulated parties to project how their actions would
change under different regulatory alternatives. By examining trends in the
behavior of regulated parties since the passage of CERCLA, and by examining
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
6-13
the projections of future behavior made by the regulated parties, qualitative
projections of behavior under the regulatory alternatives can be made. Some
of the potential sources of this type of information are:
• Chemical firms: Chemical firms will be responsible
for a large number of releases, and will probably take
much of the preventive action that is taken in response
to the regulations. These firms can provide
information on how their cleanup and preventive
activities have changed with the passage of CERCLA, and
how they might respond to changes in reportable
quantities of substances that have already been
designated, to additions to the list of hazardous
substances, or to changes in the notification
requirements.
• Insurance firms. Firms that provide insurance
against the costs of hazardous substance releases would
also have information on trends since the passage of
CERCLA and what might be expected under different
regulatory changes.
• Hazardous waste removal firms. These firms are
hired by responsible parties and governments to clean
up releases of hazardous substances and may provide
information on recent trends in their business.
6.3 FINDINGS AND ANALYSIS
The previous section presented methods of measuring benefits, costs, and
the numbers of actions to be taken under Sections 102 and 103 regulatory
alternatives. This section presents the results of our investigations thus
far into the components of benefits, costs, and numbers of actions. For each
component, the information gathered as of the end of October 1981 is
summarized. We then discuss the implications of the data gathered for further
investigation, and present our plans for gathering further information.
6.3.1 Information on Benefits
6.3.1.1 Benefits of Federal Government Involvement
This subsection presents findings of several case studies
that illustrate the potential benefits of federal government involvement in
responding to releases of hazardous substances. One problem with using these
case studies is that it is rare to find documentation of hypothetical events.
Thus, if federal officials were involved in a successful cleanup of a release,
the case writer does not usually estimate what would have happened in the
absence of federal involvement. Or, if response to a release is immediate, it
is difficult to project the extent of damage if response had been delayed.
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
6-14
Furthermore, many case studies discuss the results of unreported releases
that were not cleaned up at all, or only after lengthy delays. It would not
be expected that reporting requirements alone would alleviate the situation in
cases where vandalism, illegal dumping, or unrecognized releases occur.
However, increased public awareness of the problem may reduce the delays in
reporting releases. The cases that follow illustrate the types of benefits
(in terms of costs and damages averted) that can result from timely federal
response to a release.
• Federal expertise and resources. In October 1978, an Illinois
Gulf Central train derailed near Claxton, Kentucky, causing discharge of
various toxic chemicals and the burning of 2 tank cars containing vinyl
chloride. EPA officials immediately requested assistance from the Public
Information Assistance Team (PIAT) and the Gulf Strike Team (GST). EPA and
the Kentucky Department of Natural Resources and Environmental Protection set
up crews to conduct air monitoring on a 24-hour basis and respond to all
complaints from residents. State requests for participation of an Army
Explosives Ordnance Detachment (EOD) team encountered delays; the state was
told it "would take 4-5 days for the team to arrive. However, EPA was able to
expedite the process; an EOD team arrived on the scene within 8 hours of EPA's
request.
The cost and inconvenience of unnecessary further evacuation was avoided
when EPA conducted air monitoring tests in the nearby town of Dawson Springs.
Area residents were dissuaded from evacuating when EPA detected no toxic
chemicals in the area. EPA's knowledge of special vinyl chloride problems was
important in selecting the proper course of action. Had an attempt been made
to extinguish the burning of the vinyl chloride cars, further explosions could
have occurred, possibly damaging the other tank cars and causing extensive
contamination. The successful and prompt venting of the VC tank cars by the
EOD team accelerated the burning process, shortening the time people initially
evacuated had to remain away from their homes. Cleanup of the toxic chemicals
took a total of 9 days and only one incident of health complications as a
result of the dierailment was reported.
• Timely response. An example of an unreported release related to
us by an On-Scene Coordinator is the situation in Murry, Kentucky. In
December 1980, federal and state officials were alerted to a hazardous
substance release on the Clarkes River by fish kills extending for over 40
miles of the river. The unknown nature of the release resulted in substantial
laboratory testing costs for EPA and state agencies. The value of the lost
fish is estimated to be $18,000, and.the sport fish have not returned to the
area, harming local recreational fishing and trapping. The widespread
contamination of bottom mud with heavy metals and some organic chemicals will
require special permits for any future development of the river, including
building bridges or other near-shore structures. Such economic effects are
difficult to measure, but can be significant to the local community. The cost
of dredging part of the river and the feeder stream which held the initial
slug of contaminants, plus disposal costs of the toxic sludge, will be
substantial. Had the responsible party notified federal or state officials
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
6-15
and contained the contaminant effluent in the feeder stream to the Clarkes
River, the area of contamination would have been much smaller. Early response
would have resulted in much lower cleanup costs and comparatively negligible
economic and environmental damage.
• Government Follow-Up and Monitoring. On September 13, 1974 the
U.S. Coast Guard in Seattle received a report from the U.S. Army of a minor
oil spill as a result of an electrical transformer being dropped into the
Duwamish Waterway. No mention of PCBs was made in the initial spill report.
Three days later, a follow-up investigation by the State and U.S. Coast Guard
revealed that the transformer coolant was PCB and not oil. An EPA laboratory
team arrived at the site on September 18 to perform sampling and analysis.
EPA met with State and Army representatives to investigate alternatives and
responsibility for possible removal and disposal of the PCB material.2
• Conservation of Fund Resources and Concern for Environment. On
July 5, 1979, an explosion and fire occurred at the Drexel Chemical Company in
Memphisx Tennessee. Drexel Chemical was predominantly involved in the
formulation of pesticides and handled these C,lass B poisons: Methyl
Parathion, EPN, and DNBP (Dinoseb). Local Civil Defense, the U.S. Coast
Guard, the OSC, the EPA Emergency Response Team and other Region IV EPA
officials, and state and local officials responded to the emergency. The
initial fire caused significant air pollution concerns. Local officials
evacuated the threatened areas to reduce human exposure to the billowing
smoke. However, as firefighters battled the blaze, disposal of gallons of
contaminated water became a serious problem. An estimated 14 million gallons
of contaminated water from firefighting and broken water lines ran off from
the plant site to a nearby drainage culvert and into a creek which flowed into
Mekellan Lake.
Cleanup at the Drexel Site was performed by a commercial contractor and
was paid for by Drexel Chemical. However, company officials had initially
wondered if disposal costs were going to "break" the company. The OSC worked
with the company and state officials for one and one-half days (July 7-8) to
devise a cleanup plan that was environmentally sound yet within reason on
cost, to treat and discharge the contaminated water. In addition, although
chemical fallout from the fire's smoke appeared minimal, Drexel Chemical
initiated an investigation of the degree and effects of fallout on soil and
vegetation.*
2Willmann, J.C., J. Blazevich, and H. Snyder, "PCB Spill in the
Duwamish-Seattle, WA", Proceedings of the 1976 National Conference on Control
of Hazardous Material Spills, pp. 351-355.
3U.S. Environmental Protection Agency, Region IV, Environmental
Emergency Branch, Chemical Fire at Drexel Chemical Company, Memphis,
Tennessee, November 1980.
November 2 Draft * * * Do not quote or cite
ICF
INCORPORATED
-------
6-16
It may be possible, based on an analysis of many more cases, to arrive at
an estimate of benefits gained from a reduction in the number of improperly
cleaned up releases, including the cleanup costs saved by responding on a
timely basis. However, we suspect that the documentation of case studies,
even supplemented by interviews with participants, will be insufficient to
generate such quantitative estimates.
Another line of inquiry that will be explored is the frequency with which
federal involvement produces these benefits. It would be particularly useful
to discover the percentage of the time that federal involvement produces
benefits for different sizes of releases. If, as we suspect, these benefits
are produced mostly for large releases (where avoidance of potentially
catastrophic results requires enormous coordination and resources), this
information may be useful in selecting and assessing reportable quantity
levels.
6.3.1.2 Benefits of Increased Cleanup of Releases
As discussed earlier, increased cleanup can occur as a
result of designation regulations and regulations lowering reportable
quantities.4 Any increase over baseline activity depends on whether
releasers respond differently to releases of pollutants and contaminants than
to releases of designated substances and whether the reportable quantity
affects response decisions.
The benefits of cleaning up any release are primarily real resource
benefits to people and the environment. As stated earlier, a useful way to
measure these benefits (and one of the few feasible measures) is in terms of
the number of people whose risk of exposure to hazardous substances is
reduced. Ideally, we would estimate the number of people exposed to hazardous
substances by virtue of a reportable release and the number of people whose
risk of exposure may be reduced due to cleanup of the release. The latter
term would then be multiplied by the number of releases adjusted for the
* It is possible that even if RQs are raised, releasers will clean up
more of their releases than under the baseline simply because they are more
aware of the Superfund response authorities and liability provisions.
However, that would be a result of the passage of CERCLA and implementation of
Sections 104 and 105 rather than Sections 102 and 103.
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
6-17
probability that the release will be cleaned up adequately, and the
probability that the release would have been cleaned up under the baseline
policy.5
The actual estimation of these benefits will be a difficult task. The
most obvious difficulty is that releases may occur anywhere and there is no
simple way of obtaining estimates for the number of people potentially exposed
to releases. Our approach would be to examine data on releases in the hope of
finding "average" or model releases of different sizes for which estimates of
surrounding population at risk of exposure can be made. The model releases
would have to vary not only by size of release but by type of release (e.g.
fire, vapor explosion,, spill) and by environmental medium affected.
For each of the other terms in the analysis, similar assumptions and
estimates will need to be made. We anticipate the most serious problems
arising in the estimation of the probability that unreported releases are
cleaned up, since little information will likely be available on that subject.
6.3.1.3 Benefits Resulting from Preventive Activities
The benefits resulting from increased preventive
activities will accrue both to society and releasers. Public health will be
improved by a reduction in the number or severity of releases. It may be
possible to estimate this reduction, although the data currently available do
not permit quantitative estimates. Ideally, we would obtain estimates of the
numbers of releases of different sizes averted through preventive measures,
and then develop estimates of the reductions in numbers of people exposed to
"model" releases of different sizes.
Releasers will also benefit from a reduction in releases as a result of
preventive activities. Whether we can measure these benefits—in the form of
savings on cleanup costs, reduced liability for damages, and savings on
materials formerly wasted—will depend on the data forthcoming from industry
groups.
5In more formal terms, the benefits of the regulation can be expressed
by the following equation:
B = N x R x [p(A)-p(B)]
where
B = benefits
N = number of people whose risk of exposure is reduced due to
cleanup of a single release
R = number of releases
p(A) = probability that release is adequately cleaned up under the
regulation
p(B) = probability that release would have been adequately cleaned
up under the baseline.
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
6-18
6.3.2 Information on Costs
This section presents information on the costs of cleanup and the
costs of preventive actions. Information has not yet been collected on the
costs of the third category of actions, protective actions (insurance).
6.3.2.1 Cleanup Costs
Information on cleanup costs has been collected from the
sources of data discussed in Section 6.2, published reports, and various EPA
and Coast Guard data bases. The published reports are described in more
detail in Appendix A, and summarized in Exhibit 6-2. In general, no
systematic conclusions about cleanup costs can be derived from any of these
reports. The cost estimates for what seem to be relatively similar cleanup
operations often diverge markedly from each other. However, we were able to
draw some conclusions and obtain systematic information about costs from the
PIRS data base.
Only a small proportion of the entries in the PIRS data base contain
information on the cost of cleanup. However, the number of releases in the
PIRS data base is large enough that the total number of releases for which
cleanup cost data are available is large enough for our purposes. Between
1974 and 1980, inclusive, 121 of the PIRS entries contained data on cleanup
costs.
Exhibit 6-3 displays the distribution of cleanup costs for these 121
releases. In order to make the cost figures fox different years comparable,
all cost data were converted to constant 1972 dollars, using the GNP deflator
for nonresidential structures. Over a third of the releases cost between $100
and $500 to clean up. At the extremes, 12 releases cost less than $50, and 7
cost more than $5,000. Of the latter group of 7, two releases cost about
$30,000 each to clean up, and one cost over $1.5 million.
Inspection of the data revealed that they seemed to approximate a
lognormal distribution.6 In order to test this hypothesis, we estimated the
parameters of the distribution and compared the actual distribution to the
6The probability density function of a lognormal distribution is
1 r 1
f (x) = exp I - — (ln(x)-a)
x >J2TTb |_ 2b
The distribution is a function of two parameters, a and b. The mean of
the distribution is exp [a + b/2].
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED j
-------
6-19
EXHIBIT 6-2
SUMMARY OF PUBLISHED REPORTS ON
CLEANUP COSTS
Name of Report
Rockwell
Amson I ..
Amson III
ADL
Orlemann and Higgins
Nye
Willman, Blazevich,
and Snyder
Fullner and Crump-
Weisner
Description
Summary of many other
reports
Cleanup costs-for
different categories
of substances
Cleanup costs for
different categories
of substances
Cleanup costs for
different categories
of substances
Surfactant release
into water
Endrin release into
water
PCB release into water
Organo-chlorine pesti-
cide release into
land and water
Cost Information
Large amount of data--no
systematic conclusions can
be derived.
Soluble substances-
$1.00/lb-.; insoluble
floaters-$0.84/lb.;
insoluble sinking
substances-$22.00/lb.
Costs are an order of
magnitude greater than
earlier estimates.
Costs range from $0.16 to
$11.50/lb., depending on
the substance.
$140,000
$100,000
$150,000
$73,500
Source: See Appendix A for detailed discussion of these reports and complete
references.
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
6-20
EXHIBIT 6-3
COST OF CLEANUP OF RELEASES FROM
PIRS DATA BASE, 1974-1980
Cost
(Constant 1972 $) Number Percent of Total
500-1,000
1,000-5,000
> 5,000
0-50
50-100
100-500
12
13
47
26
16
. 7
9.9
10.7
38.8
21.5
13.2
5.8
TOTAL
121
100.0
expected distribution, given the estimated parameters and the assumption of
lognormality. We failed to reject the hypothesis that the cost data are
distributed lognormally, with a mean cost of $1,537.7
It is important to note that $1,537 is the mean of the distribution that
we fit to the data, not the mean of the data in the sample. As mentioned
above, one of the releases alone cost more than $1.5 million to clean up.
Thus, the sample mean must be well over $10,000. If the data really do follow
a lognormal distribution, the very cdstly release is~an occurrence with very
low probability. Because the sample mean is very sensitive to the values of
such outlying data points, while the mean of the estimated distribution is
less sensitive to outliers, the mean of the estimated distribution is a better
estimate of the expected value of the cost of future releases, as long as the
distribution really is lognormal.
In order to determine the extent to which our conclusions about the
lognormality of the distribution are influenced by the outlier, we excluded
the outlier from the data set, reestimated the parameters, and tested the
'Remember that the costs are in constant 1972 dollars. To obtain the
nominal mean cost for any given year, it is necessary to adjust by the
appropriate price index. Thus, $1,537 must be multiplied by 2.252 to give
$3,461, the average cost in 1980 dollars. In interpreting the results of the
statistical test, it is useful to know that "failing to reject" a hypothesis
means that, given available information, it is permissible to base analyses on
that hypothesis. Thus, in the present case, it appears that using the
lognormal distribution to approximate the size distribution of releases is
appropriate.
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
6-21
hypothesis that the remaining data fit the new distribution. Once again the
fit was a good one. We failed to reject the hypothesis that the cost data
(less the outlier) are distributed lognormally, with a me'an of $1,075. In
fact, the fit to this distribution is better than the fit to the first
distribution. As can be seen, the mean of the distribution is sensitive to
outliers, although it is nowhere near as sensitive as the mean of the data in
the sample.
The incidence of such outlying data points raises the question of whether
all the cost data are derived from the source underlying distribution.
Perhaps there are really two types of releases: lognormally distributed small
releases, including 120 of the 121 releases considered here; and large
releases, only one of which is included in this sample and whose distribution
is as yet unspecified. This question will be explored in greater detail in
the discussion of the number and quantity distribution of releases in Section
6.3.3. The comments there apply to the cost distribution as well as the
quantity distribution.
As discussed in Section 6.2, the cost of cleaning up a release is a
function of the identity of the substance, the quantity released, and
environmental factors specific to each release. A measure of the importance
of quantity alone in determining the cost of cleanup is the correlation
between quantity and cost. For this data set, the correlation between the
logarithms of quantity and cost is 0.52.8 Therefore, 27 percent (the square
of the correlation coefficient) of the variance in cost of cleanup can be
explained by variations in the quantity released. As yet, we have not
examined the multivariate relationship between cost, the identity of the
substance, and quantity. Because of the large number of substances under
consideration, far more data would have to be used in such an investigation.
Although we have not examined the multivariate relationship among cost,
the identify of the substance, and quantity, we have examined the bivariate
relationship between cost and quantity. Several of the reports summarized in
Exhibit 6-2 calculate the cost of cleanup as the cost per pound of substance
released. Under this assumption, if twice as much is released, it will cost
twice as much to clean up. But an analysis of the PIRS data on cleanup cost
and quantity of release shows that the relationship between cost and quantity
is not linear. Instead, the relationship between the logarithms of cost and
quantity is linear.
Statistics summarizing the relationship between cost and quantity are
presented in Exhibit 6-4. As can be seen, the correlation between cost and
quantity for the entire data set, 115 data points,9 is 0.48, meaning that 23
8 Exclusion of the outlier reduces the correlation to 0.48.
9Analysis of the cost distribution above was based on 121 data points.
But six of these did not contain information on quantity released, and were
discarded for this analysis.
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
6-22
percent of the variation in removal cost is explained by variation in quantity
released. However, the correlation coefficient is extremely sensitive to
outlying data points. As shown in the exhibit, removal of the most extreme
outlier, with cleanup costs of more than $1.5 million, reduces the correlation
to 0.30, and removal of a second outlier reduces the correlation to 0.02. The
regression coefficients are similarly unstable.10 For the full data set,
the slope coefficient (b) is 16.2. The removal of one outlier reduces it to
0.30, and the removal of the second outlier reduces it to 0.02.
EXHIBIT 6-4
RELATIONSHIP BETWEEN COST OF CLEANUP AND QUANTITY RELEASED
Regression Coefficients11
Number of Data Points
Correlation
Intercept
Slope
115
.48
-3955
16.2
114
.30
1022
.30
113
.02
1058
.02
As shown in Exhibit 6-5, the true relationship between cost and quantity
is much closer to being linear in the logarithms of the two variables. Using
the entire data set, the correlation between the log of cost and the log of
quantity is 0.52. That correlation is not attenuated much by the removal of
the two outliers. It follows, therefore, that the regression coefficients are
relatively insensitive to the removal of outliers.
xeIn future analyses, it may be useful to project cost by using an
equation expressing cost as a function of quantity. The regression
coefficients are needed for making such projections.
"The equation estimated is:
y = a+bx+e
where:
y =
cost, in dollars
X -
quantity, in gallons
a =
intercept
b =
slope
e =
error
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
6-23
EXHIBIT 6-5
RELATIONSHIP BETWEEN THE LOGARITHMS OF
COST OF CLEANUP AND QUANTITY RELEASED
Regression Coefficients12
Number of Data Points Correlation Intercept Slope
115 .52 4.17 .39
114 .48 4.40 .33
113 .44 4.50 .30
Assuming that the hypothesis that the data are lognormally distributed is
valid, we conclude from the above analysis that one cannot estimate cost of
cleanup by multiplying quantity released by a constant cost per pound or
gallon cleaned up. Total cost of cleanup depends not only on the total
volume of material released, but on how that volume is distributed among
releases.
Another conclusion to be drawn from the data is that the environmental
factors particular to each release are important influences on the total cost
of cleanup. Even using the lognormal specification of the variables, only 27
percent of the variation in cost is accounted for by variations in quantity
released. Undoubtedly the addition of substance identity would explain more
of the variance, but it is likely that the characteristics of the release site
and time (e.g., type of soil, contours of the land, weather) are important
determinants of the cost of the release.
6.3.2.2 Costs of Preventive Activities
We have obtained information on the cost of preventive
activities from three reports. Each of these is summarized below.
12The equation estimated is:
ln(y) = a+bln(x)+e ln(y)
ln(x)
a
b
= natural logarithm of y, in dollars
= natural logarithm of x, in gallons
= intercept
= slope
= error
November 2 Draft
* *
Do not quote or cite
ICF INCORPORATED
-------
6-24
• MRI Report13—Midwest Research Institute (MRI)
examined the economic impact of EPA's proposed
regulations to require all facilities using any of the
substances designated as hazardous under Section 311 of
the Clean Water Act to prepare and implement Spill
Prevention Control and Countermeasure (SPCC) plans to
prevent releases. The estimates in the report were
developed using the following method:
-- First, the specific hazardous substances likely to
be present for each type of industry subcategory
subject to the regulation were identified.
Hazardous substances present as raw material,
by-product, or commercial product were all
considered. Using information on the ratio of
hazardous substance produced to final output
produced, and on total production of final output,
the quantity of each hazardous substance on hand
for each industry subcategory was estimated.
-- Model plants were developed by grouping industry
subcategories with others having similar operations
and using similar hazardous substances.
Costs were estimated for each model plant for each
of the following capital and operating cost
elements:
• SPCC plan preparation;
• Secondary containment of bulk storage areas;
• Secondary containment of process tanks and
preventive measures at in-process transfer
areas;
• Secondary containment of loading/unloading
facilities;
• Preventive maintenance, inspection,
recordkeeping, and housekeeping;
• Installation of fences and lights for
facility security; and
• Employee training and program administration.
13 Cost Analysis for Proposed Hazardous Substance Pollution Prevention
Regulation (40 CFR PART 151), August 7, 1980.
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
6-25
— The incremental cost of complying with the SPCC
regulations was estimated by subtracting the cost
of complying with existing regulations and by
accounting for preventive facilities already in
place and activities already being undertaken.
The report estimated that over the period 1980-1984,
the capital costs caused by the SPCC regulation would
range between $74.6 and $210.3 million. It also
estimated that annual variable costs would range
between $5.8 and $16.3 million by 1984, when all
facilities would be covered by the regulation. Of more
importance for this analysis, the report provides
constant dollar estimates of capital and operating
costs for each of the above-mentioned cost elements for
each of 22 model plants.
• Battelle Reportlb—In support of regulations
dealing with the designation of rates of penalty and
harmful quantities for releases of hazardous substances
under Section 311 of the Clean Vater Act, Battelle
Memorial Institute developed four different
methodologies. Under one of these methodologies, the
DOHM method, the rate of penalty is equated to the cost
that would have been incurred by the discharger had he
instituted measures to prevent the release. During the
course of the methodology, Battelle concluded that the
cost of prevention of releases is unrelated to the
toxicity, dispersion, or degradation characteristics of
the substance. However, cost is a function of the
substance's corrosiveness and flammability. Therefore,
Battelle developed two sets of estimates: one for
noncorrosive, nonflammable substances, and the other
for either corrosive or flammable substances.
Battelle developed estimates for stationary sources,,
barges, and railroads (the prevention cost for trucks
was assumed to be the same as the cost for stationary
sources, because 95 percent of truck accidents involve
spillage at loading and unloading facilities). For
stationary sources, Battelle surveyed two major
chemical manufacturers. One firm reported a cost of
$1,000,000 for the installation of a complete plant
spill prevention system, including dikes, levees,
drainage ditches, sewers, sumps and pumps, wells,
1*EPA, Determination of Harmful Quantities and Rates of Penalty for
Hazardous Substances. 4 volumes (Washington,D.C.: GPO, 1975),
EPA*440/9-75-005-a, b, c, d.
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
6-26
holding tanks or ponds, monitoring equipment, roofs
oyer process holding tanks 01 ponds, monitoring
equipment, roofs over process areas, valves, concrete
pads, and other similar devices. According to this
firm, a similar system for corrosive or flammable
mateials would cost $2,000,000. Another firm estimated
the cost of the total system as $3,000,000 and reported
annual operations and maintenance costs of $80,000.
This firm did not differentiate between different types
of substances. The cost of prevention for barges was
taken to be the incremental cost of double skin barges
over single skin barges. The incremental cost was
estimated to be $24,000, in capital costs, plus $1,080
per year in operation and maintenance costs. For tank
cars, the capital cost of spill prevention was
estimated as $1,835 for one type of railroad car, and
$10,085 for another.
Although specific estimates of prevention costs were
presented in this report, the estimates were based on
only two data points. Furthermore, only the costs of a
complete system were estimated. Costs of the
individual components of the system were not provided.
The observation that corrosivity and flammability are
the only important variables in determining prevention
costs is a useful tool for future analysis.
• Amson II15 — In a paper written for the 1978
National Conference on Control of Hazardous Material
Spills, Amson and Goodier presented data on the cost of
equipment and procedures for prevention and control of
spills.16 The spill prevention costs were obtained
from a wide variety of sources: facilities in which
spill prevention and control measures had already been
implemented; organizations that provide field testing
services, electrical contracting firms, abrasive blast
cleaning firms, consulting engineers, industrial
instrumentation and control suppliers, weed control
organizations, marine contractors, and drainage and
evacuation firms. Costs were presented for eight major
15Jonathan E. Amson and J. Leslie Goodier, "An Analysis of the Economic
Impact of Hazardous Substances Regulations," Proceedings of the 1976 National
Conference on Control of Hazardous Material Spills, pp. 39-45.
lsThe data come from a report by ADL which we have not yet been able to
obtain.
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
6-27
categories of prevention and control procedures: spill
containment systems, in-plant pipelines, bulk storage
tanks, liquid level alarm systems, vehicle loading and
unloading areas, marine loading and unloading piers,
plant drainage systems, and facility security
measures. The costs were not given in terms of the
amount required for a typical plant; instead, they were
given per unit of equipment (e.g., $100 per locked
drain valve, or $120 per natural earthen dike 6' high
with 1:2 slope). In general the author stated that
because no standardization for spill prevention exists,
each individual facility requires a custom-designed
spill prevention, containment, and response program.
6.3.2.3 Further Investigation of Costs
More data on costs can be gathered from the same type of
sources summarized here. We have not identified all the studies in which cost
•in format inn appears, nor analyzed all the data bases which include cost data.
Before proceeding, it is necessary to determine what type of information
gathering and analytic efforts will be most useful for future regulatory
impact analyses.
Below we briefly discuss several paths that could be pursued. Each path
is evaluated in terms of the amount of new information that can be acquired,
the cost of obtaining that information, the utility of the information for
regulatory impact analyses, and the utility of that information for other
purposes. The options are not mutually exclusive. Indeed, it may be decided
that all of them should be pursued.
• Gather more information and conduct more analyses of
the type described above. We have not yet been able
to digest all of the information in published reports
and compilations of information about individual
releases. The analysis of more information of this
type would help determine whether our conclusions are
an artifact of the particular data analyzed, or whether
they are more generally applicable. This option would
be relatively inexpensive, and would provide
information on costs and the relationship between cost
and quantity released that would be useful for
conducting regulatory impact analyses. In addition,
knowledge of this type may be useful to EPA in
projecting the cost of responding to releases under
different scenarios, thereby helping to better manage
disbursements from the fund.
• Solicit information on costs from regulated parties
and others with knowledge of costs. This information
could be very useful in supplementing the information
already gathered. Although these sources can probably
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
6-28
not provide quantitative data, they can provide
qualitative information which might modify the
conclusions reached on the basis of quantitative
analysis. For example, they might be able to
supplement 'the information in the MRI report on the
degree to which preventive equipment and procedures are
already in place. This option would be relatively
inexpensive, because these parties must be contacted in
order to estimate the response of regulated parties to
CERCLA' These parties can also provide information on
the costs of notifying EPA when a release occurs, and
the costs of any recordkeeping and changes in
operational procedures associated with notification.
• Detailed investigation of the determinants of removal
costs. As stated above, the cost of cleaning up a
release is a function of the identity of the substance,
the quantity released, and environmental conditions
specific to each release. We have analyzed the
relationship between cost and quantity, but to analyze
and estimate statistically the relationship among all
variables would be much more costly and time-consuming.
The 115 cases included in the above analysis would not
be sufficient--many more data points are necessary to
adequately investigate all of the environmental
variables that might influence cost. In fact, it is
possible that not enough data are available to perform
such an investigation. This type of analysis is not
necessary for the preparation of a regulatory impact
analysis. But if EPA wishes to investigate in any
depth the determinants of cleanup costs, a multivariate
statistical analysis of the factors influencing cleanup
costs is essential.
• Estimate the impact of the identity of the substance
released on cleanup costs. A multivariate analysis
including only quantity and substance as determinants
of cost would be less costly than the one outlined
above. - The inclusion of each of the hundreds of
substances regulated by CERCLA would require large
amounts of data. But if all hazardous substances could
be divided into categories such as those in Amson II or
the ADL report, the differences among the categories
could be examined with only a few more data than we now
have. This option would improve our ability to project
release costs, and would be useful in conducting
regulatory impact analyses.
6.3.3 Information on Numbers of Actions
This subsection presents information on the number of actions that
might occur under the Sections 102 and 103 regulations. We present
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
6-29
information only on the number and size distribution of releases here.
Information can also be gathered on the number of facilities subject to the
regulations and the response of regulated parties.
The remainder of this subsection analyzes the total number of releases and
releases by class of substance (Section 6.3.3.1); the number of releases above
and below reportable quantities (Section 6.3.3.2); the size distribution of
releases (Section 6.3.3.3); and further investigations (Section 6.3.3.4).
6.3.3.1 Total Number of Releases and
Distribution by Class of Substance
To estimate the total number of releases, it is necessary
to combine the information from several data bases. Although each data base
includes a different set of releases, there may be an overlap among the data
bases. To avoid double-counting, those releases which are included in more
than one data base must be identified. Therefore, the first step in
estimating the total number of releases is to determine which releases are
included' in more than one data base.
Exhibit 6-6 displays the results of a comparison of PIRS to the
compilation of releases reported to EPA regions in FY 79. There was very
little duplication between these two data bases. We were able to identify
only 32 releases as appearing in both data bases, although it is possible that
24 others may have been duplicates. The information included in each data
base is insufficient to identify unambiguously a particular release, so there
will always be some uncertainty about how many releases are included in more
than one data base.
We have examined the distribution of reported releases among substances
using two data bases: PIRS and the compilation of hazardous material
incidents reported to EPA Region IV in FY 79. Both data bases contain reports
of releases from a wide range of substances. As shown in Exhibit 6-7, the 236
PIRS releases from 1980 include 50 specific chemicals, plus 56 releases listed
as "other hazardous substances". The 150 PIRS releases from 1980 include 49
different chemicals, plus 17 releases listed as "other hazardous substances".
The 323 releases from EPA Region IV include releases of 72 specific chemicals,
plus 76 listed as "other hazardous substances". There seems to be very little
continuity from year to year on the specific chemicals which are reported to
be released. For example, of the 50 chemicals listed in the 1979 PIRS data,
only 26 appear in the 1980 data. As might be expected, the chemicals whose
releases are most frequently reported are, with the exception of PCBs,
produced at the rate of billions of pounds per .year.
6.3.3.2 Number of Releases Above and Below
Reportable Quantities
Exhibit 6-8 displays information on the percentage of
releases of selected substances which are at or above the reportable
quantities. As can be seen, a substantial percentage of the reports received
are below the reportable quantities for those substances.
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
6-30
EXHIBIT 6-6
NUMBER OF RELEASES REPORTED TO EPA AND TO PIRS, FY 79
Number of
Releases 1
Regions
Reported 10
to PIRS
Double- 0
counted _
Possible 0
double-
counts
TOTAL
b/
7 8
25
30
30
11
38
91 3
5 1
4 0
10 Total
Reported 152 205 256 323 280 180 86 54 179 51
to EPA a/
3
1
(high) 162 227 278 342 317 266 88 54 182 53
• £/
(low) (102) (220) (273) (340) (313) (262) (88) (54) (181) (52)
1,766
235
32
24
1,969
(1945)
-Total number hazardous material and combined oil and hazardous
material incidents.
-^High = sum of releases reported to EPA and to PIRS minus
double-counted releases.
c /
-Low = high minus possible double-counted releases.
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
6-31
EXHIBIT 6-7
DISTRIBUTION OF RELEASES BY SPECIFIC
CHEMICAL IDENTITY, PIRS 1979-1980 AND
EPA REGION IV, FY 79
PIRS 1979
PIRS 1980
EPA Region IV
Number of
Releases
Number of
Specific
Substances
Benzene
PCBs
Phosphoric
Acid
Styrene
Sulfuric
Acid
Toluene
Xylene
Others
Number
236
50
18
2
4
13
10
17
18
154
Percent of
Total
Releases
100.0
7.6
0.8
1.7
5.5
4.2
7.2
7.6
65.3
Number
150
49
10
6
2
13
14
8
6
91
Percent of
Total
Releases
100.0
6.7
4.0
1.3
8.7
9.3
5.3
4.0
60.7
Number
323
72
3
35
14
4
30
5
3
229
Percent of
Total
Releases
100.0
0.9
10.8
4.3
1.2
9.3
1.5
0.9
70.1
November 2 Draft * ¦* * Do not quote or cite
,/Nf" INCORPORATED
ICF i
-------
6-32
EXHIBIT 6-8
PERCENTAGE OF RELEASES REPORTED AT OR
ABOVE REPORTABLE QUANTITIES FOR
SELECTED SUBSTANCES, PIRS 1979-1980
AND EPA REGION IV, FY
PIRS Region IV
Substance
m
1979
1980
FY 79
Toluene
1,000 lbs.
47%
57%
54%
(17)
(7)
(13)
Sulfuric Acid
1,000 lbs.
78%
67%
67%
(9)
(9)
(52)
Xylene .
1,000 lbs.
7%
0%
57%
(14)
(5)
(7)
Chlorosulfonic
1,000 lbs.
43%
80%
no reports
Acid
(7)
(5)
Napthalene
5,000 lbs.
50%
no reports
55%
(4)
(11)
Benzene
1,000 lbs.
11%
44%
100%
(18)
(9)
(3)
Hydrochloric
5,000 lbs.
0%
80%
46%
Acid
(3)
(5)
(24)
Styrene
1,000 lbs.
50%
54%
67%
(10)
(13)
(9)
Ammonia
100 lbs.
100%
100%
100%
(3)
(4)
(5)
Phosphoric Acid
5,000 lbs.
0%
100%
62%
(4)
(1)
(16)
TOTAL
35%
59%
62%
(89)
(58)
(140)
— The numbers in parentheses refer to the total number of reported
releases of each substance in each year. The numbers in parentheses may not
correspond to the numbers in Exhibit 6-7, because the numbers in this table
include only those releases for which quantity data are available.
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
6-33
6.3.3.3 Size Distribution of Releases
Recalling the discussion of the use of the size distribution of releases
in Section 6.2.2, here, we examine EPA data to determine the manner in which
the required reporting of releases would vary as the RQ level is altered.
Thus far we have examined the size distribution of releases from two
sources of information: PIRS and releases reported to EPA Region VII. Two
subsets of the PIRS data are analyzed here. One subset consists of the 115
cases between 1974 and 1980 for which both cost and quantity data exist.
Using that data set, we have already analyzed the cost distribution and the
interaction between cost and quantity. Here we examine only the size
distribution. The other subset consists of all cases from 1980 for which
quantity data exist. Data from EPA Region VII were chosen for analysis
because they are reportedly the most complete data available. Region VII
encourages regulated parties to report all releases, and has a computer system
for storing and retrieving reports of releases. We have received summaries of
all releases reported to Region VII between May 11 and October 7, 1981.
Analysis of the data presented below indicates that the PIRS releases and
the EPA Region VII releases other than PCB releases are distributed
lognormally. The PCB releases are distributed differently, conforming to an
exponential distribution in which the expected size of a release falls as the
size of the release increases. The average size of the PCB releases is much
smaller than the average size of the other chemical releases. The difference
between the PCB releases and other releases in terms of number and average
size causes us to conclude that a large number of releases go unreported.
However, the substantial number of reported releases which are below the
reportable quantity suggests that some overreporting also occurs.
Size Distribution - PIRS Data—The distribution of the 123 releases which
were reported to PIRS in 1980 and for which data on quantity releases were
available is summarized in Exhibit 6-9. The majority of reported releases
were between 100 and 10,000 pounds, although some very small and some very
large releases were reported. At the extremes, the quantities of 6 releases
were less than 10 pounds and the quantities of 8 releases were greater than
100,000 pounds, including one of almost 2 million pounds, one of almost 3
million pounds, and one of 8 million pounds.
EXHIBIT 6-9
SIZE DISTRIBUTION OF RELEASES FROM
PIRS DATA BASE, 1980
Size (lbs.)
Number
Percent of Total
0-10
6
4.9
10-100
22
17.9
100-1,000
34
27.6
1,000-10,000
30
24.4
10,000-100,000
23
18.7
> 100,000
8
6.5
TOTAL
123
100.0
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
6-34
Inspection of the data revealed that they seemed to approximate a
lognormal distribution. We estimated the parameters of the distribution,
assuming lognormality, and tested the hypothesis that the data fit this
distribution. We failed to reject the hypothesis that the quantity data fit a
lognormal distribution with a mean size of 95,071 pounds. It must be noted
that the estimate of the mean of the distribution is sensitive to extreme data
points. For example, the exclusion of the most extreme outlier from the data
reduces the estimate of the mean of the distribution to 65,797 pounds.
However, even if that outlier is excluded and the parameters of the
distribution reestimated, the data still fit the distribution very well.
An examination of a plot of all the data points by quantity released does
not reveal & smooth distribution. Instead, as shown in Exhibit 6-10, there
are concentrations of reports at certain quantities, such as 1, 3, 5, 20 and
500 gallons. To a certain extent, this is due to the fact that these
quantities are equal to the volume of commonly used containers. For example,
if an accident is likely to result in the complete release of the material
contained in a.5 gallon can, releases of 5 gallons are likely to be relatively
common. But the concentrations of reports at certain quantities are also due
to the natural tendency to give an approximate estimate of the quantity
released. For example, a release of between 8 and 12 gallons might be
reported as being approximately 10 gallons.
EXHIBIT 6-10
CONCENTRATION OF REPORTS AT
CERTAIN QUANTITIES, PIRS DATA BASE, 1980
Size (gallons)
Number Percent of Total
42 (1 barrel)
50
100
200
500
3,000
4,000
1
3
5
20
25
5
5
7
7
4
4
4
4
4
5
4
4
4.1
4.1
5.7
5.7
3.25
3.25
3.25
3.25
3.25
4.1
3.25
3.25
All other sizes
66
53.7
TOTAL
123
100.0
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
6-35
The size distribution of the sample of 115 PIRS data points from the
period 1974-1980 that were used in the cost analysis also have a lognormal
distribution, but it is not the same as the lognormal distribution.estimated
above. The hypothesis that the 1974-1980 data fit the distribution estimated
for the 1980 data was rejected. However, the 1974-1980 data did fit a
different lognormal distribution, one with a mean size of 9,077 pounds. The
size distribution from the 1974-1980 data set is displayed in Exhibit 6-11.
The difference between the two distributions may be due to some kind of
reporting bias. It is possible that cost data are more frequently provided
for small releases than for large releases. This conclusion is buttressed by
examining the eight data points that appear in both data sets. By comparing
Exhibit 6-12 to Exhibit 6-9, it can be seen that the releases for which both
cost and quantity data are provided in 1980 seem to be smaller than those for
which only quantity data are provided.
EXHIBIT 6-11
SIZE DISTRIBUTION OF RELEASES
PIRS DATA BASE, 1974-1980-/
Size (lbs.)
Number Percent of Total
1,000-10,000
10,000-100,000
> 100,000
0-10
10-100
100-1,000
2
19
50
30
11
3
1.7
16.5
43.5
26.1
9.6
2.6
TOTAL
115
100.0
€i/
- Only those data points for which
both cost and quantity information are avail-
able are included.
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
6-36
EXHIBIT 6-12
SIZE DISTRIBUTION OF RELEASES,
PIRS DATA BASE, 1980-/
Size (lbs.) Number Percent of Total
0-10 0 0
10-100 1 12.5
100-1,000 4 50.0
1,000-10,000 3 37.5
10,000-100,000 0 0
> 100,000 0 0
TOTAL 8 100.0
fi/
- Only those data
both cost and quantity
able are included here
points for which
information are avail-
Size Distribution - Region VII Data—There were a total of 259 hazardous
substance releases reported to EPA Region VII between May 11 and October 7,
1981. Of these 93 were reports of PCB releases, and 166 were reports of
releases of other hazardous substances. Quantity released was reported for 75
of the PCB releases, and for 104 of the other releases.
The distribution of the 104 other chemical releases in the Region VII
sample is shown in Exhibit 6-13. As with the PIRS sample, the majority of
reported releases were between 100 and 10,000 pounds. The largest release was
32,000,000 pounds, and the next largest was 1,680,000 pounds.
EXHIBIT 6-13
SIZE DISTRIBUTION OF NON-PCB
RELEASES, EPA REGION VII, 1981
Size (lbs.) Number Percent of Total
0-10 9 8.7
10-i00 15 14.4
100-1,000 27 26.0
1,000-10,000 38 36.5
10,000-100,000 11 10.6
> 100,000 4 3.8
TOTAL 104 100.0
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
6-37
Inspection of the data revealed that they seemed to approximate a
lognormal distribution. We estimated the parameters of the distribution,
assuming lognormality, and tested the hypothesis that the data fit the
distribution. We failed to reject the hypothesis that the data fit a
lognormal distribution with a mean size of 49,935 pounds. Once again, we note
that the estimate of the mean of the distribution is sensitive to outlying
data points. For example, if the most extreme outlier is excluded and the
parameters of the distribution reestimated, the mean of the new distribution
is 27,063 pounds. However, the overall shape of the distribution is not very
sensitive to outliers. Even if the extreme outlier is excluded, the data fit
the estimated distribution very well.
As shown in Exhibit 6-14, the size of reported PCB releases is
substantially different from the size of other types of reported releases.
Virtually all PCB releases are less than 5 gallons, about 36 pounds. The
largest release reported to Region VII was 90 gallons, about 645 pounds.
EXHIBIT 6-14
SIZE DISTRIBUTION OF PCB RELEASES,
EPA REGION VII, 1981
Size (gallons)
Number
Percent of Total
< 1
5
6.7
1
31
41.3
2
15
20.0
3
9
12.0
4
2
2.7
5
7
9.3
> 5
_6
8.0
TOTAL
75
100.0
Inspection of the data revealed that they seemed to approximate a
exponential distribution.17 To test this hypothesis, we estimated the
parameter of the exponential distribution and tested the fit of the data to
that distribution. The fit was very poor. However, the exclusion of the two
17The probability density function of an exponential distribution is
f(x) = a exp(-ax). The distribution is a function of one parameter, a. The
mean of the distribution is 1.
a
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
6-38
extreme outliers (80 and 90 gallons) produced a better fit. Based on the
reestimation of the distribution excluding the two outliers, we failed to
reject the hypothesis that the data were from an exponential distribution with
a mean of 2.39 gallons.
In summary, it appears that reported releases other than Region VII PCB
releases follow a lognormal distribution. Both PIRS 1980 releases and EPA
Region VII releases of chemicals other than PCBs were found to fit lognormal
distributions, albeit slightly different ones. We tested whether both data
sets could be fit to the same distribution by combining the data (see Exhibit
6-15), estimating the parameters of the lognormal distribution from the merged
data set, and testing whether the merged data set and each of the original
data sets fit the lognormal distribution based on all the data. The lognormal
distribution estimated from all the data fits the entire data set extremely
well. In addition, each of the two subsets fits this distribution. The mean
of the merged distribution equals 69,953 pounds.
EXHIBIT 6-15
SIZE DISTRIBUTION OF RELEASES,
PIRS DATA BASE, (1980) AND
NON-PCB RELEASES, EPA REGION VII, (1981)-/
Size (lbs.)
Number
Percent of Total
0-10
15
6.6
10-100
37
16.3
100-1,000
61
26.9
1,000-10,000
68
30.0
10,000-100,000
34
15.0
> 100,000
12
5.3
TOTAL
227
100.0
fl /
- The data in this exhibit include
both the data in Exhibit 6-9 (123 data points)
and Exhibit 6-13 (104 data points).
The fit between the size data and the lognormal distribution may be
masking more complex relationships. There are many sources of releases, and
each source should generate its own distinct distribution of releases. For
example, there is no reason to believe that the distribution of releases from
barge accidents is the same as the distribution of releases from leaking
valves. Even if releases from both sources followed lognormal distributions,
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
6-39
there is no reason to expect the means of the two distributions to be the
same. It is possible to determine if the size distribution of releases varies
according to the identity of the source, but this would require the analysis
of much more data than we have analyzed so far.
Perhaps the most important conclusion that emerges from the data analysis
is that a large number of releases currently go unreported. This can be seen
by focusing on the releases reported from EPA Region VII. In FY 79, only 86
releases were reported in Region VII. But since then, Region VII has
encouraged regulated parties to report all releases that occur. As a result,
in 5 months of 1981 for which data are available, 259 releases were reported
¦in Region VII. Because there is no good reason for the number of releases to
have increased dramatically since 1979, we conclude that reporting has
increased because of the passage of CERCLA, which increased the number of
subtances whose releases must be reported, and/or because of the efforts of
Region VII officials to encourage reporting.
But the Region VII data also indicate that many releases still go
unreported. According to the Region VII data, 93 of 259 reported releases, or
36 percent, are PCB releases, and the average PCB release is also far smaller
than the average release of other hazardous substances. The size distribution
of PCB releases (exponential) is also different from the size distribution of
all other releases (lognormal). A plausible explanation for these data is
that PCB releases are much more likely to be reported than releases of other
hazardous substances, because of the danger of PCBs and the publicity
surrounding this particular chemical. In particular, most small releases of
substances other than PCBs are not reported. If there are enough of these
small, unreported releases of other chemicals, the distribution of other
chemical releases is really exponential, rather than lognormal." Under this
scenario, the lognormality of the distribution of chemicals other than PCBs is
caused by underreporting.
If we are to believe that releases are not underreported in Region VII, it
is necessary to believe that PCBs, which represent only a small fraction of
all hazardous chemicals in use, are much more likely to be released than are
other hazardous chemicals. In addition, it is necessary to believe that while
releases of PCBs are usually quite small, releases of other chemicals are
usually quite large. Neither of these beliefs is plausible. Therefore, we
conclude that many releases still go unreported, even in a region which has
made great efforts to encourage the reporting of releases.
Although it appears that many releases are unreported, it also appears
that- many releases are reported that are not required to be reported under
the law. As shown in Exhibit 6-8, about half of the reported releases of 10
different chemical substances are below the reportable quantities. What
probably happens is that some firms are very diligent about reporting
releases. These firms are likely to report all releases of a hazardous
substance, whether the release is above or below the reportable quantity. In
contrast, other firms report none of their releases. This view is confirmed
by conversations with EPA officials in Region VII. For example, in Region VII
one utility reports an average of five PCB releases a week from electrical
transformers, but another utility has never reported a release.
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
6-40
6.3.3.4 Further Investigations
Given the information gathered thus far and the
conclusions that can be drawn from it, where should further data gathering and
analytical efforts be directed? Below we briefly discuss several paths that
could be pursued. We evaluate each path in terms of the amount of new
information that can be acquired, the cost of obtaining that information, the
utility of the information for regulatory impact analyses, and the utility of
that information for other purposes. The options are not mutually exclusive,
and again, it may be worthwhile to pursue all options discussed below.
• Gather more information and conduct more analyses on the number
and size of releases. We have not yet acquired and analyzed all the data on
releases. Analysis of all these data would enhance the quality of estimates
of the number of releases expected under the baseline and under various
regulatory alternatives. Moreover, additional information will enable us to
test further the estimates of the size distribution of release data. A good
estimate of the number of releases would be essential for a regulatory impact
analysis", and (estimates of the size distribution are necessary for any
regulatory impact analysis of the regulations establishing reportable
quantities for designated substances.
• Obtain information on the response of regulated parties to
possible regulatory alternatives. As discussed above, information on the
response of regulated parties would be essential for conducting any regulatory
impact analysis. Some of the questions we would seek to have answered by
regulated parties are:
What are your current procedures for responding to
hazardous materials spills?
Do you respond differently to releases of substances
designated as hazardous under CERCLA than to releases
of other substances that you may consider hazardous?
How have your response and prevention procedures
changes since the passage of CERCLA?
Were those changes made in response to the mandate of
CERCLA, or would they have been made anyway?
What did these changes cost (both capital and operating
costs)?
What changes would you make in your procedures if the
reportable quantities of hazardous substances were
changed?
Do you report releases if they are below the reportable
quantity? If so, why?
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
6-41
What changes would you make in your operating
procedures if additional substances were designated as
hazardous under CERCLA?
This type of information could -also help EPA in planning measures for the
prevention of releases.
• Obtain more information on the number of parties which may be
affected by the Sections 102 and 103 regulations. We have developed rough
estimates of these numbers from existing data, but much more could be done.
However, further efforts in this direction may be very costly. For example,
if the FINDS system is not available, it would be necessary to eliminate
duplicates manually from the lists of facilities with different types of
permits. This would be a very costly and time-consuming process.
• Conduct a detailed statistical investigation of the differences in
number and distribution among different types of releases. As discussed
above, there is no reason to believe that all types of releases are members of
one common distribution. In fact, we would expect releases from different
sources to have different size distributions. The estimation of separate
distributions for different types of releases would require much more data
than we have analyzed thus far. Such an analysis would be useful, though not
essential, for the conduct of regulatory impact analyses. However, the
information might be useful for future EPA efforts to prevent releases. By
focusing attention on the sources of releases which pose the greatest
problems, such an analysis might help to make EPA actions more efficient.
6.4 SUMMARY AND CONCLUSIONS
This chapter has presented an approach to measuring the benefits and costs
of the Sections 102 and 103 regulations and our findings from the data
analysis conducted thus far. The major conclusions from the data analysis are:
1. With the exception of PCBs, the substances most frequently reported
released are produced in billions of pounds per year. This finding tends to
confirm the intuitive conclusion that the substances produced in the largest
quantities are released most often. However, the number of releases of PCBs,
which are no longer used in new products, is very high (about 35 percent of
Region VII's reported releases). This indicates that the candidates for
designation should include not only those substances produced in large
quantities today, but also those which were previously produced and widely
used. Some pesticides, for example, whose registration for most or all uses
has been cancelled, should still be candidates for designation.
2. About half of all reports of releases are for quantities below the
reportable quantity. This implies that at least some firms have been
reporting releases without regard to the reportable quantity established by
EPA. Estimating the extent of such behavior is important for assessing the
effects of changes in reportable quantities.
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
6-42
3. In general, PCB-reported releass are much smaller than reported
releases of other chemicals, and the number of PCB-reported releases is a
large percentage of total releases in the EPA region in which reporting
appears to be most complete (Region VII). Because of the differences between
reports of PCB releases and reports of releases of other chemicals, a
statistical analysis of the release data indicates that there may be
substantial underreporting of releases of other chemicals. Thus, the data
obtained so far indicates that some releases are reported whether or not the
RQ has been reached. Other releasers seem not to be reporting releases
regardless of the RQ.
4. The relationship between the cleanup costs of releases and
quantities released is not linear, but the logarithms of cleanup cost and
quantity released appear to be linearly related. The relationship,
furthermore, is significantly affected by environmental conditions at the
release site. Analysis of data on the relationship of cleanup cost to
quantity released should be useful in establishing reportable quantities; EPA
may want to establish the RQ for some substances slightly below that level at
which the cost of cleanup of a continuing release increased dramatically for
each additional pound released. If this were the case, the government's use
of Superfund monies for immediate removals would have a high probability of
being cost-effective.
5. The information available on the consequences of cleaning up
releases that would not otherwise be cleaned up is probably insufficient to
generate quantitative estimates of the benefits of the Sections 102 and 103
regulations. Several anecdotal case studies of the costs of delayed
notification have been included in Chapter 6, but data do not appear to exist
for a comprehensive analysis of benefits as envisioned by Executive Order
12291.
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
APPENDIX A
SUMMARIES OF PUBLISHED REPORTS ON
CLEANUP COSTS
This Appendix presents more detailed summaries of data on cleanup costs
discussed in Chapter 6.
• Rockwell1—In a draft report to EPA, Rockwell International's
Environmental and Energy Systems Division presented data for a variety of
mitigation techniques. The data are of two types. One type includes the
costs of typical clean-up and mitigation activities, such as groundwater
interception, excavation, revegetation, etc. The other type includes the
costs of renting pieces of equipment used in clean-up operations, labor rates,
and prices for' certain types of services required in cleanup operations. The
data were derived from 28 published references, and from price lists of
private contractors. Rockwell derived cost curves for the first type of
cost. However, each cost curve is based on only one data point, and the
curves are constructed by assuming linearity (e.g., moving twice as much earth
costs twice as much). No attempt was made to examine how costs vary by type
of substance.
Because of the large number of cost estimates provided in the Rockwell
report, we will not cite any here. It should be noted that the costs given
for what seems to be quite similar activities often diverge.
• Amson I2—In a paper written for the 1976 National Conference on
Control of Hazardous Material Spills, Jonathan Amson estimated clean-up and
damage mitigation costs for different categories of substances spilled in
water. He used physical/chemical properties identified in the Battelle report
to divide chemicals into three groups, depending on how they behave in water:
soluble substances, insoluble floaters ("oil-like" substances), and insoluble
sinking substances. Using data on the average cost of clean-up obtained from
reports submitted to EPA, he determined that the cost of clean-up for soluble
substances was about $1.00 per pound; for insoluble floaters it was $0.84 per
pound; and for insoluble sinking substances was $22.00 per pound. He
acknowledged that these estimates are quite uncertain, although he did not
1EPA, Field Guide for Priorities and Extent of Hazardous Substance
Removal, June 30, 1981 (draft).
2Jonathan E. Amson, "An Analysis of the Economic Impact of Hazardous
Substances Regulations," Proceedings of the 1976 National Conference on
Control of Hazardous Material Spills Conference, pp. 49-54.
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
A-2
present any variance estimates. Average costs tend to decrease with the size
of the spill. In addition, such factors as terrain, temperature, wind, water
conditions, the presence of natural or floating obstructions, and the
availability of clean-up equipment and personnel all influence the cost of
clean-up.
• Amson III3—In a subsequent undated paper, Amson revised his
earlier estimates of clean-up costs. He noted since those earlier estimates
were made, costs have increased by an order of magnitude or more. This
increase is attributable to the development of better, more costly removal
techniques.
• ADL**--In 1978, ADL classified releases into 11 categories
requiring generic types of responses, and developed response cost estimates
for each of these categories. The spill categories are shown in Exhibit A-l,
and the.costs in Exhibit A-2. Note that ADL assumed that costs are fully
variable with the quantity released. Virtually all the data on which these
estimates are based were also used in the above reports, particularly Amson I
and III. Thus, the ADL estimates should not be seen as an independent
confirmation of these estimates, but as another manifestation of the same
underlying data. As noted earlier, and as ADL acknowledged, the unique
characteristics of each incident can drastically affect the response costs.
• Orlemann and Higgins6--In 1974, Orlemann and Higgins reported
that the cleanup of a 1972 surfactant spill of about 100,000 pounds cost more
than $50,000. In addition, the responsible party paid the state of Ohio
$16,762 to replace the estimated 67,000 fish killed as a result of the
release. Afterwards, the firm spent $70,000 for additional spill control
facilities and procedures, including: concrete curbing, a new drainage
system, the development of internal contingency plans, and the collection of
information about the physical, chemical and biological characteristics of all
hazardous materials in use at the facility.
3Jonathan E. Amson, "The Economic Impact of Hazardous Substance
Discharge Events," mimeo.
*EPA, Environmental Emergency Response, October 1978 (draft).
5James A. Orlemann and Donald C. Higgins, "A Major Surfactant Spill in
Retrospect," Proceedings of the 1974 National Conference on Control of
Hazardous Material Spills, pp. 222-228.
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
For Water Releases
Spill Category 1 .
Spill Category 2
Spill Category 3
Spill Category 4
Spill Category 5
Spill Category 6
Spill Category 7
For Land Releases
Spill Category 8
Spill Category 9
Spill Category 10
A-3
EXHIBIT A-l
RELEASE CATEGORIES
Liquids or solids which dissolve in the water column
and do not produce hazardous gases or vapors.
As above, but hazardous gases or vapors are produced.
Floating insoluble substances which do not produce
hazardous vapors or gases.
As above, but hazardous gases and vapors are produced.
Sinking insoluble substances which do not produce
hazardous vapors or gases (i.e., boiling points greater
than ambient).
As above, but hazardous vapors or gases are produced.
Liquids or solids which are self-reactive or reactive
with water.
Liquids or solids which are self-reactive or reactive
with water.
Compressed gas releases (include releases over water).
Liquids or solids which do not boil or quickly
evaporate.
Spill Category 11 - As above, but hazardous vapors or gases are produced.
SOURCE: EPA, Environmental•Emergency Response, October 1978 (draft),
pp. 111-39 - 111-41.
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
A-4
EXHIBIT A-2
RESPONSE COST SUMMARY
Response Cost
Spill Category ($/lb)
1 1.00
2 1.23
3 1.00
4 1.23
5 11.50
6 0.50
7 1.23
8 0.16
9 0.50
10 0.16
11 0.39
SOURCE: EPA, Environmental Emergency Response, October 1978 (draft),
p. 111-47.
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
A-5
• Nye6—In 1972, Nye reported that it had cost $100,000 the
previous year to mitigate the effects of the release of 1-1/2 lbs. of Endrin,
a pesticide, into a lake. It was necessary to filter the entire lake
(approximately 20 billion gallons of water) and dispose of 6,500 cubic yards
of sediment from the bottom and sides of the lake.
• Willman, Blazevich, and Snyder7--In 1974, 265 gallons of PCBs
were released in the Seattle harbor. To mitigate the effects of the release,
it was necessary to filter the water in the harbor, to dredge contaminated
material from the bottom, and to dispose of the contaminated material. The
total cost was $150,000.
• Fullner and Crump-Wiesner8—In 1975, the application of 2-1/2
gallons of an organo-chlorine pesticide to control termites in one house
leaked into a stream. EPA and the State of Pennsylvania spent $24,500 to
filter the water in the stream and to collect and dispose of contaminated
soil. The exterminator responsible for the release spent $49,000 to move the
house.
6William 6. Nye, "The Hazardous Material Spill Experience in Shawnee
Lake Ohio—A Care History," Proceedings of the 1972 National Conference on
Control of Hazardous Materials Spills; pp. 217-219.
7James C. Willman, Joseph Blazevich, and Harold J. Snyder, "PCB Spill in
the Duwamish-Seattle, WA," Proceedings of the 1976 National Conference on
Control of Hazardous Material Spills, pp. 345-350.
®Richard W. Fuller and Hans J. Crump-Weisner, "Use of EPA's
Environmental Emergency Response Unit in a Pesticide Spill," Proceedings of
the 1976 National Conference on Control of Hazardous Material Spills, pp.
345-350.
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
APPENDIX B
A PRELIMINARY ANALYSIS OF THE PENALTY, LIABILITY, AND
FINANCIAL RESPONSIBILITY PROVISIONS OF CERCLA
1. INTRODUCTION
In this report, we are concerned with the impacts of proposed regulations
under Sections 102 and 103 of CERCLA. The regulations will deal with
designating substances as "hazardous substances" for purposes of CERCLA and
with establishing what quantity of a particular substance will trigger
reporting requirements when that substance is released. These proposed
regulations may affect both the liability of certain persons who deal with
substances designated as "hazardous" and the financial responsibility which
such persons are required to demonstrate. The statute appears to indicate
that when a substance is designated "hazardous," its release or threatened
release subjects the person responsible to liability provisions that do not
apply to substances not designated "hazardous." Similarly, the establishment
of a reportable quantity for a particular hazardous substance subjects a
person responsible for a release of that substance to certain penalties for
failure to provide adequate notice of a reportable release. The penalty,
liability, and financial responsibility provisions that are triggered by
Sections 102 and 103 reporting and designation requirements are analyzed in
this appendix.
The statutory provisions contained in CERCLA impose costs on the persons
to whom they apply. The direct costs are those incurred by operation of the
Act -- for example, the payment of a civil penalty for failure to take some
action. Other costs, called "induced" costs, are incurred by persons in an
effort to avoid the direct costs of the Act. Thus, a person may take
additional precautions to avoid releasing hazardous substances and being
required to reimburse EPA for actions the Agency may take to cleanup the
released substances.
The analysis in this appendix has been used to help evaluate the economic
effects of the Sections 102 and 103 regulations. The analysis focuses on
three CERCLA sections:
• Section 103 imposes penalties for failure to
provide notification when a reportable quantity
(defined under Section 102 and its regulations) of a
hazardous substance is released.
• • Section 107 imposes liability for response costs
and natural resource damages and establishes limits to
such liability.
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
B-2
• Section 108 requires demonstration and maintenance
of financial responsibility by certain vessels and
facilities. Failure to comply with the requirements
of Section 108 may subject a person to penalties '
imposed under Section 109.
The practical effect of the requirements of these sections is summarized below
in Section 2 for three situations:
hazardous substance releases below the reportable
quantity;
hazardous substance releases at or above the
reportable quantity; and
pollutant or contaminant releases.
Following discussion of these situations in Section 2, the liability and
liability limitation provisions of Section 107 are analyzed (Sections 3 and 4,
respectively), followed by a discussion of Section 103 notification penalties
(Section 5), Section 108 financial responsibility requirements (Section 6),
and other sources of liability (Section 7).
2. EFFECT OF LIABILITY AND FINANCIAL RESPONSIBILITY IN THREE CASES
One of the most important characteristics of the liability and financial
responsibility regime established under CERCLA is that it appears to apply
only to designated hazardous substances. Although Section 104(a)(1)
authorizes response to the release or substantial threat of release of a
"pollutant or contaminant" that may present an imminent and substantial danger
to the public health or welfare, the Act appears not to authorize imposition
of liability on nor require evidence of financial responsibility from owners
and operators of vessels or facilities dealing only with "pollutants or
contaminants." There are three broad cases discussed below:
the situation where a hazardous substance release is
below the reportable quantity;
the situation where a hazardous substance release is
at or above the reportable quantity; and
the situation where there is a release of pollutants
or contaminants.
For each of these three cases, there are three major issues:
• Is there responsibility to report the release?
• Is there responsibility to take response action in
regard to the release?
November 2 Draft * * * Do not quote or cite 1*%^
lUr INCORPORATED
-------
B-3
• Is there liability for response costs and natural
resource damages?
These cases and issues are discussed below. No final answer, however, can be
given here to these questions since to some degree they depend on EPA's own
interpretation of its statutory mandate. The analysis presented here is
intended to assist the Agency's staff in reaching a conclusion on the exact
meaning of the statutory language, not to interpret the law on EPA's behalf.
2.1 Hazardous Substance Releases Below Reportable Quantities
There will be releases of designated hazardous substances that are
less than the reportable quantity for that substance. The responsibilities
and related liabilities in this situation are discussed in the following
paragraphs.
Is there a responsibility to report the release? Section 103(a) imposes
a reporting requirement if a designated hazardous substance is released in an
amount equal to or greater than the reportable quantity set Tinder Section
102. If the reportable quantity is not released, the owner or operator is not
required to notify the National Response Center of the release.1 There also
will be no penalties imposed for failure to report the release. One
unresolved issue is what reporting responsibilities exist when it is not known
how much of a substance was released.
Is there responsibility to take response action in regard to the release?
CERCLA does not clearly and directly impose an affirmative duty on persons to
take response actions for releases or threatened releases for which they are
responsible. Section 104 gives the President authority to take response
action. This authority is conditioned on there being no appropriate response
initiated by the responsible party. This, however, does not require such
party to respond. Section 106, on the other hand, gives authority to the
President to seek judicial relief to abate an imminent and substantial danger
to public health, welfare, or the environment due to the release or threatened
release of a hazardous substance from a facility. In addition, the President
may issue orders necessary to protect public health, welfare, or the
environment when such release or threatened release occurs. Willful failure
to comply with these orders may subject a person to a fine of up to $5,000/day
of non-compliance (Section 106(b)) and punitive damages (Section 107(c)(3)).
If a hazardous substance is released in an amount lower them a reportable
quantity, it is still possible that the release could present an imminent and
substantial danger. (It is uncertain how the government will find out about a
1Under Section 311 of the Clean Water Act, the amount of a release is
added over a 24-hour period to determine whether a reportable quantity has
been discharged. No comparable approach has been formally established under
CERCLA.
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
B-4
non-reported release in order to take the abatement actions authorized in
Section 106.) The mere fact that less than a reportable quantity has been
released does not excuse the responsible party from compliance with Section
106 judicial or executive orders to take response actions.
Is there liability for response costs and natural resource damages? A
person can be liable for response costs and damages to natural resources even
though the amount released was not a reportable quantity. As mentioned above,
the statute does not condition the President's response authority on the
release of a reportable quantity. Rather, response is conditioned on the
release or threatened release of any quantity of hazardous substance (see
Section 104(a)(1)(A)) and consistency with the National Contingency Plan.
Moreover, the authority to seek judicial relief is contingent on the existence
of an imminent and substantial danger, and the authority to issue orders is
contingent on the necessity to protect public health, welfare, and the .
environment. Nowhere are these various actions directly conditioned on the
amount of hazardous substance released. Liability for response costs, natural
resource damages, and other penalties is based on authorized government
activity not inconsistent with the National Contingency Plan. It is
conceivable (though it is not yet possible to determine how probable) that a
release of a hazardous substance in less than a reportable quantity will
involve government response or orders that responsible parties take the
necessary reponse action. In either case, it seems that CERCLA liability can
attend the release of less than a reportable quantity.
2.2 Hazardous Substance Releases At or Above the Reportable Quantity
This case is the easiest to resolve. The statute appears unambiguous
about the treatment of releases of reportable quantities.
Is there a responsibility to report the release? If a release of a
hazardous substance equals or exceeds the reportable quantity, the owner or
operator of the vessel or facility from which the release comes must
immediately notify the National Response Center.
Is there responsibility to take response action in regard to the release?
Although CERCLA does not seem directly to impose an affirmative duty on
parties to take response action when they are responsible for a release, it
seems clear from the National Contingency Plan (NCP) that there is a de
facto duty to do so. The NCP indicates that its strategy is to attempt to
have responsible parties take the necessary response actions. In many
instances, then, failure by the responsible parties to initiate response may
result in enforcement action against those parties. If Fund-financed response
is taken, of course, the responsible parties still are liable for the response
costs, any natural resource damage, and any penalties that result from the
parties' failure to take any ordered action.
It is important to note that CERCLA is only one of several potential
sources of liability for a person who releases hazardous substances (see the
discussion in Section 7, below). CERCLA, however, may be the source most
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
B-5
likely to be used if it is an easier vehicle for imposing liability than other
laws.
Is there liability for response costs and natural resource damages? The.
persons responsible for the release of a reportable quantity of a hazardous
substance are liable for response costs incurred by federal or state
governments, response costs incurred by any other person consistent with the
NCPj and any natural resource damages resulting from the release.
2.3 Releases, of Pollutants or Contaminants
This case is different from the two preceeding cases. The earlier
cases were concerned with the issue of reportable quantities, whereas this
case deals with the issue of designation. Under CERCLA, Section 102, EPA is
given authority to designate substances as "hazardous" if they are not already
so designated under the statutory authorities listed in Section 101(14). In
general, if a harmful substance is not designated "hazardous" under CERCLA, it
is considered a "pollutant or contaminant"; the definition of "pollutant or
contaminant" (Section 104(a)(2)) specifically excludes petroleum products,
natural gas, and synthetic gas. The distinction between a hazardous substance
and a pollutant/contaminant is important from the standpoint of liability and
financial responsibility. For example, the financial responsibility
requirements for facilities appear to relate only to hazardous substances--a
facility dealing only with pollutants/contaminants seems not to be required to
produce evidence of financial responsibility. As the following discussion
indicates, pollutants or contaminants appear to receive lenient treatment
under CERCLA.
Is there responsibility to report the release? There is no reporting
requirement for the owner or operator of a vessel or facility from which there
is a release only of pollutants or contaminants. Section 103(a) imposes
reporting requirements only for hazardous substance releases. Consequently,
it appears that no penalty or liability under CERCLA can follow from a failure
to report the release of a pollutant or contaminant. If the pollutant or
contaminant is commingled with hazardous substances, however, the presence of
such pollutants or contaminants should be no excuse from the reporting
requirement for the released hazardous substances.
Is there reponsibility to take response action in regard to the release?
As noted above, CERCLA itself does not appear directly to impose
responsibility for taking cleanup actions oh responsible parties. Instead,
the Act gives the President authority to order such actions and provides for
penalties if there is a failure to comply with such a Presidential order. In
practice, then, there may often be a de facto requirement that responsible
parties take response actions. The authority granted to the President in
Section 106(a), however, extends only to the release or threatened release of
hazardous substances. The President appears not to be authorized to issue
such orders regarding pollutants or contaminants, although the final sentence
of Section 106(a) might be interpreted to confer on the President authority to
order response to a release of pollutants or contaminants. That sentence
provides:
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
B-6
The President may also, after notice to the affected State,
take other action under this section including, but not
limited to, issuing such orders as may be necessary to
protect public health and welfare and the environment.
In this sentence, there is no restrictive reference to hazardous substances.
Thus, application to pollutant/contamination releases could be imputed to this
provision. Such an interpretation, however, may be unlikely. The President
is authorized to take "other action under this section" (emphasis added).
The first clauses of Section 106(a) describe when action under Section 106 may
be taken:
In addition to any other action taken by a State or local
government, when the President determines that there may be
an imminent and substantial endangerment to the public
health or welfare or the environment because of an
actual or threatened release of a hazardous
substance from a facility...(emphasis added)
In light of the language in the first part of Section 106(a), the subsequent
failure to limit specifically the President's authority to hazardous substance
releases does not seem to extend his authority to the release of pollutants or
contaminants under this section.
Is there liability for response costs and natural resource damage? One
of the interesting aspects of CERCLA is that it includes authority to respond
to the releases of certain pollutants or contaminants, but it does not appear
to include authority to recover these response costs. Section 104(a)(1)
authorizes response to an actual or threatened release of pollutants or
contaminants if the release presents an imminent and substantial danger to
public health or welfare (note that danger to the environment is not
included). Section 107(a) imposes liability for response costs and natural
resource damages, but it appears to do so only in regard to the release or
threatened release of hazardous substances. The government, therefore, can
respond to pollutant/contaminant releases or threatened releases; but, under
CERCLA, it may not always be able to recover those costs or damages caused to
natural resources.
This result seems anomalous, and it would be if CERCLA were the only
relevant authority. As discussed below in Section 7, however, there are a
number of other authorities under which liability may be imposed. Thus,
although CERCLA appears not to be useful for recovering costs or damages
related to pollutant/contaminant releases, recovery may be easier under other
legal authority.
3. SECTION 107 — LIABILITY PROVISIONS
Section 107 contains the liability provisions of CERCLA which also include
limits on the amount of liability to which a person can be subjected. Limits
to liability are discussed in Section 4 below. In general, liability is
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
B-7
imposed for two things: (1) the cost of responding to a release or threatened
release of a hazardous substance, and (2) the damages such releases inflict on
natural resources owned by the Federal Government or the States. Unlike the
precursor Senate bill, CERCLA has no provision for compensating third party
damages -- personal and property injuries.
There are three situations in which Section 107 imposes liability:
(1) There is a release or threatened release of hazardous
substances which causes response costs to be incurred,
or there is a release that damages natural resources
(Section 107(a) and (b));
(2) A person refuses to comply with a Presidential order
tinder Sections 104 or 106 (Section 107(c)(3)); and
(3) A person is grossly negligent or commits willful
misconduct in assisting a response effort (Section
i07(d)).
These provisions are summarized in Exhibit B-l and are discussed below in
detail.
3.1 Section 107(a) and (b)
Section 107(a) is the major liability provision of CERCLA. It
imposes liability for a release or threatened release of hazardous substances
that causes response costs to be incurred, or releases that damage natural
resources.2 There are four categories of persons who are potentially liable
for a release or threatened release:
* The owner/operator of the vessel from which the
release or threatened release comes;
* Any person who owned/operated a facility at the time
hazardous substances were disposed of at the facility;
* Any person who owned or possessed hazardous
substances and arranged, directly or through a
transporter, for their disposal or treatment at a
facility owned or operated by another party or entity,
if those hazardous substances are still contained at
that facility and there is a release or threatened
release of hazardous substances from that facility; and
2As defined in Section 101(16), "natural resources" covered by CERCLA
are those "belonging to, managed by, held in trust by, appertaining to, or
otherwise controlled by" the federal, state or local, or any foreign
government.
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
EXHIBIT B-l
LIABILITY PROVISIONS IN CERCLA SECTION 107
Liable Party
To Whom Liable Reason for Imposing Liability Defenses to Liability Under CERCLA Maximum Liability
(1) Owner/operator of a vessel
(otherwise subject to U.S.
jurisdiction) or facility.
(2) Any person who owned or
operated a facility at the
time any hazardous substances
were disposed of at such
facility.
(3) Any person who arranges
for disposal or treatment, or
who arranges with a trans-
porter for transport for
disposal or treatment, of
hazardous substances owned
or possessed by that person.
The disposal or treatment is
to be by another party or
entity at a facility owned or
operated by such party or
entity and containing such
hazardous substances.
(4) Any person who both
accepts or accepted hazar-
dous substances for trans-
port and selects disposal
or treatment facilities to
which to transport such
substances.
(1) U.S. Govern-
ment or State
(2) Any other
person
(1) Release or threatened
release of hazardous sub-
stances from these vessels or
facilities which cause
response costs to be
incurred.
(2) Release causes injury
to, destruction of, or loss
of natural resources.
(1) Release/threat of release and
resulting damages were caused
solely by an act of God.
(2) Release/threat of release and
resulting damages were caused
solely by an act of war.
(3) Release/threat of release and
resulting damages were caused
solely by the act/omission of a
third party, or someone other than
a third party whose act or omis-
sion occurs in connection with a
direct or indirect contractual
relationship.
(4) Any combination of (l)-(3).
(5) Damages to natural resources
and release causing the damages
occurred wholly before CERCLA*s
passage (Section 107(f)).
(6) The release is the application
of a pesticide product registered
under FIFRA (Section 107(1)).
(7) The release Is a federally
permitted release (Section 107(j)).
(8) Regarding liability for natural
resource damages, liability can be
avoided if the potentially liable
party can demonstrate: (1) the
damages in question had been speci-
fically identified as an irrevers-
ible and irretrievable commitment
of natural resources in an EIS or
some comparable environmental analysis;
(11) the decision to license or
permit this facility or project author-
ized this commitment; and (ill) the
facility was otherwise operating
within the terms of its permit or
license (Section 107(f)).
(1) All costs of
removal or remedial
actions not inconsi-
tent with the NCP.
(2) Other necessary
costs of response
consistent with the
NCP.
(3) Damages for
injury to, destruc-
tion of, or loss of
natural resources,
including the
reasonable costs of
damage assessment.
November 2 Draft
Do not quote or cite
-------
EXHIBIT B—1
LIABILITY PROVISIONS IN CERCLA SECTION 107
(Continued)
Section
Liable Party
Any person liable for a
release or threatened release
of hazardous substances.
To Whom Liable Reason for Imposing Liability Defenses to Liability Under CERCLA Maximum Liability
U.S. Government
Failure, without sufficient
cause, to provide removal/
remedial action.upon
President's order pursuant
to Sections 104 or 106
Same defenses as above.
Treble punitive
damages (based on
amount of cost
incurred by the
Fund as a result
of failure to take
the needed action) .
Also, a penalty of
$5,000/day for each
day noncompliance
continues (Section
106(b)) .
Person rendering care, assis-
tance, or advice in accordance
with the NCP or at the OSC's
direction with regard to an
incident creating danger to
public health, welfare, or
environment as the result of a
release or threatened release
of hazardous substances.
U.S. Government or
State Government
Gross negligence ("reckless,
willful, or wanton mis-
conduct") , or intentional
misconduct.
Absence of such negligence or
misconduct.
Damages as a result
of actions taken or
omitted in the
course of rendering
care, assistance,
or advice.
November 2 Draft * * * Do not quote or cite
o
3
fit
O
-------
B-10
• Any person who both accepts (or accepted) hazardous
substances for transport and selects (or selected)
disposal or treatment facilities to which to transport
these hazardous substances and from which there is a
release or threatened release of hazardous substances.
The maximum liability (but see Section 4 below which discusses limitations on
liability) to which such persons can be subjected has three components:
• all removal and remedial action costs, consistent
with the National Contingency Plan (NCP), incurred by
the Federal Government or States;
• any other "necessary" response costs incurred by
other persons, consistent with the NCP; and
• natural resource damages.
The liability for the first and third components is owed to the government
entity involved; liability for the second component is owed to the person,
other than the government entity, who incurs such costs.
Section 107 provides many defenses to liability, and these are listed in
Exhibit B-l, above. Section 107(b), however, provides the four major
defenses. No liability is imposed if the release or threat of release of
hazardous substances and any resulting damages were caused solely by:
• an act of God;
• an act of war;
• an act or omission of a third party other them an
employee or agent of the defendant, or other than a
third party whose act or omission occurs in connection
with a direct or indirect contractual relation (except
where the contractual relation exists only because of
a published tariff and acceptance of carriage by a
common carrier by rail);' and
• any combination of the preceding.
3To use this defense, the person asserting it must alsq establish by a
preponderance of the evidence: (i) that he exercised due care with respect to
the hazardous substance concerned, taking into consideration the
characteristics of these substances, in light of all relevant facts and
circumstances; and (ii) that he took precautions against the foreseeable acts
or omissions of such third party and the consequences that could foreseeably
result from such acts or omissions.
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
B-ll
One further defense to liability should be noted here — the federally
permitted release. These are releases that are allowed to occur under other
federal or state law (e.g., discharges that comply with a Section 402 permit
under the Federal Water Pollution Control Act). Section 107(j) excludes from
CERCLA liability any response costs or damages resulting from such a release.
There are many issues that arise under Section 107(a) and (b). They
include the following issues which will be discussed below in detail:
• How are pollutants and contaminants treated under
this section?
• Is a past owner of a facility liable only for the
release of substances that were disposed of while such
person owned the facility?
• Is a generator who arranges for his hazardous waste
to be disposed or treated at a facility owned by a
third party subject to liability for any release or
threatened release of hazardous substances from that
facility, or only for the release or threatened
release of those hazardous substances he arranged to
have at that facility?
• To what degree are transporters liable for releases
from facilities to which they have carried hazardous
substances for disposal or treatment?
• What is a "necessary" response by a non-governmental
entity?
• What standard of proof must be met in asserting the
four defenses provided under Section 107(b)?
How are pollutants and contaminants treated under this section? One
important feature of Section 107(a) is that it refers explicitly only to
releases of hazardous substances. When a hazardous substance is released,
liability for certain costs and damages resulting from the release are
assigned to persons who, under specified circumstances, are associated with
the substance's treatment, storage, transport, or disposal. There is no
mention of pollutants or contaminants throughout this section.
Failure to include specific mention of pollutants or contaminants in
CERCLA's liability provisions may have significant policy implications. There
are, however, two divergent readings of these possible implications:
• One reading is that persons who handle pollutants or
contaminants are favored under the law over those who
handle designated hazardous substances. Since no
reference is made to pollutant or contaminants in the
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
B-12
liability provisions, there is presumably no liability
under CERCLA for costs or damages incurred by their
release, thus freeing handlers of undesignated
pollutants or contaminants from at least one potential
source of liability. As discussed in Section 7,
below, these persons still are subject to liability
from other legal sources such as state statutes and
common law.
* A second reading emphasizes the limits to liability
that, together with the assignment of liability, are
established by Section 107 of CERCLA.k Persons who
handle undesignated pollutants or contaminants cannot
avail themselves of these limits to liability, yet
they may find themselves liable for costs or damages
incurred by a release under other legal authority (see
the discussion in Section 7). Thus, the availability
of limited liability is an incentive for transporters,
storers, treaters, or disposers of.pollutants or
contaminants to seek the addition of substances they
handle to the list of hazardous substances. Note,
however, that there also may be limits to liability
under these other legal authorities.
The second reading, however, seems flawed because it is premised on the
supposed benefit of limited liability. Under CERCLA, the liability limits
apply only to the liability imposed by CERCLA;5 if liability arises from
another source (e.g., the common law), the limitations provided in Section 107
are of no effect. As noted above, it is possible that these other sources of
liability will themselves have limits (e.g., liability under common law would
be limited to some degree by the foreseeability of the damages resulting from
a release), but such limitations must be determined independently of CERCLA.
Because CERCLA's limitation on liability does not cover the other sources of
liability, there seems to be no reason for a person to seek the designation of
a pollutant or contaminant as "hazardous." To seek designation under the
reasoning of the second reading would be to seek additional liability (albeit
limited) where before there had been no liability; no rational actor will make
this decision.
Is the past owner of a facility liable only for the release of substances
that were disposed of while such person owned the facility? Section
107(a)(2) imposes liability on 'any person who at the time of disposal of any
''For example, the liability of owners or operators of vessels that carry
hazardous substances cannot exceed $300 per gross ton or $5 million, whichever
is greater, for each release.
BSection 114(a) specifically allows states to impose additional
liability for hazardous substance releases.
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
B-13
hazardous substance owned or operated any facility at which such hazardous
substances were disposed of." The liability is imposed when response costs
are incurred as the result of a release or threatened release "of a hazardous
substance" from that facility. It is conceivable that a release from a
facility could involve hazardous substances that were not disposed of during a
past owner's tenure at the facility. The language of Section 107(a), however,
might be interpreted to impose blanket liability -- as long as a person owned
a facility at the time that some hazardous substance was disposed of at that
facility, he can be liable for the response costs occasioned by the release or
threatened release of any hazardous substance from that facility. Section
107(b), however, seems to provide a defense if liability is sought to be
imposed in such a situation. If the potentially liable party can
demonstrate6 that the release or'threatened release was caused solely by the
act or omission of a third party other than an employee, agent, or someone
else standing in a contractual relationship with the potentially liable
person, liability can be avoided.7 A past owner whose activities did not
include disposal of hazardous substances involved in a particular release or
threatened release may be successful in asserting this defense.
Is a generator who arranges for his hazardous waste to be disposed of or
treated at a facility owned by a third party subject to liability for any
release or threatened release of hazardous substances from that facility, or
only for the release or threatened release of those hazardous substances he
arranged to have at that facility? This issue is a variation of the
immediately preceding issue. Just because a person has arranged to have
hazardous substances he owns disposed of at a facility, does he become liable
for any release or threatened release of any hazardous substance from that
facility? Section 107(a) might be interpreted to permit this; if so, the
third party defense of Section 107(b)(3) may be available to a person in this
situation. It is interesting to note that a person who arranges for the
storage of his hazardous substances at a facility owned by a third person
does not seem to be liable for the release of these hazardous substances from
such facility.
To what degree are transporters liable for releases from facilities to
which they have carried hazardous substances for disposal or treatment? The
liability provision applies to transporters if they selected the facility to
which they carried hazardous substances for disposal or treatment. Thus,
transporters appear not to be liable for the release/threatened release from a
facility to which they transported hazardous substances, if the choice of the
facility was not theirs. As with earlier examples, transporters might be
liable for any release from a facility to which they have chosen to transport
6See discussion below on "What standard of proof must be met in
asserting the four defenses provided under Section 107(b)?"
'See footnote 3 above for additional conditions for asserting this
defense.
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
B-14
hazardous substances, whether or not the substances released are traceable to
the transporter. As in those earlier examples, however, a transporter may be
able to assert the third party defense provided in Section 107(b)(3).
What is a "necessary" response by a non-governmental entity? Section
107(a)(B) includes as a component of liability "any other necessary costs of
response incurred by any other person consistent with the NCP." The "any
other person" seems to refer to entities other than the U.S. government or a
state. The NCP, 1510.70(b)(1) and (2), provides for Fund payment of
"non-federal costs" for response actions that have been approved in advance by
EPA. "Non-federal costs" could be interpreted to include non-governmental
costs. It is not clear, however, that this provision of the NCP extends to
non- governmental costs. If it does, "necessary" costs might be equated with
costs that have been authorized in advance.
What standard of proof must be met in asserting the four defenses
provided under Section 107(b)? A person otherwise liable must establish by a
"preponderance of the evidence" that the release or threat of release of
hazardous substances and resulting damages are subject to the defenses set out
in 107(b). This evidentiary standard is easier to satisfy than "beyond all
reasonable doubt" and traditionally is the standard used in civil cases. A
noted authority suggests the best interpretation of this standard "to be proof
which leads the jury to find that the existence of the contested fact is more
probable than its nonexistence"8 -- here, that the defense asserted is
correct.
3.2 Section 107(c)(3)
Section 107(c)(3) authorizes up to treble punitive damages for
failure, without sufficient cause, to provide removal or remedial action when
ordered by the President to do so under Sections .104 and 106. The order
referred to under Section 104 seems to be the authorization in 104(e)(1) for
providing access to certain information. The more substantive authority is
under Section 106(a) where the President is given authority to issue orders
"necessary to protect public health and welfare and the environment." The
legislative history suggests that treble damages are authorized only when
"See McCormick's Handbook on the Law of Evidence (West: St. Paul,
1972, 2nd Edition), Section 339.
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
B-15
there is failure to take action necessitated by an imminent and substantial
danger.9 The persons who are potentially liable for these punitive damages
are those same persons who are liable under Section 107(a) for a release or
threatened release of a hazardous substance. The punitive damages are
measured by the amount of costs incurred by the Fund as a result of the
person's failure to take the ordered action.
In addition to the punitive damages authorized under this section, Section
106(b) authorizes a $5,000 penalty for each day that a person fails to comply
with a Presidential order issued under Section 106(a). The defenses to this
penalty are the same defenses that a potentially liable party can assert
against liability imposed under Section 107(a).
One issue raised by Section 107(c)(3) is what constitutes "sufficient
cause" to fail to provide the removal or remedial action ordered by the
President? The phrase is not elaborated upon further in the Act. The
legislative history suggests three situations that would fit within this
exception and thereby shield the person from having to pay punitive damages:
(1) the person does not have the capability (technical or financial) to comply,
with the order; (2) the person has a valid defense to assert against liability
under Section 107(a); or (3) the President's order was not valid.10
9During Senate consideration of the bill that became CERCLA, Senator
Simpson offered the following interpretation of what Section 106 Presidential
orders encompass and, consequently, when treble damages could be imposed:
First, I read the second sentence in Section 106(a) --
which grants the President authority to issue orders necessary
to protect the public health and welfare and the environment —
to be authority that may be exercised when the President
believes there is an imminent and substantial endangerment to
the public health or welfare or environment because of an actual
or threatened release of a hazardous substance.
Thus, I consider the language of the first sentence of
Section 106(a), limiting action to cases where the President
determines there is an imminent and substantial endangerment,
and the first sentence of Section 106(c), requiring the
publication of guidelines for using emergency response
authority, to apply only to the President's authority under
Section 106 to issue orders.
Senator Stafford, one of the primary sponsors of the bill, agreed with the
interpretation. Congressional Record, November 24, 1980, p. S 15008.
10See the remarks for Senator Stafford, Congressional Record, November
24, 1980, p. S 15008.
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
B-16
3.3 Section 107(d)
Section.107(d) is a "good Samaritan" provision -- it protects from
liability a person who, without being required to do so, is "rendering care,
assistance, or advice in accordance with the national contingency plan or at
the direction of an onscene coordinator." The protection from liability,
however, does not extend to "liability for damages as the result of gross
negligence or intentional misconduct."11 Technically, this section does not
provide independent authority for imposing liability, it merely emphasizes
that the section .does not excuse liability for. misbehavior that normally
exposes a person to liability.
In the following section, the liability limitation provisions of Section
107 are discussed.
4. LIMITS TO LIABILITY
The .liability imposed under CERCLA is generally limited. In most
instances, a person's liability for response costs and natural resource
damages is subject to a statutory ceiling. The limits of liability apply only
to liability that is imposed by CERCLA; they do not apply to liability imposed
under other law, including common law.
Section 107(c) establishes both the maximum liability to which a person is
exposed under CERCLA and the situations in which the liability limits do not
apply. Section 107(k)(l) provides limited liability through the Post-closure
Liability Trust Fund. These sections are summarized in Exhibit B-2 and
discussed in detail below.
4.1 Section 107(c)(1)
Liability limits for vessels and facilities are established in this
section. There, are four categories of liability limits:
• Vessels carrying hazardous substances as cargo or
residue — maximum liability is the greater of
$300/gross ton or $5 million;
• Any other vessel -- maximum liability is the greater
of $300/gross ton or $500,000;
• Motor vehicles, aircraft, pipelines, or rolling
stock -- maximum liability is $50 million unless the
President, by regulation, establishes some lower limit
which can be as low as $5 million. This lower limit
cannot be less than $8 million if hazardous substances
llSection 107(d) provides that for purposes of this section, gross
negligence is "reckless, willful, or wanton misconduct."
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
EXHIBIT B-2
LIABILITY LIMITATION PROVISIONS IN CERCLA SECTION 10 7
Liable Party
To Whom Li able
Maximum Liability
Any person liable under
Section 107
Those persons indicated
in Section 107
(1) For a vessel carrying hazardous substance as cargo or residue — the
greater of $300/gross ton or $5 million.
(2) Any other vessel — the greater of $300/gross ton or $500,000.
(3) Motor vehicle, aircraft, pipeline, or rolling stock — $50 million or a
lesser amount, down to $5 million, that the President may establish by
regulation; except, the lower limit is $8 million for releases into the
navigable waters of hazardous substances designated under Section
311(b) (2) (A) of the Federal Hater Pollution Control Act.
(4) Any other facility — total of all response costs plus $50 million for
any damages under Title I of the Act.
Any person liable under
Section 107
Those persons indicated
in Section 107
Owner/operator of a hazardous
waste disposal facility per-
mitted under Subtitle C of
the Solid Waste Disposal
Act (SWDA)
Those persons indicated
under Section 107 of
CERCLA, or under any
other law
The liability limitations provided by Section 107(c)(1) do not apply
in the following circumstances!
(1) The release or threat of release was the result of willful misconduct or
willful negligence within the liable person's privity or knowledge; or
(2) The primary cause of a release or threat of release was a violation
(within the liable party's privity or knowledge) of applicable safety,
construction, or operating Federal standards or regulations; or
(3) The liable person fails or refuses to provide all reasonable cooperation
or assistance requested by the responsible public official regarding
response activities under the NCP with regard to regulated carriers
subject to 49 (f.S.C. or vessels subject to 33 or 46 U.S.C.
Liability is transferred to the Post-Closure Liability Fund established under
Section 232 of CERCLA when both of the following conditions are met:
(1) Compliance with requirements of subtitle C of SWDA and its regulations
which may affect post-closure performance; and
(2) The facility has been closed in accordance with these regulations and
the conditions of the permit, and the facility and surrounding area have
been monitored as required by regulations and permit conditions for up
to 5 years after closure in order to demonstrate "no substantial
likelihood that any migration off-site or release from confinement of
any hazardous substance or other risk to public health or welfare will
occur."
November 2 Draft
Do not quote or cite
-------
B-18
designated under Section 311(b)(2)(A) of the Federal Water
Pollution Control Act are released into navigable waters; and
• Any other facility — maximum liability is the total of all
response costs plus $50 million for "any damages under this
title."
One issue raised by this section is what response costs are covered by
the fourth category of limitation. In other sections of the Act, the
reference is usually to response costs "consistent with the national
contingency plan." This phrase does not appear in Section 107(c)(1) and might
be interpreted to leave a facility owner or operator liable for costs of
response actions that are not authorized by the NCP. This, however, does not
appear to be the proper interpretation. Section 107(c)(1) refers to
"liability under this section" (i.e., Section 107); and Section 107(a) limits
liability for response costs to those costs consistent with the NCP. Note,
also, that the fourth category refers to "damages under this title." The only
damages.for which liability is imposed under CERCLA are natural resource
damages.'
4.2 Section 107(c)(2)
Limited liability is not available in some situations. Section
107(c)(2) sets out three circumstances which disqualify a person from using
the iiability limitations:
• The release or threat of release was the result of
willful misconduct or willful negligence within the
liable person's privity or knowledge;
• The primary cause of a release or threat of release
was a violation of applicable safety, construction, or
operating federal standards or regulations, and such
violation was within the liable person's privity or
knowledge; or
• The liable person fails or refuses to provide all
reasonable cooperation or assistance requested by the
responsible public official regarding response
activities under the NCP with regard to regulated
carriers subject to 49 U.S.C. or vessels subject to 33
or 46 U.S.C.
The first circumstance involves concepts drawn from tort law -- "willful
misconduct," "willful negligence," and "privity or knowledge." The first two
terms are essentially interchangeable. A noted authority in tort law has
written:
The usual meaning assigned to "willful" ... is that
the actor has intentionally done an act of an unreasonable
character in disregard of a risk known to him or so obvious
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
B-19
that he must be taken to have been aware of it, and so
great as to make it highly probable that harm would
follow. It usually is accompanied by a conscious
indifference to the consequences.
"Willful" ... conduct tends to take on the aspect of
highly unreasonable conduct, or an extreme departure from
ordinary care, in a situation where a high degree of danger
is apparent.
[To be considered "willful," misconduct or negligence]
must be more than any mere mistake resulting from
inexperience, excitement, or confusion, and more than mere
thoughtlessness or inadvertence, or simple inattention.12
The following passage seems to indicate that Congress intended the Section
107(c)(2) disqualifications to be interpreted the same as a similar provision
in Section 311(f) of the Clean Water Act:
In providing that limits of liability may be
inapplicable under certain circumstances, section 107(c)(2)
borrows language directly from section 311(f) of the Clean
Water Act. In so doing, the terms "willful negligence1 and
"willful misconduct" were inadvertently reversed. The
intent of Congress is to provide the same inapplicability
to liability limits as that provided in the Clean Water
Act. The inadvertent order reversal is not meant to impart
any different meaning then [sic] that contained in section
311(f) of the Clean Water Act.13
The phrase "privity or knowledge" is defined in Black's Law Dictionary as:
actual knowledge of the things causing or contributing to
the loss, or knowledge or means of knowledge of a condition
of things likely to produce or contribute to the loss
without adopting proper means to prevent it.
The first circumstance for which liability limits do not apply might be
summarized as when the release or threatened release is the result of reckless
behavior that the responsible person actually knew of or had the means of
discovering.
1Grosser, The Law of Torts (4th ed.) (St. Paul: West Publishing Co.,
1971), p. 185, footnotes omitted.
13Remarks of Congressman Florio, Congressional Record, December 3,
1980, p. H 11789.
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
B-20
The second circumstance also is conditioned on being within the person's
privity or knowledge. The circumstance covered is the violation of some
standard or regulation dealing with safety, construction, or operations.14
These are standards that, if followed, would have helped prevent a release.
The third circumstance seems, on its face, to be limited to regulated
carriers subject to Title 49 of the United States Code (dealing with
transportation--rail, pipeline, air, water, road) and vessels subject to
Titles 33 (dealing with navigation and navigable waters) and 46 (dealing with
shipping) of the Code. The Act's language is:
or (B) such person fails or refuses to provide all
reasonable cooperation and assistance requested by a
responsible public official in connection with response
activities under the national contingency plan with respect
to regulated carriers subject to the provisions of title 49
of the United States Code or vessels subject to the
provisions of title 33 or 46 of the United States Code,
subparagraph (A)(ii) of this paragraph shall be deemed to
refer to Federal standards or regulations. (42 USCA
9607(c)(2)(B))
The bill voted upon by the Senate and recorded in the Congressional Record
presents a slightly different version of this provision:
or (B) such person fails or refuses to provide all
reasonable cooperation and assistance requested by a
responsible public official in connection with response
activities under the national contingency plan. With
respect to regulated carriers subject to the provisions of
title 49 of the United States Code or vessels subject to
the provisions of title 33 or 46 or the United States Code,
subparagraph (A)(ii) of this paragraph shall be deemed to
refer to Federal standards or regulations.15
1"Although Section 107(c)(2)(A)(ii) does not specify that the
"applicable safety, construction, or operating standard or regulations" refer
only to federal standards, that may be the intention of Congress. Senator
Helms noted, with Senator Stafford's concurrence, that "under section
107(c)(2), the limits on liability established by section 107(c)(1) do not
apply if the primary cause of the release of a hazardous substance was a
violation, within the privity or knowledge of an owner, operator, or other
responsible person, of an applicable Federal safety, construction, or
operating standard or regulation. (Congressional Record, November 24, 1980,
p. S 15007.)
15 Congressional Record, November 24, 1980, p. S 14993.
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
B-21
Although the words are the same, the period between "national contingency
plan" and "with respect to" has been deleted in the law. The deletion seems
to change the meaning of the provision by limiting its applicability to
regulated carriers and vessels -- many facilities that might have been
included under the bill as recorded in the Congressional Record may now be'
excluded by the law. There is no indication whether this omission was
intentional; two factors suggest that it was not:
• In summarizing the modifications to the compromise
bill (which, along with the modifications, became
CERCLA), Senator Stafford refers to the language
beginning "with respect to..." in the following
manner: "Another [modification] clarifies the
language for breaching liability limits for carriers
to deal with releases where the primary cause was a
violation of Federal standards on page 42.1116 This
summary makes sense in reference to the version of
the bill in the Congressional Record; it seems to
make less sense in reference to the provision in the
Act.
• The Congressional Record version makes better
grammatical sense than the language in the Act.
In any event, the section in the Act lacks the earlier punctuation, and the
section's application seems to be much narrower than the Congress apparently
intended.
Another issue raised by the third circumstance is what constitutes a
failure or refusal "to provide all reasonable cooperation and assistance
requested by a responsible public official." The phrase is not explained in
the legislative history, and case-by-case resolution of the issue may be
required. Uncertainty about the precise application of this provision may
cause potentially liable persons to provide assistance even when the request
does not seem reasonable.
4.3 Section 107(k)(l)
This section provides a limitation of liability to owners or
operators who close a hazardous waste disposal facility that is permitted
under Subtitle C of the Solid Waste Disposal Act (SWDA). Liability for such
closed facilities can be transferred to a post-closure liability fund if two
conditions are met:
• The facility complies with any regulations of the
SWDA and its regulations that may affect post-closure
performance.
16 Congressional Record, November 24, 1980, p. S 14988.
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
B-22
• The facility has been closed in accordance with
these regulations and the conditions of the permit,
and the facility and surrounding area have been
monitored as required by regulations and permit
conditions for up to 5 years after closure in order
to demonstrate "no substantial likelihood that any
migration off-site or release from confinement of any
hazardous substance or other risk to public health or
welfare will occur".
An issue raised by this provision is whether completion of the monitoring
requirement, without a release occurring, creates an irrebuttable presumption
that there is no substantial likelihood" that a release will occur. The
Senate Report on S. 1480 states that this period is provided so that the
Administrator can test further if it is felt to be necessary or
desirable".(pp. 89-90) In addition, the report notes:
If the Adminstrator has cause to believe the facility
operator made fraudulent representation or knowingly or
willfully violated his permit, the transfer of liability to
this fund is cancelled and the Administrator is directed to
seek to recover from the operator. The operator is
released from liability only if he actually complied with
the regulations and his permit and closure order. ( Ibid.)
The passage of the monitoring period, therefore, seems neither to confer
automatic eligibility for transfer of liability to the post-closure liability
fund nor to make such a transfer irreversible if fraud has occurred.
5. SECTION 103 -- PENALTIES RELATED TO NOTIFICATION REQUIREMENTS
Section 103 of CERCLA imposes penalties on certain persons who fail to
comply with reporting and recordkeeping requirements. The penalties,
liabilities, and requirements in these sections are imposed on "persons." The
Act's definition of "person" (Section 101(21)) is very broad:
"person" means an individual, firm, corporation,
association, partnership, consortium, joint venture,
commercial entity, United States Government, State,
municipality, commission, political subdivision of a State,
or any interstate body.
Almost any conceivable entity (including transporters, generators, disposers)
dealing with hazardous substances will be affected by the penalty, liability,
and financial responsibility provisions of CERCLA.
Penalties are imposed in three situations:
• failure to provide notification about a release of
hazardous substances (Section 103(b));
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
B-23
• failure to provide notification about a facility
where hazardous substances have been stored, treated,
or disposed (Section 103(c)); and
• failure to maintain certain records for a specified
period of time (Section 103(d)(2)).
These provisions are summarized in Exhibits B-3 and B-4. For this report,
however, the provisions of Section 103(c) and (d)(2) are of limited interest.
Section 103(c) is already to have been complied with and will not be affected
by the designation and reportable quantities regulations. Section 103(d)(2)
is also subject to a different set of regulations, although designation of a
substance as hazardous is necessary for Section 103(d)(2) to apply to a
substance. Because of their limited relation to the designation and
reportable quantity regulations, these two provisions are not discussed
further. The remainder of the discussion focuses on the penalty provisions in
Section 103(b)(2).
A person in charge of a vessel or an offshore or onshore facility is
required by Section 103(a) to notify the National Response Center (NRC) when a
reportable quantity (reportable quantities are established by Section 102 and
regulations under it) of a hazardous substance is released from that vessel or
facility (such releases will be referred to as "reportable releases"). The
notification must be made immediately, as soon as the person has knowledge of
the release. Section 103(b) authorizes a penalty of up to $10,000 and
one-year imprisonment for failure to comply with Section 103(a). The penalty
applies to persons in charge of vessels if (.1) the release is into or upon the
navigable waters of the United States, the adjoining shorelines, or into or
upon the waters of the contiguous zone, or (2) the release may affect natural
resources owned by, appertaining to, or exclusively managed by the United
States. The penalty applies to a person in charge of a facility from which
there is any reportable release. Federally permitted releases17 are
excluded from this notification requirement. Several issues raised by Section
103(b) are discussed in the following paragraphs.
Who is a "person in charge"? The Act does not define this term. One
possible interpretation is that this is the person most likely to know that a
release has occurred -- i.e., the person who is operating the vessel or
facility.
What constitutes "knowledge" of a release? The knowledge requirement
could be interpreted as its common meaning -- "belief in the existence of a
17This term is defined in Section 101(10).
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
EXHIBIT B-3
PENALTY PROVISIONS IN CEHCLA SECTION 103
Section Affected Party
103(b) 1) Person In charge of a vessel
from which there is a hazardous
substance release that equals
or exceeds the reportable quan-
tities established under
Section 102.
To Whom
Responsible
U.S. Government
Reason for
Imposing Penalty
Failure, as soon as a person
has knowledge, to notify immedi-
ately the appropriate U.S. Govern-
ment agency as soon as the
person has knowledge of the
release.
Maximum
Penalty
$10,000 and one
year Imprisonment
The release must be into
navigable waters, adjoining
shoreline, or into or upon
waters of the contiguous
zone; or
The release must be one that
may affect natural resources
owned by, appertaining to,
or under the exclusive manage-
ment authority of Che U.S.
The person must otherwise be
subject to U.S. jurisdiction
at the time of the release
0
1
to
Federally permitted releases
are not covered.
2) Person in charge of facility
from which there is a hazardous
substance release that equals
or exceeds the reportable quan-
tities established under Sec-
tion 102. Federally permitted
releases are not covered.
November 2 Draft
Do not quote or cite
-------
EXHIBIT B-3
PENALTY PROVISIONS IN CERCLA SECTION 103
(continued)
Section
103(c)
103(d) (2)
o
Affected Patty
With regard to facility where
hazardous substances are or have
been stored, treated, or disposed
and which facility does not have
a permit or interim status under
Subtitle C of the Solid Waste
Disposal Act:
1) Person who owns or operates
such a facility;
2) Person who, at the time oC
disposal, owned or operated
such a facility; or
3) Person who accepted hazardous
substances for transport and
who selected such a facility.
Not affected is any facility that
would be reported solely because
of a temporary stoppage in transit
that is incidental to the transpor-
tation or that is at the ordinary
operating convenience of a common
or contract carrier.
Same as affected persons under
103(c)
To Whom
Responsible
U.S. Government
U.S. Government
Reason for
Imposing Penalty
Knowing failure'to notify EPA
of the existence of such
facilities, including speci-
fying amount and type of any
hazardous substance found at
the facility, and any known,
suspected, or likely release
of such substances from such
facility
For 50 year period—knowing
destruction, mutilation,
erasure, disposal of, con-
cealment, or otherwise ren-
dering unavailable or unread-
able, or falsifying records
required under 103(d) (1) s
— location, title, or
condition of facility
— Identity, characteristics,
quantity, origin, or con-
dition of any hazardous
substances contained or
deposited in a facility
Section 103(d) (3) permits &
waiver of the full 50 year
period, and 103(d) (4) permits
EPA to impose greater than a
50 year period.
Maximum
Penalty
1) $10,000 and 1
year imprison-
ment
2) Loss of en-
titlement to
any limitation
of liability
or any defenses
to liability
set out in
Section 107
V
I
to
Ln
$20,000 and one
year imprisonment
November 2 Draft
Do not quote or cite
-------
EXHIBIT B-4
PENALTY PROVISIONS IN CERCLA SECTION 103—EXCEPTIONS
Section Exception
103(b) Federally-permitted release excepted
from requirements of Section 103(b)
103(e) Pesticide exception to coverage
of Section 103:
103(f) Exception from the notification
(and, consequently, the penalty)
provisions of 103(a) and (b)
Conditions of Exception
Release must satisfy the requirements set forth in Section 101(10)
(1) Application of pesticide product registered under FIFRA
(2) Handling and storage of such pesticide product by agricultural
producer.
(1)
(2)
A hazardous substance release that must be reported (or is specifically exempt) under
Subtitle C of the Solid Haste Disposal Act or its regulations and which has been
reported to the NRC.
A continuous release that is stable in quantity and rate and that is either:
(i) from a facility for which sufficient notice has been given under 103(c) , or
(11) one for which notification has been given under 103(a) and (b) for a period that
is sufficient to establish the continuity, quantity, and regularity of the release.
Notification under (1) or (ii) must be given annually or at any time there is a "statistically
significant increase" in the quantity of the release above that previously reported or
occurring.
V
I
to
November 2 Draft * * * Do not quote or cite
0
¦n
z
1
nl
o
-------
B-27
fact, which coincides with truth."18 It is uncertain whether a penalty
would be imposed for failure to notify by a person who "should have known"
about a release.
What is "immediate" notification? This requirement may be interpreted
according to the everyday meaning of "immediately" -- without delay. This
requirement imposes a more stringent requirement than a phrase such as "within
a reasonable time." Thus, for example, delaying an hour before reporting a
release might subject a person to the penalties provided by this section.
Can any penalty be imposed under Section 103(b)? There is an
inconsistency in the requirement of 103(a) and the reason for imposing a
penalty under 103(b). Section 103(a) imposes the notification requirement by
directing that notification is to be given to the National Response Center
which, in turn, will notify other appropriate government agencies. Section
103(b), however, imposes liability for failure to notify the appropriate
agency. Section 103(b), then, seems to impose the penalty for failure to do
something that is not required by Section 103(a). This problem could be
eliminated if the National Response Center were considered an "agency", but
the National Response Center does not fall neatly into that category. The
intent of the drafters seems obvious, however, and this inconsistency might be
viewed by the courts as one of form and not substance — the penalty would be
imposed for failure to notify the National Response Center. It also is
possible that notification of the National Response Center would be deemed
constructive notice of the appropriate agencies because of the requirement in
103(a) that the National Response Center convey notice to all appropriate
government agencies.
Under what conditions is notice excused? There is no notification
requirement and, consequently, no penalty for failure to notify about:
• a federally permitted release (Section 103(b));
• a hazardous substance release that must be reported
(or is specifically exempt) under Subtitle C of the
Solid Waste Disposal Act or its regulations and which
has been reported to the National Response Center
(Section 103(f)(1)); or
• a continuous release that is "stable in quantity and
rate" and that is either (i) from a facility for which
sufficient notice has been given under Section 103(c)
(see the discussion below) or (ii) one for which
18Prosser, Law of Torts (St. Paul: West Publishing Company, 1971) p.
157.
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
B-28
notification has been given under Section 103(a) and
(b) for a period that is sufficient to establish the
continuity, quantity, and regularity of the release
(Section 103(f)(2),).
The concepts of "continuous release" and "stable in quantity and rate" are not
further explained in the statute and will have to be dealt with in
regulations or Agency guidance. If notice is excused under Section 103(f)(2),
there still must be at least annual notice under the conditions noted there or
notice whenever there is "any statistically significant increase" in the
quantity being released. This phrase is not explained in the Act either.19
Also excluded from the coverage of Section 103(b) (and all of Section 103)
is the application of a pesticide product registered tinder the Federal
Insecticide, Fungicide, and Rodenticide Act (FIFRA) and the handling and
storage of such pesticide products by agricultural producers. (Section
103(e)). Application refers to the "customary uses" of pesticides — i.e.,
field and domestic use.20
6. FINANCIAL RESPONSIBILITY REQUIREMENTS
Section 108 of CERCLA establishes the financial responsibility levels that
must be demonstrated by owners and operators of vessels and facilities.
Section 108(a) deals with vessels, and Section 108(b) covers facilities.
These sections are summarized in Exhibit B-5 and are discussed in detail below.
6.1 Section 108(a)
The requirements apply to owners or operators of each vessel
exceeding 300 gross tons and using any "port or place" in the U.S., or using
navigable waters or any off-shore facility. In addition to excluding vessels
less than or equal to 300 gross tons, the section excludes non-self-propelled
barges that do not carry hazardous substances as cargo. Owners or operators
of covered vessels must establish and maintain financial responsibility levels
of $300/gross ton; if the vessel carries hazardous substances as cargo, the
level of financial responsibility is the greater of $300/gross ton or $5
million.
lsNote that the statute's language requiring the at least annual
reporting refers to "notification in accordance with subsections (a) and (b)
of this paragraph..." This appears to be a printing error and probably should
read: "notification in accordance with subsections (a) and (b) of 'this
section..." Thus, the requirements being referred to are those imposed under
Section 103(a) and (b), not Section 103(f)(2)(A) and (B).)
2"See Congressional Record, p. S 15006, November 24, 1980, remarks of
Senator Cannon.
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
EXHIBIT B-5
FINANCIAL RESPONSIBILITY REQUIREMENTS IN CERCLA SECTION 108
Section Affected Party
108(a) Owner or operator of each
vessel (except non-self-
propelled barges that do
not carry hazardous sub-
stances as cargo) greater
than 300 gross tons that
uses any port or place in
the U.S. or navigable
waters or any off-shore
facility
Level of
Financial Responsibility
Establish and maintain, as per
regulations, evidence of financial
responsibility of $300/gross ton.
For a vessel carrying hazardous
substances as cargo, the level of
financial responsibility is the
greater of $300/gross ton or
$5 million.
Bow to Demonstrate
Financial Responsibility
Any one. or combination of: insurance,
guarantee, surety bond, or self-
insurance. Bond must be from a
bonding company authorized to do
business in the U.S. If an owner/
operatot owns, operates, or charters
more than one vessel subject to
Section 108(a), evidence of respon-
sibility Is required only for the
maximum liability of the largest
of the owner/operator*8 vessels
Consequences of
Failure to Demonstrate
(1) The Secretary of Transportation
shall withhold or revoke the
clearance required by Section
4197 of the United States Revised
Statutes
(2), The Secretary of Transportation,
according to regulations issued
by him, shall (1) deny entry to
any port or place in the U.S. or
navigable waters, and (ii) detain
at the port or place in the U.S.
from which the vessel is about
to depart for any other port or
place in the U.S.
(3) $10,000/day of violation. This
civil penalty also is applied to
failures to comply with denial
or detention orders issued under
(2), above. (Section 109)
W
I
M
\S>
0
T1
5 November 2 Draft * * * Do not quote or cite
1
i
o
-------
EXHIBIT B-5
FINANCIAL RESPONSIBILITY REQUIREMENTS IN CERCLA SECTION 108
(continued)
Section Affected Party
108(b) Owners or operators of
facilities in addition to
those under Subtitle C
of the Solid Haste Disposal
Act and other Federal law
Level of
Financial Responsibility
(1) To be established by regula-
tion. Level to be based on
"degree and duration of risk
associated with the production,
transportation, treatment,
storage, or disposal of hazard-
ous substances." Within 3
years of CEBCLA's passage,
classes of facilities are to
be developed using these
criteria. Financial respon-
sibility requirements are to
be promulgated no earlier than
5 years after CERCLA's passage.
These requirements are to be
develop giving priority to those
classes of facilities, owners,
and operators deemed to present
the highest level of risk of
injury.
(2) Level of financial responsi-
bility required will be adjusted
to reflect experience of dif-
ferent classes.
How to Demonstrate
Financial Responsibility
Presumably the same as under
Section 108(a). (See above.)
Consequences of
Failure to Demonstrate
$10,000/day of violation
0
1
w
o
o
11
z
¦o
(3) Financial responsibility
requirements shall be phased
in Incrementally over a 3-6
year period.
(4) Multiple owners or operators
can present consolidated
evidence of proportional
share of financial responsi-
bility.
(5) Requirements for motor
carriers are to be determined
under Section 30 of the Motor
Carrier Act of 1980, Public
Law 96-296.
O
November 2 Draft * * * Do not quote or cite
-------
B-31
Financial responsibility can be demonstrated in several ways: insurance,
guarantee, surety bond, or self-insurance. For owners or operators who own,
operate, or charter more than one vessel that is subject to Section 108(a),
financial responsibility must be demonstrated only for the maximum liability
of the largest of the owner's or operator's vessels.
Several sanctions can be imposed if a vessel owner or operator fails to
comply with the financial responsibility requirements:
• The Secretary of the Treasury will withhold or
revoke the clearance required by Section 4197 of the
United States Revised Statutes if the vessel does not
have certification that the financial responsibility
provisions have been met.
• The Secretary of Transportation, pursuant to
regulations he is to issue, will deal with a vessel
lacking certification of financial responsibility by
(1) denying it entry to any port or place in the
United States or any navigable water, and (2)
detaining it at the port or place in the U.S. from
which it is about to depart for any other port or
place in the U.S.
• Failure to comply with financial responsibility
requirements or with a denial of entry or a detention
order under Section 108 can result in a penalty of up
to $10,000/day of the violation.
The vessel financial resonsibility provisions of CERCLA are modeled after
those in Section 311(p) of the Clean Water Act. The CERCLA provision,
however, requires a demonstration of greater financial responsibility. One
issue raised by the financial responsibility requirements for both vessels and
facilities is whether insurance will be available to owners and operators.
The answer to this issue is not clear, although the phase-in period under
Section 108(b) (see discussion below) is designed to facilitate insurers'
handling of facilities' demands for this kind of coverage.
6.2 Section 108(b)
Different financial responsibility requirements are imposed on
facilities than on vessels. The affected persons are owners and operators of
facilities in addition to those under Subtitle C of the Solid Waste Disposal
Act. The Senate Report, explaining the precursor to this section, stated:
These requirements are in addition to the financial
responsibility requirements promulgated under the authority
of section 3004(6) of the Solid Waste Disposal Act. It is
not the intention of the Committee that operators of
facilities covered by section 3004(6) of that Act be
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
B-32
subject to two financial responsibility requirements for
the same dangers. The purpose of this provision is
twofold: first, to extend financial tesponsibilty
requirements to facilities and transporters who are not now
covered by any requirements under section 3004(6), and
second, to create authority for phasing in financial
responsibility requirements for facilities and
transporters.21
The Act does.not specify the level of financial responsibility that must
be demonstrated. Instead, it establishes a framework and schedule for
developing financial responsibility requirements. No requirements can be
promulgated earlier than five years after CERCLA's passage (i.e., late 1985).
However, within three years of CERCLA's passage, classes of facilities are to
be developed based on "degree and duration of risk associated with the
production, transportation, treatment, storage or disposal of hazardous
substances." The financial responsibility requirements will be developed
giving priority to the classes of facilities, owners, and operators that are
deemed to present the highest level of risk of injury. Not only must five
years elapse before these requirements are promulgated, but an additional
three to six years must be spent phasing in these requirements. The Senate
Report notes that the phasing provision "was drafted in consultation with
representatives of the domestic insurance industry to assure that, as the need
arose, commercial insurance would be available.... There will be five years
in which claims experience can be built up, then another three-year period in
which insurers can gradually enter the market11.22
The level of financial reponsibility required is supposed to be adjusted,
"when necessary", to reflect the experience of the different classes of
facilities, owners, and operators. Multiple owners or operators are permitted
to present consolidated evidence that the financial responsibility require-
ments for the facility with which they are involved are fully satisfied.
Finally, motor carriers are to satisfy the financial responsibility require-
ments established under Section 30 of the Motor Carrier Safety Act of 1980.
7. OTHER SOURCES OF LIABILITY
CERCLA does not preempt other authorities under which a person may be held
liable for the release of hazardous substances.23 Liability is something of
a legal mosaic because there are a number of different, non-exclusive ways in
which it can arise. The main sources of liability are:
21 Environmental Emergency Response Act, Report of the Committee on
Environment and Public Works, U.S. Senate, p. 92 (Report No. 96-848, July 11,
1980).
22 Ibid., pp. 92-93.
23The one exception to this rule is provided by Section 304(c) which
states that any conflict between CERCLA and section 311 of the Federal Water
Pollution Control Act will be resolved in favor of CERCLA.
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
B-33
• common law, and
• statutory.law.
Within each category there are further distinctions. For example, a firm
releasing hazardous substances may be liable under a number of different
common law theories, including:
• negligence;
• nuisance;
• trespass;
• strict liability;
• absolute liability; and
• other theories and precedents.
In each case, the kinds and amounts of damages subject to recovery, as
well as possible defenses, may be different or may vary in different states.
In some states, liability may be "joint and several" (i.e., shared), which has
procedural advantages for claimants.
Unlike regulations, which impose compliance costs with the goal of
preventing harm, liability rules are activated by the occurrence of harm and
allocate certain categories of costs to certain actors after the fact. Also,
the common law imposes no financial responsibility requirements.
Nevertheless, even uncertain prospects of being liable for costs may act
as an incentive for a firm to (1) enter or leave a line of business, and (2)
manage the risks involved by investing in such things as better safeguards and
training.
The scope and reach of common law theories of liability are determined
through case-by-case litigation. Consequently, there is much variation among
the states with respect to the issue of hazardous substance liability.
Typically, however, suits are brought by individuals for damages to private
property and health. Furthermore, there is virtually no case law relating to
liability for (1) removal and clean-up costs, or (2) damages to publicly-owned
natural resources, under common law..
Statutory liability is similarly fragmented. A significant number of
states have imposed different types of statutory liabilities on operators of
hazardous waste sites and companies involved with hazardous substances.
Likewise, under federal law, both the Clean Water Act (CWA) and CERCLA
establish liability for certain costs associated with the release of hazardous
substances;24 and financial responsibility for loss claims resulting from
24See footnote 23.
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
-------
B-34
releases is authorized not only by CERCLA and the CVA, but by the Motor
Carrier Safety Act and RCRA as well.
Thus, it is important to understand that other liability laws and
doctrines form part of the baseline for this analysis. As shown in Exhibit
B-6, the liability imposed by CERCLA complements but does not overlap common
law liability. This means that existing insurance coverage--which insures
against damage to persons and private property only--will need to be
supplemented or revised. This also means that the extension of liability
under CERCLA will impose an incremental cost burden on affected industries.
One key issue for purposes of this analysis is to recognize that Section
311(f) of the Clean Water Act is part of the "liability baseline." The
burdens and benefits of CERCLA liability are only those that go beyond this
baseline. Insurance for response cost liability under the Clean Water Act has
been made available to marine and inland waterway transporters of hazardous
substances.
EXHIBIT B-6
APPLICABILITY OF LIABILITY RULES TO DIFFERENT
CATEGORIES OF COSTS OF RELEASES
Response &
Clean-up
Costs
Natural
Resource
Damages
Private
Property
Damages
Personal
Injury
Damages
Economic
Loss
Damages
Common Law
No
No
Yes
Yes
Yes
CERCLA §107
Yes
Yes
No
No
No
Clean Water Act §311
Yes
Yes
No
No
No
Trans-Alaska
Pipeline Act
Yes
Yes
Yes
Yes
Yes
Deepwater Ports Act
Yes
Yes
Yes
Yes
Yes
State Statutes
Vary with the Provisions of State Law
November 2 Draft
* * * Do not quote or cite
ICF INCORPORATED
-------
B-35
8. SUMMARY
Liability and financial responsibility provisions are major determinants
of regulated parties' behavior under CERCLA. The liability and financial
responsibility regime of CERCLA imposes significant costs on responsible
parties who fail to comply with the requirements and provides incentives for
behavior that benefits public health and the environment. These provisions
focus exclusively on CERCLA-designated hazardous substances; persons
responsible for pollutant or contaminant releases are not required by CERCLA
to report or cleanup the release, nor are they subject to liability under
CERCLA. Other sources of liability, however, may provide sufficient incentive
for releasers of pollutants or contaminants to take prevention and cleanup
measures with regard to such releases.
November 2 Draft * * * Do not quote or cite
ICF INCORPORATED
------- |