EPA 450/5-84-QQ1
RESPONSE TO PUBLIC COMMENTS ON EPA'S
LISTING AND REGULATION OF BENZENE UNDER SECTION 112:
COMMENTS OF A GENERAL POLICY NATURE
Office of Air Quality Planning and Standards
U.S. Environmental Protection Agency
Research Triangle Park, N.C. 27711
May 1984
-------
ENVIRONMENTAL
PROTECTION
AGENCY
DALLAS, TEXAS
1Y
NOTE: Commenters are identified by the rulemaking docket numbers assigned
their submissions. The relevant dockets are as follows:
OAQPS-79-3 Benzene Listing and Maleic Anhydride Standard
A-79-27 Benzene Fugitive Emissions Standard
A-79-49 Ethylbenzene/Styrene Standard
A-80-14 Benzene Storage Vessel Standard
OAQPS-79-14 Airborne Carcinogen Policy
A number of commenters submitted multiple or duplicate comments to several of
the dockets. While EPA has made an effort to reference representative comment
sources, the citations should not be considered an exhaustive record of the
docket items addressing a particular issue.
-------
1.0 INTRODUCTION
In addition to comments on the decision to list benzene, addressed
in "Response to Public Comments on EPA's Listing of Benzene Under
Section 112" (EPA 450/5-82-003), EPA received a number of oral and written
submissions addressing the procedures followed by EPA in evaluating the
public health hazard posed by benzene, selecting sources for regulation,
and determining appropriate levels of control. Many of the procedural
comments referenced statements made by the Agency in a related rulemaking
proposal (44FR58642, October 10, 1979) to establish an EPA policy for the
identification, assessment, and regulation of airborne carcinogens.1 EPA
has not, at this time, published a final policy. Relevant comments from
the carcinogen policy record, however, are included in the discussion
below and EPA's responses reflect consideration of the comments from both
sources.
2.0 COMMENTS OF A GENERAL POLICY NATURE
The comments are divided into four major policy issues: 1) the role
of quantitative risk estimation in the regulatory process; 2) the appropriate
criteria for listing an airborne carcinogen under Section 112; 3) the appro-
priate criteria for assigning regulatory priorities to emitting sources; and
4) the requirement of best available technology (BAT) for selected source
categories and the evaluation of residual risks in determining the appropriate
level of control under section 112.
U.S. Environmental Protection Agency "Policy and Procedures for Identifying
Assessing, and Regulating Airborne Substances Posing a Risk of Cancer; Notice
of Proposed Rulemaking" 44 FR 58642, October 10, 1979.
-------
3
2.1 The Role of Quantitative Risk Estimation in the Regulatory Process
Commenters were significantly divided on the general issues of the
accuracy and reliability of quantitative risk assessment (QRA) for carcinogens
as well as on the utility of such estimates in the regulatory decision process.
Proponents of quantitative risk assessment (QRA), primarily industry and trade
association commenters, held that such assessments, when based on reliable
data, should play an important role at all stages of decisionmaking on
potential airborne carcinogens (OAQPS-79-3(Part ID-IY-D-23, IV-F-1 , IV-F-9;
OAQPS-79-14-IY-D-65,73,74,78,79,88,90a,94,97,102,104,110,116,117,1193,128,129,
130-133,135,195). Comments largely from public interest groups, State air
pollution control agencies, and private individuals, expressed concern that
the underlying uncertainties in attempting to quantify cancer risks greatly
reduces the reliability of such estimates and argues for limiting or avoiding
their use in the regulatory process (OAQPS-79-3(Part II)-IY-F-4;OAQPS-79-14-
IV-D-52,118,120,179a,184,194,G-8) and that such use is not required by the
language of the Clean Air Act (A-79-25-IY-D-31).
Commenters skeptical of the utility of QRA cited a number of reasons for
discounting the reliability of numerical risk estimates. These included:
uncertainties in dose extrapolation and scaling factors; differing
sensitivities and metabolic pathways in humans versus laboratory animals;
confounding variables, potential interactions, and synergism; latency
and unknown exposure levels in epidemiological studies; and the general
unreliability of exposure monitoring and modeling techniques (OAQPS
79-3(Part ID-IV-F-4; OAQPS-79-14-IV-D-118, IV-G-8). Organizations
taking this view suggested that EPA place less emphasis on QRA (OAQPS-79-
14-IV-D-120), attempt to characterize the uncertainty in such estimates
by presenting them as ranges (OAQPS-79-14IV-D-120,IV-D-194), and limit
-------
4
their role in the decision process to a rough indicator of relative risk
for use in the assignment of regulatory priorities (OAQPS-79-3(Part II)-
IV-F4; OAQPS-79-14-IV-D-118,120,1 79a,l 94).
Of the role of risk estimates in cost/benefit analysis, two commenters
argued that EPA lacked authority to engage in such practices under
section 112. The commenters stated further that Congress recognized the
difficulties, given the current state of scientific knowledge, with
attempting to quantify an "ample margin of safety." Congress did not
require EPA to delay regulation in the absence of such quantification or
to undertake "sham quantitative risk analysis." (A-79-27-IV-D-27;
A-80-14-IV-D-19).
EPA Response
EPA has considered the comments on the role of quantitative risk
assessment in the regulatory decision process for airborne carcinogens and
concludes that, while numerical estimates of risk offer a clear appeal over
non-quantitative data in regulatory decisionmaking, the importance ascribed
to such information must be in proportion to its reliability. EPA agrees
that QRA can provide meaningful information to the decision process but
disagrees with comments suggesting that risk assessments should be the
sole criterion for decisionmaking.
Where risk estimation is feasible and some measure of confidence is
obtainable, EPA will perform quantitative assessments for use in the ap-
propriate stages of the regulatory process. For potential airborne car-
cinogens, these stages could include: comparisons between pollutants; com-
parisons among source categories of a pollutant; selection of appropriate
control levels; and the evaluation of residual risks.
-------
5
EPA does not agree that the Agency lacks authority to consider risk
estimation or other social and economic factors in the determination of
whether an "ample margin of safety" is provided by a given control level.
First, the primary purpose of section 112 is to improve public health
and QRA, despite its limitations, is the best tool available for evaluating
the health consequences of environmental pollution. Second, EPA believes
that the Congress1 direction that standards under section 112 provide an
"ample margin of safety" does not require the total elimination of risk.
Given the serious economic consequences of total elimination of risk for
the control of non-threshold pollutants such as carcinogens, EPA feels
that, had this been the intent, Congress would have spoken more explicitly,
as it did in the Delaney Clause of the Food and Drug Act2, prohibiting
the use of any food additive found to induce cancer.
In this interpretation, EPA is following the approach enunciated by
the U.S. Suprement Court in Industrial Union Department AFL-CIO v.
American Petroleum Institute, 448 U.S. 607 (1980), in which it stated
that:
"By empowering the Secretary to promulgate standards that are
'reasonably necessary or appropriate to provide safe or healthful
employment and places for employment,' the Act implies that,
before promulgating any standard, the Secretary must make a finding
that the workplaces in question are not safe. But 'safe' is not
the equivalent of 'risk-free.' There are many activities that
we engage in every day—such as driving a car . . .--that entail
some risk of accident or material health impairment; nevertheless,
few people would consider these activities 'unsafe.1 Similarly,
a workplace can hardly be considered 'unsafe1 unless it threatens
the workers with a significant risk of harm.
Therefore, before he can promulgate any permanent health or safety
standard, the Secretary is required to make a threshold finding
that a place of employment is unsafe—in the sense that significant
risks are present ..."
221 U.S.C. 348(c)(3)(A)
-------
6
Where the presence of some residual risk is likely, EPA believes
that it is appropriate and necessary to attempt to characterize this risk
and to consider it along with other factors in the determinations of the
appropriate level of control. EPA does not agree that such consideration,
appropriately tempered by an awareness of the uncertainties in such
analysis, would delay regulation or represent "sham quantitative risk
analysis."
2.2 The Appropriate Criteria for Listing an Airborne
Carcinogen Under Section 112
A number of commenters expressed concern over the criteria for listing a
substance under section 112 as described in EPA's proposed airborne carcinogen
policy (44 FR 58642). The criteria provided that EPA would list under section
112 substances that 1) had a high probability of being human carcinogens, and
2) were emitted into the ambient air from one or more categories of stationary
sources in amounts sufficient to significantly expose human populations. The
listing decision would be based on preliminary assessments of carcinogenicity
and human exposure. Several commenters objected to the listing criteria on
the grounds that a "low hurdle" based on preliminary studies was unwarranted
and unrealistic (OAQPS-79-3(Part ID-IV-F-1 ,IV-F-9; OAQPS-79-14-IV-D-26,
90a,106,110,135). One commenter argued that EPA must conduct more than a
risk assessment to justify listing (OAQPS-79-3(Part ID-IV-F-9). A number
of commenters felt that EPA had underestimated the impact of the listing
decision itself, and should avoid listing until such time as a reasoned
and supported judgment could be made (OAQPS-79-3(Part ID-IV-F-9; OAQPS-79-
14,IV-D-77,79,97,112,116,179a,l 92).
-------
7
It was the view of five commenters that EPA should publish, and take
public comment on, the basis for listing prior to a listing decision (OAQPS-
79-14-IV-D-51 ,55,60,73,199). Six commenters suggested that EPA make public
the list of pollutants under assessment as candidates for listing (OAQPS-79-
14-IV-D-60,75,77,86,90a,157).
Commenters were divided on the use of quantitative risk assessment (QRA)
in the listing decision, with positions ranging from the use of QRA as the
listing basis (OAQPS-79-14-IV-D-94), to use as a factor in the decision
(OAQPS-79-3(Part ID-IV-F-9; OAQPS-79-14-IV-D-67,97,116,1 39,199), to the
establishment of a risk "benchmark" for listing (OAQPS-79-14-IV-0-79), and
finally, to having no appropriate role in the listing decision (OAQPS-79-14-
IY-D-100).
Two commenters requested more guidance on and a definition of
"significant" as applied to the human exposure criterion (OAQPS-79-14-IV-D-
63,IV-D-118). One commenter felt the listing decision format was not clear
(OAQPS-79-14-IY-D-104).
Two commenters suggested that the final policy should include specific
criteria for delisting as well as listing substances (OAQPS-79-14-IV-D-94,
IY-D-199).
One commenter recommended that EPA proceed to list all substances
currently regulated as carcinogens by other agencies (OAQPS-79-14-IV-0-140).
EPA Response:
In consideration of comments on the nature of the "significance"
criterion for human exposure, it remains EPA's opinion that the test of
significance is largely judgmental and does not lend itself to rigid quanti-
fication. A judgment of significance depends not only on the magnitude of
-------
8
emissions and consequent exposure levels, but also on the distribution
of emissions including the number and nature of emitting sources.
Delisting procedures are adequately described in section 112(b)(l)(B).
EPA does not see a need for further discussion or interpretation of the
del isting process.
EPA does not consider the wholesale listing of all substances currently
regulated as carcinogens by other agencies as a feasible or warranted
action. First, listing under section 112 requires a finding that a substance
"contributes to air pollution." Moreover, in EPA's judgment only substances
having a significant effect on public health should be listed. The mere
fact that a substance is regulated as a carcinogen in Pharmaceuticals,
cosmetics or food does not show that it meets these criteria.
Second, section 112 directs that the Administrator should list only those
pollutants "for which he intends to establish an emission standard." In the
absence of an assessment of the emitting sources, such intent would not be
established.
2.3 The Appropriate Criteria for Assigning Regulatory Priorities to
Emitting Sources
Most commenters supported a mechanism for assigning regulatory priorities
to categories of stationary sources emitting a hazardous air pollutant. Some
saw such a process as a means of diverting attention away from "small cancer
risks" (OAQPS-79-3(Part D-IY-D-13), others as a means of focusing regulatory
attention on the "worst offenders" (OAQPS-79-14-IV-G-6).
Commenters were generally supportive of the use of risk estimates in
priority assignment as a measure of health hazard. Three commenters felt that
risk estimates should weigh equally with the non-risk criteria (feasibility,
-------
9
ease of control) in the determination (OAQPS-79-14-IV-D-90a,190a,191). One
commenter suggested that only the extent of the health hazard be used to make
distinction (OAQPS-79-14-IV-D-115). One commenter approved the use of risk
estimates but only as a "rough measure" (OAQPS-79-14-IV-D-118).
Two commenters suggested that the risk estimates be combined with
projected control costs to permit measures of cost-effectiveness to be
factored into the priority process (OAQPS-79-3(Part D-IV-D-23; OAQPS-79-
14-IV-D-26).
One commenter maintained that EPA should develop priorities across
pollutants as well as across source categories of a particular pollutant
(OAQPS-79-14-IV-D-79).
EPA Response:
EPA envisioned the priority assignment process as serving two purposes:
first, the identification of those categories of sources which presented
potentially the greatest hazards to public health; and, second, the
identification of those sources for which the regulatory and control
dollar would go furthest in reducing the hazards. In combination, these
aims would ensure that the most important and tractable problems would be
addressed on a priority basis.
EPA agrees that non-risk criteria are important in the assignment of
priorities. Setting priorities based on estimates of health hazard alone
could result in resource-intensive regulations with little health benefit.
EPA recognizes the advantages of combining risk estimates with proposed
emission reduction and control costs in the derivation of "cost effectiveness"
figures. Such estimates would be, however, preliminary since the extent
and cost of possible control would not be well enough known at this stage to
permit firm estimates.
-------
10
EPA has considered expanding the priority process to include other
pollutants. There are difficulties with such an approach, however. The
staggered timing of listings would make it difficult to coordinate
priorities, but to the extent possible, EPA does attempt to set priorities
for regulation development across pollutants.
The availability of resources for regulatory development will influence
the assignment of priorities for source categories of a listed substance.
Actions planned or underway for source categories of previously listed
pollutants will be considered in the priority assignment process. While it
is possible that an ongoing action could be suspended to divert resources to
a higher priority project, EPA anticipates that such redirections would be
i nfrequent.
2.4 The Requirement of Best Available Technology (BAT) as the Minimum
Level of Control for Selected Source Categories
Although most commenters endorsed EPA's proposed procedure for
consideration of economic and technological feasibility in the development of
emission standards under section 112, the automatic requirement of best
available technology (BAT) was often criticized as unnecessarily rigid and
not reflective of Congressional intent. A number of commenters argued that a
level of control less stringent than BAT could be appropriate where the health
risks are low (OAQPS-79-3(Part ID-IY-D-22; OAQPS-79-14-IV-D-132, IV-D-155).
Similarly, several commenters argued that the imposition of BAT could result
in excessive control (OAQPS-79-14-IV-D-29), arbitrarily chosen (OAQPS-79-14-
IV-D-65), with no statutory support (OAQPS-79-3(Part D-IV-D-5), and no
evidence that the avoidable risks were unreasonable (OAQPS-79-14-IY-D-87).
-------
11
Several commenters suggested that health risks should be considered in
establishing BAT (OAQPS-79-14-IV-D-73,90a,155) or that cost benefit analysis
should be performed (OAQPS-79-H-IV-D-29-90a,97). One commenter recommended
EPA determine acceptable risk levels in place of a BAT minimum control level
(OAQPS-79-14-IY-D-135).
Two commenters suggested as alternatives to BAT: "reasonably available
control technology" (OAQPS-79-14-IV-D-187), and "best available retrofit
technology" (OAQPS-79-14-IV-D-195).
One commenter argued that section 112 contained the presumption of a zero-
emission standard, and contended that an evaluation of "residual risk" goes
farther than EPA is authorized in the consideration of technical and
economic factors, and argued that BAT was inadequate to address the
hazardous posed by substances for which no level of exposure could be
considered safe (A-79-27-IV-D31 ;A-80-14-IY-D-19).
EPA Response:
Section 112 provides that standards must be set so as to protect public
health with an ample margin of safety.3 Historically, EPA has responded
to this mandate by requiring for source categories determined to pose
significant cancer risks, the application of best available technology
(BAT), considering costs, non-air quality health and environmental, and
energy impacts, and then evaluating the residual risks to determine the
need for further controls.
Where a health effects threshold can be determined, the margin of
safety requirement can be met by establishing the standard at a level
that insures that the exposure threshold is highly unlikely to be exceeded.
3 Clean Air Act as amended August, 1977, Section 112(b)0 ) (B) ,p.38.
-------
12
Cases may also arise where the reductions in the estimated health risks
obtainable with the application of BAT are so small that little health
benefit would be realized.
In establishing the appropriate level of control for carcinogens,
EPA views the objective as a judgment of the extent to which the estimated
risk of cancer must be reduced before the degree of control can be considered
amply protective. Two choices are available: either the emission standards
must be set at zero to eliminate the risk of cancer altogether, or some
residual risk must be permitted. In the absence of specific direction on
this choice in section 112 and in recognition of the drastic economic
consequences that could follow a requirement to eliminate all risk from
carcinogenic emissions, EPA believes that it is not the intent of this
section to eliminate totally all risks and that section 112 standards
which permit some level of residual risk, therefore, provide an ample
margin of safety to protect public health.
Over the past several years, the Agency has explored a number of
approaches to the evaluation of residual risks and the determination of what
constitutes an amply protective standard under section 112. In EPA's
judgment, the decision that a particular residual risk is not unreasonable
cannot be based solely on the level of risk. Other factors which*influence
society's ability to estimate and to mitigate such risks must also be
considered. These include the uncertainties inherent in the estimation of
carcinogenic risk as well as the social and economic impacts of further
emission reductions.
In determining BAT for source categories regulated under section
112, the Agency first identifies alternative levels of control which have
-------
13
control may have actually been achieved by representative plants in the
source categories. The Agency determines the costs and associated impacts
of the various alternatives. The Administrator selects that alternative
which achieves the most emission reduction and/or risk reduction without
incurring unreasonably adverse impacts of one type or another. As suggested
by one of the commenters, retrofit costs are considered when determining
the appropriate control level for existing source categories. A particular
regulatory alternative may be rejected for a variety of reasons, among
them that: it is judged to result in a price increase that adversely
affects consumers to an unreasonable extent; it will result in plant
closures or unreasonably discourage the construction of new plants due to
reduced return on investment or capital inavailability; or it has reasonably
high costs for the amount of emission or risk achieved.
After selecting BAT for each source category to be regulated, EPA
evaluates the incremental reductions in health risks obtainable against
the incremental costs and economic impacts estimated to result from the
application of more stringent control alternatives. Based on this
evaluation of the risks remaining after the application of the control
technology and the impacts of further control, EPA determines whether the
residual risks are unreasonable. If not, BAT represents an appropriate
level of control that provides an ample margin of safety to protect human
health. If the residual risks are unreasonable, the standard will be set
at a more stringent level.
Although the BAT approach has been used in recent benzene rulemakings,
EPA has come to recognize that it may give limited and indirect weight to
information on exposure and health risks in determining BAT and more direct
-------
14
weight to the amount of emissions reduced. For example, in determining BAT
for emission sources, the Agency relies on estimates of the total emissions
reduced and on estimates of the average and incremental cost of reducing
those emissions. However, the Agency recognizes that emission estimates
along can sometimes be poor measures of public health risks because they
do not account for the carcinogenic potency or exposure potential of
hazardous air pollutant emissions.
In order to more directly consider health risks, the Agency is
considering changing the approach for selecting the appropriate control
levels for hazardous air pollutants. The alternative approach would
combine the current two-step process into one step. In selecting the
appropriate control technique, EPA would consider in one step the before-
and societal costs of achieving those risk reductions. The major change in
this approach would be the greater consideration of public health risks
over emission estimates in selecting controls.
EPA has considered the adoption of acceptable risk levels and cost
per life saved targets, but considers such goals less desirable than an
approach which considers not only the health risks but other economic,
energy, and environmental impacts of regualtory alternatives. Risk
targets are difficult to establish and give no weight to the feasibility
of risk reduction or other benefits of regulation.
------- |