[Federal Register Volume 62, Number 36 (Monday, February 24, 1997)]
[Rules and Regulations]
[Pages 8314-8328]
From the Federal Register Online via the Government Publishing Office [www.gpo.gov]
[FR Doc No: 97-4196]



[[Page 8313]]

_______________________________________________________________________

Part II





Environmental Protection Agency





_______________________________________________________________________



40 CFR Part 51, et al.



Credible Evidence Revisions; Final Rule

  Federal Register / Vol. 62, No. 36 / Monday, February 24, 1997 / 
Rules and Regulations  

[[Page 8314]]



ENVIRONMENTAL PROTECTION AGENCY

40 CFR Parts 51, 52, 60 and 61

[FRL-5691-2]
RIN 2020-AA27


Credible Evidence Revisions

AGENCY: Environmental Protection Agency (EPA).

ACTION: Final rule.

-----------------------------------------------------------------------

SUMMARY: In an October 22, 1993 Federal Register, EPA solicited public 
comment on a proposal to amend 40 CFR Parts 51, 52, 60 and 61 to 
eliminate language that has been read to provide for exclusive reliance 
on reference test methods as the means of demonstrating compliance with 
various emission limits under the Clean Air Act (``CAA'' or ``Act''). 
These revisions--generally referred to as the ``credible evidence'' 
revisions--were designed to clarify that non-reference test data can be 
used in enforcement actions, and to remove any potential ambiguity 
regarding this data's use for compliance certifications under Section 
114 and Title V of the Act. In the same document, EPA proposed an 
``enhanced monitoring'' rule under Section 114 and Title V. EPA 
subsequently decided to suspend development of the original enhanced 
monitoring rule and develop a compliance assurance monitoring (``CAM'') 
approach to serve the same statutory goals as the original enhanced 
monitoring proposal. Today's rulemaking finalizes the previously 
proposed credible evidence revisions to Parts 51, 52, 60 and 61. EPA 
will take final action regarding enhanced monitoring and CAM in a 
separate rulemaking.

DATES: Effective Date: April 25, 1997. Judicial Review: Under CAA 
section 307(b)(1), judicial review of this nationally applicable final 
action is available only by the filing of a petition for review in the 
U.S. Court of Appeals for the District of Columbia Circuit within 60 
days of today's publication of this rule. Under CAA section 307(b)(2), 
the regulations that are the subject of today's rule may not be 
challenged later in civil or criminal proceedings brought by EPA in 
reliance on them.

ADDRESSES: Docket. Supporting information used in developing this 
rulemaking is contained in Public Docket No. A-91-52. This docket is 
available for public inspection and copying between 8:00 a.m. and 5:30 
p.m. on weekdays, excluding federal holidays, at the EPA Air and 
Radiation Docket and Information Center, Room M-1500, Waterside Mall, 
401 M Street SW., Washington, DC 20460; telephone (202) 260-7548. A 
reasonable fee may be charged for photocopying docket materials.

FOR FURTHER INFORMATION CONTACT:
Gregory Jaffe, Air Enforcement Division (Mailcode 2242-A), Office of 
Regulatory Enforcement, U.S. Environmental Protection Agency, 401 M 
Street, SW, Washington, D.C. 20460; telephone (202) 564-2260.

SUPPLEMENTARY INFORMATION: The contents of the preamble are listed in 
the following outline:

I. Background
    A. Statutory and Regulatory Authority
    B. Benefits of the Credible Evidence Revisions
    C. Public Participation
II. Summary of Final Rule
    A. 40 CFR Part 51, Sec. 51.212
    B. 40 CFR Part 52, Sec. 52.12
    C. 40 CFR Part 52, Sec. 52.30
    D. 40 CFR Part 60, Sec. 60.11
    E. 40 CFR Part 61, Sec. 61.12
III. Major Issues
    A. Use of Credible Evidence in Enforcement Actions
    B. Use of Credible Evidence in Compliance Certifications
    C. EPA's Authority To Promulgate the Credible Evidence Revisions
    1. Statutory Authority
    2. The Kaiser Steel Decision Does Not Constrain EPA's Authority 
To Amend its Regulations
    3. Despite Commenters' Claims, Clean Air Act Case Law Does Not 
Mandate Exclusive Reference Tests
    4. The 1990 CAA Amendments Further Support EPA's Authority
    5. Commenters' Attempts To Narrow the Scope of Sections 113(e) 
and 113(a) Are Unpersuasive
    6. EPA Can Promulgate the Credible Evidence Revisions Without 
Reproposal
    D. Stringency
    1. Emissions Limits Require Continuous Compliance (Consistent 
With Any Averaging Times) Except During Periods Where Compliance is 
Specifically Excused
    2. Commenters' Advocacy of Noncontinuous Compliance Would Lead 
to Numerous Anomalies
    3. Comments Regarding Continuous Compliance Are Not Directed at 
Today's Action, but Rather at Underlying Emission Standards
    4. Enforcement Using Continuous Monitoring Data Does Not 
Increase the Stringency of Applicable Requirements
    5. Sources Must Comply Both With Good Operation and Maintenance 
Requirements and With Emission Limits
    E. SIP Call
IV. Administrative Requirements
    A. Docket
    B. Office of Management and Budget (OMB) Review
    C. Unfunded Mandates Reform Act
    D. Regulatory Flexibility Act
    E. Paperwork Reduction Act
    F. Submission to Congress and the General Accounting Office

I. Background

A. Statutory and Regulatory Authority

    The credible evidence revisions are based on EPA's long-standing 
authority under the Act, and on amplified authority provided by the 
1990 CAA Amendments. Section 113(a) of the Act authorizes EPA to bring 
an administrative, civil or criminal enforcement action ``on the basis 
of any information available to the Administrator.'' In this provision, 
which predates the 1990 CAA Amendments, Congress gave EPA clear 
statutory authority to use any available information--not just data 
from reference tests or other federally promulgated or approved 
compliance methods--to prove CAA violations. (The preamble will 
generally use the phrase ``reference tests'' to include all these 
compliance methods. Where appropriate, the phrase ``reference tests'' 
will also include test conditions specified in individual regulations.)
    In the 1990 CAA Amendments, Congress included an enforcement title 
(Title VII) to enhance EPA's compliance and enforcement authorities. 
Among other things, Congress revised Section 113(e)(1) of the Act to 
overrule a federal court decision (Kaiser Steel, discussed below) that 
had held that only specified reference test data could prove 
violations. Thus, although the pre-existing authority of Section 113(a) 
forms the principal basis for today's action, the credible evidence 
revisions are also supported by the language, history and intent of the 
1990 CAA Amendments. See also Section III.C. below.
    In addition to clarifying EPA's, states' and citizens' enforcement 
authorities under the Act, the credible evidence revisions eliminate 
any potential ambiguity regarding the use of non-reference test data as 
a basis for Title V compliance certifications. Such potential ambiguity 
could arise from comparing the draft compliance assurance monitoring 
(CAM) approach and associated Part 70 changes, which would allow 
sources to include CAM data as a basis for certifying compliance, with 
various EPA regulations that could be read on their face to specify 
reference test methods as the sole means of determining compliance.

B. Benefits of the Credible Evidence Revisions

    As a preliminary matter, EPA wishes to clearly state that this 
rulemaking merely addresses an evidentiary issue. The credible evidence 
revisions are not

[[Page 8315]]

intended to and will not serve to affect the stringency of underlying 
emission standards by amending the nature of the compliance obligation. 
This rulemaking does not amend existing emission standards nor does it 
modify generic regulations affecting the compliance obligation such as 
exceptions for startup, shutdown, and malfunctions. See, e.g., 40 CFR 
60.8(c). This regulation also does not designate any particular data as 
probative of a violation of an emission standard. Rather, this 
regulation merely removes what some have construed to be a regulatory 
bar to the admission of non-reference test data to prove a violation of 
an emission standard, no matter how credible and probative those data 
are that a violation has occurred. The credible evidence revisions do 
not affect the compliance obligation and thus do not affect the 
stringency of existing emission standards. What compliance obligation 
is imposed by any given emission standard remains an issue ultimately 
to be determined based on that emission standard and not this 
rulemaking.
    For these reasons, we do not believe that this rulemaking affects 
whether emission standards require intermittent or continuous 
compliance. However, as made clear below, and in the detailed response 
to comments document, EPA's position continues to be that an emission 
standard requires continuous compliance unless the emission standard 
specifically provides otherwise.
    Today's credible evidence revisions will benefit sources, state 
environmental agencies, EPA and the public. EPA, states and citizens 
will be able to use credible evidence to assess a source's compliance 
status and respond to noncompliance. This will help ensure that the 
government and citizens alike can respond to sources that are not 
complying with air pollutant emission standards on an ongoing basis, 
thus furthering the protection of public health and the environment. At 
the same time, sources will be able to use credible evidence for 
contesting allegations of noncompliance in enforcement actions. 
Accordingly, today's rulemaking exemplifies EPA's ``common sense'' 
approach to environmental protection, which encourages smarter, cheaper 
and more flexible means of achieving environmental goals without 
compromising the fundamental health and environmental protections 
provided by federal environmental laws.
    In the past, state regulatory authorities and EPA have relied 
primarily on infrequent on-site inspections and even more infrequent 
reference tests in order to check compliance with emission limits at 
major stationary sources. According to a September, 1990, General 
Accounting Office (GAO) report, these on-site inspections were 
performed approximately once a year; the reference tests, typically 
once every five years. ``Air Pollution: Improvements Needed in 
Detecting and Preventing Violations,'' GAO, No. GAO/RCED-90-155, 
September 1990, at 12, 19. These methods are inadequate to ensure that 
sources continuously stay within their emission limits: for example, 
Pennsylvania officials have estimated that, in comparison with 
continuous emissions monitoring, on-site inspections may be 50 times 
less likely to detect non-compliance. Id. at 18. Reference tests may 
not yield a representative emissions picture because the sources 
typically schedule, set up and run the tests themselves. This allows 
sources to ``fine tune'' their operations and emissions control 
processes prior to the tests, and generate results that may not be 
typical of day-to-day source operations. Id. at 19-20. Reference tests 
can also be expensive and burdensome: They can cost up to $100,000, and 
take a week or more to complete. See, e.g., 43 FR 7568, 7571 (1978).
    In contrast to the above approach, today's rule will make it clear 
that various kinds of information other than reference test data, much 
of which is already available and utilized for other purposes, may be 
used to demonstrate compliance or noncompliance with emission 
standards. (The preamble generally refers to this other information as 
``non-reference test data''). EPA, state agencies and industry 
routinely rely on many types of information, including engineering 
calculations, indirect estimates of emissions, and direct measurement 
of emissions by a variety of means, in order to assess compliance with 
CAA requirements. Where available, continuous emission monitoring (CEM) 
data and well-chosen parametric monitoring data, such as the operating 
temperature and air flow rate of a regenerative thermal oxidizer, 
generally provide accurate data regarding a source's compliance with 
emission limits and standards. These data also generally cover a 
greater percentage of a source's time in operation and are more 
representative of a source's ongoing compliance status than sporadic 
performance testing.
    Under today's rule both sources and potential enforcers will be put 
on the same evidentiary footing in an enforcement action. Further, 
since 1992, EPA's Part 70 operating permit regulations have allowed the 
use of this data in compliance certifications. Today's action reaffirms 
this approach, and removes any potential ambiguity regarding the use of 
such data for this purpose.
    Today's action reflects EPA's efforts to make existing regulatory 
programs work better rather than creating additional requirements. By 
ensuring greater compliance with existing emissions limits, the 
credible evidence revisions will help minimize the need for further 
requirements to achieve air quality goals. See the October, 1993, 
proposal, 58 FR 54654.

C. Public Participation

    The final credible evidence revisions were developed with the 
benefit of insight from many parties that will be affected by the 
regulations, including State and local air pollution control agencies, 
large and small industries, trade associations and environmental 
organizations. Many comments regarding credible evidence issues were 
received during the development and after the proposal of the original 
enhanced monitoring rule, in 1991 through 1995. Many additional 
comments were received after the Agency announced that it was 
continuing to go forward with the credible evidence revisions in 1996.
    To obtain the views of all interested parties at the early stages 
of developing the enhanced monitoring rulemaking, EPA published a 
notice in the Federal Register on August 8, 1991, to make available a 
Public Information Document on enhanced monitoring and to provide 
notice of a public meeting to be held on August 22, 1991, on the 
subject (56 FR 37700-37701, August 8, 1991). In response to the public 
meeting, EPA received many comments which were included in the docket 
for the proposed regulations.
    Over the next four years, EPA held over one hundred informal 
informational and discussion sessions with representatives of 
interested organizations to receive their views on enhanced monitoring, 
as well as a second informational meeting with approximately fifty 
attendees held on August 12, 1993. Following publication of the 
proposed enhanced monitoring regulations on October 22, 1993 at 58 FR 
54648, EPA conducted a public hearing in Washington, D.C., on November 
19, 1993. Testimony was given by twelve individuals, representing 
industry and environmental organizations.
    In addition, during the public comment period, which was first 
scheduled to close on December 30, 1993, and was extended until January 
31, 1994, in response to requests for

[[Page 8316]]

extension, EPA received comments from a wide variety of interested 
parties concerning the enhanced monitoring proposal, including numerous 
comments on credible evidence issues. In the fall of 1994, EPA held a 
series of informational meetings with interested parties affected by 
the rule. The Agency then reopened the public comment period on 
specific issues to solicit additional comments, and held an additional 
stakeholder meeting. In response to the reopened public comment period, 
EPA received over 200 additional comment letters.
    In April, 1995, EPA announced that it was suspending development of 
the enhanced monitoring rule while it developed the CAM approach to 
serve the same statutory goals. In a September, 1995, public draft of 
the CAM approach, EPA stated that it would hold further discussions 
with stakeholders before it proceeded to finalize the credible evidence 
revisions. On March 8, 1996, EPA announced that a public meeting on 
credible evidence issues would be held on April 2, 1996. To focus the 
meeting's discussion, EPA released a paper on March 21, 1996, entitled 
``The Use of Information Other Than Reference Test Results for 
Determining Compliance With the Clean Air Act'' (sometimes referred to 
as the ``Credible Evidence White Paper''). EPA distributed this paper 
by electronic bulletin board to the same stakeholders who were involved 
in the enhanced monitoring and CAM rulemakings, further distributed it 
to various other interested parties, and made it generally available to 
the public.
    The public meeting was held on April 2, 1996, where twenty-three 
organizations and individuals presented oral statements and written 
comments. At the meeting, EPA announced that, although the rulemaking 
docket would not formally be re-opened, additional written comments 
would be accepted for at least another 30 days. Moreover, EPA stated 
that it would meet with any interested parties to discuss the credible 
evidence rules. As a result, many additional written comments have been 
received, and numerous additional EPA/stakeholder meetings have been 
held.
    Section III of this preamble contains a description of the most 
significant public comments and EPA's responses to them. Summaries of 
other public comments on the credible evidence revisions received over 
the past five years, together with the Agency's responses, are 
available in the docket in a document entitled ``Credible Evidence 
Revisions: Detailed Response to Comments Document'' (referred to in 
this preamble as the ``Detailed Response Document'').

II. Summary of Final Rule

    The credible evidence revisions consist of various changes to 40 
CFR 51.212, 52.12, 52.30, 60.11 and 61.12. These revisions provide 
minor modifications to existing regulatory provisions to clearly allow 
for the use of any credible evidence--that is, both reference test and 
comparable non-reference test data--to prove or disprove violations of 
the Act in enforcement actions. These revisions make clear that 
enforcement authorities can prosecute actions based exclusively on any 
credible evidence, without the need to rely on any data from a 
particular reference test. The revisions also have the effect of 
eliminating any potential ambiguity regarding the use of non-reference 
test data as a basis for Title V compliance certifications. The 
credible evidence revisions do not call for the creation or submission 
of any new emissions or parametric data, but rather address the role of 
existing data in enforcement actions and compliance certifications. As 
such, today's final action is distinct and separable from the bulk of 
the proposed enhanced monitoring rule, which addressed new monitoring 
requirements.
    By clearly providing that federally approved SIP test methods or 
Agency reference test methods are not the exclusive means of 
establishing noncompliance or compliance, EPA in no way intends to 
alter the underlying emission standards. The Agency will still use the 
reference methods for exactly what they are: test methods of reference 
against which to compare information generated by means other than the 
reference tests. The National Bureau of Standards maintains a number of 
standards against which other measuring devices, used in scientific or 
commercial applications, are calibrated. Similarly, where a SIP, New 
Source Performance Standard or permit specifies EPA Method 25A, for 
example, for determining the amount of volatile organic compounds 
(``VOCs'') that are emitted, the ``other evidence'' that could 
establish compliance would have to relate to the likely measurement of 
VOCs that would be obtained by a Method 25A measurement. This could 
include, for example, consideration of key operating parameters for the 
facility as correlated with emissions during a Method 25A test.

A. 40 CFR Part 51, Sec. 51.212

    Section 51.212(c) is revised to clarify that the inclusion in a 
state implementation plan (SIP) of enforceable test methods for SIP 
emissions limits does not preclude enforcement based on other credible 
evidence or information, relevant to whether a source would have been 
in compliance with applicable requirements if the appropriate 
performance or compliance test procedures or methods had been 
performed. This revision does not affect the existing requirements in 
Secs. 51.212(a) and (b) for periodic testing and inspections, and 
establishment of a system of violation detection and investigation.
    The proposed revisions to Sec. 51.212 contained detailed lists of 
``presumptively credible evidence'' and ``presumptively credible 
monitoring methods.'' After consideration of public comments, EPA has 
decided to delete these lists because they are potentially confusing 
and unnecessary. While EPA continues to believe that the listed 
evidence and monitoring methods are indeed credible, the Agency 
recognizes that both judicial and administrative tribunals routinely 
make determinations concerning the admissibility and weight of evidence 
on a case-by-case basis.

B. 40 CFR Part 52, Sec. 52.12

    Section 52.12(c) is revised to clarify that, for purposes of 
federal enforcement, any credible evidence relevant to whether a source 
would have been in compliance with applicable requirements if the 
appropriate performance or compliance test procedures or methods had 
been performed may be used to establish whether or not SIP violations 
have occurred. As with Sec. 51.212 above, EPA has deleted the proposed 
lists of presumptively credible evidence and monitoring methods for the 
same reasons stated above. Under today's final action, where an 
emission limitation specifies a particular monitoring or testing method 
approved by EPA for use in the SIP to determine compliance, data from 
such method will continue to be the benchmark against which other 
emissions or parametric data, or engineering analyses, will be 
measured. Similarly, where there are no approved SIP methods, the test 
methods specified in part 60 of this chapter will remain the standard 
against which other such information will be evaluated.

C. 40 CFR Part 52, Sec. 52.30

    Proposed Sec. 52.30(a), which concerned compliance certifications, 
has been revised in accordance with Sec. 51.212 above, and the same 
comments apply. The enforcement-related Sec. 52.30(b) is rendered 
unnecessary by today's final Sec. 52.12(c), which effectively

[[Page 8317]]

encompasses it. Finally, the entire section has been renumbered as 
Sec. 52.33.

D. 40 CFR Part 60, Sec. 60.11

    Similar to the existing regulation, Sec. 60.11(a) states that 
compliance with Part 60 standards shall be determined in accordance 
with the applicable performance tests and performance testing 
provisions in this part. A new Sec. 60.11(g) clarifies that nothing in 
Sec. 60.11 precludes the use, including exclusive use, of any credible 
evidence or information, relevant to whether a source would have been 
in compliance with applicable requirements if the appropriate 
performance or compliance test or procedure had been performed, for 
purposes of submitting compliance certifications or establishing 
whether or not a source has violated or is in violation of any Part 60 
standard, including opacity standards.
    The first sentence in today's final Sec. 60.11(a) has been modified 
from the proposal. EPA has decided to use mandatory phrasing in the 
first sentence (``Compliance with standards * * * shall be determined 
in accordance with the applicable performance tests * * *'') as is 
included in the existing regulation, rather than adopt the permissive 
language proposed in 1993 (``Compliance with standards * * * may be 
determined by performance tests * * *). The rationale for retaining 
this mandatory language is to make clear that, although the regulation 
is being modified to clarify that it does not establish an exclusive 
method of determining compliance, the reference tests remain the 
benchmark against which other emissions or parametric data, engineering 
analyses, or other information will be evaluated. For similar reasons, 
EPA included in Sec. 60.11(g) the requirement that evidence or 
information gathered by other means than the reference tests be 
``relevant to whether a source would have been in compliance with 
applicable requirements if the appropriate performance or compliance 
test or procedure had been performed''. This phrase means that the 
evidence or information must bear on whether a facility would have been 
found to be in compliance, during the time period in question, if the 
appropriate performance test had been conducted. It does not mean that, 
to prove a violation occurred, ideal testing conditions, for example 
the sun light at a certain angle to the tester for an opacity reading, 
must exist if other credible evidence, such as continuous opacity 
monitor data, can establish that a violation occurred. These changes 
have been made in response to comments that EPA's proposal did not give 
full recognition to the role of reference tests in determining 
compliance with emission standards. Section 60.11(g) combines the 
requirements of the proposed subsections (g) and (h) with the exception 
of presumptions included in those sections which have been deleted. The 
clarifying language in Sec. 60.11(g) renders unnecessary the previously 
proposed language in Sec. 60.11(b). Accordingly, the proposed language 
for that subsection is deleted from today's rule. The proposed changes 
to subsection (e) have been deleted as unnecessary due to changes to 
subsections (a) and (g). Finally, Sec. 60.11(f) is revised so as to 
clarify that it does not countermand subsection (g).
    Under today's revisions, information generated from an appropriate 
and properly conducted test method established under the general 
provisions of Part 60 or in the applicable subpart will still generally 
be the best method for determining a source's compliance during the 
test period. Other emissions or parametric data, or engineering 
analyses, may be considered if relevant to the results that would have 
been obtained by the appropriate, properly conducted reference test 
methods.

E. 40 CFR Part 61, Sec. 61.12

    Today's revisions to Sec. 61.12 generally mirror the revisions to 
Sec. 60.11, largely for the same reasons. Section 61.12(b) remains 
unchanged from its current promulgated version because credible 
evidence has always been used to establish violations of these 
standards.

III. Major Issues

    Throughout the development of this rulemaking, various commenters 
have expressed concerns regarding the proposed rule's potential effects 
on CAA enforcement, compliance certifications and emissions standards. 
The most significant of these comments, together with EPA's responses, 
are discussed below.

A. Use of Credible Evidence in Enforcement Actions

    Commenters raised various concerns regarding the potential use of 
credible evidence in enforcement actions. Some commenters argued that 
the use of such evidence would be unconstitutional, unprecedented and 
unfair. Others expressed concern that EPA, states or citizen groups 
would use credible evidence to bring enforcement actions for 
insignificant violations. These comments are addressed below.
    Industry commenters have argued that the use of credible evidence 
in enforcement actions would violate sources' constitutional right to 
due process. Specifically, the commenters argue that EPA must 
comprehensively identify the precise types of information that can be 
used as credible evidence, or else sources will not have sufficient 
``fair warning'' regarding potential enforcement. EPA rejects this 
view. ``Fair warning'' jurisprudence holds that regulated sources must 
have adequate notice identifying ``the standards with which the agency 
expects parties to conform.'' General Electric Co. v. U.S. EPA, 53 F.3d 
1324, 1329 (D.C. Cir. 1995). Today's rule does not establish or alter 
standards with which sources regulated under the CAA must comply. 
Rather, today's rule only concerns the evidence that can be used to 
prove violations of a standard, giving full recognition to the role of 
reference test methods under the standards. The Federal Rules of 
Evidence govern the admission of evidence in all federal district court 
litigation, including CAA enforcement actions, without any discernible 
constitutional infirmity. Similar evidentiary rules govern federal 
administrative and state environmental actions. Our legal system 
provides that a federal or administrative law judge will be the 
ultimate, independent arbitrator of the evidence's admissibility and 
credibility.
    Credible evidence is far from a new concept in judicial and 
administrative actions. In private lawsuits such as contract disputes, 
and in governmental and citizen enforcement actions brought under 
environmental laws other than the CAA, litigants can and do use a wide 
variety of information to prove their claims, or to refute the claims 
of opposing parties. In all these lawsuits, the judge acts as the 
final, independent arbitrator of what constitutes credible and 
admissible evidence. Today's final rule addresses problems arising from 
certain CAA regulations, which predate the 1990 Amendments to the CAA, 
containing language that has been read to allow only a very limited 
amount of information, i.e., data from reference test methods, to be 
used as evidence of violations. As such, the rule merely corrects an 
anomaly that has been read into these regulations, and brings their 
potential enforcement into line with that of other CAA requirements 
such as the ``general duty obligations'' in 40 CFR 60.11(d) (for NSPS 
standards) and 40 CFR 61.22(c) (for National Emission Standards for 
Hazardous Air Pollutants (NESHAPs)), and with other environmental 
statutes. It should be emphasized that the determination that evidence 
or information is credible is merely a threshold determination that

[[Page 8318]]

the evidence or information in question is technically relevant, and 
therefore, legally admissible in an enforcement action. In light of 
section 113(a) providing that the Administrator may bring an 
enforcement action based on ``any information'', EPA believes that 
Congress intended this threshold to be a low one.
    Industry commenters have also argued that using credible evidence 
in enforcement actions is unfair because sources will not know what 
credible evidence may be used against them. EPA believes that this 
claim lacks merit. This issue is no different in CAA enforcement than 
in any civil or criminal matter resolved by our nation's courts. 
Further, EPA disagrees with the notion that sources will likely be 
faced with an unknown and unlimited array of evidence. To the contrary, 
with regard to sources subject to Title V permits, EPA generally 
expects that most if not all of the data that EPA would consider as 
potentially credible evidence of an emission violation at a unit 
subject to monitoring under the agency's proposed CAM rule would be 
generated through means of appropriate, well-designed parametric or 
emission monitoring submitted by the source itself and approved by the 
permitting authority, or through other requirements in the source's 
permit. Sources not subject to CAM should still be readily able to 
discern the information, for example information about the operation of 
pollution control devices, that is relevant to their compliance with 
applicable regulation.
    Some industry representatives have expressed concern that the use 
of credible evidence in compliance determinations will reveal multiple 
minor violations for which EPA, the states or citizens will bring 
lawsuits. It is not EPA's intent to foster frivolous lawsuits, and EPA 
does not expect that such lawsuits will occur as the result of today's 
action. As EPA explained in the Credible Evidence March 1996 
memorandum, EPA generally focuses its judicial enforcement resources on 
violations that (1) may threaten or result in harm to public health or 
the environment, (2) are of significant duration or magnitude, (3) 
represent a pattern of noncompliance, (4) involve a refusal to provide 
specifically requested compliance information, (5) involve criminal 
conduct, or (6) allow a source to reap an economic windfall. See March 
1996 Memorandum, p. 5.
    An examination of EPA's judicial enforcement cases over the past 
few years reveals that EPA has focused its judicial enforcement 
resources on large, significant cases rather than a large number of 
relatively minor matters. The Credible Evidence March 1996 memorandum 
contains several examples that illustrate this point. In contrast, 
EPA's approach to minor unexcused violations generally has been to 
exercise prosecutorial discretion and use tools such as notices of 
violation and administrative compliance and penalty orders. In every 
case, EPA considers the nature and extent of the violation and all 
other circumstances surrounding the violation in determining whether 
and what kind of enforcement response is appropriate. Further, for any 
type of noncompliance, EPA generally will not bring a federal 
enforcement action where a state or local permitting authority has 
taken timely and appropriate action under existing policies to resolve 
the violations. Finally, for all violations, EPA will apply all other 
existing specific enforcement policies, such as the May, 1996, Policy 
on Compliance Incentives for Small Businesses, in accordance with their 
terms. EPA does not intend to use credible evidence to change any of 
these policies.
    EPA has a balanced enforcement program that seeks to assure 
compliance using the mix of the compliance and enforcement tools 
available to it. Deterrence is also an overall goal of the program. 
Judicial enforcement against minor CAA violations generally is a lower 
enforcement priority, because EPA believes its other enforcement and 
compliance assistance tools allow it to respond to such violations 
without the need to file an action in federal court. Accordingly, in 
considering whether to bring a judicial action, or whether to use some 
other enforcement or compliance tool, EPA generally takes into 
consideration such factors as number and duration of the exceedances, 
harm or risk posed by the exceedance, potential for recurrence, the 
source's compliance history, and other circumstances surrounding the 
violation. For example, if a source were installing a new unit subject 
to an NSPS standard and had some difficulty getting the control 
equipment to operate properly after the ``shakedown'' period permitted 
before the initial performance test (see 40 CFR 60.8(a)) but solved the 
problem promptly after the test, this generally would be a low 
enforcement priority, absent other circumstances indicating a need for 
judicial action.
    These same general policies regarding EPA's use of judicial and 
administrative enforcement actions were discussed in Section I.D. of 
the August 2, 1996, CAM draft approach. Therein, EPA provided various 
specific examples of circumstances where the Agency was or was not 
likely to take compliance or enforcement action based on the 
examination of CAM data.
    Finally, the NSPS general provisions and many SIPs generally excuse 
sources from compliance with emissions limits during periods of 
startup, shutdown or malfunction. See 40 CFR 60.11(c). Some specific 
NSPS standards additionally excuse sources from compliance during 
certain operating periods. Exceedances monitored during any of these 
specifically excused periods are not violations of the emission limit. 
Moreover, some NSPS standards specify averaging periods for determining 
compliance and noncompliance. As a result, many short term emissions 
values when averaged with other values in the relevant averaging 
period, will not constitute violations. The credible evidence proposal 
does not change any of these general or specific periods of excused 
noncompliance, or any averaging periods, or any of their effects on 
compliance.
    Regarding citizen suits, in February, 1996, EPA performed a review 
of citizen enforcement actions under the Clean Water Act (CWA), and 
found that citizen enforcers generally do not focus on sporadic, 
inconsequential violations. This analysis was summarized in the 
Credible Evidence White Paper, and is included in the Air Docket. 
Although to date there have been far fewer CAA citizen suits than CWA 
citizen suits, there have been at least two notable CAA citizen cases 
involving serious violations: National Wildlife Federation v. Copper 
Range Co., Civil Action No. 2:92-CV-186 (W.D. Michigan), involving one 
of the largest sources of particulate matter in Michigan's Upper 
Peninsula, which was emitting particulates at 230 lbs/hour (over five 
times its permitted limit) and toxic air pollutants including mercury, 
arsenic, cadmium and lead; and Sierra Club v. Public Service Company, 
894 F. Supp. 1455 (D.C. Col. 1995), involving a power plant that had 
committed over 19,000 opacity emission violations, which had allegedly 
affected a nearby wilderness area. Both of these suits were ultimately 
settled (with the United States an intervenor) for multi-million dollar 
penalties and significant injunctive relief, including the installation 
of appropriate pollution controls.
    EPA notes that today's rule creates no new rights or powers for 
citizen enforcers; instead, the rule clarifies existing EPA 
regulations. Citizens have been free to use credible evidence in Clean 
Air Act enforcement, and have won at least two court cases using it. 
See Sierra Club v. PSC, cited above, and Unitek Environmental Services 
v.

[[Page 8319]]

Hawaiian Cement, Civ. No. 95-00723 (D. Hawaii 1996). Also, EPA is aware 
of no increase in citizen suits in any of the five states--Kansas, 
Iowa, Nebraska, North Dakota and Georgia--whose SIPs, based on EPA's 
SIP Call, have specifically clarified that credible evidence can be 
used for enforcement, or in those states that have credible evidence 
provisions in other parts of their state law.
    Finally, EPA takes this opportunity to further elaborate on certain 
credible evidence and enforcement issues that were discussed in the 
August, 1996, draft CAM approach preamble. Therein, EPA explained that 
``the CAM rule cannot and does not replace a source's obligation to 
comply with otherwise applicable emission limits.'' Nonetheless, as a 
practical matter, ``EPA expects that a unit that is operating within 
appropriately established indicator ranges as part of an approved CAM 
plan will, in fact, be in compliance with its applicable limits.'' (See 
draft CAM rule Sec. 64.6(c), which requires that ``the ranges shall be 
established so as to provide a reasonable assurance of compliance with 
emission limitations or standards for the anticipated range of 
operations at a pollutant-specific emissions unit.'') Such a unit 
generally will not be an enforcement target. However, if the Agency 
obtains information that the unit is in fact exceeding its applicable 
emission limit even though it is operating within its approved 
indicator ranges, the Agency will consider whether or not to take 
compliance or enforcement action in accordance with its general 
enforcement policies. Further, under the CAM approach, the source has 
such information, it would have to promptly remedy the exceedance and 
notify the permitting authority and submit a proposed permit 
modification to correct its CAM monitoring as required under draft CAM 
rule Sec. 64.3(b)(5).
    Under today's rule, the legal burdens regarding the establishment 
of violations or compliance in an enforcement action are not changed. 
The means of meeting these burdens will vary in different 
circumstances. Today's rule provides that where information (such as 
non-reference emissions data, parametric data or engineering analyses) 
is equivalent to information generated by reference test methods, the 
former may be used to establish compliance or noncompliance in an 
enforcement action. There is no need to establish that every test 
condition specified in a reference test method has been matched by a 
surrogate condition in the method used to generate the comparable 
information. Typically, reference test methods (and any additional test 
conditions specified in individual regulations) quantify the presence 
of particular physical attributes--for example, mass or concentration 
of a chemical or group of chemicals--over a specified period of time. 
As long as these two elements--quantification and specified time 
period--are retained and the data from the alternate method is related 
to the reference test, information generated by alternate methods yield 
data bearing on what the results of a reference test would have been, 
and the use of such information to establish compliance or 
noncompliance in an enforcement action will not affect the stringency 
of the underlying standard. Of course, non-reference data that is 
already quantified in the same units as the underlying standard, e.g., 
emissions data generated by properly operating and calibrated non-
reference CEMs, should generally be comparable to reference test data, 
with all specified averaging periods still applying.
    For example, Method 9, the NSPS reference method for opacity, 
requires that a trained visible emissions observer (VEO) view a smoke 
plume with the sun at a certain angle to the plume in order to properly 
illuminate it. In contrast, a continuous opacity monitor (COM) contains 
a calibrated light source that provides for accurate and precise 
measurement of opacity at all times. Notably, EPA uses COM data to 
certify and re-certify the credentials of VEOs under Method 9. 
Accordingly, since a comparable light source is provided by a COM, if 
COM data were offered in an enforcement action to prove or disprove 
opacity violations, there would be no need to establish that the sun 
was shining during the period the COM data was collected. Where a 
reference test method or test requirements in an individual regulation 
include plant operating conditions, e.g., a requirement that testing be 
conducted at a specified percentage of maximum plant capacity, this 
does not mean that the underlying standard applies only when the plant 
is operating at that capacity or that the ``other information'' would 
have to show that the plant was operating at the specified capacity 
during the period that the other ``credible evidence'' was obtained.
    Where a party seeks to introduce other sorts of information in an 
enforcement action, for example, expert testimony as to whether a unit 
was able to meet its emission limit based on the operation or 
nonoperation of its control equipment during the period of alleged 
violation, the information would still need to be relevant to reference 
test data in the sense that it must be related to reference test data 
in some fashion. In the expert testimony example, this might be 
accomplished by a qualified expert opinion that a reference test would 
have demonstrated noncompliance in these same circumstances. Finally, 
where general burdens of proof for the proponent of this information 
are reduced through statutory provisions or other means, the same 
reduced burdens will apply in circumstances where EPA uses non-
reference test data to assert noncompliance. See, e.g., CAA section 
113(e)(2).

B. Use of Credible Evidence in Compliance Certifications

    Some commenters argued that today's final action will create new 
uncertainties and burdens for sources, because sources will not know 
what information they must consider before certifying compliance with 
Title V permit requirements. Previously, these commenters argue, 
sources would have needed to consider only the results of any specified 
reference tests, whereas under the credible evidence revisions almost 
any information could be potentially relevant to determining 
compliance. Thus, as a practical matter sources would need to ``go 
through every file drawer'' and examine a great deal of additional 
information before certifying compliance. Even then, sources would not 
know whether they had reviewed all compliance information that was 
potentially credible. According to some commenters, even if the source 
determined its compliance using a reference method, the source would 
still be uncertain as to whether it could certify compliance during 
that period, because other contemporaneous information might still 
indicate noncompliance. Still other commenters argue that allowing a 
broad array of information to be considered in compliance 
certifications would render the certification requirement void for 
vagueness.
    At the outset, EPA notes that today's action merely eliminates any 
potential ambiguity or conflict between Parts 51, 52, 60, and 61 and 
Part 70 regarding the ability of sources to use non-reference test data 
in compliance certifications. Consistent with the congressional intent 
reflected in Title V and section 114(a)(3), Part 70 already 
contemplates use of non-reference test data in compliance 
certifications. There are other pending rulemakings--specifically, 
pending actions involving the CAM approach and Part 70--that are

[[Page 8320]]

proposing to modify existing Part 70 requirements to provide additional 
detail as to what information sources must consider when certifying 
compliance. Nothing in these rule revisions is meant to specify what 
degree of correlation there must be between CAM monitoring data and 
emissions violations or compliance certifications; rather this issue 
will be discussed in the CAM rulemaking.
    In addition, EPA believes that the commenters have greatly 
exaggerated the purported uncertainties and burdens in certifying 
compliance under Part 70 and notes that facilities routinely determine 
their compliance with numerous statutory or regulatory obligations 
without government imposed ``checklists.'' Under Title V, the source's 
substantive CAA obligations (i.e., the source's applicable 
requirements) are clearly set forth in the source's CAA operating 
permit.
    Contrary to the commenters' claims, sources that are certifying 
compliance using properly conducted continuous reference methods may 
generally certify compliance based solely on the continuous reference 
method data, although naturally such sole reliance would be 
inappropriate in the face of obvious contrary information or fraud as 
discussed below.
    Of course, if a source becomes aware of other material information 
that indicates that an emission unit has experienced deviations (as 
that term is defined in the draft CAM approach) or may otherwise be out 
of compliance with an applicable requirement even though the unit's 
permit-identified data indicates compliance, the source must consider 
this information, identify and address it in the compliance 
certification, and certify accordingly. This ensures, among other 
things, that sources will not certify compliance in circumstances where 
doing so would constitute a violation of CAA section 113(c) and 18 
U.S.C. Section 1001, which prohibits sources from knowingly making a 
false certification or omitting material information, or a violation of 
other prohibitions on fraud. EPA emphasizes, however, that its purpose 
here is to make clear that sources may not ignore obvious relevant 
information. EPA does not view compliance certification requirements as 
imposing a duty on the source to search out and review every possible 
document to determine its relevance on the issue of the source's 
compliance.
    Following on the above discussion, the Agency takes this 
opportunity to restate that while a Title V permit can include a 
``permit shield'' protecting it from allegations that it has failed to 
satisfy CAA monitoring requirements, such shield does not relieve the 
source of its obligation to comply with the underlying emission limits 
or other applicable requirements being monitored. In other words, even 
where a source receives a ``shield'' providing that the monitoring 
provisions set forth in its Title V permit constitute compliance with 
all monitoring requirements of the CAA, the source would not be 
shielded from allegations of noncompliance with the underlying 
substantive requirements (e.g., emission limits) being monitored even 
if the source's required monitoring failed to detect the violation. See 
also the October, 1993, proposal, 58 FR 54678.
    Industry commenters argued that allowing credible evidence in Title 
V compliance certifications would render the certification requirement 
constitutionally void for vagueness. According to these commenters, 
reference test methods are necessary to define, in a consistent and 
reproducible manner, the level of performance that constitutes 
compliance; without a reference method, an emission limit would be 
incomplete. As discussed above, EPA in no way intends to eliminate 
reference tests or to alter their methodology. Instead, these tests, 
performed as specified under EPA and state regulations, will remain the 
benchmark against which to compare other emissions or parametric data, 
or engineering analyses, regarding source compliance.
    Finally, numerous commenters argued that allowing credible evidence 
in compliance certifications and enforcement actions would disrupt the 
Title V permit process and cause substantial delays in the issuance of 
these permits because local permitting authorities would have to adjust 
many of the sources' emission limits, which the commenters contend were 
not intended to be complied with continuously. Such Title V gridlock 
could occur only if today's action in fact changed the stringency of 
emission standards.

C. EPA's Authority To Promulgate the Credible Evidence Revisions

1. Statutory Authority
    Today's rulemaking and related SIP call are based primarily on 
EPA's existing authority prior to the 1990 CAA Amendments. Section 
113(a) of the Act authorizes EPA to bring an administrative, civil or 
criminal enforcement action ``on the basis of any information available 
to the Administrator.'' This provision provides the Agency with clear 
statutory authority to use any available information to prove 
violations of requirements under the Act, and demonstrates that 
Congress did not intend to limit EPA to using reference test method 
results in bringing enforcement actions. The language of Section 
113(a), together with the fact that the Act nowhere prohibits the use 
of information other than reference test results to prove violations, 
indicates that the Act does not limit the use of any information to 
prove a violation. Therefore, by law the Agency is limited only by 
general evidentiary rules in what it can use to prove a violation 
alleged in an enforcement action.
2. The Kaiser Steel Decision Does Not Constrain EPA's Authority To 
Amend Its Regulations
    Although the Act sets no inherent limits on EPA's authority to use 
any type of information to prove a violation, some EPA regulations 
provide for specific test methods for determining compliance and have 
been read by some to constrain EPA's enforcement authority. In United 
States v. Kaiser Steel Corp., No. CV-82-2623 IH (C.D. Cal. Jan. 17, 
1984), the district court construed the language of EPA's regulations 
at 40 CFR 60.11 as limiting the admissible evidence of violations of 
opacity standards to observations utilizing Method 9, the opacity 
reference test method. Thus, when the Agency attempted to use expert 
testimony pertaining to opacity to prove the existence of violations 
without Method 9 test data, the court rejected the evidence and held 
that EPA could prove violations only on those days where the Method 9 
test was conducted. This decision--which interpreted only EPA's 
existing regulations, not the Act--was specifically overruled by 
Congress in the 1990 CAA Amendments. Today's rulemaking is intended to 
clarify that EPA's regulations do not constrain EPA to using reference 
tests to prove a violation of an emission standard. Rather, EPA retains 
its full authority under Section 113(a) to use ``any information'' as 
the basis for an enforcement action.
3. Despite Commenters' Claims, Clean Air Act Case Law Does Not Mandate 
Exclusive Reference Tests
    At least one commenter has asserted that the decision in Portland 
Cement Ass'n v. Ruckelshaus, 486 F.2d 375, 399 (D.C. Cir. 1973), cert. 
denied, 417 U.S. 921 (1974), stands for the proposition that CAA 
emission standards may be enforced only through an exclusive reference 
test method. First, the commenter relies on the court's ruling

[[Page 8321]]

that a reference test method must make measurements with ``reasonable 
accuracy'' and be ``objective.'' 486 F.2d at 401 & n. 103. Second, the 
commenter cited the court's concern with deviations between sampling 
methods used in gathering data to set an emission standard and sampling 
methods used in reference methods. The court stated that ``a 
significant difference between techniques used by the agency in 
arriving at standards, and requirements presently prescribed for 
determining compliance with standards [i.e., the reference method], 
raises serious questions about the validity of the standards.'' 486 
F.2d at 396. EPA disagrees with this reading of Portland Cement.
    These holdings, individually or together, do not support the 
conclusion that violations of an emission standard may only be 
demonstrated by an exclusive reference method. The court's statements 
regarding the reliability of reference methods were made in context of 
a challenge to an opacity standard. The industry petitioner argued that 
testing compliance with that standard, inspector observations, is 
inaccurate and therefore arbitrary. The court agreed that the evidence 
called the reliability of inspector observations into question and 
remanded to EPA for it to determine if there was a way to measure 
compliance with the standard with ``reasonable accuracy.'' In no way 
did the court imply that the opacity standard had to have an exclusive 
reference test but simply rejected the test EPA proposed to use as 
insufficiently supported.
    The Portland Cement court's discussion of a compliance method that 
differed from the test method used to develop the standard also lends 
no support to the conclusion that an exclusive test method is required. 
It is true that the court mentioned reference methods ``outlined by 
regulation.'' However, the mere description of an agency practice 
(here, the inclusion of a reference test in a regulation setting an 
emission standard) does not transform that practice into a statutory 
requirement. Moreover, the thrust of the court's remarks was to caution 
EPA that, where EPA has established by regulation a reference method 
for sources to demonstrate compliance, the best data EPA can put forth 
to show that a standard is in fact achievable is data generated by the 
reference method. The D.C. Circuit, however, has specifically rejected 
the assertion that standards can only be supported by reference test 
data. See National Lime Ass'n v. EPA, 627 F.2d 416, 446, fn.103 (D.C. 
Cir. 1980). None of this, thus, supports the commenter's claim that a 
standard's supporting data must be generated using the reference 
method, and its supposed corollary that only reference method data can 
be used to enforce the standard, especially where, as here, that other 
information must be related back to a reference test method. At best, 
the commenter's arguments would apply only in the context of an 
original standard-setting, where an emission limitation or other 
standard newly promulgated by EPA was being challenged on the basis 
that the standard's supporting data was inadequate. Today's rule sets 
no new emission or work-practice standards, and amends no existing 
ones.
    Thus, the commenter is mistaken. Neither of the two passages in 
Portland Cement cited by the commenter address whether exclusive 
reference tests are necessary, much less mandate establishment of such 
tests. Further, EPA regulations are inconsistent with the exclusivity 
argument of the commenter. For example, section 60.8(a) of Title 40 of 
the CFR provides a whole string of circumstances under which a source 
can alter or completely replace the reference test required by the 
regulation. Finally, today's final action regarding the use of non-
reference test data in enforcement is fully consistent with the court's 
requirement that reference testing be conducted in a nonarbitrary 
manner.
4. The 1990 CAA Amendments Further Support EPA's Authority
    Various provisions of the 1990 CAA Amendments provide additional 
support for EPA's position that reference tests are not the exclusive 
means of proving violations. As noted above, Congress specifically 
reversed the Kaiser Steel decision in Section 113(e) of the Amendments 
by providing that the duration of a violation may be established ``by 
any credible evidence (including evidence other than the applicable 
test method).'' The legislative history for this provision shows that 
Congress meant to clarify that in an enforcement action courts are not 
restricted to reference test method data, but may consider any evidence 
of violation or compliance admissible under relevant evidentiary rules. 
See S. Rep. No. 228, 101st Cong., 1st Sess. 1, 358 (1989) (``Senate 
Report''), reprinted in 1990 U.S. Code Cong. & Admin. News 3385, 3741 
(``Reprint'').
    Other provisions of the 1990 CAA Amendments also evidence 
Congressional intent that reference test methods should not be used as 
the exclusive means for assessing compliance with CAA emission limits. 
Most pointedly, the requirements in Section 114(a)(3) for enhanced 
monitoring and for compliance certifications based on a determination 
of whether compliance was continuous or intermittent presumes that data 
other than reference tests would be used for these purposes. As 
explained in the October, 1993, proposal, the use of non-reference test 
data is also consistent with the monitoring, compliance assurance, and 
compliance certification requirements in Sections 504(a), 504(c), and 
503(b)(2) of the Act. See 58 FR 54649-50. In addition, Section 504(b) 
of the Act grants discretionary authority to the Administrator to 
prescribe procedures and methods for monitoring, and provides that 
continuous emission monitoring systems need not be required ``if 
alternative methods are available that provide sufficiently reliable 
and timely information for determining compliance.'' In sum, Congress' 
repeated emphasis on providing reliable and timely compliance 
information is inconsistent with the notion that only data from 
infrequently performed reference tests is relevant to compliance 
certifications and enforcement actions.
5. Commenters' Attempts To Narrow the Scope of Sections 113(e) and 
113(a) Are Unpersuasive
    Several industry commenters have claimed that the legislative 
history of the 1990 CAA Amendments shows that section 113(e)(1) does 
not provide authority for today's final action. Additionally, these 
commenters have asserted that the section's legislative history upon 
which EPA has relied is ambiguous.
    In the October, 1993, proposal, EPA cited to the Senate Report's 
discussion of Section 113(e)(1). The Senate Report stated:

    This title of the bill enhances the ability of the Environmental 
Protection Agency * * * by making clear that the Agency may rely 
upon any credible evidence of violations in pursuing alleged 
violations.

Senate Report at 358, Reprint at 3741. The Report further explained:

    [T]he amendment clarifies that courts may consider any evidence 
of violation or compliance admissible under the Federal Rules of 
Evidence, and that they are not limited to consideration of evidence 
that is based solely on the applicable test method in the State 
implementation [plan] or regulation. For example, courts may 
consider evidence from continuous emission monitoring systems, 
expert testimony, and bypassing and control equipment malfunctions, 
even if these are not the applicable test methods. Thus, this 
amendment overrules the ruling in United States v. Kaiser Steel 
Corp., No. 82-2623 (C.D. Cal. January 17, 1984) to the extent that

[[Page 8322]]

the court in that case excluded the consideration of such evidence.

Senate Report at 366, Reprint at 3749. Finally, the Report notes that 
data from enhanced monitoring and compliance certifications ``will 
facilitate enforcement, due in part to the fact that such data and 
certifications can be used as evidence.'' Senate Report at 368, Reprint 
at 3751.
    The commenters, in turn, rely on the views of Senator Chafee 
regarding S. 1630, inserted into the Congressional Record at the time 
the legislation was introduced. Senator Chafee stated with regard to 
Section 113(e)(1):

    Subsection 113(e) also clarifies and confirms that once EPA 
establishes evidence of a violation using a formal test method, EPA 
can use other credible evidence to prove additional violations, or 
that violation has continued.

135 Cong. Rec. S 9650, 9655 (August 3, 1989).
    EPA believes that the best reading of the legislative history still 
supports its interpretation of Section 113(e)(1). First, there is no 
ambiguity in the Senate Report, the language of which unreservedly 
supports enforcement actions brought on the basis of non-reference test 
data. Second, EPA does not believe that Senator Chafee's floor 
statement outweighs the clear statement in the Senate Report. The 
Senate Report is a more authoritative reflection of congressional 
intent than a floor statement produced at the beginning of the 
legislative process.
    Various commenters also objected to EPA's reliance on Section 
113(a) as a basis for today's action. One commenter argued that Section 
113(a) does not preempt regulatorily specified reference test methods. 
Several commenters asserted that EPA's construction of Section 113(a) 
would render superfluous the new language in Section 113(e)(1) 
concerning credible evidence. These commenters claim that, under EPA's 
interpretation of Section 113(a), Congress could have ``fixed'' the 
Kaiser Steel decision simply by clarifying the scope of EPA's authority 
under Section 113(a).
    These various commenters have misunderstood EPA's interpretation of 
Section 113(a). EPA has not asserted that Section 113(a) preempts 
reference test methods. Rather, EPA believes that Section 113(a) 
provides authority to amend current regulations to make clear that data 
from reference test methods are not the exclusive means of establishing 
noncompliance or compliance in enforcement actions. Given this 
interpretation of Section 113(a), Congress's passage of Section 
113(e)(1) cannot be described as superfluous--particularly in light of 
the decision in Kaiser Steel.
6. EPA Can Promulgate the Credible Evidence Revisions Without 
Reproposal
    Several commenters have argued that finalization of the proposed 
changes in Parts 51, 52, 60 and 61 without first reproposing those 
changes violates the Administrative Procedure Act (APA), the CAA, and 
due process. The commenters' main argument is based on EPA's presumed 
change in course on implementing the requirement in Section 114(a)(3) 
concerning enhanced monitoring and compliance certification. As noted 
above, the changes to Parts 51, 52, 60 and 61 were proposed in the same 
rulemaking that proposed an enhanced monitoring and compliance 
certification program. Since that proposal, EPA has re-evaluated its 
approach to enhanced monitoring and has made publicly available and has 
sought comment on a revised approach--the CAM approach--for satisfying 
the same statutory goals as the original enhanced monitoring proposal. 
Some commenters contend that switching to CAM will fundamentally change 
their view of the proposed changes to Parts 51, 52, 60 and 61 because 
those proposed changes were evaluated only in terms of how they would 
be implemented under the October, 1993, proposal on enhanced 
monitoring. Until CAM is formally proposed, these commenters assert, 
they cannot give meaningful comments on the credible evidence 
revisions. Further, the commenters argue that the proposed revisions 
provided insufficient notice and opportunity to comment because EPA has 
not adequately defined the term ``credible evidence.''
    EPA believes today's rule has no procedural infirmities. EPA is 
today finalizing the enforcement-related portions of the proposal it 
made in 1993 with only minor changes.
    The commenters' claim that they cannot meaningfully comment on 
credible evidence revisions prior to proposal of the CAM approach is 
not well-taken for two reasons. First, EPA does not believe that any 
knowledge of the draft CAM approach is necessary to comment on today's 
rulemaking. In today's final rule, EPA has removed any presumptions 
regarding the credibility of any specific data. If and when the draft 
CAM approach is finally adopted, CAM data will be treated under today's 
rule like any other potential source of compliance information. Thus, 
knowledge of the draft CAM approach is not critical to commenting on 
this rulemaking. In any event, the nature of the draft CAM approach has 
been generally available in some detail since September, 1995--well 
before EPA renewed its request for comment on today's rulemaking. 
Further, EPA has sought and received additional comment on the 
enforcement consequences of the draft CAM approach by distribution of a 
revision of the CAM approach in August, 1996. The revised approach 
specifically discussed the relationship of the draft CAM approach and 
today's action.
    Second, the October, 1993, proposed rulemaking gave interested 
parties sufficient notice of the issues raised by the proposed changes 
to Parts 51, 52, 60 and 61. The Agency made clear that these revisions 
were designed to remove any potential ambiguity regarding the use of 
enhanced monitoring data in compliance certifications, and to clarify 
that any credible evidence of a violation of an emission standard was 
admissible to prove (or disprove) such a violation. See 58 FR 54677. To 
clarify that these credible evidence revisions extended beyond the data 
gathered under an enhanced monitoring program, EPA gave two specific 
examples of evidence collected outside the enhanced monitoring program 
that under the revised regulations could be used to prove a violation. 
See 58 FR 54676-54677. Thus, the October, 1993, proposal clearly put 
interested parties on notice that the credible evidence revisions were 
not merely an adjunct to the enhanced monitoring program. In fact, 
industry commenters on the October, 1993, proposal clearly understood 
the central issue posed by the proposed credible evidence changes, and 
they commented on it extensively. Today's final action promulgates 
revisions to existing regulations, and are not contingent upon future 
promulgation of the CAM approach or any other form of enhanced 
monitoring requirement.
    Neither is this rulemaking procedurally deficient for not providing 
an express regulatory definition of the term ``credible evidence''--a 
term which Congress itself inserted, without definition, into the Act. 
The issues of credibility, admissibility and weight of evidence have 
been exhaustively addressed by federal and state court evidentiary 
rules regarding evidence, and the thousands of cases decided under 
them. Today's final action defers to those regulations and makes clear 
that there are no bars in regulations under the CAA which prevent the 
use of evidence or information other than reference test methods in 
compliance certifications and enforcement actions. Of course, in 
judicial enforcement

[[Page 8323]]

proceedings, what evidence is credible and admissible will be 
determined by the court taking into account how the evidence was 
gathered and the specifics of the emission standard and any associated 
reference method.
    Finally, EPA believes that it has taken extensive steps, detailed 
in Section I.C. above, to ensure that the concerns of affected parties 
were fully aired. None of the additional public outreach actions that 
EPA undertook in 1996 were required by the APA or the CAA; instead, EPA 
undertook them voluntarily to ensure full input by interested parties 
regarding the credible evidence rules.

D. Stringency

    Industry commenters have presented several arguments in support of 
their position that this rulemaking requires sources to be in 
continuous compliance and thus would effectively increase the 
stringency of underlying requirements, including SIP limits and 
standards established by EPA under the NSPS and NESHAP programs.
    EPA believes that industry's arguments on this point are 
fundamentally wrong. It is not EPA's intent that these rules should 
increase the stringency of any applicable requirement. These rules do 
not do so because they maintain the focus of the compliance 
determination on whether or not the appropriate reference test would 
have shown a violation.
    The commenters' arguments regarding increased stringency are as 
follows: applicable requirements are accompanied by specified reference 
tests. Any departure from past practice regarding the use of these 
tests, including the use of other credible information to directly 
assess compliance, particularly on a more frequent basis, will 
inevitably change the results of an inquiry into the compliance status 
of any source compared to exclusive reliance on the infrequent 
performance of the reference tests. Therefore, industry argues, using 
credible evidence would change the underlying applicable requirements--
usually in a manner that makes them more stringent--without going 
through the necessary rulemaking procedures.
    Industry's argument hinges on the premise that adoption of an 
emission standard that includes a particular form of reference test--
one that is not required to be performed continuously as a matter of 
course--limits the compliance obligation. The scope of the compliance 
obligation is not at issue in this rulemaking. The scope of the 
compliance obligation prescribed by any particular standard shall be 
based on the emission standard and not this rulemaking. However, to 
fully respond to industry comments, and to give notice of the position 
EPA will take in future enforcement proceedings, EPA believes it is 
necessary to address in some detail the nature of the compliance 
obligation under emission standards with particular emphasis on the 
compliance obligation as it pertains to emission standards which have a 
reference test method that is not required to be performed 
continuously.
    While the bulk of the commenters' concerns were expressed with 
respect to NSPS, the same concerns also apply in most cases to NESHAPs 
and SIPs. Likewise, EPA's responses focus on NSPS, but are generally 
applicable to other emissions limits as well.
1. Emissions Limits Require Continuous Compliance (Consistent With Any 
Averaging Times) Except During Periods Where Compliance Is Specifically 
Excused
    To resolve commenters' claims of increased stringency, the nature 
of the compliance obligation facing owners and operators of sources of 
air pollution under the Act must be addressed. Under the CAA, its 
regulations, and the case law, a source's compliance with emission 
limitations must be continuous (consistent with any averaging times) 
except where a particular emission standard specifically provides for 
periods of noncompliance.
    The Statute. The Clean Air Act defines the terms ``emission 
limitation'' and ``emission standard'' as meaning ``a requirement 
established by the State or the Administrator which limits the 
quantity, rate, or concentration of emissions of air pollutants on a 
continuous basis * * *.'' CAA section 302(k) (emphasis added). In 
accordance with this clear statutory statement, the Act authorizes 
penalties for multiple days of violation should a source fail to meet 
its continuing obligation. See also CAA sections 113(e)(2) (providing 
that ``a penalty may be assessed for each day of violation,'' and 
establishing a presumption of continuing violation if certain 
conditions are met) and 113(e)(1).
    CAA Regulations. The Act's general requirement of continuous 
compliance is mirrored in the NSPS regulations, which generally require 
that sources comply with established emission limits except during 
certain defined time periods. NSPS provisions typically specify that 
compliance with stated limits is required ``on and after the date'' of 
an initial performance test conducted in accordance with 40 CFR 60.8. 
See, e.g., 40 CFR 60.502. The need for continuous compliance is also 
discussed in the preambles to numerous NSPS, including many older ones. 
For example, in proposing standards for glass manufacturing plants 
(Subpart CC), EPA stressed the need for effective monitoring to assure 
that affected facilities are ``continuing to maintain the emission 
reduction observed during the performance test.'' 48 FR 50670, 50675 
(1983). EPA has also made this point clear in publicly-available 
guidance memoranda. See Detailed Response Document at Section 4.
    In addition to requirements for continuous compliance, NSPS 
regulations also typically contain specifically excused periods of 
noncompliance. These periods confirm that compliance is required at 
other times. They also confirm the basic reasonableness of this 
compliance scheme--that is, sources must generally comply continuously 
with their numerical emission limits, but not during periods of 
specifically excused noncompliance, and only in accordance with any 
specified averaging periods. For example, for many standards, 
compliance is not required during periods of startup, shutdown or 
malfunction. This exception is contained in the NSPS general provisions 
and in individual standards. See 40 CFR 60.8(c); see also, e.g., 40 CFR 
60.46a.
    Case Law. In various judicial decisions, courts have approved of 
the basic NSPS regulatory scheme of continuous compliance accompanied 
by limited, specified exceptions for noncompliance. The courts have 
stated that the specified exceptions are needed because sources must 
comply at all other times. See, e.g., Portland Cement, 486 F.2d at 399 
(court noted EPA's then-proposed ``startup, shutdown and malfunction'' 
compliance exclusion regulation with approval, suggested that it was a 
``limited safety valve'' and stated that it imparts a construction of 
``reasonableness'' to the standards as a whole and adopts a more 
flexible system of regulation that can be had by a system devoid of 
``give''; (Essex Chemical Corp. v. Ruckelshaus, 486 F.2d. 427, 433 
(D.C. Cir. 1973), cert. denied, 416 U.S. 969 (1974) (in a challenge to 
sulfuric acid plant and coal-fired steam generator NSPS standards, the 
court again noted with approval the proposed start-up, shutdown and 
malfunction exception and remanded the rule stating that ``such variant 
provisions appear necessary to preserve the reasonableness of the 
standards as a whole and that the record does not support the ``never 
to be

[[Page 8324]]

exceeded'' standard currently in force'') (emphasis added); and Bunker 
Hill Co. v. EPA, 572 F.2d 1286, 1301-02 (9th Cir. 1977) (in challenge 
to SIP sulfur dioxide standard, court observed that EPA regulations 
required that the standard be met ``all of the time,'' and thus EPA 
must typically promulgate upset provisions to excuse noncompliance 
beyond the source's control). Similarly, the proposition that 
compliance must be continuous is reflected in numerous judicial 
decisions involving challenges to various NSPS rulemakings. In these 
cases, both the D.C. Circuit Court and industry petitioners have 
emphasized that for an emission standard to be achievable it must be 
able to be continuously complied with over wide operating ranges at 
varied facilities. See, e.g., Portland Cement, Essex Chemical, National 
Lime, and Sierra Club v. Costle, 657 F.2d 298 (D.C. Cir. 1981). In 
National Lime, for example, the lime industry's trade association 
itself complained that the data underlying the promulgated numerical 
emission standards were insufficient to show that the standards were 
``in fact achievable on a continuous basis.'' 627 F.2d at 430. In 
holding that EPA had not adequately demonstrated the achievability of 
the standards for the industry as a whole, the court explained that 
``to be achievable, we think a uniform standard must be capable of 
being met under most adverse conditions that can reasonably be expected 
to recur . . . .'' Id. at 431. In Sierra Club v. Costle, various 
electric utility companies challenged a particulate standard on the 
basis that ``the data reflect only short term performance while the 
standard requires long term continuous compliance.'' 657 F.2d at 377 
(emphasis added). This challenge was rejected by the court based on 
data showing that certain sources had ``consistently complied with the 
standard.'' Id. at 382.
2. Commenters' Advocacy of Noncontinuous Compliance Would Lead to 
Numerous Anomalies
    Some industry commenters have argued that numerous emissions 
limitations do not require continuous compliance or, alternatively, 
that ``continuous'' does not have the straightforward meaning suggested 
above. The commenters' argument centers on NSPS standards issued under 
CAA section 111. In the commenters' view, many such standards do not 
contemplate that facilities will operate in compliance on a continuous 
basis with stated emissions limits, but rather require only an initial 
demonstration of compliance with stated limits upon start-up or shortly 
thereafter. After an initial performance test, continuous compliance is 
required only with respect to operation and maintenance ``in a manner 
consistent with good air pollution control practice'' as specified in 
40 CFR 60.11(d). As to numerical emissions limits, commenters suggest 
that these must be met only on those infrequent occasions that a 
subsequent performance test is conducted. So long as any such 
performance test is passed, the source is in ``continuous'' compliance 
with numerical emissions limits without regard to whether its emissions 
in fact exceeded the numerical limit during the time between the tests, 
no matter how long that may be.
    EPA rejects this view of the nature of the obligation to comply 
with NSPS and other emission limits under the CAA. See Detailed 
Response Document. EPA and the courts have long held that emission 
limits must be complied with continuously, consistent with any 
associated averaging periods, except where a particular limit provides 
otherwise. Adopting the commenters' view of compliance would lead to 
numerous anomalies.
    In the April 2, 1996, public meeting and in follow-up written 
comments, several commenters argued that many reference test methods 
were selected specifically because they would only be performed 
infrequently--for example, on a yearly basis. These once a year tests 
would be proper for their associated emission standards, which in the 
commenters' view were intended to be complied with only 95% of the 
time. Specifically, performing a reference test once a year would yield 
``acceptable'' compliance results, because on average a source would be 
found out of compliance only 5% of the time--that is, in one in twenty 
tests, or once every twenty years. According to these commenters, 
testing for compliance more frequently would be unfair, because it 
would increase the likelihood that the source would be found out of 
compliance during periods where the standard itself contemplated 
noncompliance. In order to avoid being found in noncompliance, sources 
would have to continuously stay below their emission limits--which in 
these commenters' view would effectively increase the stringency of the 
emission standard.
    EPA disagrees with the commenters' notion that sources must meet 
their legal numerical air emission limits only seldomly. Further, EPA 
rejects as inconsistent with the Act and its underlying purposes the 
notion that sources can somehow be in routine ``compliance'' without 
staying within these limits on an ongoing basis. The fundamental goal 
of the CAA and the emission standards established under it, is to 
achieve clean air. Moreover, many emission standards, such as hazardous 
air pollutant standards under Section 112 and emission standards in 
State Implementation Plans designed to implement national ambient air 
quality standards, have a direct relationship to the protection of 
human health. Routine compliance with numerical emission standards is 
critical to achieving this goal. The commenters' view that such 
compliance is somehow not required would completely undercut these 
public health and safety goals.
    If the commenters' view was correct, any EPA or state targeting of 
a specific source by requiring the source to perform more frequent 
reference tests would be unfair and presumably illegal, because any 
such increased frequency in reference testing would destroy the 
delicate balance of frequent noncompliance and infrequent testing that 
the commenters claim is contemplated by the rules. Under this view, EPA 
and states might not be able to require an apparently violating source 
to conduct a previously unscheduled reference test, because it would 
improperly raise the source's chances of being found in noncompliance 
and thereby ``increase the stringency of the underlying standards.''
    The commenters'' argument is also inconsistent with the language, 
structure, and purpose of the CAA. For example, if the frequency of 
testing must be limited to meet the intent of the emission limits, to 
be fair to all sources EPA's regulations should have required that the 
tests be performed only at infrequent intervals. EPA's rules contain no 
such restrictions; rather, CAA section 114(a)(1)(D) grants EPA broad 
discretion to order reference tests whenever the Administrator deems it 
appropriate. Moreover, commenters'' argument is inconsistent with CAA 
section 113(e)(1), which even on its narrowest reading (note that EPA's 
reading is considerably broader) specifically provides for use of non-
reference test data to prove continuing additional days of violation 
after an initial violation is established by reference test data, and 
by CAA section 113(e)(2), which establishes a presumption of continuing 
violation after notice of the violation has been given to the source, 
provided that EPA can make a prima facie showing that ``the conduct or 
events giving rise to the violation are likely to have continued or 
recurred past the date of notice.'' This presumption continues until 
the violator ``establishes that continuous compliance has been 
achieved.''

[[Page 8325]]

Likewise, sections 114(a)(3) and 504(a)-(c) regarding enhanced 
monitoring and certification as to whether compliance is continuous or 
intermittent, and prompt reporting of deviations, are simply 
inconsistent with a regulatory regime that would require only 
occasional demonstrations of compliance with emission limits. Taken 
together, these provisions, represent a fundamental statutory rejection 
of the commenters' argument. See Detailed Response Document, Section 4, 
which discusses other reasons why these comments are without merit.
3. Comments Regarding Continuous Compliance Are Not Directed at Today's 
Action, but Rather at Underlying Emission Standards
    Industry commenters have argued that the quality and quantity of 
the data used in establishing emissions limitations, such as those 
under the NSPS and NESHAP programs, reflect a conscious decision by EPA 
that compliance with such standards would need to be demonstrated only 
periodically. It follows that requiring continuous compliance with 
stated limits at this juncture would effectively increase the 
stringency of the standards. As discussed above, EPA believes that the 
commenters' general arguments strain common sense. Commenters have 
pointed to various NSPS standards to support their views, but EPA finds 
these examples unpersuasive.
    In particular, commenters have pointed to the NSPS for kraft pulp 
mills, 40 CFR Part 60, Subpart BB, and for steam electric generators 
constructed between 1971 and 1978, Subpart D, as reflecting a general 
acknowledgment by EPA that national standards need not be complied with 
at all times. EPA believes that, to the contrary, Subparts BB and D and 
other cases demonstrate that where EPA intended to allow affected 
sources to exceed stated emissions limits, the standards in question 
expressly so provide. It is true that in the development of some NSPS 
and NESHAP standards, EPA was concerned with the limited number and 
distribution of test runs and the inherent variability in levels of 
emissions from even well-controlled facilities. Where appropriate, EPA 
addressed those concerns by adjusting the numerical value of the 
standard, providing excess emissions allowances and provisions for 
noncompliance during certain upset conditions, or through changes in 
averaging times. With other standards, EPA did not provide for any 
departure from the general requirement that compliance must be 
continuous. Examples of all these approaches, and specific responses to 
comments regarding Subparts D and BB, are provided in the Detailed 
Response Document.
    The commenters' assertions that sources cannot comply on a 
continuous basis are really directed not to the propriety of today's 
rules, but rather to the adequacy of the underlying NSPS and other 
emission standards that are not at issue in this rulemaking. To the 
extent there is any documentation that a well-run facility cannot 
comply consistently with underlying national emission standards, or 
applicable SIP requirements, such documentation would be relevant only 
to those existing standards, not to today's rule. EPA notes that 
despite several requests to commenters to identify any standards that 
cannot be complied with on a regular basis, no specific information has 
been provided to this rulemaking docket that demonstrates that well 
operated and maintained facilities employing pollution control 
technologies of the types upon which the underlying emission standards 
were based cannot comply with those standards on a continuing basis. 
The most that was submitted was a statistical re-analysis of the data 
relied upon by EPA in promulgating several emission standards and a one 
page graph purporting to show that an industrial boiler could not 
comply with the NOX emission limit at low levels.
    The agency has considered this comment concerning the Subpart D 
NOX standard carefully, as it does not intend to impose 
requirements that are impossible for well-designed sources to meet, but 
believes that this concern is largely theoretical. The information 
provided by the commenter to EPA was vague and did not prove that the 
undisclosed source could not comply with the emission standard. 
Further, if a standard was impossible to achieve under some 
circumstance, EPA and citizens are not likely to bring enforcement 
cases in such instances. In reviewing CAA enforcement actions the 
agency has been unable to identify any case where either the agency or 
a citizen sought to enforce a standard that was impossible to achieve. 
The agency was also unable to identify any case in which a defendant 
established that compliance was not possible at the time of the alleged 
violation. This appears to be the case even in those states and 
localities that have had ``credible evidence'' rules for years.
    Additionally, should it be determined that a standard could not be 
met during some relatively infrequent or inconsequential period of 
source activity, the potential for significant adverse impact on that 
source is remote. The agency has previously expressed its policy that, 
generally, judicial enforcement is not the appropriate vehicle to 
redress sporadic, infrequent violations with no environmental 
consequence. Further, it is unlikely that a citizen could prevail in 
enforcing a theoretically impossible standard since Courts will not 
issue an injunction where there is nothing to be done. Similarly, where 
one cannot establish that a source failed to act in a manner required 
by law a significant penalty will not be imposed by the courts. The 
agency is not aware of any situation in which it has filed, and one 
should not anticipate large numbers of citizen suits being filed, where 
there is nothing the source could have done or could do to achieve a 
greater degree of compliance. Moreover, the courts today have 
additional tools, including fee awards and sanctions available under 
the Federal Rules of Civil Procedure and other statutes to address 
meritless suits.
    In further response to these industry comments, EPA has included in 
the record a 1993 study conducted by EPA Region V that shows that 
almost all (95%) of sources with sulfur dioxide CEMs were meeting their 
federal and state sulfur dioxide emission limits approximately 97% of 
the time, with excess emission periods totaling only 3%. See Region V 
Study, Figure 2. Because this 3% figure included excess emissions 
recorded during periods in which compliance is specifically not 
required, such as startup and shutdown, the percentage of operating 
time in noncompliance with the standard is even smaller and may mean 
that most sources are in compliance all the time. EPA Region V sources 
with continuous opacity monitors showed similar results: the average 
source's percentage of opacity exceedances was less than 2%, with 95% 
of sources at or below approximately 4%. See Study, Figure 1. As with 
the sulfur dioxide data, opacity exceedances during periods of startup, 
shutdown and other excused periods were not excluded. Accordingly, the 
percentage of actual noncompliance with opacity limits was even 
smaller. Note that these figures are for the average (50th percentile) 
and worst (95th percentile) facilities. The best run facilities have 
fewer excess emissions reports.
    Additional CEM data from EPA Region V that focused specifically on 
exceedances from NSPS Subpart D SO2 emission standards shows 
similar results. This data shows that Subpart D sources report few or 
no excess SO2 emissions. Approximately two-thirds of the sources 
report no excess emissions

[[Page 8326]]

at all, during any three month reporting period. Further, since 1990, 
the vast majority of sources (95%) have reported total excess emissions 
averaging less than 2.5% of operating time; since 1993, less than 1.7%. 
Since these figures include all excess emission periods, including 
periods that are probably excused, the actual SO2 exceedance rates 
were even lower.
    These data show that there are not ``fundamental flaws'' in the 
subject standards such that the standard cannot be met. Indeed, the 
data demonstrate that most sources do comply all or nearly all of the 
time.
    If the regulated community believes that a standard cannot be met 
across some meaningful range of normal operating conditions, or if 
specific exemptions beyond those currently provided are proper, we 
believe the appropriate action is for the affected industry to file a 
petition for amendment of the standard at issue or propose more 
specific permit conditions so that the matter can be fully assessed and 
addressed through the regulatory process. However, the information 
submitted by the commenters does not show that there currently exists a 
significant ``impossibility'' issue that is so widespread as to 
outweigh the benefits of the proposed rule.
4. Enforcement Using Continuous Monitoring Data Does Not Increase the 
Stringency of Applicable Requirements
    Industry commenters have argued that the stringency of emission 
standards will be increased if enforceable data is obtained more 
frequently than has been ordinarily obtained in the past through 
reference testing. Further, the commenters argue that direct 
enforceability of this data would contradict EPA's stated positions in 
adopting standards under the NSPS and NESHAP programs because EPA 
intended that continuous monitoring would only show compliance with 
good operation and maintenance procedures, i.e., general duty 
requirements, and would not be otherwise used in enforcement. (See, 
e.g., 38 FR 10820 (1973) (preamble to proposed startup, shutdown and 
malfunction regulation); 43 FR 7571 (1978) (preamble to final kraft 
pulp mill standards).
    Because the NSPS and NESHAP emission standards must be met 
continuously, consistent with any averaging times and except during 
periods where compliance is specifically excused, any more frequent or 
continuous monitoring of the standards and any enforcement based on 
violations uncovered thereby have no effect on the stringency of the 
standards. To take a simple analogy, allowing the use of radar guns or 
increasing the number of police checking for speeding may raise the 
chance that a speeder will be detected, but this does not alter the 
legal stringency of a posted speed limit.
    In some early NSPS, the agency required the installation of what 
were styled ``indicator monitors'' and provided policy guidance that 
such monitoring data would not be used as the sole basis of enforcement 
actions absent further rulemaking. 38 FR 10820. To the extent that the 
CAA Amendments of 1990 did not supersede this policy statement, today's 
action is that future rulemaking. These policy statements, like today's 
rulemaking, pertain only to the kinds of evidence EPA uses to prove 
violations. The policy change that was contemplated in our 1993 
proposal and 1996 memorandum are supported by technological advances in 
the accuracy and reliability of continuous emission monitors, 
deficiencies in EPA's previous practices identified by GAO and others, 
and the language and intent of the Act and the 1990 CAA Amendments.
    EPA's past statements regarding limitations on the use of data 
derived from continuous monitoring methods for purposes of enforcing 
standards were motivated in part by concerns over the cost and 
availability of such methods and their ability to accurately determine 
compliance. See, e.g., National Lime, 627 F.2d at 450 (responding to 
petitioners' argument that there was no adequately demonstrated 
technology for monitoring opacity, EPA stated that the continuous 
monitoring data would not be used to determine compliance with the 
opacity standard but ``to keep a check on the operation and maintenance 
of the control equipment,'' and that the monitors were reliable enough 
to perform this limited function). For example, in the 1973 startup, 
shutdown and malfunction regulation proposal, EPA noted that while 
continuous monitoring data would not, at that time, be used to 
determine compliance as a general matter, such data could be used if 
``approved as [an] equivalent or alternative method for performance 
testing.'' 38 FR 10820. Indeed, the NSPS general provisions have long 
provided that in lieu of performance tests using reference methods, a 
source could demonstrate compliance using an approved equivalent or 
alternative method, and that EPA can waive reference tests where the 
source has otherwise satisfactorily demonstrated compliance. See 40 CFR 
60.8(b).
    Since the 1970s, the availability, cost and accuracy of methods 
that enable determinations of compliance on a continuous basis has 
improved markedly. See, e.g., 1990 GAO report at 19, 22-23 (1986 and 
1988 EPA studies showed CEM data highly reliable); Continuous Emission 
Monitoring, 1993, Jahake, Thomas Publishing Co. For these reasons, EPA 
believes it is appropriate as a technical matter to allow information 
derived from these methods to be used in compliance certifications and 
enforcement actions. In fact, more recent national standards issued by 
EPA provide for determining and enforcing compliance directly by use of 
continuous monitoring data.
5. Sources Must Comply Both With Good Operation and Maintenance 
Requirements and With Emission Limits
    Industry commenters have claimed that as to the NSPS program, the 
only goal of the program was to insure that best demonstrated 
technology was employed, such that once an initial reference test 
demonstrated that compliance with the standards could be achieved, it 
need not be demonstrated thereafter, and that an affected source's only 
ongoing obligation was its ``general duty'' to employ good operation 
and maintenance practices to minimize emissions in accordance with 40 
CFR 60.11(d).
    EPA agrees that proper operation and maintenance of an emissions 
unit and any associated pollution controls in accordance with 40 CFR 
60.11(d) is vital to complying with emission standards. However, while 
it is true that sources have a continuing duty to employ good 
operations and maintenance practices, this duty does not substitute for 
the sources' obligation to comply with its emission limits. The two 
obligations, while related, are separate requirements in the NSPS 
regulations and in legal effect.
    EPA has made these points plain as far back as 1973 in the proposed 
NSPS startup, shutdown and malfunction rulemaking:

    It is anticipated that the initial performance test and 
subsequent performance tests will ensure that equipment is installed 
which will permit the standards to be attained and that such 
equipment is not allowed to deteriorate to the point where the 
standards are no longer maintained. In addition, the proposed 
regulation requires that the plant operator use maintenance and 
operating procedures designed to minimize emissions in excess of the 
standard.

38 FR 10820 (1973) (emphasis added). This preamble text clearly states 
both that proper equipment maintenance is vital to remaining within an 
emission

[[Page 8327]]

standard (otherwise equipment would deteriorate to the point where 
standards were not met) and that the general operation and maintenance 
obligation is a separate regulatory requirement. Additional discussion 
of the distinction between the emission limits and good operating 
practice requirements can be in the Detailed Response Document. These 
statements make it clear that good operating practices requirements are 
separate and distinct from the need to continuously comply with 
emissions limits.

E. SIP Call

    In the October, 1993, proposal, EPA announced that it planned to 
call for States to amend their applicable implementation plans to 
ensure that owners or operators may use enhanced monitoring (or other 
monitoring approved for the source pursuant to part 70) for compliance 
certification purposes, and that data from this monitoring, along with 
any other credible evidence, may be used as evidence of a violation of 
an applicable plan. 58 FR 54660. In December, 1993, and February, 1994, 
the Office of Air and Radiation's Stationary Source Compliance 
Division, the division then responsible for writing and implementing 
the enhanced monitoring rules, issued memoranda to EPA's Regional 
offices instructing them to conduct the SIP call. As of September, 
1996, fifteen states and local air pollution control districts, 
together with the Commonwealth of Puerto Rico, had responded to the 
call and submitted SIP amendments for EPA approval. Kansas, Iowa, 
Nebraska, North Dakota, Georgia and Puerto Rico had received approval; 
the other states and districts' revisions were pending.
    For substantially the same reasons that allow EPA to go forward 
with today's final rule, EPA has the authority to initiate and continue 
this SIP call. EPA's decision to forego the enhanced monitoring 
approach in favor of the CAM proposal has no effect on the basic goals 
of the SIP call, which are to clarify that non-reference test data can 
be used in enforcement actions, and to remove any potential ambiguity 
regarding this data's use for Title V compliance certifications.
    Today's action ensures that the evidentiary rules for CAA 
violations are consistent in all fifty states. EPA has surveyed those 
states that have responded to the SIP call and has determined that the 
credible evidence changes have not created the difficulties forecast by 
the commenters.

IV. Administrative Requirements

A. Docket

    Today's final rulemaking action is subject to Section 307(d) of the 
Act. Accordingly, EPA has established a docket (No. A-91-52), which 
consists of an organized and complete file of all information submitted 
to, or otherwise considered by, EPA in the development of today's 
action and the CAM approach. The docket includes all memoranda and 
studies cited by EPA in this preamble. The principal purposes of the 
docket are: (1) to allow interested parties a means to identify and 
locate documents so that they can effectively participate in the 
rulemaking process, and (2) to serve as the record in case of judicial 
review. The docket is available for public inspection at EPA's Air 
Docket, which is listed under the ADDRESSES section of this document.

B. Office of Management and Budget (OMB) Review

    Today's rulemaking is not a ``significant regulatory action'' 
because the revisions make only evidentiary changes and do not impose 
any additional implementation costs on regulated sources. Nevertheless, 
EPA submitted this final rule to OMB for review. Changes made in 
response to OMB suggestions and recommendations will be documented in 
the public record.

C. Unfunded Mandates Reform Act

    Under section 202 of the Unfunded Mandates Reform Act of 1995 
(UMRA), Public Law 104-4, EPA generally must prepare a budgetary impact 
statement to accompany any proposed or final rule that includes a 
Federal mandate that may result in expenditure by State, local, or 
tribal governments in the aggregate, or by the private sector, of $100 
million or more. Before promulgating a rule for which such a statement 
is needed, section 205 of the UMRA generally requires EPA to identify 
and consider a reasonable number of regulatory alternatives and adopt 
the least costly, most cost-effective or least burdensome alternative 
that achieves the objectives of the rule. The provisions of section 205 
do not apply when they are inconsistent with applicable law. Section 
203 requires the Agency to establish a plan for obtaining input from 
and informing, educating, and advising any small governments that may 
be significantly or uniquely affected by the rule.
    EPA has determined that this rule does not contain a Federal 
mandate that may result in expenditures of $100 million or more for 
State, local, and tribal governments, in the aggregate, or the private 
sector in any one year. Today's rulemaking makes only evidentiary 
changes and does not impose any additional costs on regulated sources 
or State, local, or tribal governments. For the same reason, these 
evidentiary changes will not significantly or uniquely affect small 
governments. Accordingly, this rulemaking is not subject to the 
requirements of sections 202, 203, and 205 of the UMRA.

D. Regulatory Flexibility Act

    EPA has determined that it is not necessary to prepare a regulatory 
flexibility analysis in connection with this final rule. EPA has also 
determined that this rule will not have a significant economic impact 
on a substantial number of small entities. As explained above, this 
rulemaking does not impose any additional implementation costs on small 
or large entities.

E. Paperwork Reduction Act

    The information collection requirements for the proposed enhanced 
monitoring rule were previously submitted for approval to OMB under the 
Paperwork Reduction Act (PRA), 44 U.S.C. 3501 et seq. In contrast, 
today's rule does not contain any information collection requirements 
subject to OMB review under the PRA.

F. Submission to Congress and the General Accounting Office

    Under 5 U.S.C. 801(a)(1)(A) of the Small Business Regulatory 
Enforcement Fairness Act of 1996 (Pub. L. 104-121, 110 Stat. 847), EPA 
submitted a report containing this rule and other required information 
to the U.S. Senate, the U.S. House of Representatives and the 
Controller General of the General Accounting Office prior to 
publication of this rule in today's Federal Register. For the same 
reasons that this rulemaking is not a ``significant regulatory action'' 
under Executive Order 12866, this rule is not a ``major rule'' as 
defined by 5 U.S.C. 804(2).

List of Subjects

40 CFR Part 51

    Environmental protection, Air pollution control.

40 CFR Part 52

    Air pollution control.

40 CFR Part 60

    Air pollution control.

40 CFR Part 61

    Air pollution control.


[[Page 8328]]


    Dated: February 13, 1997.
Carol M. Browner,
Administrator, U.S. Environmental Protection Agency.
    For the reasons set out in the preamble, 40 CFR Chapter I is 
amended as follows:

PART 51--REQUIREMENTS FOR PREPARATION, ADOPTION, AND SUBMITTAL OF 
IMPLEMENTATION PLANS

    1. The authority citation for part 51 is revised to read as 
follows:

    Authority: 42 U.S.C. 7401, 7411, 7412, 7413, 7414, 7470-7479, 
7501-7508, 7601, and 7602.

    2. Section 51.212 is amended by revising paragraph (c) to read as 
follows:


Sec. 51.212  Testing, inspection, enforcement, and complaints.

* * * * *
    (c) Enforceable test methods for each emission limit specified in 
the plan. For the purpose of submitting compliance certifications or 
establishing whether or not a person has violated or is in violation of 
any standard in this part, the plan must not preclude the use, 
including the exclusive use, of any credible evidence or information, 
relevant to whether a source would have been in compliance with 
applicable requirements if the appropriate performance or compliance 
test or procedure had been performed. As an enforceable method, States 
may use:
    (1) Any of the appropriate methods in appendix M to this part, 
Recommended Test Methods for State Implementation Plans; or
    (2) An alternative method following review and approval of that 
method by the Administrator; or
    (3) Any appropriate method in appendix A to 40 CFR part 60.

PART 52--APPROVAL AND PROMULGATION OF IMPLEMENTATION PLANS

    1. The authority citation for part 52 is revised to read as 
follows:

    Authority: 42 U.S.C. 7401 et seq.

    2. Section 52.12 is amended by revising paragraph (c) to read as 
follows:


Sec. 52.12  Source surveillance.

* * * * *
    (c) For purposes of Federal enforcement, the following test 
procedures and methods shall be used, provided that for the purpose of 
establishing whether or not a person has violated or is in violation of 
any provision of the plan, nothing in this part shall preclude the use, 
including the exclusive use, of any credible evidence or information, 
relevant to whether a source would have been in compliance with 
applicable requirements if the appropriate performance or compliance 
test procedures or methods had been performed:
    (1) Sources subject to plan provisions which do not specify a test 
procedure and sources subject to provisions promulgated by the 
Administrator will be tested by means of the appropriate procedures and 
methods prescribed in part 60 of this chapter unless otherwise 
specified in this part.
    (2) Sources subject to approved provisions of a plan wherein a test 
procedure is specified will be tested by the specified procedure.
    3. Subpart A is amended by adding a new Sec. 52.33 to read as 
follows:


Sec. 52.33  Compliance certifications.

    (a) For the purpose of submitting compliance certifications, 
nothing in this part or in a plan promulgated by the Administrator 
shall preclude the use, including the exclusive use, of any credible 
evidence or information, relevant to whether a source would have been 
in compliance with applicable requirements if the appropriate 
performance or compliance test had been performed.
    (b) For all federal implementation plans, paragraph (a) of this 
section is incorporated into the plan.

PART 60--STANDARDS OF PERFORMANCE FOR NEW STATIONARY SOURCES

    1. The authority citation for part 60 is revised to read as 
follows:

    Authority: 42 U.S.C. 7401, 7411, 7413, 7414, 7416, 7601 and 
7602.

    2. Section 60.11 is amended by revising paragraphs (a) and (f) and 
by adding paragraph (g) to read as follows:


Sec. 60.11  Compliance with standards and maintenance requirements.

    (a) Compliance with standards in this part, other than opacity 
standards, shall be determined in accordance with performance tests 
established by Sec. 60.8, unless otherwise specified in the applicable 
standard.
* * * * *
    (f) Special provisions set forth under an applicable subpart shall 
supersede any conflicting provisions in paragraphs (a) through (e) of 
this section.
    (g) For the purpose of submitting compliance certifications or 
establishing whether or not a person has violated or is in violation of 
any standard in this part, nothing in this part shall preclude the use, 
including the exclusive use, of any credible evidence or information, 
relevant to whether a source would have been in compliance with 
applicable requirements if the appropriate performance or compliance 
test or procedure had been performed.

PART 61--NATIONAL EMISSION STANDARDS FOR HAZARDOUS AIR POLLUTANTS

    1. The authority citation for part 61 is revised to read as 
follows:

    Authority: 42 U.S.C. 7401, 7412, 7413, 7414, 7416, 7601 and 
7602.

    2. Section 61.12 is amended by revising paragraph (a) and adding 
paragraph (e) to read as follows:


Sec. 61.12  Compliance with standards and maintenance requirements.

    (a) Compliance with numerical emission limits shall be determined 
in accordance with emission tests established in Sec. 61.13 or as 
otherwise specified in an individual subpart.
* * * * *
    (e) For the purpose of submitting compliance certifications or 
establishing whether or not a person has violated or is in violation of 
any standard in this part, nothing in this part shall preclude the use, 
including the exclusive use, of any credible evidence or information, 
relevant to whether a source would have been in compliance with 
applicable requirements if the appropriate performance or compliance 
test had been performed.

[FR Doc. 97-4196 Filed 2-21- 97; 8:45 am]
BILLING CODE 6560-50-P