[Federal Register: December 31, 2002 (Volume 67, Number 251)]
[Rules and Regulations]               
[Page 80185-80289]
From the Federal Register Online via GPO Access [wais.access.gpo.gov]
[DOCID:fr31de02-26]                         
 


[[Page 80185]]


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


Part III










Environmental Protection Agency










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






40 CFR Parts 51 and 52






Prevention of Significant Deterioration (PSD) and Nonattainment New 
Source Review (NSR); Final Rule and Proposed Rule




[[Page 80186]]




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


ENVIRONMENTAL PROTECTION AGENCY


40 CFR Parts 51 and 52


[AD-FRL-7414-5]
RIN 2060-AE11


 
Prevention of Significant Deterioration (PSD) and Nonattainment 
New Source Review (NSR): Baseline Emissions Determination, Actual-to-
Future-Actual Methodology, Plantwide Applicability Limitations, Clean 
Units, Pollution Control Projects


AGENCY: Environmental Protection Agency (EPA).


ACTION: Final rule.


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


SUMMARY: The EPA is revising regulations governing the New Source 
Review (NSR) programs mandated by parts C and D of title I of the Clean 
Air Act (CAA or Act). These revisions include changes in NSR 
applicability requirements for modifications to allow sources more 
flexibility to respond to rapidly changing markets and to plan for 
future investments in pollution control and prevention technologies. 
Today's changes reflect EPA's consideration of discussions and 
recommendations of the Clean Air Act Advisory Committee's (CAAAC) 
Subcommittee on NSR, Permits and Toxics, comments filed by the public, 
and meetings and discussions with interested stakeholders. The changes 
are intended to provide greater regulatory certainty, administrative 
flexibility, and permit streamlining, while ensuring the current level 
of environmental protection and benefit derived from the program and, 
in certain respects, resulting in greater environmental protection.


EFFECTIVE DATE: This final rule is effective on March 3, 2003.


ADDRESSES: Docket. Docket No. A-90-37, containing supporting 
information used to develop the proposed rule and the final rule, is 
available for public inspection and copying between 8 a.m. and 4:30 
p.m., Monday through Friday (except government holidays) at the Air and 
Radiation Docket and Information Center (6102T), Room B-108, EPA West 
Building, 1301 Constitution Avenue, NW., Washington, DC 20460; 
telephone (202) 566-1742, fax (202) 566-1741. A reasonable fee may be 
charged for copying docket materials. Worldwide Web (WWW). In addition 
to being available in the docket, an electronic copy of this final rule 
will also be available on the WWW through the Technology Transfer 
Network (TTN). Following signature, a copy of the rule will be posted 
on the TTN's policy and guidance page for newly proposed or promulgated 
rules: http://www.epa.gov/ttn/oarpg.


FOR FURTHER INFORMATION CONTACT: Ms. Lynn Hutchinson, Information 
Transfer and Program Integration Division (C339-03), U.S. EPA Office of 
Air Quality Planning and Standards, Research Triangle Park, North 
Carolina 27711, telephone 919-541-5795, or electronic mail at 
hutchinson.lynn@epa.gov, for general questions on this rule. For 
questions on baseline emissions determination or the actual-to-
projected-actual applicability test, contact Mr. Dan DeRoeck, at the 
same address, telephone 919-541-5593, or electronic mail at 
deroeck.dan@epa.gov. For questions on Plantwide Applicability 
Limitations (PALs), contact Mr. Raj Rao, at the same address, telephone 
919-541-5344, or electronic mail at rao.raj@epa.gov. For questions on 
Clean Units, contact Mr. Juan Santiago, at the same address, telephone 
919-541-1084, or electronic mail at santiago.juan@epa.gov. For 
questions on Pollution Control Projects (PCPs), contact Mr. Dave 
Svendsgaard, at the same address, telephone 919-541-2380, or electronic 
mail at svendsgaard.dave@epa.gov.


SUPPLEMENTARY INFORMATION:


Regulated Entities


    Entities potentially affected by this final action include sources 
in all industry groups. The majority of sources potentially affected 
are expected to be in the following groups.


------------------------------------------------------------------------
          Industry group              SIC a              NAICSb
------------------------------------------------------------------------
Electric Services................          491  221111, 221112, 221113,
                                                 221119, 221121, 221122
Petroleum Refining...............          291  32411
Chemical Processes...............          281  325181, 32512, 325131,
                                                 325182, 211112, 325998,
                                                 331311, 325188
Natural Gas Transport............          492  48621, 22121
Pulp and Paper Mills.............          261  32211, 322121, 322122,
                                                 32213
Paper Mills......................          262  322121, 322122
Automobile Manufacturing.........          371  336111, 336112, 336712,
                                                 336211, 336992, 336322,
                                                 336312, 33633, 33634,
                                                 33635, 336399, 336212,
                                                 336213
Pharmaceuticals..................          283  325411, 325412, 325413,
                                                 325414
------------------------------------------------------------------------
a Standard Industrial Classification
b North American Industry Classification System.


    Entities potentially affected by this final action also include 
State, local, and tribal governments that are delegated authority to 
implement these regulations.
    Outline. The information presented in this preamble is organized as 
follows:


I. Overview of Today's Final Action
    A. Background
    B. Introduction
    C. Overview of Final Actions
    1. Determining Whether a Proposed Modification Results in a 
Significant Emissions Increase
    2. CMA Exhibit B
    3. Plantwide Applicability Limitations (PALs)
    4. Clean Units
    5. Pollution Control Projects (PCPs)
    6. Major NSR Applicability
    7. Enforcement
    8. Enforceability
II. Revisions to the Method for Determining Whether a Proposed 
Modification Results in a Significant Emissions Increase
    A. Introduction
    B. What We Proposed and How Today's Action Compares
    C. Baseline Actual Emissions For Existing Emissions Units Other 
than EUSGUs
    D. The Actual-to-projected-actual Applicability Test
    E. Clarifying Changes to WEPCO Provisions for EUSGUs
    F. The ``Hybrid'' Applicability Test
    G. Legal Basis for Today's Action
    H. Response to Comments and Rationale for Today's Actions
III. CMA Exhibit B
IV. Plantwide Applicability Limitations (PALs)
    A. Introduction
    B. Relevant Background
    C. Final Regulations for Actuals PALs
    D. Rationale for Today's Final Action on Actuals PALs
V. Clean Units


[[Page 80187]]


    A. Introduction
    B. Summary of 1996 Clean Unit Proposal
    C. Final Regulations for Clean Units
    D. Legal Basis for the Clean Unit Test
    E. Summary of Major Comments and Responses
VI. Pollution Control Projects (PCPs)
    A. Description and Purpose of This Action
    B. What We Proposed and How Today's Action Compares To It
    C. Legal Basis for PCP
    D. Implementation
VII. Listed Hazardous Air Pollutants
VIII. Effective Date for Today's Requirements
IX. Administrative Requirements
    A. Executive Order 12866--Regulatory Planning and Review
    B. Executive Order 13132--Federalism
    C. Executive Order 13175--Consultation and Coordination with 
Indian Tribal Governments
    D. Executive Order 13045--Protection of Children from 
Environmental Health Risks and Safety Risks
    E. Unfunded Mandates Reform Act of 1995
    F. Regulatory Flexibility Act (RFA), as Amended by the Small 
Business Regulatory Enforcement Fairness Act of 1996 (SBREFA), 5 
U.S.C. 601 et seq.
    G. Paperwork Reduction Act
    H. National Technology Transfer and Advancement Act of 1995
    I. Congressional Review Act
    J. Executive Order 13211--Actions Concerning Regulations That 
Significantly Affect Energy Supply, Distribution, or Use
X. Statutory Authority
XI. Judicial Review


I. Overview of Today's Final Action


A. Background


    We \1\ proposed revisions to the NSR rules in a notice published in 
the Federal Register on July 23, 1996 (61 FR 38250). On July 24, 1998, 
we published a notice (63 FR 39857) to solicit further comment on two 
specific aspects of the proposed revisions. Today's Federal Register 
action announces EPA's final action on the proposed revisions for 
baseline emissions determinations, the actual-to-future-actual 
methodology, actuals PALs, Clean Units, and PCPs. We have not made 
final determinations on any other proposed changes to the regulations.
---------------------------------------------------------------------------


    \1\ In this preamble the term ``we'' refers to EPA and the term 
``you'' refers to major stationary sources of air pollution and 
their owners and operators. All other entities are referred to by 
their respective names (for example, reviewing authorities.)
---------------------------------------------------------------------------


    Today's actions finalize these changes to the regulations for both 
the approval and promulgation of implementation plans and requirements 
for preparation, adoption, and submittal of implementation plans 
governing the NSR programs mandated by parts C and D of title I of the 
Act. We also proposed conforming changes to 40 CFR (Code of Federal 
Regulations) part 51, appendix S, and part 52.24. Today we have not 
included the final regulatory language for these regulations. It is our 
intention to include regulatory changes that conform appendix S and 40 
CFR 52.24 to today's final rules in any final regulations that set 
forth an interim implementation strategy for the 8-hour ozone standard. 
We intend to finalize changes to these sections precisely as we have 
finalized requirements for other parts of the program. Because these 
are conforming changes and the public has had an opportunity for review 
and comment, we will not be soliciting additional comments before we 
finalize them.
    The major NSR program contained in parts C and D of title I of the 
Act is a preconstruction review and permitting program applicable to 
new or modified major stationary sources of air pollutants regulated 
under the Act. In areas not meeting health-based National Ambient Air 
Quality Standards (NAAQS) and in ozone transport regions (OTR), the 
program is implemented under the requirements of part D of title I of 
the Act. We call this program the ``nonattainment'' NSR program. In 
areas meeting NAAQS (``attainment'' areas) or for which there is 
insufficient information to determine whether they meet the NAAQS 
(``unclassifiable'' areas), the NSR requirements under part C of title 
I of the Act apply. We call this program the Prevention of Significant 
Deterioration (PSD) program. Collectively, we also commonly refer to 
these programs as the major NSR program. These regulations are 
contained in 40 CFR 51.165, 51.166, 52.21, 52.24, and part 51, appendix 
S.
    The NSR provisions of the Act are a combination of air quality 
planning and air pollution control technology program requirements for 
new and modified stationary sources of air pollution. In brief, section 
109 of the Act requires us to promulgate primary NAAQS to protect 
public health and secondary NAAQS to protect public welfare. Once we 
have set these standards, States must develop, adopt, and submit to us 
for approval a State Implementation Plan (SIP) that contains emission 
limitations and other control measures to attain and maintain the NAAQS 
and to meet the other requirements of section 110(a) of the Act.
    Each SIP is required to contain a preconstruction review program 
for the construction and modification of any stationary source of air 
pollution to assure that the NAAQS are achieved and maintained; to 
protect areas of clean air; to protect Air Quality Related Values 
(AQRVs) (including visibility) in national parks and other natural 
areas of special concern; to assure that appropriate emissions controls 
are applied; to maximize opportunities for economic development 
consistent with the preservation of clean air resources; and to ensure 
that any decision to increase air pollution is made only after full 
public consideration of all the consequences of such a decision.
    For newly constructed, ``greenfield'' sources, the determination of 
whether an activity is subject to the major NSR program is fairly 
straightforward. The Act, as implemented by our regulations, sets 
applicability thresholds for major sources in nonattainment areas 
[potential to emit (PTE) above 100 tons per year (tpy) of any pollutant 
subject to regulation under the Act, or smaller amounts, depending on 
the nonattainment classification] and attainment areas (100 or 250 tpy, 
depending on the source type). A new source with a PTE at or above the 
applicable threshold amount ``triggers,'' or is subject to, major NSR.
    The determination of what should be classified as a modification 
subject to major NSR presents more difficult issues. The modification 
provisions of the NSR program in parts C and D are based on the 
definition of modification in section 111(a)(4) of the Act: the term 
``modification'' means ``any physical change in, or change in the 
method of operation of, a stationary source which increases the amount 
of any air pollutant emitted by such source or which results in the 
emission of any air pollutant not previously emitted.'' That definition 
contemplates that, first, you will determine whether a physical or 
operational change will occur. If so, then you will proceed to 
determine whether the physical or operational change will result in an 
emissions increase over baseline levels.
    The expression ``any physical change * * * or change in the method 
of operation'' in section 111(a)(4) of the Act is not defined. We have 
recognized that Congress did not intend to make every activity at a 
source subject to the major NSR program. As a result, we have 
previously adopted several exclusions from what may constitute a 
``physical or operational change.'' For instance, we have specifically 
recognized that routine maintenance, repair and replacement, and 
changes in hours of operation or in the production rate are not 
considered a physical change or change in the method of


[[Page 80188]]


operation within the definition of major modification.\2\
---------------------------------------------------------------------------


    \2\ See 40 CFR 52.21(b)(2).
---------------------------------------------------------------------------


    We have likewise addressed the scope of the statutory definition of 
modification by excluding all changes that do not result in a 
``significant'' emissions increase from a major source.\3\ This 
regulatory framework applies the major NSR program at existing sources 
to only ``major modifications'' at major stationary sources.
---------------------------------------------------------------------------


    \3\ See 40 CFR 52.21(b)(23).
---------------------------------------------------------------------------


    One key attribute of the major NSR program in general is that you 
may ``net'' modifications out of review by coupling proposed emissions 
increases at your source with contemporaneous emissions reductions. 
Thus, under regulations we promulgated in 1980, you may modify, or even 
completely replace, or add, emissions units without obtaining a major 
NSR permit, so long as ``actual emissions'' do not increase by a 
significant amount over baseline levels at the plant as a whole.
    Applicability of the major NSR program must be determined in 
advance of construction and is pollutant-specific. In cases involving 
existing sources, this requires a pollutant-by-pollutant determination 
of the emissions change, if any, that will result from the physical or 
operational change. Our 1980 regulations implementing the PSD and 
nonattainment major NSR programs thus inquire whether the proposed 
change constitutes a ``major modification,'' that is, a physical change 
or change in the method of operation ``that would result in a 
significant net emissions increase of any pollutant subject to 
regulation under the Act.'' A ``net emissions increase'' is defined as 
the increase in ``actual emissions'' from the particular physical or 
operational change (taking into account the use of emissions control 
technology and restrictions on hours of operation or rates of 
production where such controls and restrictions are enforceable), 
together with your other contemporaneous increases or decreases in 
actual emissions.\4\ In order to trigger applicability of the major NSR 
program, the net emissions increase must be ``significant.'' \5\
---------------------------------------------------------------------------


    \4\ In approximate terms, ``contemporaneous'' emissions 
increases or decreases are those that have occurred between the date 
5 years immediately preceding the proposed physical or operational 
change and the date that the increase from the change occurs. See, 
for example, Sec.  52.21(b)(3)(ii).
    \5\ Once a modification is determined to be major, the PSD 
requirements apply only to those specific pollutants for which there 
would be a significant net emissions increase. See, for example, 
Sec.  52.21(j)(3) (BACT) and Sec.  52.21(m)(1)(b) (air quality 
analysis).
---------------------------------------------------------------------------


    Before today's changes, our regulations generally defined actual 
emissions as ``the average rate, in tpy, at which the unit actually 
emitted the pollutant during a 2-year period which precedes the 
particular date and which is representative of normal source 
operation.'' The reviewing authorities will allow use of a different 
time period ``upon a determination that it is more representative of 
normal source operation.'' We have historically used the 2 years 
immediately preceding the proposed change to establish a source's 
actual emissions. However, in some cases we have allowed use of an 
earlier period.
    With respect to changes at existing sources, a prediction of 
whether the physical or operational change would result in a 
significant net increase in your actual emissions following the change 
was thus necessary. In part, this involved a straightforward and 
readily predictable engineering judgment--how would the change affect 
the emission factor or emissions rate of the emissions units that are 
to be changed.
    Before today's changes, the regulations provided that when your 
emissions unit, other than an electric utility steam generating unit 
(EUSGU), ``has not begun normal operations,'' actual emissions equal 
the PTE of the unit. When you have not begun normal operations 
following a change, you must assume that your source will operate at 
its full capacity year round, that is, at its full emissions potential. 
This is referred to as the actual-to-potential test. You may avoid the 
need for an NSR permit by reducing your source's potential emissions 
through the use of enforceable restrictions to pre-modification actual 
emissions levels plus an amount that is less than ``significant''.
    In 1992, we promulgated revisions to our applicability regulations 
creating special rules for physical and operational changes at EUSGUs. 
See 57 FR 32314 (July 21, 1992).\6\ In this rule, prompted by 
litigation involving the Wisconsin Electric Power Company (WEPCO) and 
commonly referred to as the ``WEPCO rule,'' we adopted an actual-to-
future-actual methodology for all changes at EUSGUs except the 
construction of a new electric generating unit or the replacement of an 
existing emissions unit. Under this methodology, the actual annual 
emissions before the change are compared with the projected actual 
emissions after the change to determine if a physical or operational 
change would result in a significant increase in emissions. To ensure 
that the projection is valid, the rule requires the utility to track 
its emissions for the next 5 years and provide to the reviewing 
authority information demonstrating that the physical or operational 
change did not result in an emissions increase.
---------------------------------------------------------------------------


    \6\ The regulations define ``electric utility steam generating 
units'' as any steam electric generating unit that is constructed 
for the purpose of supplying more than one-third of its potential 
electric output capacity and more than 25 megawatts (MW) of 
electrical output to any utility power distribution system for sale. 
See, for example, Sec.  51.166(b)(30).
---------------------------------------------------------------------------


    In promulgating the WEPCO rule, we also adopted a presumption that 
utilities may use as baseline emissions the actual annual emissions 
from any 2 consecutive years within the 5 years immediately preceding 
the change.
    In attainment areas, once major NSR is triggered, you must, among 
other things, install best available control technology (BACT) and 
conduct modeling and monitoring as necessary. If your source is located 
in a nonattainment area, you must install technology that meets the 
lowest achievable emissions rate (LAER), secure emissions reductions to 
offset any increases above baseline emission levels, and perform other 
analyses.


B. Introduction


    Today's final regulations were proposed as part of a larger 
regulatory package on July 23, 1996 (61 FR 38250). That package 
proposed a number of changes to our existing major NSR requirements. 
(Please refer to the outline of that proposed rulemaking for a complete 
list of changes that were proposed to our existing regulations.) On 
July 24, 1998, we published a Federal Register Notice of Availability 
(NOA) that requested additional comment on three of the proposed 
changes: determining baseline emissions, actual-to-future-actual 
methodology, and PALs. Following the 1996 proposals, we held two public 
hearings and more than 50 stakeholder meetings. Environmental groups, 
industry, and State, local, and Federal agency representatives 
participated in these many discussions.
    In May 2001, President Bush's National Energy Policy Development 
Group issued findings and key recommendations for a National Energy 
Policy. This document included numerous recommendations for action, 
including a recommendation that the EPA Administrator, in consultation 
with the Secretary of Energy and other relevant agencies, review NSR 
regulations, including administrative interpretation and 
implementation. The recommendation requested that we issue a report to 
the President on the impact of the regulations on investment


[[Page 80189]]


in new utility and refinery generation capacity, energy efficiency, and 
environmental protection.
    In response, in June 2001, we issued a background paper giving an 
overview of the NSR program. This paper is available on the Internet at 
http://www.epa.gov/air/nsr-review/background.html. We solicited public 
comments on the background paper and other information relevant to the 
New Source Review 90-day Review and Report to the President. During our 
review of the NSR program, we met with more than 100 groups, held four 
public meetings around the country, and received more than 130,000 
written comments. Our report to the President and our recommendations 
in response to the energy policy were issued on June 13, 2002. A copy 
of this information is available at http://www.epa.gov/air/nsr-review/. 
We expect that our recommendations in response to the energy policy 
will be reflected in the future in various programs and regulatory 
actions. Today's actions implement several of those recommendations.
    Today, we are finalizing five actions that we previously proposed 
in 1996 (three of which were re-noticed in the 1998 NOA). We are not 
taking final action on any of the remaining issues in the 1996 proposal 
at this time. We have not decided what final action we will take on 
those issues.


C. Overview of Final Actions


    Today we are taking final action on five changes to the NSR program 
that will reduce burden, maximize operating flexibility, improve 
environmental quality, provide additional certainty, and promote 
administrative efficiency. These elements include baseline actual 
emissions, actual-to-projected-actual emissions methodology, PALs, 
Clean Units, and PCPs. We are also codifying our longstanding policy 
regarding the calculation of baseline emissions for EUSGUs. In 
addition, we are responding to comments we received on a proposal to 
adopt a methodology, developed by the American Chemistry Council 
(formerly known as the Chemical Manufacturers Association (CMA)) and 
other industry petitioners, to determine whether a source has 
undertaken a modification based on its potential emissions. We are 
including a new section in today's final rules that outlines how a 
major modification is determined under the various major NSR 
applicability options and clarifies where you will find the provisions 
in our revised rules. Finally, we have codified a new definition of 
``regulated NSR pollutant'' that clarifies which pollutants are 
regulated under the Act for purposes of major NSR.
    This section briefly introduces each improvement. Detailed 
discussions of the improvements are found in sections II through VII of 
this preamble.
1. Determining Whether a Proposed Modification Results in a Significant 
Emissions Increase
    Today we are finalizing two changes to our existing major NSR 
regulations that will affect how you calculate emissions increases to 
determine whether physical changes or changes in the method of 
operation trigger the major NSR requirements. First, we have a new 
procedure for determining ``baseline actual emissions.'' That is, the 
relevant terminology for calculating pre-change emissions for most 
applications is now ``baseline actual emissions'' rather than ``actual 
emissions.'' You may use any consecutive 24-month period in the past 10 
years to determine your baseline actual emissions. Second, we are 
supplementing the existing actual-to-potential applicability test with 
an actual-to-projected-actual applicability test for determining if a 
physical or operational change at an existing emissions unit will 
result in an emissions increase. Notwithstanding the new test, you will 
still have the ability to conduct an actual-to-potential type test 
within the new actual-to-projected-actual applicability test. In this 
case, you will not be subject to recordkeeping requirements that are 
being established and would otherwise apply as part of the new actual-
to-projected actual applicability test.
    For EUSGUs, we are making several changes to the existing 
procedures and are codifying our current policy for calculating the 
baseline actual emissions. That is, the baseline actual emissions for 
EUSGUs is the average rate, in tpy, at which that unit actually emitted 
the pollutant during a 2-year (consecutive 24-month) period within the 
5-year period immediately preceding when the owner or operator begins 
actual construction. We are also retaining the option that allows the 
use of a different time period if the reviewing authority determines it 
is more representative of normal source operation.
2. CMA Exhibit B
    As described in section I.C.1 above, we have decided to adopt an 
actual-to-projected-actual methodology, combined with a revised process 
to determine baseline emissions, to use in determining when sources are 
considered to have made a modification and are thereby subject to NSR. 
We are not adopting the methodology based on potential emissions as 
discussed in the CMA Exhibit B proposal. See section III of this 
preamble for a discussion of the comments we received on this proposal 
and our responses.
3. Plantwide Applicability Limitations
    A PAL is a voluntary option that will provide you with the ability 
to manage facility-wide emissions without triggering major NSR review. 
We believe that the added flexibility provided under a PAL will 
facilitate your ability to respond rapidly to changing market 
conditions while enhancing the environmental protection afforded under 
the program.
    Today we are promulgating a PAL based on plantwide actual 
emissions. If you keep the emissions from your facility below a 
plantwide actual emissions cap (that is, an actuals PAL), then these 
regulations will allow you to avoid the major NSR permitting process 
when you make alterations to the facility or individual emissions 
units. In return for this flexibility, you must monitor emissions from 
all of your emissions units under the PAL. The benefit to you is that 
you can alter your facility without first obtaining a Federal NSR 
permit or going through a netting review. A PAL will allow you to make 
changes quickly at your facility. If you are willing to undertake the 
necessary recordkeeping, monitoring, and reporting, a PAL offers you 
flexibility and regulatory certainty.
4. Clean Units
    We are promulgating a new type of applicability test for emissions 
units that are designated as Clean Units. The new applicability test 
recognizes that when you go through major NSR review and install BACT 
or LAER, you may make any changes to the Clean Unit without triggering 
an additional major NSR review, if the project at a Clean Unit does not 
cause the need for a change in the emission limitations or work 
practice requirements in the permit for the unit that were adopted in 
conjunction with BACT or LAER and the project would not alter any 
physical or operational characteristics that formed the basis for the 
BACT or LAER determination. If the project causes the need for a change 
in the emission limitations or work practice requirements in the permit 
for the unit adopted in conjunction with BACT or LAER or would alter 
any physical or operational characteristics that formed the basis for 
the BACT or LAER determination, you lose Clean Unit status. You may 
still proceed with the project without triggering major NSR


[[Page 80190]]


review, if the increase is not a significant net emissions increase. 
Emissions units that have not been through major NSR may still qualify 
for Clean Unit status if they demonstrate that the emissions control 
level is comparable to BACT or LAER. Clean Unit status will be valid 
for up to a 10-year period. The new applicability test does not exclude 
consideration of physical changes or changes in the method of operation 
of Clean Units from major NSR, but rather changes the way emissions 
increases are calculated for these changes. This new applicability test 
therefore protects air quality, creates incentives for sources to 
install state-of-the-art controls, provides flexibility for sources, 
and promotes administrative efficiency.
5. Pollution Control Projects
    Today's rule contains a new list of environmentally beneficial 
technologies that qualify as PCPs for all types of sources. 
Installation of a PCP is not subject to the major modification 
provisions. An owner or operator installing a listed PCP automatically 
qualifies for the exclusion if there is no adverse air quality impact--
that is, if it will not cause or contribute to a violation of NAAQS or 
PSD increment, or adversely impact an AQRV (such as visibility) that 
has been identified for a Federal Class I area by a Federal Land 
Manager (FLM) and for which information is available to the general 
public. The PCPs that are not listed in today's rules may also qualify 
for the PCP Exclusion if the reviewing authority determines on a case-
specific basis that a non-listed PCP is environmentally beneficial when 
used for a particular application. Also, in the future, we may add to 
the listed PCPs through a rulemaking that provides for public notice 
and opportunity for comment. The PCP Exclusion allows sources to 
install emissions controls that are known to be environmentally 
beneficial. These provisions thus offer flexibility while improving air 
quality.
6. Major NSR Applicability
    We have briefly described the new provisions for baseline actual 
emissions, actual-to-projected-actual methodology, PALs, and Clean 
Units. Sections II, IV, and V describe the new provisions in detail. 
These provisions offer major new changes to NSR applicability, 
especially regarding how a major modification is determined. The major 
NSR applicability provisions have developed over time and therefore 
have been added to the NSR rules in a piecemeal fashion. In today's 
final rules we are including a new section that outlines how a major 
modification is determined under the various major NSR applicability 
options and clarifies where you will find the provisions in our revised 
rules. For each applicability option, we describe how a major 
modification is determined in detail. You'll find this new 
applicability ``roadmap'' in Sec. Sec.  51.165(a)(2), 51.166(a)(7), and 
52.21(a)(2). To summarize, the various provisions for major 
modifications are now as follows.
    [sbull] Actual-to-projected-actual applicability test for all 
existing emissions units. (Including an actual-to-potential option)
    [sbull] Actual-to-potential test for any new unit, including 
EUSGUs.
    [sbull] The Clean Unit Test for existing emissions units with Clean 
Unit status.
    [sbull] The hybrid test for modifications with multiple types of 
emissions units. (Used when a physical or operational change affects a 
combination of more than one type of unit.)
    We describe actuals PALs, which are an alternative way of complying 
with major NSR, in section IV of this preamble. If you have a PAL, as 
long as you are complying with the PAL requirements, any physical or 
operational changes are not major modifications.
    We have revised the definition of major modification to clarify 
what has always been our policy--that determining whether a major 
modification has occurred is a two-step process. The new definition of 
major modification is ``any physical change in or change in the method 
of operation of a major stationary source that would result in: (1) A 
significant emissions increase of a regulated NSR pollutant; and (2) a 
significant net emissions increase of that pollutant from the major 
stationary source.'' We have also revised the definitions of actual 
emissions, emissions unit, net emissions increase, and construction. We 
have deleted the word ``actual'' as related to emissions from the 
definition of ``construction.'' This change was necessary because of 
how the definition of ``actual emissions'' is used in the final rule, 
but the deletion is not intended to change any meaning in the term 
``construction.'' We have added new definitions for baseline actual 
emissions, projected actual emissions, project, and significant 
emissions increase. These revisions and additions implement our new 
provisions for major modifications under the actual-to-projected-actual 
applicability test, actual-to-potential test, Clean Unit Test, and 
hybrid test. You will find a complete discussion of the Clean Unit 
Test, including how modifications to Clean Units are treated, in 
section V of this preamble. The other tests are discussed in section 
II.
    ``Actual emissions,'' as the term has been historically applied, 
will still be used to determine air quality impacts (for example, 
compliance with NAAQS, PSD increments, and AQRVs) and to compute the 
required amount of emissions offsets.
    To further clarify major NSR applicability in one location, we have 
moved Sec.  51.166(i)(1) through (3) and Sec.  52.21(i)(1) through (3) 
into the new applicability sections at Sec.  51.166(a)(7) and Sec.  
52.21(a)(2). These provisions clarify that you must obtain a permit 
before you begin construction (including for major modifications), that 
the provisions apply for each regulated NSR pollutant that your source 
emits, and that the provisions apply to any source located in the area 
designated as attainment or unclassifiable (for Sec. Sec.  51.166 and 
52.21).
    We have also added a new definition for reviewing authority that 
clarifies who has authority to implement major NSR programs. Reviewing 
authority means the State air pollution control agency, local agency, 
other State agency, Indian tribe, or other agency authorized by the 
Administrator to carry out a permit program under Sec. Sec.  51.165 and 
51.166, or the Administrator in the case of EPA-implemented permit 
programs under Sec.  52.21.
7. Enforcement
    As noted above, today we are taking final action on five changes to 
the NSR program that create alternative means of determining NSR 
applicability for projects that begin actual construction after these 
provisions become effective in your jurisdiction. If you are 
subsequently determined not to have met any of the obligations of these 
new alternatives (for example, failure to meet emissions or 
applicability limits, properly project emissions, and/or properly 
implement the PCP Exclusion or Clean Unit Test), you will be subject to 
any applicable enforcement provisions (including the possibility of 
citizens' suits) under the applicable sections of the Act. Sanctions 
for violations of these provisions may include monetary penalties of up 
to $27,500 per day of violation, as well as the possibility of 
injunctive relief, which may include the requirement to install air 
pollution controls.
8. Enforceability
    This rule uses several terms related to enforceability of 
particular provisions. A requirement is ``legally enforceable'' if some 
authority has the right to enforce the restriction. Practical 
enforceability for a source-specific permit will be


[[Page 80191]]


achieved if the permit's provisions specify: (1) A technically-accurate 
limitation and the portions of the source subject to the limitation; 
(2) the time period for the limitation (hourly, daily, monthly, and 
annual limits such as rolling annual limits); and (3) the method to 
determine compliance, including appropriate monitoring, recordkeeping, 
and reporting. For rules and general permits that apply to categories 
of sources, practicable enforceability additionally requires that the 
provisions: (1) Identify the types or categories of sources that are 
covered by the rule; (2) where coverage is optional, provide for notice 
to the permitting authority of the source's election to be covered by 
the rule; and (3) specify the enforcement consequences relevant to the 
rule.\7,\ \8\ ``Enforceable as a practical matter'' will be achieved if 
a requirement is both legally and practically enforceable.
---------------------------------------------------------------------------


    \7\ See memorandum, ``Release of Interim Policy on Federal 
Enforceability of Limitations on Potential to Emit,'' signed by John 
Seitz and Robert Van Heuvelen, Jan. 22, 1996 at 5-6 and Attachment 
4, available on the Web as http://www.epa.gov/rgytgrnj/programs/artd/air/title5/t5memos/pottoemi.pdf.
 More detailed guidance on 
practical enforceability is contained in the memorandum.
    \8\ The Agency has frequently used the term ``practicably 
enforceable'' and ``practical enforceability,'' interchangeably. 
There is no difference in the meaning of these terms.
---------------------------------------------------------------------------


    Note that we continue to require offsets to be federally 
enforceable. ``Federal enforceability'' means that not only is a 
requirement practically enforceable, as described above, but in 
addition, ``EPA must have a direct right to enforce restrictions and 
limitations imposed on a source to limit its exposure to Act 
programs.'' \9\ Also note that, for computing baseline actual emissions 
for use in determining major NSR applicability or for establishing a 
PAL, you must consider ``legally enforceable'' requirements. A 
requirement will be legally enforceable if the Administrator, State, 
local or tribal air pollution control agency has the authority to 
enforce the requirement irrespective of its practical enforceability.
---------------------------------------------------------------------------


    \9\ See generally memorandum, ``Options for Limiting the 
Potential to Emit (PTE) of a Stationary Source Under Section 112 and 
Title V of the Clean Air Act,'' signed by John Seitz and Robert Van 
Heuvelen, Jan. 25, 1995, at 2-3.
---------------------------------------------------------------------------


    In our existing regulations that are unamended by today's action, 
the term ``federally enforceability'' still appears. In 1995, the court 
in Chemical Manufacturers Ass'n v. EPA remanded the definition of PTE 
in the major NSR program to EPA. No. 89-1514 (D.C. Cir. Sept. 150 
1995). Because the court vacated the requirements in the nationwide 
rules, the term federal enforceability as it relates to PTE is not in 
effect (pending final rule making by the Agency) in the Federal rules. 
The decision, however, did not address the term ``federally 
enforceable'' as used in SIPs, because that issue was not before the 
court.


II. Revisions to the Method for Determining Whether a Proposed 
Modification Results in a Significant Emissions Increase


A. Introduction


    Today we are finalizing two sets of amendments to our existing 
major NSR regulations that provide another way in which you may 
calculate emissions increases to determine whether certain types of 
physical changes or changes in the method of operation (physical or 
operational changes) of an existing emissions unit trigger the major 
NSR requirements.\10\ The first set of amendments relates to the way in 
which you will determine your baseline actual emissions for such 
emissions units in accordance with a new definition of ``baseline 
actual emissions.'' See, for example, new Sec.  52.21(b)(48). We will 
be allowing you to use any consecutive 24-month period during the 10-
year period prior to the change to determine your baseline actual 
emissions for existing emissions units (other than EUSGUs). The second 
set of amendments replaces the existing actual-to-potential and actual-
to-representative-actual-annual emissions applicability tests for 
existing emissions units (including EUSGUs) with an actual-to-
projected-actual applicability test for determining if a physical or 
operational change will result in an emissions increase at such units. 
(Notwithstanding this new test, the actual-to-potential methodology is 
still available at your option under the new applicability tests.) The 
new procedure for determining your pre-change baseline actual emissions 
will not apply to EUSGUs.\11\ Instead, for EUSGUs we are retaining the 
existing procedures for determining the baseline actual emissions.\12\ 
See, for example, existing Sec.  52.21(b)(33). We are also affirming 
our current method used for calculating the baseline actual emissions 
for EUSGUs (allowing any consecutive 2 years in the past 5 years, or 
another more representative period) by codifying it in the NSR 
regulations. See, for example, new Sec.  52.21(b)(48).
---------------------------------------------------------------------------


    \10\ By definition, the modification of an existing source is 
potentially subject to major NSR only if that existing source is 
``major.'' In addition, when an existing ``minor'' source makes a 
physical or operational change that by itself is major, that change 
constitutes a major stationary source that is subject to major NSR. 
See, for example, Sec.  52.21(b)(1)(c).
    \11\ For NSR purposes, the definition of ``electric utility 
steam generating unit'' means any steam electric generating unit 
that is constructed for the purpose of supplying more than one-third 
of its potential electric output capacity and more than 25 MW 
electrical output to any utility power distribution system for sale. 
Any steam supplied to a steam distribution system for the purpose of 
providing steam to a steam electric generator that would produce 
electrical energy for sale is also considered in determining the 
electrical energy output capacity of the affected facility. See, for 
example, Sec.  52.21(b)(31). Reference in this notice to utility 
units is meant to include all emissions units covered by this 
definition.
    \12\ We promulgated special applicability rules for physical and 
operational changes at EUSGUs in 1992. See 57 FR 32314 (July 21, 
1992).
---------------------------------------------------------------------------


    For existing emissions units other than EUSGUs, the changes we are 
making to the method for calculating a unit's baseline actual emissions 
will apply only for the following three purposes.
    [sbull] For modifications, to determine a modified unit's pre-
change baseline actual emissions as part of the new actual-to-
projected-actual applicability test.
    [sbull] For netting, to determine the pre-change baseline actual 
emissions of an emissions unit that underwent a physical or operational 
change within the contemporaneous period.
    [sbull] For PALs, to establish the PAL emissions cap.
    Today's new procedures for calculating baseline actual emissions 
and for the actual-to-projected-actual applicability test should not be 
used when determining a source's actual emissions on a particular date 
as may be used for other NSR-related requirements. Such requirements 
include, but are not limited to, air quality impacts analyses (for 
example, compliance with NAAQS, PSD increments, and AQRVs) and 
computing the required amount of emissions offsets. For each of these 
requirements, the existing definition of ``actual emissions'' continues 
to apply. This is discussed in greater detail in section II.D.9.
    We believe that these changes will greatly improve the major NSR 
program by responding to industry concerns with our existing 
methodology without compromising air quality. One common complaint 
about the current emissions baseline process is that you have a limited 
ability to consider the operational fluctuations associated with normal 
business cycles when establishing baseline actual emissions unless your 
reviewing authority agrees that another period is ``more representative 
of normal source


[[Page 80192]]


operation.'' \13\ By extending the time period from which you may 
establish your baseline actual emissions, the new procedures should 
reflect the emissions levels that occur during a normal business cycle, 
without requiring you to demonstrate to your reviewing authority that 
another period is ``more representative of normal source operations.''
---------------------------------------------------------------------------


    \13\ The definition of ``actual emissions'' requires that a 
unit's actual emissions be based on a consecutive 24-month period 
immediately preceding the particular change. Also, however, it 
directs the reviewing authority to allow the use of another time 
period upon a determination that it is more representative. This 
procedure continues to be appropriate under the pre-existing 
regulation and for other NSR purposes, such as determining a 
source's ambient impact against the PSD increments, and we continue 
to require its use for such purposes.
---------------------------------------------------------------------------


    Commenters also believe that the current methodology requires many 
changes made to existing equipment to go through major NSR, without 
taking into account operating history, even when such changes will not 
result in increased pollution to the environment. Our new applicability 
requirements address these commenters' concerns and will focus limited 
resources more effectively.
    We are also modifying the way you may determine whether emissions 
at existing units (including EUSGUs) will increase, by allowing you to 
use projected actual emissions for purposes of this determination. 
Under this approach, in circumstances where there is a reasonable 
possibility that a project that is not part of a major modification may 
result in a significant increase of a regulated NSR pollutant, before 
beginning actual construction, you may choose to make and record a 
projection of post-change emissions of that pollutant from changed 
units.\14\
---------------------------------------------------------------------------


    \14\ Note that we plan, in the near future, to issue a Notice of 
Proposed Rulemaking that will address the issue of 
``debottlenecking.'' In today's rulemaking, we do not intend to 
change current requirements related to ``debottlenecking.'' Use of 
the term ``changed unit'' should not be interpreted as a change to 
those requirements.
---------------------------------------------------------------------------


    To make this projection, you must use the maximum annual rate at 
which the changed units are projected to emit the pollutant in any of 
the 5 calendar years following the time the unit resumes regular 
operations after the project (or 10 years if the project increases the 
unit's design capacity or potential to emit the regulated NSR 
pollutant). You then use these projections to calculate whether the 
project will result in a significant emissions increase. In making this 
calculation, you could exclude any emissions that the unit could have 
accommodated before the change and that are unrelated to the project. 
You could also exclude emissions resulting from increased utilization 
due to demand growth that the unit could have accommodated before the 
change.
    With respect to the covered changes, if you use this procedure, you 
are required to track post-change annual emissions of the units in tpy 
for the next 5 years (or 10 years if the project increases the unit's 
design capacity or potential to emit the regulated NSR pollutant). At 
the end of each year, if post-change annual emissions exceed the 
baseline actual emissions by a significant amount, and differ from your 
projections, you must submit a report to the reviewing authority with 
that information within 60 days after the end of the year.
    Instead of relying on projected actual emissions, you may instead 
elect to use the unit's PTE, in tpy. In that case, you need not track 
or report post-change emissions.
    We are also revising the procedures for projecting future emissions 
for EUSGUs to conform with these new procedures and consolidate the 
EUSGU and non-EUSGU procedures into a single set of provisions. As a 
result of our 1992 rulemaking, EUSGUs have available to them a similar 
set of procedures. We believe the procedures we are implementing for 
other units represent a sensible refinement of the rules we promulgated 
in 1992 and that we should make these procedures available to all 
existing units. We do, however, impose two requirements on EUSGUs 
beyond those we impose on other units. First, with respect to covered 
projects, EUSGUs that project post-change emissions will have to submit 
a copy of their projections to their reviewing authority before 
beginning actual construction. You will not be required to obtain any 
kind of determination from the reviewing authority before proceeding 
with construction. Second, we are requiring that if you project post-
change emissions for your EUSGUs, you must send a copy of your tracked 
emissions to your reviewing authority, without regard to whether these 
emissions have increased by a significant amount or exceed your 
projections. The effect of this consolidation is that we make minor 
changes to the existing procedures for EUSGUs. For example, you must 
project emissions for EUSGUs on a 12-month basis, rather than the 
current approach of projecting average annual emissions for the 2 years 
immediately following the change. Also, you need only make and report a 
projection for EUSGUs when there is a reasonable possibility that the 
given project may result in a significant emissions increase.
    By allowing you to use today's new version of the actual-to-
projected-actual applicability test to evaluate modified existing 
emissions units, we expect that fewer projects will trigger the major 
NSR permitting requirements. Nonetheless, we believe that the 
environment will not be adversely affected by these changes and in some 
respects will benefit from these changes. The new test will remove 
disincentives that discourage sources from making the types of changes 
that improve operating efficiency, implement pollution prevention 
projects, and result in other environmentally beneficial changes. 
Moreover, the end result is that State and local reviewing authorities 
can appropriately focus their limited resources on those activities 
that could cause real and significant increases in pollution.
    In addition, today's changes provide benefits to the public and the 
environment through the improved recordkeeping and reporting 
requirements as discussed above. We believe that these added 
recordkeeping and reporting measures will provide the information 
necessary for reviewing authorities to assure that such changes are 
made consistent with the CAA requirements. The new rule also does not 
affect the way in which a source's ambient air quality impacts are 
evaluated. Altogether, we believe that today's regulatory amendments 
focus on the types of changes occurring at existing emissions units 
that are more likely to result in significant contributions to air 
pollution.


B. What We Proposed and How Today's Action Compares


1. July 23, 1996 Notice of Proposed Rulemaking (NPRM)
    In 1996, we proposed to amend the NSR rules to allow States to use, 
among other things, a new test as an alternative to the actual-to-
potential test for determining the applicability of the NSR 
requirements when you wish to make modifications at an existing major 
stationary source. The proposed test was intended to apply exclusively 
to modifications of existing emissions units at major stationary 
sources--not to new emissions units. As described more completely 
below, the proposed test involved changes to the procedures for 
calculating an emissions unit's pre-change (baseline) actual emissions 
and post-change (future) actual emissions. The method would have also 
required you to monitor and report future emissions from certain 
modified


[[Page 80193]]


emissions units, based on the monitoring and reporting requirements 
adopted under the WEPCO amendments.
    Baseline actual emissions. In our 1996 NPRM, we proposed to change 
the definition of baseline emissions from the average annual rate of 
actual emissions during the 2-year period preceding the date of the 
modification to the annual rate associated with the highest level of 
utilization from any consecutive 12-month period during the 10-year 
period preceding the date of the modification, adjusted for any more 
stringent limits that may have been imposed since the end of the 12-
month period selected. The proposed method was intended to be used for 
calculating baseline actual emissions for any existing emissions unit, 
including EUSGUs, by replacing both the original method (that was part 
of the actual-to-potential test) and the 2-in-5-years method (as 
adopted under the WEPCO for modified EUSGUs).
    As indicated above, the proposed procedure also would have required 
you to take into account any legally enforceable constraints imposed on 
the facility since the selected 12-month time frame, and currently in 
effect. Thus, you would generally have been required to calculate the 
modified emissions unit's baseline actual emissions by using the 
appropriate utilization level from the selected 12-month period, in 
combination with the emissions unit's current enforceable emission 
factors. Such enforceable emission factors would have included current 
Federal and State limits, such as RACT (Reasonably Available Control 
Technology), MACT (Maximum Achievable Control Technology), BACT, LAER, 
and New Source Performance Standards (NSPS), as well as enforceable 
limits resulting from any voluntary reductions you may have taken (for 
example, for netting, offsets, or Emission Reduction Credits (ERCs)). 
Also, you would have had to consider any operational constraints that 
are enforceable, such as production limits, fuel use limits, or limits 
to the number of hours per day or days per year at which the unit 
modified, or affected by such modification, could operate.
    Finally, we indicated that it was not our intent to extend the 5-
year contemporaneous period (for considering creditable emissions 
increases and decreases as part of the netting calculus), even if we 
established a 10-year baseline look back period.
    Post-change actual emissions. In the 1996 proposal, we proposed to 
extend the availability of the actual-to-future-actual emissions 
method, established under the WEPCO amendments exclusively for EUSGUs, 
to predict the future actual emissions from any emissions unit 
undergoing a physical or operational change. Thus, we proposed 
extending availability of the definition of ``representative actual 
annual emissions'' to all emissions units undergoing a physical or 
operational change. This definition would have provided the basis for 
you to project an emissions unit's future actual emissions, excluding 
any emissions increases caused by demand growth or other independent 
factors, when determining whether the change at issue will increase 
emissions over the baseline levels.\15\
---------------------------------------------------------------------------


    \15\ This method, as well as the WEPCO amendments as a whole, 
was limited to modifications of existing EUSGUs and did not apply to 
the addition of a new emissions unit or the replacement of an 
existing unit.
---------------------------------------------------------------------------


    The proposal also retained the WEPCO provision requiring that, for 
any modified emissions unit using the actual-to-future-actual test, you 
must submit annually for 5 years after the change sufficient records to 
demonstrate that the change has not resulted in a significant emissions 
increase over the baseline levels. As a safeguard, the WEPCO rule also 
provides that this tracking period could be extended to 10 years when 
the reviewing authority is concerned that the first 5 years will not be 
representative of normal source operation. We sought comments on 
numerous issues, including whether any changes should be made to the 5-
year tracking requirement or to the demand growth exclusion in the 
event that we decided to broaden use of the actual-to-future-actual 
test for modifications to any existing emissions unit.
2. July 24, 1998 Notice of Availability
    In 1998, we announced that comments received on the 1996 proposal 
and changed circumstances had caused us to ask whether we should 
reconsider some of the aspects of the proposed changes to the ``major 
modification'' applicability test. The 1998 NOA set forth for public 
comment an additional applicability test. In brief, the alternative 
presented for additional comment would have: (1) Retained the actual-
to-future-actual test for EUSGUs and applied it to all source 
categories; (2) made binding for a 10-year period the emissions levels 
used in projecting future actual emissions following the modification 
for all source categories; and (3) eliminated the demand growth 
exclusion for calculating a modified emissions unit's future actual 
emissions.
    Consistent with the 1996 NPRM, this alternative methodology would 
have applied to any existing emissions unit at a major stationary 
source for which you might plan a non-routine physical or operational 
change. The methodology would have required you first to determine 
which emissions units were being changed, or were affected by the 
change, then to calculate those units' baseline actual emissions based 
on the highest consecutive 12 months of source operation during the 
past 10 years, adjusted to reflect current emission factors.
    The second step involved the forecast of future emissions resulting 
from the physical or operational change. Under this calculation of 
future actual emissions, one would not have been allowed to exclude 
predicted capacity utilization increases that were due to demand 
growth. If the difference between the pre-change and post-change actual 
emissions equaled or exceeded the significant emissions rate defined 
for a particular pollutant, major NSR would have been triggered (unless 
you took enforceable limits to keep the increase below significant 
levels or were otherwise able to net out of review using creditable, 
contemporaneous emissions increases and decreases occurring at your 
facility). If the difference between baseline and future actual 
emissions did not exceed the applicable significant emissions rate, 
your facility would not be subject to major NSR, but you would have 
been required to accept a temporary emissions cap based on the 
predicted future actual emissions for each affected pollutant at the 
emissions units being modified or affected by the modification.
    The temporary cap would have become an enforceable condition of a 
preconstruction permit. Also, the sole purpose of the temporary cap 
would have been to make sure that the physical or operational change 
did not result in a significant emissions increase, and the cap would 
have applied to those emissions units for at least 10 years after the 
changes were completed. You would also have been required to supply 
information annually to demonstrate that the future actual emissions 
did not exceed the applicable emissions caps during the 10-year period 
following the modification.
3. Summary of Major Changes in the Final Rule
    Today's action amends the existing NSR regulations to provide you 
with a common applicability test for all existing emissions units--the 
actual-to-projected-actual applicability test. This test has changed in 
some ways from both the 1996 NPRM and the 1998 NOA. As described in 
greater detail in sections


[[Page 80194]]


II.C and II.D below, the key features of the methodology are as 
follows.
    [sbull] If you are an existing emissions unit (other than an 
EUSGU), you will determine the pre-change (baseline) actual emissions 
by calculating an average annual emissions rate, in tpy, using any 
consecutive 24 months during the 10-year period immediately preceding 
the change. This rate must be adjusted downward to reflect any legally 
enforceable emission limitations imposed after the selected baseline 
period.
    [sbull] We are codifying the ``2-in-5-years'' presumption for 
calculating the baseline actual emissions for EUSGUs.
    [sbull] If you are an existing emissions unit (including EUSGUs), 
you will estimate post-change emissions (projected actual emissions), 
in tpy, to reflect any increase in annual emissions that may result 
from the proposed change. You should exclude, in calculating any 
increase in emissions that results from the particular project, that 
portion of the unit's emissions following the project that an existing 
unit could have accommodated during the baseline period and that is 
also unrelated to the particular project, including any increased 
utilization due to product demand growth. You must make the projection 
before you begin actual construction. When using this method, you must 
record the projection and certain other information in circumstances 
where there is a reasonable possibility that a change may result in a 
significant emissions increase. In addition, EUSGUs must send a copy of 
the projections and other information to your reviewing authority 
before beginning actual construction.
    [sbull] If, for a project at an existing emissions unit (other than 
an EUSGU) at a major stationary source, you elect to project your post-
change emissions, we are also requiring you to maintain information on 
these emissions, for 5 years following a physical or operational 
change, or in some cases for 10 years depending on the nature of the 
change. If your annual emissions exceed the baseline actual emissions 
by a significant amount and also exceed your projection, you must 
report this information to your reviewing authority within 60 days 
after the end of the year.
    [sbull] If you project post-change emissions for EUSGUs, you must 
report these emissions to your reviewing authority within 60 days after 
the end of the year without regard to whether such emissions exceed the 
baseline actual emissions or projected actual emissions for a period of 
5 years (or in some cases 10 years, depending on the nature of the 
change).
    [sbull] Instead of projecting your post-change emissions, for all 
existing emissions units you may instead project post-change emissions 
on the basis of each unit's post-change PTE. If you use this method, 
you need not record your projections or track or report post-change 
emissions.
    As discussed earlier, our prior regulations provide that when your 
emissions unit, other than an EUSGU, ``has not begun normal operations, 
``actual emissions equal the PTE of the unit. There have been 
considerable number issues raised with this approach. For example, 
using PTE as a measure of post-change emissions automatically 
attributes all possible emissions increases to the change. There are 
many cases, however, where this simply is not true. Moreover, when the 
actual-to-potential test is applied, it is automatically assumed that 
the emissions unit has not begun normal operations after the change 
period. In many such cases, however, the changed unit as a practical 
matter will function essentially as it did before the change. We are, 
therefore, allowing all existing emissions units to use an actual-to-
projected-actual applicability test. Accordingly, we are generally 
eliminating the term ``begun normal operations'' from the determination 
of whether a change results in a significant emissions increase.\16\
---------------------------------------------------------------------------


    \16\ We do make use of the term ``resumes regular operations'' 
(as opposed to ``normal operations'') in the final rule, but that 
term has a very different meaning and we are using it for an 
entirely different purpose. Specifically, we are not using the term 
for purposes of determining whether a change results in a 
significant emissions increase. Rather, we use it only to identify 
the date on which the owner or operator must begin tracking 
emissions of changed units when using the actual-to-projected-actual 
method.
---------------------------------------------------------------------------


    For essentially the same reasons, while our 1992 rules did not 
authorize use of projections in evaluating whether replacement of an 
existing emissions unit (which we understood to require application of 
the NSPS 50 percent cost threshold) constitutes a major modification, 
upon reflection we have decided this exception to the availability of 
the actual-to-projected-actual applicability test is also unnecessary. 
In our 1980 rulemaking, we decided against applying PSD to 
``reconstruction,'' even of entire sources, on the grounds that, as to 
existing sources that would not otherwise be subjected to PSD review as 
a major modification (i.e., such source would not cause a significant 
net emissions increase), changes that had no emission consequences 
should not be subject to PSD regardless of their magnitude.\17\
---------------------------------------------------------------------------


    \17\ The 1980 rulemaking also discussed that ``reconstruction'' 
would have only been applied on a plantwide basis and EPA believed 
that there would be few instances of plantwide reconstructions.
---------------------------------------------------------------------------


    In addition, we now believe that, as with modified units, the fact 
that replacement units are replacing similar units with a record of 
historical operational data provides sufficient reasons to believe that 
a projection of future actual emissions can be sufficiently reliable 
that an up-front emissions cap based on PTE is unnecessary. In other 
words, a source replacing a unit should be able to adequately project 
and track emissions for the replacement unit based, in part, on the 
operating history of the replaced unit. In contrast, sources adding 
``new'' units that do not qualify as replacement units must project 
that the future emissions of the new unit equal its PTE, effectively 
applying the ``actual-to-potential'' test because there is no relevant 
historical data that could be used to establish an actual emissions 
baseline or projection of future actual emissions for such new units.
    For these reasons, we have eliminated the requirement that replaced 
or reconstructed units be evaluated as to whether they constitute major 
modifications on an actual-to-potential basis. Instead, you may compare 
an emission unit's baseline actual emissions with your projected actual 
emission in measuring whether the replacement or reconstruction has 
resulted in a significant emissions increase. You must treat these 
emissions units as modifications only if the replacement or 
reconstruction of the unit results in a signficant increase so 
measured.\18\
---------------------------------------------------------------------------


    \18\ For simplicity, we state this rule without addressing 
whether the replacement or reconstruction has resulted in a 
significant net emissions increase, but under our two-step approach 
for evaluating whether a change constitutes a major modification, a 
significant net emissions increase would of course also be required. 
We have also retained the term ``representative of normal 
operations'' in the context of an EUSGU's option to seek use of a 
different baseline period, but there the question whether to seek 
such use is at the source's option, obviating many of the 
difficulties with it in other contexts.


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


[[Page 80195]]


C. Changes to the Procedures for Calculating the Pre-Change Baseline 
Actual Emissions for Existing Emissions Units Other Than EUSGUs


1. Under Today's New Requirements, How Should I Calculate the Pre-
Change Baseline Actual Emissions for an Existing Emissions Unit That Is 
Not an EUSGU?
    When you calculate the baseline actual emissions for an existing 
emissions unit (other than an EUSGU), you may select any consecutive 24 
months of source operation within the past 10 years. Using the relevant 
source records for that 24-month period, including such information as 
the utilization rate of the equipment, fuels and raw materials used in 
the operation of the equipment, and applicable emission factors, you 
must be able to calculate an average annual emissions rate, in tpy, for 
each pollutant emitted by the emissions unit that is modified, or is 
affected by the modification.
    The new requirements prohibit you from counting as part of the 
baseline actual emissions any pollution levels that are not allowed 
under any legally enforceable limitations and that apply at the time of 
the project. Therefore, you must identify the most current legally 
enforceable limits on your emissions unit. If these legally enforceable 
emission limitations and operating restrictions are more stringent than 
those that applied during the 24-month period, you must adjust downward 
the average annual emissions rate that you calculated from the 
consecutive 24-month period to reflect these current restrictions. (See 
section II.C.5 of this preamble for further discussion of the 
adjustment that you may need to make.)
    In summary, when the average annual emissions rate that you 
originally calculated is still legally achievable (see discussion 
below), then your baseline actual emissions will be the same as the 
average annual emissions rate calculated from the 24-month period. If 
it is not, you must adjust it downward so that it does not reflect 
emissions that are no longer legally allowed.
2. Can Existing Emissions Units (Other Than EUSGUs) Still Use a ``More 
Representative Time Period'' for Selecting the Baseline Actual 
Emissions?
    No, under today's new requirements neither you nor your reviewing 
authority will have the authority to select another period of time from 
which to calculate your baseline actual emissions. You must select a 
24-month period within the 10-year period before the physical or 
operational change.
3. From What Point in Time Is the 10-Year Look Back Measured?
    If you believe that you will need either a major or minor NSR 
permit to proceed with your proposed physical or operational change, 
then you must use the 10-year period immediately preceding the date on 
which you submit a complete permit application. If, however, you 
believe that the physical or operational change(s) you plan to make 
will not result in either a significant emissions increase from the 
project or a significant net emissions increase at your major 
stationary source (that is, your project will not be a major 
modification), and you are not otherwise required to obtain a minor NSR 
permit before making such change, then you must use the 10-year period 
that immediately precedes the date on which you begin actual 
construction of the physical or operational change.
4. What if, for an Existing Emissions Unit (Other Than an EUSGU), I Do 
Not Have Adequate Documentation for Its Operation for the Past 10 
Years?
    Your ability to use the full 10 years of the look back period will 
depend upon the availability of relevant data for the consecutive 24-
month period you wish to select. The data must adequately describe the 
operation and associated pollution levels for the emissions units being 
changed. If you do not have the data necessary to determine the units' 
actual emission factors, utilization rate, and other relevant 
information needed to accurately calculate your average annual 
emissions rate during that period of time, then you must select another 
consecutive 24-month period within the 10-year look back period for 
which you have adequate data.
5. For an Existing Unit (Other Than EUSGUs), When Must I Adjust My 
Calculation of the Pre-Change Baseline Actual Emissions?
    Today's amendments require you to adjust the average annual 
emissions rate derived from the selected 24-month period under certain 
circumstances. Specifically, you must adjust downward this average 
annual rate if any legally enforceable emission limitations, including 
but not limited to any State or Federal requirements such as RACT, 
BACT, LAER, NSPS, and National Emission Standards for Hazardous Air 
Pollutants (NESHAP), restrict the emissions unit's ability to emit a 
particular pollutant or to operate at levels that existed during the 
selected 24-month period from which you calculate the average annual 
emissions rate. For example, assume that during the selected 
consecutive 24-month period you burned fuel oil and you were subjected 
to a sulfur limit of 2 percent sulfur (by weight). Today, you are only 
allowed to burn fuel oil with a sulfur content of 0.5 percent or less. 
Consequently, you would be required to adjust your preliminary 
calculation of baseline actual emissions for sulfur dioxide 
(SO2) (that is, substitute the lower sulfur limit into the 
emissions calculation, yielding a 75 percent reduction in the emissions 
rate from the initial calculation) to reflect the current restriction 
allowing only 0.5 percent sulfur in fuel oil. The original average 
annual utilization rate would not be adjusted unless a more stringent 
legally enforceable operational limitation has since been imposed that 
restricts that rate.
    You must also adjust for legally enforceable emission limitations 
you may have voluntarily agreed to, such as limits you may have taken 
in your permit for netting, emissions offsets, or the creation of ERCs. 
Also, you must adjust your emissions from the 24-month period if a raw 
material you used during the baseline period is now prohibited. For 
example, you may have used a paint with a high solvent concentration 
during a portion of the consecutive 24-month period. Today, you are 
prohibited from using that particular paint. You must then adjust your 
emissions rate to reflect the raw material restriction.
6. How Should I Calculate the Baseline Actual Emissions for Emissions 
Units (Other Than EUSGUs) That Use Multiple Fuels or Raw Materials?
    For an emissions unit that is capable of burning more than one type 
of fuel, you must relate the current emission factors to the fuel or 
fuels that were actually used during the selected 24-month period. For 
example, when calculating the baseline actual emissions for an 
emissions unit that burned natural gas for a portion of the 24-month 
period and fuel oil for the remainder, you must retain that fuel 
apportionment (for example, natural gas to fuel oil ratio), but you 
must also use the current legally enforceable emission factors for 
natural gas and fuel oil, respectively, to calculate the baseline 
actual emissions. If, however, you are no longer allowed or able to use 
one of those fuel types, then you must make your calculations assuming 
use of the currently allowed fuel for the entire 24-month period. You 
must use the same approach for emissions units that use multiple 
feedstock or raw materials, which may vary in use during the unit's 
ongoing production process.


[[Page 80196]]


7. How Should I Calculate the Baseline Actual Emissions for 
Construction Projects That Involve Multiple Units?
    Today's new requirements require that you select the same single 
consecutive 24-month period within the 10-year look back period to 
calculate the baseline actual emissions for all existing emissions 
units that will be changed. See, for example, new Sec.  
52.21(b)(48)(ii)(e). The result will be that the baseline actual 
emissions for each affected pollutant will be based on the same 
consecutive 24-month period as well.
    You will have the option to select the single 24-month period that 
best represents the collective level of operation (and emissions) for 
your existing emissions units.
    If a particular existing emissions unit did not yet exist during 
the 24-month period you select to calculate the baseline actual 
emissions, you must count that emissions unit's emissions rate as zero 
for that full period of time. If an emissions unit operated for only a 
portion of the particular 24-month period that you select, you must 
calculate its average annual emissions rate using an emissions rate of 
zero for that portion of time when the unit was not in operation.
    For new emissions units (a unit that has existed for less than 2 
years) that will be changed by the project, the baseline actual 
emissions rate is zero if you have not yet begun operation of the unit, 
and is equal to the unit's PTE once it has begun to operate.
8. Am I Able To Apply Today's Changes for Calculating the Baseline 
Actual Emissions to Other Major NSR Requirements?
    No, as stated in section II.A, you are only allowed to use the new 
baseline methodology in today's rule for three specific purposes 
involving existing emissions units as follows.
    [sbull] For modifications, to determine a modified unit's pre-
change baseline actual emissions as part of the new actual-to-
projected-actual applicability test
    [sbull] For netting, to determine the pre-change actual emissions 
of an emissions unit that underwent a physical or operational change 
within the contemporaneous period. You may select separate baseline 
periods for each contemporaneous increase or decrease.
    [sbull] For PALs, to establish the PAL level.
    If you determine that the modification of your source is a major 
modification, you must revert to using the existing definition of 
``actual emissions'' to determine your source's actual emissions on a 
particular date to satisfy all other NSR permitting requirements, 
including any air quality analyses (for example, compliance with NAAQS, 
PSD increments, AQRVs) and the amount of emissions offsets required.
    For example, when you must determine your source's compliance with 
the PSD increments following a major modification, you must still use 
the allowable emissions from each emissions unit that is modified, or 
is affected by the modification. An existing source's contribution to 
the amount of increment consumed should be based on that source's 
actual emissions rate from the 2 years immediately preceding the date 
of the change, although the reviewing authority shall allow the use of 
another 2-year period if it determines that such period is more 
representative of that source's normal operation. See, for example, 
Sec.  52.21(b)(21)(ii).
    Also, any determination of the amount of emissions offset that must 
be obtained by a major modification subject to the nonattainment NSR 
requirements under Sec.  51.165(a) should be based on calculations 
using the existing definitions of ``actual emissions'' and ``allowable 
emissions.'' See new Sec.  51.165(a)(3)(ii)(H).


D. The Actual-to-Projected-Actual Applicability Test for Physical or 
Operational Changes to Existing Emissions Units Including EUSGUs


1. How are post-change actual emissions calculated under today's 
revised rule?
    Today, we are amending the major NSR rules to enable you to use an 
applicability test that is similar to the applicability test that 
currently applies to EUSGUs (that is, the actual-to-representative-
actual-annual emissions test). The new test allows you to project the 
post-change emissions of all modified existing emissions units 
(including EUSGUs) in the same manner. That is, under today's new 
provisions for non-routine physical or operational changes to existing 
emissions units, rather than basing a unit's post-change emissions on 
its PTE, you may project an annual rate, in tpy, that reflects the 
maximum annual emissions rate that will occur during any one of the 5 
(or in some circumstances 10) years immediately after the physical or 
operational change. The first year begins on the day the emissions unit 
resumes regular operation following the change and includes the 12 
months after this date. This projection of the unit's annual emissions 
rate following the change is defined as the ``projected actual 
emissions'' (see, for example, Sec.  52.21(b)(48)), and will be based 
on your maximum annual rate in tons per year at which you are projected 
to emit a regulated NSR pollutant, less any amount of emissions that 
could have been accommodated during the selected 24-month baseline 
period and is not related to the change. Accordingly, you will 
calculate the unit's projected actual emissions as the product of: (1) 
The hourly emissions rate, which is based on the emissions unit's 
operational capabilities following the change(s), taking into account 
legally enforceable restrictions that could affect the hourly emissions 
rate following the change(s); and (2) the projected level of 
utilization, which is based on both the emissions unit's historical 
annual utilization rate and available information regarding the 
emissions unit's likely post-change capacity utilization. In 
calculating the projected actual emissions, you should consider both 
the expected and the highest projections of the business activity that 
you expect could be achieved and that are consistent with information 
your company publishes for business-related purposes such as a 
stockholder prospectus, or applications for business loans. From the 
initial calculation, you may then make the appropriate adjustment to 
subtract out any portion of the emissions increase that could have been 
accommodated during the unit's 24-month baseline period and is 
unrelated to the change. Once the appropriate subtractions have been 
made, the final value for the projected actual emissions, in tpy, is 
the value that you compare to the baseline actual emissions to 
determine whether your project will result in a significant emissions 
increase.
    The adjustment to the projected actual emissions allows you to 
exclude from your projection only the amount of the emissions increase 
that is not related to the physical or operational change(s). In 
comparing your projected actual emissions to the units' baseline actual 
emissions, you only count emissions increases that will result from the 
project. For example, as with the electric utility industry, you may be 
able to attribute a portion of your emissions increase to a growth in 
demand for your product if you were able to achieve this higher level 
of production during the consecutive 24-month period you selected to 
establish the baseline actual emissions, and the increased demand for 
the product is unrelated to the change.
    For Clean Units, if a given project can be constructed and operated 
at a Clean Unit without causing the emissions unit


[[Page 80197]]


to lose its Clean Unit status, then no emissions increase will occur.
    For new units, however, you must continue to calculate post-change 
emissions on the basis of a unit's PTE.
2. Will My Projection of Projected Actual Emissions Become an 
Enforceable Emission Limitation as Suggested in the 1998 NOA?
    No, we did not adopt such a requirement. If you have an existing 
emissions unit and your project results in an increase in annual 
emissions that exceeds the baseline actual emissions by a significant 
amount, and differs from your projection of post-change emissions that 
you were required to calculate and maintain records of, then you must 
report this increase to your reviewing authority within 60 days after 
the end of the year. Since modified EUSGUs are required to report their 
post-change annual emissions to the reviewing authority annually, any 
occurrence of a significant increase will be covered under that report 
for the affected calendar year. See section II.D.6 of this preamble for 
a more detailed discussion of the reporting requirements.
3. How Do I Determine How Long My Post-Change Emissions Will Be Tracked 
To Ensure That My Project Is Not a Major Modification?
    Generally, your projected actual emissions must be tracked against 
your facility's post-change emissions for 5 years following resumption 
of regular operations whether you are an EUSGU or other type of 
existing emissions unit. We will presume that any increases that occur 
after 5 years are not associated with the physical or operational 
changes. However, you may be required to track emissions for a longer 
period of time under the following circumstances. If you are an 
existing emissions unit and one of the effects of your physical or 
operational change(s) is to increase a unit's design capacity or PTE, 
you must track your emissions for a period of 10 years after the 
completion of the project. This extended period allows for the 
possibility that you could end up using the increased capacity more 
than you projected and such use might lead to significant emissions 
increases.
4. What Are the Reporting and Recordkeeping Requirements for Projects?
    Reporting and recordkeeping for a project is required when three 
criteria are met: (1) You elect to project post-change emissions rather 
than use PTE; (2) there is a reasonable possibility that the project 
will result in a significant emissions increase; and (3) the project 
will not constitute a major modification. In such circumstances, you 
must document and maintain a record of the following information: a 
description of the project; an identification of emissions units whose 
emissions could increase as a result of the project; the baseline 
actual emissions for each emissions unit; and your projected actual 
emissions, including any emissions excluded as unrelated to the change 
and the reason for the exclusion. In addition, if your project increase 
is significant, you must record your netting calculations if you use 
emissions reductions elsewhere at your major stationary source to 
conclude that the project is not a major modification. For covered 
projects, you must record this information before beginning actual 
construction. If you are an EUSGU, you must also send this information 
to your reviewing authority before beginning actual construction. Note, 
however, that if you chose to use potential emissions as your 
projection of post-change emissions, you are not required to maintain a 
record of this decision.
    In addition, today's final rules require you to maintain emissions 
data for all emissions units that are changed by the project. You must 
maintain this information for 5 years, or 10 years if applicable. The 
information you must maintain may include continuous emissions 
monitoring data, operational levels, fuel usage data, source test 
results, or any other readily available information of sufficient 
accuracy for the purpose of determining an emissions unit's post-change 
emissions.
    If you are an EUSGU, you must report this information to your 
reviewing authority within 60 days after the end of any year in which 
you are required to generate such information. Other existing units 
must report to the reviewing authority any increase in the post-change 
annual emissions rate when that rate: (1) Exceeds the baseline actual 
emissions by a significant amount, and (2) differs from the projection 
that was calculated before the change. See, for example, new Sec.  
52.21(r)(6)(iii).
    In addition to the reporting requirements discussed above, you are 
also obligated to ensure that the necessary emissions information you 
are required to maintain is available for examination upon request by 
the reviewing authority or the general public.
5. How Do Today's Changes Affect the Netting Methodology for Existing 
Emissions Units (Other Than EUSGUs)?
    If your calculations show that a significant emissions increase 
will result from a modification, you have the option of taking into 
consideration any contemporaneous emissions changes that may enable you 
to ``net out'' of review, that is, show that the net emissions increase 
at the major stationary source will not be significant. The 
contemporaneous time period will not change under the Federal PSD 
program as a result of today's action. That is, creditable increases 
and decreases in emissions that have occurred between the date 5 years 
before construction of the particular change commences and the date the 
increase from that change occurs are contemporaneous. See Sec.  
52.21(b)(3)(ii). States will continue to have some discretion in 
defining ``contemporaneous'' for their own NSR programs.
    Although we are not changing our definition of ``contemporaneous,'' 
today's action allows existing emissions units (other than EUSGUs) to 
calculate the baseline actual emissions for each contemporaneous event 
using the 10-year look back period. That is, you can select any 
consecutive 24-month period during the 10-year period immediately 
preceding the change occurring in the contemporaneous period to 
determine the baseline actual emissions for each creditable emissions 
change. Generally, for each emissions unit at which a contemporaneous 
emissions change has occurred, you should use the 10-year look back 
period relevant to that change.\19\ When evaluating emissions increases 
from multi-unit modifications, if more than one emissions unit was 
changed as part of a single project during the contemporaneous period, 
you may select a separate consecutive 24-month period to represent each 
emissions unit that is part of the project. In any case, the calculated 
baseline actual emissions for each emissions unit must be adjusted to 
reflect the most current emission limitations (including operational 
restrictions) applying to that unit. ``Current'' in the context of a 
contemporaneous emissions change refers to limitations on emissions and 
source operation that existed just prior to the date of the 
contemporaneous change.
---------------------------------------------------------------------------


    \19\ Your ability to use the full 10 years for calculating any 
contemporaneous emissions change is contingent upon the availability 
of valid and sufficient source information for the selected 24-month 
period. See, for example, new Sec.  52.21(b)(48)(ii)(f).
---------------------------------------------------------------------------


E. Clarifying Changes to WEPCO Provisions for EUSGUs


    The method you use to calculate the baseline actual emissions for 
an existing EUSGU to determine whether there is a


[[Page 80198]]


significant emissions increase from a physical or operational change at 
an EUSGU, and to determine whether a significant net emissions increase 
will occur at the major stationary source, will not change as a result 
of today's final rulemaking. The rule provides that for an existing 
EUSGU you may calculate the baseline actual emissions as the average 
annual emissions (tpy) of the emissions unit using any 2-year period 
out of the 5 years immediately preceding the modification. (This was 
set out as a presumption in the preamble for the 1992 WEPCO 
amendments.) This rule recognizes the ordinary variability in demand 
for electricity. See, for example, new Sec.  52.21(b)(21)(ii).
    For example, a cold winter or hot summer will result in high levels 
of demand while a relatively mild year will produce lower demand. By 
allowing a utility to use any consecutive 2 years within the past 5, 
the rule recognizes that electricity demand and resultant utility 
operations fluctuate in response to various factors such as annual 
variability in climatic or economic conditions that affect demand, or 
changes at other plants in the utility system that affect the dispatch 
of a particular plant. By allowing utilities to use as a baseline any 
consecutive 2 years in the last 5 years, these types of fluctuations in 
operations can be more realistically considered.
    The reviewing authority shall allow the use of a different time 
period upon a determination that it is more representative of normal 
source operation.
    In an August 6, 2001 letter,\20\ we addressed the issue of whether 
combined cycle gas turbines (the gas turbines and waste heat recovery 
components) came within the definition of ``electric utility steam 
generating units'' for the purpose of determining whether such units 
are eligible to use the WEPCO ``applicability test.'' The letter 
concluded that ``steam generating units'' include not only electric 
utility plants with boilers, but also plants with combined cycle gas 
turbines if the combined cycle gas turbine systems supply more than 
one-third of their potential electric output capacity and more than 25 
MW electrical output to any utility power distribution system for sale. 
Consequently, qualifying combined cycle gas turbines must also use the 
2-in-5-years baseline method.
---------------------------------------------------------------------------


    \20\ Letter from John S. Seitz, Director, Office of Air Quality 
Planning and Standards, to Patrick M. Raher, August 6, 2001.
---------------------------------------------------------------------------


    Finally, today's rules provide the same method for EUSGUs that will 
exist for all other existing emissions units to project post-change 
emissions following a physical or operational change to a unit. In the 
1996 proposal, we proposed a range of options for addressing the 
applicability of changes that are made to existing emissions units, 
including the option of extending the actual-to-future-actual test, 
then available only to utilities, to all source categories. While we 
have decided to leave the WEPCO rules intact in most respects, we 
believe that it is reasonable and appropriate to establish a consistent 
method for sources to use for projecting the post-change emissions that 
will result from a physical or operational change to an existing 
emissions unit. Therefore, under today's new rules, the current method 
of basing the projection on the 2 years following the change to an 
EUSGU is being replaced with the method available to all other existing 
units, under which you project a unit's post-change emissions as the 
maximum annual rate that the unit will emit in any one of the 5 years 
following resumption of regular operations.


F. The ``Hybrid'' Applicability Test for Projects Affecting Multiple 
Types of Emissions Units


1. When Does the Hybrid Applicability Test Apply to You?
    The hybrid applicability test applies if you plan a project (or 
series of related projects) that will affect emissions units of two or 
more of the following types.
    [sbull] Existing emissions units
    [sbull] New emissions units
    [sbull] Clean Units
2. How Do I Determine Whether My Project Will Result in a Significant 
Emissions Increase Under the Hybrid Test?
    For the first two types of emissions units listed above that are 
affected by the project, calculate the emissions increase as we have 
discussed previously in this preamble. That is, use the actual-to-
projected-actual applicability test for existing units and the actual-
to-potential test for new emissions units.
    Clean Units are discussed fully in section V of this preamble. If a 
given project can be constructed and operated at a Clean Unit without 
causing the emissions unit to lose its Clean Unit status, no emissions 
increase shall be deemed to occur at that Clean Unit. If a given 
project would cause the emissions unit to lose its Clean Unit status, 
then the increase in emissions should be calculated as if the emissions 
unit is not a Clean Unit.
    After you calculate the emissions increase for each relevant unit, 
total the increases across all the emissions units of all types. If 
this total emissions increase equals or exceeds the level defined as 
significant for the regulated NSR pollutant in question, the project 
will result in a significant emissions increase for that pollutant. 
You'll find the regulatory language for determining whether a project 
will result in a significant emissions increase at Sec. Sec.  
51.165(a)(2)(vii)(D), 51.166(a)(7)(vi)(d), and 52.21(a)(2)(vi)(d).
    In section II.C.8 of this preamble, we indicate that the baseline 
actual emissions for all units that are not EUSGUs that are changed by 
a project must be calculated based on the same consecutive 24-month 
period within the previous 10 years. The same principle applies under 
the hybrid test, but it can be slightly more complicated if both EUSGUs 
and non-EUSGUs are involved. In this case, you must use the same 
baseline period for all emissions units affected by the project. This 
baseline period must be selected so as to meet the requirements for 
both EUSGUs and non-EUSGUs. Thus, you must select a 2-year period out 
of the previous 5 years for your baseline period, as required for 
EUSGUs (and within the requirements for non-EUSGUs). If you wish to use 
another period that you believe is more representative (as allowed for 
EUSGUs), the entire period must fall within the previous 10 years (as 
required for non-EUSGUs).
3. How Do I Determine the Net Emissions Increase From My Project Under 
the Hybrid Test?
    If you conclude that a significant emissions increase will result 
from the proposed project, you have the option of taking into 
consideration any contemporaneous emissions changes that may enable you 
to ``net out'' of review, that is, show that the net emissions increase 
at the major stationary source will not be significant. The netting 
analysis is carried out under the hybrid test just as it is under the 
other applicability tests. Refer to section II.D.7 of this preamble for 
a discussion of netting methodology.


G. Legal Basis for Today's Action


    The Act defines modification for the purposes of PSD and 
nonattainment NSR through cross-reference to the NSPS definition of 
``modification.'' The NSPS definition states that a modification 
``means any physical change in, or change in the method of operation 
of, a stationary source which increases the amount of any air


[[Page 80199]]


pollutant emitted by such source or which results in the emission of 
any air pollutant not previously emitted.'' CAA section 111(a)(4), 42 
U.S.C. 7411(a)(4). The Act is silent, however, on the issue of how one 
is to determine whether a physical or operational change increases the 
amount of any air pollutant emitted by the source.
    Accordingly, EPA is exercising its discretion in interpreting and 
providing clarity to this issue. We believe that the rules set forth 
today are ``a permissible construction of the statute.'' Chevron 
U.S.A., Inc. v. NRDC, 467 U.S. 843-4 (1984). The reviewing court should 
defer to it. Id. at 837.
    In the NSPS program, we determine whether there has been an 
``increase in any air pollutant emitted'' by the source by comparing 
its maximum hourly achievable emissions before and after the change. 
EPA and the courts have recognized, however, that the NSR programs and 
the NSPS programs have different goals,\21\ and thus, we have utilized 
different emissions tests in the NSR programs. Prior to today, the 
regulations applied an actual-to-future-actual applicability test for 
EUSGUs and an actual-to-potential applicability test for all other 
emissions units. Today, we are establishing a new applicability test 
for calculating emissions increases for ``Clean Units'' and an actual-
to-projected-actual applicability test for all other emissions units. 
We believe that establishing an actual-to-projected-actual 
applicability test for all emissions units is a reasonable 
interpretation of the phrase ``increase of any pollutant emitted.'' 
\22\
---------------------------------------------------------------------------


    \21\ See, for example, WEPCO Rule, 57 FR 32316 (``fundamental 
distinctions between the technology-based provisions of NSPS and the 
air quality-based provisions of NSR''). See also ASARCO Inc. v. EPA, 
578 F.2d 319 (D.C. Cir. 1978).
    \22\ The explanation of the applicability test for ``Clean 
Units'' is discussed in section V.
---------------------------------------------------------------------------


H. Response to Comments and Rationale for Today's Actions


    We received numerous comments on our proposed rule regarding the 
calculation of the baseline actual emissions and the actual-to-future-
actual test. Some of the significant comments and our responses to them 
are provided below. A complete set of comments and our responses can be 
found in the Technical Support Document located in the docket for this 
rulemaking.
1. Why Are We Extending the Look Back Period for Determining the 
Baseline Actual Emissions to 10 Years?
    Most commenters generally support our proposal to allow owners and 
operators to use a 10-year look back period to determine the baseline 
actual emissions for modifications at any existing emissions unit. 
Commenters have various reasons for supporting or opposing the proposed 
approach. Many supporters agree that extending the baseline look back 
period to 10 years would simplify current regulations and provide 
certainty to sources who otherwise would have to demonstrate to the 
reviewing authority that a period other than the 2 years immediately 
preceding the proposed change was more representative of normal source 
operation. Some commenters support the proposal because it would 
prevent the perceived confiscation of underused capacity at sources 
that have had low utilization rates for an extended period. These 
commenters agree that a 10-year look back period is more likely to 
afford a source a baseline actual emissions calculation that best 
reflects representative source operating conditions and would also 
account for fluctuations in the business cycle.
    Some commenters criticize the proposed 10-year look back period as 
being too long. These commenters recommend either a 5-year or 2-year 
look back period. One of these commenters states that the 10-year look 
back creates the opportunity for a source to increase production to the 
10-year maximum, and prevents the State or local air regulators from 
addressing the increase in emissions. Thus, the commenter believes that 
sources would be allowed to use historic emissions levels that are 
higher than current levels to establish the baseline actual emissions. 
Some commenters add that the proposed change would not reduce program 
complexity.
    Some commenters believe that instead of extending the period for 
establishing baseline actual emissions, the test for establishing 
modifications should be changed. According to the commenters, the 
problem is not that the current system does not go back far enough to 
set a fair actual emissions baseline, but that the methodology does not 
account for the fact that most emissions units are operating at an 
activity level much lower than the allowed activity level. The 
commenters believe that many of the real problems associated with the 
current major modification applicability test would be eliminated if 
the procedure was modified in an equitable manner.
    A commenter also adds that EPA may also want to include provisions 
that prevent a source from applying the new definition of actual 
emissions in a way that would retroactively enable the source to 
reverse a previous major modification determination and to eliminate 
any emissions reduction previously required for that major 
modification.
    We continue to believe that it is reasonable and appropriate to 
adopt the new method for establishing a modified unit's baseline actual 
emissions. It is important to understand the difference between the 
purpose of the new procedure, which uses the 10-year look back, and the 
existing procedure under the pre-existing definition of ``actual 
emissions'' at Sec.  52.21(b)(21(ii), which generally requires the use 
of an average annual emissions rate based on the 2-year period 
immediately preceding a particular date. The latter procedure is 
designed to estimate a source's actual emissions at a particular time 
and continues to be appropriate for such things as estimating a 
source's impact on air quality for PSD increment consumption.
    On the other hand, the new baseline procedure is specifically 
designed to allow a source to consider a full business cycle in 
determining whether there will be an emissions increase from a physical 
or operational change. Generally, a source's operations over a business 
cycle cover a range of operating (and emissions) levels--not simply a 
single level of utilization. The new procedure recognizes that market 
fluctuations are a normal occurrence in most industries, and that a 
source's operating level (and emissions) does not remain constant 
throughout a source's business cycle. The use of a 24-month period 
within the past 10 years to establish an average annual rate is 
intended to adjust for unusually high short-term peaks in utilization.
    Consequently, the new procedure ensures that a source seeking to 
make changes at its facility at a time when utilization may not be at 
its highest can use a normal business cycle baseline by allowing the 
source to identify capacity actually used in order to determine an 
average annual emissions rate from which to calculate any projected 
actual emissions resulting from the change.
    With respect to the commenters' general concerns that a 10-year 
look back period is too long, we sought to better understand what time 
period best represents an industry's normal business cycle. Therefore, 
we contracted for a study of several industries in 1997.\23\ This study 
found that, for the


[[Page 80200]]


industries analyzed, business cycles differ markedly by industry, and 
may vary greatly both in duration and intensity even within a 
particular industry. Nevertheless, we concluded from the study that 10 
years of data is reasonable to capture an entire industry cycle. 
Comments from various industries support a conclusion that a 10-year 
look back period is a fair and representative time frame for 
encompassing a source's normal business cycle.
---------------------------------------------------------------------------


    \23\ ``Business Cycles in Major Emitting Source Industries.'' 
Eastern Research Group; September 25, 1997. This study examined the 
business fluctuations for nine source categories described as CAA 
major emitting sources. Industry business cycles were examined using 
industry output data for the years 1982 to 1994 inclusive, based on 
the Office of Management and Budget's SIC codes for individual 
industries (OMB, 1987).
---------------------------------------------------------------------------


    We believe that the use of a 10-year look back period will help 
provide certainty to the process and eliminate the ambiguity and 
confusion that occurred when an applicant and the reviewing authority 
disagreed on what time frame provides the period most representative of 
normal source operation. The new requirements also provide certainty to 
the look back period, since there is no opportunity to select another 
period of time outside this 10-year period. (See additional discussion 
in section II.E.2.) In addition, we have placed certain restrictions on 
when the full 10-year look back period may be used. (See section 
II.E.3.)
    With regard to the concern that industry may try to apply the new 
requirements retroactively to undo current restrictions on existing 
sources, we want to reiterate that the new procedures do not apply 
retroactively to existing NSR permits or changes that sources have made 
in the past. Prior applicability determinations on major modifications 
and the control requirements that currently apply to sources remain 
valid and enforceable and have to be adjusted for in the calculation of 
baseline actual emissions. However, as part of the transition process 
for implementing the new provisions, we do intend to allow permit 
applicants to withdraw any permit applications submitted for review 
under the part 52 Federal PSD permit program so that they may re-
evaluate their projects in light of the new requirements. States may 
allow for the same type of transition process under their own NSR 
programs.
    Finally, we considered whether we should change the length of the 
look back period for EUSGUs for establishing the actual emissions 
baseline period to be consistent with the 10-year look back period we 
are adopting for other existing emissions units. The data we collected 
to support the 1992 rule changes show that allowing EUSGUs to use any 
2-year period out of the preceding 5 years is a sufficient period of 
time to capture normal business cycles at an EUSGU. We do not believe 
that any information received during the public comment period for this 
final rule adequately supports a different conclusion. Thus, we have 
decided to retain the 2-in-5-years baseline period for EUSGUs. However, 
for consistency with the baseline period for other existing emissions 
units, we have specified that the 2-year period is a consecutive 24-
month period.
2. Why Do the New Requirements Not Provide Discretion for the Reviewing 
Authority To Consider Another Time Period More Representative of Normal 
Operation for Non-EUSGUs?
    Several commenters oppose our proposed elimination of the reviewing 
authority's discretion to allow a different representative period 
(outside of the 10-year period), because they argue certain sources 
(for example, emissions units placed in cold reserve due to reduced 
demand) require this flexibility. Some commenters say the discretion 
should be given to the reviewing authority, while other commenters 
wanted the discretion given directly to source owners and operators. 
Instead of the discretion to use an alternate period, one commenter 
prefers that all sources should be required to show that they have 
selected a representative period that precedes the most recent 2-year 
period.
    We believe that use of a fixed 10-year look back period provides 
the desired clarity and certainty to the process of selecting an 
appropriate utilization/emissions level that is representative of a 
source's normal operation. A bounded 10-year look back provides 
certainty to the regulated community that may be undermined by an 
option to allow an unbounded alternative period as well.
3. Why Are We Placing Restrictions on the Use of a 10-Year Look Back 
for Setting the Baseline Actual Emissions?
    Numerous commenters responded to our concern that many sources 
might lack accurate records for the full 10-year look back period, and 
to our request for comments on the need to condition the full use of 
the 10-year period upon the accuracy and completeness of available 
data, as well as the need to establish specific criteria for accuracy, 
completeness, and recordkeeping when using older data. A number of 
commenters generally support limiting full use of the 10-year look back 
period to situations in which adequate emissions and/or capacity 
utilization data are available. Some commenters also recommend that EPA 
issue minimum criteria to reduce the number of case-by-case 
determinations and help reviewing authorities avoid debates with 
sources on what constitutes sufficient data.
    On the other hand, one commenter recommends that we not adopt a 
variable look back period based on the quality of the older data 
because it would ``add considerable uncertainty and protracted debate 
to the process. . . .'' If, however, we choose to limit the look back 
period based on the quality of older data, then this commenter and 
several others prefer provisions allowing for case-by-case decisions by 
State or local reviewing authorities over specific criteria established 
by EPA.
    Today's amendments condition the full use of the new 10-year look 
back period on the accuracy and completeness of your records of 
emissions and capacity utilization, with respect to the 24-month period 
you select, for any emissions unit that undergoes a physical or 
operational change. See, for example, new Sec.  52.21(b)(48)(f). As 
with all emissions calculations, accuracy and completeness are central 
elements for applicability determinations. In many cases, sources 
presently maintain accurate records on emissions and operations for 
only 3 to 5 years. Thus, we think it is appropriate to limit use of the 
full 10-year look back period when you do not have adequate data for 
the time period you wish to select. However, this limitation should be 
alleviated over time as sources begin to maintain records for longer 
periods to accommodate the 10-year look back opportunity.
    We also agree that adequacy of any given data should be left to the 
case-by-case judgment of individual reviewing authorities. The type of 
data necessary to determine emissions will vary drastically from source 
category to source category and from process to process within a source 
category. At this time, we are not able to issue generic criteria that 
would apply to all types of industries.
    We are further restricting your use of the 10-year look back for 
emissions units that are located in nonattainment areas and OTRs. In 
such cases, you are precluded from using any portion of the 10-year 
look back that precedes November 15, 1990--the date of the 1990 CAA 
Amendments--to establish baseline actual emissions for those units. 
This limit on the use of the 10-year look back is consistent the intent 
of the 1996 NPRM, which was originally proposed to apply to the use of 
the 10-year look back for any modification of an existing facility in a 
nonattainment


[[Page 80201]]


area or OTR. See 61 FR 38259 (July 23, 1996). However, because we are 
now beyond the point where the November 15, 1990 limit is relevant to 
modifications, we are only applying this limitation in the netting 
context with respect to emissions units changed within the 
contemporaneous period.
4. Why Were Changes Made to the Proposed Approach for Establishing 
Baseline Actual Emissions Using a 10-Year Look Back?
    Commenters raise specific questions about how to use the 10-year 
look back to calculate an emissions unit's baseline actual emissions. 
Several commenters are concerned about how the utilization rate would 
be considered in the calculation. For example, some commenters support 
the proposal to allow sources to use their highest capacity achieved 
during any consecutive 12 months, because it provides improved 
flexibility in establishing a capacity level that is representative of 
normal operations. However, other commenters object to using the 12 
months with the highest utilization. These commenters argue that the 
use of production rates can be unworkable because there is not always a 
clear relationship between production rate and emissions. In addition, 
reliable records may not be available to determine the highest 
production rates. As an alternative, commenters suggest using emissions 
from any 12-month period in the preceding 10 years, adjusted to reflect 
current rules, or allowing the source to use any 12-month period of its 
choice.
    A related issue raised by commenters is whether to require any 
current Federal, State, or voluntary limit to be included in the 
establishment of the baseline actual emissions. Some commenters say 
these provisions would penalize sources that complied with other 
regulatory requirements or chose to implement pollution prevention 
programs. Commenters are particularly concerned that sources be given 
credit for voluntary reductions. However, other commenters support 
including all of these factors in the baseline to better represent 
actual emissions and avoid inconsistencies between emissions units that 
have permits and those that do not. Commenters also raise specific 
questions about how the calculation would include the effect of other 
emission limitations.
    As described earlier, we have decided to require the use of a 
consecutive 24-month period within the 10-year look back instead of the 
proposed 12-month period to calculate the baseline actual emissions for 
any emissions unit that undergoes a physical or operational change, or 
is affected by such change. The longer 24-month period allows you to 
reference levels of utilization achieved in the past, but also 
eliminates the potential problem associated with short-term peaks that 
do not truly represent the unit's normal operation. In this respect, 
the use of a 24-month period is consistent with the pre-existing 
approach for calculating actual emissions.
    With respect to commenters' concerns about being required to use 
the period of highest utilization, our reference in the proposal 
preamble to selecting the period of highest utilization was based on 
our general assumption that the period of maximum utilization also 
represents the period of highest pollution levels for the unit of 
concern. However, you are not required to select the period of highest 
utilization. The choice of which consecutive 24-month period within the 
10-year window to use is up to you. The two restrictions on the 
selection of the appropriate consecutive 24-month period, as described 
earlier, are the availability of adequate and complete source records 
for the unit of concern and the limit on using dates earlier than 
November 15, 1990 for contemporaneous emissions changes in 
nonattainment areas and OTRs.
    We agree with the concerns expressed by some commenters that the 
baseline actual emissions calculated from the consecutive 24-month 
period selected could yield a higher pollution level than a unit is 
currently allowed to emit. We do not believe that we should allow a 
source to take credit for baseline actual emissions that exceed the 
current, legally allowable emissions rate. Consequently, the new 
requirements require you to determine whether any legally enforceable 
limitations currently exist that would prevent the affected unit from 
emitting a pollutant at the levels calculated from the 24-month 
baseline period. The approach that we have adopted allows you to 
reference plant capacity that has actually been used, but not pollution 
levels that are not legally allowed at the time the modification is to 
occur. You will be required to make adjustments for voluntary 
reductions that you may have taken only to the extent that the 
reductions resulted from conditions that are legally enforceable 
limitations.
5. How Does the Change in the Baseline Period Affect Related 
Requirements Regarding Protection of Air Quality?


a. How Does the Extended Baseline Period Conform With the Special 
Modification Provisions Under Sections 182(c) and (e) of the Act?


    Most commenters feel the proposed extension of the look back period 
fits within the design and intent of the special modification 
procedures set forth in sections 182(c) and (e) of the Act, applicable 
in serious, severe, and extreme ozone nonattainment areas. However, one 
commenter representing State and local air pollution control agencies 
considers the new requirements to be in significant conflict with the 
special modification procedures contained in those sections of the Act. 
The commenter indicates that this conflict could be resolved by 
deferring to relevant requirements for modifications in serious, 
severe, and extreme areas. The commenter adds that while NSR programs 
are tools to attain and maintain compliance with the NAAQS, they should 
not be available to undermine specific statutory and SIP requirements 
designed to resolve nonattainment problems.
    We disagree with the commenter's concern that the use of a 10-year 
look back period to implement sections 182(c) and (e) of the Act for 
purposes of establishing a modified unit's baseline emissions will 
undermine any statutory or SIP requirements designed to address 
nonattainment problems. The two sections establish special procedures 
for determining whether a proposed modification of a major stationary 
source of ozone in a serious, severe, or extreme ozone nonattainment 
area will be subject to major NSR under part D of the Act. The Act is 
silent on the issue of how one is to determine whether a physical or 
operational change increases the amount of a pollutant for a changed 
emissions unit. We believe, therefore, that we have the authority to 
establish a regulatory procedure for making the required determinations 
concerning emissions increases resulting from physical or operational 
changes.
    In light of the fact that the 10-year look back period may be used 
for emissions units (other than EUSGUs) that are involved in 
contemporaneous emissions changes (for netting purposes), it should be 
noted that the new requirements prohibit the use of the look back 
period earlier than November 15, 1990. Consequently, for emissions 
units whose contemporaneous emissions changes occurred before November 
15, 2000, the consecutive 24-month period selected


[[Page 80202]]


for calculating the baseline actual emissions relevant to the 
contemporaneous emissions change cannot include a date prior to 
November 15, 1990. It should be pointed out, however, that for 
modifications involving emissions of volatile organic compounds (VOC) 
in areas classified as ``extreme,'' the statutory language is clear 
that the increase in emissions resulting from the change is not 
required to be a significant increase, but rather that ``any increase'' 
that is projected using the new actual-to-projected-actual 
applicability test will trigger the applicable NSR requirements.


b. Will the Longer Look Back Period Related to the Baseline Actual 
Emissions Protect Short-term Increments and NAAQS?


    Some commenters express concerns that the opportunity to take 
credit for older baseline actual emissions would result in adverse 
environmental consequences. One commenter specifically indicates that 
the proposed baseline actual emissions determination process, involving 
a 10-year look back, would allow significant increases in emissions to 
escape the ambient impact review requirements otherwise required by 
NSR.
    Today's new rule modifies the way your NSR applicability 
determinations are made for changes made to existing emissions units. 
The new rule does not affect the way in which a source's ambient air 
quality impacts are evaluated. Compliance with the NAAQS is 
accomplished with air quality dispersion models using maximum allowable 
emission limitations (or federally enforceable permit limits) combined 
with operating factors, which consider either design capacity or actual 
operating factors averaged over the most recent 2 years of operation, 
from all modeled sources.\24\ In addition, any increase in actual 
emissions, based on the existing definition of ``actual emissions,'' 
consumes PSD increment whether it occurs through normal source 
operation or as a result of a physical or operational change. As 
mentioned earlier, the existing definition of ``actual emissions'' 
continues to apply with regard to all NSR requirements other than the 
new source applicability tests. See, for example, new Sec.  
52.21(b)(21)(i). Thus, we do not believe there is a basis for 
concluding that the use of a longer look back period for determining a 
modified emissions unit's baseline actual emissions (for purposes of 
determining whether a physical or operational change will result in a 
significant emissions increase) will cause any adverse environmental 
impacts.
---------------------------------------------------------------------------


    \24\ Guidance for modeling NAAQS compliance under the PSD 
program is set forth in EPA's Guideline on Air Quality Models 
contained in appendix W of 40 CFR part 51. This guidance is 
incorporated by reference both in the Federal PSD regulations and in 
the minimum requirements for SIPs under the part 51 PSD regulations.
---------------------------------------------------------------------------


6. Why Was the Contemporaneous Period for Netting Not Also Changed to a 
10-Year Look Back Period?
    In the 1996 NPRM, we indicated th