[Federal Register Volume 72, Number 108 (Wednesday, June 6, 2007)]
[Proposed Rules]
[Pages 31372-31399]
From the Federal Register Online via the Government Publishing Office [www.gpo.gov]
[FR Doc No: E7-10459]



[[Page 31371]]

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

Part II





Environmental Protection Agency





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



40 CFR Parts 51 and 52



 Prevention of Significant Deterioration New Source Review: Refinement 
of Increment Modeling Procedures; Proposed Rule

  Federal Register / Vol. 72, No. 108 / Wednesday, June 6, 2007 / 
Proposed Rules  

[[Page 31372]]


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

ENVIRONMENTAL PROTECTION AGENCY

40 CFR Parts 51 and 52

[EPA-HQ-OAR-2006-0888; FRL-8320-7]
RIN 2060-AO02


Prevention of Significant Deterioration New Source Review: 
Refinement of Increment Modeling Procedures

AGENCY: Environmental Protection Agency (EPA).

ACTION: Proposed rule.

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

SUMMARY: Under the requirements of the Clean Air Act (Act), the New 
Source Review (NSR) program includes Prevention of Significant 
Deterioration (PSD) measures, which protect air quality in areas that 
currently have clean air. For some pollutants, the PSD program protects 
clean air through a system of ``increments.'' These increments specify 
the maximum extent to which the ambient concentration of these 
pollutants may be allowed to increase above the legally defined 
baseline concentration in an area with clean air. In this rulemaking, 
we propose to refine several aspects of the method that may be used to 
calculate an increase in concentration for increment purposes. These 
refinements are intended to clarify how States and regulated sources 
may calculate increases in concentrations for the purposes of 
determining compliance with the PSD increments.

DATES: Comments. Written comments must be received on or before August 
6, 2007.
    Public Hearing. If anyone contacts EPA requesting to speak at a 
public hearing by June 26, 2007, we will hold a public hearing 
approximately 30 days after publication in the Federal Register. 
Additional information about the hearing would be published in a 
subsequent Federal Register notice.

ADDRESSES: Comments. Submit your comments, identified by Docket ID No. 
EPA-HQ-OAR-2006-0888, by one of the following methods:
     http://www.regulations.gov: Follow the online instructions 
for submitting comments.
     E-mail: [email protected].
     Fax: (202) 566-1741.
     Mail: Environmental Protection Agency, EPA Docket Center 
(EPA/DC), Air and Radiation Docket, Mail Code 6102T, 1200 Pennsylvania 
Avenue, NW., Washington, DC 20460. Please include 2 copies.
     Hand Delivery: EPA Docket Center, (Air Docket), EPA/DC, 
EPA West, Room 3334, 1301 Constitution Ave., NW., Washington, DC. Such 
deliveries are only accepted during the Docket's normal hours of 
operation, and special arrangements should be made for deliveries of 
boxed information.
    Instructions. Direct your comments to Docket ID No. EPA-HQ-OAR-
2006-0888. EPA's policy is that all comments received will be included 
in the public docket without change and may be made available online at 
http://www.regulations.gov, including any personal information 
provided, unless the comment includes information claimed to be 
Confidential Business Information (CBI) or other information whose 
disclosure is restricted by statute. Do not submit information that you 
consider to be CBI or otherwise protected through http://www.regulations.gov or e-mail. The http://www.regulations.gov Web site 
is an ``anonymous access'' system, which means EPA will not know your 
identity or contact information unless you provide it in the body of 
your comment. If you send an e-mail comment directly to EPA without 
going through http://www.regulations.gov, your e-mail address will be 
automatically captured and included as part of the comment that is 
placed in the public docket and made available on the Internet. If you 
submit an electronic comment, EPA recommends that you include your name 
and other contact information in the body of your comment and with any 
disk or CD-ROM you submit. If EPA cannot read your comment due to 
technical difficulties and cannot contact you for clarification, EPA 
may not be able to consider your comment. Electronic files should avoid 
the use of special characters, any form of encryption, and be free of 
any defects or viruses. For additional instructions on submitting 
comments, go to section I.B of the SUPPLEMENTARY INFORMATION section of 
this document.
    Docket. All documents in the docket are listed in the http://www.regulations.gov index. Although listed in the index, some 
information is not publicly available, i.e., CBI or other information 
whose disclosure is restricted by statute. Certain other material, such 
as copyrighted material, is not placed on the Internet and will be 
publicly available only in hard copy form. Publicly-available docket 
materials are available either electronically in http://www.regulations.gov or in hard copy at the Air Docket, EPA/DC, EPA 
West, Room 3334, 1301 Constitution Ave., NW., Washington, DC. The 
Public Reading Room is open from 8:30 a.m. to 4:30 p.m., Monday through 
Friday, excluding legal holidays. The telephone number for the Public 
Reading Room is (202) 566-1744, and the telephone number for the Air 
Docket is (202) 566-1742.

FOR FURTHER INFORMATION CONTACT: Ms. Jessica Montanez, New Source 
Review Group, Air Quality Policy Division (C504-03), U.S. Environmental 
Protection Agency, Research Triangle Park, NC 27711, telephone number: 
(919) 541-3407; fax number: (919) 541-5509, or electronic mail e-mail 
address: [email protected].

SUPPLEMENTARY INFORMATION: The information presented in this preamble 
is organized as follows:

I. General Information
    A. Does this action apply to me?
    B. What should I consider as I prepare my comments for EPA?
    C. Where can I get a copy of this document and other related 
information?
    D. How can I find information about a possible hearing?
II. Background
    A. What is the PSD program?
    B. What are PSD increment analyses?
    C. Why do we need to clarify the method for analyzing increment 
consumption?
    D. What are the Clean Air Act requirements related to 
increments?
III. Summary of This Proposed Action
IV. Proposed Clarifications Regarding the Effect of the Draft New 
Source Review Workshop Manual
V. Proposed Clarifications and Changes to Increment Modeling 
Procedures
    A. What kind of emissions consume or expand the PSD increment?
    B. How are emissions estimated for sources that consume 
increment?
    C. What meteorological models and data should be used in 
increment consumption modeling?
    D. What are my documentation and data and software availability 
requirements?
VI. Implementation Issues
    A. Is there a need for States to make revisions to their SIPs?
    B. When would these policies be put into effect?
VII. Statutory and Executive Order Reviews
    A. Executive Order 12866--Regulatory Planning and Review
    B. Paperwork Reduction Act
    C. Regulatory Flexibility Analysis
    D. Unfunded Mandates Reform Act
    E. Executive Order 13132--Federalism
    F. Executive Order 13175--Consultation and Coordination With 
Indian Tribal Governments
    G. Executive Order 13045--Protection of Children From 
Environmental Health and Safety Risks
    H. Executive Order 13211--Actions That Significantly Affect 
Energy Supply, Distribution, or Use
    I. Executive Order 12898--Federal Actions To Address 
Environmental Justice in Minority Populations and Low-Income 
Populations
    J. National Technology Transfer and Advancement Act
VIII. Statutory Authority

[[Page 31373]]

I. General Information

A. Does this action apply to me?

    Entities potentially affected by this proposed action include 
owners and operators of emission sources in all industry groups, as 
well as the EPA and State, local, and tribal governments that are 
delegated authority to implement these regulations. The majority of 
sources potentially affected are expected to be in the following 
groups:

------------------------------------------------------------------------
          Category                  NAICS\a\           Industry group
------------------------------------------------------------------------
Industry....................  221111, 221112,       Electric services.
                               221113, 221119,
                               221121, 221122.
                              32411...............  Petroleum refining.
                              325181, 32512,        Industrial inorganic
                               325131, 325182,       chemicals.
                               211112, 325998,
                               331311, 325188.
                              32511, 325132,        Industrial organic
                               325192, 325188,       chemicals.
                               325193, 32512,
                               325199.
                              32552, 32592, 32591,  Miscellaneous
                               325182, 32551.        chemical products.
                              211112..............  Natural gas liquids.
                              48621, 22121........  Natural gas
                                                     transport.
                              32211, 322121,        Pulp and paper
                               322122, 32213.        mills.
                              322121, 322122......  Paper mills.
                              336111, 336112,       Automobile
                               336712, 336211,       manufacturing.
                               336992, 336322,
                               336312, 33633,
                               33634, 33635,
                               336399, 336212,
                               336213.
                              325411, 325412,       Pharmaceuticals.
                               325413, 325414.
Federal government..........  924110..............  Administration of
                                                     Air and Water
                                                     Resources and Solid
                                                     Waste Management
                                                     Programs.
State/local/tribal            924110..............  Administration of
 Government.                                         Air and Water
                                                     Resources and Solid
                                                     Waste Management
                                                     Programs.
------------------------------------------------------------------------
\a\ North American Industry Classification System.

    This table is not intended to be exhaustive, but rather provides a 
guide for readers regarding entities likely to be affected by this 
action. If you have any questions regarding the applicability of this 
action to a particular entity, contact the person listed in the 
preceding FOR FURTHER INFORMATION CONTACT section.

B. What should I consider as I prepare my comments for EPA?

1. Submitting Confidential Business Information (CBI)
    Do not submit Confidential Business Information to EPA through 
http://www.regulations.gov or e-mail. Clearly mark the part or all of 
the information that you claim to be CBI. For CBI information in a disk 
or CD ROM that you mail to EPA, mark the outside of the disk or CD ROM 
as CBI and then identify electronically within the disk or CD ROM the 
specific information claimed as CBI, a copy of the comment that does 
not contain the information claimed as CBI must be submitted for 
inclusion in the public docket. Information so marked will not be 
disclosed except in accordance with procedures set forth in 40 Code of 
Federal Regulations (CFR) part 2.
2. Suggestions for Preparing Your Comments
    When submitting comments, remember to:
     Identify the rulemaking by docket number and other 
identifying information (subject heading, Federal Register date and 
page number).
     Follow directions. The agency may ask you to respond to 
specific questions or organize comments by referencing a CFR part or 
section number.
     Explain why you agree or disagree; suggest alternatives 
and substitute language for your requested changes.
     Describe any assumptions and provide any technical 
information and/or data that you used.
     If you estimate potential costs or burdens, explain how 
you arrived at your estimate in sufficient detail to allow for it to be 
reproduced.
     Provide specific examples to illustrate your concerns, and 
suggest alternatives.
     Explain your views as clearly as possible, avoiding the 
use of profanity or personal threats.
     Make sure to submit your comments by the comment period 
deadline identified.

C. Where can I get a copy of this document and other related 
information?

    In addition to being available in the docket, an electronic copy of 
this proposal will also be available on the World Wide Web. Following 
signature by the EPA Administrator, a copy of this notice will be 
posted in the regulations and standards section of our NSR (New Source 
Review) home page located at http://www.epa.gov/nsr.

D. How can I find information about a possible hearing?

    Persons interested in presenting oral testimony should contact Ms. 
Pam Long, New Source Review Group, Air Quality Policy Division (C504-
03), U.S. EPA, Research Triangle Park, NC 27711, telephone number (919) 
541-0641 or e-mail [email protected] at least 2 days in advance of the 
public hearing. Persons interested in attending the public hearing 
should also contact Ms. Long to verify the time, date, and location of 
the hearing. The public hearing will provide interested parties the 
opportunity to present data, views, or arguments concerning this 
action.

II. Background

A. What is the PSD program?

    Part C of title I of the Act contains the requirements for a 
component of the major NSR program known as the PSD program. This 
program sets forth procedures for the preconstruction review and 
permitting of new and modified major stationary sources of air 
pollution locating in areas meeting the National Ambient Air Quality 
Standards or ``NAAQS'' (``attainment'' areas) and areas for which there 
is insufficient information to classify an area as either attainment or 
nonattainment (``unclassifiable'' areas).
    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

[[Page 31374]]

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 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 
(including visibility) in certain national parks, wilderness areas, 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. Most States have SIP-approved major NSR programs; however 
there are some States that instead implement the Federal PSD program at 
40 CFR 52.21 through delegation.\1\
---------------------------------------------------------------------------

    \1\ Where a State does not have a SIP-approved program and 
chooses not to accept delegation of the Federal PSD program, EPA 
implements the PSD requirements as the reviewing authority within 
that jurisdiction. In addition, we implement the PSD program in 
Indian country until such time as a Tribe elects to adopt, and we 
approve, a Tribal Implementation Plan (TIP) that contains a PSD 
program that meets the requirements of the Act.
---------------------------------------------------------------------------

    The applicability of the PSD program to a particular source must be 
determined in advance of construction and is pollutant specific. Once a 
source is determined to be subject to PSD, among other requirements, it 
must undertake a series of analyses to demonstrate that it will use the 
best available control technology (BACT) and will not cause or 
contribute to a violation of any NAAQS or any maximum allowable ambient 
pollutant concentration increase (increment). In cases where the 
source's emissions may adversely affect an area classified as Class I, 
additional review is conducted to protect the increments and special 
attributes of such an area defined as ``air quality related values'' 
(AQRVs).
    When the reviewing authority reaches a preliminary decision to 
authorize construction of a proposed new major source or major 
modification, it must provide notice of the preliminary decision and an 
opportunity for comment by the general public, industry, and other 
persons that may be affected by the major source or major modification. 
After considering and responding to the comments, the reviewing 
authority may issue a final determination on the construction permit in 
accordance with the PSD regulations.

B. What are PSD increment analyses?

1. Framework for Increment Analyses
    Under section 165(a)(3) of the Act, a PSD permit applicant must 
demonstrate that emissions from the proposed construction and operation 
of a facility ``will not cause, or contribute to, air pollution in 
excess of any * * * maximum allowable increase or maximum allowable 
concentration for any pollutant* * *.'' The ``maximum allowable 
increase'' of an air pollutant that is allowed to occur above the 
applicable baseline concentration for that pollutant is known as the 
PSD increment. The maximum allowable concentration is the ceiling 
established by adding the PSD increment to the baseline concentration. 
By establishing the maximum allowable increase in a particular area, an 
increment defines ``significant deterioration.''
    Increments have been established for three pollutants--Sulfur 
Dioxide (SO2), Particulate Matter (PM), and Nitrogen Dioxide 
(NO2)--and for a variety of averaging periods, which 
correspond to the averaging periods for the NAAQS for those pollutants. 
In addition, all attainment and unclassifiable areas are classified as 
Class I, Class II, or Class III, and different increment levels apply 
in each type of area. Class I areas include certain national parks, 
wilderness areas, and other natural areas of special concern; the 
smallest increments are specified for these areas. Nearly all other 
areas in the United States are currently classified as Class II, where 
higher increments are specified. States and Tribes have the authority 
to redesignate Class II areas to Class III (with still higher 
increments) to promote development, but, to date, none have chosen to 
do so. States and Tribes also may redesignate Class II areas to Class I 
to provide additional protection; some Tribes have done so. The 
increments are codified at 40 CFR 51.166(c) and 52.21(c). The current 
increment values are shown below in Table 1.

                    Table 1. Current Increment Values
------------------------------------------------------------------------
                                                              Maximum
                                                             allowable
                                                             increase
                        Pollutant                           (micrograms
                                                             per cubic
                                                              meter)
------------------------------------------------------------------------
                                 Class I
------------------------------------------------------------------------
Particulate matter:
    PM-10, annual arithmetic mean.......................               4
    PM-10, 24-hr. maximum...............................               8
Sulfur dioxide:
    Annual arithmetic mean..............................               2
    24-hr. maximum......................................               5
    3-hr. maximum.......................................              25
Nitrogen dioxide:
    Annual arithmetic mean..............................             2.5
------------------------------------------------------------------------
                                Class II
------------------------------------------------------------------------
Particulate matter:
    PM-10, annual arithmetic mean.......................              17
    PM-10, 24-hr. maximum...............................              30

[[Page 31375]]

 
Sulfur dioxide:
    Annual arithmetic mean..............................              20
    24-hr. maximum......................................              91
    3-hr. maximum.......................................             512
Nitrogen dioxide:
    Annual arithmetic mean..............................              25
------------------------------------------------------------------------
                                Class III
------------------------------------------------------------------------
Particulate matter:
    PM-10, annual arithmetic mean.......................              34
    PM-10, 24-hr. maximum...............................              60
Sulfur dioxide:
    Annual arithmetic mean..............................              40
    24-hr. maximum......................................             182
    3-hr. maximum.......................................             700
Nitrogen dioxide:
    Annual arithmetic mean..............................              50
------------------------------------------------------------------------

    For PSD baseline purposes, a baseline area for a particular 
pollutant emitted from a source includes the attainment or 
unclassifiable area in which the source is located as well as any other 
attainment or unclassifiable area in which the source's emissions of 
that pollutant are projected (by air quality modeling) to result in an 
ambient concentration increase of at least 1 [mu]g/m3 
(annual average). See, e.g., 40 CFR 52.21(b)(15)(i). Once the baseline 
area is established, subsequent major sources undergoing PSD review in 
that area must address the fact that a portion of the available 
increment may already have been consumed by previous emissions 
increases.
    Three dates related to the PSD baseline concept are important in 
calculating the amount of increment consumed by pollutant emissions 
from the major source undergoing PSD review and other applicable 
emissions increases and decreases in a particular baseline area. In 
general, the submittal date of the first complete PSD permit 
application in a particular area is the operative ``baseline date.'' 
\2\ On or before the date of the first complete PSD application, most 
emissions are considered to be part of the baseline concentration. Most 
emissions increases that occur after the baseline date will be counted 
toward the amount of increment consumed. Similarly, emissions decreases 
after the baseline date expand the amount of increment that is 
available.
---------------------------------------------------------------------------

    \2\ Baseline dates are pollutant specific. That is, a complete 
PSD application establishes the baseline date only for those 
regulated NSR pollutants that are projected to be emitted in 
significant amounts (as defined in the regulations) by the 
applicant's new source or modification. Thus, an area may have 
different baseline dates for different pollutants.
---------------------------------------------------------------------------

    In actuality, there are two baseline dates that are related to the 
determination of how much increment is being consumed in a particular 
baseline area. These two dates, described below, are necessary to 
properly account for the emissions that are to be counted toward 
increment consumed in accordance with the statutory definition of 
``baseline concentration'' in section 169(4) of the Act. The statutory 
definition provides that the baseline concentration of a pollutant for 
a particular baseline area is generally the air quality at the time of 
the first application for a PSD permit in the area. Consequently, any 
increases in actual emissions occurring after that date (with some 
possible exceptions that we will discuss later) would be considered to 
consume the applicable PSD increment. However, the statutory definition 
also provides that ``[E]missions * * * from any major emitting facility 
on which construction commenced after January 6, 1975 shall not be 
included in the baseline and shall be counted in pollutant 
concentrations established under this part.''
    To make this distinction between the date when emissions changes in 
general (i.e., from both major and minor sources) count in the 
increment and the date when emissions resulting from the construction 
at a major stationary source count in the increment, we established the 
terms ``minor source baseline date'' and ``major source baseline 
date,'' respectively. See 40 CFR 51.166(b)(14) and 52.21(b)(14). 
Accordingly, the ``minor source baseline date'' is the date on which 
the first complete application for a PSD permit is filed in a 
particular area. Any change in actual emissions after that date counts 
in the PSD increment for that area. The ``major source baseline date'' 
is thus named because it is the date after which actual emissions 
associated with construction at a major stationary source affect the 
available PSD increment. In accordance with the statutory definition of 
``baseline concentration,'' the PSD regulations define a fixed date to 
represent the major source baseline date for each pollutant for which 
an increment exists. Congress defined the major source baseline date 
for the statutory increments for PM and SO2 as January 6, 
1975. For the NO2 increments, which we promulgated in 1988 
under our authority to establish an increment system under section 
166(a) of the Act, the major source baseline date was selected as 
February 8, 1988--the date on which we proposed increments for 
NO2.
    Finally, the PSD regulations set out the third date that is 
relevant to the PSD baseline concept. These regulations provide that 
the earliest date on which the minor source baseline date can be 
established is the date immediately following the ``trigger date'' for 
the pollutant-specific increment. See, e.g., 40 CFR 52.21(b)(14)(ii). 
For PM and SO2, Congress defined the applicable trigger date 
as August 7, 1977--the date of the 1977 amendments to the Act

[[Page 31376]]

when the original statutory increments were established by Congress. 
For NO2, we selected the trigger date as February 8, 1988--
the date on which we proposed increments for NO2. See 53 FR 
40656, 40658; October 17, 1988.
    Under this approach, the baseline concentration is not actually 
established for a PSD baseline area until after the ``minor source 
baseline date'' is established by the submission of the first PSD 
permit application for a source whose emissions would affect a given 
baseline area. Although major source emissions may consume increment 
prior to this date, they are not factored into the calculation until 
the minor source baseline date is triggered.
    Once the minor source baseline date associated with the first 
proposed new major stationary source or major modification in an area 
is established, the new emissions from that source consume a portion of 
the increment in that area, as do any subsequent emissions increases 
that occur from any source in the area. When the maximum pollutant 
concentration increase defined by the increment has been reached, 
additional PSD permits cannot be issued until sufficient amounts of the 
increment are ``freed up'' via emissions reductions that may occur 
voluntarily, e.g., via source shutdowns, or via control requirements 
imposed by the reviewing authority. Moreover, the air quality in a 
region cannot deteriorate to a level in excess of the applicable NAAQS, 
even if all the increment has not been consumed. Therefore, new or 
modified sources located in areas where the air pollutant concentration 
is near the level allowed by the NAAQS may not have full use of the 
amount of pollutant concentration increase allowed by the increment.
2. General Approach to Increment Analyses
    The EPA and the States have generally used an emissions inventory 
and modeling approach to identify the degree to which an increment has 
been consumed or will be consumed by major source construction. Ambient 
monitoring has not been used to establish baseline concentrations or to 
evaluate increment consumption because ambient measurements reflect 
emissions from all sources, including those that should be excluded 
from the measurements. We have not necessarily required the 
identification of a specific baseline concentration but rather have 
focused on measuring the change in concentration from the legally 
established baseline date to the time of the analysis. For example, in 
the preamble to the 1978 PSD regulation (43 FR 26388, 26400; June 19, 
1978), we stated the following:

    The regulations promulgated today no longer suggest that the 
baseline concentration be formally established. The Administrator 
feels that increment consumption can be best tracked by tallying 
changes in emissions levels of sources contributing to the baseline 
concentration and increases in emissions due to new sources. Data to 
establish baseline air quality in an absolute sense would be needed 
only if increment consumption were to be tracked using ambient 
measurements. Thus, to implement the air quality increment approach, 
the reviewing authority needs to verify that all changes from 
baseline emissions rates (decreases or increases as appropriate) in 
conjunction with the increased emissions associated with approved 
new source construction will not violate an applicable increment * * 
*.

This method has made it easier to comply with the statutory provisions 
(discussed below in section II.D of this preamble) excluding certain 
increases in emissions at major sources from the baseline concentration 
and allowing other emissions to be excluded from increment consumption.
    Even with that said, we believe that it would also be acceptable 
and consistent with the Act for a State to use an approach of 
establishing an actual baseline concentration using an initial baseline 
emissions inventory. The State could then calculate the consumed 
increment by revising the inventory to include the relevant emissions 
increases and decreases as discussed above.
3. Agency Guidance and Specific Approaches Used in Practice
    Over time, the Agency developed some recommended approaches that 
reviewing authorities could use to determine whether changes in 
emissions rates and increases in emission associated with new 
construction since the baseline date have or have not increased 
concentrations above the increments. Our recommendations have generally 
been described in modeling guidelines and guidance documents, while the 
PSD regulations in 40 CFR 51.166 and 52.21 contained only a few basic 
requirements for the increment analysis.
    Some of our recommendations for the increment analysis have been 
included in the ``Guideline on Air Quality Models,'' which is located 
in appendix W to 40 CFR part 51. Appendix W provides modeling 
guidelines for sources and reviewing authorities under a variety of 
Clean Air Act programs. The PSD regulations cite appendix W and state 
that all PSD air quality modeling should be based on the ``applicable 
models, data bases, and other requirements'' specified there. See 40 
CFR 51.166(l) and 52.21(l). Although appendix W is incorporated by 
reference in the PSD regulations, we have continued to refer to this as 
a ``guideline'' and used language in the guideline to indicate that it 
does not mandate specific procedures in all cases. See, In re: Prairie 
State Generating Company, PSD Permit Appeal No. 05-05, slip. op. at 132 
(EAB August 24, 2006) (``Appendix W is replete with references to 
`recommendations,' `guidelines,' and reviewing authority discretion.'') 
It is also important to keep in mind that appendix W provides 
guidelines for other types of regulatory applications, not just PSD 
increment analyses. As a result, not all the recommendations included 
in appendix W are applicable to an analysis of increment consumption 
under the PSD program. Care must be taken to evaluate whether certain 
recommendations are appropriate for the particular circumstances of 
each increment analysis.
    We also included some suggestions for the increment analysis in the 
1990 draft ``New Source Review Workshop Manual'' (draft NSR Manual).\3\ 
This draft document addressed many aspects of PSD permitting, including 
the increment analyses. However, we made clear on the very first page 
that this manual was not intended to establish binding regulatory 
requirements. Draft NSR Manual at 1 (Preface). In addition, we never 
finalized the 1990 draft of the NSR Manual and accordingly never 
intended for the manual itself to establish final EPA policies or 
interpretations of our NSR regulations. Nevertheless, many people have 
looked to this document for guidance and have sometimes improperly 
construed the draft NSR Manual to contain requirements that must be 
followed.
---------------------------------------------------------------------------

    \3\ This document is often referred to as the ``Puzzle Book'' 
due to the depiction of jigsaw puzzle pieces on its cover.
---------------------------------------------------------------------------

    The EPA's Environmental Appeals Board (``Board'') has sometimes 
referenced the draft NSR Manual as a reflection of our thinking on 
certain PSD issues, but the Board has been clear that the draft NSR 
Manual is not a binding Agency regulation. See, In re: Indeck-Elwood, 
LLC, PSD Permit Appeal No. 03-04, slip. op. at 10 n. 13 (EAB Sept. 27, 
2006); In re: Prairie State Generating Company, PSD Permit Appeal No. 
05-05, slip. op. at 7 n. 7 (EAB Aug 24, 2006). In these and other 
cases, the Board also considered briefs filed on behalf of the Office 
of Air and Radiation that provided more current information on the 
thinking of the EPA headquarters program office on specific PSD issues

[[Page 31377]]

arising in particular cases. Thus, the Board has looked to the draft 
NSR Manual as one resource to consider in developing Agency positions 
through case-by-case adjudications, while recognizing that the draft 
NSR Manual does not itself contain binding requirements.
    Other non-binding EPA guidance letters or memoranda that have 
addressed increment consumption analyses are discussed in more detail 
below in the context of discussion on specific issues.
    Based largely on prior EPA guidance, the approach that has 
generally been used in States and EPA Regional Offices for increment 
analyses has involved the following four steps:
    1. Determine the 1 [mu]g/m3 ``significant impact area'' 
for the particular pollutant for which the new major source or major 
modification is undergoing PSD review. (If the source is subject to an 
increment analysis for more than one pollutant, each analysis is 
carried out independently).
    2. Identify the other sources in the vicinity of the new or 
modified source whose emissions affect the significant impact area.
    3. Estimate the emissions from those sources that consume 
increment.
    4. Model the change in emissions to get a concentration change, and 
compare that concentration change to the applicable increment.

    The actual increment analysis that a proposed new or modified 
source undergoing PSD review must complete will depend on the area 
impacted by the source's new emissions.
    We have provided approved air quality models and guidelines for 
sources to use to project the air quality impact of each pollutant 
(over each averaging period) for which an increment analysis must be 
done. In addition, we established significant impact levels for each 
pollutant under the nonattainment major NSR program that have also been 
used under the PSD program to identify levels below which the source's 
modeled impact is regarded as de minimis. See 40 CFR 51.165(b) and part 
51, appendix S, section III.A.\4\ In the event that a source's modeled 
impacts of a particular pollutant are below the applicable significant 
impact level at all ambient air locations modeled, i.e., de minimis 
everywhere, EPA policy provides that no further modeling analysis is 
required for that pollutant. Our policy has been that when a 
preliminary screening analysis based on the significant impact level is 
sufficient to demonstrate that the source's emissions will not cause or 
contribute to a violation of the increment, there is no need for a full 
impacts analysis involving a cumulative evaluation of the emissions 
from the proposed source and other sources affecting the area.
---------------------------------------------------------------------------

    \4\ The cited regulations actually apply to sources located in a 
PSD area, which must demonstrate that they will not cause or 
contribute to a violation of the NAAQS in an adjacent nonattainment 
area. This demonstration may be made by showing that the emissions 
from the PSD source alone are below the significant impact levels 
set forth in 40 CFR 51.165(b)(2). Based on EPA interpretations and 
guidance, these significant impact levels have also been widely used 
in the PSD program to define the extent of the impact area where an 
increment analysis must be performed. We proposed to codify these 
significant impact levels for use in the PSD program in 1996 as part 
of a comprehensive proposal to revise the major NSR regulations. See 
61 FR 38250, 38325, July 23, 1996. We have not yet taken final 
action on this proposal.
---------------------------------------------------------------------------

    Within the impact area of a source that does have a significant 
impact, increment consumption is calculated using the source's proposed 
emissions increase, along with other emissions increases or decreases 
of the particular pollutant from other sources that would consume 
increment and which have occurred since the minor source baseline date 
established for that area. (For major sources, emissions increases or 
decreases resulting from construction as defined at 40 CFR 51.166(b)(8) 
and 40 CFR 52.21(b)(8) that have occurred since the major source 
baseline date consume or expand increment). Thus, an emissions 
inventory of sources whose emissions consume or expand the available 
increment in the area must be compiled. The inventory includes not only 
sources located directly in the impact area, but sources outside the 
impact area that affect the air quality within the impact area. Section 
IV.A.1 of this preamble discusses the types of sources that are to be 
included in the emissions inventory for increment analyses.
    The inventory of emissions includes emissions from increment-
affecting sources at two separate time periods--the baseline date and 
the current period of time. For each source that was in existence on 
the relevant baseline date (major source or minor source), the 
inventory includes the source's actual emissions on the baseline date 
and its current actual emissions. The change in emissions over these 
time periods represents the emissions that consume increment (or, if 
emissions have gone down, expand the available increment). For sources 
constructed since the relevant baseline date, all their current actual 
emissions consume increment and are included in the inventory.
    An emissions inventory must be prepared for each averaging period 
for which an increment has been specified for the pollutant under 
review. In many cases, direct emissions data are not available for some 
or all averaging periods, and actual emissions must be estimated. This 
can be particularly challenging for existing sources where the baseline 
emissions must be determined and the baseline date is well in the past. 
The approach generally used per EPA guidance has been to base the 
annual emissions inventory on the actual measured emissions or actual 
hours of operation, fuel usage, raw materials used, etc., while basing 
the emissions inventory for shorter averaging periods on the maximum 
emissions over each averaging period as determined from available data 
(again, emission measurements, operating hours, fuel or materials 
consumption, etc.).
    When the inventory of emissions has been compiled, computer 
modeling is used to determine the change in ambient concentration that 
will result from these emissions when combined with the proposed 
emissions increase from the new major source or major modification that 
is undergoing PSD review. The modeling has generally been guided by the 
``Guideline on Air Quality Models'' (40 CFR part 51, appendix W), which 
includes provisions on air quality models and the meteorological data 
input into these models.
    Two possible approaches have been used to predict the change in air 
pollutant concentration using models. One approach is to make a single 
model run after calculating the difference in emissions from the 
baseline date to the current period of time. An alternative approach is 
to make two model runs (one based on an inventory of baseline emissions 
and the second based on an inventory of current actual emissions) and 
calculate the difference between them.
    The model output (expressed as a change in concentration) for each 
relevant averaging period is then compared to the corresponding 
allowable PSD increment. If the model results indicate that the 
increment(s) will not be exceeded, the reviewing authority may issue a 
PSD permit to the source. Except as discussed below, if the modeling 
shows that the source would cause or contribute to a violation of a PSD 
increment,\5\ the reviewing authority

[[Page 31378]]

may not issue the permit. The source may revise its permit application 
to reduce its proposed emissions, or it may mitigate the impact of its 
emissions through obtaining offsetting emission reductions from other 
sources in the emissions inventory.
---------------------------------------------------------------------------

    \5\ The proposed source is deemed to ``cause or contribute to'' 
an increment violation if the modeling shows that the impact 
attributable to the source at the time and place of the violation is 
greater than the relevant significant impact level.
---------------------------------------------------------------------------

    If the modeling shows only an increment violation in a Class I 
area, the source has the opportunity to apply for a ``variance'' from 
the Federal Land Manager (FLM) that has responsibility for that Class I 
area. If the source successfully demonstrates to the FLM that emissions 
from the source will not have an adverse effect on the AQRVs of the 
Class I area, and to the reviewing authority that the emissions will 
not violate a set of higher increment levels specified in the Act 
(generally equal to the Class II increments), the reviewing authority 
may issue a PSD permit to the source. The source may further appeal to 
the Governor and the President in certain situations. These variances 
are discussed in greater detail in section IV.A.2 of this preamble.

C. Why do we need to refine the method for analyzing increment 
consumption?

    We have never adopted detailed regulations establishing a specific 
methodology that sources and reviewing authorities must use to 
calculate an increase in concentrations for purposes of determining 
compliance with the PSD increments. Instead, increment analyses have 
been conducted by States and EPA Regional Offices based on the 
guidelines and guidance discussed in the previous section. In the 
absence of definitive requirements, sources and reviewing authorities 
have attempted to apply the available guidance to a wide range of 
situations. Differing interpretations and approaches have resulted, 
along with controversy over how binding the guidelines and guidance are 
on reviewing authorities and who (EPA or the reviewing authorities) has 
the ultimate discretion to determine which approaches are reasonable 
for a specific increment analysis. With this proposal, we intend to 
provide greater clarity on several issues.
    One push for greater clarity has come from the Western States Air 
Resources Council (WESTAR) PSD Reform Workgroup, with participants from 
Western States, the U.S. National Park Service, U.S. Fish and Wildlife 
Service, U.S. Forest Service, and the U.S. Bureau of Land Management 
and consultation by EPA. The workgroup was formed in early 2004 to 
develop recommendations to improve the effectiveness of the PSD 
program. The goal of the WESTAR effort was to propose changes to the 
PSD program that would result in a more practical program, 
significantly reducing constraints in the current program that they 
viewed as limiting State and local agencies' abilities to address 
cumulative incremental consumption and Class I AQRV analysis and 
protection, some of which were identified in a letter to EPA.\6\ While 
the purpose of today's notice is focused on refining increment analysis 
procedures, we are considering broader changes to the program as a 
separate rulemaking to address additional concerns that WESTAR and 
others have raised.\7\
---------------------------------------------------------------------------

    \6\ ``Recommendations for Improving the Prevention of 
Significant Deterioration Program.'' Stuart A. Clark, President, 
Western States Air Resources Council, May 19, 2005.
    \7\ In addition to WESTAR's recommendations, we received 
comments from the Northeast States for Coordinated Air Use 
Management (NESCAUM) on the WESTAR recommendations in a letter and 
attachment from Arthur N. Marin, Executive Director of NESCAUM, 
October 18, 2005.
---------------------------------------------------------------------------

    A major point raised by WESTAR is that States need to consult early 
and often in order to agree in advance on modeling protocols to enable 
consistency between the States in performing the analyses and to ensure 
equity in application of the analysis. WESTAR further recommended that 
we take steps to ensure that EPA Regional Offices, in partnership with 
States and FLMs, operate consistently among themselves in inter-
jurisdictional contexts and develop data and methods that will better 
enable inter-jurisdictional analysis. WESTAR stressed that a balance is 
needed between providing States with case-by-case, cross-jurisdictional 
PSD increment analysis flexibility and providing the national or 
regional standardization necessary to ensure equity among States, 
simplify cross-jurisdictional analysis, and facilitate coordination 
with FLMs. The WESTAR report also noted a lack of clarity and sometimes 
narrow interpretations of the definition of actual emissions used for 
purposes of calculating point source emissions for inclusion in 
emissions inventories for PSD analyses. All of the WESTAR workgroup 
representatives agreed that it is desirable to bring greater clarity 
and consistency to approaches for conducting refined analyses, 
particularly related to approaches for calculating point source 
emissions. Today's notice is a step toward achieving that balance 
between case-by-case flexibility and inter-jurisdictional consistency.

D. What are the Clean Air Act requirements related to increments?

    The PSD increments are established under sections 163 and 166 of 
the Act. In section 163 of the Act, Congress adopted specific numerical 
increments for particulate matter and sulfur dioxide in each of the 
three classes of PSD baseline areas (i.e., Class I, II, and III, as 
described above in section II.B.1). In 1990, Congress created section 
166(f) of the Act which authorized us to substitute increments based on 
the PM10 indicator for the original particulate matter 
increments contained in section 163. Consistent with this provision, we 
substituted PM10 increments for the increments based on 
total suspended particulate matter in a 1993 rulemaking (58 FR 51622, 
June 3, 1993). In section 166(a) of the Act, Congress directed and 
authorized EPA to promulgate additional increments for nitrogen oxides 
and other pollutants. We promulgated increments for NO2 in 
1988 and reaffirmed those increments in a 2005 rulemaking (53 FR 40656, 
Oct. 17, 1988; 70 FR 59582, Oct. 12, 2005).
    The Act does not directly specify how to determine an increase in 
concentrations for purposes of determining compliance with the PSD 
increments. Section 163(b) of the Act provides that ``the maximum 
allowable increase in concentrations of sulfur dioxide and particulate 
matter over baseline concentration of such pollutants shall not 
exceed'' specified amounts for each pollutant. See CAA sections 
163(b)(1)-(3). The Act does not define an ``increase in 
concentrations'' for purposes of section 163. Likewise, section 
165(a)(3) prohibits permitting a source that causes or contributes to 
``air pollution in excess of any maximum allowable increase or maximum 
allowable concentrations,'' but does not specify how EPA is to 
determine that air pollution would exceed the allowable increase or 
concentration. Section 166 of the Act directs EPA to promulgate 
pollutant-specific PSD regulations which contain ``specific numerical 
measures against which permit applications may be evaluated'' and 
indicates that such measures ``may contain air quality increments.'' 
See CAA sections 166(a), (c), (d). However, there is no further 
guidance in section 166 concerning the method to be used to measure an 
increase in air pollutant concentrations for purposes of evaluation 
against the PSD increments.
    We have found some guidance in the Act in the definition of 
``baseline concentration,'' which we interpret to support our view that 
an increase in concentration for increment purposes

[[Page 31379]]

should be determined on the basis of actual emissions. Section 169(4) 
of the Act defines ``baseline concentration'' as ``the ambient 
concentration levels which exist at the time of the permit 
application.'' The opinion of the United States Court of Appeals for 
the District of Columbia Circuit in Alabama Power v. Costle interpreted 
section 169(4) in a manner that supports establishing the PSD baseline 
concentration using actual emissions. 636 F.2d 323, 375-381 (D.C. Cir. 
1980). Since emissions that consume increment are not included in the 
baseline, we have long recognized that an increase in concentration 
(the consumption of increment) is directly related to baseline 
concentration (45 FR 52676, 52718, Aug. 7, 1980). In light of these 
considerations, we reached the following conclusion:

    Since the Alabama Power decision and the statute both provide 
that actual air quality be used to determine baseline 
concentrations, but provide no guidance on increment consumption 
calculations, EPA has concluded that the most reasonable approach, 
consistent with the statute, is to use actual source emissions, to 
the extent possible, to calculate increment consumption or 
expansion.

See 45 FR 52676, 52718 (Aug. 7, 1980). We expressly incorporated the 
definition of ``actual emissions'' into the regulatory definition of 
``baseline concentration'' (40 CFR 51.166(b)(13) and 52.21(b)(13)). In 
this definition of ``baseline concentration,'' the term ``actual 
emissions'' is referenced both in the provision describing how to 
determine the baseline concentration and in the provision identifying 
emissions that affect the maximum allowable increases (the increment). 
See, e.g., 40 CFR 51.166(b)(13)(ii). The term ``actual emissions'' is 
itself defined in 40 CFR 51.166(b)(21) and 52.21(b)(21).
    The Act also provides some direction concerning the increment 
consumption analysis by identifying particular sources whose emissions 
are counted against the maximum allowable increases and listing 
categories of sources whose emissions may be excluded from the 
increment consumption analysis. In the statutory definition of 
``baseline concentration,'' section 169(4) of the Act specifies that 
``[e]missions of sulfur oxides and particulate matter from any major 
emitting facility on which construction commenced after January 6, 
1975, shall not be included in the baseline and shall be counted 
against the maximum allowable increases in pollutant concentrations 
established under this part.'' This provision makes clear that 
emissions of these pollutants from new or modified major sources that 
commence construction between 1975 and the baseline date for a given 
area shall be counted against the increments and thus are considered to 
``consume'' increment. In addition, section 163(c) authorizes States to 
exclude certain pollution concentrations from the increment consumption 
analysis. This provision authorizes States to ``promulgate rules 
providing that for purposes of determining compliance with the maximum 
allowable increases in ambient concentrations of an air pollutant, the 
following concentrations of such pollutants shall not be taken into 
account.'' The concentrations identified are those attributable to (1) 
fuel switches required under other laws (15 U.S.C. 792 or 16 U.S.C. 
791a); (2) construction or other temporary emission-related activities; 
and (3) new sources outside the United States. The PSD regulations 
reflect these provisions of sections 163(c) and 169(4) of the Act.
    The existing PSD regulations reflect these specific requirements of 
the Act. As discussed earlier, we implemented the last sentence of 
section 169(4) by establishing two separate baseline dates--the major 
source baseline date and the minor source baseline date. See 40 CFR 
51.166(b)(14) and 52.21(b)(14). We implemented section 163(c) of the 
Act by promulgating 40 CFR 51.166(f), which is discussed further below.
    Within the boundaries described above, we read the Act to provide 
EPA with fairly broad discretion to establish regulations concerning 
the approach to be used to measure an increase in concentration for 
purposes of assessing consumption of PSD increments. Since the Act does 
not define ``increase in concentration'' for increment purposes, we 
interpret the Act to grant EPA discretion to develop a method for 
measuring this increase, so long as that method is reasonable and 
consistent with the limited requirements described above. The absence 
of specific direction in the Act concerning how to calculate an 
increase in concentration for increment purposes is similar to the gap 
in the Act concerning how to calculate an increase in emissions for 
purposes of identifying a major modification. With respect to the 
latter issue, the DC Circuit has recently observed that ``In enacting 
the NSR program, Congress did not specify how to calculate `increases' 
in emission, leaving EPA to fill that gap while balancing the economic 
and environmental goals of the statute.'' New York v. EPA, 413 F.3d 3, 
27 (Jan. 25, 2005). We believe Congress intended a similar result with 
respect to ``increases'' in concentration under the increment 
provisions of the PSD side of the NSR program. As observed by the court 
in Alabama Power, ``Congress expected EPA to use `administrative good 
sense' in establishing the baseline and calculating exceedances.'' See 
Alabama Power, 636 F.2d at 380. In this rulemaking, we propose to 
exercise our rulemaking discretion on this topic and provide additional 
guidance to States and regulated sources on how to calculate increases 
in concentrations for purposes of determining compliance with the PSD 
increments.

III. Summary of This Proposed Action

    This action proposes clarifications in eight areas related to 
increment analyses. They are summarized below:
     Effect of the 1990 draft ``New Source Review Workshop 
Manual.'' Discussed in detail in section IV; no regulatory revisions.
     Treatment of sources that have previously received a Class 
I area FLM variance in subsequent increment consumption modeling. 
Discussed in detail in section V.A; regulatory revisions in 40 CFR 
51.166(f)(2) and 52.21(f)(2).
     Data used to estimate emissions. Discussed in detail in 
section V.B.1; regulatory revisions in 40 CFR 51.166(f)(1) and 
52.21(f)(1).
     Time period of emissions used to model pollutant 
concentrations. Discussed in detail in section V.B.2; regulatory 
revisions in 40 CFR 51.166(f)(1) and 52.21(f)(1).
     Actual emissions rates used to model short-term increment 
compliance. Discussed in detail in section V.B.3; regulatory revisions 
in 40 CFR 51.166(f)(1) and 52.21(f)(1).
     Meteorological data and processing. Discussed in detail in 
section V.C.1; no regulatory revisions.
     Years of meteorological data. Discussed in detail in 
section V.C.2; no regulatory revisions.
     Documentation and data and software availability. 
Discussed in detail in section V.D; no regulatory revisions.

IV. Proposed Clarifications Regarding the Effect of the Draft New 
Source Review Workshop Manual

    To avoid future misunderstandings concerning the effect of the 
draft 1990 New Source Review Workshop Manual (draft NSR Manual), we 
propose in this action to make clear that the draft NSR Manual is not a 
binding regulation and does not by itself establish final EPA policy or 
authoritative interpretations of EPA regulations under the New Source 
Review Program. As discussed above, because this document was never 
finalized, we never intended for the manual to establish final agency 
policy

[[Page 31380]]

or authoritative interpretations of EPA's NSR regulations. Furthermore, 
in many areas the positions reflected in the document have become 
outdated and superseded by statutory amendments, rulemakings, 
additional guidance memoranda, and adjudications by the Administrator 
and the EPA Environmental Appeals Board.
    Notwithstanding this proposed clarification concerning the effect 
of the draft NSR Manual, we recognize that some of the views expressed 
in the draft NSR Manual may have been promulgated in EPA regulations or 
adopted by the Agency as final policy statements or interpretations in 
other actions taken before or after the release of the draft NSR Manual 
in 1990. On some topics, the draft NSR Manual compiled pre-existing EPA 
policy and interpretations, but on other matters the document expressed 
proposed policies or interpretations that were never finalized by the 
Agency. To the extent EPA subsequently or previously adopted a view 
expressed in the draft NSR Manual through other action that was clearly 
final, those positions may have achieved the status of final policies 
or interpretations, but positions that are only expressed in the draft 
NSR Manual should not be considered to be a final EPA policy or 
interpretation.
    With respect to the increment analysis that is the subject of this 
rulemaking action, we are proposing to establish regulations that 
supersede many of the recommended approaches for conducting the 
increments analysis set forth in the draft NSR Manual and other EPA 
guidance documents, as discussed in more detail below. However, we are 
not proposing in this action to supersede or change specific policies 
or interpretations not discussed in this notice that EPA may have 
adopted in final form prior to or after the development of the draft 
NSR Manual.
    With respect to the draft NSR Manual as a whole, we are only 
proposing to clarify that the 1990 draft of the NSR Manual does not by 
itself establish final policies or interpretations of the EPA. To the 
extent such policies or interpretations are reflected in other action 
or documents that were issued in a final form (such as rulemakings, 
guidance memorandum, or adjudications by the Administrator or the 
Environmental Appeals Board), EPA will continue to follow them unless 
the Agency has otherwise indicated that it no longer adheres to such 
policies or interpretations. For example, it remains EPA's policy to 
use the five-step, top-down process to satisfy the Best Available 
Control Technology (``BACT'') requirements when PSD permits are issued 
by EPA and delegated permitting authorities, and we continue to 
interpret the BACT requirement in the Clean Air Act and EPA regulations 
to be satisfied when BACT is established using this process, as it has 
been described in decisions of the Environmental Appeals Board. 
However, notwithstanding this policy and the interpretations of the 
BACT requirement reflected in EPA adjudications, EPA has not 
established the top-down BACT process as a binding requirement through 
regulation.
    We request comment on this proposal to clarify that the draft NSR 
Manual is not a binding regulation and does not independently reflect 
or establish a final statement of EPA policy or an authoritative 
interpretation of EPA regulations.

V. Proposed Refinements to Increment Modeling Procedures

A. What kind of emissions consume or expand the PSD increment?

1. What types of sources are included in increment consumption 
modeling?
    In defining ``baseline concentration,'' the PSD regulations also 
spell out the emissions sources that must be included in an increment 
analysis. Specifically, in 40 CFR 51.166(b)(13)(ii) and 
52.21(b)(13)(ii), the regulations indicate that the following emissions 
are not included in the baseline concentration, but instead affect the 
available increment:
     Actual emissions from any major stationary source on which 
construction commenced after the major source baseline date.
     Actual emissions increases and decreases at any stationary 
source occurring after the minor source baseline date.

Thus, the sources that affect available increment, and therefore must 
be included in an increment analysis are: (1) Major sources that have 
increased or decreased actual emissions after the major source baseline 
date as a result of construction of a new source, a physical or 
operational change to an existing source, or shutdown of an existing 
source; and (2) any source that has had an increase or decrease in 
actual emissions since the minor source baseline date. The latter 
includes major sources, minor sources, and area sources that have been 
constructed since the minor source baseline date (i.e., new sources) or 
have experienced a change in actual emissions since the minor source 
baseline date (i.e., existing sources that have been modified or have 
changed their capacity utilization or hours of operation).
    For many years, we have interpreted the PSD regulations to require 
increases and decreases in mobile source emissions to be included in 
the increment consumption analysis. See, e.g., 53 FR 40656, 40662 
(October 17, 1988). However, we understand that many States have not 
consistently accounted for mobile source emissions in their increment 
analyses. To make clear that mobile source emissions need to be 
included in an analysis of increment consumption, we are proposing to 
amend the reference to ``any stationary source'' in 40 CFR 
51.166(b)(13)(ii)(b) and 52.21(b)(13)(ii)(b) of our regulations to make 
explicit that actual emissions increases or decreases that consume or 
expand increment are not limited solely to stationary source emissions.
    Despite prior inconsistencies, EPA has generally not second-guessed 
state increment assessments after they are completed or PSD permits 
have been issued. Thus, to the extent a state has neglected to account 
for mobile source emissions in prior increment analysis, EPA does not 
intend for this technical amendment to require those states to revisit 
those increment assessments or previously-issued permits. These states 
should simply include mobile source emissions in their next permit 
review or periodic review of increment consumption and factor those 
results into future permitting decisions or planning strategies.
    The existing regulations also specify that ``secondary emissions'' 
are to be included in an increment analysis. See 40 CFR 51.166(k) and 
52.21(k). Secondary emissions are defined as emissions which occur as a 
result of the construction or operation of a major source or 
modification, but do not come from the major source itself. They 
include emissions from any offsite support facility which would not be 
constructed or increase emissions except as a result of the 
construction of the major source or modification that is undergoing PSD 
review. Secondary emissions must be specific, well defined, 
quantifiable, and impact the same general area as the major source or 
modification that is under review. See 40 CFR 51.166(b)(18) and 
52.21(b)(18).
    We have also codified an exemption to these general principles in 
40 CFR 51.166(f) of the PSD regulations. This provision authorizes SIPs 
to exclude from increment consumption those sources in the four 
categories listed in section 163(c) of the Act. The regulations also 
allow States to exclude concentrations attributable to temporary 
increases in emissions from sources affected by SIP revisions approved 
by

[[Page 31381]]

EPA. See 40 CFR 51.166(f)(1)(v). When we promulgated increments for 
NO2, 40 CFR 51.166(f) became applicable to the increments 
for that pollutant as well. Thus, emissions attributable to sources or 
actions listed in 40 CFR 51.166(f) may not consume increment if a State 
has promulgated regulations approved by EPA that exclude such emissions 
from the increment consumption analysis. We have not included a 
companion provision in 40 CFR 52.21 because we read section 163(c) of 
the Act to apply only to States with approved PSD programs in their 
State implementation plans.
2. How is a source with a Class I area Federal Land Manager variance 
treated in subsequent increment consumption modeling?
    We propose to add a category of sources that may be excluded from 
the increment consumption analysis in a specialized circumstance 
described in the Clean Air Act. We propose to establish that sources 
that have been permitted based in part on a variance issued by a 
Federal Land Manager (FLM) for a Class I area may be excluded from the 
increment consumption analysis for the Class I increment in the area 
for which the variance was issued.
    Background. Under section 165(d) of the Act, when a proposed source 
subject to permitting has the potential to adversely impact a Class I 
area, an additional review is required to assess whether the source 
will adversely impact Air Quality Related Values (AQRVs) in the Class I 
area. The AQRV review provisions of section 165(d) provide another 
layer of protection against significant deterioration in Class I areas 
on top of the protection provided by increments.\8\ Although any area 
may be designated to be a Class I area, such areas are generally 
national parks and wilderness areas of a certain size that are required 
to be Class I areas under the Act. See section 162(a) of the Act.
---------------------------------------------------------------------------

    \8\ ``A second test of protection is provided in specified 
Federal land areas (Class I areas), such as national parks and 
wilderness areas; these areas are also subjected to a review process 
based on the effect of pollution on the area's air quality related 
values.'' S. Rep. 95-127, at 17, 4 LH at 1401.
---------------------------------------------------------------------------

    The Act does not define AQRVs or identify specific AQRVs other than 
visibility. See section 165(d)(2)(B) of the Act. However, AQRVs are 
generally understood to encompass the purposes for which lands have 
been preserved, to the extent those purposes may be affected by air 
quality. In legislative history to the Act, AQRVs are described as 
follows:

    The term ``air quality related values'' of Federal lands 
designated as class I includes the fundamental purposes for which 
such lands have been established and preserved by the Congress and 
the responsible Federal agency. For example, under the 1916 Organic 
Act to establish the National Park Service (16 U.S.C. 1), the 
purpose of such national park lands ``is to conserve the scenery and 
the natural and historic objects and the wildlife therein and to 
provide for the enjoyment of the same in such manner and by such 
means as will leave them unimpaired for the enjoyment of future 
generations.''

See S. Rep. 95-127 at 36, reprinted at 3 LH at 1410. In 1996, we 
proposed to adopt the following definition of AQRV:

    Air quality related values means visibility or a scenic, 
cultural, physical, biological, ecological, or recreational resource 
that may be affected by a change in air quality, as defined by the 
Federal Land Manager for Federal lands, or by the applicable State 
or Indian Governing Body for nonfederal lands.

See 61 FR 38250, 38332, July 23, 1996. We have not yet taken final 
action to adopt this definition.
    The Act provides that the FLM charged with responsibility for 
managing a Class I area has an ``affirmative responsibility'' to 
protect the AQRVs in the area. See section 165(d)(2)(B) of the Act. 
Section 165(d) establishes a procedure under which the FLM may object 
to or concur in the issuance of a PSD permit based on the impact, or 
lack thereof, that new emissions may have on any affected AQRV that the 
FLM has identified. If the proposed source's emissions do not cause or 
contribute to a violation of a Class I increment (satisfying the 
requirement in section 165(a)(3) of the Act), the FLM may nevertheless 
prevent issuance of the permit by demonstrating to the satisfaction of 
the reviewing authority that the source or modification will have an 
adverse impact on AQRVs. See section 165(d)(2)(C)(ii) of the Act. 
Conversely, if the proposed source will cause or contribute to a 
violation of a Class I increment, the reviewing authority may not issue 
the permit unless the owner or operator demonstrates to the 
satisfaction of the FLM that the emissions from the proposed facility 
will have no adverse impact on the AQRVs of the Class I area. See 
section 165(d)(2)(C)(iii) of the Act. Under this procedure, the 
compliance status of the increment determines whether the FLM or the 
permit applicant has the burden of satisfactorily demonstrating whether 
or not the proposed source's emissions would have an adverse impact on 
AQRVs.\9\ The FLM has the burden of demonstrating an adverse impact 
when the Class I increment is not exceeded. However, if the proposed 
source causes or contributes to a violation of the Class I increment, 
the permit applicant must convince the FLM to certify that the proposed 
source will not have an adverse impact on AQRVs.
---------------------------------------------------------------------------

    \9\ ``The class I increment is a test for determining where the 
burden of proof lies and is an index of changes in air quality. It 
is not the final determinant for approval or disapproval of a permit 
application.'' S. Rep. 95-127 at 35.
---------------------------------------------------------------------------

    This certification by the FLM is known as a ``variance'' under 40 
CFR 51.166(p) and 52.21(p) of the PSD regulations. The process for 
issuance of a variance was originally applied only in the context of 
the statutory increments for PM and SO2 based on section 
165(d) of the Act, but we have, by rulemaking, extended the AQRV review 
procedures set forth in Sec. Sec.  51.166(p) and 52.21(p) to cover 
NO2. See 70 FR 59583, October 12, 2005; 53 FR 40656, October 
17, 1988.
    In the case of the 24-hour and 3-hour increments for 
SO2, the Act provides an additional process through which 
the permit applicant may request that the Governor of a State issue a 
variance or appeal to the President to issue the variance if the FLM 
does not concur with the Governor's conclusion. See section 
165(d)(2)(D) of the Act. If the FLM does not initially issue a variance 
under section 165(d)(2)(C), the Governor may issue a variance subject 
to the concurrence of the FLM, if the Governor finds, after public 
notice and hearing, that a facility cannot be constructed because of a 
short-term increment for SO2 and that the variance will not 
adversely affect AQRVs. See section 165(d)(2)(D)(i) of the Act; 40 CFR 
51.166(p)(5) and 52.21(p)(6). If the FLM does not concur with the 
Governor's decision to issue the variance, the dispute is submitted to 
the President for resolution. The President may grant the variance if 
he finds that a variance is in the national interest. See section 
165(d)(2)(D)(ii) of the Act; 40 CFR 51.166(p)(6) and 52.21(p)(7).
    Under both of these variance provisions, the variance cannot issue 
unless the permit contains emissions limitations sufficient to prevent 
violations of alternative increments that are established for the 
specific permitting action due to the variance. In the case of an FLM 
variance issued under section 165(d)(2)(C), the alternative increments 
are equal to the Class II increments in most instances. In the unique 
case of the 3-hour increment for SO2, the Act requires use 
of an increment of 325 [mu]g/m\3\ (a level between the Class I and 
Class II increments) for SO2 for the 3-hour averaging 
period. See section 165(d)(2)(C)(iv) of the Act; 40

[[Page 31382]]

CFR 51.166(p)(4) and 52.21(p)(5). We also applied this approach to 
NO2 by adding a cap of 25 [mu]g/m\3\ (equal to the 
NO2 Class II increment) to the regulations. See 53 FR 3704; 
see 40 CFR 51.166(p)(4) and 52.21(p)(5). Although the short-term Class 
II increments may ordinarily be violated one time per year, the Act 
suggests that when the Class II increment applies under the Class I 
variance provisions in section 165(d)(2)(C), no violations of the Class 
II increment are permissible. See section 163(a) of the Act.
    In the case of a gubernatorial or presidential variance for the 
short term SO2 increments, the Act establishes another set 
of alternative increments at a level between the Class I and Class II 
increments for the 24-hour and 3-hour averaging periods. See section 
165(d)(2)(D)(iii) of the Act. This provision includes separate 
alternative increments for permitting actions receiving a variance in 
low and high terrain areas. Id. In addition to requiring emissions 
limitations sufficient to assure these alternative increments are not 
exceeded, this portion of the Act also specifies that the permit must 
``assure that such emissions will not cause or contribute to 
concentrations which exceed the otherwise applicable maximum allowable 
increases for periods of exposure of 24 hours or less on more than 18 
days during any annual period.'' Id. We interpret the ``otherwise 
applicable maximum allowable increases'' to describe the Class I 
increments and thus understand this provision to allow 18 exceedances 
of the Class I increment per year after a variance has been issued 
under section 165(d)(2)(D).
    In contrast to section 165(d)(2)(D)(iii), the FLM variance 
provisions in section 165(d)(C)(iv) that refers primarily to the Class 
II increments does not discuss an ``otherwise applicable maximum 
allowable increase'' or identify an allowable number of days on which 
such an increment might be exceeded. This omission leaves some 
ambiguity concerning whether the Class I increment should continue to 
apply in the Class I area for which a variance has been issued by the 
FLM under section 165(d)(2)(C) based upon a certification that the 
emissions from a proposed facility will not have an adverse impact on 
AQRVs. Since Congress has not directly spoken to this issue, we propose 
to add provisions to the PSD regulations to clarify how a reviewing 
authority should account for these variances when evaluating compliance 
with the Class I increment when a source has previously been issued a 
variance.
    Proposed Action. To address this issue, we propose to add a new 
provision in 40 CFR 51.166(f) stating that the emissions of any source 
that were permitted after receiving a Class I increment variance from 
an FLM need not be included in the consumption analysis for the Class I 
increment for the area for which the variance was issued under section 
165(d)(2)(C) of the Act. However, we propose that the emissions of such 
source continue to be accounted for in the analysis of compliance with 
the alternative Class II increments that are applied in the Class I 
area after the issuance of a variance. As noted above, in the case of 
SO2, the alternative increment is not the Class II increment 
but a level between the Class I and Class II increments.
    We interpret section 165(d)(2)(C) of the Act to allow this 
additional exclusion, not contained in section 163(c) of the Act, from 
the increment consumption analysis for emissions that an FLM has 
considered and certified to not have an adverse impact on AQRVs. 
However, this is a narrow exclusion that applies only with respect to 
the Class I increment in those areas for which a variance has been 
issued. We do not read section 165(d)(2)(C) to authorize such emissions 
to be excluded from an analysis of compliance with the Class II 
increments (or the alternative 3-hour SO2 increment).
    In Class I areas, the key criterion for determining whether a 
permit may issue is the effect of a project on AQRVs. The Class I 
increment is important, but the terms of sections 165(d)(2)(C)(ii) and 
165(d)(2)(C)(iii) make clear that AQRVs actually control whether a 
permit should be issued or not. As discussed above, the increment 
determines who has the burden of demonstrating the degree of impact on 
AQRVs, but ultimately the degree of impact on AQRVs is the controlling 
standard in such areas. Exceedances of the increment are allowed so 
long as the source can demonstrate to the satisfaction of the FLM that 
a source will not have an adverse impact on AQRVs. An exceedance of 
Class I increment creates a presumption that AQRVs within the affected 
impact area will also be adversely affected, but that presumption may 
be rebutted. Likewise, the absence of an increment exceedance creates a 
presumption that there is no adverse impact on AQRVs within the 
affected impact area, but that presumption may also be rebutted if the 
FLM provides evidence sufficient to convince the reviewing authority 
that emissions from a proposed source will have an adverse impact on 
AQRVs. Thus, based on the interplay of sections 165(d)(2)(C)(ii) and 
165(d)(2)(C)(iii), we interpret the Act to establish AQRVs, rather than 
the Class I increment, as the controlling standard in Class I areas. 
AQRVs are always applicable in Class I areas, regardless of the status 
of the Class I increment.
    However, AQRVs are the controlling benchmark only to the extent 
that AQRVs provide more protection than the Class II increments (or a 
lower figure in the case of the 3-hour averaging time for 
SO2). Section 165(d)(2)(C)(iv) indicates that, although a 
permit may be issued where AQRVs are not adversely impacted, such 
permit must ensure that the Class II increments are not exceeded. We 
interpret this provision to mean that the Class II increment cannot 
ever be exceeded in a Class I area, notwithstanding the degree of 
impact on AQRVs. So, reading sections 165(d)(2)(C)(ii)-(iv) together, 
we interpret the Act to establish AQRVs and the Class II increments to 
be the air quality standards that ultimately determine whether a permit 
may be issued for a source potentially affecting a Class I area. The 
Class I increment serves to establish a presumption of harm or the 
absence of harm to AQRVs, but does not ultimately control whether a 
permit may be issued.
    While it is clear that AQRVs and the Class II increments ultimately 
control whether a particular permit may be issued, the Act does not 
specify what role the Class I increment has to play on an ongoing basis 
after a variance has been issued. To obtain a variance, the applicant 
must rebut the presumption that AQRVs will be adversely impacted by an 
increase in concentrations in excess of the Class I increment. Once 
that presumption has been rebutted for a particular area, the Class I 
increment may no longer be representative of the degree of impact on 
AQRVs for that area. If the Class I increment has been exceeded but 
there is no adverse impact on AQRVs, this indicates that the Class I 
increment is not a reliable predictor of adverse impacts on AQRVs in a 
particular area.
    Thus, the question arises as to whether the Class I increment 
should remain applicable in a Class I area after the issuance of a 
variance. Section 165(d)(2)(C) does not address this issue. Although 
section 165(d)(2)(D)(iii) says that the ``otherwise applicable'' 
increment may not be exceeded more than 18 days per year in the case of 
a gubernatorial or presidential variance, section 165(d)(2)(C)(iv) does 
not refer to any ``otherwise applicable'' increment in the context of 
an FLM variance. The other parts of section 165(d)(2)(C) also fail to 
address this issue.

[[Page 31383]]

    One approach we have considered is to construe the silence in 
section 165(d)(2)(C) as an indication that Congress did not intend to 
permit violations of the Class I increment for any additional days 
beyond the one day per year allowed in the case of the 24-hour and 3-
hour increments. Under this interpretation, a variance under section 
165(d)(2)(C) would be considered only to be a variance from the ``cause 
or contribute'' standard in section 165(a)(3) of the Act for purposes 
of an individual permit application. An applicant would be relieved of 
the obligation to demonstrate that a proposed source does not cause or 
contribute to a violation of the Class I increment if the applicant can 
demonstrate that the source will not adversely affect AQRVs. However, 
under this view, the variance would not necessarily relieve the 
reviewing authority or State air quality planning agency from the 
obligation to ensure that the SIP contains measures to protect the 
Class I increment. The source might receive its permit based on the 
variance from section 165(a)(3) for a particular Class I area, but the 
State would remain obligated to comply with 40 CFR 51.166(a)(3) of the 
PSD regulations and take subsequent action to amend the SIP to correct 
the exceedance of the Class I increment caused by the source that 
received the variance.
    The latter interpretation appears to be supported by a statement 
from the DC Circuit's opinion in Alabama Power v. Costle. In this 
decision, the Court upheld the language cited above (40 CFR 
51.166(a)(3)) that requires a State to revise its SIP to correct a 
violation of the increment.\10\ Some of the Petitioners in that case 
had argued that EPA could not require a State to remedy a Class I 
increment violation, because section 165(d) allowed a waiver of the 
Class I increment in certain circumstances. The court reconciled the 
variance provision and the language in Sec.  51.166(a)(3) as follows:

    \10\At the time of that decision, this language was contained in 
Sec.  51.24(a)(3) of EPA's regulations. See 636 F.2d at 361 n. 92.

    Industry petitioners also rely on those sections of the Act that 
provide for waiver provisions which, conceivably, could allow 
increments to be exceeded. The waiver has vitality and recognition 
in that facilities granted special consideration under these 
provisions are, in effect, treated as facilities operating in 
compliance with the provisions of the Act. But the totality of 
facilities in compliance, as a group, may be subject to measures 
necessary to cope with a condition of pollutants exceeding the PSD 
---------------------------------------------------------------------------
maximum.

See 636 F.2d at 363.
    We have previously acknowledged that this may be a permissible way 
to reconcile the FLM variance provision with the requirement in Sec.  
51.166(a)(3) to amend SIPs to remedy an increment exceedance. In 
correspondence sent to the State of North Dakota, the Director of EPA's 
Office of Air Quality Planning and Standards recommended the approach 
suggested by the Alabama Power opinion. The letter stated the 
following:

    In the case of a Class I increment violation, a source may be 
granted a variance under certain conditions. First, the source must 
demonstrate to the FLM, and the FLM certify to the State, that the 
source will not adversely impact any Class I AQRVs. Second, the 
State must revise its SIP to correct increment violations ([Act] 
Section 161 and 163, 40 CFR 51.166(a)(3)).

See Letter from John Seitz, EPA/OAQPS, to Francis Schwindt, North 
Dakota Dept. of Health (December 12, 2001). EPA Region 8 followed this 
recommendation in comments submitted to North Dakota in 2002. See EPA 
Comments on North Dakota Department of Health's Proposed Determination 
Regarding the Adequacy of the SIP to Protect PSD Increments for Sulfur 
Dioxide (May 24, 2002).
    Since the time of these recommendations, we have evaluated this 
issue further and now recognize that there may be more than one 
permissible reading of the Act on this issue. The approach that we 
suggested in 2001 (amending the SIP to eliminate the Class I increment 
exceedance after the permit issues) would effectively require the 
source seeking the variance to obtain offsets from other sources 
affecting the Class I increment. If section 165(d)(2)(C) is read to 
require that a variance source obtain offsets, there would be no need 
for that proposed source to demonstrate that its emissions would not 
have an adverse impact on AQRVs. This would render the AQRV provisions 
in section 165(d)(2)(C) of the Act meaningless where the increment is 
exceeded because one would not need to consider AQRVs and obtain the 
variance in the first place if offsetting emissions reductions were 
obtained. Furthermore, where a single source consumes the entire 
increment but does not adversely impact AQRVs, the issuance of a 
variance would have no effect because a SIP could not be tightened to 
obtain reductions from any other source to remedy the increment 
exceedance. In this circumstance the State would have no choice but to 
tighten or revoke the permit of the variance source immediately after 
the permit was issued. We do not believe Congress intended such a 
result. In light of these considerations, we are proposing to refine 
our interpretation of section 165(d)(2)(C) with respect to the role of 
the Class I increment after a variance has been issued under section 
165(d)(2)(C).
    Another possible approach would be to read section 165(d)(2)(C)(iv) 
to call for the Class II increments to substitute for the Class I 
increment on an ongoing basis after a variance is issued. We might 
construe the absence of any discussion of an ``otherwise applicable'' 
increment in this section of the Act to mean that Congress did not 
intend for the Class I increment to have continuing effect in the area 
after the variance was issued. Since Congress did not specify the 
number of days on which the ``otherwise applicable'' increment could be 
exceeded per year (as it did in section 165(d)(2)(D)(iii)), one 
interpretation is that this information was not needed because Congress 
did not intend for the Class I increments to apply after it was 
demonstrated that the Class I increment was not a reliable predictor of 
the degree of impact on AQRVs in a particular Class I area. Under this 
approach, the Class II increments (plus the unique 3-hour 
SO2 increment) would continue to provide an upper bound on 
emissions growth to protect the Class I area while AQRVs remained in 
effect to protect against site-specific impacts that are not adequately 
represented by the Class I increment. However, under this Class II 
increment substitution approach, the Class I increment would no longer 
be available as a tool to determine who has the burden of proof to 
demonstrate the degree of impact on AQRVs.
    In this action, we are proposing a compromise approach that retains 
the Class I increment for the purpose of establishing the burden of 
proof in the AQRV analysis but does not require a SIP to be amended to 
offset the contribution of sources that have received a variance 
because they do not adversely affect AQRVs. We propose to accomplish 
this effect by allowing States to exclude the emissions from sources 
receiving an FLM variance from the Class I increment consumption 
calculation. The emissions of the variance source must continue to be 
considered for purposes of determining compliance with the Class II 
increments, but they would no longer be considered relevant to the 
Class I increment assessment after a variance has been issued. The 
Class I increment would remain in effect with respect to the emissions 
of other sources, and could not be exceeded on any additional days. The 
emissions of sources that have

[[Page 31384]]

not received a variance would continue to count against the Class I 
increment.
    For example, assume that an impact area for a proposed new source 
contains four sources that currently consume the SO2 
increment for the 3-hour averaging period--two of which have FLM 
variances and two of which do not. There are no other increment 
consuming or expanding sources in the impact area. For the 3-hour 
averaging period for SO2, the Class I increment is 25 [mu]g/
m\3\ and the alternative increment that applies after issuance of an 
FLM variance in this area is 325 [mu]g/m\3\.\11\ Assume that the two 
sources with variances consume 4 [mu]g/m\3\ each, for a total of 8 
[mu]g/m\3\. Assume that the two sources without variances consume 10 
[mu]g/m\3\ each, for a total of 20 [mu]g/m\3\. Under this scenario, if 
a new source applies for a permit, under this proposed rule the new 
source must combine its emissions with the emissions from the other two 
sources without variances and not exceed, for the Class I area of 
impact, 25 [mu]g/m\3\. Thus, the new source can consume up to 5 [mu]g/
m\3\ (i.e., 25 [mu]g/m\3\ minus 20 [mu]g/m\3\) of the available Class I 
increment for SO2 without assuming the burden of obtaining a 
third variance by demonstrating to the FLM that the source will not 
have an adverse impact on AQRVs in the Class I area.
---------------------------------------------------------------------------

    \11\As previously noted, the 3-hour averaging period for 
SO2 is unique in that the Act specifies an increment for 
purposes of the FLM variance (325 [mu]g/m\3\) that is different from 
the corresponding Class II increment (512 [mu]g/m\3\).
---------------------------------------------------------------------------

    Under this hypothetical example, because two sources in the area 
have previously obtained variances and shown that the Class I increment 
is not necessarily a reliable indicator of impacts on AQRVs, an 
alternative increment of 325 [mu]g/m\3\ now applies in the Class I area 
for all sources. The proposed source must combine its emissions with 
that of all 4 sources and not exceed a concentration increase of 325 
[mu]g/m\3\. Since the other four sources consume 28 [mu]g/m\3\, the new 
source can consume up to 297 [mu]g/m\3\ (i.e., 325 [mu]g/m\3\ minus 28 
[mu]g/m\3\) of the available increment for SO2.\12\
---------------------------------------------------------------------------

    \12\ The increment consumption estimates for all existing 
sources are based on modeling of their actual emissions, while the 
consumption estimate for the new source is based on modeling of its 
potential to emit (PTE).
---------------------------------------------------------------------------

    Furthermore, the AQRV test remains applicable to the ultimate 
decision as to whether the permit may be issued for the new source. 
Even though the new source, combined with the two existing sources 
without variances, may not cause or contribute to an exceedance of the 
Class I increment, the permit could nevertheless be denied if the FLM 
convinces the reviewing authority that the new source will have an 
adverse impact on AQRVs in the affected Class I area.
    Since a variance will not be issued unless the Class I area FLM 
certifies that the emissions from a proposed source will not have an 
adverse impact on AQRVs, it is reasonable to omit the emission of such 
source from the increment consumption analysis for the Class I 
increment on an ongoing basis. A source issued a variance does not 
adversely impact AQRVs, which as discussed above, is the critical and 
adaptable test Congress established for protecting site-specific 
concerns in Class I areas. Each successive source that impacts the 
Class I area would still have to show that it does not harm the AQRVs 
to receive a permit. The Class I increment would remain relevant as an 
indicator for assessing when other sources may have an adverse impact 
on AQRVs. If sources other than the variance source cause an exceedance 
of the Class I increment, the next source to apply for a permit 
affecting the area will have the burden of demonstrating to the FLM 
that the proposed source's emissions do not adversely affect AQRVs. If 
the emissions of the proposed source and other sources that have not 
received a variance do not consume the Class I increment, then the FLM 
will bear the burden of convincing the reviewing authority that the 
proposed source will adversely impact AQRVs. Plus, the alternative 
increments (generally the Class II increments) apply to limit the 
overall increase in concentrations caused by all sources affecting the 
Class I area.
    This approach is a permissible reading of the Clean Air Act that 
reconciles some apparent inconsistencies in the statutory scheme. Even 
when a variance is issued under section 165(d)(2)(C), the Act does not 
expressly allow the Class I increment to be exceeded on any additional 
days. If this omission were read strictly to preclude any additional 
days of violation of the increment, this would be inconsistent with 
allowing a variance because the strict reading would preclude any 
additional days of a Class I increment violation, even those caused by 
a variance source. The issuance of a variance would appear to require 
at least a temporary variance from the Class I increment, even if the 
SIP still has to be amended at a later date to correct the violation, 
but that would be inconsistent with a strict reading of section 
165(d)(2)(C)(iv) to preclude additional violations of the Class I 
increment. If section 165(d)(2)(C)(iv) is read to require that the 
Class II increment permanently supersede the Class I increment, an 
unlimited number of additional days of Class I increment violations 
would be permitted and the burden shifting effect of the Class I 
increment would be lost. Our proposed approach of excluding the 
emissions of variance sources from the Class I analysis appears to be 
the best way to avoid authorizing any additional days of Class I 
increment violations while retaining the role of the Class I increment 
as a tool to determine who has the burden in the AQRV analysis.
    Because of the differences between section 165(d)(2)(C) and 
165(d)(2)(D), we do not propose to apply this same exclusion to 
variances issued under section 165(d)(2)(D). Instead of allowing an 
exclusion from the Class I increment consumption analysis, it appears 
that Congress opted in section 165(d)(2)(D) to apply the otherwise 
applicable Class I increment and instead to allow that increment to be 
exceeded on 18 days per year instead of the normal limit of 1 day per 
year.
    We also propose to use this rule as an opportunity to correct a 
typographical error in the provisions of our rules addressing the FLM 
variances. The cross references contained within 40 CFR. 51.166(p) and 
52.21(p) incorrectly refer to paragraph (q) of these provisions. We 
propose to amend these provisions so they reflect the correct cross-
references to portions of paragraph (p).

B. How are emissions estimated for sources that consume increment?

    To model the expected change in concentration of pollutants above 
the baseline, one needs to identify the emissions of those sources that 
are included in the increment consumption analysis. As noted earlier, 
the PSD regulations call for this analysis to be based on the actual 
emissions of sources. The baseline concentration is generally based on 
``actual emissions * * * representative of sources in existence on the 
applicable minor source baseline date.'' See 40 CFR 51.166(b)(13)(i)(a) 
and 52.21(b)(13)(i)(a). The concentration after the minor source 
baseline date is generally based on ``actual emissions increases and 
decreases * * * at any stationary source occurring after the minor 
source baseline date.'' See 40 CFR 51.166(b)(13)(i)(b) and 
52.21(b)(13)(ii)(b). There are certain exceptions to these general 
principles for emissions of major sources, but the basic methodology 
involves identifying the actual emissions of sources on the minor 
baseline date and actual emissions increases and decreases after

[[Page 31385]]

the minor source baseline date at sources existing on the minor source 
baseline date and increases attributable to the addition of new sources 
since that time.
    In practice, an assessment of increment consumption in accordance 
with these requirements has generally involved compiling an actual 
emissions inventory for two separate time periods. The first part of 
the inventory generally contains actual emissions as of the minor 
source baseline. However, for major sources that experienced changes in 
emissions resulting from construction (as defined at 40 CFR 
51.166(b)(8) and 40 CFR 52.21(b)(8)) after the major source baseline 
date, the emissions as of the major source baseline date would be used. 
The second part of the inventory contains actual emissions as of the 
time of a periodic review of increment compliance or the review of a 
pending PSD permit. In the case of a PSD permit review, the second part 
of the inventory contains the projected emissions of the proposed 
source. The existing PSD regulations contain a definition of the term 
``actual emissions'' in 40 CFR 51.166(b)(21) and 52.21(b)(21). This 
definition is expressly incorporated into the definition of ``baseline 
concentration'' which establishes the basic parameters described above 
for determining the change in concentration since the baseline date.
    In this action, we are proposing to adopt a revised definition of 
``actual emissions'' that will address the methodology for quantifying 
emissions as of the baseline date and emissions that consume increment. 
Rather than revising the existing definition of actual emissions in 40 
CFR 51.166(b)(21) and 52.21(b)(21) which may continue to be used for 
other purposes under the PSD program, we propose to promulgate a new 
definition of ``actual emissions'' in 40 CFR 51.166(f) and 52.21(f) 
that will apply only to the analysis of increment consumption and be 
easier to find among other provisions pertaining to the increment 
consumption analysis. We also request comment on whether we could also 
repeal the existing definition of actual emissions in 40 CFR 
51.166(b)(21) and 52.21(b)(21) without affecting other elements of the 
PSD program.
1. Data and Calculation Methods Used to Establish Actual Emissions
    We propose to add language to the PSD regulations to clarify that a 
reviewing authority has discretion to use its best professional 
judgment when determining the actual emissions of sources as of the 
baseline date and at subsequent periods of time, particularly where 
there is limited data available from which to determine actual 
emissions. We propose to establish a general standard for the 
sufficiency of data and calculation methods on which actual emissions 
may be based, but also request comment on WESTAR's recommendation that 
EPA establish a menu of permissible data types and calculation methods 
from which each reviewing authority may select.
    Background. Because direct measurement of the emissions from a 
stack may not be available, the emissions of baseline and increment 
consuming sources must often be derived from other data that is 
available. The current regulations applicable to increment consumption 
analyses specify that ``actual emissions shall be calculated using the 
unit's actual operating hours, production rates, and types of materials 
processed, stored, or combusted during the selected time period.'' See 
40 CFR 51.166(b)(21) and 52.21(b)(21). This general requirement adopted 
in the PSD regulations in 1980 presumed the availability of reliable 
and consistent records on operating hours, production rates, and 
materials composition.
    However, the experience of EPA and many States in implementing the 
PSD program since this time has shown that the accuracy and reliability 
of the available data may be questionable or may vary significantly 
over the time period of the emissions estimate. For PSD baseline dates 
that are many years in the past, information on actual source 
operations may be sketchy or lacking altogether. Furthermore, the 
composition of raw materials, such as the sulfur content of coal, may 
change over time and might be reliably estimated for an annual average 
value, but may be significantly higher during a shorter period of time 
within that year or when a maximum value is determined.
    There may also be cause to choose among various calculation 
methodologies for a given emissions estimate. For example, annual 
emission rates could be calculated based on continuous operation (24 
hours per day, 365 days per year). If a source does not operate 
continuously, whether by design or permit limitation, the annual 
emissions could be based on the limitation. Due to scheduled shutdowns 
and maintenance, sources rarely operate at design or permit limits, and 
in such cases actual operating hours could be used. However, there will 
be situations when data on operating hours are not available and some 
other estimate of operation must be determined. The choice of which 
data to use in a particular circumstance, particularly where there is 
more than one set of data that could be used or more than one 
methodology, has generated substantial uncertainty in the context of 
the PSD program. This uncertainty also extends to how gaps in the data 
are handled, such as when data are unavailable or are available for 
only a subset of a group of similar sources.
    Other than the language quoted above from the definition of 
``actual emissions'' calling for emissions to be calculated based on 
actual operating hours, production rates, and materials composition, 
the PSD regulations have not included any criteria for reviewing 
authorities to use to determine actual emissions. We have provided more 
specific guidance for demonstrations of compliance with the NAAQS under 
the PSD program in table 8-2 of appendix W, but this table was not 
developed for purposes of increment consumption analysis. Section 
8.1.2.i. currently recommends only that ``NAAQS compliance 
demonstrations in a PSD analysis should follow the emission input data 
shown in Table 8-2.'' We do not believe our recommendations in Table 8-
2 can be readily extended to increment consumption analyses because of 
differences in the increment consumption analysis. Unlike the NAAQS 
analysis, increment consumption assessments have generally focused on 
changes in emissions, rather than absolute concentrations, and often 
must account for emissions that occurred many years earlier on the 
applicable baseline date.
    We do not necessarily read the Act to call for the same degree of 
precision in the increment consumption analysis as a determination of 
compliance with the NAAQS. Under the constraints imposed by Congress, 
the increment analysis is in many ways an artificial assessment because 
the actual emissions as of the date of the first PSD permit application 
in an area must be adjusted. This adjustment accounts for emissions 
increases resulting from construction (as defined at 40 CFR 
51.166(b)(8) and 40 CFR 52.21(b)(8)) at major sources in the area that 
occurred prior to that date. CAA section 169(4). In addition, the 
actual emissions of some sources may be omitted from the analysis 
altogether under section 163(c) of the Act. Because Congress required 
or permitted these adjustments to the calculation of baseline 
concentrations and concentrations after the baseline date, we believe 
the method used to determine increment consumption should endeavor to 
provide a representative indication of the relative magnitude by which 
air quality

[[Page 31386]]

concentrations have changed over time, but is not necessarily required 
to provide an exact prediction of the change in air quality 
concentrations from one date to another.
    Proposed Action. To address the uncertainty in how to determine 
actual emissions for increment consumption purposes, we propose to 
codify a policy that gives the reviewing authority discretion to select 
the data and emissions calculation methodologies that are reliable, 
consistent, and representative of actual emissions. The cornerstone of 
such a policy is that emissions estimates used to establish baseline 
concentrations and increment consumption or expansion must be supported 
by the available record and be rationally-based. This policy would give 
reviewing authorities the discretion to use the best available 
information and to make reasonable judgments as to the reliability of 
that information for determining actual emissions, particularly when 
estimating emissions for baseline dates in distant years for which very 
little useful data may be available. In addition, this policy would 
seek to ensure a reliable estimate of the change in air quality 
concentrations by encouraging reviewing authorities to evaluate the 
degree of change by comparing consistent data types or concentration 
predictions (i.e., to conduct an ``apples'' to ``apples'' comparison of 
the change in emissions or concentrations). We believe that this 
flexible approach is preferable to a rigid requirement to use a 
specific type of data or calculation method because of uncertainty over 
the exact type and quality of data that will be available in each 
instance.
    This policy is consistent with existing recommendations in appendix 
W and EPA guidance. Section 8.0.a. of appendix W currently states that 
``[t]he most appropriate data available should always be selected for 
use in modeling analyses.'' This approach is consistently applied 
throughout appendix W wherein the reviewing authority is given 
discretion to approve the selection of input data for air quality 
models.
    We have generally given reviewing authorities substantial leeway 
within the PSD program to select data and emissions calculation 
methodologies that they believe are representative of actual emissions. 
We recognize that where the available data are poor, substantial 
judgment must be used to estimate actual emissions. Once the reviewing 
authority has selected data and emissions calculation methodologies 
according to general guidelines, we typically have not second-guessed 
their choices. In particular, we have not required reviewing 
authorities to select data or methodologies that we might consider 
``more reasonable'' or ``more representative'' than those they have 
chosen.
    We propose to give each reviewing authority the responsibility to 
verify and approve the data used, and to assure that it meets a basic 
standard of reliability, consistency, and representativeness. In light 
of the fact that many recommendations in section 8.0 of appendix W are 
not necessarily applicable to the increment analysis, we propose to 
make clear that this standard will control over the recommendations in 
appendix W.
    We request comment on this policy, and on the regulatory language 
proposed at 40 CFR 51.166(f)(1)(iv) and 52.21(f)(1)(iv) to codify this 
policy. In addition, we request comment on whether additional guidance 
or limitations should be articulated and codified for estimating 
emissions that make up the baseline concentration or consume increment.
    Request for comment on WESTAR recommendation. In its May 2005 
recommendations, WESTAR expressed the view that EPA should ``afford 
reviewing authorities some flexibility to ensure that analyses 
accommodate considerations such as data availability and accuracy.'' 
However, WESTAR also asked us ``to encourage consistency, 
predictability, and regulatory certainty with regard to approaches for 
preparing emissions inventories for refined PSD analyses.''
    In order to achieve these goals, WESTAR recommended a two-step 
approach. The first step would be for EPA to develop a ``menu'' of 
acceptable emissions calculation approaches for both short-term and 
annual PSD analyses. The second step would allow the reviewing 
authority to select what they believed to be the most appropriate 
option from the menu based on a set of guiding principles. The 
reviewing authority would be able to use calculation approaches not 
included in the menu provided that they can demonstrate that the 
approach is consistent with the Act and NSR regulations, as well as the 
principles included in step two. According to WESTAR's report, this 
two-step approach would help alleviate the current lack of clarity and 
narrow interpretations of the definition of actual emissions used for 
emissions inventories in PSD analyses.
    WESTAR's report identifies various types of data that might be used 
in the menu. These data types are discussed in more detail below in the 
context of the more specific issue of short-term emissions estimates.
    WESTAR also provided guiding principles that could be used in 
selecting among the menu items. These principles are the following:
     Maximize the accuracy of the method(s) in reflecting the 
actual status of air quality during each time period associated with 
applicable standards;
     Conform to the Act, Federal PSD rules, and other 
applicable laws and rules;
     Ensure consistency between emissions calculation methods 
used for sources in the baseline emissions inventory and the current 
emissions inventory;
     Ensure that selected methods are practical given the 
availability of reviewing authority access to the emissions data;
     Support fairness and consistency in how emissions are 
calculated for various source types across and within States; and
     Support key air quality management objectives that States 
and EPA are seeking to achieve, such as encouraging sources' use of 
continuous emissions monitoring systems (CEMS) and discouraging sources 
from seeking more permitted air quality increment than they need.
    We request comment on WESTAR's proposed approach. For more 
information, we encourage you to review the WESTAR recommendations that 
can be found in the docket for this rulemaking. We also request comment 
on any other aspect of selecting data and calculation methodologies for 
emissions inventories for PSD analyses.
2. Time Period of Emissions Used To Model Pollutant Concentrations
    In this action, we are also proposing amendments to clarify the 
time periods to be used for emissions from sources included in the 
calculation of the baseline concentration and the change in 
concentration after the baseline date. In general, we have called for 
the modeling change in concentration to be based on the emissions rates 
from increment consuming sources over the 2 years immediately preceding 
a particular date. However, there are circumstances when another period 
of time may be more representative of actual emissions as of a 
particular date. This rulemaking is intended to clarify those 
circumstances when it is permissible to use another period of time to 
represent actual emissions as of a particular date for purposes of 
calculating the change in concentration used to evaluate consumption of 
PSD increments.

[[Page 31387]]

    Background. Since source operations are inherently variable over 
time, the NSR regulations do not require that ``actual emissions'' on a 
particular date be based only on the emissions occurring on that single 
date. Instead, the regulations generally require that the baseline 
concentration be based on an average of the emissions observed over the 
2 years prior to the baseline date (40 CFR 51.166(b)(21)(ii) and 
52.21(b)(21)(ii)). However, we have long recognized an exception to 
this general rule, which provides that a different period of time may 
be used when another period of time is more representative of normal 
source operations (40 CFR 51.166(b)(21)(ii) and 52.21(b)(21)(ii)).
    The original definition of ``actual emissions'' was used in several 
different ways under the NSR program. In addition to being incorporated 
in the definition of ``baseline concentration'' and thus used for 
purposes of determining consumption of increment, this definition of 
``actual emissions'' has also been applied for the purpose of 
identifying the change in emissions attributable to the modification of 
a major source. An existing major source is subject to NSR if it 
engages in a major modification which is defined to mean ``any physical 
change in or change in the method of operation of a major stationary 
source that would result in a significant emissions increase * * * and 
a significant net emissions increase of that pollutant from the major 
stationary source.'' See 40 CFR 51.166(b)(2) and 52.21(b)(2). Prior to 
2002, the definition of ``actual emissions'' in 40 CFR 51.166(b)(21) 
and 52.21(b)(21) applied to determine the actual emissions of the 
source prior to the change and after the change.
    In 2002, we adopted a new definition of ``baseline actual 
emissions'' that is now used to determine actual emissions before a 
change for purposes of determining whether a source is proposing a 
major modification that requires a preconstruction permit. This 
definition allows non-utility units to identify pre-change emissions 
using any 2-year period in the 10 years preceding and requires electric 
utilities to use any consecutive 2 years in the last 5 years. We 
adopted this new definition to reflect the emissions levels that occur 
during a normal business cycle, without requiring sources to 
demonstrate to the reviewing authority that another period is more 
representative of normal source operation. See 67 FR 80191-92. However, 
in that rulemaking, we made clear that original ``actual emissions'' 
definition continues to apply for other purposes under the PSD program. 
We observed that the existing definition of actual emissions 
``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.'' See 67 FR 80192, footnote 13; 67 FR 80196.
    Prior to 2002, when determining the baseline actual emissions at a 
source experiencing a modification that might trigger NSR, we applied 
the ``more representative of normal source operations'' exception in 40 
CFR 51.666(b)(21) and 52.21(b)(21) in a narrow set of circumstances. 
For example, in 1999, the Administrator addressed this issue in 
response to a petition to object to issuance of a title V operating 
permit and observed that EPA ``has applied its discretion narrowly in 
assigning representative periods other than the 2 years immediately 
preceding the physical or operational change.'' See Order Responding to 
Petitioner's Request That Administrator Object to Issuance of State 
Operating Permit, In the Matter of Monroe Electric Generating Plant 
Entergy Louisiana, Petition No. 6-99-2. In a draft 1990 guidance 
document, the agency observed that normal source operations ``may be 
affected by strikes, retooling, major industrial accidents, and other 
catastrophic occurrences.'' NSR Workshop Manual at A.39. Based on these 
examples, we have sometimes looked for evidence of a ``catastrophic 
occurrence'' before permitting an alternative period to be used to 
establish the actual emissions of a source prior to a modification. For 
example, in a 1992 memorandum, the Director of the Air Quality 
Management Division (AQMD) concluded that the exception should not be 
invoked for a source that had been idle for 10 years due to economic 
reasons and had not demonstrated that operations of the plant were 
disrupted by catastrophic occurrences or other extraordinary 
circumstances. The director identified strikes and major industrial 
accidents as examples of catastrophic occurrences. Memo from John 
Calcagni, AQMD, to David Kee, Region V (August 11, 1992). Although we 
have, in our discretion, applied the definition in 40 CFR 51.166(b)(21) 
and 52.21(b)(21) narrowly, we did not amend these regulations to 
restrict application of the ``normal source operation'' exception in 
the definition of ``actual emissions'' to only catastrophic 
occurrences. In recent years, we have moved away from this approach in 
rulemaking actions.
    In the process of establishing the new definition of ``baseline 
actual emissions'' for applicability purposes, we observed that the 
more representative or normal source operation provision ``has been a 
source of confusion and uneven implementation.'' See 61 FR 38259, July 
23, 1996. This observation was based on our experience with identifying 
increases in emissions for purposes of determining whether a source was 
proposing to undergo a major modification and required a permit. We 
were not concerned at that time about the application of this exception 
in the context of the PSD increment analysis. However, we have since 
discovered that the legacy of implementing the ``normal source 
operation'' exception in the context of NSR applicability has had a 
collateral effect of fostering confusion in those circumstances, such 
as PSD increment analyses, where the ``actual emissions'' definition in 
40 CFR 51.166(b)(21) and 52.21(b)(21) continues to apply. Recently, the 
question has arisen as to whether the guidance we provided on the 
``more representative of normal source operations'' exception in the 
applicability context should also be applied in the context of 
increment consumption analysis. As a result of this question, we have 
been reviewing the issue, and propose to clarify our position in this 
rulemaking.
    Proposed Action. In this action, we are proposing to establish a 
new definition of ``actual emissions'' (applicable only to the 
increment consumption analysis) which clarifies the circumstances when 
it is permissible, in the context of an increment consumption analysis, 
to determine actual emissions for increment consuming sources using a 
period of time other than the 2 years immediately preceding the 
relevant date. We propose to codify this element of the new definition 
in 40 CFR 51.166(f)(1)(iv) and 52.21(f)(1)(iv) of the PSD regulations.
    This issue has arisen most recently in the context of determining 
the actual emissions of sources as of the baseline date. However, we 
recognize that this issue could also arise when seeking to establish 
the ``present day'' inventory of emissions increases or decreases after 
the baseline date. Under existing regulations, the same definition of 
actual emissions applies in each instance. Our proposed definition of 
``actual emissions'' for the increment consumption analysis is intended 
to apply to both sides of the ledger in order to provide consistency. 
We believe the same principles should apply when determining emissions 
as of the baseline date and the present day.

[[Page 31388]]

    The proposed revisions are intended to address three primary 
issues. First, we propose to clarify that one is not required to 
demonstrate the occurrence of a catastrophic event in order to 
determine actual emissions on the basis of a period other than the 2 
years immediately preceding the date in question. Second, we seek to 
clarify that there can be circumstances where emissions increases 
occurring after the baseline date or due to increases in hours of 
operation or capacity utilization may be more representative of normal 
source operation. Third, we are clarifying that when an alternative 
(more representative) time period other than the 2 years before the 
particular date is used to reflect actual emissions, that alternative 
time period must be representative of source emissions (within an 
expected range of variability) as of the particular date and cannot be 
based on emissions experienced because of a change in the normal 
operations of that source after that date.
    With respect to the first issue (whether a ``catastrophic 
occurrence'' must be shown), we have historically approached the 
``normal source operation'' exception differently in the context of the 
PSD increment analysis. The guidance in which we have looked for 
evidence of ``catastrophic occurrences'' only addressed the subject of 
baseline actual emissions prior to a modification and did not discuss 
how to determine the emissions of sources on the PSD baseline date for 
increment purposes. As discussed further below, in the context of the 
PSD baseline concentration, we have not previously limited the 
application of the ``normal source operation'' exception to those 
circumstances where a source experienced a malfunction or catastrophic 
event. In the context of increments, we have recognized that the 
``normal source operation'' exception may apply in other kinds of 
circumstances where it can be shown that source emissions in the 24 
months preceding the baseline date are not representative of its normal 
operations at the time of the baseline date.
    We do not believe it is appropriate to define ``actual emissions'' 
as narrowly in the context of PSD increment consumption analysis as it 
had been applied in the context of PSD applicability determinations 
before 2002. Although we have looked for evidence of ``catastrophic 
occurrences'' to establish that another time period is more 
representative of actual emissions prior to a modification, we do not 
believe this fact alone justifies using a similar approach for 
identifying representative periods of actual emissions in the context 
of a PSD increment analysis. The modification context in which this 
approach was once used is different from the increment consumption 
context. The former involves the initial determination of whether a PSD 
permit is required, and evaluates only an increase in emissions from a 
single source resulting from a proposed change. By contrast, an 
increment compliance assessment is performed after it is clearly 
established that a source must obtain a PSD permit (or may be done in a 
periodic review when no permit is pending) and evaluates a change in 
air pollutant concentration using modeling and emissions data inputs 
for multiple sources. We believe the differing nature of the increments 
analysis justifies a different approach.
    As to the second issue described above, our proposal to sometimes 
allow emissions after the baseline date to be used to calculate the 
baseline concentration is consistent with our historic interpretation 
of the ``normal source operation'' exception in the context of the 
increment consumption analysis. In our original PSD regulations after 
the 1977 Amendments to the Act, we considered emissions increases 
attributable to increases in hours of operation or capacity utilization 
to be a part of the baseline concentration (rather than increment 
consuming increases) if the source was allowed to operate at that level 
in 1977 and could have reasonably been expected to make those increases 
at the time. See 43 FR 26400, June 19, 1978. However, in 1980, we 
eliminated the automatic inclusion of these emissions in the baseline 
concentration. Instead, we chose to address the issue on a case-by-case 
basis when it could be demonstrated that emissions attributable to 
increased utilization were more representative of normal source 
operation under the definition of ``actual emissions.'' When we adopted 
this change, we said that ``if a source can demonstrate that its 
operation after the baseline date is more representative of normal 
source operation than its operation preceding the baseline date, the 
definition of actual emissions allows the reviewing authority to use 
the more representative period to calculate the source's actual 
emissions contribution to the baseline concentration.'' See 45 FR 
52714, Aug. 7, 1980. We continue to view this to be an appropriate 
policy and propose regulatory language to make this explicit in the 
regulations.
    Identifying ``actual emissions'' based on representative emissions 
as of the PSD baseline date is consistent with the opinion of the D.C. 
Circuit in the Alabama Power case. In that decision, the court noted 
the following:

    Congress did not intend a simple measurement of air quality on a 
day with atypical conditions to control calculation of the baseline. 
Reasonable efforts to ascertain the actual but usual concentration 
levels, as of the date of the first applicable for a permit, are 
required.

See Alabama Power, 636 F.2d at 380 n. 44. We believe that the proposed 
definition of ``actual emissions'' for increment consumption purposes 
is consistent with Congressional intent, as described by the court. It 
is reasonable to allow a showing that a period other than the 24 months 
prior to the baseline date are representative of the ``usual'' 
concentration levels at the time of the baseline date where emissions 
after the baseline date can be shown to represent the ``usual'' or 
``normal'' concentration levels. As observed by the court in Alabama 
Power, ``Congress expected EPA to use `administrative good sense' in 
establishing the baseline and calculating exceedances.'' See Alabama 
Power, 636 F.2d at 380. We have considered this approach to make good 
sense since 1980. Although emissions after a baseline date may 
sometimes be reflected in the baseline concentration, this has 
historically been a narrow exception because, in general, increases in 
emissions that occur after the baseline date consume increment. See 40 
CFR 51.166(b)(13) and 52.21(b)(13); see also draft NSR Manual at C.35 
and C.48.
    With respect to the third issue listed above, while we propose to 
clarify that emissions after the baseline date may sometimes be used to 
represent actual emissions as of the baseline date, we must also 
emphasize that this is permissible only in limited circumstances. We 
propose to include language in our new definition that limits the 
circumstances under which post-baseline date emissions can be 
considered representative of normal source operations for purposes of 
establishing the baseline concentration. Such a limitation is needed to 
ensure that the increment system continues to function as intended to 
prevent significant deterioration from actual increases in emissions 
after the baseline concentration is established. We seek to ensure that 
real increases in emissions that are outside of a normal range of 
variability will continue to be regarded as consuming increment, while 
recognizing that due to the normal variability in source operations, 
some apparent increases in emissions are justifiably included in the 
baseline where they are representative of the emissions experienced by 
a source as of

[[Page 31389]]

the baseline date. We believe that increases in emissions that are not 
attributable to the normal variability of source operations at a 
particular time are actual increases that should be counted as 
consuming the available increment.
    Under the Act and applicable case law, it is clear that the 
emissions that make up the baseline concentration must be 
representative of air pollutant concentration levels at the time of the 
baseline date. Section 169(4) of the Act defines baseline concentration 
as the ``ambient air concentration levels which exist at the time of 
the first application for a permit.'' In the Alabama Power decision, 
the court observed that the baseline concentration is tied to first 
permit application because Congress intended permitting authorities to 
use actual data to establish baseline or make permit applicants collect 
data at the appropriate time. See 636 F.2d at 375-76. In defining 
baseline concentration, we have required a baseline concentration to be 
based on ``actual emissions * * * representative of sources in 
existence on the applicable minor source baseline date.'' See 40 CFR 
51.166(b)(13)(i)(a).
    Our proposed approach should not be construed to allow emissions 
estimates as of the baseline date to be based on operations over the 
entire life of a source or a period of operations that is not 
representative of operations as of a particular date. Actual emissions 
as of a particular date must be representative of normal operations 
(which include an expected range of variability) during the applicable 
time period. For example, when estimating sulfur dioxide emissions from 
a coal-fired electric generating unit, we do not believe it is 
appropriate to use the weighted average sulfur content for coal from 
any period over the life of the mine supplying the facility. However, 
we recognize that there may be some variability in the sulfur content 
of the coal used by a source at the time a baseline date is 
established. For example, if the baseline date were some time in the 
1970s, we believe it would be appropriate for the emissions from this 
source to be based on a weighted average sulfur content for coal used 
by the source in the 1970s. However, we would not consider it 
appropriate for the source to use a weighted average of sulfur content 
from coal used in the 1990s to represent the composition of coal 
combusted in the 1970s, unless it can be shown that the composition of 
coal used in the 1990s is in fact representative of the coal the source 
actually used in the 1970s. Our intent is to revise the regulation to 
codify the approach reflected in our Memorandum of Understanding with 
North Dakota which calls for using the sulfur content of coal consumed 
during a unit's baseline normal source operations, rather than the 
sulfur content averaged over the entire life of a mine or any period of 
operations in the life of the source that is not representative of 
operations on a particular date.
    This approach is consistent with language in the existing 
definition of ``actual emissions,'' which provides that ``[a]ctual 
emissions shall be calculated using the unit's actual operating hours, 
production rates, and types of materials processed, stored, or 
combusted during the selected time period.'' See 40 CFR 
51.166(b)(21)(ii) and 52.21(b)(21)(ii). The selected time period under 
this provision should be either the 24 months before the particular 
date or an alternative period that is shown to be more representative.
    In order to ensure consistent measurement of increases in air 
pollutant concentration, we believe it is also appropriate to also 
apply the ``normal source operation'' exception in the context of the 
emissions inventory for the present day period. As applied to the 
present day inventory of emissions, if a source experiences lower than 
normal emissions in the 2 years preceding the review, more 
representative emissions should also be used in the present day 
inventory to avoid undercounting actual emissions increases.
    Thus, we propose to revise the regulatory language to allow actual 
emissions used in an increment consumption analysis to be computed 
based on the operations of a source during a time other than the 24 
months preceding a particular date upon a determination that such 
period is more representative of normal source operation as of the 
particular date if a credible demonstration can be made that the unit's 
operations in the 24 months preceding the date were not typical of 
operations as of the particular date. A period after the particular 
date may be used, but only if such period is more representative of 
normal source operations as of the particular date. Operations 
occurring prior to a particular date would not be considered 
representative of normal source operations for a particular date if 
they permanently ceased more than 24 months prior to that date. Under 
the proposed regulation, the alternative time period that is used to 
compute actual emissions must be another consecutive 24-month period 
unless two non-consecutive 12-month periods are demonstrated to be more 
representative of normal source operation under the criteria in the 
regulation.
3. Actual Emissions Rates Used to Model Short-Term Increment Compliance
    We also propose in this rule to clarify how one should derive 
source emissions rates of less than 1 year for sources contributing to 
the baseline concentration and increment consumption when evaluating 
compliance with the short-term (24-hour and 3-hour) increments for PM 
and SO2. Increments for a 24-hour averaging time are 
currently in place for both PM and SO2. The 3-hour averaging 
time is only used for the SO2 increments. Based on recent 
experience and the recommendations of WESTAR, we believe that we need 
to provide additional guidance to States and regulated entities 
concerning how to determine actual emissions for purposes of modeling 
the concentration changes over the 3-hour and 24-hour averaging times.
    Background. The definition of actual emissions in 40 CFR 
51.166(b)(21) and 52.21(b)(21) does not directly address how one is to 
determine actual emissions when modeling pollutant concentrations 
averaged over periods less than 1 full year. Under the current 
provision, actual emissions are identified using an annual average in 
tons per year. However, this section does not directly address how to 
determine actual emissions over shorter time periods, such as the 24-
hour or 3-hour averaging times that are used for some of the PSD 
increments.
    In draft guidance prepared in 1990, we recommended that sources and 
reviewing authorities use the ``maximum actual emissions rate'' for 
short-term averaging periods. See draft NSR Manual at C.49. We 
indicated that ``the maximum rate is the highest occurrence for that 
averaging period during the previous two years of operation.'' Id. We 
recommended using this maximum rate for both the current and the 
baseline time periods. Id. This was consistent with guidance that had 
been provided by at least one EPA Regional Office as far back as 1981. 
See Memorandum from Thomas W. Devine, Region IV, to State and Local Air 
Directors, ``Policy Determinations Regarding PSD Questions'' (July 31, 
1981).
    In practice, however, we have since come to recognize that there is 
often not sufficient data available to determine the maximum short-term 
emissions rate over a 2-year period. This type of determination will 
typically require CEMS. For PSD baseline dates

[[Page 31390]]

established in the 1970s and 1980s, these data are especially difficult 
to find. As a result of this difficulty, some States and EPA Regional 
Offices have allowed calculation of an average short-term rate using an 
average rate calculated from annual emissions in situations where 
short-term maximum actual emissions data are not available.
    Proposed Action. We propose to promulgate a new definition of 
``actual emissions'' applicable to the PSD increment analysis that 
specifically addresses how to derive short-term emissions rates when 
modeling the change in concentration for the 24-hour and 3-hour 
averaging periods used in increments for some pollutants. We propose to 
add a provision that allows permitting authorities to use their 
discretion to use data that promotes consistency in the analysis and 
does not bias the analysis in favor of one group of sources over 
another. Under this approach, an average short-term rate may be used if 
the reviewing authority finds this to be the best way to promote 
consistency and avoid bias. Maximum short-term rates may continue to be 
used where sufficient data are available, but need not be used in all 
circumstances. Although we have historically called for use of maximum 
short-term rates, some stakeholders have suggested that the modeled 
change in concentration may be overly conservative when increment 
consumption modeling is based on maximum emissions rates from all 
sources that consume increment. We understand it may not be reasonable 
to expect that increment-consuming sources will all be operating at 
their maximum short-term emissions rates at exactly the same time. If 
we were to require the use of maximum emissions rates in all instances, 
this would mandate that PSD modeling always be conducted using a 
scenario that is not necessarily representative of actual emissions or 
concentrations. As the court said in Alabama Power, EPA should use 
``reasonable efforts to ascertain the actual but usual concentration 
levels'' and ``administrative good sense in establishing the baseline 
and calculating exceedances.'' See Alabama Power, 636 F.2d at 380, 380 
n.44. Since it may be unusual for all increment consuming sources to 
all be operating at their maximum emissions rates at the same time, we 
believe that ``administrative good sense'' dictates that we permit 
average emissions rates to be used as well. However, we are not 
proposing to preclude use of a maximum rate where a reviewing authority 
or source wishes to conduct a more conservative screening analysis or 
considers a maximum rate more appropriate under the circumstances for 
all sources or just for certain sources in the inventory. In many 
cases, combining the average emissions rates of all increment consuming 
sources in an emissions inventory may produce a more representative 
picture of the degree of change in short-term pollution concentration 
over time.
    A more representative indication of the change in emissions is 
produced by using a consistent set of data. If actual short-term 
emissions rate or hourly operations data are only available from some 
sources in an inventory, the analysis could be biased by mixing these 
data with averages calculated from annual operational data. However, if 
the reviewing authority derives short-term emissions rates by averaging 
annual data from all sources in the inventory, this may provide a 
representative depiction of the change in emissions over time. 
Likewise, if reliable and consistent maximum or short-term rate data 
are available for all sources in the inventory, this could provide a 
representative assessment of the change in maximum rates over time. We 
are proposing to establish a standard that allows sources to select a 
consistent data set and to otherwise forgo using some maximum or actual 
short-term data that may be available, but is incomplete and would 
potentially bias the overall analysis when combined with data of a 
different type that must be used to complete the assessment. At the 
same time, we are not proposing to preclude reviewing authorities from 
mixing data of different types where they consider it appropriate and 
this technique produces a representative analysis.
    In addition, fairness also dictates that we allow use of average 
short-term emissions rates and not require use of maximum emission 
rates in all cases. If maximum emissions rates may be used when data 
are available but averages are used when the data are insufficient, the 
analysis may be biased against the sources that have maximum emissions 
rate data. We want to encourage the use of CEMS that have been shown to 
be reliable and want to avoid a policy that inadvertently discourages 
the development and use of CEMS. Where most sources in an area are 
using CEMS to track emissions, the maximum rate approach may be more 
equitable, but this may not be the case in all areas. Thus, we propose 
to give the reviewing authority discretion to use available data and to 
achieve equitable treatment across sources and consistency in the 
analysis.
    Request for Comment on WESTAR Recommendations. As part of its 
general approach of establishing a menu of available data and 
calculation methodologies, WESTAR has recommended that EPA establish a 
more extensive list of permissible data sources and methods for 
determining short-term emissions rates. For calculating short-term 
actual emission rates where CEMS data are available, WESTAR recommended 
that the menu include, with no implications of a hierarchy:
     Use short-term maximum emissions for the entire plant over 
a 2-year period;
     Determine maximum short-term emissions from each source at 
the facility;
     Determine short-term emission rates and sort them, then 
determine representative rates, such as an upper percentile, as the 
single short-term emission rate for modeling;
     Use CEMS data to determine actual emissions as defined by 
rule and explained by EPA in the preamble to the 1980 PSD rule 
revisions; or
     Use hour-by-hour CEMS data in the model.

In situations where CEMS data are not available, WESTAR recommended 
that the menu for calculating short-term actual emission rates include, 
with no implications of a hierarchy:
     Average 2 years of actual annual emissions representing 
normal operations surrounding the baseline date and date of analysis 
for current emissions, and divide by annual operating hours;
     Calculate emissions from production data for the 2 years 
prior to the baseline date or date of analysis for current emissions 
(emissions calculated using valid emissions factors and methods);
     Use 2 years of emissions data, which may be before or 
after the baseline dates, which have a similar facility configuration 
that would be representative of baseline emissions; or
     Use of allowable emission rates, including use of 
regulatory limits, where appropriate.

We request comment on whether we should expand the proposed options for 
short-term emissions rate calculation to include elements from WESTAR's 
menu.
4. Use of Allowable Emissions Rates
    We have always allowed a reviewing authority or source to conduct a 
more conservative screening analysis using allowable emissions rates 
which are

[[Page 31391]]

typically higher than actual emissions rates. We propose to preserve 
that option under the new definition, but we are modifying the language 
from the prior definition slightly to make clear that we do not intend 
to mandate the use of allowable emissions, only to allow it at the 
discretion of the source or reviewing authority.
5. Emissions From a New or Modified Source
    When an increment consumption analysis is performed in the context 
of a pending permit application to demonstrate that a new or modified 
source will not cause or contribute to an exceedance of the increment, 
the analysis must include the emissions from the new or modified source 
when it begins operations after the permitted construction is complete. 
In the past, we have required such emissions to be based on the 
potential to emit of the new or modified source. However, in reforms to 
the NSR program completed in 2002, we allowed modified sources to use 
projected actual emissions in calculating whether the change resulted 
in a significant net increase in emissions. See 67 FR 80290 (December 
31, 2002). For the same reasons discussed in that rulemaking, we 
propose to adopt revised language for purposes of the increment 
consumption assessment that requires the use of projected actual 
emissions for a modified source. We propose to continue requiring the 
increment assessment to be based on the potential to emit of a new 
source that has not begun normal operations as of the date of the 
assessment.

C. What meteorological models and data should be used in increment 
consumption modeling?

    In addition to information on emissions from sources in the 
relevant area, one also needs meteorological data to evaluate 
consumption of the PSD increments. Meteorological data are a necessary 
input to the air quality dispersion models that are used to identify 
the change in concentration relative to a pollutant-specific baseline 
date. This change in concentration is then compared to the increments 
to demonstrate compliance. Adequate and appropriate meteorological data 
are a critical input for dispersion models \13\ in characterizing the 
state of the atmosphere in terms of the transport and diffusion of 
airborne pollutants within the modeling domain. Appendix W contains a 
list of meteorological data types and meteorological processors that 
are appropriate for various applications of preferred dispersion 
models.
---------------------------------------------------------------------------

    \13\ Dispersion models are mathematical formulations that 
describe the fundamental processes that occur in the atmosphere. 
These processes, for example, include emission, transport, and 
chemical reaction of pollutants.
---------------------------------------------------------------------------

    Recent experience with PSD increment modeling exercises has raised 
questions regarding the adequacy of the current EPA guidance to the 
States and regulated community concerning the appropriateness of 
certain types of meteorological data and the amount of data that should 
be obtained for certain dispersion model applications, including PSD 
increment analyses. We discuss these issues below in light of existing 
guidance, and seek comment on the need for modification and/or 
development of additional guidance.
1. Types of Meteorological Data and Processing
    Traditionally, dispersion model applications have utilized 
meteorological inputs derived from the direct processing of National 
Weather Service (NWS) observation data or meteorological data collected 
as part of a site-specific measurement program. However, prognostic 
meteorological models and other tools are available to project 
meteorological conditions in order to fill gaps in site-specific 
observational data. Recent experience suggests there may be a need for 
us to clarify the circumstances when it is permissible and appropriate 
to use meteorological data derived from prognostic meteorological 
models in dispersion model simulations such as a PSD increment 
consumption analysis.
    Prognostic meteorological models use fundamental equations of 
momentum, thermodynamics, and moisture to determine the evolution of 
specific meteorological variables from a given initial state. These 
models can characterize meteorological conditions at times and 
locations where observational data do not exist. Photochemical grid-
based air quality models, which require consistent input parameters 
distributed over an even grid in time and space, routinely utilize data 
output from prognostic meteorological models. Examples of prognostic 
meteorological models are:
     MM5--Penn State University/National Center for Atmospheric 
Research.
     WRF--Weather Research and Forecasting Model, NOAA/NCAR.
     RUC--Rapid Update Cycle, NOAA Rapid Refresh Development 
Group.
    In addition, diagnostic processors such as CALMET can format 
meteorological model output data for input into dispersion models. 
These diagnostic processors often can incorporate meteorological 
observation data into the process, resulting in a field of 
meteorological data that effectively blends the ground-truth of 
observations with the dynamics of the meteorological model. This data 
assimilation process frequently takes place within the prognostic 
meteorological models themselves. Run-time parameters may be set in the 
diagnostic processors to vary the influence observations may have on 
the resulting data set.
    Appendix W identifies criteria for judging the adequacy and 
appropriateness of such meteorological input data for dispersion 
modeling applications, including the spatial (i.e., space) and temporal 
(i.e., time) representativeness of the data for the specific 
application and the ability of the individual meteorological parameters 
selected to properly characterize the transport and diffusion 
conditions based on the formulations of a specific dispersion model. 
Meteorological data may be considered adequate and appropriate for a 
particular dispersion model or application, but that determination does 
not necessarily imply the adequacy and appropriateness of the data for 
other dispersion models or other applications of the same model. The 
proper judgment of adequacy and appropriateness of meteorological data 
requires expert knowledge of each of the main components--the 
meteorological observation data; the meteorological processor; and the 
dispersion model formulations and data requirements.
    Appendix W lists specific factors to consider when determining 
whether or not a set of meteorological data is representative for a 
particular dispersion model application. These include the proximity of 
the meteorological monitoring site to the area of interest, the 
complexity of the terrain in the area, the exposure of the 
meteorological monitoring site, and the period of data collected. 
Additional factors may be important depending on the requirements of 
specific models. For example, surface characteristics of the 
meteorological observation location, depending on land use and land 
cover characteristics, as well as terrain type and elevation, are 
required for input to AERMET, the meteorological processor for the 
AERMOD dispersion model.\14\

[[Page 31392]]

These surface characteristics have a significant impact on the boundary 
layer \15\ parameters that are required for input into the AERMOD 
model, and therefore have an impact on the resulting air quality 
results. The determination of representativeness for AERMOD therefore 
requires consideration of the potential impact of differences in 
surface characteristics between the meteorological monitoring site and 
the surface characteristics that generally describe the area upon which 
the air quality model simulation is focused.
---------------------------------------------------------------------------

    \14\ AERMOD is a steady-state plume dispersion model for 
assessment of pollutant concentrations from a variety of sources. 
AERMOD simulates transport and dispersion from multiple point, area, 
or volume sources based on an up-to-date characterization of the 
atmospheric boundary layer. Sources may be located in rural or urban 
areas, and receptors may be located in simple or complex terrain. 
AERMOD accounts for building wake effects (i.e., plume downwash) 
based on the PRIME building downwash algorithms. The model employs 
hourly sequential preprocessed meteorological data to estimate 
concentrations for averaging times from 1 hour to 1 year (also 
multiple years). AERMOD is designed to operate in concert with two 
pre-processor codes: AERMET processes meteorological data for input 
to AERMOD, and AERMAP processes terrain elevation data and generates 
receptor information for input to AERMOD.
    \15\ The boundary layer is the layer of the atmosphere closest 
to the Earth's surface.
---------------------------------------------------------------------------

    For long-range transport modeling assessments or assessments 
involving complex winds that require non-steady-state dispersion 
modeling \16\ appendix W allows, and in fact encourages, the use of 
prognostic mesoscale \17\ meteorological models to provide input data 
into dispersion model simulations. See 40 CFR part 51, appendix W, 
paragraph 8.3(d). However, proper use of output from these prognostic 
meteorological models in dispersion model applications requires expert 
judgment, and acceptance of such data is contingent on the concurrence 
of the appropriate reviewing authorities. Appendix W further indicates 
that mesoscale meteorological fields should be used in conjunction with 
available NWS or comparable meteorological observations within and near 
the modeling domain.
---------------------------------------------------------------------------

    \16\ Non-steady-state dispersion modeling is the one that 
accounts for spatial and temporal variability in meteorological 
parameters.
    \17\ Mesoscale is the meteorological phenomena with a horizontal 
extent from a few to several hundred kilometers.
---------------------------------------------------------------------------

    In this action, we are proposing to provide additional guidelines 
for determining the appropriateness of prognostic meteorological model 
output data for use in dispersion models. We propose that a 
determination of appropriateness would involve a process equal in rigor 
to that already used to review prognostic meteorological model output 
data for use in photochemical grid modeling applications at the 
regional scale. We believe that our existing guidance for ozone, 
PM2.5, and regional haze SIP modeling provides a useful 
basis for the process by which the State may allow use of certain data 
sets created by prognostic meteorological models as input into 
dispersion model applications provided these data sets are determined, 
by using this process, to be appropriate. Currently, acceptable quality 
of meteorological inputs derived from prognostic meteorological models 
would be demonstrated by statistical comparison of the prognostic model 
output to observations for key meteorological parameters, which may 
include temperature, water vapor mixing ratio, wind speed and direction 
(surface-level and aloft), clouds/radiation, precipitation, and the 
depth and evolution of vertical mixing. Identification of key 
meteorological parameters may depend on the type of model and the 
temporal and spatial scale of the application.
    When making a determination of the representativeness of 
meteorological inputs derived from prognostic models, it is important 
to consider the influences of observations both in the meteorological 
model and in any subsequent processing of the prognostic model outputs 
when comparing the output to observations as part of the evaluation. 
For example, a portion of the meteorological observations may be set 
aside (i.e., not used in the data assimilation process) for evaluation 
purposes. However, it is important to emphasize that a statistical 
comparison of the meteorological observation data to the output of the 
diagnostic processor, or even of the prognostic meteorological models, 
can only be one part of any determination of appropriateness. A 
phenomenological evaluation, a generally qualitative comparison focused 
on the specific meteorological phenomena of importance to a specific 
application, can be used together with the more quantitative 
comparisons of specific parameters to provide a more complete 
assessment of the representativeness of meteorological data. Additional 
technical factors that may need to be considered in the determination 
of appropriateness include:
     Selection of geographic domains and time periods;
     Influence of boundary and initial conditions;
     Technical options governing the meteorological model 
calculations; and
     Data assimilation parameters.

Guidance for consideration of these factors can be found in ``Guidance 
on the Use of Models and Other Analyses for Demonstrating Attainment of 
Air Quality Goals for Ozone, PM2.5, and Regional Haze,'' 
draft version 3.2, September 2006 \18\ (referred to hereafter as ``the 
Draft Guidance''). However, this guidance concerns regional-scale 
photochemical grid model applications. We request comment on how these 
and other factors may be considered in a determination of 
appropriateness of meteorological data derived from prognostic 
meteorological models for use in dispersion modeling applications. As 
explained in the Draft Guidance, regional-scale photochemical grid 
model applications require the above factors to be considered with 
regard to prognostic meteorological model output, and additionally 
require consideration of other factors specific to photochemical grid 
modeling.
---------------------------------------------------------------------------

    \18\ Available at http://www.epa.gov/scram001/guidance/guide/draft_final-pm-O3-RH.pdf.
---------------------------------------------------------------------------

    While meteorological model input that has been accepted for use in 
photochemical grid modeling may generally be acceptable for application 
in dispersion modeling inasmuch as the specifics of the meteorological 
model simulation are concerned, there are additional factors specific 
to dispersion modeling that must be considered. For example, the 
particular portion of the meteorological model output used in 
dispersion modeling must be considered in terms of its appropriateness 
for that particular dispersion model. Keeping in mind that the grid 
model is designed to produce a consistent set of parameters covering a 
large geographic area, we must consider the effects of extracting a few 
geographic points, from as few as only one grid cell in the entire 
model domain, and applying that very small subset of data from a 
greater dataset that was designed to be used in total.
    For example, meteorological model simulations are influenced by 
input data assigned to the boundary grid cells in the domain (i.e., 
boundary conditions) as well as to all grid cells within the domain at 
the initial time step (i.e., initial conditions). There are appropriate 
techniques that may be applied to model simulations to substantially 
reduce the influence of initial and boundary conditions for 
photochemical grid modeling.
    Boundary conditions, however, are incorporated into the 
meteorological model at each time step, and therefore the effect of the 
boundary conditions is evident throughout the meteorological model 
simulation. To reduce the effect of these assigned boundary conditions, 
we propose the area of interest be selected from an area substantially 
within the model simulation domain, for example, at least six grid 
cells from the boundary. We also propose to include in any review, a 
thorough

[[Page 31393]]

description of the techniques used to extract data from a larger grid, 
even if the meteorological data have been approved for use in a 
photochemical grid model application, if the extraction is performed 
using a tool or technique not listed in appendix W as part of a 
preferred modeling system.
2. Years of Meteorological Data
    In addition to clarifying the process and guidance for determining 
the circumstances under which it may be appropriate to input data from 
prognostic meteorological models into dispersion modeling, we believe 
it is also necessary to clarify guidance on the number of years of 
prognostic meteorological model output data that are necessary for a 
representative dispersion model simulation. With respect to the number 
of years of meteorological observation data that should be used for 
dispersion modeling, appendix W currently states the following:
     Five years of representative NWS meteorological 
observation data are required--the most recent, readily available 5-
year period is preferred.
     At least 1 year of site-specific meteorological data is 
required--as many as 5 years are preferred.
    See 40 CFR part 51, appendix W, paragraph 8.3.1.2(a). However, with 
respect to prognostic meteorological data, appendix W states that for 
long-range transport modeling and for other assessments involving non-
steady-state dispersion modeling to account for complex flows, less 
than 5, but at least 3, years of data from prognostic meteorological 
models may be used, and that the years need not be consecutive. See 
paragraph 8.3.1.2(d). We believe that our current guidance provides 
adequate discretion to the State to determine which and how many years 
(but no less than 3 years) should be used with regard to meteorological 
model output appropriate for the dispersion model application. 
Consistent with appendix W, this approach is integrated with the 
process described in the preceding section for determining 
appropriateness of prognostic meteorological model output. When a State 
is developing a set of data years for dispersion modeling, we propose 
to allow the State to consider any data years that it has determined to 
be appropriate using the process described above even if those data 
years were not produced by the same exact meteorological model 
configuration and simulation. However, we also propose that the State 
must further determine that a particular set of data years can be 
modeled to produce an appropriate depiction of the air quality issue at 
hand.
3. Evaluating the Appropriateness of Data Years From Prognostic 
Meteorological Models for Modeling Worst-Case Impacts
    For applications in which the modeling approach is designed to 
model worst-case impacts, we propose that the State should determine 
whether or not a set of years is appropriate based upon meteorological/
climatological representativeness, and additionally determine whether 
or not that set of years is appropriate to simulate the worst-case 
conditions required of the application. Keeping in mind worst-case 
conditions might not be discernable until simulated through a 
dispersion model, the term ``worst-case'' does not describe a set of 
worst-case meteorology, but rather a set of meteorology that when 
modeled, produces a worst-case depiction of air quality. This 
relationship may not be apparent on simple inspection of only the 
meteorological data set.
    That a particular data set sufficiently represents the 
meteorological observations for a given area for a given time period, 
based upon statistical analyses, may not be proof enough to determine 
that the particular data set is most appropriate for a dispersion 
application, especially when conducting worst-case applications. 
Additionally, a set of prognostic meteorological model output might be 
appropriate for dispersion modeling generally, but the portion of the 
data extracted for the specific dispersion model application should 
still be examined for appropriateness. While we do not explicitly 
propose a three-step process for determining appropriateness, these 
three individual examinations--appropriateness of the prognostic 
meteorological model output in general, appropriateness (meteorological 
representativeness) of the extracted data set, and appropriateness of 
the data set for the dispersion model application--are each a necessary 
part of the overall determination of appropriateness, especially in 
replacing data years of processed meteorological observations. Of 
course, once a particular data set/subset is determined appropriate, we 
do not anticipate re-examining that data set for use in other 
dispersion modeling provided the modeling applications and modeling 
domains are similar.
    We request comment on continuing the current path, based upon 
appendix W's guidance that previous years of meteorological data which 
have been used as the basis for permit emission limitations should be 
added to any subsequent period of meteorological data used for 
dispersion modeling. See 40 CFR part 51, appendix W, paragraph 
8.3.1.2(c). We will also accept comments on alternative methods for 
determining appropriate years of meteorological data including the use 
of data sets of processed observations, prognostic meteorological model 
output, or combinations of both.

D. What are my documentation and data and software availability 
requirements?

    Appendix W currently provides recommendations (see paragraph 3.1.1) 
regarding documentation and software availability for preferred 
modeling techniques that are listed in appendix W. (The preferred 
models are found in appendix A to appendix W, and are sometimes 
referred to as ``Appendix A models.'') The purpose of these 
recommendations includes fostering consistency in the application of 
dispersion models, minimizing the burden on applicants related to 
acquiring and setting up modeling applications, and providing 
transparency regarding model formulations, model performance, and model 
input requirements. These appendix W recommendations regarding 
documentation and software availability for preferred modeling 
techniques include that the ``model and its code cannot be 
proprietary.'' See paragraph 3.1.1(b)(vi) of appendix W.
    Application of the non-proprietary requirement to data developed 
for input into or use by a preferred model, or to other software used 
to process input data for a preferred model, is not explicitly 
addressed in appendix W. However, a strict requirement to be non-
proprietary is currently not applied to alternative models (paragraph 
3.2) that may be selected for use on a case-by-case basis, subject to 
the approval of the appropriate reviewing authority. Rather, the focus 
of recommendations related to the use of alternative models is on a 
demonstration and documentation of model performance that is equivalent 
or superior to the preferred model and, for cases where there is no 
preferred model, a scientific peer review and documentation and 
demonstration of the theoretical basis for the applicability of the 
alternative model. In addition, proprietary software interfaces to 
simplify the setup and analysis of Appendix A models have been 
developed by several commercial vendors, and have been in common usage 
for more than a decade. Such commercial software interfaces have not 
been subjected to a requirement to make the proprietary code available 
to the

[[Page 31394]]

public or the reviewing authority. However, demonstrations of 
equivalency may be, and have been, required of such proprietary 
interfaces, in keeping with paragraph 3.2.2(c) of appendix W.
    With technical advances and the increased use of more sophisticated 
methodologies for developing the required meteorological inputs for 
preferred modeling techniques, and in particular the use of prognostic 
meteorological model outputs in the development of spatial and 
temporally varying meteorology for long-range transport modeling 
applications with the preferred CALPUFF model, it is appropriate to 
address the adequacy and appropriateness of existing guidance for these 
emerging modeling technologies. Given the critical impact that the 
processed meteorological data have on such modeling applications, basic 
requirements for technical documentation and performance demonstration 
are certainly necessary. However, we believe that the existing guidance 
provided for alternative modeling techniques adequately addresses these 
concerns. The existing guidance implies a certain discretion and 
latitude for the reviewing authority in defining the specific data and 
documentation requirements necessary to make its determination of the 
acceptability of an alternative modeling technique for a given 
application. However, such requirements should be technically 
appropriate and avoid imposing an unnecessary burden on the applicant. 
In the case of meteorological data inputs for dispersion models, many 
of the relevant issues and requirements for such data are also 
discussed above in section IV.C of this preamble.
    In the special case of proprietary data that may be used in the 
development of model inputs, we believe that it is currently within the 
discretion of the State to require some independent review of the 
proprietary data by an oversight agency, if such a review is deemed 
critical to the overall assessment of the appropriateness of data for a 
particular modeling application. Another option within the discretion 
of the State would be for the State itself to conduct the review, 
provided that proprietary information and trade secrets are protected 
under a system that is equivalent to EPA's rules for requesting non-
disclosure of Confidential Business Information (CBI) submitted to the 
Agency. See 40 CFR part 2. Provided that any appropriate and necessary 
reviews can be conducted by an independent body or the State reviewing 
authority with protection against disclosure of CBI, we do not believe 
it is necessary to require such proprietary data to be made available 
to the general public or to wholly preclude reliance on the data in 
regulatory modeling applications.
    In the case of software, the focus of the determination of 
acceptability by the reviewing authority should be on the adequacy of 
the technical documentation and performance demonstrations that are 
required to support the use of such software. More specifically in the 
case of proprietary software, the reproducibility of the data or model 
simulation may be an important component of the documentation to ensure 
confidence in the modeling results, and the applicant should facilitate 
such a demonstration when required. Additional documentation regarding 
the quality assurance procedures used in the development of the 
proprietary software may also be relevant to supporting the integrity 
and accuracy of the results.
    We believe that the current text of appendix W adequately defines 
the documentation and software availability requirements related to 
both preferred and alternative modeling techniques. We request comment 
on whether additional guidance is needed to clarify these requirements 
as they apply to the use of proprietary software and/or data to develop 
input for an Appendix A modeling application for PSD increment 
consumption.

VI. Implementation Issues

A. Is there a need for States to make revisions to their SIPs?

    As described in this notice, with these regulations we are 
proposing to refine certain aspects of PSD increment analyses to 
provide greater clarity to States and regulated sources on how to 
calculate increases in concentrations for purposes of determining 
compliance with the PSD increments. Once we finalize these proposed 
regulations, we intend to encourage States to incorporate them for the 
sake of consistency and clarity, and to make their SIPs consistent with 
the proposed rule amendments. This would be a relatively easy task 
given that SIP changes resulting from other upcoming NSR rulemakings 
(e.g., rules for electric generating units (EGUs); corn milling; 
potential to emit (PTE); and aggregation, debottlenecking, and project 
netting) will likely be required in roughly the same time period. 
However, we believe that SIP changes would not necessarily be required 
in order for reviewing authorities to begin conducting PSD increment 
analyses consistent with these regulations because EPA's prior 
recommendations have not been binding on States. We are specifically 
seeking comment on the need for SIP revisions or any viable 
alternatives for implementing the changes for these proposed increment 
analysis provisions.

B. When would these policies be put into effect?

    We propose to make the proposed regulations effective 60 days from 
promulgation.

VII. Statutory and Executive Order Reviews

A. Executive Order 12866: Regulatory Planning and Review

    Under Executive Order (EO) 12866 (58 FR 51735, October 4, 1993), 
this action is a ``significant regulatory action'' because it is likely 
to raise novel legal or policy issues arising out of legal mandates, 
the President's priorities, or the principles set forth in the 
Executive Order. Accordingly, EPA submitted this action to the Office 
of Management and Budget (OMB) for review under EO 12866 and any 
changes made in response to OMB recommendations have been documented in 
the docket for this action.

B. Paperwork Reduction Act

    This action does not impose any new information collection burden. 
We are not proposing any new paperwork requirements (e.g., monitoring, 
reporting, recordkeeping) as part of this action. Although we are 
refining our existing regulations and policy on the analysis of PSD 
increment consumption, the proposed regulations do not contain new 
paperwork requirements for permit applicants or reviewing authorities. 
The PSD increment analysis is already required under existing EPA 
regulations. The OMB has previously approved the information collection 
requirements contained in the existing PSD program regulations (40 CFR 
51.166 and 52.21) under the provisions of the Paperwork Reduction Act, 
44 U.S.C. 3501 et seq. and has assigned OMB control number 2060-0003, 
EPA ICR number 1230.17. A copy of the OMB approved Information 
Collection Request (ICR) may be obtained from Susan Auby, Collection 
Strategies Division; U.S. Environmental Protection Agency (2822T); 1200 
Pennsylvania Ave., NW., Washington, DC 20460 or by calling (202) 566-
1672.
    Burden means the total time, effort, or financial resources 
expended by persons to generate, maintain, retain, or disclose or 
provide information to or for a Federal agency. This includes the time 
needed to review instructions; develop,

[[Page 31395]]

acquire, install, and utilize technology and systems for the purposes 
of collecting, validating, and verifying information, processing and 
maintaining information, and disclosing and providing information; 
adjust the existing ways to comply with any previously applicable 
instructions and requirements; train personnel to be able to respond to 
a collection of information; search data sources; complete and review 
the collection of information; and transmit or otherwise disclose the 
information.
    An agency may not conduct or sponsor, and a person is not required 
to respond to a collection of information unless it displays a 
currently valid OMB control number. The OMB control numbers for EPA's 
regulations in 40 CFR are listed in 40 CFR part 9.

C. Regulatory Flexibility Analysis

    The Regulatory Flexibility Act (RFA) generally requires an agency 
to prepare a regulatory flexibility analysis of any rule subject to 
notice and comment rulemaking requirements under the Administrative 
Procedure Act or any other statute unless the agency certifies that the 
rule will not have a significant economic impact on a substantial 
number of small entities. Small entities include small businesses, 
small organizations, and small governmental jurisdictions.
    For purposes of assessing the impacts of this action on small 
entities, small entity is defined as: (1) A small business as defined 
by the Small Business Administration's (SBA) regulations at 13 CFR 
121.201; (2) a small governmental jurisdiction that is a government of 
a city, county, town, school district or special district with a 
population of less than 50,000; and (3) a small organization that is 
any not-for-profit enterprise which is independently owned and operated 
and is not dominant in its field.
    After considering the economic impacts of this action on small 
entities, I certify that this action will not have a significant 
economic impact on a substantial number of small entities. This action 
will not impose any new requirements on small entities. The increment 
consumption analysis is already required under existing PSD regulations 
and the proposed refinements to our existing regulations and policy are 
not expected to increase the economic impact of this analysis on 
regulated entities. We continue to be interested in the potential 
impacts of the proposed rule on small entities and welcome comments on 
issues related to such impacts.

D. Unfunded Mandates Reform Act

    Title II of the Unfunded Mandates Reform Act of 1995 (UMRA), Public 
Law 104-4, establishes requirements for Federal agencies to assess the 
effects of their regulatory actions on State, local, and tribal 
governments and the private sector. Under section 202 of the UMRA, EPA 
generally must prepare a written statement, including a cost-benefit 
analysis, for proposed and final rules with ``Federal mandates'' that 
may result in expenditures to State, local, and tribal governments, in 
the aggregate, or to the private sector, of $100 million or more in any 
1 year. Before promulgating an EPA rule for which a written statement 
is needed, section 205 of the UMRA generally requires EPA to identify 
and consider a reasonable number of regulatory alternatives and adopt 
the least costly, most cost-effective or least burdensome alternative 
that achieves the objectives of the rule. The provisions of section 205 
do not apply when they are inconsistent with applicable law. Moreover, 
section 205 allows EPA to adopt an alternative other than the least 
costly, most cost-effective or least burdensome alternative if the 
Administrator publishes with the final rule an explanation as to why 
that alternative was not adopted. Before EPA establishes any regulatory 
requirements that may significantly or uniquely affect small 
governments, including tribal governments, it must have developed under 
section 203 of the UMRA a small government agency plan. The plan must 
provide for notifying potentially affected small governments, enabling 
officials of affected small governments to have meaningful and timely 
input in the development of EPA regulatory proposals with significant 
Federal intergovernmental mandates, and informing, educating, and 
advising small governments on compliance with the regulatory 
requirements.
    This proposed action contains no Federal mandates (under the 
regulatory provisions of Title II of the UMRA) for State, local, or 
tribal governments or the private sector. The PSD increment consumption 
analysis is already required under existing regulations. In this 
rulemaking, we are only proposing to refine our existing regulations 
and policy on how this analysis may be conducted and are not imposing 
any additional analytical requirements. Thus, this action is not 
subject to the requirements of sections 202 and 205 of the UMRA.
    In addition, we have determined that this rule contains no 
regulatory requirements that might significantly or uniquely affect 
small governments. As discussed above, this proposal would not impose 
any new requirements on small governments.

E. Executive Order 13132--Federalism

    Executive Order 13132, entitled ``Federalism'' (64 FR 43255, August 
10, 1999), requires EPA to develop an accountable process to ensure 
``meaningful and timely input by State and local officials in the 
development of regulatory policies that have federalism implications.'' 
``Policies that have federalism implications'' is defined in the 
Executive Order to include regulations that have ``substantial direct 
effects on the States, on the relationship between the national 
government and the States, or on the distribution of power and 
responsibilities among the various levels of government.''
    Under section 6(b) of Executive Order 13132, EPA may not issue a 
regulation that has federalism implications, that imposes substantial 
direct compliance costs, and that is not required by statute, unless 
the Federal government provides the funds necessary to pay the direct 
compliance costs incurred by State and local governments, or EPA 
consults with State and local officials early in the process of 
developing the proposed regulation. Under section 6(c) of Executive 
Order 13132, EPA may not issue a regulation that has federalism 
implications and that preempts State law, unless the Agency consults 
with State and local officials early in the process of developing the 
proposed regulation.
    The EPA has concluded that this proposed rule may have federalism 
implications. The proposed rule establishes Federal standards for the 
administration of the PSD program by State reviewing authorities. 
However, the proposed rule does not impose additional requirements on 
State reviewing authorities because a PSD increment analysis is already 
required under existing regulations. In addition, EPA proposes in this 
action to make clear that States have discretion to use their best 
judgment in conducting elements of the increment consumption analysis. 
Thus, this rule will not impose substantial direct compliance costs on 
State or local governments, nor will it preempt State law. Thus, the 
requirements of sections 6(b) and 6(c) of the Executive Order do not 
apply to this rule.
    Consistent with EPA policy, EPA nonetheless consulted with several 
State officials and representatives of State governments early in the 
process of developing the proposed regulation to permit them to have 
meaningful and timely input into its development. As

[[Page 31396]]

discussed above, this proposal has been informed by the recommendations 
of the Western States Air Resources Council (WESTAR) PSD Reform 
Workgroup, which is an organization that includes State officials who 
have sought greater clarity in methodologies for evaluating consumption 
of the PSD increment. In addition, EPA has also been consulting for 
several years with State officials in North Dakota about the parameters 
for the increment consumption analysis.
    In the spirit of Executive Order 13132, and consistent with EPA 
policy to promote communications between EPA and State and local 
governments, EPA specifically solicits comment on this proposed rule 
from State and local officials.

F. Executive Order 13175--Consultation and Coordination With Indian 
Tribal Governments

    Executive Order 13175, entitled ``Consultation and Coordination 
with Indian Tribal Governments'' (65 FR 13175, November 9, 2000), 
requires EPA to develop an accountable process to ensure ``meaningful 
and timely input by tribal officials in the development of regulatory 
policies that have tribal implications.'' The EPA has concluded that 
this proposed rule may have tribal implications. However, it will 
neither impose substantial direct compliance costs on tribal 
governments, nor preempt Tribal law.
    By refining our existing regulations and policy, this proposal may 
affect how reviewing authorities determine increment consumption on the 
tribal lands that have been redesignated to Class I or are in the 
process of being redesignated to Class I. For that reason, EPA will 
provide an opportunity for meaningful and timely involvement in this 
action by consulting, during the period between proposal and 
promulgation, with tribal officials from the six Tribes whose 
reservations have been redesignated from Class II to Class I or are in 
the process of being so redesignated. In addition, EPA specifically 
solicits additional comment on this proposed rule from all tribal 
officials.

G. Executive Order 13045--Protection of Children From Environmental 
Health and Safety Risks

    Executive Order 13045, entitled ``Protection of Children from 
Environmental Health Risks and Safety Risks'' (62 FR 19885, April 23, 
1997), applies to any rule that: (1) Is determined to be ``economically 
significant'' as defined under Executive Order 12866; and (2) concerns 
an environmental health or safety risk that EPA has reason to believe 
may have a disproportionate effect on children. If the regulatory 
action meets both criteria, the Agency must evaluate the environmental 
health or safety effects of the planned rule on children, and explain 
why the planned regulation is preferable to other potentially effective 
and reasonably feasible alternatives considered by the Agency.
    This proposed rule is not subject to the Executive Order because it 
is not economically significant as defined in Executive Order 12866, 
and because the Agency does not have reason to believe the 
environmental health or safety risks addressed by this action present a 
disproportionate risk to children. The proposed rule does not impose 
any new regulatory or analytical requirements, but simply refines 
existing regulations and policy with respect to the PSD increment 
consumption analysis that is currently required. The public is invited 
to submit or identify peer-reviewed studies and data, of which the 
Agency may not be aware, that may be pertinent to the effect of this 
proposed rule on children.

H. Executive Order 13211--Actions That Significantly Affect Energy 
Supply, Distribution, or Use

    This rule is not a ``significant energy action'' as defined in 
Executive Order 13211, ``Actions Concerning Regulations That 
Significantly Affect Energy Supply, Distribution, or Use'' (66 FR 28355 
(May 22, 2001)) because it is not likely to have a significant adverse 
effect on the supply, distribution, or use of energy. Further, we have 
concluded that this rule is not likely to have any adverse energy 
effects because it does not impose any new requirements on sources that 
supply, distribute, or use energy. The proposed rule does not establish 
additional regulatory or analytical requirements, but simply refines 
existing regulations and policy with respect to the PSD increment 
consumption analysis that is currently required.

I. Executive Order 12898--Federal Actions To Address Environmental 
Justice in Minority Populations and Low-Income Populations

    Executive Order (EO) 12898 (59 FR 7629 (Feb. 16, 1994)) establishes 
federal executive policy on environmental justice. Its main provision 
directs Federal agencies, to the greatest extent practicable and 
permitted by law, to make environmental justice part of their mission 
by identifying and addressing, as appropriate, disproportionate high 
and adverse human health or environmental effects of its programs, 
policies, and activities on minorities and low-income populations in 
the United States.
    The EPA has determined that this proposed rule would not have 
disproportionately high and adverse human health or environmental 
effects on minority or low-income populations because it does not 
affect the level of protection provided to human health or the 
environment. The proposed rule does not establish or eliminate 
regulatory or analytical requirements, but simply refines existing 
regulations and policy with respect to the PSD increment consumption 
analysis that is currently required.

J. National Technology Transfer and Advancement Act

    Section 12(d) of the National Technology Transfer and Advancement 
Act of 1995 (NTTAA), Public Law 104-113, 12(d) (15 U.S.C. 272 note) 
directs EPA to use voluntary consensus standards in its regulatory 
activities unless to do so would be inconsistent with applicable law or 
otherwise impractical. Voluntary consensus standards are technical 
standards (for example, materials specifications, test methods, 
sampling procedures, and business practices) that are developed or 
adopted by voluntary consensus standards bodies. The NTTAA directs EPA 
to provide Congress, through OMB, explanations when the Agency decides 
not to use available and applicable voluntary consensus standards.
    This proposed rulemaking does not involve technical standards. 
Therefore, EPA did not consider the use of any voluntary consensus 
standards.

VIII. Statutory Authority

    The statutory authority for this action is provided by sections 
163, 166, 169(4), and 301(a) of the Act as amended (42 U.S.C. 7473, 
7476, 7479(4), and 7601(a)). This notice is also subject to section 
307(d) of the CAA (42 U.S.C. 7607(d)).

List of Subjects

40 CFR Part 51

    Environmental protection, Administrative practice and procedures, 
Air pollution control, Intergovernmental relations.

40 CFR Part 52

    Environmental protection, Administrative practice and procedures, 
Air pollution control, Intergovernmental relations.


[[Page 31397]]


    Dated: May 24, 2007.
Stephen L. Johnson,
Administrator.
    For the reasons stated in the preamble, title 40, chapter I of the 
Code of Federal Regulations is proposed to be amended as set forth 
below.

PART 51--[AMENDED]

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

    Authority: 23 U.S.C. 101; 42 U.S.C. 7401-7671q.

Subpart I--[Amended]

    2. Section 51.166 is amended as follows:
    a. By revising paragraph (b)(13);
    b. By revising paragraph (b)(21)(i);
    c. By revising paragraph (f);
    d. By removing from paragraph (p)(5)(i) the cross reference to 
``(q)(4)'' and adding in its place ``(p)(4)'';
    e. By removing from paragraphs (p)(5)(iii) and (p)(6)(iii) the 
cross reference to ``(q)(7)'' and adding in its place ``(p)(7)''; and
    f. By removing from paragraph (p)(7) the cross reference to 
``(q)(5) or (6)'' and adding in its place ``(p)(5) or (6)''.
    The revisions read as follows:


Sec.  51.166  Prevention of significant deterioration of air quality.

* * * * *
    (b) * * *
    (13)(i) Baseline concentration means that ambient concentration 
level that exists in the baseline area at the time of the applicable 
minor source baseline date. A baseline concentration is determined for 
each pollutant for which a minor source baseline date is established 
and shall include:
    (a) The actual emissions, as defined in paragraph (f)(1) of this 
section, representative of sources in existence on the applicable minor 
source baseline date, except as provided in paragraph (b)(13)(ii) of 
this section; and
    (b) The allowable emissions of major stationary sources that 
commenced construction before the major source baseline date, but were 
not in operation by the applicable minor source baseline date.
    (ii) The following will not be included in the baseline 
concentration and will affect the applicable maximum allowable 
increase(s):
    (a) Actual emissions, as defined in paragraph (f)(1) of this 
section, from any major stationary source on which construction 
commenced after the major source baseline date; and
    (b) Actual emissions increases and decreases, as defined in 
paragraph (f)(1) of this section, at any source (including stationary, 
mobile, and area sources) occurring after the minor source baseline 
date.
* * * * *
    (21)(i) Actual emissions means the actual rate of emissions of a 
regulated NSR pollutant from an emissions unit, as determined in 
accordance with paragraphs (b)(21)(ii) through (iv) of this section, 
except that this definition shall not apply for calculating whether a 
significant emissions increase has occurred, for establishing a PAL 
under paragraph (w) of this section, or for determining consumption of 
ambient air increments. Instead, paragraphs (b)(40), (b)(47), and 
(f)(1) of this section shall apply for those purposes.
* * * * *
    (f) Methods for determining increment consumption.
    (1) Actual emissions. For purposes of determining consumption of 
the ambient air increments set forth in paragraph (c) of this section, 
the plan shall define ``actual emissions'' in accordance with 
paragraphs (f)(1)(i) through (vii) of this section.
    (i) Actual emissions shall be calculated based on information that, 
in the judgment of the reviewing authority, provides the most reliable, 
consistent, and representative indication of the emissions from a unit 
or group of units in an increment consumption analysis as of the 
baseline date and on subsequent dates. In general, actual emissions for 
a specific unit should be calculated using the unit's actual operating 
hours, production rates, and types of materials processed, stored, or 
combusted during the selected time period. However, where records of 
actual operating hours, production rates, and composition of materials 
are not available or are incomplete, the reviewing authority shall use 
its best professional judgment to estimate these parameters from 
available information in accordance with the criteria in this 
paragraph. When available and consistent with the criteria in this 
paragraph, data from continuous emissions monitoring systems may be 
used.
    (ii) In general, when evaluating consumption of an increment 
averaged over an annual time period, actual emissions as of a 
particular date in an increment consumption analysis (the applicable 
baseline date or the current time period) shall equal the average rate, 
in tons per year, at which the unit actually emitted the pollutant 
during a consecutive 24-month period which precedes the particular date 
and which is representative of normal source operation.
    (iii) When evaluating consumption of an increment averaged over a 
period of less than 1 year (i.e., 24-hour or 3-hour averaging), actual 
emissions as of a particular date in an increment consumption analysis 
(the applicable baseline date or the current time) may equal the 
average rate, for the applicable averaging time, at which the unit 
actually emitted the pollutant during a consecutive 24-month period 
which precedes the particular date. The average rate may be calculated 
by dividing an annual rate by the number of hours the unit was actually 
operating over the annual period. The reviewing authority may use an 
actual maximum rate over a 24-month period when sufficient data are 
available to produce a consistent, reliable, and representative 
analysis of the change in emissions from baseline to the current time 
period.
    (iv) The reviewing authority may allow actual emissions to be based 
on a different time period than the 24 months preceding a particular 
date upon a determination that such period is more representative of 
normal source operation as of the particular date, based upon credible 
information showing that the unit's operations in the 24 months 
preceding the date were not typical of operations as of the particular 
date. A period after the particular date may be used, but only if such 
period is more representative of normal source operations as of the 
particular date. Operations occurring prior to a particular date are 
not representative of normal source operations for a particular date if 
they permanently ceased more than 24 months prior to that date. The 
different time period shall be a consecutive 24-month period unless two 
non-consecutive 12-month periods are demonstrated to be more 
representative of normal source operation as described above.
    (v) The reviewing authority may use source-specific allowable 
emissions for the unit instead of the actual emissions of the unit.
    (vi) For any modified emissions unit that has not resumed normal 
operations on the date of an increment consumption analysis, the actual 
emissions on the date the source begins operation shall equal the 
projected actual emissions of the unit on that date. For any new 
emissions unit that has not begun normal operations on the date of an 
increment consumption analysis, the actual emissions on the date the 
new source begins operations shall equal the potential to emit for that 
source.
    (vii) To the extent any requirement of this paragraph (f)(1) 
conflicts with a recommendation in appendix W of this part, paragraph 
(f)(1) shall control.

[[Page 31398]]

    (2) Exclusions from increment consumption. (i) The plan may provide 
that the following concentrations shall be excluded in determining 
compliance with a maximum allowable increase:
    (a) Concentrations attributable to the increase in emissions from 
stationary sources which have converted from the use of petroleum 
products, natural gas, or both by reason of an order in effect under 
section 2(a) and (b) of the Energy Supply and Environmental 
Coordination Act of 1974 (or any superseding legislation) over the 
emissions from such sources before the effective date of such an order;
    (b) Concentrations attributable to the increase in emissions from 
sources which have converted from using natural gas by reason of 
natural gas curtailment plan in effect pursuant to the Federal Power 
Act over the emissions from such sources before the effective date of 
such plan;
    (c) Concentrations of particulate matter attributable to the 
increase in emissions from construction or other temporary emission-
related activities of new or modified sources;
    (d) The increase in concentrations attributable to new sources 
outside the United States over the concentrations attributable to 
existing sources which are included in the baseline concentration;
    (e) Concentrations attributable to the temporary increase in 
emissions of sulfur dioxide, particulate matter, or nitrogen oxides 
from stationary sources which are affected by plan revisions approved 
by the Administrator as meeting the criteria specified in paragraph 
(f)(2)(iii) of this section; and
    (f) Concentrations attributable to sources that obtained a permit 
based on a variance issued pursuant to paragraph (p)(4) of this 
section, but only with respect to the Class I increment in the area for 
which the variance was issued. Concentrations attributable to such 
sources shall continue to be included in determining compliance with 
the maximum allowable increase set forth in paragraphs (p)(4).
    (ii) If the plan provides that the concentrations to which 
paragraph (f)(2)(i)(a) or (b) of this section refers shall be excluded, 
it shall also provide that no exclusion of such concentrations shall 
apply more than 5 years after the effective date of the order to which 
paragraph (f)(2)(i)(a) of this section refers, or the plan to which 
paragraph (f)(2)(i)(b) of this section refers, whichever is applicable. 
If both such order and plan are applicable, no such exclusion shall 
apply more than 5 years after the later of such effective dates.
    (iii) For purposes of excluding concentrations pursuant to 
paragraph (f)(2)(i)(e) of this section, the Administrator may approve a 
plan revision that:
    (a) Specifies the time over which the temporary emissions increase 
of sulfur dioxide, particulate matter, or nitrogen oxides would occur. 
Such time is not to exceed 2 years in duration unless a longer time is 
approved by the Administrator.
    (b) Specifies that the time period for excluding certain 
contributions in accordance with paragraph (f)(2)(iii)(a) of this 
section, is not renewable;
    (c) Allows no emissions increase from a stationary source which 
would:
    (1) Impact a Class I area or an area where an applicable increment 
is known to be violated; or
    (2) Cause or contribute to the violation of a national ambient air 
quality standard;
    (d) Requires limitations to be in effect the end of the time period 
specified in accordance with paragraph (f)(2)(iii)(a) of this section, 
which would ensure that the emissions levels from stationary sources 
affected by the plan revision would not exceed those levels occurring 
from such sources before the plan revision was approved.
* * * * *

PART 52--[AMENDED]

    3. The authority citation for part 52 continues to read as follows:

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

Subpart A--[Amended]

    4. Section 52.21 is amended as follows:
    a. By revising paragraph (b)(13);
    b. By revising paragraph (b)(21)(i);
    c. By adding paragraph (f);
    d. By removing from paragraph (p)(6) the cross reference to 
``(q)(4)'' and adding in its place ``(p)(5)'';
    e. By removing from paragraphs (p)(6) and (p)(7) the cross 
reference to ``(q)(7)'' and adding in its place ``(p)(8)''; and
    f. By removing from paragraph (p)(8) the cross reference to 
``(q)(5) or (6)'' and adding in its place ``(p)(6) or (7)''.
    The addition and revisions read as follows:


Sec.  52.21  Prevention of significant deterioration of air quality.

* * * * *
    (b) * * *
    (13)(i) Baseline concentration means that ambient concentration 
level that exists in the baseline area at the time of the applicable 
minor source baseline date. A baseline concentration is determined for 
each pollutant for which a minor source baseline date is established 
and shall include:
    (a) The actual emissions, as defined in paragraph (f)(1) of this 
section, representative of sources in existence on the applicable minor 
source baseline date, except as provided in paragraph (b)(13)(ii) of 
this section; and
    (b) The allowable emissions of major stationary sources that 
commenced construction before the major source baseline date, but were 
not in operation by the applicable minor source baseline date.
    (ii) The following will not be included in the baseline 
concentration and will affect the applicable maximum allowable 
increase(s):
    (a) Actual emissions, as defined in paragraph (f)(1) of this 
section, from any major stationary source on which construction 
commenced after the major source baseline date; and
    (b) Actual emissions increases and decreases, as defined in 
paragraph (f)(1) of this section, at any source (including stationary, 
mobile, and area sources) occurring after the minor source baseline 
date.
* * * * *
    (21)(i) Actual emissions means the actual rate of emissions of a 
regulated NSR pollutant from an emissions unit, as determined in 
accordance with paragraphs (b)(21)(ii) through (iv) of this section, 
except that this definition shall not apply for calculating whether a 
significant emissions increase has occurred, for establishing a PAL 
under paragraph (aa) of this section, or for determining consumption of 
ambient air increments. Instead, paragraphs (b)(41), (b)(48), and 
(f)(1) of this section shall apply for those purposes.
* * * * *
    (f) Methods for determining increment consumption--(1) Actual 
emissions. For purposes of determining consumption of the ambient air 
increments set forth in paragraph (c) of this section, the term 
``actual emissions'' shall be defined in accordance with paragraphs 
(f)(1)(i) through (vii) of this section.
    (i) Actual emissions shall be calculated based on information that, 
in the judgment of the Administrator, provides the most reliable, 
consistent, and representative indication of the emissions from a unit 
or group of units in an increment consumption analysis as of the 
baseline date and on subsequent dates. In general, actual emissions for 
a specific unit should be calculated using the unit's actual operating 
hours, production rates, and types of materials processed, stored, or 
combusted during the selected time period. However, where records of

[[Page 31399]]

actual operating hours, production rates, and composition of materials 
are not available or are incomplete, the Administrator shall use his or 
her best professional judgment to estimate these parameters from 
available information in accordance with the criteria in this 
paragraph. When available and consistent with the criteria in this 
paragraph, data from continuous emissions monitoring systems may be 
used.
    (ii) In general, when evaluating consumption of an increment 
averaged over an annual time period, actual emissions as of a 
particular date in an increment consumption analysis (the applicable 
baseline date or the current time period) shall equal the average rate, 
in tons per year, at which the unit actually emitted the pollutant 
during a consecutive 24-month period which precedes the particular date 
and which is representative of normal source operation.
    (iii) When evaluating consumption of an increment averaged over a 
period of less than one year (i.e., 24-hour or 3-hour averaging), 
actual emissions as of a particular date in an increment consumption 
analysis (the applicable baseline date or the current time) may equal 
the average rate, for the applicable averaging time, at which the unit 
actually emitted the pollutant during a consecutive 24-month period 
which precedes the particular date. The average rate may be calculated 
by dividing an annual rate by the number of hours the unit was actually 
operating over the annual period. The Administrator may use an actual 
maximum rate over a 24-month period when sufficient data are available 
to produce a consistent, reliable, and representative analysis of the 
change in emissions from baseline to the current time period.
    (iv) The Administrator may allow actual emissions to be based on a 
different time period than the 24 months preceding a particular date 
upon a determination that such period is more representative of normal 
source operation as of the particular date, based upon credible 
information showing that the unit's operations in the 24 months 
preceding the date were not typical of operations as of the particular 
date. A period after the particular date may be used, but only if such 
period is more representative of normal source operations as of the 
particular date. Operations occurring prior to a particular date are 
not representative of normal source operations for a particular date if 
they permanently ceased more than 24 months prior to that date. The 
different time period shall be a consecutive 24-month period unless two 
non-consecutive 12-month periods are demonstrated to be more 
representative of normal source operation as described above.
    (v) The Administrator may use source-specific allowable emissions 
for the unit instead of the actual emissions of the unit.
    (vi) For any modified emissions unit that has not resumed normal 
operations on the date of an increment consumption analysis, the actual 
emissions on the date the source begins operation shall equal the 
projected actual emissions of the unit on that date. For any new 
emissions unit that has not begun normal operations on the date of an 
increment consumption analysis, the actual emissions on the date the 
new source begins operations shall equal the potential to emit for that 
source.
    (vii) To the extent any requirement of this paragraph (f)(1) 
conflicts with a recommendation in 40 CFR part 51, appendix W, 
paragraph (f)(1) shall control.
    (2) Exclusions from increment consumption. In determining 
compliance with the maximum allowable increase, the Administrator shall 
exclude concentrations attributable to sources that obtained a permit 
based on a variance issued pursuant to paragraphs (p)(5) of this 
section, but only with respect to the Class I increment in the area for 
which the variance was issued. Concentrations attributable to such 
sources shall continue to be included in determining compliance with 
the maximum allowable increases set forth in paragraph (p)(5).
* * * * *
[FR Doc. E7-10459 Filed 6-5-07; 8:45 am]
BILLING CODE 6560-50-P