[Federal Register Volume 83, Number 221 (Thursday, November 15, 2018)]
[Rules and Regulations]
[Pages 57324-57333]
From the Federal Register Online via the Government Publishing Office [www.gpo.gov]
[FR Doc No: 2018-24820]


=======================================================================
-----------------------------------------------------------------------

ENVIRONMENTAL PROTECTION AGENCY

40 CFR Parts 51 and 52

[EPA-HQ-OAR-2003-0064; FRL-9986-47-OAR]
RIN 2060-AP80


Prevention of Significant Deterioration (PSD) and Nonattainment 
New Source Review (NNSR): Aggregation; Reconsideration

AGENCY: Environmental Protection Agency (EPA).

ACTION: Final action; lifting of administrative stay and announcement 
of effective date.

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

SUMMARY: In this action, the Environmental Protection Agency (EPA) is 
concluding the reconsideration of an earlier action that the EPA 
published on January 15, 2009, titled ``Prevention of Significant 
Deterioration (PSD) and Nonattainment New Source Review (NSR): 
Aggregation and Project Netting.'' The 2009 action--hereafter referred 
to as ``2009 NSR Aggregation Action''--clarified implementation of the 
New Source Review (NSR) permitting program under the Clean Air Act (CAA 
or Act) with respect to treating related physical or operational 
changes as a single ``modification'' for the purpose of determining NSR 
applicability at a stationary source. On April 15, 2010, the EPA 
proposed to revoke the 2009 NSR Aggregation Action. After a review of 
the public comments received on that proposal, the EPA has now decided 
to not revoke the 2009 NSR Aggregation Action. The EPA is, therefore, 
retaining the interpretation set forth in the 2009 NSR Aggregation 
Action, while not adopting any changes to the relevant rule text. At 
the same time, the EPA is using this present action to clarify the 
implications of the 2009 NSR Aggregation Action for EPA-approved 
permitting programs. This action also lifts the administrative stay and 
announces the effective date of the 2009 NSR Aggregation Action.

DATES: This action is effective on November 15, 2018.

ADDRESSES: The EPA has established a docket for this action, identified 
by Docket ID No. EPA-HQ-OAR-2003-0064. All documents in the docket are 
listed in the http://www.regulations.gov website. Although listed in 
the index, some information may not be publicly available, e.g., 
Confidential Business Information 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. Publicly available docket materials are available 
electronically in http://www.regulations.gov.

FOR FURTHER INFORMATION CONTACT: For further general information on 
this action, contact Mr. Dave Svendsgaard, Office of Air Quality 
Planning and Standards (OAQPS), Air Quality Policy Division, U.S. EPA, 
Mail Code 504-03, 109 T.W. Alexander Drive, Research Triangle Park, NC 
27711; by telephone at (919) 541-2380; or by email at 
[email protected].

SUPPLEMENTARY INFORMATION: 

I. General Information

A. Does this action apply to me?

    Entities potentially affected directly by this action include 
sources in all industry categories. Entities potentially affected by 
this action also include state, local and tribal air pollution control 
agencies (air agencies) responsible for permitting sources pursuant to 
the NSR program.

B. 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 Federal Register document will be posted at https://www.epa.gov/nsr.

[[Page 57325]]

C. How is this document organized?

    The information presented in this document is organized as follows:

I. General Information
    A. Does this action apply to me?
    B. Where can I get a copy of this document and other related 
information?
    C. How is this document organized?
II. Background
    A. What is New Source Review?
    B. What is project aggregation?
    C. Regulatory History
III. This Action
    A. Overview
    B. Retaining the 2009 NSR Aggregation Action
    C. Completing the Reconsideration Proceeding
    D. Lifting the Administrative Stay; Announcement of Effective 
Date
IV. Environmental Justice Considerations
V. Executive Order 12866: Regulatory Planning and Review and 
Executive Order 13563: Improving Regulation and Regulatory Review
VI. Judicial Review
VII. Statutory Authority

II. Background

A. What is New Source Review?

    The NSR program is a preconstruction permitting program that 
requires certain stationary sources of air pollution to obtain permits 
prior to beginning construction. The NSR permitting program applies 
both to new construction and to modifications of existing sources, 
regardless of whether the source is in an area where the national 
ambient air quality standards (NAAQS) have been exceeded (nonattainment 
area) or if the source is in an area where the NAAQS have not been 
exceeded (attainment or unclassifiable area). New construction and 
modifications that emit ``regulated NSR pollutants'' \1\ over certain 
thresholds are subject to major NSR requirements, while smaller 
emitting sources and modifications may be subject to minor NSR 
requirements or be excluded from NSR altogether.
---------------------------------------------------------------------------

    \1\ 40 CFR 51.165(a)(1)(xxxvii), 40 CFR 51.166(b)(49), 40 CFR 
52.21(b)(50).
---------------------------------------------------------------------------

    Major NSR permits for sources that are located in attainment or 
unclassifiable areas are referred to as Prevention of Significant 
Deterioration (PSD) permits. These permits can also cover pollutants 
for which there are no NAAQS. Major NSR permits for sources located in 
nonattainment areas and that emit pollutants above the specified 
thresholds for which the area is in nonattainment are referred to as 
nonattainment NSR (NNSR) permits. The pollutant(s) at issue and the air 
quality designation of the area where the facility is located or 
proposed to be built determine the specific permitting requirements. 
The CAA requires sources subject to PSD to meet emission limits based 
on Best Available Control Technology (BACT) as specified by CAA section 
165(a)(4), and sources subject to NNSR to meet Lowest Achievable 
Emissions Rate (LAER) pursuant to CAA section 173(a)(2). Other 
requirements to obtain a major NSR permit vary depending on whether it 
is a PSD or NNSR permit.
    A new stationary source is subject to major NSR requirements if its 
potential to emit (PTE) a regulated NSR pollutant exceeds statutory 
emission thresholds.\2\ If it exceeds the applicable threshold, the NSR 
regulations define it as a ``major stationary source.'' \3\ An existing 
major stationary source triggers major NSR permitting requirements when 
it undergoes a ``major modification,'' which occurs when a source 
undertakes a physical change or change in method of operation (i.e., a 
``project'') that would result in (1) a significant emissions increase 
from the project, and (2) a significant net emissions increase from the 
source (i.e., a source-wide ``netting'' analysis that considers 
creditable emission increases and decreases occurring at the source as 
a result of other projects over a 5-year contemporaneous period). See, 
e.g., 40 CFR 52.21(b)(2)(i) and 40 CFR 52.21(b)(52). For this two-step 
process, the NSR regulations define what emissions rate constitutes 
``significant'' for each NSR pollutant. See 40 CFR 51.165(a)(1)(x), 40 
CFR 51.166(b)(23), and 40 CFR 52.21(b)(23).
---------------------------------------------------------------------------

    \2\ For PSD, the statute uses the term ``major emitting 
facility'' which is defined as a stationary source that emits, or 
has a PTE, at least 100 tons per year (TPY) if the source is in one 
of 28 listed source categories--or at least 250 TPY if the source is 
not--of ``any air pollutant.'' CAA 169(1). For NNSR, the emissions 
threshold for a major stationary source is 100 TPY, although lower 
thresholds may apply depending on the degree of the nonattainment 
problem and the pollutant. 40 CFR 51.165(a)(1)(iv)(A).
    \3\ 40 CFR 51.165(a)(1)(iv), 40 CFR 51.166(b)(1)(i), 40 CFR 
52.21(b)(1)(i).
---------------------------------------------------------------------------

    In many cases, these requirements of the major NSR program (or 
equivalent requirements) are formally adopted by a state or local air 
agency, and the agency submits a revised state implementation plan 
(SIP) to the EPA for approval. The EPA's regulations provide for the 
minimum requirements of these programs. Upon EPA approving the SIP, the 
air agency becomes the ``permitting authority'' for major NSR permits 
for sources within its boundaries. When a state or local air agency is 
not the permitting authority, either the EPA issues the major NSR 
permits or a state or local air agency issues the major NSR permits on 
behalf of the EPA by way of a delegation agreement. For sources located 
in Indian country, the EPA is currently the only permitting authority 
for major NSR. Currently, state and local air agencies issue the vast 
majority of major NSR permits each year.
    New sources and modifications that do not require a major NSR 
permit may instead require a minor NSR permit prior to construction. 
Minor NSR permits are almost exclusively issued by state and local air 
agencies, although the EPA issues minor NSR permits in some areas of 
Indian country. Minor NSR requirements are approved into a SIP in order 
to achieve and maintain the NAAQS. See CAA section 110(a)(2)(C). The 
CAA and EPA's regulations are less prescriptive regarding minimum 
requirements for minor NSR, so air agencies generally have more 
flexibility in designing their minor NSR programs.

B. What is project aggregation?

    As described in the preceding section, the EPA's implementing 
regulations for NSR establish a two-step process for determining major 
NSR applicability for projects at stationary sources. To be subject to 
major NSR requirements, the project must result in both (1) a 
significant emissions increase from the project (the determination of 
which is called ``Step 1'' of the NSR applicability analysis, or 
``project emissions accounting''); and (2) a significant net emissions 
increase at the stationary source, taking account of emission increases 
and emission decreases attributable to other projects undertaken at the 
stationary source within a specific time frame (called ``Step 2'' of 
the NSR applicability analysis, or ``contemporaneous netting''). This 
approach to applicability makes it necessary to accurately define what 
constitutes the ``project'' under review to ensure that the proper 
emissions increase resulting from the project is used when comparing it 
with the applicable NSR significance threshold at Step 1 of the NSR 
applicability analysis.\4\ Otherwise, a source could

[[Page 57326]]

conceivably carve up a higher-emitting project into two or more lower-
emitting ``projects'' and avoid triggering major NSR requirements.\5\ 
``Project aggregation,'' therefore, ensures that nominally-separate 
projects occurring at a source are treated as a single project for NSR 
applicability purposes where it is unreasonable not to consider them a 
single project.\6\
---------------------------------------------------------------------------

    \4\ In this notice, we use the terms ``project,'' ``changes,'' 
and ``activities'' interchangeably in referring to physical or 
operational changes that occur at a facility. In some cases, 
particularly in using the term ``activities,'' we are actually 
referring to ``sub-projects'' that are nominally separate in scope 
but are nevertheless related to other sub-projects such that they 
all are part of a larger single project when determining NSR 
applicability. It is important to note that our use of the term 
``activities'' in this notice is not intended to imply that every 
``activity'' at a plant is a physical or operational change. The EPA 
recognizes that there are numerous activities undertaken at a 
facility, of which only a subset will constitute ``changes'' under 
the NSR regulations.
    \5\ Emission changes from separate projects (not included under 
Step 1 as falling within the project under review) are considered at 
Step 2 if they are ``contemporaneous'' and ``otherwise creditable'' 
under the NSR regulations. See 40 CFR 52.21(b)(3).
    \6\ It is not permissible to seek to circumvent NSR by securing 
several minor NSR permits for individual projects with the effect of 
avoiding major NSR requirements for what is actually a single 
project.
---------------------------------------------------------------------------

    As with certain other aspects of the NSR program, determining what 
constitutes the ``project'' is a case-by-case decision that is both 
site-specific and fact-driven. There is no pre-determined list of 
activities that should be aggregated for a given industry or 
industries. It is, therefore, necessary to establish criteria for 
determining when nominally-separate activities are considered one 
project under NSR. The EPA has specifically sought to develop 
principles for aggregating changes such that a project is appropriately 
defined by the source, so that the emission increases attributable to 
the project are accurately quantified for purposes of analyzing NSR 
applicability. Over the years, the EPA articulated its policy on 
project aggregation through a series of statutory and regulatory 
interpretations contained in EPA letters and memoranda, the most 
commonly cited being a 1993 EPA memorandum regarding NSR applicability 
for activities that had occurred at a 3M facility in Minnesota.\7\
---------------------------------------------------------------------------

    \7\ Memorandum from John B. Rasnic, Director, Stationary Source 
Compliance Division, OAQPS, to George T. Czerniak, Chief, Air 
Enforcement Branch, EPA Region 5, titled, ``Applicability of New 
Source Review Circumvention Guidance to 3M--Maplewood, Minnesota'' 
(June 17, 1993) (hereinafter ``3M Memorandum'').
---------------------------------------------------------------------------

    To date, the EPA's focus in formulating criteria for project 
aggregation has been to ensure that NSR is not circumvented through 
some artificial separation of activities at Step 1 of the NSR 
applicability analysis where it would be unreasonable for the source to 
consider them to be separate projects. However, in a March 13, 2018, 
memorandum \8\ on the topic of ``project emissions accounting,'' the 
EPA broached the question of whether it might also somehow be possible 
for a source to circumvent NSR through some wholly artificial grouping 
of activities to include decreases in emissions as part of Step 1 of 
the NSR applicability analysis--i.e., assessing whether a project by 
itself results in a significant emissions increase before reaching Step 
2, where one then determines whether there will be a significant net 
emissions increase by taking into account all contemporaneous increases 
and decreases across the source. While we \9\ have been mindful of this 
question in deciding to employ the project aggregation criteria 
described in this action, we intend to address more fully this scenario 
in the context of a subsequent rulemaking action on the topic of 
project emissions accounting.
---------------------------------------------------------------------------

    \8\ Memorandum from E. Scott Pruitt, Administrator, to Regional 
Administrators, titled, ``Project Emissions Accounting Under the New 
Source Review Preconstruction Permitting Program'' (March 13, 2018) 
(hereinafter ``Project Emissions Accounting Memorandum'').
    \9\ In this preamble, the terms ``we'', ``our'' and ``us'' refer 
to the EPA.
---------------------------------------------------------------------------

C. Regulatory History

1. The 2009 NSR Aggregation Action
    On January 15, 2009, the EPA published a final action--which we are 
calling the ``2009 NSR Aggregation Action''--that described the 
principles of project aggregation that we would apply when determining 
whether a source had unreasonably segregated a single project into 
multiple projects, thereby circumventing the NSR permitting 
requirements.\10\ We had initially proposed in 2006 to establish 
principles for project aggregation through an amendment to the NSR 
regulations.\11\ However, because of the difficulty of creating a 
bright line to determine when activities should be aggregated, we 
ultimately decided not to adopt the proposed changes to the regulations 
and elected instead to pursue a less prescriptive approach by 
describing, in the 2009 action, the EPA's interpretation of the 
existing regulations and a policy for applying that interpretation 
going forward.
---------------------------------------------------------------------------

    \10\ Prevention of Significant Deterioration (PSD) and 
Nonattainment New Source Review (NSR): Aggregation and Project 
Netting (74 FR 2376; January 15, 2009).
    \11\ Prevention of Significant Deterioration (PSD) and 
Nonattainment New Source Review (NSR): Debottlenecking, Aggregation, 
and Project Netting (71 FR 54235; September 14, 2006).
---------------------------------------------------------------------------

    The 2009 NSR Aggregation Action called for sources and reviewing 
authorities to aggregate emissions from nominally-separate activities 
when they are ``substantially related'' for the purpose of determining 
whether they are a single modification resulting in a significant 
emissions increase under NSR at Step 1.\12\ This ``substantially 
related'' criterion is based on an interpretation of the term 
``project'' contained in the major NSR regulations.\13\ The action also 
included a statement that the EPA would, as a matter of policy, 
establish a rebuttable presumption that activities that occurred more 
than three years apart are not ``substantially related'' and therefore, 
generally, should not be aggregated for purposes of determining whether 
they are a single modification at Step 1.
---------------------------------------------------------------------------

    \12\ See 74 FR 2378 (``When there is no technical or economic 
relationship between activities or where the relationship is not 
substantial, their emissions need not be aggregated for NSR 
purposes.'' (emphasis added)). That is, mere relatedness is not 
sufficient to upend the source's definition of its project, but 
sources cannot circumvent NSR by artificially separating a series of 
emissions-increasing projects into separate projects that fall below 
the significance thresholds.
    \13\ See, e.g., 40 CFR 52.21(b)(52).
---------------------------------------------------------------------------

    The 2009 NSR Aggregation Action retained the existing rule text 
defining the term ``project''--i.e., ``a physical change in, or change 
in method of operation of, an existing major stationary source''--and 
interpreted this rule text to mean that sources and permitting 
authorities should combine emissions only when nominally-separate 
changes are ``substantially related.'' While acknowledging the case-
specific nature of a project aggregation determination, the 2009 NSR 
Aggregation Action described the factors that should be considered when 
evaluating whether changes are substantially related, including 
technical or economic dependence. It also offered examples of what it 
means to be substantially related, and it referenced examples provided 
in EPA's 2006 proposed rule on project aggregation to further amplify 
EPA's meaning of the term. Thus, in many respects, the ``substantially 
related'' interpretation in the 2009 NSR Aggregation Action was 
intended to encompass principles for aggregating projects that were 
similar to those that the EPA had proposed in 2006, but ultimately 
concluded should not be prescriptively defined in a regulation because 
of the difficulty of developing a bright line for determining when 
activities should be aggregated.
    The 2009 NSR Aggregation Action specifically addressed the timing 
element of project aggregation decisions in multiple ways. It affirmed 
that timing alone should not be a basis for aggregating projects 
because the appropriate basis for aggregation is whether there is a 
substantial technical or economic relationship. It further explained 
that activities that occur simultaneously should not be presumed to be 
substantially related, although it is reasonable to presume that 
activities

[[Page 57327]]

closer in time are more likely to be substantially related than 
activities separated by larger time frames. Thus, it affirmed that the 
timing between activities remains important from a standpoint of 
framing the analysis of whether a substantial technical or economic 
relationship exists.
    The 2009 NSR Aggregation Action also expressed that the farther 
apart projects are timed, the less likely they are to be substantially 
related, since the activities would likely be part of distinct planning 
and capital-funding cycles. It stated ``the passage of time provides a 
fairly objective indicator of nonrelatedness between physical or 
operational changes. Specifically, the greater the time period between 
activities, the less likely that a deliberate decision was made by the 
source to split an otherwise `significant' activity into two or more 
smaller, non-major activities.'' 74 FR 2380.
    To this end, the 2009 NSR Aggregation Action affirmed that timing 
could be a basis to not aggregate separate projects, and it established 
a policy of applying a rebuttable presumption against aggregating 
projects that occur 3 or more years apart. The EPA justified its 
selection of 3 years as the presumptive timeframe in part by reasoning 
that it ``is long enough to ensure a reasonable likelihood that the 
presumption of independence will be valid, but is short enough to 
maintain a useful separation between relevant construction cycles, 
consistent with industry practice. For example, in the case of electric 
utilities, a commenter explained that companies plan and schedule major 
turbine outages every four to five years.'' Id. However, the EPA did 
note that this presumptive timeframe may be rebutted in certain 
circumstances. For instance, the 2009 NSR Aggregation Action noted that 
where there is ``evidence that a company intends to undertake a phased 
capital improvement project'' where the activities ``have a substantial 
economic relationship,'' this would likely overcome the presumption 
that those activities should not be aggregated. Id.
    With regard to implementing the 3-year presumption, the EPA stated 
``the time period separating physical or operational changes should be 
calculated based on time of approval (i.e., minor NSR permit issuance). 
If a permit has not been, or will not be, issued for the physical or 
operational changes, the time period should be based on when 
construction commences on the changes.'' 74 FR 2381.
    The EPA also explained that a statement within the 3M Memorandum 
was potentially vulnerable to misapplication and did not properly 
reflect the ``substantially related'' criterion. The 3M Memorandum 
stated the following:

    Some minimum level of research activity and commensurate 
emissions, source-wide, perhaps could be expected from year to year, 
as would be expected to keep the 3M plant productive or operable. 
These emissions and thereby modifications cannot be presumed to be 
independent given the plant's overall basic purpose to support a 
variety of research and development activities. Therefore, even 
though each research project may have been individually conceived 
and separately funded, it is appropriate to look at the overall 
expected research activity in assessing NSR applicability and 
enforcement. 3M Memorandum at 5 (emphasis added).

    In the 2009 NSR Aggregation Action, the EPA expressed concern with 
this statement from the 3M Memorandum, saying ``it could be interpreted 
to imply that almost any activity is related to any other activity at 
that source simply because they are both capital investments and 
support the company's goal to make a profit.'' 74 FR 2376, 2379. The 
suggestion that all changes consistent with the ``overall basic 
purpose'' of the plant can and should be aggregated is inconsistent 
with the interpretation of ``project'' to include only those changes 
that have a substantial relationship. While the EPA did not, in the 
2009 NSR Aggregation Action, find such a broad approach to project 
aggregation was often applied after the 3M determination, we 
nevertheless had concerns that it did not represent an appropriate 
criterion for aggregating projects for NSR purposes and could be 
misapplied. Thus, in the 2009 NSR Aggregation Action, we maintained 
that two nominally separate projects are not substantially related if 
they are only related to the extent that they both support the source's 
``overall basic purpose.''
    In summarizing what it means for projects to be substantially 
related, the 2009 NSR Aggregation Action provided that ``in most cases, 
activities occurring in unrelated portions of a major stationary source 
(e.g., a plant that makes two separate products and has no equipment 
shared among the two processing lines) will not be substantially 
related. The test of a substantial relationship centers around the 
interrelationship and interdependence of the activities, such that 
substantially related activities are likely to be jointly planned 
(i.e., part of the same capital improvement project or engineering 
study), and occur close in time and at components that are functionally 
interconnected.'' 74 FR 2378. The 2009 NSR Aggregation Action added, 
``[t]o be `substantially related,' there should be an apparent 
interconnection--either technically or economically--between the 
physical and/or operational changes, or a complementary relationship 
whereby a change at a plant may exist and operate independently, 
however its benefit is significantly reduced without the other 
activity. We note that these factors are not necessarily determinative 
of a substantial relationship, but are merely indicators that may 
suggest that two or more activities are likely to be substantially 
related and, therefore, candidates for aggregation.'' Id.
2. Reconsideration and Administrative Stay
    On January 30, 2009, the Natural Resources Defense Council (NRDC) 
submitted a petition for reconsideration of the 2009 NSR Aggregation 
Action (the ``NRDC Petition''). In response to the NRDC Petition, on 
February 13, 2009, the EPA convened a proceeding for reconsideration as 
provided for under the CAA section 307(d)(7)(B), finding that the 
petitioner had raised objections to the action that arose after the 
comment period and that were of central relevance to the action.
    To allow time to complete the reconsideration prior to the 2009 NSR 
Aggregation Action becoming effective, the EPA announced a 90-day 
administrative stay of the action. See 74 FR 7284 (February 13, 2009). 
The EPA subsequently completed an action to further delay the effective 
date until May 18, 2010. See 74 FR 22693 (May 14, 2009). On May 18, 
2010, the EPA invoked APA section 705 to stay the action indefinitely 
pending the proceedings for judicial review or the completion of 
reconsideration. These stays were intended to allow the EPA the time to 
take comment on issues that were in question and complete any revisions 
of the action that became necessary as a result of the reconsideration 
process.
    As part of the reconsideration proceeding, on April 15, 2010, the 
EPA published a proposed reconsideration of the 2009 NSR Aggregation 
Action (the ``2010 Reconsideration Proposal'').\14\ 75 FR 19567. At the 
time, the EPA considered whether some of the points

[[Page 57328]]

raised by the NRDC petition might demonstrate potential flaws in the 
process and with fundamental aspects of the 2009 NSR Aggregation 
Action, including the legal basis, state adoption and implementation, 
and the clarity of the ``substantially related'' criterion. In the 2010 
Reconsideration Proposal, we expressed agreement with the petitioner on 
a number of fronts, invited comment on all issues raised in the NRDC 
petition, and proposed a preferred option to revoke the 2009 NSR 
Aggregation Action. The 2010 Reconsideration Proposal also referenced a 
number of the past determinations on project aggregation. See 75 FR 
19570-1.
---------------------------------------------------------------------------

    \14\ In the 2010 Reconsideration Proposal, the EPA described the 
2009 action as the ``NSR Aggregation Amendments.'' However, since 
this action did not ``amend'' the NSR regulations, but rather laid 
out an interpretation of our current regulations and described a 
policy on timing for aggregation, the 2009 action is more 
appropriately described, as it is described herein, as the 2009 NSR 
Aggregation Action.
---------------------------------------------------------------------------

    The EPA received a total of 27 comments on our 2010 Reconsideration 
Proposal. Of those commenters, 20 represented industry parties, three 
represented state and local air agencies, one represented a tribal 
government agency, one represented a federal agency, one represented an 
environmental advocacy group, and one was a private citizen.
3. Characterizing the 2009 NSR Aggregation Action
    In the history of actions that the EPA has taken regarding its 
project aggregation policy since 2006, the EPA has variously described 
the 2009 NSR Aggregation Action as a ``rule,'' ``interpretation,'' and 
``policy.'' However, we are now mindful that these terms may be used to 
refer to three distinct types of agency action that have varying 
degrees of legal effect and can be changed through different types of 
procedures. National Mining Association v. McCarthy, 758 F.3d 243, 251-
52 (D.C. Cir. 2014). As is explained below, the distinction between the 
proper procedures for changing rules, interpretations, and policies 
were not as clear to the agency in 2009 and 2010 as they are today. 
Recent court decisions have provided more clarity regarding the 
distinction between these types of actions and the means through which 
an agency can change them. In order to clarify how state and local 
permitting authorities may apply the principles for project aggregation 
that the EPA articulated in 2009, in this final action we seek to 
address any confusion regarding the nature of that 2009 action.
    We begin by defining what we understand each of these terms to mean 
when they are used in the discussion that follows. We use the term 
``rule'' to describe a ``legislative rule,'' which is ``[a]n agency 
action that purports to impose legally binding obligations or 
prohibitions on regulated parties--and that would be the basis for an 
enforcement action for violations of those obligations or 
requirements.'' National Mining, 758 F.3d at 251-52. We use the term 
``interpretation'' to describe ``an agency action that merely 
interprets a prior statute or regulation, and does not itself purport 
to impose new obligations or prohibitions or requirements on regulated 
parties.'' Id. Following the language in the APA, courts have used the 
term ``interpretive rule'' to describe this type of action. Id. Here, 
however, we use the term ``interpretation'' to more clearly distinguish 
such an action from a legislative rule. Finally, a ``policy'' or 
``statement of policy'' is ``an agency action that merely explains how 
the agency will enforce a statute or regulation--in other words, how it 
will exercise its broad enforcement discretion or permitting discretion 
under some extant statute or rule.'' Id.
    In 2006, we proposed a rule (meaning a legislative rule) that would 
have changed the text in the Code of Federal Regulations. We included 
in the preamble an explanation of what we intended that proposed 
regulatory text to mean. 71 FR 54235 (September 14, 2006). In that 
Federal Register document, we referred to the action as a ``proposed 
rule.'' Id.; see also 71 FR at 54245 (``We are proposing to add our 
aggregation policy to our NSR regulations . . .'').
    In 2009, we took ``final action'' in the matter. That is, we 
completed the action begun in 2006, while not changing the regulatory 
text itself. 74 FR 2376. In retaining the existing regulatory text 
defining the term ``project,'' we said that the action we were taking 
``interprets that rule text.'' Id. The interpretation offered in the 
2009 NSR Aggregation Action was that a ``project,'' which the 
regulatory text defines to mean ``a physical change in, or change in 
the method of operation of, an existing major stationary source,'' 40 
CFR 52.21(b)(53) (emphasis added), includes those activities that are 
``substantially related.'' 74 FR 2377. This portion of the 2009 NSR 
Aggregation Action was an interpretation.\15\ Although we had proposed 
to adopt a legislative rule in 2006 and to reflect that in amended 
regulatory text, we made a final decision in 2009 not to adopt any 
legislative rule or to amend the text of the NSR regulations. Instead, 
we chose to announce an interpretation of the existing regulations that 
drew from EPA's prior experience on the topic of project aggregation, 
but which to some degree altered the aggregation policy that the EPA 
had previously articulated in past guidance memoranda and letters.
---------------------------------------------------------------------------

    \15\ As explained above, courts follow the APA in referring to 
this type of action as an ``interpretive rule,'' but we refer to it 
herein simply as an ``interpretation'' to more clearly distinguish 
such an action from a legislative rule.
---------------------------------------------------------------------------

    In 2009, we also discussed our intention to apply a rebuttable 
presumption that activities separated by more than 3 years would not be 
considered substantially related. This section of the action is best 
understood as a statement of policy, as we were describing how we 
intended to exercise our discretion under the NSR regulations, as we 
interpreted them. We justified the 3-year presumption as a commonsense 
approach, in that we believed that in practice once 3 years had passed, 
``it is difficult to argue that th[e activities] are substantially 
related and constitute a single project.'' 74 FR 2380. But recognizing 
that there may be situations that would warrant an exception to this 
approach, we indicated that the 3-year presumption would be rebuttable. 
We indicated our view that it would be allowable and appropriate for 
other permitting authorities to ``also adopt this presumptive timeframe 
as guidance for their sources.'' 74 FR 2381.
    The 2009 action, thus, contained both an interpretation of the 
existing regulations and a statement of policy on how we intended to 
implement that interpretation. It is for this reason that we refer to 
it as the 2009 NSR Aggregation Action. However, when reconsidering that 
2009 action, we were not sufficiently clear in the 2010 Reconsideration 
Proposal regarding the nature of the action we were reconsidering. At 
times, we described the 2009 action as a ``final rule,'' and called it 
the ``NSR Aggregation Amendments,'' which could be read to suggest that 
we considered the 2009 NSR Aggregation Action, despite the lack of 
regulatory text changes, to somehow be a legislative rule, or something 
that ``amended'' the existing regulations.
    Much of the confusion stemmed from the fact that at the time we 
took these actions, judicial precedent in the United States Court of 
Appeals for District of Columbia Circuit (D.C. Circuit) provided that, 
where an agency had given a definitive interpretation to one of its own 
legislative rules, the agency could not thereafter change that 
interpretation without providing notice and an opportunity to comment. 
Paralyzed Veterans of America v. D.C. Arena L.P., 117 F.3d 579 (D.C. 
Cir. 1997). In part because of this precedent, we were persuaded in 
2010 that we should provide an opportunity for the public to comment on 
the 2009 interpretation, which could have been viewed as a

[[Page 57329]]

change from the interpretation that the EPA had articulated in 2006 and 
earlier. In addition, since we understood the Paralyzed Veterans 
opinion to require a notice-and-comment rulemaking process when an 
agency wished to change a regulatory interpretation (which, under the 
APA, would constitute the issuance of an ``interpretive rule,'' or, as 
we refer to it herein, an ``interpretation''), and because the 2009 NSR 
Aggregation Action had completed a notice-and-comment rulemaking 
process that had originally proposed to amend rule text, we chose in 
the 2010 Reconsideration Proposal to apply the procedures for 
reconsidering a ``legislative rule.''
    The United States Supreme Court has since abrogated the Paralyzed 
Veterans doctrine, ruling that it was inconsistent with the APA, which 
by its plain terms does not require agencies to go through a notice-
and-comment process in issuing an interpretive rule. Perez v. Mortgage 
Bankers Association, 135 S. Ct. 1199 (2015). Because the 2009 NSR 
Aggregation Action did not impose legally binding obligations or 
prohibitions on regulated entities or state permitting authorities, it 
was not a legislative rule. Since the 2009 NSR Aggregation Action was a 
combination of interpretation and policy statement, it could have been 
issued by the EPA without following notice-and-comment rulemaking 
procedures. 5 U.S.C. 553(b); 42 U.S.C. 7607(d)(1). Further, to the 
extent the interpretation reflected therein is a change from a prior 
interpretation, after the Supreme Court decision in Mortgage Bankers, 
it is now clear that an agency may also change such an interpretation 
of its regulations without the need to publish notice in the Federal 
Register and solicit public comment. However, because the EPA has been 
using notice-and-comment rulemaking procedures up to this point, the 
EPA believes it is prudent, but not required, in order to retain the 
interpretation of the NSR regulations with regard to project 
aggregation that we published in 2009, that we publish this document in 
the Federal Register. This procedure also allows us to complete the 
reconsideration proceeding and lift the indefinite administrative stay 
of the 2009 NSR Aggregation Action. We also believe that it is prudent 
to respond to those comments we received during the reconsideration 
process.

III. This Action

A. Overview

    In this action, we are taking final action on reconsideration of 
the issues for which we asked for comment in the 2010 Reconsideration 
Proposal. The proposal invited comment on all issues alleged in the 
petition for reconsideration, including the following: Lack of adequate 
opportunity for notice and comment on the final action; legal 
inconsistency with a prior court decision; lack of demonstrated need 
for a policy change; and lack of clarity over state plan adoption of 
the action.
    This action addresses all of the petitioner's issues. Moreover, to 
the extent that commenters lacked an adequate notice-and-comment 
opportunity in the development of the 2009 NSR Aggregation Action, the 
reconsideration process has addressed this deficiency by inviting 
comment in 2010 on the issues raised by the petitioner. This action (1) 
takes final action on the 2010 Reconsideration Proposal and retains the 
2009 NSR Aggregation Action without adopting any changes to the rule 
text or the interpretation and statement of policy contained therein; 
(2) completes the CAA section 307 reconsideration proceeding on the 
2009 NSR Aggregation Action to address any potential notice-and-comment 
deficiency; and (3) lifts the APA section 705 stay of the 2009 NSR 
Aggregation Action. The conclusions reached and expressed in this final 
action are based on careful review of the public comments on the 2010 
Reconsideration Proposal.\16\
---------------------------------------------------------------------------

    \16\ In the docket for this action, we are making available a 
document, ``Response to Public Comments for Prevention of 
Significant Deterioration (PSD) and Nonattainment New Source Review 
(NSR): Aggregation; Reconsideration'', in which the EPA responds to 
the public comments received on the 2010 Reconsideration Proposal.
---------------------------------------------------------------------------

    This final decision on reconsideration of the 2009 NSR Aggregation 
Action does not finalize the 2010 Reconsideration Proposal's preferred 
option to revoke the 2009 NSR Aggregation Action's interpretation and 
policy. Upon reviewing public comments, after further deliberation, and 
taking account of the Administration's priorities and policy goals, the 
EPA has concluded that the interpretation and policy in the 2009 NSR 
Aggregation Action should be retained.\17\ We believe the 2009 NSR 
Aggregation Action articulates a reasonable standard for aggregating 
related projects and is consistent with the CAA and our regulations.
---------------------------------------------------------------------------

    \17\ See Presidential Memorandum on Streamlining Permitting and 
Reducing Regulatory burdens for Domestic Manufacturing (82 FR 8667; 
January 24, 2017); Executive Order 13777 on Enforcing the Regulatory 
Reform Agenda (82 FR 12285, March 1, 2017).
---------------------------------------------------------------------------

    With regard to the petitioner's concern about how the 2009 NSR 
Aggregation Action applies to EPA-approved permitting programs, we 
affirm our decision in 2009 not to revise the current rule text, and 
instead to conclude that the terms ``project'' and ``a physical change 
in, or change in method of operation of'' in the existing NSR 
regulations can be reasonably interpreted as already incorporating the 
``substantially related'' test set forth in the 2009 preamble. Because 
the 2009 NSR Aggregation Action did not amend the rule text, state and 
local air agencies with approved state implementation plans (SIPs) are 
not required to amend those plans to adopt this interpretation that 
projects should be aggregated when ``substantially related.'' If state 
and local agencies want to adopt this interpretation, we believe that 
in most cases this interpretation can be applied without formal 
adoption into their rules. We encourage state and local air agencies to 
follow this interpretation to ensure greater national consistency in 
making NSR applicability determinations, though state and local air 
agencies with approved SIPs can continue to apply their own 
interpretation of the scope of a ``project.''
    Consistent with comments received on the EPA's 2006 proposed rule, 
commenters on the 2010 Reconsideration Proposal raised concerns with 
the clarity of our prior policy on project aggregation, which was 
developed over time through a number of post hoc site-specific 
applicability determinations. We anticipate the 2009 NSR Aggregation 
Action will reduce any confusion over our past policy and provide 
sources and regulators with increased clarity when determining whether 
projects should be aggregated for NSR purposes. The EPA believes the 
principles outlined in the 2009 NSR Aggregation Action will not only 
help to achieve greater national consistency in project aggregation 
determinations but will also streamline NSR permitting by reducing the 
time needed to assess whether nominally-separate physical and 
operational changes should be aggregated for NSR applicability 
purposes.
    As this action officially completes our reconsideration proceeding, 
we are also lifting the APA section 705 stay and announcing the 
effective date of the 2009 NSR Aggregation.

[[Page 57330]]

B. Retaining the 2009 NSR Aggregation Action

1. An Interpretation Is Needed
    As explained earlier in this document, the EPA's past position on 
project aggregation--prior to the 2009 NSR Aggregation Action--was not 
established through a rule or through a single, comprehensive policy 
statement. Rather, the policy had been articulated by the EPA through a 
number of site-specific determinations, many of which were issued after 
the activities subject to the determination had already occurred. 
Navigating this collection of EPA statements, capturing their salient 
points, and determining whether and how to apply their rationale to new 
determinations with different fact patterns was arguably a challenge 
for sources and permitting authorities over the years. Such an approach 
lacked clarity for sources and permitting authorities, making it 
sometimes difficult to understand the overall policy so they could 
effectively apply it prospectively.
    There is a substantive distinction between making case-by-case 
determinations after-the-fact and making case-by-case determinations 
prospectively--i.e., as part of a permitting applicability review--for 
NSR purposes. Many post hoc determinations are made with an eye to 
determining whether the requirements of NSR were circumvented, whereas 
prospective determinations are made with the purpose of giving sources 
an opportunity to evaluate modifications during the planning or 
preconstruction phase in order to determine whether a planned or 
proposed modification requires a PSD or NNSR permit, so as not to 
circumvent the NSR process. While the underlying criteria for assessing 
whether to group multiple activities as a single project should be the 
same regardless of whether the determination is prospective or post 
hoc, a post hoc determination is often very specific to the industry 
and the individual fact pattern under consideration, and therefore 
applying the determination's rationale prospectively, while potentially 
informative, could be misapplied to situations involving different 
industries or having different fact patterns. The 2009 NSR Aggregation 
Action also recognized the limitations of having a policy that is based 
on the specific fact patterns of past determinations: ``the decision to 
aggregate or disaggregate activities is highly case-dependent, such 
that letters and memoranda that opine on whether to aggregate a 
particular set of activities at one facility are not necessarily 
transferable to a decision to aggregate a similar set of activities but 
with a slightly different set of circumstances at a different plant.'' 
74 FR 2377.
    Previous agency statements can be taken out of context or 
misunderstood when reviewing projects having a different set of facts. 
For example, while the 3M Memorandum was considered by some as the 
EPA's guiding policy on project aggregation, parties could certainly 
misconstrue portions of that statement to suggest that all projects 
occurring within the same timeframe should be aggregated, or that all 
projects occurring at a facility should be aggregated as long as they 
contribute to the source's ``overall basic purpose.'' Such an 
approach--i.e., to aggregate projects simply because they may occur 
close in time or may support the same overall purpose of the facility--
fails to take proper account of the actual interrelationship of 
activities. Meanwhile, in other parts of the 3M Memorandum, the EPA's 
statements clearly indicate that, in order to justify aggregating 
activities for purposes of major NSR, the reasonable approach is to 
determine whether those activities are related in some meaningful way: 
e.g., ``[a]uthorities should scrutinize [permit] applications that 
relate to the same process or units . . .''; ``two or more related 
minor changes over a short time period should be studied for possible 
circumvention.'' 3M Memorandum at 3 (emphasis added). We consequently 
do not believe that a broader approach to aggregating activities--i.e., 
based on their contribution to a plant's overall purpose--is an 
accurate characterization of the EPA's view at the time of the 3M 
determination. Furthermore, we do not believe it reflects EPA's view in 
any other statement made by the agency over the years.
    We noted in the 2010 Reconsideration Proposal that ``in reviewing 
the record for the NSR Aggregation Amendments, we find that the only 
factual support for the contention that our historic approach caused 
confusion was anecdotal,'' and that the ``parties supporting a change 
in policy failed to provide us with any characterization of the overall 
level of uncertainty or other problems resulting from the existing 
policy on aggregation.'' 75 FR 19572. However, after further 
consideration, the EPA finds this to be an insufficient basis for 
changing or revoking the 2009 NSR Aggregation Action. So-called 
``anecdotal'' evidence is nevertheless still evidence of which the 
agency can properly take account if, in its judgment, it finds it to be 
meaningful. Indeed, the criticism of relying on ``anecdotes'' suggests 
that examples of problems offered in public comments should be ignored. 
The EPA is required to take into account the comments submitted. 
Furthermore, merely because the overall level of uncertainty 
demonstrated by public comments cannot be characterized--a given entity 
would not necessarily know whether others were as uncertain as they 
were--does not serve to demonstrate that the 2009 NSR Aggregation 
Action was unwarranted. We believe that the evidence before the EPA in 
2009 and the agency's own extensive permitting experience, coupled with 
statements from public commenters in this reconsideration proceeding, 
clearly indicates that the EPA's prior policy on project aggregation 
lacked clarity and promoted confusion. The 2009 NSR Aggregation Action 
provides a more concise formulation for how to interpret the scope of a 
project and provides clarity for permitting authorities, regulated 
entities, and the public.
    Finally, the 2010 Reconsideration Proposal states that ``[w]hile 
the [2009 NSR Aggregation Action] may, in some respects, appear clearer 
than our previous policy, we are not convinced that it achieved enough 
additional clarity to improve the process of making aggregation 
assessments by sources and reviewing authorities. . . .'' 75 FR 19573. 
After further consideration, we now believe that providing clarity in a 
single document is a better approach than continuing the previous 
policy that was based on a host of EPA letters and memoranda, which 
collectively provided less clarity. We recognize there will continue to 
be ``gray areas'' that sources and permitting authorities will 
ultimately have to work through in deciding whether or not to aggregate 
a set of changes at a facility. But this is attributable to the 
inherent nature of such decisions, not to some deficiency in the 2009 
NSR Aggregation Action. That does not mean that the EPA should abandon 
the clarity it attempted to provide in that action.
2. ``Substantially Related'' Is an Appropriate Standard
    As noted above, the EPA continues to believe that there is a need 
for some criteria for determining when nominally-separate changes 
should be considered a single ``project'' for purposes of determining 
NSR applicability. It remains necessary to draw a line between those 
activities that are to be considered a single ``physical or operational 
change'' and those that are not. In this action, we are affirming that 
the 2009 NSR Aggregation Action's

[[Page 57331]]

``substantially related'' test is an appropriate standard for project 
aggregation.
    As explained elsewhere in this document, the nature of the project 
aggregation determination is case-specific, which means it is 
inherently difficult to establish a bright line standard: Such a 
standard may be reasonable when conducting an evaluation of project 
scope in one situation, but could prove to be unreasonable or 
unworkable when applied in other situations. This case-by-case aspect 
necessitates that the EPA establish a reasonable general principle to 
apply, and we believe the ``substantially related'' criterion is an 
appropriate principle for concluding that claimed separate projects are 
a single project for NSR applicability purposes. We believe the 
substantially related criterion is sound from a policy and 
implementation perspective.
    The 2009 NSR Aggregation Action effectively addresses certain past 
EPA statements in relation to implementing the ``substantially 
related'' test for future project aggregation determinations. The 2009 
NSR Aggregation Action outlined the role of timing--specifically, that 
timing alone is not determinative of whether activities are 
substantially related and that, as a policy matter, activities 
separated in time by three or more years may be presumed to be not 
substantially related. The 2009 NSR Aggregation Action also rejected 
the use of an ``overall basic purpose'' criterion for aggregating 
physical or operational changes, since it could have been read to 
constitute an open-ended standard, resulting in the unreasonable or 
improper aggregation of unrelated activities.
    Importantly, we do not believe the 2009 NSR Aggregation Action 
reflects a major shift in policy from EPA's prior policy on project 
aggregation. To the contrary, we believe that in many ways the 2009 NSR 
Aggregation Action clarifies and supplements previous statements of 
policy. For example, in the case of timing, the 3M Memorandum suggested 
that when minor NSR permit applications occur ``over a short time 
period (e.g., 1 year or 18 months), the modifications may require major 
new source review.'' 3M Memorandum at 4 (emphasis added). Thus, the 3M 
Memorandum never said timing was the sole criterion or otherwise 
conclusive. Rather, timing was a reason to look more closely at the 
relevant activities' ``intrinsic relationship with each other (physical 
proximity, stages of production process, etc.) and their impact on 
economic viability of the plant (scheduling down time in light of 
production targets, economies of scale, etc.).'' Id. Similarly, the 
2009 NSR Aggregation Action said that ``whether a physical or 
operational change is dependent on another for its viability is still a 
relevant factor in assessing whether the changes should be 
aggregated,'' and ``substantially related activities are likely to be 
jointly planned (i.e., part of the same capital improvement project or 
engineering study), and occur close in time and at components that are 
functionally interconnected.'' 74 FR 2378.
    In addition, the ``substantially related'' criterion is not 
materially different from the factors the agency has considered in 
previous project aggregation decisions. Over time, the EPA has used 
various terms and phrases--e.g., ``intrinsic relationship'' as was used 
in the 3M Memorandum--to describe the basis for why multiple nominally-
separate changes at a source should be treated as a single project for 
NSR applicability purposes. The term ``substantially related'' is, 
therefore, little more than a functional synonym for other terms that 
the EPA has historically used to characterize its project aggregation 
policy. While sources and permitting authorities making project 
aggregation determinations may continue to use the EPA's previous 
terms, and may rely on other terms or phrases going forward, we believe 
that the terminology used should ultimately express a standard for 
determining whether the activities are or are not substantially 
related. Thus, we believe ``substantially related'' works effectively 
as an umbrella term to include these previous descriptors for analyzing 
the relationship between projects that warrant aggregation.
    Finally, the matter of defining the scope of a project was raised, 
in a different context, in the Project Emissions Accounting Memorandum 
issued on March 13, 2018. There, we observed that, as general matter, 
the source itself is responsible for defining the scope of its own 
project, subject to the limitation that the source cannot seek to 
circumvent NSR by characterizing the proposed project in a way that 
would separate a single project into multiple projects. We further 
pointed out that, ``[s]ubject to the equivalent understanding that it 
might be possible [for a source] to circumvent NSR through some wholly 
artificial grouping of activities, the EPA does not interpret its NSR 
regulations as directing the agency to preclude a source from 
reasonably defining its proposed project broadly, to reflect multiple 
activities.'' \18\
---------------------------------------------------------------------------

    \18\ Project Emissions Accounting Memorandum at 9 (emphasis 
added).
---------------------------------------------------------------------------

    In the Project Emissions Accounting Memorandum, we noted that EPA 
was then evaluating whether to undertake a future notice-and-comment 
rulemaking to implement, through changes to the regulatory text itself, 
the interpretation of the NSR applicability provisions set forth in the 
memorandum. At such time as we proceed with that rulemaking, we will 
look to provide further guidance with respect to properly accounting 
for the scope of a project in which a source is seeking to take account 
of emission decreases at Step 1 of the NSR applicability analysis. 
Meanwhile, in advance of that rulemaking, we take the opportunity here 
to clarify that, as a general matter, it is neither necessary nor 
appropriate to take into consideration such matters as whether emission 
decreases attributable to a particular activity are ``integral'' to the 
overall project, as had once been proposed by a petroleum refinery to 
the EPA.\19\ Our current view is that the concerns regarding the real 
possibility that NSR might be circumvented through some artificial 
separation of activities where it would be unreasonable to consider 
them separate projects--i.e., the concerns which the 2009 NSR 
Aggregation Action is intended to address--are not so obviously 
presented by the situation where a source itself is choosing to group 
together, as a single project, activities to which a projected 
emissions decrease is attributable.\20\ In a future rulemaking to 
clarify, through regulatory text changes, the interpretation set forth 
in the Project Emissions Accounting Memorandum, the EPA will be taking 
comment on whether our current view of this issue is reasonable, 
whether the ``substantially related'' criterion described here may 
speak to this issue, and other related matters.
---------------------------------------------------------------------------

    \19\ Letter from Steven C. Riva, U.S. EPA Region 2, to Kathleen 
Antoine, HOVENSA, LLC, ``Re: Emission Decreases Integral to 
Projects'' (June 7, 2010) (``EPA, by this letter, is not opining on 
the merits of HOVENSA's analysis regarding the underlying basis for 
`integral to the project' approach.'').
    \20\ Indeed, the EPA views this latter situation as one where 
sources could potentially be incentivized to seek out emission 
reductions that might otherwise be foregone entirely--e.g., because 
of perceived complexity with contemporaneous netting under Step 2 of 
the NSR applicability analysis.
---------------------------------------------------------------------------

3. Legal Basis Is Sound
    We believe the 2009 NSR Aggregation Action is legally supportable 
and makes sense for sometimes difficult case-by-

[[Page 57332]]

case determinations required for assessing whether to aggregate 
nominally-separate projects. Contrary to the petitioner's argument, the 
use of the term ``substantially related'' would not create a carve-out 
from the scope of the statutory definition of ``modification.''
    Drawing on arguments made by NRDC in its petition, in 2010 we had 
postulated, while ``[m]uch of the emphasis'' of New York v. EPA, 443 
F.3d 880 (D.C. Cir. 2006) (New York II) and other cases had been on 
whether the EPA ``could exclude small changes from being considered 
potential modifications as defined in the Act,'' the court's reasoning 
in New York II also applies to a rule that would split apart one change 
into separate changes in order to limit the applicability of NSR.'' 75 
FR 19571. The D.C. Circuit's New York II decision had focused on 
whether the EPA's amendment to the ``routine maintenance, repair and 
replacement'' provision of the NSR regulations which provided that a 
specifically defined category of ``equipment replacement'' projects did 
not constitute a ``physical change or change in the method of 
operation,'' was lawful. The court in New York II held that it was not 
lawful, opining that the EPA ``must apply NSR whenever a source 
conducts an emissions-increasing activity that fits within one of the 
ordinary meanings of physical change.'' 443 F.3d at 885.
    In the 2010 Reconsideration Proposal, we said we then read the D.C. 
Circuit's opinion as ``requir[ing] EPA to aggregate any group of small 
changes'' that were ``sufficiently related to `fit[] within one of the 
ordinary meanings of `physical change.' '' 75 FR 19571. In this regard, 
we said that we ``agree[d] with [NRDC's] contention that, to the extent 
that our `substantially related' interpretation,'' as set forth in the 
2009 NSR Aggregation Action, would ``exclude meanings that fit within a 
reasonable understanding of the ordinary meaning of `any physical 
change,' '' that interpretation would ``impermissibly narrow the scope 
of CAA section 111(a)(4).'' Id. We sought comment on this analysis of 
the statute and New York II.
    Upon further consideration and after reviewing the public comments 
on this reconsideration proposal, the agency does not read New York II 
as supportive of the notion that the ``substantially related'' 
interpretation set forth in the 2009 NSR Aggregation Action is somehow 
contrary to the language of CAA section 111(a)(4). While we had 
previously suggested that there might be some weight to NRDC's argument 
that the `` `aggregation of nominally separate changes that are not 
substantially related' also may be within an ordinary meaning of 
physical change,'' 75 FR 19571, citing NRDC Petition at 5-6 (emphasis 
in original), we do not now perceive any merit in NRDC's assertion.
    With NRDC's arguments in mind, the agency at one point read New 
York II as suggesting that the CAA ``prohibits EPA from picking and 
choosing among meanings of the phrase `any physical change . . . or 
change in the method of operation' if it would result in omitting a 
common meaning that would subject an emission increase to review.'' 75 
FR 19571. Based on this, we were concerned that, ``[i]f `substantially 
related' would omit an ordinary, common meaning of physical change that 
would bring an emissions-increasing project under review, then the 
definition would eliminate a type of physical change that Congress 
intended to cover (i.e., the change that consists of the group of 
nominally-separate changes that comprise a project but do not qualify 
as `substantially related').'' Id. Thus, we reasoned at the time 
``that, to the extent that [the] `substantially related' interpretation 
would exclude meanings that fit within a reasonable understanding of 
the ordinary meaning of `any physical change,' '' then the 2009 NSR 
Aggregation Action ``would impermissibly narrow the scope of CAA Sec.  
111(a)(4).'' Id.
    We now believe that such concerns were unwarranted. Upon further 
consideration, we do not view New York II, properly understood, as 
providing support for the proposition that a ``common meaning'' of a 
single ``change'' would include multiple changes, much less multiple, 
separate changes that are not substantially related, such as changes 
which are undertaken at a source at different times, or undertaken for 
different purposes, or which are otherwise unrelated to each other. 
That is, the EPA's current view is that nothing in New York II 
supports, much less compels, a reading of the CAA under which all 
``nominally-separate changes'' are deemed to ``comprise'' a single 
``project,'' where those changes are not substantially related. 
Nevertheless, under the interpretation reflected in the 2009 NSR 
Aggregation Action, multiple changes that are ``substantially related'' 
are to be considered to be one project for purposes of determining NSR 
applicability.
    Finally, to the extent that NRDC argues that the aggregation of 
activities that are not substantially related into one activity that 
fits within the ordinary meaning of a physical change--and not 
aggregating those changes to compare to the significance level would 
violate New York II--it has provided no examples where that may be the 
case and have not followed the reasoning of their argument to its 
logical conclusion. This argument would require the EPA to prove a 
negative: That whatever interpretation or policy on aggregation we 
adopted would not exclude any level of aggregated activities that fit 
within the ordinary meaning of a physical change. This impossible task 
would mean that even the EPA aggregation policy prior to the 2009 NSR 
Aggregation Action was in violation of New York II because it allowed a 
facility to sometimes disaggregate activities when, if aggregated, they 
would fall within the ordinary meaning of physical change. A better 
approach to defining the scope of the ordinary meaning of physical 
change is to provide, as we did in the 2009 NSR Aggregation Action, a 
principle for source owners or operators to follow, here the 
``substantially related'' principle, when defining the scope of ``a 
physical change in, or change in method of operation of,'' pursuant to 
40 CFR 52.21(b)(52), in a particular case.
4. Adoption Is Not Mandatory
    We acknowledge that, by not making any changes to the regulatory 
text, as had been proposed, it may have been somewhat unclear to some 
whether state and local air agencies have to adopt or implement the 
elements of the 2009 NSR Aggregation Action, and, if so, how they 
should do so. In the 2010 Reconsideration Proposal, we expressed our 
agreement with ``NRDC's assertion that the state and local 
implementation requirements of the NSR Aggregation Amendments are 
unclear,'' and that the ``question of whether a SIP amendment is 
required when the CFR remains unchanged is likely to cause confusion 
for reviewing authorities and other stakeholders.'' 75 FR 19572. Taking 
account of this confusion, the agency considered that it ``added 
support for our preferred position in this notice, which is to revoke'' 
the 2009 NSR Aggregation Action. Id.
    We now find such concerns over potential ``confusion'' to have been 
overstated. In the Response to Comments document for the 2009 NSR 
Aggregation Action (2009 RTC), the agency had specifically noted that 
``[s]ince we are not promulgating the proposed rule regulatory changes, 
we are not adding NSR minimum program elements that would require 
states to modify their SIP.'' 2009 RTC at 56. The agency continued that 
it would ``begin applying the interpretations laid out in the final 
action to activities that postdate actions after the effective date of 
the final rulemaking notice.'' Id. ``At

[[Page 57333]]

that time,'' the EPA explained, states ``may also begin applying EPA's 
interpretations to the extent they do not conflict with their approved 
SIPs.'' Id. We now believe it is likely that state and local permitting 
authorities would have understood this straightforward explanation.
    Further, as previously discussed, determining whether a source has 
sought to circumvent NSR by failing to treat nominally-separate 
activities as a single project is inherently case-specific and fact-
dependent. Given this, it is not reasonable to imagine that perfect 
clarity could ever be achieved. To the extent, however, that the 2009 
NSR Aggregation Action, in setting forth both the ``substantially 
related'' interpretation and the EPA's policy for applying that 
interpretation, provides some meaningful guidance to sources and to 
state and local permitting authorities, we fail to understand how 
revoking the 2009 NSR Aggregation Action would serve to promote 
clarity.
    Indeed, in this regard, we believe in most cases that sources and 
state and local air agencies already implement a standard that is 
similar to the substantially related standard. To the extent that a 
state or local air agency desires to formally adopt the 2009 NSR 
Aggregation Action, the EPA will provide support to those agencies to 
process SIP submittals and issue approvals, as warranted. In most 
cases, however, we do not think changes in state plans would be needed 
to implement this interpretation.

C. Completing the Reconsideration Proceeding

    We believe that this final action addresses the concerns raised by 
the petitioner with respect to the 2009 NSR Aggregation Action--e.g., 
adequate notice and logical outgrowth, the legal underpinnings of the 
action, state adoption, and our need to change or clarify our 
aggregation policy. Accordingly, this action concludes the 
reconsideration proceeding of the 2009 NSR Aggregation Action.

D. Lifting the Administrative Stay; Announcement of Effective Date

    On May 18, 2010, after a series of temporary administrative stays 
of the 2009 NSR Aggregation Action, the EPA exercised the provisions of 
the APA section 705 to postpone the effectiveness of the action ``until 
judicial review is no longer pending or the EPA completes the 
reconsideration process.'' 75 FR 27644. Since this action concludes the 
reconsideration proceeding, and we have affirmed the legal consistency 
and policy appropriateness of the 2009 NSR Aggregation Action, we are 
hereby lifting the indefinite administrative stay and announcing the 
effective date of the action. The effective date of the 2009 NSR 
Aggregation Action, published in the Federal Register on January 15, 
2009 (74 FR 2376), and delayed on February 13, 2009 (74 FR 7284), May 
14, 2009 (74 FR 22693), and May 18, 2010 (75 FR 27643), begins again on 
November 15, 2018.

IV. Environmental Justice Considerations

    We believe that this action does not have any effect on 
environmental justice communities. Through this action, the EPA is 
affirming its interpretation that its current NSR regulations allow for 
the 2009 NSR Aggregation Action and, as such, no increased burden is 
expected for source owners, permitting authorities, or environmental 
justice communities.

V. Executive Order 12866: Regulatory Planning and Review and Executive 
Order 13563: Improving Regulation and Regulatory Review

    This action is a significant action that was submitted to the 
Office of Management and Budget (OMB) for review. Any changes made in 
response to OMB recommendations have been documented in the docket.

VI. Judicial Review

    Section 307(b)(1) of the CAA indicates which Federal Courts of 
Appeal have venue for petitions of review of final agency actions by 
the EPA under the CAA. This section provides, in part, that petitions 
for review must be filed in the U.S. Court of Appeals for the District 
of Columbia Circuit (i) when the agency action consists of ``nationally 
applicable regulations promulgated, or final actions taken, by the 
Administrator'' or (ii) when such action is locally or regionally 
applicable, if ``such action is based on a determination of nationwide 
scope or effect and if in taking such action the Administrator finds 
and publishes that such action is based on such a determination.''
    This action completes the reconsideration proceeding and makes 
effective the 2009 NSR Aggregation Action. The 2009 NSR Aggregation 
Action is an interpretation of NSR rule language that applies in every 
state and territory in the United States where EPA is the permitting 
authority. Therefore, to the extent that this action is a ``final 
action,'' it is ``nationally applicable'' within the meaning of CAA 
section 307(b)(1).
    Under section 307(b)(1) of the Act, to the extent that this action 
is judicially reviewable, petitions for judicial review of this action 
must be filed in the United States Court of Appeals for the District of 
Columbia Circuit by January 14, 2019.

VII. Statutory Authority

    The statutory authority for this action is provided by section 
301(a) of the CAA as amended (42 U.S.C. 7601(a)). This document is also 
subject to section 307(d) of the CAA (42 U.S.C. 7407(d)).

    Dated: November 7, 2018.
Andrew R. Wheeler,
Acting Administrator.
[FR Doc. 2018-24820 Filed 11-14-18; 8:45 am]
 BILLING CODE 6560-50-P


