PRECEDENTIAL

       UNITED STATES COURT OF APPEALS
            FOR THE THIRD CIRCUIT
                 _____________

               Nos. 21-3023 & 22-1012
                   _____________

      CENTER FOR BIOLOGICAL DIVERSITY,
                   Petitioner

                          v.

 UNITED STATES ENVIRONMENTAL PROTECTION
AGENCY; ADMINISTRATOR OF THE UNITED STATES
    ENVIRONMENTAL PROTECTION AGENCY

            CITY OF PHILADELPHIA;
        PENNSYLVANIA DEPARTMENT OF
         ENVIRONMENTAL PROTECTION,
                    Intervenors
                 _____________

        On Petition for Review of Actions of the
     United States Environmental Protection Agency
                    _____________

                Argued: April 13, 2023
                   ______________

Before: CHAGARES, Chief Judge, SCIRICA and AMBRO,
                  Circuit Judges
                  (Filed: July 25, 2023)
                     _____________


Alexa M. Carreno [ARGUED]
Jeremy D. McKay
Environmental and Animal Defense
501 South Cherry Street
Suite 1100
Denver, CO 80246

      Counsel for Appellant

Melissa Hoffer
United States Environmental Protection Agency
Office of General Counsel
1200 Pennsylvania Avenue NW
Washington, DC 20460

Jeffrey Hughes [ARGUED]
United States Department of Justice
Environment & Natural Resources Division
P.O. Box 7611
Ben Franklin Station
Washington, DC 20044

      Counsel for Appellee

Zachary G. Strassburger
City of Philadelphia
Law Department
1515 Arch Street




                              2
Philadelphia, PA 19102

      Counsel for Intervenor City of Philadelphia

Robert A. Reiley
Pennsylvania Department of Environmental Protection
Office of Chief Counsel
9th Floor
400 Market Street
Rachel Carson State Office Building
Harrisburg, PA 17101

Jesse C. Walker [ARGUED]
Office of Attorney General of Pennsylvania
Bureau of Regulatory Counsel
RCSOB 9th Floor
P.O. Box 8464
Department of Environmental Protection
Harrisburg, PA 17105

      Counsel for Intervenor Pennsylvania Department of
      Environmental Protection

                      _____________

                OPINION OF THE COURT
                    _____________




CHAGARES, Chief Judge.




                             3
       The Center for Biological Diversity (the “Center”)
challenges the Environmental Protection Agency’s (“EPA”)
approval of certain air pollution control technology for use at
various Pennsylvania industrial facilities. The Center argues
that the EPA violated the Clean Air Act by focusing
exclusively on emissions from those facilities instead of
examining their impact upon air quality more generally. The
Center also claims that, even if the EPA is permitted to base its
approvals on an emissions-only analysis, the agency
incorrectly concluded that emissions would not be increased
by Pennsylvania’s pollution control technologies at issue here.
Because we interpret the relevant statutory provisions to permit
the EPA’s chosen emissions-based approach, and because the
Center’s alternative challenges are procedurally and
substantively deficient, we will deny the Center’s consolidated
petitions for review.

                               I.

        We embark first on an acronym-filled journey through
this case’s factual and procedural history. This dispute has its
origins in the Clean Air Act, a statute meant “to protect and
enhance the quality of the Nation’s air resources so as to
promote the public health and welfare and the productive
capacity of its population” and “to encourage and assist the
development and operation of regional air pollution prevention
and control programs.” 

(b)(1), (4). The
statute directs the EPA to set and periodically revise national
ambient air quality standards (“NAAQS”) for certain
pollutants. 

(a)–(d). The NAAQS constitute
air quality benchmarks toward which states must work by
reducing their pollution levels. 

.




                               4
       Although the EPA sets the NAAQS, individual states
are afforded discretion in the creation and implementation of
plans to achieve the EPA’s targets for reduction of air
pollutants. To this end, states must at various times submit
state implementation plans (“SIPs”) that “specify the manner
in which [NAAQS] will be achieved and maintained” within
that state. 

(a). The EPA then reviews
whether the SIP in question meets the Clean Air Act’s
requirements, in which case the agency “shall” approve it. 

(k)(3). A similar principle governs situations
where a state revises a pre-existing SIP: the EPA “shall not
approve a revision of a plan if the revision would interfere with
any applicable requirement concerning attainment and
reasonable further progress . . . .” 

(l).
“Attainment” in § 7410(l) refers to attainment of any NAAQS,
not just the one for which a SIP or SIP revision has been
submitted.

        The relevant pollutant here is ozone. Over the past
several decades, the EPA has revised the ozone NAAQS to
make it more rigorous. When such a revision is made, the EPA
must assess whether a geographic area is compliant with the
updated NAAQS. 

(d)(1)(A)–(B). If a region
does not meet the updated NAAQS, it is deemed to be in
“nonattainment” and is subject to increasingly stringent
requirements depending upon the severity of its air quality
problems. See 

(d)(1)(A). Certain states in
the Northeastern United States are additionally subject to
stricter ozone pollution requirements by virtue of their location
in what the Clean Air Act terms the “Ozone Transport Region,”
a geographic area with properties that may render these states’
pollution control strategies interdependent. 42 U.S.C. § 7511c.




                               5
        Pennsylvania is both part of the Ozone Transport
Region and has several areas within it that are in nonattainment
with the 1997 and 2008 iterations of the ozone NAAQS. As a
result, the Clean Air Act required it to submit a SIP addressing
the updated 2008 ozone NAAQS. See 

(a)(1)–
(2). In particular, Pennsylvania’s SIP was required to impose
Reasonably Available Control Technology (“RACT”) for pre-
existing major sources of volatile organic compounds (“VOC”)
and nitrous oxides (“NOx”), both of which contribute to ozone
formation. RACT is “the lowest emission limitation that a
particular source is capable of meeting by the application of
control technology that is reasonably available considering
technological and economic feasibility.” Sierra Club v. United
States Env’t Prot. Agency, 

 (3d Cir. 2020)
(cleaned up).

        In May 2016, Pennsylvania submitted its SIP to the
EPA to satisfy the state’s RACT planning requirements for
VOCs and NOx. See 

.97–129.100 (2016).
It included two types of RACT provisions: (1) “presumptive”
RACT, and (2) source-specific RACT. The presumptive
RACT provision, 

, establishes broad NOx
and VOC emissions limits for certain types of stationary
machines, such as process heaters, combustion turbines, and
cement kilns. The relevant source-specific RACT provision,
meanwhile, permits a facility that cannot meet the presumptive
RACT requirement to propose an alternative RACT
requirement specific to its facility. 

.

       The EPA approved the presumptive RACT portion of
Pennsylvania’s SIP revision but only conditionally approved
the source-specific RACT provisions. 

. It
conditioned final approval of the source-specific RACT rule




                               6
on “further information on specific sources.” 

 Pennsylvania
then proposed RACT variances to seventeen major NOx and
VOC emitting facilities and submitted them to the EPA as
revisions to its SIP. While these revisions were pending final
EPA approval, the Center submitted comments to the agency
objecting to the variances. The comments are substantively
identical and read in relevant part as follows:

      The Clean Air Act [§ 7410(l)] “analysis” in the
      proposed rule is inadequate. The fact that these
      rules reduce NOx emissions in no way
      established that the reduced NOx emissions will
      not cause or contribute to a 2010 1-hour NOx
      NAAQS violation. We have modeled numerous
      sources of NOx emissions in the oil and gas and
      other industry with annual NOx emissions much
      lower than the sources in this RACT rule and
      found them to cause or contribute to 2010 1-hour
      NOx NAAQS violations. For example, attached
      is a modeling report for a well pad which caused
      NOx NAAQS violations. Therefore, EPA must
      undertake a modeling analysis of at least the
      following to determine if they cause or
      contribute to 2010 1-hour NOx NAAQS
      violations.

Joint Appendix (“JA”) 552, 554. Additionally, the Center
appended to its comments an analysis of emissions from a
Colorado facility using flare control technology.

       Notwithstanding the Center’s objections, the EPA
approved Pennsylvania’s SIP revisions. It did so via two
separate rules. The first rule approved Pennsylvania’s SIP




                              7
revision containing RACT determinations at eight major NOx
and/or VOC emitting facilities. 

 (Sept. 1,
2021). The second rule contained RACT determinations for
nine additional facilities. 

 (Nov. 1, 2021).
The EPA based these approvals on its view that none of the
seventeen variances would increase emissions. It compared
what the revised SIPs would allow with the prior emissions
limits in each facility’s permit and concluded that “the status
quo in . . . emissions had been maintained, if not improved, and
that there is no need to conduct the modeling suggested by the
[Center].” 

,909–12; 

,171–77.

       The Center subsequently submitted a petition for
reconsideration to the EPA, though it only sought
reconsideration of the first rule approving Pennsylvania’s SIP
revisions. It did not submit a petition for reconsideration in
connection with the second rule. The Center’s petition for
reconsideration claims that the EPA erred when it concluded
that certain of the source-specific RACT does not involve any
NOx emissions because certain RACT used to control VOCs
emits NOx. This petition for reconsideration remains pending
before the EPA.

       The Center now seeks this Court’s review of both rules
approving Pennsylvania’s SIP revisions, which are the
consolidated petitions before us today.

                              II.1




1
 We have jurisdiction to review “[a] petition for review of the
Administrator’s action in approving or promulgating any




                               8
        We must uphold the EPA’s approval of a SIP revision
unless it is “arbitrary, capricious, an abuse of discretion, or
otherwise not in accordance with law.” 

(2)(A).
While our review is deferential, “the agency cannot reach
whatever conclusion it likes and then defend it with vague
allusions to its own expertise; instead, the agency must support
its conclusion with demonstrable reasoning based on the facts
in the record.” Sierra Club, 

. A court must be
careful, however, to avoid “substitut[ing] its judgment for that
of the agency.” Motor Vehicle Mfrs. Ass’n of U.S., Inc. v.
State Farm Mut. Auto. Ins. Co., 

 (1983). A
court must therefore defer to the agency’s expertise if it can
discern “a rational connection between the facts found and the
choice made.” Sierra Club, 

 (quoting
Prometheus Radio Project v. FCC, 

, 389–90 (3d
Cir. 2004)).2


implementation plan under section 7410” of the Clean Air Act.

(b)(1).
2
  There remains an elephant in the room: whether, and to what
extent, Chevron deference specifically plays a role in this
analysis. See Chevron, U.S.A., Inc. v. NRDC, 

(1984). The Chevron decision requires that if “Congress has
directly spoken to the precise question at issue . . . [,] the court
. . . must give effect to the unambiguously expressed intent of
Congress.” 

 at 842–43. If, however, “the statute is silent or
ambiguous with respect to the specific issue,” a reviewing
court must defer to the agency’s interpretation if it is
reasonable. 

 at 843–44. Other Courts of Appeals dealing
with similar questions related to EPA approval of SIP revisions
have utilized the Chevron deference framework and held in
favor of the EPA by concluding that § 7410(l)’s use of the term




                                 9
“interfere” was ambiguous and deeming the agency’s
interpretation of the provision to be permissible. See, e.g.,
Indiana v. EPA, 

 (7th Cir. 2015); Alabama
Env’t Council v. EPA, 

 (11th Cir. 2013);
Kentucky Res. Council, Inc. v. EPA, 

 (6th
Cir. 2006).

The EPA here conspicuously makes no mention of Chevron in
its briefing and skirted the issue at oral argument, despite
relying heavily on the foregoing Chevron-based case law to
support its position. Such tiptoeing is, perhaps, not an
accident. The Supreme Court recently granted a writ of
certiorari in Loper Bright Enters. v. Raimondo, which presents,
in part, the question of whether the Court should overrule
Chevron. See No. 22-451, 

, at *1 (U.S. May
1, 2023). This follows several years of opinions in which the
Court has moved away from the doctrine in its administrative
law jurisprudence. See Thomas B. Griffith & Haley N.
Proctor, Deference, Delegation, and Divination: Justice Breyer
and the Future of the Major Questions Doctrine, 132 Yale L.J.
Forum 693, 714–18 (2022) (recapping the Supreme Court’s
recent “retreat from Chevron”). Given the EPA’s decision here
to eschew reliance on this doctrine, we will look instead to the
aforementioned general principles of deference inherent in
arbitrary and capricious review to guide us here. See Fed.
Commc’ns Comm’n v. Prometheus Radio Project, 

, 1158 (2021) (describing arbitrary-and-capricious review
as “deferential” and observing that “[a] court simply ensures
that the agency has acted within a zone of reasonableness and,
in particular, has reasonably considered the relevant issues and
reasonably explained the decision”).




                              10
                              III.

       The Center argues that the EPA erred in two different
ways in approving Pennsylvania’s SIP revisions. It claims first
that the EPA’s decision to consider the revisions via an
emissions-based analysis — instead of examining air quality
more generally — violated § 7410(l)’s statutory mandate.
Second, the Center argues that, even if § 7410(l) permits the
EPA to focus exclusively on emissions, the agency wrongly
concluded that Pennsylvania’s SIP revisions would not
increase emissions. We will address these arguments in turn
below.

                              A.

       We first examine whether the EPA acted arbitrarily and
capriciously in violation of § 7410(l) when it used an
emissions-based analysis to evaluate and approve
Pennsylvania’s SIP revisions. The relevant statutory language
provides:

      The Administrator shall not approve a revision of a plan
      if the revision would interfere with any applicable
      requirement concerning attainment and reasonable
      further progress (as defined in section 7501 of this title),
      or any other applicable requirement of this chapter.

(l). The EPA concluded that Pennsylvania’s
SIP revisions satisfied this language — that the revisions did
not “interfere with . . . [NAAQS] attainment” — because it
found that “[e]missions are not expected to increase, and will
likely decrease” under the source-specific RACT the revisions
imposed. 

; see also 

.




                              11
The Center, meanwhile, argues that this emissions-based
analysis did not satisfy § 7410(l) because the statute centers on
“attainment” of the NAAQS, and NAAQS are air quality, not
emissions, standards. Emissions contribute to, but are not the
same as, air quality; so, to the Center, “[t]he only way for EPA
to know the effect on air quality that would result from the
approval of a SIP would be for it to perform some analysis of
ambient air quality beyond a calculation of the emissions.”
Center Reply at 15.
       We hold that the EPA’s decision to conduct an
emissions-only analysis was not arbitrary and capricious.
Beginning first with the relevant statutory language, § 7410(l)
cabins its reach to only those SIP revisions that “interfere with”
NAAQS attainment. Determining whether such interference
will occur is an inquiry centered upon the specific relationship
between the instrument doing the potential interfering (here, a
SIP revision) and its effect (in this case, on air quality).
Although there are many factors that generally contribute to air
quality, including emissions, topography, atmospheric
conditions, and smokestack composition to name a few, a
particular SIP revision may only affect a subset of these
variables. If that revision leaves other air quality variables
unchanged, it makes sense for the EPA to eschew a
comprehensive air quality analysis in favor of a tailored
approach focused on the specific variables implicated by the
revision. Put simply, different types of SIP revisions pose
different risks of air quality interference and § 7410(l) permits
the EPA to adjust its analysis accordingly.

        The EPA’s emissions-based analysis fulfilled its
statutory duty here because there is no evidence in the record
to suggest that Pennsylvania’s SIP revisions affected, or could
affect, any other air quality factor. The revisions proposed




                               12
certain source-specific RACT; that is, technologies that limit
the level of emissions being released from a pollution source.
Beyond emissions levels, the record contains no evidence of
any other air quality variable that could have been affected by
Pennsylvania’s SIP revisions. Indeed, emissions were the sole
factor upon which the Center based its objections in its
comments to the EPA, and neither party has suggested in
briefing or at oral argument that any other air quality variable
besides emissions might change as a result of the SIP revisions.
We therefore deem the EPA’s decision to conduct an
emissions-only assessment here to be within the “zone of
reasonableness” required, given that it “reasonably considered
the relevant issue[]” — emissions.3 Prometheus, 141 S. Ct. at
1158.

        The Center’s proposed construction of § 7410(l), by
contrast, reads too much into the provision’s relatively narrow
strictures. Nothing in § 7410(l) suggests that the EPA must

3
   Our conclusion that the EPA acted reasonably here is
bolstered by the deference we owe to an agency’s expertise-
based factual determinations. See Sw. Pa. Growth All. v.
Browner, 

, 117 (3d Cir. 1997) (“A reviewing
court must generally be at its most deferential when reviewing
factual determinations within an agency’s area of special
expertise.”). The EPA’s decision to focus exclusively on
emissions as the relevant air quality variable when analyzing
whether the Pennsylvania SIP revisions “interfere[d]” with air
quality is precisely such a factual determination within its area
of expertise. The agency is best positioned to determine which
air quality variables are implicated by, and thus must be
analyzed for, a given SIP revision, and we owe that decision
deference.




                               13
conduct an air quality analysis in every instance. Such silence
weighs against concluding that the EPA’s analysis was arbitrary
and capricious. The Center’s focus on the fact that one cannot
determine air quality based on emissions alone is beside the
point because, again, § 7410(l) does not require the EPA to
assess air quality generally but rather to analyze the specific
relationship between the proposed SIP revision and NAAQS
attainment. This is a different, narrower inquiry with
parameters that depend entirely on the nature of the SIP
revision and its particular effect on air quality. The more
limited nature of this inquiry comports with common sense —
there is no need for the EPA to conduct a comprehensive air
quality analysis if there is no evidence that other, non-
emissions factors will be changed by a particular SIP revision.
See Indiana v. EPA, 

 (7th Cir. 2015)
(affirming the EPA’s emissions-based analysis in part because
petitioner “ha[d] not shown that the agency’s conclusion [that
the revision did not interfere with attainment] would have been
any different” had the SIP included air quality modeling).

        Section 7410(l) is not a one-size-fits-all provision. Just
as it does not require an air quality analysis in every instance,
so too could there be circumstances in which an emissions-only
analysis is insufficient. To reiterate, the precise variables that
must be analyzed to satisfy § 7410(l) necessarily depend on the
nature of the SIP revision in question and the particular
interference risk it poses.4 Some revisions may include

4
  The EPA explicitly recognized as much in its rulemaking
approving the Pennsylvania SIP revisions, in which it observed
that “the level of rigor needed for any [§ 7410(l)]
demonstration will vary depending on the nature and
circumstances of the revision” before explaining in detail why




                               14
changes that affect different aspects of the air quality equation
instead of, or in addition to, emissions. The record here
suggests that emissions were the sole air quality variable
implicated by Pennsylvania’s SIP revisions. It was therefore
not arbitrary or capricious under § 7410(l) for the EPA to use
an emissions-based analysis here.5

                               B.

       The Center argues in the alternative that, even if §
7410(l) permitted the EPA to use an emissions-based approach
to analyze Pennsylvania’s SIP revisions, the agency acted


it settled on its emissions-based approach here. 

. This is further evidence that the EPA here has
“considered the relevant issues and reasonably explained the
decision” and thus has not acted arbitrarily or capriciously. See
Prometheus, 141 S. Ct. at 1158.
5
  In so holding, we join several other Courts of Appeals that, in
similar but not identical contexts, affirmed the EPA’s use of
emissions-based analyses to evaluate SIPs. See, e.g., Indiana,

; WildEarth Guardians v. EPA, 

, 1073–74 (9th Cir. 2014); Alabama Env’t Council, 

; Ky. Res. Council, 

. Most of these
courts, as previously noted, held the agency’s emissions-based
approach to be permissible based on Chevron deference. The
sole exception is WildEarth Guardians, in which the Court of
Appeals for the Ninth Circuit recognized, as here, that “nothing
in [the SIP at issue] weakens or removes any pollution controls.
And even if the [SIP] merely maintained the status quo, that
would not interfere with the attainment or maintenance of the
NAAQS.” 

.




                               15
arbitrarily and capriciously by concluding that those revisions
did not increase emissions. First, the Center contends that the
EPA compared those revisions’ changed emissions levels to
the wrong emissions baseline; failure to use the appropriate
baseline — presumptive RACT — meant that the EPA
erroneously approved SIP revisions that increased emissions.
Second, the Center claims that certain of the control
technologies approved by the EPA emit NOx pollutants of their
own, leading to increased emissions that were not accounted
for in the agency’s analysis. Both arguments fall short for the
reasons explained below.

                               1.

       Recall that Pennsylvania’s initial SIP included (1)
presumptive RACT that would apply as the default technology
standard for certain NOx and VOC pollution sources, and (2)
separate provisions allowing facilities to propose their own
source-specific RACT variances for their particular facility or
group of facilities. The Center claims that the emissions limits
associated with the presumptive RACT constituted the baseline
to which the EPA should have compared Pennsylvania’s
subsequent source-specific RACT variances. The EPA did not
use the presumptive RACT baseline and instead used
emissions limits contained in the previously applicable permits
for those various facilities. In the Center’s view, this led the
agency to approve certain control technologies that resulted in
higher levels of emissions than would be permitted under the
presumptive RACT, in violation of § 7410(l).

       The Center’s argument as to the presumptive RACT
baseline falls short in light of the plain language of the
Pennsylvania regulatory scheme. Pennsylvania’s revised SIP,




                              16
as previously noted, contemplates multiple ways in which
pollution sources within the state can implement RACT. The
first part of the relevant language, enshrining presumptive
RACT as the Pennsylvania emissions control technology
standard for many NOx and VOC sources, provides:

       (a) a source . . . located at a major NOx emitting facility
       or major VOC emitting facility . . . shall comply with
       the applicable presumptive RACT requirement . . .
       beginning with the specified compliance date as
       follows, unless an alternative compliance schedule is
       submitted and approved under . . . § 129.99 . . . .

. The associated provision § 129.99
provides in relevant part that:

       (a) the owner or operator of an air contamination source
       subject to 129.97 . . . located at a major NOx emitting
       facility or major VOC emitting facility . . . that cannot
       meet the applicable presumptive RACT requirement . .
       . may propose an alternative RACT requirement . . . .

.

       Read together, these provisions demonstrate that
presumptive RACT cannot be the emissions baseline. The
purpose of § 129.99’s self-described “alternative RACT
requirement” carve-out from § 129.97’s presumptive RACT
baseline is that the facilities subject to § 129.99 “cannot meet
the applicable presumptive RACT requirement.” Id. (emphasis
added). In other words, the applicable portion of § 129.99 only
permits facilities that cannot comply with presumptive RACT
to obtain the sort of source-specific, case-by-case RACT




                               17
determinations that the Center challenges here. It would make
little sense, then, to hold those same facilities to the very
presumptive RACT baseline that the statute contemplates them
being unable to meet. See Griffin v. Oceanic Contractors, Inc.,

 (1982) (“[I]nterpretations of a statute which
would produce absurd results are to be avoided if alternative
interpretations consistent with the legislative purpose are
available.”). It was therefore not arbitrary or capricious for the
EPA to use prior permitting standards, instead of presumptive
RACT, as the emissions baseline for its § 7410(l) comparative
emissions analysis.

                               2.

       The Center’s final argument is that the EPA acted
arbitrarily and capriciously when it concluded that certain of
Pennsylvania’s proposed emissions control technologies do not
emit NOx. It claims, specifically, that some of the EPA-
approved technologies that limit VOC emissions actually emit
NOx pollutants in the process, and that the EPA failed to
account for these emissions in its analysis. But before we can
reach the merits, we must examine the procedural hurdles
affecting our ability to consider this aspect of the consolidated
petitions for review.

       First, the Center forfeited its challenge here since
neither of its comments to the EPA mentioned the risk of
Pennsylvania’s source-specific RACT increasing NOx
emissions. See Sw. Pa. Growth. All. v. Browner, 

,
112 (3d Cir. 1997) (“Generally, federal appellate courts do not
consider issues that have not been passed on by the agency . . .
whose action is being reviewed.”) (cleaned up); see generally
Barna v. Bd. of Sch. Directors of Panther Valley Sch. Dist.,




                               18


 (3d Cir. 2017). The Center, however, claims
that it preserved the NOx emissions argument via computer
modeling data that it referenced in and attached to its
comments. The modeling shows that flares, a control
technology for VOC emissions that was not used in
Pennsylvania’s SIP revisions, created NOx violations. Such an
indirect reference to the risk of increased NOx emissions from
VOC control technology does not suffice to prevent forfeiture
here because it does not meet 

(d)(7)(B)’s
requirement that an objection to an agency’s rule be raised
“with reasonable specificity” in order to be preserved for
judicial review. An objection like this — raised by implication
only, via a tangentially related study attached to a comment
that otherwise makes no mention of the objection in question
(and in fact could be read to accept that NOx emissions would
be reduced) — does not satisfy the “reasonable specificity”
requirement because it cannot be deemed to have alerted the
agency to the alleged increased NOx emissions. See Tex Tin
Corp. v. EPA, 

 (D.C. Cir. 1991) (“An
objection must be made with sufficient specificity reasonably
to alert the agency.”).

       Second, we must consider the effect of the Center’s
petition for reconsideration currently pending before the EPA,
which asks the agency to revisit the first of its two rules
approving Pennsylvania’s SIP revisions. That petition for
reconsideration for the first time directly and explicitly alerts
the agency to the Center’s concerns that some of
Pennsylvania’s proposed VOC control technology itself emits
NOx. The petition for reconsideration’s pending status limits
our ability to review part of the Center’s argument. A court
may not consider matters raised for the first time in a petition
for reconsideration while that petition remains pending before




                               19
the relevant agency. Util. Air Regul. Grp. v. EPA, 

 (D.C. Cir. 2014) (“Objections raised for the first time
in a petition for reconsideration must await EPA’s action on
that petition.”). Because the EPA has not yet resolved the
Center’s petition for reconsideration, we cannot review the
Center’s claims pertaining to NOx emissions from VOC
control technologies raised for the first time within it.

      Procedural considerations prevent us from reviewing
the merits of the Center’s claim that the EPA erroneously
approved certain VOC control technologies that release NOx
emissions.6 We thus hold that, for each facility covered by the


6
  We cannot, however, ignore the agency’s admission of error
with respect to one Pennsylvania facility in particular: the
Roystone Compressor Station (“Roystone”). The EPA admits
that it approved the use of a thermal oxidizer at Roystone to
limit VOC emissions that does increase NOx emissions, even
though the agency’s approval stated that there were no NOx
emissions from the proposed RACT at the facility. This error
is undoubtedly concerning. The EPA’s thorough identification
and consideration of each air quality variable implicated by a
given SIP revision is a key prerequisite of a reasonable §
7410(l) analysis — otherwise, it cannot accurately ascertain
whether the revision will “interfere with . . . [NAAQS]
attainment,” as the statute requires. An emissions-based
assessment that accounts only for the emissions being
controlled by the relevant technology and not for the emissions
being released by the control process itself could indicate that
the agency may have “entirely failed to consider an important
aspect of the problem.” See Motor Vehicle Mfrs. Ass’n, 

. Fortunately, consideration of the Roystone station




                              20
petition for reconsideration presently pending before the EPA,
the Center’s petition for review is denied without prejudice to
any subsequent objections resulting from the EPA’s resolution
of the reconsideration process. As for the remainder of the
facilities not covered by the petition for reconsideration, we
hold that the Center has forfeited its claim by failing to raise its
concerns regarding NOx emissions from VOC control
technologies with reasonable specificity; its petition for review
will therefore be denied as to these facilities as well.

                                IV.

       For the foregoing reasons, we will deny the Center’s
consolidated petitions for review.




remains pending before the EPA as part of the Center’s petition
for reconsideration.




                                21